HIF-1 expression follows microvascular loss in advanced murine adriamycin nephrosis
http://www.100md.com
《美国生理学杂志》
The University of Sydney at Westmead Millennium Institute, Sydney, New South Wales, Australia
Institute of Veterinary Physiology, Vetsuisse Faculty, University of Zurich, Zurich, Switzerland
ABSTRACT
Cellular hypoxia has been proposed as a major factor in the pathogenesis of chronic renal injury, yet to date there has been no direct evidence to support its importance. Therefore, we examined cortical hypoxia in an animal model of chronic renal injury (murine adriamycin nephrosis; AN) by assessing nuclear localization of the oxygen-dependent -subunit of hypoxia-inducible factor-1 (HIF-1) in animals 7, 14, and 28 days after adriamycin. Results were assessed in conjunction with quantitation of the cortical microvasculature (by CD34 immunostaining) and cortical expression of VEGF. Cortical apoptosis was also examined by terminal deoxynucleotidyl transferase dUTP nick-end labeling staining. A dramatic and significant increase in nuclear localization of HIF-1 was seen 28 days after adriamycin in the context of severe glomerular and tubulointerstitial damage. Areas of nuclear HIF-1 staining did not colocalize with areas of cellular apoptosis. AN was also associated with a significant attenuation of the peritubular capillaries that was significant at 14 and 28 days after adriamycin. Cortical VEGF expression fell in a stepwise manner from day 7 until day 28 after adriamycin. In conclusion, these data are consistent with a significant increase in cellular hypoxia occurring in the advanced stages of murine AN. Increased cortical hypoxia was preceded by significant reductions in both the number of peritubular capillaries (i.e., oxygen supply) and the angiogenic cytokine VEGF. Apart from providing the first direct evidence for cellular hypoxia in a model of chronic renal disease, these results suggest that a primary dysregulation of angiogenesis may be the cause of increased hypoxia in this model.
capillary; chronic; tubular
CELLULAR HYPOXIA IN VITRO has been shown to have effects potentially relevant to the pathogenesis of chronic renal injury. Hypoxia has profibrotic effects in vitro, invoking the transcription of matrix proteins (28, 29, 35), profibrotic growth factors (15, 28, 35), and endothelin-1 (ET-1) (23, 34). In addition, distinct hypoxia-dependent pathways may exist for the induction of apoptosis (6, 43). Since renal fibrosis and tubular atrophy are inevitable components of the tubulointerstitial disease that accompanies all progressive renal diseases, it has been suggested that cellular responses to hypoxia may represent a "final common pathway" for the development of these pathological changes (9, 10).
To date, supportive evidence for this theory has been indirect. Progressive renal diseases are associated with both a progressive loss of vascular structures (4, 21, 22, 26, 30) and an excess of vasoconstrictor substances such as ANG II and ET-1 (33, 52), both of which may decrease oxygen supply to the renal parenchyma. In conjunction with these changes, there are alterations in renal oxygen demand due to the processes of hypertrophy, atrophy, and loss of tissue components, all of which may be present in variable degrees. The net effect in terms of the timing and severity of tubular hypoxia, therefore, remains to be elucidated.
We previously reported a progressive reduction in binding of the hypoxia marker EF5 in murine adriamycin nephrosis, a model of human chronic renal disease (20a). Although this result could be interpreted as indicating less hypoxia in this model, we were unable to correct for alterations in delivery or binding of the EF5 marker occurring as a consequence of renal injury. To further assess hypoxia in this model, we examined a potentially more reliable marker of cellular hypoxia [nuclear localization of the -subunit of hypoxia-inducible factor-1 (HIF-1)]. HIF-1 is the prototypical factor involved in transcriptional responses to cellular hypoxia (46). While the -subunit of HIF-1 is hypoxia independent, the -subunit is stabilized in an oxygen-dependent manner (reviewed in Ref. 18). Stabilization of HIF-1 allows it to bind to the -subunit and translocate to the nucleus, where it induces the transcription of multiple genes involved in homeostasis and disease pathogenesis. Detection of nuclear HIF-1 is therefore recognized as a marker of cellular hypoxia (20, 44, 50) that reflects local oxygen supply/demand ratios within the tissue.
Nuclear HIF-1 localization was examined at different time points in the course of murine adriamycin nephrosis by quantitative immunohistochemistry and Western blotting. The degree of nuclear HIF-1 accumulation was interpreted in the context of changes to the cortical microcirculation and measurement of the angiogenic growth factor VEGF.
MATERIALS AND METHODS
Murine Adriamycin Nephrosis
BALB/c mice were cared for in the Department of Animal Care, Westmead Hospital, under the ethical guidelines outlined in the Code of Practice for the Care and Use of Animals for scientific purposes established by the Australian National Health and Medical Research Council (NHMRC). Six-week-old male inbred BALB/c mice, weighing 22–25 g, were kept under standard conditions with unrestricted access to food and water in a 12:12-h light-dark cycle. Murine adriamycin nephrosis (AN) (47) was established by a single injection of adriamycin (doxorubicin hydrochloride, Pharmacia and Upjohn, Perth, Australia) at a dose of 11 μg/g into the tail vein of 6-wk-old male BALB/c mice. Animals (6–8/time point) were killed at 7, 14, and 28 days after adriamycin injection. At death, samples of serum and bladder urine were collected. The kidneys were rapidly removed and transferred to a metal plate cooled on ice for sectioning. The kidney poles were snap frozen in liquid nitrogen and stored at –70°C. The remaining kidney was sectioned coronally. One-half of the kidneys were placed in cryomoulds, embedded in optimum cooling temperature compound (Tissue-Tek, Sakura Finetechnicals, Torrance, CA) by immersion in liquid nitrogen-cooled isopentane, and stored at –70°C. The other half was immersed in 10% neutral-buffered formalin for 24 h, dehydrated in graded alcohols, and embedded in paraffin. Creatinine was measured in serum and urine samples by the Jaffé reaction using a Hitachi 747 multianalyzer (Tokyo, Japan). The same machine was used to measure serum albumin by the bromocresyl green method and urine protein using the benzethonium chloride turbidimetric method.
Comparative Histology
Quantitative histological measurements were obtained from the outer cortex using a modification of previously published methods (40, 47). Five-micrometer paraffin-embedded sections were stained with periodic acid-Schiff (PAS). Thirty random cortical images were taken using an SV-microdigital camera (Sound Vision, Framingham, MA) fitted to an Axioskop microscope (Zeiss, Oberkochen, Germany). Blinded analysis was then made using image-analysis software (Optimas version 6.5, Media Cybernetics, Seattle, WA).
Twenty outer cortical glomeruli sectioned through the hilum were analyzed. The outline of the glomerular tuft was traced with a mouse for an estimate of tuft volume. As an index of glomerulosclerosis (GS%), a 24-bit color threshold was used to detect PAS-positive material within the glomerular tuft, and the proportion of PAS staining within the tuft was expressed as a percentage.
The volume fraction of the interstitium was estimated in 10 nonoverlapping fields using an 8 x 11 counting grid. The proportion of interstitial grid intersections was calculated and expressed as a percentage.
Tubular measurements were obtained from 50 randomly selected cortical tubules/animal. Using line morphometry, the tubular diameter was estimated as the shortest axis through the center of the imaged tubule. Tubular cell height was assessed using line morphometry by measuring the radial distance between the basement membrane and the lumen of a line drawn through the center of the largest nucleus visible in the tubular cross section of interest.
Immunostaining for HIF-1
A well-characterized chicken polyclonal anti-HIF-1 antibody (5) was used as a primary antibody. Five-micrometer cryosections were fixed in 4% paraformaldehyde in PBS (pH 7.4) for 10 min and incubated with a 1:50 concentration of primary antibody overnight at 4°C. A peroxidase-conjugated rabbit anti-chicken IgY antibody (1:100, Pierce) was used as a secondary antibody and peroxidase-conjugated goat anti-rabbit IgG (Dako, Carpinteria, CA) as a tertiary antibody. Antibody binding was localized using nickel-enhanced diaminobenzidine (Pierce). Kidney and liver tissue from mice exposed to in vivo hypoxia (10% O2, 6 h) was used as a positive control.
Double immunostaining was performed to look for evidence of HIF-1 nuclear staining within interstitial inflammatory cells. HIF-1 staining was performed as above with the substitution of a Texas red-conjugated goat anti-rabbit antibody (Molecular Probes) as the tertiary antibody. Following HIF-1 staining, sections were incubated with rat monoclonal antibodies against murine macrophages (clone F4/80, 1:100 dilution, BD Biosciences, San Diego, CA), murine CD4 cells (clone RM4–5, 1:100, BD Biosciences), and murine CD8 cells (clone 53–6.7, 1:100, BD Biosciences) for 1 h at room temperature. A fluorescein-conjugated goat Fab' fragment to rat immunoglobulins (1:50, Serotec, Oxford, UK) was used as a secondary fluorescent reagent for the inflammatory cells.
Cortical nuclear staining of HIF-1 was analyzed semiquantitatively using image-analysis software (Optimas version 6.5). Ten random cortical images (each measuring 257 x 214 μm) were taken from each animal. Using the software, positively stained nuclei within the positive control tissue (hypoxic mouse liver) were used to set a 24-bit color "threshold" that recognized positively stained nuclei within the positive control image. When a suitable threshold was selected, it was then applied to the experimental slides to enable an unbiased comparison of the number of positive nuclei within the samples.
Western Blotting for HIF-1
Nuclear proteins were extracted from snap-frozen samples of renal cortex as previously described (40). Aliquots (25 μg) of nuclear proteins were resolved on an 8% SDS-polyacrylamide gel and transferred to a nitrocellulose membrane. The chicken IgY anti-HIF-1 (1:100) was used as a primary antibody and peroxidase-conjugated rabbit anti-chicken IgY antibody (1:1,000, Pierce) as the secondary antibody. Sites of antibody binding were revealed by immersion in an enhanced chemiluminescent substrate (ECL, Amersham) and exposure to autoradiographic film (Amersham). Comparative densitometry was performed using a Molecular Dynamics personal densitometer SI (Amersham) and ImageQuant image-analysis software (Amersham).
Colocalization of HIF-1 and Terminal Deoxynucleotidyl Transferase dUTP Nick-End Labeling Staining
Apoptosis within tissue samples was examined by terminal deoxynucleotidyltransferase dUTP nick-end labeling (TUNEL) staining. HIF-1 staining was performed as above using a Texas red-conjugated anti-rabbit antibody (Molecular Probes) as the tertiary antibody. TUNEL staining was then performed using a commercially available kit (in situ cell death detection kit, Roche Diagnostics, Mannheim, Germany). Fluorescent microscopy was performed using a Leica DM-LB fluorescent microscope (Wetzlar, Germany) fitted with a digital camera (Spot RT, Diagnostic Instruments, Sterling Heights, MI).
Assessment of Microvasculature
Immunostaining for murine CD34. The distribution of the endothelial cell marker CD34 (7, 12, 39) within the cortex was examined by immunohistochemistry and image analysis. Five-micrometer paraffin-embedded sections were dewaxed. Antigen retrieval was performed for 20 min at 95–99°C using a commercial target retrieval buffer (Dako). A rat monoclonal antibody to mouse CD34 (1:400, MEC 14.7, Cedarlane Laboratories) was used for 1 h at room temperature. Biotin-conjugated rabbit anti-rat immunoglobulin (Dako) was used as a secondary antibody. Sites of secondary antibody binding were demonstrated using a streptavidin-peroxidase conjugate (Dako) with diaminobenzidine as the enzyme substrate.
Peritubular capillary density was estimated using an adaptation of previously published methods (21). Ten random outer cortical fields (measuring 257 x 214 μm) were obtained from each animal. The capillary rarefaction index (CRI) was then measured using image-analysis software (Optimas version 6.5). Using a 10 x 12 counting grid, the proportion of grid squares that did not contain CD34-staining capillaries was calculated and expressed as a percentage.
Western blotting for CD34. Aliquots (50 μg) of cytoplasmic proteins were resolved on an 8% SDS-polyacrylamide gel and transferred to a nitrocellulose membranes. The MEC14.7 antibody was used at a concentration of 1:2,000, and biotinylated rabbit anti-rat IgG (1:2,000, Dako) was used as the secondary antibody. Sites of antibody binding were revealed by immersion in streptavidin-HRP (1:3,000, Dako) before chemiluminescent development. Comparative densitometry was performed as described above.
Assessment of Cortical VEGF
Immunostaining for VEGF. Four-micrometer paraffin-embedded sections were incubated with a 1:50 dilution of goat polyclonal antibody against VEGF (P20, Santa Cruz Biotechnology, Santa Cruz, CA). Preliminary staining of acetone-fixed cryosections showed VEGF staining in the glomerular mesangium, cortical tubules, and vascular smooth muscle. The degree of glomerular VEGF staining was less prominent in paraffin-embedded sections but was restored if antigen retrieval was performed. Antigen retrieval was not performed for routine staining due to the minimal contribution of glomerular VEGF to the total.
Peroxidase-conjugated rabbit anti-goat antibody (1:200, Dako) was used as the secondary antibody and peroxidase-conjugated goat anti-rabbit antibody (1:100, Dako) as the tertiary antibody. Cortical VEGF staining was analyzed semiquantitatively using image-analysis software (Optimas version 6.5) using an adaptation of published methods (21). Ten random images (measuring 257 x 214 μm) were obtained from the outer cortex of each animal. Using the software, a 24-bit color threshold was used to identify positively stained areas in each image. The area of each image represented by positively stained areas was then calculated by the software and expressed as a percentage.
Western blotting for VEGF. Aliquots (50 μg) of cytoplasmic proteins were resolved on a 12% SDS-polyacrylamide gel and transferred to nitrocellulose membranes. Staining was performed using a 1:100 dilution of the primary antibody and peroxidase-conjugated rabbit anti-goat Ig (Dako, 1:2,000) as the secondary antibody. Comparative densitometry was performed as described above.
Statistical Comparisons
Statistical comparisons between animal groups were performed using a computer-based statistical package (SPSS version 8, SPSS, Chicago, IL). Statistical significance between groups was tested using one-way ANOVA and the least squares method of post hoc analysis. A P value of 0.05 was considered to represent statistical significance.
RESULTS
Functional and Pathological Effects of Murine AN
As previously described (47), these animals developed progressive renal disease characterized by proteinuria, hypoalbuminemia, and impaired renal function (Table 1). Three mice died after adriamycin treatment (2 in the 28-day group and 1 in the 14-day group). Measured values for serum creatinine at 7 days after adriamycin (5.8 ± 5.5 μM) were outside the linear range of the analyzer used (stated as 9–2,000 μM) and were excluded from analysis. Serum creatinine levels were statistically greater than control at 28 days after adriamycin. AN animals developed proteinuria, as evidenced by a stepwise increase in the urine protein/creatinine ratio that was significantly greater than control after 7 days.
View this table:
Pathologically, this model was associated with progressive glomerular, tubular, and interstitial injury, all of which were severe at 28 days after adriamycin (Table 1). Glomerular changes observed included progressive glomerular enlargement and an increase in glomerulosclerosis. Progressive tubular atrophy was reflected by loss of the luminal brush border and a stepwise reduction in tubular cell height. The interstitium was expanded with a mononuclear cell infiltrate and increased PAS-positive matrix.
Immunostaining and Western Blotting for HIF-
Strong nuclear staining of HIF-1 was seen in the kidney and liver tissue from animals exposed to 10% oxygen (Fig. 1A). In the normal kidney, there was faint staining in cortical nuclei, predominantly in distal tubules (Fig. 1B). Glomeruli demonstrated no nuclear staining, and proximal tubules showed faint nuclear staining only. No nuclear staining was seen in the medulla or papilla.
The pattern of nuclear staining of HIF-1 was significantly altered in AN. Although HIF-1 staining was not obviously different from control at days 7 and 14 after adriamycin, a dramatic increase in nuclear HIF-1 staining was seen in the renal cortex at 28 days after adriamycin (Fig. 1C). Nuclear HIF-1 staining was most apparent in areas of extensive tubular dilatation and interstitial inflammation. In these areas, nuclear HIF-1 staining was seen in both tubular and interstitial cells. Double fluorescent staining for HIF-1 and inflammatory cell-surface markers showed nuclear HIF-1 staining within macrophages, CD4 lymphocytes, and CD8 lymphocytes in cortical samples from animals obtained 28 days after adriamycin (Fig. 2).
Semiquantitative analysis of the number of positive nuclei per unit of cortical area reflected these observations (Fig. 1D). There was a dramatic increase in the number of positively stained nuclei at the 28-day time point (191 ± 115 vs. 9 ± 17 nuclei/unit area, P < 0.01). An identical pattern was seen when nuclear protein extracts were probed for HIF-1 (Fig. 1E).
Colocalization of HIF-1 and TUNEL Staining
TUNEL staining revealed apoptotic nuclei within the middle epithelial layers of the positive control tissue (murine esophagus) consistent with the normal process of keratinocyte maturation (Fig. 3A). Double staining revealed occasional apoptotic nuclei within the epithelial layers of the esophagus that also stained positive for HIF-1, although the majority of apoptotic nuclei were negative (Fig. 3A). No apoptosis was visible in the normal kidney (not shown). Apoptotic nuclei were detectable after adriamycin. There were scanty apoptotic nuclei at day 7 after adriamycin, with apoptotic nuclei at days 14 and 28 after adriamycin largely confined to occasional epithelial cells lining atrophic tubules and sloughed nuclei present within the tubular lumen (Fig. 3, B and C). Double staining of the AN samples did not demonstrate areas of overlap of the HIF-1 or apoptotic stains (Fig. 3, B and C).
Assessment of Cortical Microvasculature
CD34 immunostaining of the normal renal cortex revealed a dense capillary network that in some areas appeared to completely encircle individual tubules (Fig. 4A). The glomerular capillary network also exhibited positive staining. A dense intertubular capillary network was also noted in the medulla and papilla (not shown). AN was associated with significant changes in the density of the cortical microvasculature. There was a significant reduction in visible capillaries from day 14 that was most prominent 28 days after adriamycin (Fig. 4B). The reduction in capillaries was particularly prominent in areas of interstitial expansion and tubular atrophy. In keeping with these observations, there was a significant increase in the capillary rarefaction index after adriamycin (Fig. 4C). Microvascular density was reduced threefold at 14 days after adriamycin (P < 0.01) and fivefold at 28 days (P < 0.01). Western blotting of protein extracts (Fig. 4D) showed a similar stepwise reduction in cortical CD34 expression that was significant at 14 and 28 days after adriamycin.
Immunostaining and Western Blotting for VEGF
Extensive tubular staining for VEGF was seen in the normal kidney, with accentuation of distal tubules (Fig. 5A). There was a reduction in tubular VEGF staining following adriamycin, particularly in areas of tubular damage or atrophy (Fig. 5B). Semiquantitative analysis of cortical VEGF expression reflected these changes. There was a significant reduction in cortical VEGF detectable at 7 days after adriamycin (Fig. 5C), with further stepwise reductions at later time points. Western blotting of cortical proteins also showed a similar decremental pattern of VEGF expression (Fig. 5D).
DISCUSSION
The results presented in this paper describe in detail the time course of cortical HIF-1 expression in murine AN, a robust model of human focal glomerulosclerosis. Increased expression and nuclear localization of HIF-1 were seen in the late stages of this model and were localized to tubular and interstitial cells in areas of extensive parenchymal injury. This finding represents the first direct evidence of hypoxia in an animal model of chronic renal disease. Increased HIF-1 expression in these experiments followed significant reductions in both peritubular capillary density (a surrogate measure of oxygen supply) and reduced cortical expression of the angiogenic growth factor VEGF. Areas of cellular hypoxia did not colocalize with areas of tubular apoptosis in this model.
Two separate methods were used to compare nuclear localization of HIF-1 in the current experiments (immunostaining and Western blotting). Consistent with previous reports, we found minimal nuclear staining of HIF-1 in the normal kidney (44, 45, 51). Cortical nuclear HIF-1 was significantly increased by whole animal hypoxia as well as in the later stages of AN. This pattern of nuclear localization is consistent with increased stabilization of HIF-1 in the advanced stages of this model. The cellular fate of HIF-1 is determined by the activity of a redox-sensitive pathway (reviewed in Ref. 18). In the presence of oxygen, the -subunit undergoes prolyl hydroxylation and is targeted for proteasomic degradation. In the absence of oxygen, this hydroxylation does not occur, and after translocation to the nucleus, HIF-1 is free to bind to the -subunit (the aryl hydrocarbon nuclear receptor translocator) to form the HIF-1 complex that, in turn, interacts with hypoxia-responsive elements in the promoter region of multiple genes. Other than hypoxia, a number of other stimuli may potentially affect the stabilization of HIF-1 in this model, including the proinflammatory molecules TNF-, IL-1 (2, 17), and nitric oxide (NO) (19, 37). It is noteworthy, however, that the effects of TNF-, IL-1, and NO on HIF-1 stabilization are based on in vitro experiments. A consistent difficulty with the interpretation of these experimental systems is that these molecules may well exert their in vitro effects by influencing oxygen consumption, weakening the case for a direct effect of these stimuli on HIF-1 stabilization. Furthermore, extrapolation of the results obtained with the use of NO donors in vitro is made more complex by the fact that the production of NO in vivo is an oxygen-dependent process (16), and exogenous NO donors may themselves disrupt the redox-dependent process of HIF-1 stabilization in a way that is not relevant to NO produced in vivo (19, 37). The effects of endogenous TNF-, IL-1, and NO on HIF-1 stabilization in the current experiments are therefore uncertain.
The presence of increased nuclear HIF-1 in the later stages of this model is strongly suggestive of a significant increase in cellular hypoxia in advanced disease. Before these data were gathered, experimental evidence to support the presence of "chronic renal hypoxia" has been based largely on indirect findings, including the observed reduction in peritubular capillaries (4, 21, 22, 26, 30) and extrapolation of the profibrotic effects of hypoxia in vitro (28, 29, 35). The principal advantage of a HIF-1-based method of hypoxia detection is that it gives direct evidence of transcriptionally relevant hypoxia at a cellular level. The HIF-1 pathway is activated in vivo when oxygen demand outweighs supply (18). This delicate balance cannot be inferred from microelectrode measurements that have previously been regarded as the "gold standard" for direct hypoxia measurement.
The finding of increased HIF-1 stabilization in AN warrants a consideration of whether HIF-1-dependent stimuli may contribute to the pathogenesis of renal injury in this model. Although hypoxia in vitro induces a wide range of profibrotic effects, including the stimulation of profibrotic growth factors (15, 28, 35), matrix proteins (28, 29, 35), and matrix stabilizers, it was notable in the current experiments that nuclear HIF-1 was increased only in the later stages of this model, at a time when interstitial fibrosis was already present. This temporal association makes it unlikely that hypoxia contributes significantly to the development of interstitial fibrosis in this model. However, hypoxia could contribute to pathological changes occurring in the late stages of AN. Our finding that apoptosis was more extensive in the later stages of this model in conjunction with the possible existence of HIF-1-dependent pathways for apoptosis (6) prompted us to examine for evidence of colocalization of HIF-1 and a nuclear marker of apoptosis (TUNEL staining). Using a previously described fluorescent technique (49), we found evidence of colocalization of HIF-1 and TUNEL staining in nuclei of the esophageal epithelium, a tissue in which both cellular hypoxia (24, 38) and apoptosis (48) occur under normoxic conditions. In the kidney sections, apoptotic nuclei were most prominent in the later stages of AN, with TUNEL labeling of occasional tubular epithelial cells and intraluminal cellular casts. Although HIF-1 staining was detectable in the same sections, there was no colocalization of HIF-1 or apoptosis in the kidney samples. This result suggests that HIF-1-dependent pathways are unlikely to contribute significantly to tubular apoptosis in this model but may instead represent an attempted homeostatic response to relative oxygen deficiency within the tubular cells in advanced AN.
To investigate the potential causes of hypoxia in this model, we measured peritubular capillary density as a surrogate measure of oxygen supply. AN was associated with a significant reduction in CD34-labeled peritubular capillaries, consistent with similar findings in other chronic renal diseases (4, 21, 22, 26, 30). This reduction in capillary density preceded the detection of increased cortical hypoxia and would suggest that a reduction in oxygen supply plays a role in the development of hypoxia. Consistent with this hypothesis was the finding that the areas of greatest capillary rarefaction at day 28 after adriamycin were those with the greatest degree of tubulointerstitial injury and correlated morphologically with the areas within which HIF-1 nuclear staining was seen. Two methods (CD34 immunostaining and Western blotting) were used to assess the cortical microvasculature in this model. Western blotting was performed because the method used for calculation of the capillary rarefaction index assumes that the volume of tissue assessed in the normal kidney is unchanged in AN. This model is associated with tubular dilatation, tubular loss, and interstitial expansion, all of which affect the volume of tissue to a variable amount. The reduction of CD34 protein in AN by Western blotting (which is not subject to the same limitations) was further evidence that the cortical microvasculature is reduced in this model.
A further result from these experiments is that cortical expression of the angiogenic cytokine VEGF was significantly reduced as the disease progressed, consistent with results obtained in other models of chronic renal injury (8, 21, 31). This result may seem contradictory to the finding of increased hypoxia in this model, particularly since VEGF is both induced by hypoxia and transcriptionally activated by HIF-1 (11, 25, 27). There are several explanations for this apparent disparity. First, HIF-1 is not the only known transcriptional activator of VEGF, with experimental evidence suggesting a role for SP-1 (1, 41) as well as the oncogene ras (3, 32). VEGF transcription is also inhibited by the anti-oncogenes p53 (36) and p73 (42). Also relevant are previous experiments showing that VEGF expression is reduced in cutaneous models of inflammation at a time when HIF-1 expression is increased (2) and the finding that VEGF production in response to hypoxia was reduced in the presence of macrophage-derived cytokines in vitro (21). These latter findings are potentially relevant to the reduction of VEGF observed in the current experiments, since AN is associated with a prominent mononuclear infiltrate (47). It is therefore likely that the finding of reduced VEGF in AN (as well as in other renal disease models) is multifactorial in origin.
Whatever the cause, the finding that VEGF is reduced in AN has potential relevance to the increased HIF-1 expression that occurs in the late stages of this model. If cellular hypoxia is caused by a reduction in oxygen supply engendered by microvascular loss in this model, it is also possible, in turn, that the reduction in peritubular capillaries may in fact be secondary to reduced local production of VEGF, since VEGF is an important survival factor for endothelial cells (13, 14). In support of this, the reductions in cortical VEGF seen in AN were observed before a detectable reduction in peritubular capillaries, which, in turn, preceded a detectable increase in cellular hypoxia.
In summary, we showed a significant increase in nuclear HIF-1 localization in murine AN, consistent with the presence of increased cellular hypoxia in the later stages of this model. HIF-1-dependent mechanisms do not appear to be the predominant cause of apoptosis in this model, and the timing of the increase in nuclear HIF-1 would argue against a significant role for hypoxia in either parenchymal injury or interstitial fibrosis in this model. The observed reductions in both the density of peritubular capillaries and expression of the principal angiogenic growth factor VEGF suggest that the increase in nuclear HIF-1 in AN is secondary to a disruption of mechanisms involved in the maintenance of the normally extensive capillary network within the kidney.
GRANTS
This work was supported by grants from the National Health and Medical Research Council of Australia and the Swiss National Science Foundation.
ACKNOWLEDGMENTS
In vivo hypoxia experiments to provide positive controls for HIF-1 staining were performed in collaboration with Dr. Scott Geller and Prof. Jonathon Stone from the Department of Retinal Biology, the University of Sydney.
FOOTNOTES
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
REFERENCES
Akiyama H, Tanaka T, Maeno T, Kanai H, Kimura Y, Kishi S, and Kurabayashi M. Induction of VEGF gene expression by retinoic acid through Sp1-binding sites in retinoblastoma Y79 cells. Invest Ophthalmol Vis Sci 43: 1367–1374, 2002.
Albina JE, Mastrofrancesco B, Vessella JA, Louis CA, Henry WL Jr, and Reichner JS. HIF-1 expression in healing wounds: HIF-1 induction in primary inflammatory cells by TNF-. Am J Physiol Cell Physiol 281: C1971–C1977, 2001.
Arbiser JL, Moses MA, Fernandez CA, Ghiso N, Cao Y, Klauber N, Frank D, Brownlee M, Flynn E, Parangi S, Byers HR, and Folkman J. Oncogenic H-ras stimulates tumor angiogenesis by two distinct pathways. Proc Natl Acad Sci USA 94: 861–866, 1997.
Bohle A, Gise H, Mackensen-Haen S, and Stark-Jakob B. The obliteration of the post glomerular capillaries and its influence upon the function of both glomeruli and tubuli. Klin Wochenschr 59: 1043–1051, 1981.
Camenisch G, Tini M, Chilov D, Kvietikova I, Srinivas V, Caro J, Spielmann P, Wenger RH, and Gassmann M. General applicability of chicken egg yolk antibodies: the performance of IgY immunoglobulins raised against the hypoxia-inducible factor 1. FASEB J 13: 81–88, 1999.
Carmeliet P, Dor Y, Herbert JM, Fukumura D, Brusselmans K, Dewerchin M, Neeman M, Bono F, Abramovitch R, Maxwell P, Koch CJ, Ratcliffe P, Moons L, Jain RK, Collen D, Keshert E, and Keshet E. Role of HIF-1 in hypoxia-mediated apoptosis, cell proliferation and tumour angiogenesis. Nature 394: 485–490, 1998.
Choi YG, Baranowska-Daca E, Barrios R, Nguyen V, Nguyen C, and Truong LD. Peritubular capillary (PTC) loss mediates tubular loss and interstitial fibrosis (Abstract). J Am Soc Nephrol 10: 70A, 1999.
Choi YJ, Chakraborty S, Nguyen V, Nguyen C, Kim BK, Shim SI, Suki WN, and Truong LD. Peritubular capillary loss is associated with chronic tubulointerstitial injury in human kidney: altered expression of vascular endothelial growth factor. Hum Pathol 31: 1491–1497, 2000.
Fine LG, Bandyopadhay D, and Norman JT. Is there a common mechanism for the progression of different types of renal diseases other than proteinuria Towards the unifying theme of chronic hypoxia. Kidney Int 57, Suppl 75: S22–S26, 2000.
Fine LG, Orphanides C, and Norman JT. Progressive renal disease: the chronic hypoxia hypothesis. Kidney Int Suppl 65: S74–S78, 1998.
Forsythe JA, Jiang BH, Iyer NV, Agani F, Leung SW, Koos RD, and Semenza GL. Activation of vascular endothelial growth factor gene transcription by hypoxia-inducible factor 1. Mol Cell Biol 16: 4604–4613, 1996.
Garlanda C, Berthier R, Garin J, Stoppacciaro A, Ruco L, Vittet D, Gulino D, Matteucci C, Mantovani A, Vecchi A, and Dejana E. Characterization of MEC 14.7, a new monoclonal antibody recognizing mouse CD34: a useful reage for identifying and characterizing blood vessels and hematopoietic precursors. Eur J Cell Biol 73: 368–377, 1997.
Gerber HP, Dixit V, and Ferrara N. Vascular endothelial growth factor induces expression of the antiapoptotic proteins Bcl-2 and A1 in vascular endothelial cells. J Biol Chem 273: 13313–13316, 1998.
Gerber HP, McMurtrey A, Kowalski J, Yan M, Keyt BA, Dixit V, and Ferrara N. Vascular endothelial growth factor regulates endothelial cell survival through the phosphatidylinositol 3'-kinase/Akt signal transduction pathway. Requirement for Flk-1/KDR activation. J Biol Chem 273: 30336–30343, 1998.
Gleadle JM, Ebert BL, Firth JD, and Ratcliffe PJ. Regulation of angiogenic growth factor expression by hypoxia, transition metals, and chelating agents. Am J Physiol Cell Physiol 268: C1362–C1368, 1995.
Grisham MB, Jourd'Heuil D, and Wink DA. I. Physiological chemistry of nitric oxide and its metabolites: implications in inflammation. Am J Physiol Gastrointest Liver Physiol 276: G315–G321, 1999.
Hellwig-Burgel T, Rutkowski K, Metzen E, Fandrey J, and Jelkmann W. Interleukin-1 and tumor necrosis factor- stimulate DNA binding of hypoxia-inducible factor-1. Blood 94: 1561–1567, 1999.
Hopfl G, Ogunshola O, and Gassmann M. HIFs and tumors—causes and consequences. Am J Physiol Regul Integr Comp Physiol 286: R608–R623, 2004.
Huang LE, Willmore WG, Gu J, Goldberg MA, and Bunn HF. Inhibition of hypoxia-inducible factor 1 activation by carbon monoxide and nitric oxide. Implications for oxygen sensing and signaling. J Biol Chem 274: 9038–9044, 1999.
Jiang BH, Semenza GL, Bauer C, and Marti HH. Hypoxia-inducible factor 1 levels vary exponentially over a physiologically relevant range of O2 tension. Am J Physiol Cell Physiol 271: C1172–C1180, 1996.
Kairaitis LK, Kock C, Wang Y, Tay C, and Harris DCM. Time course of cortical tubular hypoxia in murine adriamycin nephrosis. J Am Soc Nephrol 12: A4270, 2001.
Kang DH, Joly AH, Oh SW, Hugo C, Kerjaschki D, Gordon KL, Mazzali M, Jefferson JA, Hughes J, Madsen KM, Schreiner GF, and Johnson RJ. Impaired angiogenesis in the remnant kidney model. I. Potential role of vascular endothelial growth factor and thrombospondin-1. J Am Soc Nephrol 12: 1434–1447, 2001.
Kim YG, Suga SI, Kang DH, Jefferson JA, Mazzali M, Gordon KL, Matsui K, Breiteneder-Geleff S, Shankland SJ, Hughes J, Kerjaschki D, Schreiner GF, and Johnson RJ. Vascular endothelial growth factor accelerates renal recovery in experimental thrombotic microangiopathy. Kidney Int 58: 2390–2399, 2000.
Kramer BK, Bucher M, Sandner P, Ittner KP, Riegger GA, Ritthaler T, and Kurtz A. Effects of hypoxia on growth factor expression in the rat kidney in vivo. Kidney Int 51: 444–447, 1997.
Laughlin KM, Evans SM, Jenkins WT, Tracy M, Chan CY, Lord EM, and Koch CJ. Biodistribution of the nitroimidazole EF5 (2-[2-nitro-1H-imidazol-1-yl]-N-(2,2,3,3,3-pentafluoropropyl) acetamide) in mice bearing subcutaneous EMT6 tumors. J Pharmacol Exp Ther 277: 1049–1057, 1996.
Liu Y, Cox SR, Morita T, and Kourembanas S. Hypoxia regulates vascular endothelial growth factor gene expression in endothelial cells. Identification of a 5' enhancer. Circ Res 77: 638–643, 1995.
Ljunquist A. The intrarenal arterial pattern in the normal and diseased human kidney. Acta Med Scand 174, Suppl 10: 5–34, 1963.
Minchenko A, Bauer T, Salceda S, and Caro J. Hypoxic stimulation of vascular endothelial growth factor expression in vitro and in vivo. Lab Invest 71: 374–379, 1994.
Norman JT, Clark IM, and Garcia PL. Hypoxia promotes fibrogenesis in human renal fibroblasts. Kidney Int 58: 2351–2366, 2000.
Norman JT, Clark IM, and Garcia PL. Regulation of TIMP-1 expression by hypoxia in kidney fibroblasts. Ann NY Acad Sci 878: 503–505, 1999.
Ohashi R, Kitamura H, and Yamanaka N. Peritubular capillary injury during the progression of experimental glomerulonephritis in rats. J Am Soc Nephrol 11: 47–56, 2000.
Ohashi R, Shimizu A, Masuda Y, Kitamura H, Ishizaki M, Sugisaki Y, and Yamanaka N. Peritubular capillary regression during the progression of experimental obstructive nephropathy. J Am Soc Nephrol 13: 1795–1805, 2002.
Okada F, Rak JW, Croix BS, Lieubeau B, Kaya M, Roncari L, Shirasawa S, Sasazuki T, and Kerbel RS. Impact of oncogenes in tumor angiogenesis: mutant K-ras up-regulation of vascular endothelial growth factor/vascular permeability factor is necessary, but not sufficient for tumorigenicity of human colorectal carcinoma cells. Proc Natl Acad Sci USA 95: 3609–3614, 1998.
Ong A and Fine L. Tubular-derived growth factors and cytokines in the pathogenesis of tubulointerstitial fibrosis: implications for human renal disease progression. Am J Kidney Dis 23: 205–209, 1994.
Ong AC, Jowett TP, Firth JD, Burton S, Karet FE, and Fine LG. An endothelin-1 mediated autocrine growth loop involved in human renal tubular regeneration. Kidney Int 48: 390–401, 1995.
Orphanides C, Fine LG, and Norman JT. Hypoxia stimulates proximal tubular cell matrix production via a TGF-1-independent mechanism. Kidney Int 52: 637–647, 1997.
Pal S, Datta K, and Mukhopadhyay D. Central role of p53 on regulation of vascular permeability factor/vascular endothelial growth factor (VPF/VEGF) expression in mammary carcinoma. Cancer Res 61: 6952–6957, 2001.
Palmer LA, Gaston B, and Johns RA. Normoxic stabilization of hypoxia-inducible factor-1 expression and activity: redox-dependent effect of nitrogen oxides. Mol Pharmacol 58: 1197–1203, 2000.
Parliament MB, Wiebe LI, and Franko AJ. Nitroimidazole adducts as markers for tissue hypoxia: mechanistic studies in aerobic normal tissues and tumour cells. Br J Cancer 66: 1103–1108, 1992.
Pillebout E, Burtin M, Yuan HT, Briand P, Woolf AS, Friedlander G, and Terzi F. Proliferation and remodeling of the peritubular microcirculation after nephron reduction: association with the progression of renal lesions. Am J Pathol 159: 547–560, 2001.
Rangan GK, Wang Y, Tay YC, and Harris DC. Inhibition of nuclear factor-B activation reduces cortical tubulointerstitial injury in proteinuric rats. Kidney Int 56: 118–134, 1999.
Reisinger K, Kaufmann R, and Gille J. Increased Sp1 phosphorylation as a mechanism of hepatocyte growth factor (HGF/SF)-induced vascular endothelial growth factor (VEGF/VPF) transcription. J Cell Sci 116: 225–238, 2003.
Salimath B, Marme D, and Finkenzeller G. Expression of the vascular endothelial growth factor gene is inhibited by p73. Oncogene 19: 3470–3476, 2000.
Sowter HM, Ratcliffe PJ, Watson P, Greenberg AH, and Harris AL. Hif-1-dependent regulation of hypoxic induction of the cell death factors bnip3 and nix in human tumors. Cancer Res 61: 6669–6673, 2001.
Stroka DM, Burkhardt T, Desbaillets I, Wenger RH, Neil DA, Bauer C, Gassmann M, and Candinas D. HIF-1 is expressed in normoxic tissue and displays an organ-specific regulation under systemic hypoxia. FASEB J 15: 2445–2453, 2001.
Talks KL, Turley H, Gatter KC, Maxwell PH, Pugh CW, Ratcliffe PJ, and Harris AL. The expression and distribution of the hypoxia-inducible factors HIF-1 and HIF-2 in normal human tissues, cancers, and tumor-associated macrophages. Am J Pathol 157: 411–421, 2000.
Wang GL and Semenza GL. Purification and characterization of hypoxia-inducible factor 1. J Biol Chem 270: 1230–1237, 1995.
Wang Y, Wang YP, Tay YC, and Harris DC. Progressive adriamycin nephropathy in mice: sequence of histologic and immunohistochemical events. Kidney Int 58: 1797–1804, 2000.
Wetscher GJ, Schwelberger H, Unger A, Offner FA, Profanter C, Glaser K, Klingler A, Gadenstaetter M, and Klinger P. Reflux-induced apoptosis of the esophageal mucosa is inhibited in Barrett's epithelium. Am J Surg 176: 569–573, 1998.
Yu R, Gao L, Jiang S, Guan P, and Mao B. Association of HIF-1 expression and cell apoptosis after traumatic brain injury in the rat. Chin J Traumatol 4: 218–221, 2001.
Zagzag D, Zhong H, Scalzitti JM, Laughner E, Simons JW, and Semenza GL. Expression of hypoxia-inducible factor 1 in brain tumors: association with angiogenesis, invasion, and progression. Cancer 88: 2606–2618, 2000.
Zhong H, De Marzo AM, Laughner E, Lim M, Hilton DA, Zagzag D, Buechler P, Isaacs WB, Semenza GL, and Simons JW. Overexpression of hypoxia-inducible factor 1 in common human cancers and their metastases. Cancer Res 59: 5830–5835, 1999.
Zoja C, Morigi M, Figliuzzi M, Bruzzi I, Oldroyd S, Benigni A, Ronco P, and Remuzzi G. Proximal tubular cell synthesis and secretion of endothelin-1 on challenge with albumin and other proteins. Am J Kidney Dis 26: 934–941, 1995.(Lukas Karolis Kairaitis, )
Institute of Veterinary Physiology, Vetsuisse Faculty, University of Zurich, Zurich, Switzerland
ABSTRACT
Cellular hypoxia has been proposed as a major factor in the pathogenesis of chronic renal injury, yet to date there has been no direct evidence to support its importance. Therefore, we examined cortical hypoxia in an animal model of chronic renal injury (murine adriamycin nephrosis; AN) by assessing nuclear localization of the oxygen-dependent -subunit of hypoxia-inducible factor-1 (HIF-1) in animals 7, 14, and 28 days after adriamycin. Results were assessed in conjunction with quantitation of the cortical microvasculature (by CD34 immunostaining) and cortical expression of VEGF. Cortical apoptosis was also examined by terminal deoxynucleotidyl transferase dUTP nick-end labeling staining. A dramatic and significant increase in nuclear localization of HIF-1 was seen 28 days after adriamycin in the context of severe glomerular and tubulointerstitial damage. Areas of nuclear HIF-1 staining did not colocalize with areas of cellular apoptosis. AN was also associated with a significant attenuation of the peritubular capillaries that was significant at 14 and 28 days after adriamycin. Cortical VEGF expression fell in a stepwise manner from day 7 until day 28 after adriamycin. In conclusion, these data are consistent with a significant increase in cellular hypoxia occurring in the advanced stages of murine AN. Increased cortical hypoxia was preceded by significant reductions in both the number of peritubular capillaries (i.e., oxygen supply) and the angiogenic cytokine VEGF. Apart from providing the first direct evidence for cellular hypoxia in a model of chronic renal disease, these results suggest that a primary dysregulation of angiogenesis may be the cause of increased hypoxia in this model.
capillary; chronic; tubular
CELLULAR HYPOXIA IN VITRO has been shown to have effects potentially relevant to the pathogenesis of chronic renal injury. Hypoxia has profibrotic effects in vitro, invoking the transcription of matrix proteins (28, 29, 35), profibrotic growth factors (15, 28, 35), and endothelin-1 (ET-1) (23, 34). In addition, distinct hypoxia-dependent pathways may exist for the induction of apoptosis (6, 43). Since renal fibrosis and tubular atrophy are inevitable components of the tubulointerstitial disease that accompanies all progressive renal diseases, it has been suggested that cellular responses to hypoxia may represent a "final common pathway" for the development of these pathological changes (9, 10).
To date, supportive evidence for this theory has been indirect. Progressive renal diseases are associated with both a progressive loss of vascular structures (4, 21, 22, 26, 30) and an excess of vasoconstrictor substances such as ANG II and ET-1 (33, 52), both of which may decrease oxygen supply to the renal parenchyma. In conjunction with these changes, there are alterations in renal oxygen demand due to the processes of hypertrophy, atrophy, and loss of tissue components, all of which may be present in variable degrees. The net effect in terms of the timing and severity of tubular hypoxia, therefore, remains to be elucidated.
We previously reported a progressive reduction in binding of the hypoxia marker EF5 in murine adriamycin nephrosis, a model of human chronic renal disease (20a). Although this result could be interpreted as indicating less hypoxia in this model, we were unable to correct for alterations in delivery or binding of the EF5 marker occurring as a consequence of renal injury. To further assess hypoxia in this model, we examined a potentially more reliable marker of cellular hypoxia [nuclear localization of the -subunit of hypoxia-inducible factor-1 (HIF-1)]. HIF-1 is the prototypical factor involved in transcriptional responses to cellular hypoxia (46). While the -subunit of HIF-1 is hypoxia independent, the -subunit is stabilized in an oxygen-dependent manner (reviewed in Ref. 18). Stabilization of HIF-1 allows it to bind to the -subunit and translocate to the nucleus, where it induces the transcription of multiple genes involved in homeostasis and disease pathogenesis. Detection of nuclear HIF-1 is therefore recognized as a marker of cellular hypoxia (20, 44, 50) that reflects local oxygen supply/demand ratios within the tissue.
Nuclear HIF-1 localization was examined at different time points in the course of murine adriamycin nephrosis by quantitative immunohistochemistry and Western blotting. The degree of nuclear HIF-1 accumulation was interpreted in the context of changes to the cortical microcirculation and measurement of the angiogenic growth factor VEGF.
MATERIALS AND METHODS
Murine Adriamycin Nephrosis
BALB/c mice were cared for in the Department of Animal Care, Westmead Hospital, under the ethical guidelines outlined in the Code of Practice for the Care and Use of Animals for scientific purposes established by the Australian National Health and Medical Research Council (NHMRC). Six-week-old male inbred BALB/c mice, weighing 22–25 g, were kept under standard conditions with unrestricted access to food and water in a 12:12-h light-dark cycle. Murine adriamycin nephrosis (AN) (47) was established by a single injection of adriamycin (doxorubicin hydrochloride, Pharmacia and Upjohn, Perth, Australia) at a dose of 11 μg/g into the tail vein of 6-wk-old male BALB/c mice. Animals (6–8/time point) were killed at 7, 14, and 28 days after adriamycin injection. At death, samples of serum and bladder urine were collected. The kidneys were rapidly removed and transferred to a metal plate cooled on ice for sectioning. The kidney poles were snap frozen in liquid nitrogen and stored at –70°C. The remaining kidney was sectioned coronally. One-half of the kidneys were placed in cryomoulds, embedded in optimum cooling temperature compound (Tissue-Tek, Sakura Finetechnicals, Torrance, CA) by immersion in liquid nitrogen-cooled isopentane, and stored at –70°C. The other half was immersed in 10% neutral-buffered formalin for 24 h, dehydrated in graded alcohols, and embedded in paraffin. Creatinine was measured in serum and urine samples by the Jaffé reaction using a Hitachi 747 multianalyzer (Tokyo, Japan). The same machine was used to measure serum albumin by the bromocresyl green method and urine protein using the benzethonium chloride turbidimetric method.
Comparative Histology
Quantitative histological measurements were obtained from the outer cortex using a modification of previously published methods (40, 47). Five-micrometer paraffin-embedded sections were stained with periodic acid-Schiff (PAS). Thirty random cortical images were taken using an SV-microdigital camera (Sound Vision, Framingham, MA) fitted to an Axioskop microscope (Zeiss, Oberkochen, Germany). Blinded analysis was then made using image-analysis software (Optimas version 6.5, Media Cybernetics, Seattle, WA).
Twenty outer cortical glomeruli sectioned through the hilum were analyzed. The outline of the glomerular tuft was traced with a mouse for an estimate of tuft volume. As an index of glomerulosclerosis (GS%), a 24-bit color threshold was used to detect PAS-positive material within the glomerular tuft, and the proportion of PAS staining within the tuft was expressed as a percentage.
The volume fraction of the interstitium was estimated in 10 nonoverlapping fields using an 8 x 11 counting grid. The proportion of interstitial grid intersections was calculated and expressed as a percentage.
Tubular measurements were obtained from 50 randomly selected cortical tubules/animal. Using line morphometry, the tubular diameter was estimated as the shortest axis through the center of the imaged tubule. Tubular cell height was assessed using line morphometry by measuring the radial distance between the basement membrane and the lumen of a line drawn through the center of the largest nucleus visible in the tubular cross section of interest.
Immunostaining for HIF-1
A well-characterized chicken polyclonal anti-HIF-1 antibody (5) was used as a primary antibody. Five-micrometer cryosections were fixed in 4% paraformaldehyde in PBS (pH 7.4) for 10 min and incubated with a 1:50 concentration of primary antibody overnight at 4°C. A peroxidase-conjugated rabbit anti-chicken IgY antibody (1:100, Pierce) was used as a secondary antibody and peroxidase-conjugated goat anti-rabbit IgG (Dako, Carpinteria, CA) as a tertiary antibody. Antibody binding was localized using nickel-enhanced diaminobenzidine (Pierce). Kidney and liver tissue from mice exposed to in vivo hypoxia (10% O2, 6 h) was used as a positive control.
Double immunostaining was performed to look for evidence of HIF-1 nuclear staining within interstitial inflammatory cells. HIF-1 staining was performed as above with the substitution of a Texas red-conjugated goat anti-rabbit antibody (Molecular Probes) as the tertiary antibody. Following HIF-1 staining, sections were incubated with rat monoclonal antibodies against murine macrophages (clone F4/80, 1:100 dilution, BD Biosciences, San Diego, CA), murine CD4 cells (clone RM4–5, 1:100, BD Biosciences), and murine CD8 cells (clone 53–6.7, 1:100, BD Biosciences) for 1 h at room temperature. A fluorescein-conjugated goat Fab' fragment to rat immunoglobulins (1:50, Serotec, Oxford, UK) was used as a secondary fluorescent reagent for the inflammatory cells.
Cortical nuclear staining of HIF-1 was analyzed semiquantitatively using image-analysis software (Optimas version 6.5). Ten random cortical images (each measuring 257 x 214 μm) were taken from each animal. Using the software, positively stained nuclei within the positive control tissue (hypoxic mouse liver) were used to set a 24-bit color "threshold" that recognized positively stained nuclei within the positive control image. When a suitable threshold was selected, it was then applied to the experimental slides to enable an unbiased comparison of the number of positive nuclei within the samples.
Western Blotting for HIF-1
Nuclear proteins were extracted from snap-frozen samples of renal cortex as previously described (40). Aliquots (25 μg) of nuclear proteins were resolved on an 8% SDS-polyacrylamide gel and transferred to a nitrocellulose membrane. The chicken IgY anti-HIF-1 (1:100) was used as a primary antibody and peroxidase-conjugated rabbit anti-chicken IgY antibody (1:1,000, Pierce) as the secondary antibody. Sites of antibody binding were revealed by immersion in an enhanced chemiluminescent substrate (ECL, Amersham) and exposure to autoradiographic film (Amersham). Comparative densitometry was performed using a Molecular Dynamics personal densitometer SI (Amersham) and ImageQuant image-analysis software (Amersham).
Colocalization of HIF-1 and Terminal Deoxynucleotidyl Transferase dUTP Nick-End Labeling Staining
Apoptosis within tissue samples was examined by terminal deoxynucleotidyltransferase dUTP nick-end labeling (TUNEL) staining. HIF-1 staining was performed as above using a Texas red-conjugated anti-rabbit antibody (Molecular Probes) as the tertiary antibody. TUNEL staining was then performed using a commercially available kit (in situ cell death detection kit, Roche Diagnostics, Mannheim, Germany). Fluorescent microscopy was performed using a Leica DM-LB fluorescent microscope (Wetzlar, Germany) fitted with a digital camera (Spot RT, Diagnostic Instruments, Sterling Heights, MI).
Assessment of Microvasculature
Immunostaining for murine CD34. The distribution of the endothelial cell marker CD34 (7, 12, 39) within the cortex was examined by immunohistochemistry and image analysis. Five-micrometer paraffin-embedded sections were dewaxed. Antigen retrieval was performed for 20 min at 95–99°C using a commercial target retrieval buffer (Dako). A rat monoclonal antibody to mouse CD34 (1:400, MEC 14.7, Cedarlane Laboratories) was used for 1 h at room temperature. Biotin-conjugated rabbit anti-rat immunoglobulin (Dako) was used as a secondary antibody. Sites of secondary antibody binding were demonstrated using a streptavidin-peroxidase conjugate (Dako) with diaminobenzidine as the enzyme substrate.
Peritubular capillary density was estimated using an adaptation of previously published methods (21). Ten random outer cortical fields (measuring 257 x 214 μm) were obtained from each animal. The capillary rarefaction index (CRI) was then measured using image-analysis software (Optimas version 6.5). Using a 10 x 12 counting grid, the proportion of grid squares that did not contain CD34-staining capillaries was calculated and expressed as a percentage.
Western blotting for CD34. Aliquots (50 μg) of cytoplasmic proteins were resolved on an 8% SDS-polyacrylamide gel and transferred to a nitrocellulose membranes. The MEC14.7 antibody was used at a concentration of 1:2,000, and biotinylated rabbit anti-rat IgG (1:2,000, Dako) was used as the secondary antibody. Sites of antibody binding were revealed by immersion in streptavidin-HRP (1:3,000, Dako) before chemiluminescent development. Comparative densitometry was performed as described above.
Assessment of Cortical VEGF
Immunostaining for VEGF. Four-micrometer paraffin-embedded sections were incubated with a 1:50 dilution of goat polyclonal antibody against VEGF (P20, Santa Cruz Biotechnology, Santa Cruz, CA). Preliminary staining of acetone-fixed cryosections showed VEGF staining in the glomerular mesangium, cortical tubules, and vascular smooth muscle. The degree of glomerular VEGF staining was less prominent in paraffin-embedded sections but was restored if antigen retrieval was performed. Antigen retrieval was not performed for routine staining due to the minimal contribution of glomerular VEGF to the total.
Peroxidase-conjugated rabbit anti-goat antibody (1:200, Dako) was used as the secondary antibody and peroxidase-conjugated goat anti-rabbit antibody (1:100, Dako) as the tertiary antibody. Cortical VEGF staining was analyzed semiquantitatively using image-analysis software (Optimas version 6.5) using an adaptation of published methods (21). Ten random images (measuring 257 x 214 μm) were obtained from the outer cortex of each animal. Using the software, a 24-bit color threshold was used to identify positively stained areas in each image. The area of each image represented by positively stained areas was then calculated by the software and expressed as a percentage.
Western blotting for VEGF. Aliquots (50 μg) of cytoplasmic proteins were resolved on a 12% SDS-polyacrylamide gel and transferred to nitrocellulose membranes. Staining was performed using a 1:100 dilution of the primary antibody and peroxidase-conjugated rabbit anti-goat Ig (Dako, 1:2,000) as the secondary antibody. Comparative densitometry was performed as described above.
Statistical Comparisons
Statistical comparisons between animal groups were performed using a computer-based statistical package (SPSS version 8, SPSS, Chicago, IL). Statistical significance between groups was tested using one-way ANOVA and the least squares method of post hoc analysis. A P value of 0.05 was considered to represent statistical significance.
RESULTS
Functional and Pathological Effects of Murine AN
As previously described (47), these animals developed progressive renal disease characterized by proteinuria, hypoalbuminemia, and impaired renal function (Table 1). Three mice died after adriamycin treatment (2 in the 28-day group and 1 in the 14-day group). Measured values for serum creatinine at 7 days after adriamycin (5.8 ± 5.5 μM) were outside the linear range of the analyzer used (stated as 9–2,000 μM) and were excluded from analysis. Serum creatinine levels were statistically greater than control at 28 days after adriamycin. AN animals developed proteinuria, as evidenced by a stepwise increase in the urine protein/creatinine ratio that was significantly greater than control after 7 days.
View this table:
Pathologically, this model was associated with progressive glomerular, tubular, and interstitial injury, all of which were severe at 28 days after adriamycin (Table 1). Glomerular changes observed included progressive glomerular enlargement and an increase in glomerulosclerosis. Progressive tubular atrophy was reflected by loss of the luminal brush border and a stepwise reduction in tubular cell height. The interstitium was expanded with a mononuclear cell infiltrate and increased PAS-positive matrix.
Immunostaining and Western Blotting for HIF-
Strong nuclear staining of HIF-1 was seen in the kidney and liver tissue from animals exposed to 10% oxygen (Fig. 1A). In the normal kidney, there was faint staining in cortical nuclei, predominantly in distal tubules (Fig. 1B). Glomeruli demonstrated no nuclear staining, and proximal tubules showed faint nuclear staining only. No nuclear staining was seen in the medulla or papilla.
The pattern of nuclear staining of HIF-1 was significantly altered in AN. Although HIF-1 staining was not obviously different from control at days 7 and 14 after adriamycin, a dramatic increase in nuclear HIF-1 staining was seen in the renal cortex at 28 days after adriamycin (Fig. 1C). Nuclear HIF-1 staining was most apparent in areas of extensive tubular dilatation and interstitial inflammation. In these areas, nuclear HIF-1 staining was seen in both tubular and interstitial cells. Double fluorescent staining for HIF-1 and inflammatory cell-surface markers showed nuclear HIF-1 staining within macrophages, CD4 lymphocytes, and CD8 lymphocytes in cortical samples from animals obtained 28 days after adriamycin (Fig. 2).
Semiquantitative analysis of the number of positive nuclei per unit of cortical area reflected these observations (Fig. 1D). There was a dramatic increase in the number of positively stained nuclei at the 28-day time point (191 ± 115 vs. 9 ± 17 nuclei/unit area, P < 0.01). An identical pattern was seen when nuclear protein extracts were probed for HIF-1 (Fig. 1E).
Colocalization of HIF-1 and TUNEL Staining
TUNEL staining revealed apoptotic nuclei within the middle epithelial layers of the positive control tissue (murine esophagus) consistent with the normal process of keratinocyte maturation (Fig. 3A). Double staining revealed occasional apoptotic nuclei within the epithelial layers of the esophagus that also stained positive for HIF-1, although the majority of apoptotic nuclei were negative (Fig. 3A). No apoptosis was visible in the normal kidney (not shown). Apoptotic nuclei were detectable after adriamycin. There were scanty apoptotic nuclei at day 7 after adriamycin, with apoptotic nuclei at days 14 and 28 after adriamycin largely confined to occasional epithelial cells lining atrophic tubules and sloughed nuclei present within the tubular lumen (Fig. 3, B and C). Double staining of the AN samples did not demonstrate areas of overlap of the HIF-1 or apoptotic stains (Fig. 3, B and C).
Assessment of Cortical Microvasculature
CD34 immunostaining of the normal renal cortex revealed a dense capillary network that in some areas appeared to completely encircle individual tubules (Fig. 4A). The glomerular capillary network also exhibited positive staining. A dense intertubular capillary network was also noted in the medulla and papilla (not shown). AN was associated with significant changes in the density of the cortical microvasculature. There was a significant reduction in visible capillaries from day 14 that was most prominent 28 days after adriamycin (Fig. 4B). The reduction in capillaries was particularly prominent in areas of interstitial expansion and tubular atrophy. In keeping with these observations, there was a significant increase in the capillary rarefaction index after adriamycin (Fig. 4C). Microvascular density was reduced threefold at 14 days after adriamycin (P < 0.01) and fivefold at 28 days (P < 0.01). Western blotting of protein extracts (Fig. 4D) showed a similar stepwise reduction in cortical CD34 expression that was significant at 14 and 28 days after adriamycin.
Immunostaining and Western Blotting for VEGF
Extensive tubular staining for VEGF was seen in the normal kidney, with accentuation of distal tubules (Fig. 5A). There was a reduction in tubular VEGF staining following adriamycin, particularly in areas of tubular damage or atrophy (Fig. 5B). Semiquantitative analysis of cortical VEGF expression reflected these changes. There was a significant reduction in cortical VEGF detectable at 7 days after adriamycin (Fig. 5C), with further stepwise reductions at later time points. Western blotting of cortical proteins also showed a similar decremental pattern of VEGF expression (Fig. 5D).
DISCUSSION
The results presented in this paper describe in detail the time course of cortical HIF-1 expression in murine AN, a robust model of human focal glomerulosclerosis. Increased expression and nuclear localization of HIF-1 were seen in the late stages of this model and were localized to tubular and interstitial cells in areas of extensive parenchymal injury. This finding represents the first direct evidence of hypoxia in an animal model of chronic renal disease. Increased HIF-1 expression in these experiments followed significant reductions in both peritubular capillary density (a surrogate measure of oxygen supply) and reduced cortical expression of the angiogenic growth factor VEGF. Areas of cellular hypoxia did not colocalize with areas of tubular apoptosis in this model.
Two separate methods were used to compare nuclear localization of HIF-1 in the current experiments (immunostaining and Western blotting). Consistent with previous reports, we found minimal nuclear staining of HIF-1 in the normal kidney (44, 45, 51). Cortical nuclear HIF-1 was significantly increased by whole animal hypoxia as well as in the later stages of AN. This pattern of nuclear localization is consistent with increased stabilization of HIF-1 in the advanced stages of this model. The cellular fate of HIF-1 is determined by the activity of a redox-sensitive pathway (reviewed in Ref. 18). In the presence of oxygen, the -subunit undergoes prolyl hydroxylation and is targeted for proteasomic degradation. In the absence of oxygen, this hydroxylation does not occur, and after translocation to the nucleus, HIF-1 is free to bind to the -subunit (the aryl hydrocarbon nuclear receptor translocator) to form the HIF-1 complex that, in turn, interacts with hypoxia-responsive elements in the promoter region of multiple genes. Other than hypoxia, a number of other stimuli may potentially affect the stabilization of HIF-1 in this model, including the proinflammatory molecules TNF-, IL-1 (2, 17), and nitric oxide (NO) (19, 37). It is noteworthy, however, that the effects of TNF-, IL-1, and NO on HIF-1 stabilization are based on in vitro experiments. A consistent difficulty with the interpretation of these experimental systems is that these molecules may well exert their in vitro effects by influencing oxygen consumption, weakening the case for a direct effect of these stimuli on HIF-1 stabilization. Furthermore, extrapolation of the results obtained with the use of NO donors in vitro is made more complex by the fact that the production of NO in vivo is an oxygen-dependent process (16), and exogenous NO donors may themselves disrupt the redox-dependent process of HIF-1 stabilization in a way that is not relevant to NO produced in vivo (19, 37). The effects of endogenous TNF-, IL-1, and NO on HIF-1 stabilization in the current experiments are therefore uncertain.
The presence of increased nuclear HIF-1 in the later stages of this model is strongly suggestive of a significant increase in cellular hypoxia in advanced disease. Before these data were gathered, experimental evidence to support the presence of "chronic renal hypoxia" has been based largely on indirect findings, including the observed reduction in peritubular capillaries (4, 21, 22, 26, 30) and extrapolation of the profibrotic effects of hypoxia in vitro (28, 29, 35). The principal advantage of a HIF-1-based method of hypoxia detection is that it gives direct evidence of transcriptionally relevant hypoxia at a cellular level. The HIF-1 pathway is activated in vivo when oxygen demand outweighs supply (18). This delicate balance cannot be inferred from microelectrode measurements that have previously been regarded as the "gold standard" for direct hypoxia measurement.
The finding of increased HIF-1 stabilization in AN warrants a consideration of whether HIF-1-dependent stimuli may contribute to the pathogenesis of renal injury in this model. Although hypoxia in vitro induces a wide range of profibrotic effects, including the stimulation of profibrotic growth factors (15, 28, 35), matrix proteins (28, 29, 35), and matrix stabilizers, it was notable in the current experiments that nuclear HIF-1 was increased only in the later stages of this model, at a time when interstitial fibrosis was already present. This temporal association makes it unlikely that hypoxia contributes significantly to the development of interstitial fibrosis in this model. However, hypoxia could contribute to pathological changes occurring in the late stages of AN. Our finding that apoptosis was more extensive in the later stages of this model in conjunction with the possible existence of HIF-1-dependent pathways for apoptosis (6) prompted us to examine for evidence of colocalization of HIF-1 and a nuclear marker of apoptosis (TUNEL staining). Using a previously described fluorescent technique (49), we found evidence of colocalization of HIF-1 and TUNEL staining in nuclei of the esophageal epithelium, a tissue in which both cellular hypoxia (24, 38) and apoptosis (48) occur under normoxic conditions. In the kidney sections, apoptotic nuclei were most prominent in the later stages of AN, with TUNEL labeling of occasional tubular epithelial cells and intraluminal cellular casts. Although HIF-1 staining was detectable in the same sections, there was no colocalization of HIF-1 or apoptosis in the kidney samples. This result suggests that HIF-1-dependent pathways are unlikely to contribute significantly to tubular apoptosis in this model but may instead represent an attempted homeostatic response to relative oxygen deficiency within the tubular cells in advanced AN.
To investigate the potential causes of hypoxia in this model, we measured peritubular capillary density as a surrogate measure of oxygen supply. AN was associated with a significant reduction in CD34-labeled peritubular capillaries, consistent with similar findings in other chronic renal diseases (4, 21, 22, 26, 30). This reduction in capillary density preceded the detection of increased cortical hypoxia and would suggest that a reduction in oxygen supply plays a role in the development of hypoxia. Consistent with this hypothesis was the finding that the areas of greatest capillary rarefaction at day 28 after adriamycin were those with the greatest degree of tubulointerstitial injury and correlated morphologically with the areas within which HIF-1 nuclear staining was seen. Two methods (CD34 immunostaining and Western blotting) were used to assess the cortical microvasculature in this model. Western blotting was performed because the method used for calculation of the capillary rarefaction index assumes that the volume of tissue assessed in the normal kidney is unchanged in AN. This model is associated with tubular dilatation, tubular loss, and interstitial expansion, all of which affect the volume of tissue to a variable amount. The reduction of CD34 protein in AN by Western blotting (which is not subject to the same limitations) was further evidence that the cortical microvasculature is reduced in this model.
A further result from these experiments is that cortical expression of the angiogenic cytokine VEGF was significantly reduced as the disease progressed, consistent with results obtained in other models of chronic renal injury (8, 21, 31). This result may seem contradictory to the finding of increased hypoxia in this model, particularly since VEGF is both induced by hypoxia and transcriptionally activated by HIF-1 (11, 25, 27). There are several explanations for this apparent disparity. First, HIF-1 is not the only known transcriptional activator of VEGF, with experimental evidence suggesting a role for SP-1 (1, 41) as well as the oncogene ras (3, 32). VEGF transcription is also inhibited by the anti-oncogenes p53 (36) and p73 (42). Also relevant are previous experiments showing that VEGF expression is reduced in cutaneous models of inflammation at a time when HIF-1 expression is increased (2) and the finding that VEGF production in response to hypoxia was reduced in the presence of macrophage-derived cytokines in vitro (21). These latter findings are potentially relevant to the reduction of VEGF observed in the current experiments, since AN is associated with a prominent mononuclear infiltrate (47). It is therefore likely that the finding of reduced VEGF in AN (as well as in other renal disease models) is multifactorial in origin.
Whatever the cause, the finding that VEGF is reduced in AN has potential relevance to the increased HIF-1 expression that occurs in the late stages of this model. If cellular hypoxia is caused by a reduction in oxygen supply engendered by microvascular loss in this model, it is also possible, in turn, that the reduction in peritubular capillaries may in fact be secondary to reduced local production of VEGF, since VEGF is an important survival factor for endothelial cells (13, 14). In support of this, the reductions in cortical VEGF seen in AN were observed before a detectable reduction in peritubular capillaries, which, in turn, preceded a detectable increase in cellular hypoxia.
In summary, we showed a significant increase in nuclear HIF-1 localization in murine AN, consistent with the presence of increased cellular hypoxia in the later stages of this model. HIF-1-dependent mechanisms do not appear to be the predominant cause of apoptosis in this model, and the timing of the increase in nuclear HIF-1 would argue against a significant role for hypoxia in either parenchymal injury or interstitial fibrosis in this model. The observed reductions in both the density of peritubular capillaries and expression of the principal angiogenic growth factor VEGF suggest that the increase in nuclear HIF-1 in AN is secondary to a disruption of mechanisms involved in the maintenance of the normally extensive capillary network within the kidney.
GRANTS
This work was supported by grants from the National Health and Medical Research Council of Australia and the Swiss National Science Foundation.
ACKNOWLEDGMENTS
In vivo hypoxia experiments to provide positive controls for HIF-1 staining were performed in collaboration with Dr. Scott Geller and Prof. Jonathon Stone from the Department of Retinal Biology, the University of Sydney.
FOOTNOTES
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
REFERENCES
Akiyama H, Tanaka T, Maeno T, Kanai H, Kimura Y, Kishi S, and Kurabayashi M. Induction of VEGF gene expression by retinoic acid through Sp1-binding sites in retinoblastoma Y79 cells. Invest Ophthalmol Vis Sci 43: 1367–1374, 2002.
Albina JE, Mastrofrancesco B, Vessella JA, Louis CA, Henry WL Jr, and Reichner JS. HIF-1 expression in healing wounds: HIF-1 induction in primary inflammatory cells by TNF-. Am J Physiol Cell Physiol 281: C1971–C1977, 2001.
Arbiser JL, Moses MA, Fernandez CA, Ghiso N, Cao Y, Klauber N, Frank D, Brownlee M, Flynn E, Parangi S, Byers HR, and Folkman J. Oncogenic H-ras stimulates tumor angiogenesis by two distinct pathways. Proc Natl Acad Sci USA 94: 861–866, 1997.
Bohle A, Gise H, Mackensen-Haen S, and Stark-Jakob B. The obliteration of the post glomerular capillaries and its influence upon the function of both glomeruli and tubuli. Klin Wochenschr 59: 1043–1051, 1981.
Camenisch G, Tini M, Chilov D, Kvietikova I, Srinivas V, Caro J, Spielmann P, Wenger RH, and Gassmann M. General applicability of chicken egg yolk antibodies: the performance of IgY immunoglobulins raised against the hypoxia-inducible factor 1. FASEB J 13: 81–88, 1999.
Carmeliet P, Dor Y, Herbert JM, Fukumura D, Brusselmans K, Dewerchin M, Neeman M, Bono F, Abramovitch R, Maxwell P, Koch CJ, Ratcliffe P, Moons L, Jain RK, Collen D, Keshert E, and Keshet E. Role of HIF-1 in hypoxia-mediated apoptosis, cell proliferation and tumour angiogenesis. Nature 394: 485–490, 1998.
Choi YG, Baranowska-Daca E, Barrios R, Nguyen V, Nguyen C, and Truong LD. Peritubular capillary (PTC) loss mediates tubular loss and interstitial fibrosis (Abstract). J Am Soc Nephrol 10: 70A, 1999.
Choi YJ, Chakraborty S, Nguyen V, Nguyen C, Kim BK, Shim SI, Suki WN, and Truong LD. Peritubular capillary loss is associated with chronic tubulointerstitial injury in human kidney: altered expression of vascular endothelial growth factor. Hum Pathol 31: 1491–1497, 2000.
Fine LG, Bandyopadhay D, and Norman JT. Is there a common mechanism for the progression of different types of renal diseases other than proteinuria Towards the unifying theme of chronic hypoxia. Kidney Int 57, Suppl 75: S22–S26, 2000.
Fine LG, Orphanides C, and Norman JT. Progressive renal disease: the chronic hypoxia hypothesis. Kidney Int Suppl 65: S74–S78, 1998.
Forsythe JA, Jiang BH, Iyer NV, Agani F, Leung SW, Koos RD, and Semenza GL. Activation of vascular endothelial growth factor gene transcription by hypoxia-inducible factor 1. Mol Cell Biol 16: 4604–4613, 1996.
Garlanda C, Berthier R, Garin J, Stoppacciaro A, Ruco L, Vittet D, Gulino D, Matteucci C, Mantovani A, Vecchi A, and Dejana E. Characterization of MEC 14.7, a new monoclonal antibody recognizing mouse CD34: a useful reage for identifying and characterizing blood vessels and hematopoietic precursors. Eur J Cell Biol 73: 368–377, 1997.
Gerber HP, Dixit V, and Ferrara N. Vascular endothelial growth factor induces expression of the antiapoptotic proteins Bcl-2 and A1 in vascular endothelial cells. J Biol Chem 273: 13313–13316, 1998.
Gerber HP, McMurtrey A, Kowalski J, Yan M, Keyt BA, Dixit V, and Ferrara N. Vascular endothelial growth factor regulates endothelial cell survival through the phosphatidylinositol 3'-kinase/Akt signal transduction pathway. Requirement for Flk-1/KDR activation. J Biol Chem 273: 30336–30343, 1998.
Gleadle JM, Ebert BL, Firth JD, and Ratcliffe PJ. Regulation of angiogenic growth factor expression by hypoxia, transition metals, and chelating agents. Am J Physiol Cell Physiol 268: C1362–C1368, 1995.
Grisham MB, Jourd'Heuil D, and Wink DA. I. Physiological chemistry of nitric oxide and its metabolites: implications in inflammation. Am J Physiol Gastrointest Liver Physiol 276: G315–G321, 1999.
Hellwig-Burgel T, Rutkowski K, Metzen E, Fandrey J, and Jelkmann W. Interleukin-1 and tumor necrosis factor- stimulate DNA binding of hypoxia-inducible factor-1. Blood 94: 1561–1567, 1999.
Hopfl G, Ogunshola O, and Gassmann M. HIFs and tumors—causes and consequences. Am J Physiol Regul Integr Comp Physiol 286: R608–R623, 2004.
Huang LE, Willmore WG, Gu J, Goldberg MA, and Bunn HF. Inhibition of hypoxia-inducible factor 1 activation by carbon monoxide and nitric oxide. Implications for oxygen sensing and signaling. J Biol Chem 274: 9038–9044, 1999.
Jiang BH, Semenza GL, Bauer C, and Marti HH. Hypoxia-inducible factor 1 levels vary exponentially over a physiologically relevant range of O2 tension. Am J Physiol Cell Physiol 271: C1172–C1180, 1996.
Kairaitis LK, Kock C, Wang Y, Tay C, and Harris DCM. Time course of cortical tubular hypoxia in murine adriamycin nephrosis. J Am Soc Nephrol 12: A4270, 2001.
Kang DH, Joly AH, Oh SW, Hugo C, Kerjaschki D, Gordon KL, Mazzali M, Jefferson JA, Hughes J, Madsen KM, Schreiner GF, and Johnson RJ. Impaired angiogenesis in the remnant kidney model. I. Potential role of vascular endothelial growth factor and thrombospondin-1. J Am Soc Nephrol 12: 1434–1447, 2001.
Kim YG, Suga SI, Kang DH, Jefferson JA, Mazzali M, Gordon KL, Matsui K, Breiteneder-Geleff S, Shankland SJ, Hughes J, Kerjaschki D, Schreiner GF, and Johnson RJ. Vascular endothelial growth factor accelerates renal recovery in experimental thrombotic microangiopathy. Kidney Int 58: 2390–2399, 2000.
Kramer BK, Bucher M, Sandner P, Ittner KP, Riegger GA, Ritthaler T, and Kurtz A. Effects of hypoxia on growth factor expression in the rat kidney in vivo. Kidney Int 51: 444–447, 1997.
Laughlin KM, Evans SM, Jenkins WT, Tracy M, Chan CY, Lord EM, and Koch CJ. Biodistribution of the nitroimidazole EF5 (2-[2-nitro-1H-imidazol-1-yl]-N-(2,2,3,3,3-pentafluoropropyl) acetamide) in mice bearing subcutaneous EMT6 tumors. J Pharmacol Exp Ther 277: 1049–1057, 1996.
Liu Y, Cox SR, Morita T, and Kourembanas S. Hypoxia regulates vascular endothelial growth factor gene expression in endothelial cells. Identification of a 5' enhancer. Circ Res 77: 638–643, 1995.
Ljunquist A. The intrarenal arterial pattern in the normal and diseased human kidney. Acta Med Scand 174, Suppl 10: 5–34, 1963.
Minchenko A, Bauer T, Salceda S, and Caro J. Hypoxic stimulation of vascular endothelial growth factor expression in vitro and in vivo. Lab Invest 71: 374–379, 1994.
Norman JT, Clark IM, and Garcia PL. Hypoxia promotes fibrogenesis in human renal fibroblasts. Kidney Int 58: 2351–2366, 2000.
Norman JT, Clark IM, and Garcia PL. Regulation of TIMP-1 expression by hypoxia in kidney fibroblasts. Ann NY Acad Sci 878: 503–505, 1999.
Ohashi R, Kitamura H, and Yamanaka N. Peritubular capillary injury during the progression of experimental glomerulonephritis in rats. J Am Soc Nephrol 11: 47–56, 2000.
Ohashi R, Shimizu A, Masuda Y, Kitamura H, Ishizaki M, Sugisaki Y, and Yamanaka N. Peritubular capillary regression during the progression of experimental obstructive nephropathy. J Am Soc Nephrol 13: 1795–1805, 2002.
Okada F, Rak JW, Croix BS, Lieubeau B, Kaya M, Roncari L, Shirasawa S, Sasazuki T, and Kerbel RS. Impact of oncogenes in tumor angiogenesis: mutant K-ras up-regulation of vascular endothelial growth factor/vascular permeability factor is necessary, but not sufficient for tumorigenicity of human colorectal carcinoma cells. Proc Natl Acad Sci USA 95: 3609–3614, 1998.
Ong A and Fine L. Tubular-derived growth factors and cytokines in the pathogenesis of tubulointerstitial fibrosis: implications for human renal disease progression. Am J Kidney Dis 23: 205–209, 1994.
Ong AC, Jowett TP, Firth JD, Burton S, Karet FE, and Fine LG. An endothelin-1 mediated autocrine growth loop involved in human renal tubular regeneration. Kidney Int 48: 390–401, 1995.
Orphanides C, Fine LG, and Norman JT. Hypoxia stimulates proximal tubular cell matrix production via a TGF-1-independent mechanism. Kidney Int 52: 637–647, 1997.
Pal S, Datta K, and Mukhopadhyay D. Central role of p53 on regulation of vascular permeability factor/vascular endothelial growth factor (VPF/VEGF) expression in mammary carcinoma. Cancer Res 61: 6952–6957, 2001.
Palmer LA, Gaston B, and Johns RA. Normoxic stabilization of hypoxia-inducible factor-1 expression and activity: redox-dependent effect of nitrogen oxides. Mol Pharmacol 58: 1197–1203, 2000.
Parliament MB, Wiebe LI, and Franko AJ. Nitroimidazole adducts as markers for tissue hypoxia: mechanistic studies in aerobic normal tissues and tumour cells. Br J Cancer 66: 1103–1108, 1992.
Pillebout E, Burtin M, Yuan HT, Briand P, Woolf AS, Friedlander G, and Terzi F. Proliferation and remodeling of the peritubular microcirculation after nephron reduction: association with the progression of renal lesions. Am J Pathol 159: 547–560, 2001.
Rangan GK, Wang Y, Tay YC, and Harris DC. Inhibition of nuclear factor-B activation reduces cortical tubulointerstitial injury in proteinuric rats. Kidney Int 56: 118–134, 1999.
Reisinger K, Kaufmann R, and Gille J. Increased Sp1 phosphorylation as a mechanism of hepatocyte growth factor (HGF/SF)-induced vascular endothelial growth factor (VEGF/VPF) transcription. J Cell Sci 116: 225–238, 2003.
Salimath B, Marme D, and Finkenzeller G. Expression of the vascular endothelial growth factor gene is inhibited by p73. Oncogene 19: 3470–3476, 2000.
Sowter HM, Ratcliffe PJ, Watson P, Greenberg AH, and Harris AL. Hif-1-dependent regulation of hypoxic induction of the cell death factors bnip3 and nix in human tumors. Cancer Res 61: 6669–6673, 2001.
Stroka DM, Burkhardt T, Desbaillets I, Wenger RH, Neil DA, Bauer C, Gassmann M, and Candinas D. HIF-1 is expressed in normoxic tissue and displays an organ-specific regulation under systemic hypoxia. FASEB J 15: 2445–2453, 2001.
Talks KL, Turley H, Gatter KC, Maxwell PH, Pugh CW, Ratcliffe PJ, and Harris AL. The expression and distribution of the hypoxia-inducible factors HIF-1 and HIF-2 in normal human tissues, cancers, and tumor-associated macrophages. Am J Pathol 157: 411–421, 2000.
Wang GL and Semenza GL. Purification and characterization of hypoxia-inducible factor 1. J Biol Chem 270: 1230–1237, 1995.
Wang Y, Wang YP, Tay YC, and Harris DC. Progressive adriamycin nephropathy in mice: sequence of histologic and immunohistochemical events. Kidney Int 58: 1797–1804, 2000.
Wetscher GJ, Schwelberger H, Unger A, Offner FA, Profanter C, Glaser K, Klingler A, Gadenstaetter M, and Klinger P. Reflux-induced apoptosis of the esophageal mucosa is inhibited in Barrett's epithelium. Am J Surg 176: 569–573, 1998.
Yu R, Gao L, Jiang S, Guan P, and Mao B. Association of HIF-1 expression and cell apoptosis after traumatic brain injury in the rat. Chin J Traumatol 4: 218–221, 2001.
Zagzag D, Zhong H, Scalzitti JM, Laughner E, Simons JW, and Semenza GL. Expression of hypoxia-inducible factor 1 in brain tumors: association with angiogenesis, invasion, and progression. Cancer 88: 2606–2618, 2000.
Zhong H, De Marzo AM, Laughner E, Lim M, Hilton DA, Zagzag D, Buechler P, Isaacs WB, Semenza GL, and Simons JW. Overexpression of hypoxia-inducible factor 1 in common human cancers and their metastases. Cancer Res 59: 5830–5835, 1999.
Zoja C, Morigi M, Figliuzzi M, Bruzzi I, Oldroyd S, Benigni A, Ronco P, and Remuzzi G. Proximal tubular cell synthesis and secretion of endothelin-1 on challenge with albumin and other proteins. Am J Kidney Dis 26: 934–941, 1995.(Lukas Karolis Kairaitis, )