Hypoxia-inducible factor signaling in the development of tissue fibrosis

Abstract

Capillary rarefaction is a hallmark of fibrotic diseases and results in reduced blood perfusion and oxygen delivery. In the kidney, tubulointerstitial fibrosis, which leads to the destruction of renal tissue and the irreversible loss of kidney function, is associated with hypoxia and the activation of Hypoxia-Inducible-Factor (HIF) signaling. HIF-1 and HIF-2 are basic-helix-loop-helix transcription factors that allow cells to survive in a low oxygen environment by regulating energy metabolism, vascular remodeling, erythropoiesis, cellular proliferation and apoptosis. Recent studies suggest that HIF activation promotes epithelial to mesenchymal transition (EMT) and renal fibrogenesis. These findings raise the possibility that the spectrum of HIF activated biological responses to hypoxic stress may differ under conditions of acute and chronic hypoxia. Here we discuss the role of HIF signaling in the pathogenesis and progression of chronic kidney disease.

Introduction

Fibrosis leads to the destruction of tissues and irreversible loss of normal tissue function. Tissue fibrosis as a maladaptive accumulation of extracellular matrix (ECM) is a common pathway in organs that sustain chronic injuries and is the result of a multifaceted, multilayered cellular response that involves epithelial to mesenchymal transition (EMT), the activation of fibroblasts to produce ECM, recruitment of inflammatory cells, and cellular regeneration at sites of damage.

In the kidney, tubulointerstitial fibrosis is a poor prognostic indicator and represents a common final pathway in the development of end-stage renal disease (ESRD) irrespective of the underlying cause. Morphologically, tubulointerstitial fibrosis is associated with tubular atrophy, glomerulosclerosis, interstitial inflammation and loss of peritubular capillaries, which together with concomitant functional changes such as post-glomerular hypertension, vasoconstriction, and proteinuria lead to loss of nephron mass.

Hypoxia resulting from capillary loss and decreased blood flow, has long been thought to play an active role in the progression of chronic kidney disease (CKD).^1–3 Key mediators of global cellular adaptation to hypoxia are Hypoxia-Inducible Factors (HIFs), HIF-1 and HIF-2 being the most extensively studied. HIF-1 and HIF-2 (here collectively referred to as HIF unless individually mentioned) are members of the Per-ARNT-Sim (PAS) family of basic helix-loop-helix (bHLH) transcription factors and consist of an oxygen sensitive α-subunit and a constitutively expressed β-subunit, also known as the Aryl Hydrocarbon-Receptor Nuclear Translocator (ARNT) or simply HIF-β. As global regulators of oxygen homeostasis, HIF transcription factors facilitate both oxygen delivery and adaptation to oxygen deprivation by regulating angiogenesis, erythropoiesis, anaerobic glycolysis, as well as cellular proliferation and apoptosis (for a detailed overview see Wenger⁴ or Schofield and Ratcliffe⁵). Although the acute activation of HIF signaling results in cytoprotection and has been shown to improve cell survival in ischemia-reperfusion injuries of different organs such as the heart and kidney,^6,7 the functional role of HIF under chronic disease conditions is not well understood. One major interest of our laboratory is to understand the biological consequences of hypoxic HIF activation in the development and progression of chronic kidney disease. Here we discuss recent findings that investigated the role of HIF signaling in the development of renal fibrosis.

HIF Activation in Fibrotic Kidney Disease

Despite a very large renal blood flow (~20% of total cardiac output), kidneys, as a consequence of their unique vascular architecture, have to carry out complex and energy consuming cellular transport functions under markedly reduced oxygen tension, which can be as low as 10 mmHg in certain anatomical regions.⁸ This makes this organ particularly susceptible not only to acute ischemic injuries but also to chronic disease processes that affect tissue oxygenation.⁹ Hypoxia has been detected in several chronic kidney disease models (reviewed in ref. 7), and can be attributed to capillary rarefaction, vasoconstriction and luminal narrowing of atherosclerotic arteries resulting in reduced blood perfusion, and to a decrease in oxygen carrying capacity as a consequence of anemia. In a recent study we detected HIF-1α immunohistochemically in an experimental mouse model of renal fibrosis and in human tissue samples from CKD patients with diabetic and IgA nephropathy (Fig. 1 and Higgins et al.,¹⁰). In this study the number of cells expressing HIF-1α in diabetic kidneys (glomerular and tubular cells) correlated with the degree of tissue injury and fibrosis. Genome-wide gene expression analysis of microdissected renal biopsy tissues from control and diabetic patients (glomeruli were excluded) detected that approximately 45 confirmed or putative HIF regulated genes were differentially expressed in either renal tubules and/or interstitial cells.¹⁰ Taken together our findings provided correlative evidence for the activation of HIF signaling in CKD tissues. They furthermore suggested that the hypoxic activation of HIF may be relevant to the development and progression of kidney fibrosis. This was subsequently investigated in mice that were HIF-1α deficient in proximal tubule epithelial cells, and that were subjected to experimental obstructive nephropathy, which is characterized by the development of severe fibrosis over the course of 2 weeks.

Figure 1.

Detection of nuclear HIF-1α in paraffin embedded renal biopsy tissue sections from a patient with IgA nephropathy. (A) Masson Trichrome (MT) stain. Blue staining indicates the presence of collagen. (B) Immunohistochemical stain for HIF-1α in an adjacent section. Arrows point to tubular cells expressing HIF-1α. (C) Immunohistochemical stain for FSp1. Arrows point to FSp1-positive cells associated with fibrotic areas. Magnification ×200.

HIF-1 Regulates EMT and Promotes Renal Fibrosis

We found that hypoxia through activation of HIF-1 enhanced EMT in vitro (Fig. 2) and that genetic ablation of epithelial HIF-1α in a mouse model of renal fibrosis resulted in a decrease of fibroblast-specific protein 1 (FSp1) expressing transitioned cells, which was associated with a reduction in ECM accumulation. Pharmacological inihibition of lysysl-oxidases phenocopied the effects of genetic inactivation of HIF-1α on cell motility in vitro and on fibrogenesis in vivo suggesting that lysyl oxidase genes are downstream of HIF and play an important role in the regulation of epithelial cell motility and EMT, and the development of fibrosis in vivo.¹⁰ EMT is a major contributor to the development of renal fibrosis and is characterized by the disassembly of cell to cell contacts such as E-cadherin based adherens junctions.^11–13 This leads to cell-cell separations associated with re-organization of the actin cytoskeleton, ultimately resulting in the generation of fibroblast-like cells, which express mesenchymal markers such as vimentin and FSp1¹⁴ de novo, and display increased motility and invasiveness.¹¹ Renal epithelial cells contribute largely to the development of renal fibrosis, as they increase and remodel ECM when stimulated with TGFβ1, angiotensin II and other cytokines,^15,16 or when they transition into myofibroblasts as a result of the EMT process.^11,12

Figure 2.

Hypoxia induces a mesenchymal phenotype in primary renal epithelial cells. Primary cells were lineage-tagged with a lacZ reporter using Cre-loxP mediated recombination and de novo expression of fibroblast marker FSp1 was monitored by immunoflourescence. Hypoxia (5 days, 1% O₂) increased the percentage of FSp1 expressing cells. (A–C) Shown is immunoflourescence staining for lacZ in red (A) and for FSp1 in green (B). Lineage-tagged epithelial cells, which stain for both lacZ and FSp1 and fluoresce in yellow are shown in (C). De novo expression of FSp1 indicates that these cells have undergone transition to a mesenchymal phenotype. (D and E) LacZ and FSp1 immunofluorescence stain of primary cells exposed to normoxia (D) and hypoxia (E). The percentage of epithelial cells that express FSp1 de novo increases under hypoxia (yellow fluorescence) as illustrated in (E). For technical details of this analysis see Higgins et al.,¹⁰ magnification ×400.

The importance of HIF signaling in the regulation of EMT is furthermore underpinned by studies in cancer cells that were defective in the von Hippel-Lindau (VHL) tumor suppressor. Functional loss of the VHL gene product results in constitutive activation of HIF signaling, as it is required for proteasomal degradation of HIF-α under normoxia. In VHL-deficient renal cancer cell lines Krishnamachary et al., reported that HIF-1 activation downregulated E-cadherin, led to the loss of cell-cell adhesion and promoted epithelial to mesenchymal transition through the induction of transcriptional repressors TCF3, ZFHX1A and ZFHX1B/SIP1.¹⁷ Similarly, Evans et al., demonstrated that knockdown of pVHL resulted in E-cadherin suppression via HIF-dependent induction of E2 box-dependent transcriptional repressors Snail and SIP1¹⁸ and Esteban et al., suggested that HIF-2 may have a more potent suppressive effect on E-cadherin expression than HIF-1.¹⁹ While our studies provide experimental evidence that HIF-1 activation promotes fibrosis by enhancing the transition of epithelial cells towards a mesenchymal phenotype, we did not observe upregulation of the same transcriptional repressors in hypoxic primary renal epithelial cells, which may be a reflection of cell-type and context-dependent differences in gene regulation. However we identified lysysl oxidase genes as downstream targets of HIF, which may functionally interact with transcriptional repressors that regulate epithelial cell plasticity. Although initially identified by their ability to crosslink collagen and elastin fibers, lysyl oxidase (LOX) and lysyl oxidase-like (LOXL) proteins have been shown to carry out intracellular functions as well and display a range of biological activities that extend beyond extracellular cross-linkage. These include the regulation of ras- and NFκB-signaling, possibly through the modification of DNA-histone and histone-histone interactions.^20,21 Recent reports indicated that LOXL2 is capable of promoting EMT through functional interaction with the transcriptional repressor Snail,²² which has been shown to induce renal fibrosis when overexpressed in transgenic mice.²³ HIF-1 induction of LOX and LOXL2 has also been shown to promote migration in human breast and cervical cancer cells, and to regulate E-cadherin expression in renal cell cancer.^24,25 Thus, it is plausible that a HIF-1 mediated increase in lysyl oxidase gene expression promotes fibrosis by enabling transition towards an activated ECM producing fibroblast-like phenotype. In keeping with this notion, we found increased LOXL2 expression in renal biopsy tissues from patients with CKD underscoring the relevance of lysyl oxidases to the pathogenesis of renal fibrosis. Defining the exact contribution of HIF-activation to the regulation of LOX and LOXL2 gene expression in the context of chronic renal disease is currently part of an ongoing effort in our laboratory.

The molecular events underlying EMT are complex and HIF's role in EMT may involve functional interactions with other signaling pathways such as the TGFβ,^26,27 or the Notch pathway.²⁸ For example, TGFβ1 induces the zinc-finger repressors Snail and Slug,²⁹ the two-handed E-box-binding zinc-finger protein SIP-1,³⁰ which is regulated by HIF in renal cancer cells¹⁸ and the bHLH repressor HEY1.³¹ HIF-1α has furthermore been shown to interact with the Notch intracellular domain (NICD) functionally and as a result of increased Notch activity maintains embryonic stem cells in a de-differentiated state.²⁸ Increased Notch activity from enhanced Jagged 1 ligand expression has also been detected in fibrotic kidneys, and therefore may be relevant to the pathogenesis of renal fibrosis.³² Thus, it is conceivable that the hypoxic induction of renal epithelial de-differentiation and initiation of EMT may also involve a HIF-mediated upregulation of certain Notch target genes, including transcriptional repressors HES and HEY.³¹ In support of this notion, we observed a HIF-1 dependent upregulation of Notch target gene HEY1 in hypoxic primary renal epithelial cells.¹⁰ Therefore the potential role of HIF/Notch or HIF/TGFβ functional interactions in renal fibrogenesis warrants further investigation.

Hypoxia Regulation of ECM Turnover and Cross-Talk with TGFβ

Aside from promoting EMT, hypoxia may drive fibrogenesis through a direct transcriptional increase in the expression of collagen genes or gene products that are directly involved in the regulation of ECM turnover. Hypoxia induces collagen I, and decreases matrix-metallopeptidase 2 (MMP-2) in renal epithelial cells,³³ and increases plasminogen activator inhibitor-1 (PAI-1),³⁴tissue-inhibitor of metalloproteinase-1 (TIMP-1)³³ and connective tissue growth factor (CTGF)³⁵ through a HIF mediated transcriptional response. Transcriptional activation of oxygen-sensitive gene expression may also result from a synergistic co-operation between HIF and non-HIF pathways, such as the TGFβ1/SMAD3 signaling pathway. This has been shown for the regulation of vascular growth factor (VEGF),²⁶endoglin³⁶ and erythropoietin (EPO).³⁷ Further support for this notion comes from the observation that hypoxia and TGFβ1 synergize with regard to the production of certain collagens in fibroblasts.^38,39 TGFβ1 is a member of the TGFβ superfamily, which encompasses other multifunctional cytokines such as bone morphogenetic proteins, activins and inhibins. TGFβ1 expression has a strong correlation with tissue fibrosis and is largely responsible for the observed increase in ECM deposition in fibrotic diseases through stimulation of pro-fibrogenic gene expression in a wide variety of cell types. It has been identified as a main contributing factor to the development of renal fibrosis in a variety of renal disease settings, most notably diabetic nephropathy,⁴⁰ and regulates gene expression through the activation of SMAD transcription factors. Although the molecular basis of a functional interaction between hypoxia, HIF and TGFβ signaling is not well understood at this point, the most simplest explanation being that TGFβ1 levels increase in response to hypoxia,^33,41 close proximity of HIF and SMAD binding DNA sequences in regulatory regions of certain hypoxia-sensitive genes would suggest that both pathways could interact at a transcriptional level as suggested for the regulation of VEGF expression.²⁶ Although a direct role for HIF has yet to be demonstrated, hypoxia also increases SMAD3 mRNA levels and promotes the thrombospondin dependent release of latent TGFβ2, thus activating TGFβ signaling,²⁷ suggesting that hypoxia and/or HIF may affect the TGFβ signaling pathway at multiple levels.

HIF Activation and Clinical Outcome of Renal Injury

Work by our laboratory and others suggested that activation of HIF signaling in resident kidney cells promotes the development of renal fibrosis by at least three mechanisms: (a) by direct transcriptional regulation of gene products that control ECM turnover, (b) by functional co-operation with TGFβ1, which is a potent profibrotic factor and (c) by promoting EMT (Fig. 3). Hypoxic activation of HIF-1 therefore exerts a disease promoting effect, which is in contrast to its proposed tissue-protective role in acute ischemic injuries. This raises the possibility that the spectrum of HIF activated stress responses may differ between conditions of acute and chronic hypoxia and implies that HIF-1 must have context- and/or cell type-dependent biological functions with regard to clinical disease outcome. From a teleological perspective one could speculate that a HIF-1 mediated adaptation to chronic hypoxia in highly differentiated renal epithelial cells, which are typically engaged in complex ATP consuming cellular transport functions, may be geared towards promoting a de-differentiated phenotype to conserve energy resources and to reduce oxygen-demand, whereas HIF-mediated responses in an acute hypoxia setting may be mainly geared towards protecting differentiated epithelial cells from cell death.

Figure 3.

Schematic suggesting hypoxia-regulated biological processes and mechanisms through which HIF activation may promote renal fibrogenesis. HIF signaling appears to play a pivotal role in the regulation of inflammatory responses,⁵⁵ however the role of HIF activation with regard to renal injury mediated by inflammatory cells is unclear. Because of its important role in the initiation and maintenance of kidney injury we have included `inflammation' in this schematic.

Activation of HIF before or at the onset of renal injury using HIF-α stabilizing agents, such as cobalt chloride, L-mimosine or other HIF prolyl-hydroxylase inhibitors has been shown to ameliorate acute ischemic renal failure^42,43 and to reduce injury in Thy-1 nephritis and in a remnant kidney model most likely through the induction of cytoprotective factors and by promoting angiogenesis.^44,45 Conversely, genetic ablation of HIF-1α or HIF-2α worsened acute ischemic renal injuries.^43,46 Although it is not exactly clear which cell type conferred protection in these models, HIF has been shown to upregulate the expression of gene products that have tissue-protective functions in hypoxic renal injury. These include VEGF,^47,48 heme oxygenase 1 (HO-1)^49–52 and EPO.⁵³ Interestingly VEGF and also HO-1 were found to be decreased in tubulointerstitial tissues from patients with diabetic nephropathy,^10,54 which may indicate that HIF-mediated gene regulation is modulated by factors that are specific to a certain chronic disease microenvironment. In addition to epithelial HIF-1α stabilization, we found that HIF-2α was detectable in interstitial cells by immunohistochemical staining of fibrotic kidneys. The role of renal HIF-2, which under hypoxic conditions is normally expressed in endothelium, glomerular cells and EPO-producing peritubular fibroblast-like interstitial cells, and other oxygen-sensitive transcription factors such as NFκB in the fibrotic process is unclear at this point and remains to be investigated.

Concluding Remarks

Recent studies provide a molecular basis for the long-standing theory that hypoxia plays an active role in the progression of fibrotic kidney disease and identify epithelial HIF-1 as an oxygen-regulated transcription factor that is capable of promoting fibrogenesis through increased expression of extracellular matrix modifying factors, lysyl oxidase genes and by facilitating EMT. These findings have immediate clinical implications as they strongley encourage therapies that aim at improving tissue oxygenation to halt the progression of fibrosis and raise interesting questions with regard to functional interactions of the HIF pathway with other signaling pathways that are relevant to the regulation of epithelial cell plasticity and ECM turnover. The notion of cell-type and context-dependent HIF biological functions, including those in immune and inflammatory cells, will stimulate further studies into the role of HIF signaling in acute and chronic ischemic injuries.

Abbreviations

EMT

epithelial to mesenchymal transition

HIF

hypoxia-inducible factor

CKD

chronic kidney disease

ECM

extracellular matrix