Abstract
ABSTRACT. Oxygen tensions in the kidney are heterogeneous, and their changes presumably play an important role in renal physiologic and pathophysiologic processes. A family of hypoxia-inducible transcription factors (HIF) have been identified as mediators of transcriptional responses to hypoxia, which include the regulation of erythropoietin, metabolic adaptation, vascular tone, and neoangiogenesis. In vitro, the oxygen-regulated subunits HIF-1α and -2α are expressed in inverse relationship to oxygen tensions in every cell line investigated to date. The characteristics and functional significance of the HIF response in vivo are largely unknown. High-amplification immunohistochemical analyses were used to study the expression of HIF-1α and -2α in kidneys of rats exposed to systemic hypoxia bleeding anemia, functional anemia (0.1% carbon monoxide), renal ischemia, or cobaltous chloride (which is known to mimic hypoxia). These treatments led to marked nuclear accumulation of HIF-1α and -2α in different renal cell populations. HIF-1α was mainly induced in tubular cells, including proximal segments with exposure to anemia/carbon monoxide, in distal segments with cobaltous chloride treatment, and in connecting tubules and collecting ducts with all stimuli. Staining for HIF-1α colocalized with inducible expression of the target genes heme oxygenase-1 and glucose transporter-1. HIF-2α was not expressed in tubular cells but was expressed in endothelial cells of a small subset of glomeruli and in peritubular endothelial cells and fibroblasts. The kidney demonstrates a marked potential for upregulation of HIF, but accumulation of HIF-1α and HIF-2α is selective with respect to cell type, kidney zone, and experimental conditions, with the expression patterns partly matching known oxygen profiles. The expression of HIF-2α in peritubular fibroblasts suggests a role in erythropoietin regulation.
Sufficient oxygenation is a prerequisite for organ function. However, oxygen delivery to organs and tissue oxygen tensions within organs vary considerably. The kidney is characterized by an interesting paradox with respect to its oxygen supply. Although blood flow is high in relation to organ weight and the arteriovenous oxygen difference is small, shunt diffusion of oxygen and heterogeneous utilization lead to marked oxygen gradients (1,2). Oxygen supply to the renal medulla barely exceeds demand, and medullary oxygen tensions are approximately 10 mmHg (3–6). Cortical oxygen tensions are more heterogeneous but are also frequently less than the venous oxygen tensions (4,6–8).
The effects of oxygen on cellular functions of the kidney are poorly understood. High rates of oxygen consumption in the proximal tubule and thick ascending limb, together with limited oxygen supply, are thought to be responsible for the high sensitivity to ischemic injury (2,9,10). A physiologic function directly related to renal oxygen tensions is the production of erythropoietin (EPO) by peritubular cortical fibroblasts (11–13). Regulation of EPO occurs at the mRNA level but, because of the lack of an appropriate in vitro system, the control of EPO gene expression in the kidney remains incompletely characterized. Studies of EPO regulation in hepatoma cells led to the identification of a family of hypoxia-inducible transcription factors (HIF) (14,15). In addition to EPO, the HIF system regulates several target genes that have important functions in renal physiologic and pathophysiologic processes, including energy metabolism [glucose transporters (GLUT) and glycolytic enzymes], vasomotor regulation [nitric oxide synthases, heme oxygenase-1 (HO-1), and endothelins], angiogenic growth (vascular endothelial growth factor and platelet-derived growth factor), matrix metabolism (collagens, collagen prolyl hydroxylases, matrix metalloproteinases, and transforming growth factor-β isoforms), and apoptosis/cell survival decisions (NIP3 and Nix) (14–16). Therefore, it might be predicted that an understanding of the patterns of HIF regulation in the kidney would provide important insights into processes mediated by these molecules.
HIF is a heterodimer composed of an α-subunit and a β-subunit. Although HIF-β is constitutively expressed, its two dimerization partners, HIF-1α and -2α, are rapidly degraded in the presence of oxygen, via the ubiquitin-proteasome system. The von Hippel-Lindau gene product is the recognition component of a multiprotein E3-ubiquitin-ligase complex that captures HIF α-chains that have undergone enzymatic hydroxylation of specific prolyl residues (17,18). The oxygen dependence of HIF prolyl hydroxylation results in HIF α-subunit accumulation during hypoxia. In vitro, virtually every cell line responds to hypoxia with increases in the levels of HIF-1α and -2α, as well as of HIF target genes, suggesting that HIF are ubiquitous effectors of cellular responses to hypoxia (19,20). When HeLa cells in suspension were exposed to progressive decreases in pericellular oxygen levels, continuous increases in HIF protein levels and DNA binding activity were observed, with half-maximal induction at 1.5 to 2% O2 (21). This range corresponds to oxygen tensions of approximately 10 to 15 mmHg, which are well within the range of values observed in the renal medulla under baseline conditions and in other parts of the kidney during hypoxia.
To assess the potential roles of HIF-1α and -2α in the renal response to hypoxia, we studied the expression of both proteins in the kidneys of rats exposed to hypoxia, anemia, functional anemia induced by carbon monoxide (CO), total or subtotal ischemia, or cobaltous chloride (CoCl2) (which is known to mimic hypoxic effects). Our data indicate that both HIF α-subunits are induced in the kidney under these conditions but their expression is restricted to distinct cell populations, with the patterns differing depending on the type of stimulus.
Materials and Methods
Animals
The study was approved by the institutional review board for the care of animal subjects and was performed in accordance with National Institutes of Health guidelines. Male Sprague-Dawley rats (Winkelmann, Borchen, Germany) were used at weights of 200 to 280 g (n = 3 to 5 for each time point and experimental condition).
Induction of Anemia
A femoral artery catheter was inserted under anesthesia, and the hematocrit level was decreased by repetitive blood drawing and substitution of normal saline solution for a period of 45 min, until a value of 0.16 was reached. Animals were allowed to regain consciousness and were euthanized after 3 h.
Exposure to Hypoxia and CO
An air-tight Plexiglas cabinet was used to expose animals to premixed gasses. For the induction of normobaric hypoxia, animals were exposed for 1 or 5 h to 8% O2/92% N2. For the induction of functional anemia, animals were exposed for 0.5, 1, or 5 h to normal air supplemented with 0.1% CO.
Treatment with CoCl2
CoCl2 hexahydrate was dissolved in distilled water and injected subcutaneously twice, at a dose of 30 mg/kg, with a dosing interval of 12 h. Animals were euthanized 6 h after the second injection.
Induction of Renal Ischemia
For induction of total renal ischemia, the left renal artery was clamped for 0.5 or 1 h after a midline laparotomy. In a separate group of animals, a branch of the left renal artery was ligated for 1 d, to induce renal infarction. In sham-operated animals, the left renal artery was dissected from the vein but not clamped. The abdomen was sutured, and the animals were allowed to regain consciousness until euthanasia.
Tissue Preparation
For collection of kidneys, animals received intraperitoneal injections of sodium pentobarbital (Sanofi, Hannover, Germany), at a dose of 0.05 g/kg. In experiments with exposure to low ambient O2 levels or CO, animals were returned to the chamber after injection and the chamber was flushed with the respective gas mixture. After the onset of anesthesia, kidneys were generally perfusion-fixed in situ, as described (22). In brief, a polyethylene tube was inserted into the infrarenal aorta and the inferior vena cava was incised. Perfusion was performed with 330 mosmol/L sucrose in phosphate-buffered saline (PBS) for 10 s at a constant pressure of 240 mmHg, followed by freshly prepared 3% paraformaldehyde in PBS (pH 7.4) at 240 mmHg for 1.5 min and at 100 mmHg for 3.5 min and then by sucrose/PBS to stop fixation. Additional kidneys from animals euthanized by cervical dislocation were fixed by immersion. After fixation, specimens were transferred to 330 mosmol/L sucrose in PBS with 0.02% sodium azide; 1 to 7 d later, specimens were embedded in paraffin.
Immunohistochemical Analyses and Anti-HIF Antibody Characterization
Paraffin sections (4 μm) were dewaxed in xylene, rehydrated in a series of ethanol washes, and placed in distilled water before staining procedures. Slides were coated with 3-aminopropyl-tri-ethoxysylane. For detection of HIF isoforms, monoclonal mouse anti-human HIF-1α antibody (α67; Novus Biologicals, Littleton, CO) and polyclonal rabbit anti-mouse HIF-2α antibodies (PM8 and PM9, obtained from two different rabbits immunized with a peptide containing amino acids 337 to 439 of mouse HIF-2α) were used. Specific staining of each HIF α-isoform was confirmed in immunoblots (20) by using in vitro transcribed and translated mouse HIF-1α and HIF-2α (TnT T7; Promega, Madison, WI) and homogenates of rat endothelial cells (RBE 4; kindly provided by Hugo Marti, Zurich, Switzerland) exposed to hypoxia in vitro (1% oxygen, 4 h) (Figure 1). For immunohistochemical analyses, α67 was used at a dilution of 1:6000 and PM8 and PM9 were used at dilutions of 1:3000. Additional primary antibodies were anti-chicken calbindin D-28K (1:30,000; Sigma Chemical Co., St. Louis, MO) as a marker for connecting tubules (23), polyclonal rabbit anti-rat Tamm-Horsfall protein (1:3000; a gift from John Hoyer, Philadelphia, PA) as a marker for thick ascending limbs (24), polyclonal rabbit anti-mouse thiazide-sensitive cotransporter (1:6000; a gift from David H. Ellison, Denver, CO) as a marker for distal convoluted tubules (25), monoclonal mouse anti-rat CD31 (1:200; Serotec, Oxford, UK) for staining of endothelial cells, polyclonal rabbit anti-mouse HO-1 (1:60,000; Stressgen, Victoria, Canada), and polyclonal rabbit anti-human Glut-1 (1:10,000; Biotrend, Golden, CO). Detection of bound antibodies was performed by using biotinylated secondary anti-mouse or anti-rabbit antibodies and a catalyzed signal amplification system (Dako, Hamburg, Germany) based on the streptavidin-biotin-peroxidase reaction, according to the instructions provided by the manufacturer. Antigen retrieval was performed for 90 s in preheated Dako target retrieval solution, using a pressure cooker. All incubations were performed in a humidified chamber. Between incubations, specimens were washed two to four times in buffer (50 mM Tris-HCl, 300 mM NaCl, 0.1% Tween-20, pH 7.6). Control samples included samples from normoxic or sham-operated animals, samples prepared with the omission of primary antibodies, and samples prepared with the use of preimmune serum from animals immunized against HIF-2α.
Figure 1. Antibody specificity of α67 [monoclonal mouse anti-human hypoxia-inducible transcription factor-1α (HIF-1α)] and PM9 (polyclonal rabbit anti-mouse HIF-2α). Immunoblots of in vitro transcribed and translated mouse HIF-1α (IVTT-1α) and HIF-2α (IVVT-2α) using reticulocyte lysates and homogenates of RBE cells under normoxic (No) and hypoxic (Hy) conditions, are presented. Lysate, unprogrammed reticulocyte lysates, used as negative control samples. Primary antibodies were diluted 1:1000 for immunoblotting.
Ultrastructural Preembedding Histochemical Analyses
With the use of a modified standard protocol (26), 15-μm sections were maintained in 2-ml glass vessels and subjected to the same procedures as the sections mounted on glass slides. Anti-mouse HIF-2α antibodies were diluted 1:1000 and incubation times were increased fourfold, compared with the protocol provided by the manufacturer (Dako). After staining, the sections were treated with 1.5% glutaraldehyde and 1% OsO4, dehydrated with a graded ethanol series, and flat-embedded in Epon 812. Semithin sections were stained with Richardson’s reagent, and additional ultrathin sections were processed for electron microscopy as described (26).
Signal Analysis
Signals were analyzed with a Leica DMRB microscope (Leica, Bensheim, Germany), using differential interference contrast. Photographs were digitally recorded by means of a Visitron system (Visitron, Puchheim, Germany).
Reagents
Unless otherwise indicated, chemicals were obtained from Sigma.
Results
Normoxia
In normoxic animals, immunohistochemical analyses for HIF-1α revealed no significant staining in the renal medulla or the cortex (Figure 2a and data not shown). Immunohistochemical analyses for HIF-2α occasionally demonstrated weak cytoplasmic staining of some tubular cells in the cortex and medulla and of medullary interstitial cells after prolonged reaction with diaminobenzidine. An identical weak staining pattern was observed with preimmune serum, however, and staining was thus considered to be nonspecific.
Figure 2. Expression of HIF-1α in the papilla of rats exposed to carbon monoxide (CO) for 5 h. (a) Normoxic control sample, showing no staining at baseline. (b) Sample showing marked upregulation after CO exposure. (c and d) Higher-magnification views of the areas of the papillary tip and middle papilla outlined in b. Arrowheads, interstitial fibroblasts. Magnifications: ×20 in a and b; 160 inc; and ×220 ind. 3, thin limb; 9, medullary collecting duct.
Normobaric Hypoxia
To activate the HIF system, we first exposed animals to 8% O2 for 1 or 5 h. Irrespective of the mode of kidney fixation (immersion or perfusion), no signals for HIF-1α or -2α could be detected after 1 h. After 5 h, nuclear staining for HIF-1α was observed in cortical tubules, papillary collecting ducts, and papillary interstitial cells when kidneys were immersion-fixed (data not shown). Because of the lower morphologic resolution, compared with perfusion fixation, the identity and precise distribution of the positively staining cells were difficult to define. Because signals were not detectable after perfusion fixation, we hypothesized that the unavoidable reoxygenation period of a few minutes might have been sufficient to allow HIF degradation. This assumption is based on experience with tissue cultures, in which the half-lives of HIF-1α and -2α after reoxygenation were observed to be only a few minutes (20,21).
Anemia and CO Exposure
Comparison of Stimuli.
To overcome the potential loss of signal, animals were phlebotomized or exposed to CO to reduce their oxygen-carrying capacity and tissue oxygenation (27–30). Because of the much higher affinity of hemoglobin for CO, compared with O2, exposure to 0.1% CO results in approximately 50% CO-hemoglobin; this effect slowly ceases when animals are returned to normal air. Acute anemia and CO exposure for 5 h resulted in pronounced and similar staining patterns for both HIF-1α and -2α, but CO exposure was a better tolerated and more reproducible stimulus. The signal intensity was slightly stronger after immersion fixation, but the signal distribution was independent of the mode of fixation. We therefore proceeded with perfusion-fixed kidneys from CO-exposed animals, to identify the cell types expressing HIF (Figures 2 to 6).
Figure 3. Expression of HIF-1α in the cortex of rats exposed to CO for 5 h. (b and c, d and e, and f and g) Pairs of serial sections of a medullary ray (b and c) and the cortical labyrinth (d to g) were stained for HIF-1α (b, d, and f) and markers used for identification of different nephron segments, namely Tamm-Horsfall protein (THP) for identification of the medullary thick ascending limb (c), thiazide-sensitive cotransporter (TSC) for identification of distal convoluted tubules (e), and calbindin D-28K (Cb) for identification of connecting tubules (g). 1, proximal tubule, pars convoluta; 5, cortical thick ascending limb; 6, distal convoluted tubule; 7, connecting tubule; 8, cortical collecting duct; G, glomerulus. Magnification, ×250.
Figure 4. Expression of HIF-1α in the outer medulla of rats exposed to CO for 5 h. (a) Area at the corticomedullary border, showing preferential expression of HIF-1α in perivenous tissue. (b) Section of the inner stripe of the outer medulla, showing staining for HIF-1α in thin limbs, thick ascending limbs, and collecting ducts. (c) Cross-section of the inner stripe of the outer medulla, showing that HIF induction increases with increasing distance from vascular bundles (VB). 3, thin limb; 4, medullary thick ascending limb; 9, medullary collecting duct; A, arcuate artery; V, arcuate vein. Magnifications: ×120 in a, ×250 in b, ×80 in c.
Figure 5. Expression of HIF-2α in kidneys of rats exposed to CO for 5 h. (a) Cortical labyrinth. (b) Outer stripe. (c and d) Inner stripe of the outer medulla. (e) Papilla. Peritubular cells in the cortex stain positive. In the outer medulla, both interstitial cells outside medullary rays (open arrows) and capillary endothelial cells within vascular bundles (black arrows) stain positive. 1, proximal tubule, pars convoluta; 4, medullary thick ascending limb; 8, cortical collecting duct; 9, medullary collecting duct; G, glomerulus. Magnifications: ×100 in a; ×220 in b, c, and d; ×120 in e.
Figure 6. Evidence for expression of HIF-2α in glomerular and peritubular endothelial cells and peritubular fibroblasts. (a and b) Consecutive sections stained for HIF-2α (a) and the endothelial marker CD31 (b). Arrowheads, corresponding regions. HIF-2α-positive cells sometimes appear in clusters, with the number exceeding the possible number of peritubular endothelial cells. Conversely, some peritubular capillaries are clearly distant from HIF-2α-positive cells in the same area. (c and d) Semithin sections showing HIF-2α staining in glomerular endothelial cells (c, arrows), peritubular endothelial cells (d, black arrow), and peritubular fibroblasts (d, open arrow). (e) Transmission electron-microscopic views of the peritubular space, showing staining for HIF-2α in both peritubular endothelial cells (black arrows) and peritubular fibroblasts (open arrow). (f) Higher-magnification view of the area outlined in e, showing collagen bundles as a characteristic of fibroblasts (open arrow). Magnifications: ×250 in a and b; ×400 in c and d; ×4000 in e; ×20,000 in f.
HIF-1α.
Staining for HIF-1α after 5 h of CO exposure was observed in cell nuclei in both the cortex and medulla. The number and intensity of HIF-1α-positive cells varied in different zones of the kidney, with staining being most pronounced in the papilla. Within each zone, expression was restricted to subsets of cells (Table 1 and Figures 2 to 4).
Summary of HIF-1α and -2α expression patternsa
In the cortex, HIF-1α was expressed only in tubular cells, with strongest expression in the S2 segment (Figure 3a). Slightly more than one-half of the S2 cross-sections were positive and, within each tubular cross-section, approximately one-half of the nuclei stained for HIF-1α. In S1 and S3 segments, only single cells stained positive. Distal tubular cells in the cortex were usually negative (Figure 3d), whereas connecting tubules and cortical collecting ducts regularly stained positive (Figure 3, a, b, and f). There was no overlap between staining for HIF-1α and that for Tamm-Horsfall protein (Figure 3, b and c) or the thiazide-sensitive cotransporter (Figure 3, d and e), but overlap was observed with counterstaining for the connecting tubule marker calbindin (Figure 3, f and g).
Within the deep cortex, signals consistently accumulated in the vicinity of arcuate veins (Figure 4a). In the outer stripe of the outer medulla, collecting ducts stained positive, whereas S3 segments of proximal tubules very rarely demonstrated signals and thick ascending limbs were negative. In the inner stripe, in addition to collecting duct cells, thick ascending limb segments and occasionally thin limbs were positive (Figure 4b). Cross-sections revealed that HIF staining increased with increasing distance from vascular bundles (Figure 4c).
In the deeper medulla, collecting ducts remained positive but, in contrast to other areas of the kidney, interstitial cells also expressed HIF-1α, with the proportion of positive cells increasing toward the tip of the papilla (Figure 2, b to d). Thin tube-like structures were also positive. On the basis of their hairpin appearance, some of those structures could be identified as thin limbs of the loop of Henle, but the possibility that some were capillaries, with endothelial cells expressing HIF-1α, cannot be excluded. Strong signals were also observed in papillary surface endothelial cells covering the papilla (Figure 2, b and c).
HIF-2α.
Staining for HIF-2α revealed an entirely different staining pattern, with virtually no overlap (Table 1 and Figure 5). HIF-2α was not detected in epithelial cells of any tubular segment, but some cells were stained in a small subgroup of glomeruli (<10%) (Figure 5a) and peritubular interstitial cells in the cortex and medulla frequently stained positive (Figure 5, b to d). In the inner stripe of the outer medulla, positive cells included some capillary endothelial cells of the vasa recta (Figure 5d). Overall, HIF-2α signal density decreased from the inner medulla toward the papillary tip, with the latter being virtually devoid of staining (Figure 5e). The staining patterns obtained with the two different antisera (PM8 and PM9) were identical.
To identify the glomerular and interstitial cell types staining for HIF-2α, we attempted double-labeling either for HIF-2α and CD31 (as a marker for endothelial cells) or for HIF-2α and 5′-ectonucleotidase (as a marker for peritubular fibroblasts) (22). Unfortunately, the anti-5′-ectonucleotidase antibody did not stain paraffin sections and the specimen heating required for HIF detection resulted in a loss of signal. Three other approaches were therefore pursued, i.e., staining for HIF-2α and the endothelial antigen CD31 on consecutive sections and analysis of semithin sections by light microscopy and of ultrathin sections by electron microscopy after preembedding histochemical analysis. Labeling for HIF-2α and CD31 in the peritubular interstitium was only partially overlapping (Figure 6, a and b). This finding suggested that at least some of the cells expressing HIF-2α were nonendothelial, because they seemed to be located either outside capillaries or in clusters surrounded by smaller numbers of CD31-positive cells. Semithin sections demonstrated that nuclear staining for HIF-2α occurred in glomerular and peritubular endothelial cells and in peritubular interstitial cells that, on the basis of their location and triangular shape, seemed to be peritubular fibroblasts (Figure 6, c and d). This observation was confirmed by electron microscopy (Figure 6, e and f), which demonstrated positive nuclei of endothelial cells bulging into capillary lumina and of fibroblasts located between tubules and capillaries.
Study of animals after 0.5 or 1 h of CO exposure revealed that the HIF-1α signal commenced in the papilla and subsequently reached the collecting ducts and the cortical labyrinth. With increasing stimulation time, both the numbers of cells staining positive for HIF-1α and -2α and the signal intensity increased (data not shown).
CoCl2
The induction patterns for HIF-1α and -2α with CoCl2 stimulation were similar to the patterns with CO stimulation, insofar as HIF-1α was primarily induced in tubular cells and in interstitial cells in the inner medulla (Figure 7), whereas HIF-2α occurred in peritubular and some glomerular cells only (Table 1). However, tubular distribution for HIF-1α in response to CoCl2 differed markedly from that observed in response to CO. Proximal tubular cells were negative, and marked induction of HIF-1α was observed in 70 to 80% of distal tubular cross-sections (Figure 7, a to c). In further contrast to CO exposure, there was no predominance of perivenous upregulation of HIF-1α in the deep cortex (Figure 7d). Moreover, collecting ducts of the lower papilla were negative in animals treated with CoCl2 (Figure 7e).
Figure 7. Expression of HIF-1α in the cortical labyrinth (a to c), the deep cortex (d), and the papillary tip (e) in rats treated with CoCl2. (c) Consecutive section of b, stained for the thiazide-sensitive cotransporter (TSC) for identification of distal convoluted tubules. It should be noted that, in contrast to CO stimulation (Figure 3, d and e), HIF-1α expression is colocalized with thiazide-sensitive cotransporter expression. Staining of the deep cortex (d) shows no preferential expression around arcuate veins, as observed with CO stimulation (compare with Figure 4a). 1, proximal tubule, pars convoluta; 6, distal convoluted tubule; 9, medullary collecting duct; A, arcuate artery; V, arcuate vein. Magnifications: ×250 in a; ×200 in b and c; ×120 in d; ×160 in e.
Renal Ischemia
HIF-1α.
Total renal ischemia also induced HIF-1α and -2α, and the staining patterns for both subunits were similar, although not identical, to those observed with CO. In the cortex, ischemia induced HIF-1α in some cells in a small number of glomeruli (Figure 8b). As with CO stimulation, connecting tubules and collecting ducts, but not proximal tubules, appeared positive (Figure 8, a and c). However, because perfusion was less effective, the morphologic resolution was lower and the possibility that some of the tubular signals were derived from cells other than those of connecting tubules or collecting ducts cannot be excluded with certainty. Total ischemia also induced HIF-1α in papillary tubular and interstitial cells (Figure 8, d to f). Staining in the papilla was not homogeneous. Staining began in a band in the outermost zone beneath the papillary surface (Figure 8d) and proceeded toward more central areas (Figure 8e). One day after the induction of renal infarction via ligation of one branch of the renal artery, marked upregulation of HIF-1α was observed in a band of cells in the direct vicinity of necrotic tissue (Figure 8, g and h).
Figure 8. HIF-1α expression in kidneys rendered ischemic by renal artery occlusion for 30 min (d) or 60 min (a to c, e, and f) or renal artery branch ligation for 24 h (g and h). A few tubular cells (a) and some glomeruli (b) in the cortical labyrinth and collecting ducts in the medullary rays (c) stain positive. In the papilla (d to f), expression can be observed in collecting ducts and interstitial cells, starting in superficial areas at 30 min and progressing toward the center after 1 h. (f) Higher-magnification view of the area outlined in d. In infarcted kidneys, HIF-1α expression can be observed in cells located adjacent to necrotic tissue (arrowheads) (g and h show the same section stained for HIF-1α before and after counterstaining with hematoxylin and eosin). 2, proximal tubule, pars recta; 5, cortical thick ascending limb; 8, cortical collecting duct; 9, medullary collecting duct; G, glomerulus. Magnifications: ×250 in a, c, and f; ×450 in b; ×60 in d and e; ×150 in g and h.
HIF-2α.
The HIF-2α induction pattern observed after renal artery clamping was similar to those observed in response to the other two stimuli investigated. However, staining seemed to occur more frequently in glomeruli and in endothelial cells, including the vasa recta of the outer stripe and periglomerular arterioles (Table 1). Signal distribution and intensity were not different in immersion-fixed kidneys.
Comparison of HIF-1α Staining with Target Gene Expression
To test for the functional significance of HIF induction in tubular cells, consecutive tissue sections were stained for HIF-1α and two different target genes that have the potential to attenuate hypoxic injury and are known to be inducible by HIF in vitro, i.e., HO-1 and GLUT-1 (31,32). HO-1 expression was not observed with normoxia (data not shown) but increased markedly in the proximal tubules of animals exposed to CO; the expression pattern colocalized with that for HIF induction. The cells with the most pronounced nuclear staining for HIF-1α frequently demonstrated the strongest expression of HO-1 (Figure 9, a and b). Expression of GLUT-1 under normoxic conditions was confined to the distal tubules and collecting ducts, as described previously (33) (data not shown). During hypoxia, GLUT-1 expression was also strongly upregulated in proximal tubules expressing HIF-1α (Figure 9, c and d).
Figure 9. HIF-1α expression and staining for heme oxygenase-1 (HO-1) and glucose transporter-1 (GLUT-1) in two pairs of consecutive sections (a/b and c/d) from the renal cortex. Sections were processed such that adjacent surfaces were stained and “mirrored” after digital recording (“back-to-back” processing). The almost complete overlap between tubular cells expressing HIF-1α and HO-1 or GLUT-1 should be noted. GLUT-1 expression in distal tubules not expressing HIF-1 (stars) was constitutive and not different from that in normoxic animals. Magnification, ×150.
Discussion
This work is the first analysis in the kidney and the first comparative study of the cellular expression of the two oxygen-regulated subunits, 1α and 2α, of HIF with different stimuli in vivo. It demonstrates that the system is operative in the kidney and that both subunits are inducible. For specific detection of the two isoforms in immunohistochemical analyses, a potent signal amplification system was used. Nevertheless, no signals were observed in animals maintained with ambient oxygen tensions, despite the well known heterogeneity of renal oxygen tensions (3–6,8). Marked upregulation occurred, however, in nuclei of restricted renal cell populations under all conditions of systemic or local hypoxia investigated, indicating very specific, rather than widespread, activation of HIF, with a considerable amplitude of modulation.
The selective expression of HIF-1α and -2α seems to reflect a combination of intrinsic cellular capabilities and microenvironmental stimulation. Clear discrimination between the two factors is inevitably difficult, but several observations support a dominant role for either intrinsic or extrinsic factors in certain cells and under certain conditions. The most striking observation indicating the involvement of intrinsic determinants was that, independent of the type of stimulus, the cell populations expressing HIF-1α and -2α were consistently different. Whereas HIF-1α was predominantly expressed in tubular cells, HIF-2α expression was largely confined to cells within the peritubular interstitium (Figures 2 to 8). This finding suggests that the vast majority of cells that respond with HIF activation under certain conditions express only one of the two isoforms, indicating an important difference between the in vivo situation and regulation in cell lines (which usually demonstrate hypoxic induction of both isoforms) (20). The expression of HIF-2α in glomerular and peritubular endothelial cells confirms that this isoform plays a predominant role in the hypoxic adaptation of endothelial cells, which was postulated when the protein was identified and termed “endothelial PAS” protein (34). However, in agreement with subsequent in vitro work (20), our findings indicate that HIF-2α is not an endothelium-specific transcription factor. Cumulative evidence from the staining of consecutive sections with an endothelial marker and fine-structural analyses clearly demonstrated additional staining of peritubular fibroblasts (Figure 6).
Another observation indicating a specific cellular predisposition for HIF activation was the predominant expression of HIF-1α in connecting tubules and collecting ducts. Irrespective of known differences in local oxygen tensions in the cortex and medulla, collecting duct cells were positive in all zones of the kidney, independent of the type of stimulation (Figures 2, c and d, 3, a and b, 4b, and 8, c and f). This induction was highly selective, because other nephron segments in the immediate vicinity were usually negative. The mechanism of this preferential expression remains unclear, but upregulation in ischemic kidneys suggests that it does not essentially depend on urine flow and regular transport activity.
Other structures and cell populations demonstrated nuclear accumulation of HIF only in certain locations and only under some conditions investigated, which suggests that local determinants were more important for those responses. In glomeruli, for example, HIF-1α was not induced in response to anemia or CO but was induced after renal artery clamping (Figures 3a and 8b). This finding could indicate that, under conditions in which arterial oxygen tension was normal, local oxygen tensions in glomeruli might not have been sufficiently low for HIF-1α induction. Stimulus-dependent activation of the HIF system was also observed in many tubular cell types. Upregulation of HIF-1α in distal convoluted tubules in response to CoCl2, for example (Figure 7b), is presumably related to local uptake and accumulation (35,36). In support of this assumption, selective apical uptake of cobalt, presumably via a carrier-mediated influx process, was demonstrated in the distal tubule-derived MDCK cell line (37). After exposure to CO, activation in particular tubular segments often seemed to be in keeping with oxygen gradients. A perivenous distribution of positively staining tubules was observed near the corticomedullary junction (Figure 4a), and HIF expression in the inner medulla increased with increasing distance from vascular bundles (Figure 4c). Within the thick ascending limb, positive cells were more commonly observed in the inner stripe, compared with the outer stripe, of the outer medulla (Figure 4b), which corresponds to an established gradient of hypoxia sensitivity (2). Further evidence for local oxygen gradients determining HIF induction was obtained after renal infarction, with staining occurring at the border of tissue necrosis throughout different kidney zones (Figure 8, g and h).
However, consideration of different nephron segments also suggests that the hypoxic thresholds for HIF activation vary widely among different cell types. Within proximal tubules, for example, upregulation of HIF-1α was much more frequent in the convoluted part (S1 and S2 segments) than in the straight part (S3 segment), although, because of local oxygen gradients, straight rather than convoluted proximal tubules are more sensitive to hypoxic injury (9,38,39). Although cobalt strongly induced HIF-1α in distal convoluted tubules, CO stimulation seemed insufficient to induce HIF-1α in these segments, even when adjacent proximal tubules and collecting ducts were positive. Furthermore, the normally hypoxic papillae did not stain positive in unstimulated animals, despite the fact that normal papillary oxygen tension values are probably lower than those achieved in the cortex of stimulated animals. Given the low basal level, papillary oxygen tensions are thought to exhibit little further reduction in response to reduced renal oxygen supply (3–6). Nevertheless, strong upregulation of HIF-1α was inducible in papillary tubular and interstitial cells. Therefore, it seems likely that the cell type determines not only the ability to express one or the other HIF α-isoform but also the level of hypoxia at which the system is activated.
Irrespective of the reasons for selective induction of HIF-1α and -2α in different parts of the kidney, the specific immunohistochemical procedures developed in this study provide powerful tools for elucidation of HIF function in vivo, including differences in HIF-1- versus HIF-2-dependent target gene activation. The most well defined oxygen-dependent function of the kidney is the production of EPO (41), and we previously identified peritubular cortical fibroblasts as the cellular sites of hormone production (12,13). Three of the experimental conditions used in this investigation, i.e., anemia, CO, and CoCl2, strongly induce EPO mRNA in the kidney, whereas renal ischemia is only a weak stimulus for EPO production (42). Parallelism with the expression of HIF-2α in cortical peritubular cells, as observed in this study, indicates a role for HIF-2α in EPO regulation. In contrast, for two other hypoxia-inducible genes (HO-1 and GLUT-1) (31,43), we observed clear colocalization with HIF-1α expression in distinctive tubular cells (Figure 9). This colocalization also provides a strong indication that HIF protein, as demonstrated in this study by means of immunohistochemical analysis, is transcriptionally active and mediates physiologically relevant adaptation.
Although HIF expression was restricted to certain cell populations under each condition investigated, it is remarkable that the majority of renal cell types were observed to express one of the two isoforms under at least one of the conditions (Table 1). This indicates a widespread capability for altered gene expression in the kidney in response to hypoxia. There is increasing consideration of the role that hypoxia might play not only in acute renal failure but also in the progression of chronic renal disease (44). However, establishing a link between altered cell biologic processes and renal tissue oxygen profiles remains difficult. Studies of HIF expression in renal pathologic conditions, in comparison with the patterns established in this study, could prove useful in providing further insight into the role of hypoxia in the progression of renal lesions. Given the fact that HIF is a key regulator of genes that mediate adaptation to hypoxia, it is tempting to speculate that the level of HIF induction not only indicates cellular hypoxia but also modulates the extent of hypoxic injury.
Acknowledgments
Drs. Rosenberger and Mandriota contributed equally to this work. Financial support was provided by the German Research Foundation (Grant EC 87-3) and a Marie Curie individual fellowship to Dr. Mandriota.
- © 2002 American Society of Nephrology
References
Jump to section
More in this TOC Section
Cited By...
- Proximal Tubular Oxidative Metabolism in Acute Kidney Injury and the Transition to CKD
- Hypoxia-Inducible Factor and Oxygen Biology in the Kidney
- Anemia and Incident End-Stage Kidney Disease
- Oxygen battle in the gut: Hypoxia and hypoxia-inducible factors in metabolic and inflammatory responses in the intestine
- Erythropoietin production by the kidney and the liver in response to severe hypoxia evaluated by Western blotting with deglycosylation
- SINHCAF/FAM60A links SIN3A function to the hypoxia response and its levels are predictive of cancer patient survival
- SINHCAF/FAM60A and SIN3A specifically repress HIF-2{alpha} expression
- Hypoxia-inducible factor prolyl hydroxylase inhibitor roxadustat (FG-4592) protects against cisplatin-induced acute kidney injury
- Endothelial Epas1 Deficiency Is Sufficient To Promote Parietal Epithelial Cell Activation and FSGS in Experimental Hypertension
- BTS guideline for oxygen use in adults in healthcare and emergency settings
- Does Renal Repair Recapitulate Kidney Development?
- MicroRNA-489 Induction by Hypoxia-Inducible Factor-1 Protects against Ischemic Kidney Injury
- Erythropoietin Synthesis in Renal Myofibroblasts Is Restored by Activation of Hypoxia Signaling
- MicroRNA-687 Induced by Hypoxia-Inducible Factor-1 Targets Phosphatase and Tensin Homolog in Renal Ischemia-Reperfusion Injury
- Epigenetic Changes Induced by Hypoxia-Inducible Factor: a Long Way Still To Go as a Target for Therapy?
- HIF Hydroxylase Pathways in Cardiovascular Physiology and Medicine
- Sleep-Disordered Breathing in Patients with Polycystic Liver and Kidney Disease Referred for Transcatheter Arterial Embolization
- Chronic Hypoxia-Inducible Transcription Factor-2 Activation Stably Transforms Juxtaglomerular Renin Cells into Fibroblast-Like Cells In Vivo
- Mechanism of Erythropoietin Regulation by Angiotensin II
- Hypoxia-Inducible Factor-1{alpha} Causes Renal Cyst Expansion through Calcium-Activated Chloride Secretion
- Tubular von Hippel-Lindau Knockout Protects against Rhabdomyolysis-Induced AKI
- Deletion of von Hippel-Lindau Protein Converts Renin-Producing Cells into Erythropoietin-Producing Cells
- Magnetic resonance imaging with hyperpolarized [1,4-13C2]fumarate allows detection of early renal acute tubular necrosis
- Orphan Nuclear Receptor Nur77 Promotes Acute Kidney Injury and Renal Epithelial Apoptosis
- Aberrant Tubuloglomerular Feedback and HIF-1{alpha} Confer Resistance to Ischemia after Subtotal Nephrectomy
- Impaired Wound Healing in Hypoxic Renal Tubular Cells: Roles of Hypoxia-Inducible Factor-1 and Glycogen Synthase Kinase 3{beta}/{beta}-Catenin Signaling
- Distinct deregulation of the hypoxia inducible factor by PHD2 mutants identified in germline DNA of patients with polycythemia
- Hypoxia-Inducible Transcription Factors Stabilization in the Thick Ascending Limb Protects against Ischemic Acute Kidney Injury
- Sulfatide-Reactive Natural Killer T Cells Abrogate Ischemia-Reperfusion Injury
- Hepatic HIF-2 regulates erythropoietic responses to hypoxia in renal anemia
- Chronic Kidney Disease: An Inherent Risk Factor for Acute Kidney Injury?
- MicroRNA-519c Suppresses Hypoxia-Inducible Factor-1{alpha} Expression and Tumor Angiogenesis
- Oxygen Tension Regulates Pancreatic {beta}-Cell Differentiation Through Hypoxia-Inducible Factor 1{alpha}
- Donor treatment with a PHD-inhibitor activating HIFs prevents graft injury and prolongs survival in an allogenic kidney transplant model
- HIF in Kidney Disease and Development
- Mutual Regulation of Hypoxia-Inducible Factor and Mammalian Target of Rapamycin as a Function of Oxygen Availability
- Induction of Pyruvate Dehydrogenase Kinase-3 by Hypoxia-inducible Factor-1 Promotes Metabolic Switch and Drug Resistance
- BTS guideline for emergency oxygen use in adult patients
- Imaging of the hypoxia-inducible factor pathway: insights into oxygen sensing
- HIF Activation Protects From Acute Kidney Injury
- Inhibition of Hypoxia Inducible Factor Hydroxylases Protects Against Renal Ischemia-Reperfusion Injury
- Hypoxia, via stabilization of the hypoxia-inducible factor HIF-1{alpha}, is a direct and sufficient stimulus for brain-type natriuretic peptide induction
- Is congenital secondary erythrocytosis/polycythemia caused by activating mutations within the HIF-2{alpha} iron-responsive element?
- The Ternary Complex Factor Net Is Downregulated by Hypoxia and Regulates Hypoxia-Responsive Genes
- Protective Role of Hypoxia-Inducible Factor-2{alpha} against Ischemic Damage and Oxidative Stress in the Kidney
- Acute postnatal ablation of Hif-2{alpha} results in anemia
- New agents that stimulate erythropoiesis
- Immunohistochemical Detection of Hypoxia-Inducible Factor-1{alpha} in Human Renal Allograft Biopsies
- The Heat Is On: An Expanding Role for Hypoxia-Inducible Factors in Kidney Transplantation
- Preconditional Activation of Hypoxia-Inducible Factors Ameliorates Ischemic Acute Renal Failure
- Cytoprotective Effects of Hypoxia against Cisplatin-Induced Tubular Cell Apoptosis: Involvement of Mitochondrial Inhibition and p53 Suppression
- Renal Cyst Development in Mice with Conditional Inactivation of the von Hippel-Lindau Tumor Suppressor
- Chronic Hypoxia and Tubulointerstitial Injury: A Final Common Pathway to End-Stage Renal Failure
- Hypoxia Increases Group IIA Phospholipase A2 Expression under Inflammatory Conditions in Rat Renal Mesangial Cells
- Differential Gene Up-Regulation by Hypoxia-Inducible Factor-1{alpha} and Hypoxia-Inducible Factor-2{alpha} in HEK293T Cells
- HIF-2{alpha} regulates murine hematopoietic development in an erythropoietin-dependent manner
- Hypoxic stress tolerance of the blind subterranean mole rat: Expression of erythropoietin and hypoxia-inducible factor 1{alpha}
- Evidence of Tubular Hypoxia in the Early Phase in the Remnant Kidney Model
- Differential Roles of Hypoxia-Inducible Factor 1{alpha} (HIF-1{alpha}) and HIF-2{alpha} in Hypoxic Gene Regulation
- HIF-1: An Oxygen Response System with Special Relevance to the Kidney
- Erythropoietin and the Erythropoietin Receptor Are Expressed by Papillary Thyroid Carcinoma fromChildren and Adolescents.: Expression of Erythropoietin Receptor Might Be a Favorable Prognostic Indicator
- Induction of Renoprotective Gene Expression by Cobalt Ameliorates Ischemic Injury of the Kidney in Rats
- Podocyte Expression of Hypoxia-Inducible Factor (HIF)-1 and HIF-2 during Glomerular Development
- The Breathing Kidney