Rho GTPases in kidney physiology and diseases

ABSTRACT

Rho family GTPases are molecular switches best known for their pivotal role in dynamic regulation of the actin cytoskeleton, but also of cellular morphology, motility, adhesion and proliferation. The prototypic members of this family (RhoA, Rac1 and Cdc42) also contribute to the normal kidney function and play important roles in the structure and function of various kidney cells including tubular epithelial cells, mesangial cells and podocytes. The kidney’s vital filtration function depends on the structural integrity of the glomerulus, the proximal portion of the nephron. Within the glomerulus, the architecturally actin-based cytoskeleton podocyte forms the final cellular barrier to filtration. The glomerulus appears as a highly dynamic signalling hub that is capable of integrating intracellular cues from its individual structural components. Dynamic regulation of the podocyte cytoskeleton is required for efficient barrier function of the kidney. As master regulators of actin cytoskeletal dynamics, Rho GTPases are therefore of critical importance for sustained kidney barrier function. Dysregulated activities of the Rho GTPases and of their effectors are implicated in the pathogenesis of both hereditary and idiopathic forms of kidney diseases. Diabetic nephropathy is a progressive kidney disease that is caused by injury to kidney glomeruli. High glucose activates RhoA/Rho-kinase in mesangial cells, leading to excessive extracellular matrix production (glomerulosclerosis). This RhoA/Rho-kinase pathway also seems involved in the post-transplant hypertension frequently observed during treatment with calcineurin inhibitors, whereas Rac1 activation was observed in post-transplant ischaemic acute kidney injury.

KEYWORDS:

• Rho GTPases
• podocytes
• cytoskeletal dynamics
• nephropathies
• diabetes

Introduction

The mammalian kidney orchestrates the elimination of metabolic wastes found in blood, a function intimately related to its key roles in general fluid homoeostasis and osmoregulation. The kidney is one of the most highly differentiated organs in the body. At the conclusion of embryologic development, nearly 30 different cell types form a multitude of filtering capillaries and segmented nephrons enveloped by a dynamic interstitium. This cellular diversity modulates complex processes: endocrine functions, regulation of blood pressure and intraglomerular haemodynamics, solute and water transport, acid-base balance, and removal of drug metabolites. Urine is produced and concentrated along the length of nephrons, the basic unit of kidneys, particularly along medullary nephron segments. An adult human kidney is considered to contain an average of 1 million and even up to 2.5 million nephrons [1]. A nephron is functionally subdivided into a filtration unit called the renal corpuscle or glomerulus and a segmented tubular resorption compartment. The glomerulus of the mammalian kidney, the filtration unit where the first step of urine formation occurs, a highly developed vascular bed that acts as a filter allowing a filtrate of small molecules (such as water, sugars, electrolytes and small proteins) to pass through a barrier that retains high molecular weight proteins and cells in the blood circulation. Distinct cell types form this filtration barrier which consists of three different layers: (1) fenestrated glomerular endothelial cells, (2) glomerular basement membrane and (3) visceral glomerular epithelial cells [2]. Fenestrated endothelial cells line the capillary loops and lie in close contact with mesangial cells. Podocytes (or renal visceral epithelial cells) have a complex cellular architecture consisting of cell body, major processes that extend outward from their cell body, forming interdigitated FPs that enwrap the glomerular capillaries. These specialized pericytes are connected by specialized intercellular junctions known as slit diaphragms and separated from the endothelial compartment by the glomerular basement membrane. Both the slit diaphragms and the glomerular basement membrane are indispensable components of the glomerular filtration barrier, which prevents protein loss from the plasma into the urine [for review, see e.g. 3–6].

Podocytes, which represent the final barrier to the loss of serum proteins into the urine, have complex actin cytoskeleton dynamics [7]. Because of their limited proliferation capacity, podocyte injuries are considered as irreversible, in particular in glomerular diseases. These cells have a large cell body located in the urinary space, primary processes, secondary processes, and tertiary processes that are ‘feet-like’ and known as ‘foot processes’ (FP). The cytoskeleton of podocytes is critical to their ability to form FP and slit diaphragm junctions that ultimately determine glomerular permselectivity. The mesangial cell occupies a central position in the renal glomerulus. Besides podocytes, tubular epithelial cells are highly specialized cells with a functional polarization. Collectively, these cells and different sections of the nephrons are enveloped by an interstitium which interacts with the tubular system via multiple communication pathways [8].

The small guanosine triphosphatases (GTPases) of the Rho family are master regulators of actin cytoskeletal dynamics and thus facilitate changes in cell shape, motility, adhesion, migration, proliferation, polarity, cell cycle progression, gene expression as well as apoptosis. Several Rho GTPases, particularly RhoA but also Cell division cycle 42 (Cdc42) and Rac1, are highly expressed in the renal cortex [9,10]. RhoGTPases regulate actin cytoskeleton plasticity and thus facilitate changes in cell shape, motility, adhesion, polarity, cell cycle progression, and gene expression. The Rho kinases (ROCKs) are a family of serine-threonine kinases that serves as key downstream effectors for Rho GTPases. Two mammalian ROCK homologs, ROCK1 and ROCK2, have been identified, which share 92% identity of amino acid sequence in their kinase domain [11]. Both proteins are structurally composed of an N-terminal kinase domain followed by a coiled-coil region that contains a Rho-binding domain (RBD). At the C-terminal region of both isomers, there is a pleckstrin homology (PH) domain, which folds back onto the N-terminal region and inhibits their kinase activity under basal conditions. When Rho proteins (RhoA, RhoB or RhoC) are activated to bind to the RBD domain or when the auto-inhibitory region is cleaved and removed by caspase-3 or Granzyme-B, ROCK is activated [for references, see e.g. 12]. ROCKs function as signalling hubs controlling cytoskeletal dynamics, cell migration, stress fibre formation, gene expression, proliferation, and survival. ROCKs are constitutively active in renal circulation [13]. Rac and Cdc42 regulate actin dynamics through activation of two types of actin nucleators, Wiskott–Aldrich syndrome protein (WASP) and WASP-family verprolin-homologous protein (WAVE) family proteins and Diaphanous-related formins (DRFs) [14]. WASP and WAVE family proteins are able to activate the Arp2/3 complex, leading to rapid actin polymerization. The p21 Activated Kinases (PAKs), a family of serine threonine kinases consisting of six members (PAK 1–6), function downstream of Cdc42 and Rac [see e.g. 15]. PAKs play key roles in the control of a number of fundamental cellular processes by phosphorylating their downstream substrates.

Such ability to modulate a wide range of biological processes allows RhoGTPases to play critical roles in renal physiology, and their dysregulated activities are implicated in the pathogenesis of both hereditary and idiopathic forms of renal diseases (nephropathies).

1. Rho GTPases in renal functions

Members of the Rho GTPase family of intracellular signalling molecules are master regulators of actin cytoskeletal dynamics. Hence, they play an important role in the structure and function of various kidney cells including tubular epithelial cells, mesangial cells and podocytes and in their multiple communication pathways.

Involvement of Rho-GTPases and their regulatory proteins in glomerular podocyte function

At the glomerulus (the filtration unit of the kidney where the first step of urine formation occurs), the plasma is filtered into the urinary space through the filtration barrier, which consists of three different layers: (1) fenestrated glomerular endothelial cells, (2) glomerular basement membrane and (3) visceral glomerular epithelial cells [2]. Podocytes (or renal visceral epithelial cells) have a complex cellular architecture consisting of cell body, major processes that extend outward from their cell body, forming interdigitated FPs that enwrap the glomerular capillaries. Interdigitated FPs of podocytes interlink with slit diaphragms, which form cell-cell contacts in mature podocytes. Both the slit diaphragms and the glomerular basement membrane are indispensable components of the glomerular filtration barrier, which prevents protein loss from the plasma into the urine.

Primary processes and FPs are highly supported by a complex cytoskeletal architecture made of an entanglement of microtubules, intermediate filaments and actin. Studies concerning the molecular composition of podocyte cell-cell contacts identified a series of adhesion proteins in the slit diaphragm (see Figure 1), including nephrin, Fat1, cadherin, catenin, intracellular adapters (Zonula Occludens-1, Podocin, CD2-associated protein (CD2AP), and membrane proteins such as the Ca²⁺-permeable transient receptor potential channel 6 (TRPC6) [see 16–20]. As in other cell types, these adhesion molecules are intimately associated with actin filaments of podocytes [21–23]. Such complex networks not only accommodate podocytes to the constantly changing environment but are also crucial hubs of signal transduction within podocytes. The podocyte function is regulated by Rho GTPases which act as molecular switches controlling activation of multiple downstream effector molecules. RhoA, together with Cdc42 and Rac1, have been proposed to regulate podocyte cytoskeletal rearrangements [24].

Figure 1. Maintenance of podocyte architecture by Rho GTPases.

The figure depicts the “slit-diaphragm constituted of cell-cell junctions at the baso-lateral membrane of podocytes, and the protein complexes driving the activation and control of RhoA, Cdc42 and Rac1 GTPases. Cdc42 controls polarization of podocytes through Par-6 complex while RhoA, activated by TRPC6, maintains cytoskeleton integrity, helped by Ubiquitin-ligase Smurf inhibition by Synaptopodin-long form (Syno-long). Synaptopodin also chelates IRSpI3 preventing it to interact with Wave 2 complex, and then inhibiting the Rac1 pathway. In the context of Rac1 activation, podocyte architecture is modified, and motility is activated.

By specifically deleting podocyte RhoGTPases, it was demonstrated that mice lacking Cdc42 developed congenital nephropathy and died as a result of renal failure within 2 weeks after birth, whereas mice lacking Rac1 or RhoA in podocytes were overtly normal and lived to adulthood [25,26]. However, after this initial phase, RhoA and Rac1 seem to play more important roles in podocyte biology. Rac1 plays a key role in actin lamellipodia induction and cell-matrix adhesion while RhoA is responsible for stress fibre formation. Podocyte-specific expression of constitutively active Rac1 induced rapid onset of proteinuria with focal foot process (FP) effacement [27,28] whereas, on the contrary, mice with podocyte-specific Rac1 deletion show no kidney dysfunction and were protected from protamine-induced FP effacement [25]. In insect nephrocytes, which are structurally and functionally homologous to mammalian kidney podocytes, a tight balance of Rac1 and Cdc42 activities is essential to maintain the specialized architecture and function of this cell [29].

In many biological systems, including podocytes, RhoA and Rac1 antagonize each other’s activation and function [30]. RhoA is considered to promote stress fibre formation, which is adaptive in podocytes, whereas Rac1 and Cdc42 would promote lamellipodia and filopodia formation, respectively, which are correlated with podocyte injury [31–33]. In transgenic mice, both RhoA activation as well as RhoA inhibition specifically in glomerular podocytes caused albuminuria and FP effacement, either RhoA activation or inhibition indeed had similar adverse effects on glomerular filtration barrier function and reduced podocyte synaptopodin expression, but the mechanisms of these detrimental effects appeared to be different, suggesting that in podocytes Rho activity must be tightly regulated to maintain podocyte function [34]. In murine podocytes, RhoA deficiency was suggested to reduce the mRNA and protein expression of YAP, considered as an anti-apoptosis protein in podocytes [35].

In vitro, treating human podocytes with interleukin-13 (IL-13) resulted in phosphorylation of Vav1. Vav (Vav1, Vav2 and Vav3) are Rho GTPase guanine nucleotide exchange factors (GEFs, see below section ‘Factors influencing the activities of renal Rho-GTPases’) activated by tyrosine phosphorylation [36]. This phosphorylation led to Rac1 activation and actin cytoskeleton rearrangement, which was abrogated in Vav1 knockdown podocytes [37]. Conversely, Vav1^−/- mice models showed no direct renal affection, but Vav3-/- mice presented renal dysfunction linked to deficient blood pressure control by GABAergic cells [38]. In animal models, alterations in Rac1 activity have shown conflicting findings, beneficial or detrimental to the podocyte depending on the type of podocyte injury involved [see 39]. In contrast, in injured podocytes, the presence of Rac1 was seen to promote the activity of mammalian target of rapamycin (mTOR) pathway for glomerular repair to protect the kidneys from developing glomerulosclerosis [40].

In mouse, the lack of Cdc42 or neuronal WASP protein (N-WASP), its downstream effector, caused podocyte apoptosis and proteinuria both in vitro and in vivo by decreasing the mRNA and protein expression of Yes-associated protein (YAP), which had been regarded as an anti-apoptosis protein in podocyte [41].

Podocyte effector and partner proteins of RhoGTPases

Synaptopodin [named from the protein’s associations with postsynaptic densities and dendritic spines and with renal podocytes; 42] is the founding member of a unique class of proline-rich actin-associated proteins that are expressed in highly dynamic cell compartments, such as telencephalic dendrites and renal podocytes. Synaptopodin exists in three isoforms, neuronal Synpo-short (685 AA), renal Synpo-long (903 AA), and Synpo-T [181 AA; 43]. All three isoforms specifically interact with α-actinin and elongate α-actinin–induced actin filaments. In mice, gene silencing of Synpo-T impairs stress-fibre formation in podocytes [43], suggesting a functional link between synaptopodin and RhoA signalling. Mechanistically, synaptopodin was found to synchronize podocyte actin dynamics and cell migration by blocking Smurf-1-mediated ubiquitination of RhoA, protecting RhoA against proteasomal degradation. This, in turn, promotes the formation of stress fibres [31]. Moreover, synaptopodin directly binds to IRSp53 (a signalling intermediate in Rac1- and Cdc42-induced actin dynamics through Wave2-Arp2/3), thereby suppressing Cdc42:IRSp53:Mena-initiated filopodia formation [44].

In mouse podocytes, synaptopodin interacts with endogenously expressed mechanosensitive large-conductance Ca²⁺-activated K⁺ (BK_Ca) channels and plays a role in regulating their steady-state expression on the cell surface. This effect is blocked by inhibition of Rho signalling in HEK293T cells and in podocytes [45]. Moreover, the current density through BK_Ca channels in podocyte cell lines appears to be directly proportional to the amount of active Rho in the cells [45]. In murine podocytes, synaptopodin can suppress Rac1 signalling by blocking the Vav2-mediated activation of Rac1 [46]. In a murine experimental model of immune-mediated podocyte injury, internalization of the podocyte slit membrane proteins nephrin and synaptopodin was prevented by ROCK inhibition [47].

The Transient Receptor Potential Channels (TRPC) family, composed of seven structurally related channels (TRPC1-7) is a regulator of the intracellular Ca²⁺ concentration in many tissues, including kidney. TRPC1, TRPC3, TRPC4, TRPC5, and TRPC6 were reported to be expressed in podocytes, but up to now, only TRPC3, TRPC5 and TRPC6 channels have been functionally and pharmacologically shown to be involved in Ca²⁺ entry in the podocytes [for references, see 48].

TRPC5 and TRPC6 channels were shown to act as antagonistic regulators of actin remodelling and cell motility in fibroblasts and kidney podocytes. In mouse podocytes, TRPC5 is in a molecular complex with Rac1, whereas TRPC6 is in a molecular complex with RhoA. TRPC5-mediated Ca²⁺ influx induces Rac1 activation, eliciting stress fibre disassembly and, thereby, promoting cell migration, whereas TRPC6-mediated Ca²⁺ influx increases RhoA activity, enhancing stress fibre formation and cell adhesion and thereby inhibiting cell migration [49]. TRPC6 mutations lead to an autosomal dominant form of human kidney disease characterized histologically by focal and segmental glomerulosclerosis [50,51]. Epidermal growth factor (EGF) induces incorporation of functional TRPC5 channels into the plasma membrane of podocytes in a Rac1 and phosphatidylinositol 4-phosphate 5-kinase-dependent manner [52]. This corresponds to a reciprocal activation of Rac1 and TRPC5, inducing a positive loop for activation of both partners. Chronic-specific inhibition of TRPC5 suppressed severe proteinuria and prevented podocyte loss in a transgenic rat model of Focal Segmental Glomerulo-Sclerosis [FSGS; 53]. However, in transgenic mouse models, genetic gain in function studies revealed that TRPC5 did not cause or aggravate glomerular barrier injury and proteinuric kidney disease [54].

Rhophilin-1, as Rho-binding protein, has long been known as a RhoA effector [55], but the biologic function of which is poorly unknown. Rhophilin-1 (one of the two Rhophilin isoforms) appears as a key determinant of podocyte cytoskeleton architecture. This podocyte-specific protein was suggested to control actin cytoskeletal dynamics by primarily restricting RhoA activity [56]. Rhophilin-1 has inhibitory effects on Rho-dependent phosphorylation of the myosin regulatory light chain and stress fibre formation [57]. Rhophilin-1 controls actin cytoskeletal dynamics by primarily restricting RhoA activity, required for the maintenance of podocyte FP architecture as well as the functional integrity of the glomerular filtration barrier [56].

Functions of Rho-GTPases in mesangial cells

Mesangial cells and their matrix form the central stalk of the glomerulus and constitute with glomerular endothelial cells and podocytes a functional unit. Mesangial cells are of mesenchymal origin and have many properties in common with vascular smooth muscle cells. They constitute a population of irregularly shaped cells which possess high levels of actin and myosin. They are in continuity with the extraglomerular mesangium and the juxtaglomerular apparatus. They generate and embed in their own extracellular matrix, different in composition from the glomerular basement membrane. The composition and amount of mesangial matrix are tightly controlled in health but can be markedly altered during disease [for references, see e.g. 58, 59]. They have contractile properties generated by anchoring filaments to glomerular basement membrane opposite podocyte foot processes and at the paramesangial angles, thereby assisting in the maintenance of capillary organization and convolution. On the capillary lumen side, mesangial cells are in direct contact with endothelial cells without an intervening basement membrane. They contract upon exposure to a number of vasoactive agents such as Ang. II and endothelin.

It is thought that one of mesangial cell functions is to regulate the blood flow through selected capillary loops. Mesangial cells show immunomodulatory properties, are capable of phagocytosis and are involved in a number of pathologic conditions affecting the kidney. Thrombospondin-1 (TSP-1), a member of the family of matricellular glycoproteins involved in the regulation of cellular functions in physiological and pathophysiological settings, is synthesized by mesangial cells. TSP-1 is known for its anti-angiogenic activity and its ability to activate latent TGF-β [see 60]. Rac-1 was seen to be essential for TSP-1 activation [60].

Abnormal remodelling of the extracellular matrix causes glomerular mesangial expansion leading to the scarring of glomeruli (glomerulosclerosis). The proliferation of glomerular mesangial cells is a hallmark of glomerular injury progression [61,62]. Once activated, mesangial cells express chemokines (as monocyte chemoattractant protein 1, MPC-1) and accelerate accumulation of macrophages, leading to progression of glomerular injury [63]. A variety of signalling molecules and pathways are associated with mesangial matrix proliferation, including RhoA, highly expressed in the renal cortex [9], able to enhance actin cytoskeleton reorganization and endothelial cell barrier permeability. Various pro-inflammatory mediators promote the production of mesangial cytokines that propagate the inflammatory response. TNF-α was seen to promote monocyte chemotaxis towards mesangial cells through a mechanism involving regulation of the expression of MCP-1 via a Rho-kinase/p38 MAPK-dependent pathway [64].

Rho protein regulation of renal tubular epithelial cell function

Epithelial tubular cells of normal adult kidneys possess primary cilia [65] in both mature nephrons and the kidney collecting system. Primary cilia are specialized non-motile organelles consisting of an axoneme with nine microtubule doublets anchored to the plasma membrane which play an important role in the control of normal kidney development [see e.g. 66].

Both in vitro and in vivo evidence suggest a crucial role for Cdc42 in kidney tubule development by controlling cytoskeletal dynamics due to its ability to interact with N-WASP and PAKs [see e.g. 67]. [68], reported that Cdc42 was necessary for primary ciliogenesis in renal tubule epithelial cells. Mice lacking Cdc42 specifically in kidney tubular epithelial cells died of renal failure within weeks of birth [69]. Dynamin binding protein (also known as Tuba), a Cdc42-specific GEF, plays a critical role in ciliogenesis and nephrogenesis [70]. Tuba was suggested to activate Cdc42 in the ciliary region and activated Cdc42 to localize the exocyst (an evolutionarily conserved octameric protein complex) which target and dock vesicles carrying membrane proteins from the trans-Golgi network to the plasma membrane [70].

The Hippo pathway is a highly conserved kinase cassette that regulates tissue growth in metazoans by controlling the activity of YAP and its paralog transcriptional coactivator with PDZ-binding motif (TAZ), closely related transcriptional co-activators that control expression of genes that promote cell proliferation and inhibit apoptosis [see 71]. The Hippo pathway is connected to renal disease and development [see 72. When the Hippo kinases Mst and Lats are active, YAP and TAZ are phosphorylated and excluded from the nucleus. In the murine kidney, nephrogenic inactivation of Cdc42 leads to loss of YAP-dependent gene expression, morphological defects and abnormalities in nephron gene expression, suggesting that Cdc42 acts upstream of YAP in nephron progenitor cells to promote gene expression required to establish and shape nephrons [73]. In mice, knockdown of Fat1 cadherin (which influences Wnt and Hippo signalling) was seen to reduce active Rac and Cdc42 and to induce defects in the formation of the tubular cell lumen [74].

Signalling via the Rho GTPases provides crucial regulation of numerous cell polarization events, including apicobasal (AB) polarity, polarized cell migration, polarized cell division and neuronal polarity. There are a multitude of processes that are important for AB polarization, including lumen formation, apical membrane specification, cell-cell junction assembly and maintenance, as well as tissue polarity. Several emerging common themes have already been highlighted, as (i) the need for Rho GTPase activities to be carefully balanced in both a spatial and temporal manner, (ii) the existence of signalling feedback loops and crosstalk to create robust cellular responses, and (iii) the frequent multifunctionality that exists among AB polarity regulators [75]. In porcine proximal tubular cells, ATP depletion induced RhoA inactivation during ischaemia, and during recovery RhoA activity was required for maintaining normal cellular architecture and function [76].

Epithelial sodium channels (ENaCs) are particularly expressed in the cortical collecting duct in the kidney, where they finely tune Na⁺ reabsorption from the lumen and play an important role in maintaining salt-water homoeostasis and regulating blood pressure [for review, see e.g. 77]. Several studies revealed that, among Rho family members, Cdc42 had no effect on ENaC current density whereas RhoA and Rac1 were able to increase the channel activity [see 78]. The regulation of Rac1 seems to be critical for physiological regulation of epithelial ENaCs via WAVE1/2 proteins [79].

Key role of kidney in the regulation of arterial blood pressure

A large body of experimental and physiological evidence have established the kidney as the organ chiefly responsible for maintaining normal blood pressure. The renin–angiotensin system (RAAS) plays a central role in the control of arterial blood pressure and sodium water homoeostasis. Renin is a proteolytic enzyme secreted by the juxtaglomerular apparatus of the kidney as an inactive form (pro-renin) and released into circulation. When renal blood flow is reduced, this precursor is converted into rennin by juxtaglomerular cells and released into circulation. It then catalyses the cleavage of the glycoprotein angiotensinogen, generating angiotensin I (Ang I) which is cleaved by the angiotensin-converting enzyme to produce angiotensin II (Ang II), the main effector in the RAAS. Most of the intrarenal Ang II is locally generated, rather than derived from circulating Ang I or Ang II [80,81]. Ang II is normally present in renal interstitial fluid in concentrations much higher (approximately 1000-fold higher in the rat) than in the systemic circulation), providing strong evidence that local angiotensin effects regulate renal function independently of systemic angiotensin [82].

Ang II is a pleiotropic hormone that activates a wide spectrum of signalling responses via the AT1 receptors 1 and 2 (AT1R and AT2R) that mediate its physiological control of blood pressure [83]. Most of its diverse actions via AT1R, which couple to classical _Gq/11 protein and signal through multiple downstream signals (including protein kinase C (PKC), extracellular signal-regulated kinase (ERK)1/2, Raf, tyrosine kinases, etc.) but also, via G_12/13 proteins, stimulates ROCK, which causes vascular contraction and hypertrophy [for references see e.g. 84]. AT1R-mediated activation also triggers activation of nicotinamide adenine dinucleotide phosphate (NADPH, the reduced form of NADP+) oxidase, a major source of reactive oxygen species (ROS) involved in the activation of pro-inflammatory transcription factors and stimulation of small G proteins such as Ras, Rac and RhoA [85]. Moreover, NADPH oxidase-derived ROS production demands active Rac1 to be bound to a cytosolic activator of most NADPH oxidase isoforms [see e.g. 86].

AT1R stimulation by Ang II leads to the reorganization of the actin cytoskeleton through several mechanisms [87,88]. One of them is AT1-R/Scr tyrosine kinase interaction, that activates phospholipase PLCγ1, which in turn activates Rac1, leading to a reduction of α-actinin-4 and ezrin/radixin/moesin (ERM) phosphorylation. The main function of α-actinin-4 and ERM is to regulate the attachment of membrane proteins to the actin cytoskeleton; therefore, those changes lead to cytoskeletal rearrangement, retraction of FPs, and slit diaphragm protein redistribution [87].

The enhanced release of Ang II leads to increased oxidative stress and free oxygen radicals production by the NADPH oxidase but also to reduce NO bioavailability [89]. Through different signalling pathways, all these agents finally affect vascular smooth muscle cell contraction, which is essentially determined by two factors, the intracellular Ca²⁺ concentration and the Ca²⁺ sensitivity of the contractile apparatus. Podocytes were seen to respond to Ang II in vivo, which caused an increase of their intracellular calcium activity via AT1R in rat glomeruli [90]. Ca²⁺ sensitivity of the contractile apparatus is mainly regulated by ROCK in a way that ROCK activation increases Ca²⁺ sensitivity whereas ROCK inhibition decreases the sensitivity to Ca²⁺, a process termed ‘Ca²⁺ sensitization’ [91]. Both cell culture experiments and studies using isolated vessel preparations demonstrated that the constrictor effects of Ang II are mainly mediated by ROCK-activation [92–94]. In vivo, overexpression of AT1R in rat podocytes induces proteinuria and glomerular disease [95].

Intrarenal autoregulatory mechanisms allow maintaining renal blood flow and glomerular filtration rate at varying arterial pressure. Myogenic vasoconstriction, an autoregulatory function of small arteries, is sustained by increased Ca²⁺ sensitivity, mediated by PKC and Rho/ROCK that favours a positive balance between myosin light-chain kinase and phosphatase [for review, see e.g. 96]. As emphasized by these authors, the effects of Ang II on the myogenic mechanism in the kidney are controversial.

Factors influencing the activities of renal Rho-GTPases

A number of factors are known to regulate Rho-GTPases [for general overview, see e.g. 97], including a variety of GEFs [82 identified in humans, 98], GTPase-activating proteins [GAPs, 66 known in humans, 99] and guanine nucleotide dissociation inhibitors [GDIs, three in humans, 100] which act to control the ratio of the GTP- and GDP-bound forms [101] but mutation analysis of patients with various forms of kidney diseases revealed new functional networks which regulate Rho-GTPase activities and which can be altered in disease-causing changeovers of the activity of these proteins. It was for example the case of mutations in six genes (MAGI2, TNS2, DCL1, CDK20, ITSN1 and ITNS2) encoding proteins which interact and form complexes involved in the regulation of RhoA and Cdc42 [102]. [103], identified mutations in genes encoding the nuclear pore complex proteins NUP107, NUP85 and NUP133 in patients with steroid-resistant nephrotic syndrome (see below) and demonstrated that knockdown of any of these three proteins in immortalized human podocytes led to the activation of Cdc42 and alterations of the podocyte’s cytoskeleton. The expression of Ca²⁺/calmodulin-dependent protein kinase IV (CaMK4) is increased in the glomeruli of patients with systemic lupus erythematosus [104], where it affects podocyte motility through an enhancement of expression of activated Rac1 and a suppression of the expression of activated RhoA [105]. Mutations in the anillin (an F-actin binding protein that modulates cell motility of podocytes) gene cause familial forms of FSGS (see below) cause hyperactivation of the PI3K/AKT/mTOR/Rac1 signalling pathway, which drives the reported dysmotility phenotypes [106].

2. Rho GTPases and kidney diseases

Glomeruli are susceptible to diseases induced by genetic mutations, infection, atherosclerosis, hypertension, diabetes, and autoimmune pathologies. Regardless of the initiating cause, renal fibrosis is the common pathogenic pathway that leads to progressive injury and organ dysfunction. This pathway involves several key steps [see 107], among them (i) an interstitial inflammatory response, (ii) the appearance of a unique interstitial cell population of myofibroblasts, (iii) the loss of the regenerative abilities of tubular epithelial cells and (iv) the fall of interstitial capillary integrity.

The first step is characterized by an interstitial infiltrate of macrophages, the density of which correlates inversely with kidney survival. It is tightly correlated with renal dysfunction, histological damage, and poor prognosis [108,109]. Animal studies demonstrated induction of kidney injury by pathogenic mediators of macrophages, improvement of renal injury and function by depletion of macrophages, and acceleration of renal injury by repletion of macrophages [see 110]. Accumulation of macrophages occurs via both the recruitment of circulating monocytes into sites of inflammation and local proliferation [see 111]. Macrophage colony-stimulating factor (M-CSF), a glycoprotein that regulates macrophage survival, proliferation, and differentiation, is required for macrophage proliferation. It is constitutively expressed in glomerular mesangial cells, tubular epithelial cells, and endothelial cells [see 111]. The Rho/ROCK pathway was found to be an important regulator of M-CSF production in vitro and in vivo [111].

Under the influence of mechanical forces and cytokines produced by cells that have infiltrated the interstitium, resident fibroblasts undergo activation and myofibroblasts are generated from bone-marrow-derived cells, pericytes and endothelial cells through the process of epithelial-mesenchymal transition [EMT, 112]. Sensu stricto, the term ‘EMT’ refers to the conversion of terminally differentiated epithelia (i.e. renal tubular cells) into cells with a mesenchymal phenotype, but a study in human biopsies from patients with chronic allograft nephropathy revealed that a significant part of interstitial mesenchymal cells was of bone marrow origin [see 113].

This leads to extracellular matrix deposition, progressive chronic kidney disease and ultimately renal tubulointerstitial fibrosis [see e.g. 114]. Activation of small RhoGTPases is considered as a key step in the mechanism of EMT [115]. Fasudil, a widely used inhibitor of the Rho/Rho-ROCKs-signalling pathway, which inhibits ROCK activity by competitively combining ATP sites of ROCK catalytic domain, was seen to attenuate the high glucose-induced EMT [116]. Benidipine, a triple calcium channel blocker, simultaneously blocking L, T, and N type channels, also inhibits ROCK1 [see 117]. The reduction in EMT and renal interstitial fibrosis in diabetic rat kidney by benidipine was ascribed to the inhibition of ROCK1 activity [117].

Acute kidney injury (AKI) is a heterogeneous syndrome characterized by an abrupt loss of kidney function associated with high morbidity and mortality. Incomplete AKI recovery can result in long-term functional deficits and has been recognized as a major contributor to chronic kidney disease (CKD) characterized by a gradual loss of renal function over a period of months to years. In addition to renal deficiency, CKD is also a major risk multiplier in patients with diabetes, hypertension, heart diseases, and stroke. Several studies have shown the interdependent relationship between AKI and CKD: AKI frequently results in permanent kidney damage (i.e. CKD) whereas CKD and proteinuria have been demonstrated to be risk factors for AKI [for recent review, see e.g. 118, 119].

As most other epithelial cells, renal epithelial cells display a polarized morphology, essential for their function. After AKI, these cells, normally mitotically quiescent, can however rapidly re-enter the cell cycle and their proliferation to ultimately replace the neighbouring cells that died as a result of the insult [120]. However, after this initially beneficial phase, the repair process becomes pathogenic through a fibrotic phase, in which connective tissues replace normal parenchymal tissue [121]. The reason for this transition is not well understood, but seems to involve cell-cycle specific factors, autophagy failure, endoplasmic reticulum stress, oxidative stress, and the loss of unknown ‘regenerative signals’ [see 107]. This fibrosis results in a permanent alteration of renal capillary density leading to a vicious cascade of hypoxia-oxidant stress that accentuates injury and fibrosis. Rac1, critical for ROS generation by TGF-β1, appears to serve as an upstream regulator of several noncanonical pathways that facilitate the activation of an array of fibrotic genes [for references, see 122].

Rho signalling pathway in diabetic nephropathy

Diabetic nephropathy (DN) or diabetic kidney disease refers to the deterioration of kidney function seen in chronic type 1 and type 2 diabetes mellitus patients. Components of the diabetic milieu such as hyperglycaemia, advanced glycation end products (ACEs), ROS, activation of hexosamine pathway, and oxidized LDL have all been shown to stimulate Rho-GTPase activity [see 123]. Excess intracellular glucose has been seen to activate cellular signalling pathways such as diacylglycerol-protein kinase C pathway, ACEs, polyol pathway, hexosamine pathway and oxidative stress [see e.g. 124], and many studies have linked these pathways to key steps in the development of non‐stoppable scarring process of renal glomerulus, known as glomerulosclerosis. In addition to these metabolic pathways, ROCKs have been linked to various steps in the ultra-structural damage of DN by inducing endothelial dysfunction, mesangial excessive extracellular matrix production, podocyte abnormality, and tubule-interstitial fibrosis. In addition to high glucose, ROCKs are stimulated by other components of the diabetic milieu, such as advanced glycation endproducts, ROS, hexosamine pathway and oxidized LDL in vascular and renal cells [for review, see e.g. 125, 126].

RhoA and diabetic nephropathy

RhoA, highly expressed in renal cortex, was seen to be activated in the renal cortex of diabetic rats [127], and high glucose levels to activate RhoA in rat renal mesangial cells [128]. In vitro studies revealed that high glucose mediated RhoA/ROCK activation in mesangial cells, causing a collagen IV accumulation via cytoskeletal remodelling, leading to mesangial matrix expansion, an effect significantly attenuated by fasudil and Y-27,632, selective ROCK inhibitors [129,130]. In cultured mesangial cells, ROCKs were suggested to regulate intracellular translocation of transcriptional factors (particularly RelA/p65), likely through re-organization of actin stress fibres [111]. In both mouse isolated glomeruli and cultured mesangial cells, knockdown by siRNA against ROCK1 and ROCK2 showed that ROCK2 but not ROCK1 controls the development of diabetic renal injury [131]. Hyperglycaemia-induced activation of RhoA also results in increased expression of pro-inflammatory (TNF-α and IL-1β) and fibrosis (TGF-β1) cytokines [see e.g. 132].

Pioglitazone is an efficient medication used in diabetes in several countries, active on type 2 DN [for recent review, see e.g. 133]. Although being a peroxisome proliferator-activated receptor gamma (PPARγ) agonist, it has also a strong inhibitory effect on Rho/ROCK pathway through tyrosine phosphatase SHP-2 [134]. This suggests a probable deleterious action of the RhoA/ROCK pathway in DN.

The Roundabout (ROBO) family of receptors and their ligands, known as SLIT glycoproteins, originally identified as important axon guidance molecules, are widely expressed in many cell lines including kidney. Homozygous mutants lacking either SLIT2 or its receptor ROBO2 died shortly after birth from kidney abnormalities [135]. SLIT-ROBO-Rho GTPase-activating proteins 2a (SRGAP2a), primarily localized at podocytes, is downregulated in patients with DN [see 136]. In cultured human podocytes, SRGAP2a was seen to inactivate RhoA and Cdc42 (but not Rac1) and thus to decrease podocyte motility and FP effacement [136].

Local tissue levels of Ang II in patients with glomerular diseases are much higher than in healthy subjects. Elevated Ang II in diabetes can transform a podocyte from a dynamically stable state to an adaptively migratory state. Ang II–induced activation of Rac1 activation was seen to increase ROS production, which causes a decrease in RhoA, leading to a migratory phenotype that contributes to chronic podocyte loss [87]. However, in cultured rat podocytes treated with Ang II, increments in RhoA and its downstream effector ROCK2 were observed, accompanied with altered morphology, redistribution of actin and increased phosphorylation of MLC phosphorylation [137]. The renoprotective effects of Ang II inhibitors in the treatment of chronic kidney diseases have been extensively studied [for recent reviews, see for example 138, 139].

In cultured rat glomerular mesangial cells, the RhoA/ROCK pathway was suggested to upregulate inflammatory genes, particularly nuclear factor κB (NF-κB), an important signalling pathway in the pathogenesis of DN [140]. The critical role of RhoA in DN both in vitro and in vivo was ascribed to a reduced nuclear protein expression of YAP [141]. Deletion of ROCK1 protected against the development of albuminuria in the model of streptozotocin-induced diabetic kidney disease [142], an effect consistent with the protective action of the ROCK inhibitor fasudil in the same mouse model [143] and in a spontaneous insulin-resistant diabetic rat model [144]. ROCK blockade was seen to attenuate the progression of diabetic glomerulosclerosis via downregulation of HIF-1α [145]. A significant body of evidence shows that the Notch signalling pathway plays a central role in podocyte apoptosis, a key step in the onset of diabetic nephropathy. In a murine podocyte cell line, ROCK was identified as a critical regulator of Notch signalling [146].

Mitochondria play a key role as major regulators of cellular energy homoeostasis but, in the context of mitochondrial dysfunction, are involved in a variety of metabolic diseases, including in the diabetic milieu [see e.g. 147]. In vitro studies have implicated mitochondrial fission as a key mediator of increased ROS production and cellular apoptosis under hyperglycaemic conditions [147]. ROCK1 was reported to play a critical role in the progression of DN by triggering mitochondrial fission [148].

Cdc42 and diabetic nephropathy

Cdc42 is a crucial factor during the progression of diabetes. It indeed not only actively participates in the process of insulin synthesis but also regulates the insulin granule mobilization and cell membrane exocytosis via activating a series of downstream factors [for recent review, see e.g. 149] but it is also involved in hyperglycaemia-linked pathologies, including DN. The reorganization of the cytoskeletal network mediated by Rac1 and Cdc42 was seen to play a major role in the cellular responses to podocyte-specific proteins [150,151]. In mouse models, the activation of N-Methyl-d-aspartate receptors plays an important role in diabetic kidney disease by reducing Cdc42-GTP activation [152]. In a murine podocyte cell line, increases in intracellular cAMP concentrations positively modulated the integrity of the cell-cell contacts, with activation of Rac1 and Cdc42 and inhibition of RhoA [153]. Podocytes exposed to Ang II or TGF-β1 exhibited a substantial cytoskeletal rearrangement and significant loss of stress fibres that extended across the entire cell (so called ‘arch F-actin’) through increasing the activity of Cdc42 [154].

A major component of the insulin signalling pathway, phosphoinositide 3-kinase (PI3K), simultaneously regulates Rac1 and Cdc42. Its inhibition by wortmannin, a furanosteroid metabolite of the fungi Penicillium funiculosum, Talaromyces wortmannii, was shown to improve DN in the early stages through the Rac1 and Cdc42 pathway [155]. In cultured rat podocytes, PI3K-dependent phosphorylation of Akt increased Rac1 activity and altered actin cytoskeleton with decreased stress fibres and increased lamellipodia [156]. Mechanistically, the significant downregulation of podocyte SRGAP2a in patients with DN would increase the active forms of Cdc42 and RhoA, leading to a greater podocyte migration and renal injury [136].

Normal expression of Cdc42 is essential (in vivo podocyte-specific deletion of Cdc42 leads to congenital nephrotic syndrome, glomerulosclerosis and death [26], but Cdc42 also appears as one of the factors which promote DN, particularly in podocyte injuries via the disturbed function of Cdc42 in podocyte cytoskeleton through a wide range of receptors and enzymes.

Rac1 and diabetic nephropathy

Ras and Rac-1 GTPases were reported to modulate the high-glucose- and glycooxidative-product-induced oxidative damage of mesangial cells and DN [157]. In rat mesangial cells, high glucose raised oxidative stress and progressively promoted apoptotic activity by activating GSK-3ß and inhibiting Wnt5a/ß-catenin signalling [157]. In diabetic rats, insulin administration reduced Rac1 overexpression [158].

In the type 1 diabetic mouse model, podocyte-specific deletion of Rac1 resulted in morphological alteration in podocytes, and the induction of apoptosis or decreased expression of the slit diaphragm proteins by hyperglycaemic stimuli were associated with the progression of DN [159]. In contrast, in STZ-induced DN mice, podocyte-specific Rac1 depletion was seen to attenuate diabetic podocyte injury through the blockade of Rac1/PAK1/p38/β-catenin cascade [160]. High-glucose stimulation increased Rac1 activity and mineralocorticoid receptor transcriptional activity in cultured mesangial cells, and Rac inhibition conferred beneficial effects on nephropathy of obesity-related diabetic mice [161]. Lysophosphatidic acid (LPA) is a critical regulator that induces mesangial cell proliferation. LPA levels are increased in the glomerulus of the kidney in diabetic mice [162]. LPA-induced hyperproliferation in mouse kidney was attenuated by the inhibition of Rac1 activity [163].

Kidney Rho proteins and arterial hypertension

The kidney is a target of hypertension drug therapies and the role of Rho GTPases in the control of this hypertension needs to be clarified in terms of mechanisms and therapeutic options. The RhoA/ROCK pathway is now widely known [for review, see e.g. 164] to play important roles in many cellular functions, including contraction, motility, proliferation, and apoptosis, and its excessive activity induces oxidative stress and promotes the development of cardiovascular diseases. 164,also highlighted the role of Rho GTPases in the regulation of blood pressure by the central nervous system. In rat preglomerular microvascular smooth muscle cells, the RhoA/ROCK inhibitors Y27632 and RKI-1447 partially inhibited voltage-dependent L-type Ca²⁺ channels signalling and substantially dilated the starting diameter of afferent arterioles [165]. With regard to the role on the kidney and its involvement in the regulation of blood pressure and in sodium metabolism, the activity of Rac1 plays an essential role via its direct relationship with the mineralocorticoid receptors [166,167]. In vitro, RhoA/ROCK signalling was shown to control the distribution and the activity of the sodium-hydrogen exchanger isoform 3 (NHE3), a transporter that participates in the renal reabsorption of Na⁺ under physiological conditions [168]. In vivo, high blood pressure in spontaneously hypertensive rats is associated with a decreased renal Na⁺ excretion and a significant increase in NHE3 activity [169]. These parameters are reversed by treatment with the ROCK inhibitor Y27632. A high-salt status was seen to act synergistically with aldosterone to activate renal Rac1 in salt-sensitive hypertension, leading to high blood pressure and renal damage through potentiating mineralocorticoid receptor signalling [170].

Rho GTPases and nephrotic syndrome

Nephrotic syndrome (NS) is caused by malfunction of the kidney glomerular filter, resulting in proteinuria, hypoalbuminemia, and oedema. NS is classified by its response to steroid treatment into steroid-sensitive nephrotic syndrome and steroid-resistant nephrotic syndrome categories. While the disease mechanisms of NS are still poorly understood, features of cell migration of podocytes seem implicated in its pathogenesis. In this concept, a stationary, sessile podocyte phenotype would represent the physiologic mode whereas a migratory phenotype would represent the nephrotic mode, and the podocyte health requires a well-controlled balance between the two extremes [33].

Rac1 activation in podocytes was observed in kidney biopsies from nephrotic syndrome patients, where it causes a spectrum of diseases ranging from minimal change disease to FSGS [27]. Hyper-activation of Rac1 via a doxycycline-inducible constitutively active form of Rac1 in murine podocytes caused proteinuria and FP effacement through p38 mitogen–activated protein kinase (MAPK)–dependent pathways [27,171]. Children with NS are typically treated with an empiric course of glucocorticoid therapy; this treatment was seen to reduce proteinuria-associated Rac1 overactivity [172]. Trio, a GEF for Rac1, is expressed in human podocytes and is significantly upregulated in glomeruli of patients with FSGS, where it might cause a pathological Rac1 hyperactivation [173].

The harmful effects of Rho GTPases in glomerular diseases are ascribed in main part to their deleterious effects on the actin cytoskeleton of podocytes, resulting in FP effacement and proteinuria. This change in podocyte architecture is remarkably dynamic [174]. If the molecular mechanisms that produce these changes in cytoskeletal dynamics remain incompletely understood, Rho GTPases are then thought to be central mediators of podocyte dysfunction in nephrotic syndrome. Mice lacking Rho GDI-α developed massive proteinuria mimicking nephrotic syndrome, leading to death as a result of renal failure [175].

Rho GTPases and congenital nephropathies

A substantial part of inherited and sporadic forms of FSGS (a scarring of the kidney’s filtration unit due to loss of podocytes, one of the most common glomerular syndromes) are caused by mutations in genes that encode regulators of Rho GTPase activities. Mutations in these genes, including the GTPase-activating protein (GAP), Rho-GAP 24 [ARHGAP24, a GAP for Rac1 which controls the RhoA-Rac1 signalling balance; 176], the Rho GDP Dissociation Inhibitor alpha [ARHGDIA, which encodes a GDI for Cdc42 and Rac1; 177, 178] or intersectin-1 and 2 [ITSN-1 and −2, which have Cdc42-GEF activity; 102] were shown to cause congenital nephrotic syndrome.

Mutations in genes encoding ARHGAP24 result in excess Rac1 signalling in podocytes [176]. A mutation in ARHGDIA leading to a loss-of-function of RhoGDIα, a negative Rho regulator, and resulting in the hyperactivation of Rho-GTPases and impaired cell motility, was observed in patients with an autosomal recessive congenital nephrotic syndrome [178]. Three mutations in ARHGDIA were also reported in patients with heritable nephrotic syndrome; Rac1 was markedly hyperactivated in podocytes in the three cases while the activation of RhoA and Cdc42 was modest and variable [179]. Genetic ablation of ARHGDIA in mice also resulted in renal abnormalities, including heavy albuminuria and podocyte damage, associated with RhoGDIα deficiency, increased Rac1 (but not RhoA) and mineralocorticoid receptor signalling in the kidney [167]. In zebrafish, ARHGDIA knockdown resulted in an exaggerated migratory phenotype in podocytes ascribed to the activation of Rac1/Cdc42 (but not RhoA), reversed by Rac1 inhibitors [177]. Excess Rac1 signalling leads to the vesicular insertion of TRPC5 ion channels into the podocyte plasma membrane [52] and results in transient Ca²⁺ influx into the podocyte, and further Rac1 activation, feeding a circuit that promotes cytoskeletal remodelling of podoctes [49,180].

In mouse podocytes, the expression of RhoGDIα was significantly reduced after the loss of Kindlin-2 (an important component of cell-matrix adhesions), resulting in the dissociation of Rac1 from RhoGDIα, leading to Rac1 hyperactivation and increased motility of podocytes [181]. In another study, the loss of Kindlin-2 in podocytes led to high levels of RhoA activation and concomitantly increased actomyosin contractility, associated with plasma membrane blebbing [182]. 183,suggested that the interaction between RhoGDIα and ezrin (an actin-binding protein highly expressed in glomerular podocytes and proximal tubules in kidneys) would be important for the regulation of Rho-GTPase activity in glomerular podocytes.

Autosomal dominant polycystic kidney disease (ADPKD) is an inherited disorder caused by mutations in either the PKD1 or PKD2 gene, encoding polycystin 1 and 2, respectively (PKD1 or PKD2), frequently leading to end-stage renal disease [184] and affects one in 500–1000 humans. In a recent study, the Hippo signalling effector YAP and its transcriptional target, c-Myc, were identified as critical mediators of cystic kidney pathogenesis after PKD1 is inactivated [185]. In addition, a signalling pathway involving the Rho GEF LARG, RhoA, its effector ROCK, and myosin light chain (MLC) was identified as a critical signalling module between PKD1 and YAP [185].

In Alport syndrome, hereditary glomerulonephritis with hearing loss and eye abnormalities, a stretch-mediated activation of Rac1 and Cdc42 was seen to induce mesangial filopodial invasion of the glomerular capillary loops [186].

RhoGTPases and other nephropathies

The HIV-1 accessory protein Nef is considered to play an important role in the development of a podocyte phenotype in HIV-1 associated nephropathy. Nef was seen to inhibit RhoA activity through stimulation of p190RhoAGAP and to simultaneously activate Rac1, allowing Nef to influences podocyte morphology [187]. According to [188], Nef would compromise cytoskeletal integrity of human podocytes via both a direct interaction with actin and an enhancement of the expression of small GTPases Rac1, Cdc42 and Rif. The HIV-1 transactivator of transcription (Tat), combined with fibroblast growth factor-2 (FGF-2), can induce the dedifferentiation and proliferation of cultured human podocytes [189]. These authors observed that Tat induced cytoskeletal changes in podocytes through the stimulation of the RhoA/phospho-Myosin Light Chain 2 (pMLC₂) pathways.

The mice on a high fat diet not only develop obesity but also display renal histological changes; including glomerular hypercellularity and increased mesangial matrix, which paralleled the increase in albuminuria, together with enhanced ROCK activity [see e.g. 190]. Obesity is associated with increased aldosterone (a potent mineralocorticoid that promotes renal sodium retention) production in humans and increased aldosterone and mineralocorticoid receptor levels in mice [191]. Several studies have documented a potential role of aldosterone in the pathogenesis of renal injury [190]. In obese mice, the elevation in renal tissue aldosterone contents and the activation of the mineralocorticoid receptor-mediated signalling pathway was seen to contribute to obesity-associated nephropathy via Rho/ROCK activation [190].

Rhabdomyolysis occurs when trauma, drug overdose, fever, or inflammation causes skeletal muscle destruction and disintegration, leading to leakage of potentially toxic cellular contents into the systemic circulation. AKI is one of its severe, potentially life-threatening complications. In a rat rhabdomyolysis model of AKI, early application of the ROCK inhibitor fasudil reduced rhabdomyolysis-induced AKI [192].

Surprisingly, even though ROCK inhibitors showed positive effects in a variety of animal models of nephropathy [for references, see 13], ROCK1^−/− mice were not protected against renal fibrosis in the unilateral ureteral obstruction model [193]. ROCK1 deletion was suggested to enhance TGF-β/Smad signalling in the obstructed kidney, which can explain, at least in part, why ROCK1 deletion failed to limit renal fibrosis [193].

RhoGTPases and cancer treatment-induced nephrotoxicity

In clinical oncology, nephrotoxic effects of targeted therapeutics, such as protein kinase inhibitors, is a growing concern as many of these agents are used as long as clinical benefits are being observed. The multi-kinase inhibitor dasatinib was seen to induce nephrotoxicity through altered actin cytoskeleton of podocytes by affecting RhoGTPase signalling, leading to injurious cellular biomechanics without apparent hypertension [194]. In the transformed human embryonic kidney cell line 293 T, the anticancer phytochemical rocaglamide A did not only exert cytotoxic and cytostatic effects but also inhibited tumour cell migration via inhibition of the activities of Rho GTPases RhoA, Rac1 and Cdc42 [195].

Rho downstream effectors in autoimmune diseases

In addition to their action on actin cytoskeleton organization, cell adhesion and motility, the ROCKs also regulate cell proliferation, differentiation, and apoptosis. Their deregulation now appears to be linked to pathophysiology in several autoimmune disorders, suggesting that it may serve as a common pathogenic pathway in autoimmunity [see 196]. Dysregulated production of interleukin (IL)-17 and IL-21 is considered to play a key pathogenic role in many autoimmune disorders [197], such as systemic lupus erythematosus (SLE) and rheumatoid arthritis, among others [198]. Lupus nephritis, an inflammation of the kidneys, is a major cause of mortality in SLE patients. The involvement of the ROCK signalling pathway in the development of kidney disease is supported by the protective action of ROCK inhibitors in a variety of animal models of nephropathy [for references, see e.g. 13].

SLE is characterized by autoantibody production and abnormal T cells that infiltrate tissues (e.g. kidneys, central nervous system, joints, skin, and cardiovascular system) through not well-known mechanisms. It was suggested that ROCK-mediated actin polymerization was responsible for increased cap formation, adhesion, and migration of SLE T lymphocytes [199]. Selective inhibition of ROCK2 was suggested to represent a potentially important therapeutic regimen for the treatment of human autoimmune disorders [197]. The deregulated production of IL-17 and IL-21, as well as the inflammatory and autoantibody responses observed in a lupus mice model were indeed ameliorated by administration of the non-selective ROCK inhibitor fasudil [200]. A majority of SLE patients exhibited higher levels of ROCK activity than healthy controls [198,201]. The selective inhibition of ROCK2 down-regulated the production of IL-17 and IL-21 by T cells from these patients [198].

Pleiotropic beneficial effects of statins

The 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors or statins are potent inhibitors of cholesterol biosynthesis. Both experimental and clinical evidence suggests that their beneficial effects may extend beyond their cholesterol-lowering effects to involve so-called pleiotropic effects, including amelioration of kidney disease [see e.g. 128, 129, 202]. The mechanism underlying some of these pleiotropic effects is the inhibition of the rate-limiting enzyme of the l-mevalonate pathway, allowing statins to reduce isoprenyl residues (farnesyl or geranylgeranyl) which are required for the correct attachment of different small GTPases [203], including Rho, Rac and Cdc42 to lipid membranes.

In diabetic mice, atorvastatin attenuates renal damage through ROS- and RhoA-dependent mechanisms [204]. In in vitro cell cultures and kidney slices and in Brattleboro rats, simvastatin induced aquaporin-2 membrane accumulation in the collecting duct in kidney cortex and the outer medullary region, which results in increased water reabsorption and urinary concentration in vivo, an effect ascribed to a significant downregulation of the RhoA activity [205].

Statin treatment was found to inhibit both tumour initiation and progression of clear cell renal carcinoma tumours in mice, an effect ascribed to the disruption of GTPase isoprenylation and partially through the inhibition of Rho/ROCK1 signalling [206]. The high-glucose-induced Rho/ROCK activation was attenuated by pitavastatin in murine cultured podocytes [207]. Rho proteins thus appear as one of the molecular targets of statins through the inhibitory effects of statins on RhoA activation, which might play an important role in statin-induced renoprotection mechanisms.

3. Rho GTPases in renal transplantation

Kidney transplantation is the optimal treatment of patients with end-stage renal disease who would otherwise require dialysis. Renal transplant recipients remain at an increased risk of fatal and non-fatal cardiovascular events compared to the general population, although at lower rates than those patients on maintenance haemodialysis. Mortality with a functioning graft is most commonly due to cardiovascular disease and cardiovascular-related deaths, which therefore remains an important therapeutic target to prevent graft failure. Moreover, transplantation is also accompanied by post-transplant diabetes mellitus, which remains an important therapeutic target to prevent graft failure. In addition, following kidney transplantation, recurrent or de novo glomerulonephritis in the renal allograft is an important cause of premature allograft failure [for ref., see e.g. 208].

Post-transplant hypertension in kidney transplant recipients

Most of the kidney transplanted patients [70–90% of them; 209] develop arterial hypertension after renal transplantation, a factor which significantly impact on the long term outcomes in the kidney transplant recipients. The pathogenesis of this hypertension is multifactorial, but one of the most important factors is the introduction in immunosuppressive therapy of calcineurin inhibitors (tacrolimus or ciclosporin), with or without corticosteroids. Among the involved mechanisms, a key role is played by the activation of intrarenal RAAS [see e.g. 210]. A ROCK-mediated vascular smooth muscle cell constrictor tone was suggested to be responsible for human post-transplant hypertension [211]. Ciclosporin A (CsA) and CsA/sirolimus (CsA/SRL) combination were reported to activate RhoA and its effectors (particularly ROCK) in renal proximal tubular cells, leading to cytoskeleton rearrangement [212].

In addition, as the activation of ROCK is also known to downregulate the endothelial NO synthase [211,213], so that enhanced ROCK activity may also contribute to the reduced bioavailability of NO observed in post-transplant hypertension [210].

Post-transplant ischaemic acute kidney injury (AKI)

Ischaemia-reperfusion injury (IRI) is one of the major causes of high morbidity, disability, and mortality in the world. Ischaemia-reperfusion (IR), which occurs during renal transplantation is one of the factors that affect the outcome of renal transplantation. To bridge the gap between supply and demand for organ donation, increasing use of suboptimal deceased donors (high risk, extended criteria, marginal donors) and uncontrolled non-heart-beating donors is now the accepted practice. Kidneys from such donors are exposed to much greater ischaemic damage before recovery and show reduced chances for proper early as well as long-term function. Consequently, it is of paramount to have an in depth understanding of the mechanisms governing kidney response to IR if one wants to define pertinent biomarkers or to elaborate targeted therapeutic interventions.

Post-transplant AKI, secondary to IRI, is a major problem influencing on the short and long term graft and patient survival. AKI frequently indeed leads to chronic kidney disease. In a rat renal ischaemia-reperfusion model [214], observed that Rac1 is the only Rho GTPase to undergo a translocation from membranes to the soluble fraction during reperfusion and suggested this could be related to ROS generation in reperfused proximal tubules. Pharmacologic inhibition of Rac1 reduced oxidant stress and ischaemic injury in vivo in a murine model of kidney IRI [215].

In the Large White pig, a model of human AKI, Rac1 activity is increased in ischaemic kidneys and the elimination of Rac1 in smooth muscle cells prevents renal injury associated with IR [216]. In mice, pivotal role of Rac1 in renal fibrosis after IRI was ascribed to regulation of expression of profibrotic cytokines and chemokines, bone-marrow derived M2 macrophage recruitment, and the transition of M2 macrophages to myofibroblasts. IL-4 treatment activated Rac1 [217]. Another approach using conjugation of the ROCK inhibitor Y27632 with lysozyme markedly reduced renal inflammation and renal lymphangiogenesis during acute transplant rejection [218]. A previous study had investigated the potential activity of Y27632-lysozyme in the IRI model and demonstrated both enhanced activity and increased selectivity of the tubular-targeted ROCK inhibitor. In addition, the renally targeted Y27632-lysozyme conjugate strongly inhibited tubular damage, inflammation, and fibrogenesis induced by IRI [219].

Indirect RhoH involvement in acute and chronic kidney transplant rejection

RhoH (also known as translocation three four (TTF) is encoded by a haematopoiesis-specific Rho-related gene and its expression is restricted to haematopoiesis‐ related organs (bone marrow, thymus and spleen) and in particular to haematopoietic progenitor cells as well as fully differentiated myeloid and lymphoid cells [220,221]. As this atypical Rho GTPase has no functional intrinsic GTPase activity, it is thought to be constitutively active and controlled only at the transcriptional level, and RhoH expression is indeed regulated in lymphocytes [221].

In murine allogenic kidney transplantation, RhoH deficiency was seen to cause a significant 75% reduction of acute and chronic transplant rejection accompanied by 75% lower alloantigen-specific antibody levels and significantly better graft function [222]. The numbers of peripheral T cells have been found to be markedly reduced in RhoH‐deficient mice [223]. But the reduced acute and chronic transplant rejection was independent of the lower T-cell numbers in RhoH-deficient recipients since injection of equal numbers of RhoH-deficient or control T cells into kidney transplanted mice with SCID led again to a significant 60% reduction of rejection [222].

Emerging evidence suggest a critical role of inflammatory cells to the pathogenesis of early peritransplant IRI and subsequent allograft rejection immune cascade [for review, see e.g. 224, 225]. Recently, the role of RhoA as a key regulator of innate and adaptive immunity was discussed [226]. The contribution of RhoA for the primary functions of innate immune cell types, namely neutrophils, macrophages, and conventional dendritic cells (DC) has been established. In activated DC, which constitute the most potent antigen-presenting cells of the immune system, RhoA is also important for the presentation of pathogen-derived antigen and the formation of an immunological synapse between DC and antigen-specific T cells as a prerequisite to induce adaptive T cell responses. Interestingly, the use of Abs to block CD1d in DC cells induced a decrease of IRI [227]. It must be noted that CD1d membrane exposition is negatively regulated by ROCK [228]. In T cells and B cells as the effector cells of the adaptive immune system, Rho signalling is pivotal for activation and migration. In leukocytes, polarization and migration are involved in immune interactions. Interconnected processes are regulated by Ezrin, radixin and moesin proteins (ERMs) and it was suggested that Rho GTPases could be activated in the lymphocyte leading edge, regulating the actin polymerization required for cell advance, and inactivated in the posterior pole, inducing uropod formation [229]. A recent review focused on these mechanisms involved in the immune synapse in macrophages/neutrophils and lymphocytes, respectively, which represent essential aspects of the effector immune response [230].

Non-HLA antibodies in kidney allograft rejection

Pathogenic alloantibodies specific to organ donor human leukocyte antigens (HLAs) have been involved in most cases of antibody-mediated rejection in solid organ transplant recipients. However, the production of non-HLA autoantibodies can also occur before transplant in the form of natural autoantigens. In contrast with HLAs, constitutively expressed on the cell surface of the allograft endothelium, these autoantigens are usually cryptic. Tissue damage associated with IR, vascular injury and/or inflammation can cause the release of sequestered autoantigens, resulting in the presentation of cryptic self-determinants and thereby triggering an autoimmune process at the site of the graft. Patients who received a kidney from a deceased donor presented decreased graft survival in the presence of a specific non-HLA antibody against Rho GDP‐dissociation inhibitor 2 [ARHGDIB; 231, 232].

Conclusion, key importance of Rho GTPase activity to kidney function

The kidneys perform several essential functions in vertebrates including hormone secretion, blood pressure regulation, maintenance of glucose homoeostasis, and urine formation. The latter process begins at the level of the glomerulus, the most proximal portion of the nephron, the kidney’s functional unit. The integrity of the actin cytoskeleton is of critical importance for maintenance of renal glomerular architecture by mesangial cells, as well as the normal glomerular filtration function by podocytes. Recent decades have tremendously improved our knowledge concerning the structure, function and regulation of the podocyte cytoskeleton, and RhoGTPases are master regulators of actin cytoskeletal dynamics. The concept of the kidney filtration barrier has changed from one of a static sieve into one of a highly dynamic structure regulated through the motility of podocyte FPs. Rho GTPases and their downstream effectors then appear as key players in the regulation of cytoskeletal rearrangement in podocytes and renal function. In addition, they are involved in a variety of ailments and different compartments of the kidney: the nephron, the vascular compartment, and the immune compartment. Overall, the consistent data obtained from human genetic analyses, whole-animal models, and cell-based studies provide some reassurances and cautionary notes regarding the potential for targeting Rho GTPases for the treatment of renal diseases. The agreement of findings between species and model systems indeed confers confidence in the fidelity of these models and their relevance to human diseases. However, given the broad range of biological functions of Rho GTPases, potential therapeutic targets must be rigorously investigated to identify signalling pathways with minimal ‘off-target’ effects.

We also need a more detailed understanding of how the cytoskeleton itself monitors several critical cellular processes such as localized transcription, vesicle trafficking, mitochondrial dynamics, and even nuclear transcription. A better understanding of underlying mechanisms and involved agents would help in the design and rational development of targeted therapeutic approaches.

Disclosure of potential conflicts of interest

No potential conflict of interest was reported by the author(s).