Hypoxia-inducible factor signaling in the development of kidney fibrosis

Despite the very large blood flow (~20% of total cardiac output), the kidneys, which carry out complex and energy consuming cellular transport functions, operate under markedly reduced oxygen tension, with regional oxygen levels ranging from 10 to 60 mmHg. A reduction in renal oxygenation occurs in most chronic kidney diseases (CKD) irrespective of etiology. This is due to a combination of several pathophysiological and morphologic changes, which are typically associated with chronic kidney injury. These include increased oxygen demand from hyperfiltration and tubular hypertrophy, capillary rarefaction, glomerular injury, luminal narrowing of atherosclerotic vessels, as well as vascular constriction due to altered expression of vasoactive factors and signaling molecules (e.g. angiotensin II, endothelin, nitric oxide). The resulting reduction in renal oxygen availability is furthermore exacerbated by extra-cellular matrix (ECM) expansion, which limits oxygen diffusion, and by renal anemia [1, 2]. Blood oxygen level-dependent (BOLD) MRI, molecular and histological techniques, as well as measurements of renal oxygen levels with microelectrodes have been used to assess tissue oxygenation in chronic kidney diseases, including diabetic and IgA nephropathy, obstructive nephropathy, fibrosis associated with 5/6 nephrectomy and anti-Thy1 glomerulonephritis (for an overview of these studies see [1–4]).

Consistent with decreased renal oxygenation in CKD is the increased expression of the oxygen-sensitive α-subunit of hypoxia-inducible factor (HIF)-1 in renal biopsy material from patients with CKD [5, 6]. The heterodimeric basic helix-loop-helix transcription factors HIF-1 and HIF-2 are key mediators of cellular adaptation to hypoxia, and belong to the PAS {PER/aryl-hydrocarbon-receptor nuclear translocator (ARNT)/single minded (SIM)} family of transcription factors. They consist of an oxygen-sensitive α-subunit and a constitutively expressed β-subunit, which is also known as the aryl hydrocarbon receptor nuclear translocator (ARNT), and facilitate both oxygen delivery and cell survival by stimulating erythropoiesis, angiogenesis and anaerobic energy metabolism. HIF-1 and HIF-2 are furthermore involved in the regulation of biological processes that are relevant to wound healing, tissue repair and fibrogenesis, such as extracellular matrix synthesis and turnover, cell adhesion and migration, and epithelial to mesenchymal transition (EMT) [7–12].

HIF heterodimers activate gene transcription in response to hypoxia by binding to specific DNA sequences, which are known as hypoxia-response elements (HREs) and by recruiting transcriptional co-activators such as CBP/p300 (Figure 1). While HIF-α is continuously synthesized, it is rapidly degraded under normoxia, keeping HIF signaling at minimal levels when oxygen tension is in normal range. HIF degradation under normoxia requires hydroxylation of specific proline residues within the oxygen-dependent degradation domain of HIF-α, enabling interaction with the von Hippel-Lindau tumor suppressor pVHL, which functions as the substrate recognition component of an E3 ubiquitin ligase complex [13, 14]. HIF hydroxylation depends on the presence of molecular oxygen, ferrous iron and ascorbate, and is carried out by 2-oxoglutarate-dependent dioxygenases (prolyl-4-hydroxylase domain (PHD) proteins). Three major HIF-hydroxylating enzymes have been identified, PHD1, 2 and 3, of which PHD2 is most important for normoxic HIF degradation [14]. A second hypoxic switch operates in the carboxy-terminal transactivation domain of HIF-α with the hydroxylation of an asparagine residue. Under hypoxic conditions asparagine hydroxylation is inhibited and CBP/p300 recruitment facilitated, enabling increased levels of transcription [14].

While HIF has been shown to be cytoprotective in acute kidney injury [3], work by our laboratory and others now suggests that prolonged activation of HIF signaling in renal epithelial cells enhances maladaptive responses, which lead to fibrosis and tissue destruction. Specifically, our laboratory has identified epithelial HIF-1α as a promoter of kidney fibrosis, and has demonstrated that HIF-1 activation stimulates collagen accumulation and inflammatory cell recruitment in experimental models of CKD [5, 15].

Renal epithelial cells contribute to the development of kidney fibrosis, as they increase and remodel ECM when stimulated with TGF-β1, angiotensin II and other cytokines [29, 30], or when they transition into myofibroblasts as a result of EMT [23, 24]. Increased HIF expression has been found in animal models of CKD and in renal biopsy material from patients with diabetic nephropathy and other forms of renal disease, where it correlates with disease severity [5, 31, 32]. Our laboratory has identified epithelial HIF-1 as a promoter of renal fibrosis in experimental UUO [5]. The biological outcome of activated HIF signaling, however, is different under conditions of acute renal hypoxia [3]. HIF-1α can be detected in the nucleus of renal tubular epithelial cells, where it dimerizes with HIF-1β to form transcriptionally active HIF-1. Whereas HIF-1α is expressed in renal epithelial and endothelial cells, HIF-2α is expressed in erythropoietin (EPO)-producing renal interstitial fibroblasts, endothelial and glomerular cells [3]. Genetic studies in mice have demonstrated that reduced expression of either HIF-1α or HIF-2α worsens clinical outcome of acute ischemic kidney injury [33, 34]. While the cell types that contribute to HIF-mediated cytoprotection remain to be determined, endothelial HIF-2 appears to ameliorate renal ischemia reperfusion injury through increased expression of ROS scavenging enzymes, such as superoxide dismutase (SOD) [33].

In the context of chronic renal injury the biological consequences of HIF activation are different, and negatively impact clinical outcome, that is, promote fibrosis, ultimately leading to the development of end stage renal disease. This may occur through increased expression of i) ECM-modifying genes, such as LOX and PAI-1, ii) functional co-operation with transforming growth factor (TGF)-β1, iii) promotion of epithelial to mesenchymal transition (EMT), and iv) through modulation of renal inflammation [35]. Hypoxia induces collagen I, decreases matrix-metallopeptidase 2 (MMP-2) in renal epithelial cells [36], and increases PAI-1 [18], tissue-inhibitor of metalloproteinase-1 (TIMP-1) [36] as well as connective tissue growth factor (CTGF) [16] through HIF-mediated transcriptional responses. Synergistic co-operation between HIF and non-HIF pathways, such as the pro-fibrotic TGF-β1/SMAD3 signaling pathway, may further enhance the expression of HIF-regulated genes in the CKD setting; HIF-1 and TGF-β1 have been shown to co-regulate vascular endothelial growth factor (VEGF) [37], endoglin [38] and EPO [39]. The notion of HIF and TGF-β1 working together in the context of CKD is furthermore supported by the observation that hypoxia synergizes with TGF-β1 with regard to the production of certain collagens [40, 41].

While hypoxia is the main stimulus for HIF activation, signaling molecules with key roles in the pathogenesis of CKD, such as angiotensin II (Ang II), have also been shown to activate HIF-1 via prolyl-4-hydroxylase inhibition. PHD inhibition in this setting results from increased ROS generation and diminished intracellular ascorbate levels, leading to the oxidation of a critical iron atom that is positioned in the center of the PHD catalytic domain [42]. Although AngII-generated ROS is largely derived from activation of NADPH oxidase [43], others have suggested that AngII-induced HIF-1α stabilization may be mediated by mitochondrial ROS [44]. In this context, recent studies in renal interstitial fibroblasts have proposed that HIF-1 functions downstream of an Ang II-induced pro-fibrotic signaling cascade, which stimulates EMT and the production of collagen [45].

There is growing experimental evidence that hypoxia via HIF promotes EMT. This occurs through i) modulation of activity of EMT inducers or their receptors (e.g. TGF-β1; c-met proto-oncogene [28]), ii) regulation of expression levels and activity of EMT-associated signaling molecules and downstream effectors (e.g. Notch and β-catenin [46, 47]), and iii) via regulation of expression and activity levels of EMT-inducing transcriptional repressors, such as snail (SNAI1) [48]. In the context of CKD, renal tubular epithelial cells loose apico-basal polarity, become motile and acquire a mesenchymal phenotype. It has been proposed that they are able to migrate into the interstitium, where they together with resident cells would produce ECM as myofibroblasts [24]. Irrespective of cellular origin, the accumulation of myofibroblasts closely correlates with the degree of interstitial damage and the risk of disease progression. Earlier in vivo studies using genetically tagged renal epithelial cells have suggested that up to 36% of interstitial myofibroblasts are EMT-derived, whereas approximately 15% stem from bone marrow, with the remainder being of resident fibroblast origin [21]. The degree to which EMT contributes to the renal myofibroblast pool, however, is intensely debated and remains unclear [49]. More recent studies indicate that pericytes [50] and/or endothelial cells [51] are major sources of ECM-producing myofibroblasts.

HIF-1 induction of LOX and LOXL-2 has been shown to promote migration in primary renal epithelial, and in breast and cervical cancer cells, which was associated with decreased E-cadherin expression [5, 25]. Although initially identified by their ability to crosslink collagen and elastin fibers, LOX and LOXL proteins also carry out intracellular functions and display a range of biological activities that extend beyond ECM cross-linkage [26]. In a recent report Peinado et al. demonstrated that LOXL2 and LOXL3 have the potential to regulate EMT by stabilizing and promoting the activity of transcriptional repressor SNAI1 [52].

Our finding that HIF activation in epithelial cells promotes renal fibrosis has immediate clinical implications, as it encourages therapies that aim at improving tissue oxygenation to retard disease progression. Because HIF regulates multiple biological processes, which include erythropoiesis and iron metabolism, its systemic pharmacological inhibition is not desirable. Certain HIF regulated proteins, however, may represent better therapeutic targets to slow progression of CKD. Pharmacological inihibition of lysyl oxidases phenocopied the effects of genetic HIF-1 inactivation on cell motility and fibrogenesis, suggesting that lysyl oxidases are important contributors to the pathogenesis of renal fibrosis. In keeping with this notion, increased LOXL2 expression was found in renal biopsy tissues from patients with CKD underscoring its relevance to the pathogenesis of CKD irrespective of etiology (Figure 5). Whether pharmacological inhibition of lysyl oxidases is feasible in clinical practice to slow progression of CKD certainly warrants further investigation.