Divergent Expression Patterns for Hypoxia-Inducible Factor-1 and Aryl Hydrocarbon Receptor Nuclear Transporter-2 in Developing Kidney
Paul B. Freeburg and
Dale R. Abrahamson
Department of Anatomy and Cell Biology, University of Kansas Medical Center, Kansas City, Kansas
Correspondence to Dr. Dale R. Abrahamson, Department of Anatomy and Cell Biology, University of Kansas Medical Center, MS 3038, 3901 Rainbow Boulevard, Kansas City, KS 66160. Phone: 913-588-7000; Fax: 913-588-2710; E-mail: dabrahamson{at}kumc.edu
The hypoxia-inducible factors (HIF) are / heterodimeric transcriptionfactors of the basic helix-loop-helix-Per-Arnt-Sim (bHLH-PAS)superfamily and are chiefly responsible for cellular adaptationto oxygen deprivation. HIF function relies on the stabilizationof the subunit. When oxygen tension falls, HIF- subunits translocateto the nucleus and, upon dimerization with HIF-, activate transcriptionof target genes, including vascular endothelial growth factor,vascular endothelial growth factor receptor-1 and -2, and WT-1,which are vital for kidney development. HIF- subunits are stableregardless of oxygen concentration and constitutively translocateto the nucleus. It was shown previously that HIF-1 protein expressionis nearly ubiquitous in newborn kidney and that HIF-1 dimerizeswith either HIF-1 or -2. Here it is shown that aryl hydrocarbonreceptor nuclear transporter-2 (ARNT2/HIF-2) also heterodimerizedwith HIF-1 and -2. ARNT2/HIF-2 protein was highly expressedin newborn kidney but decreased significantly with age, whereasHIF-1 levels remained relatively constant. By immunohistochemicalanalysis, widespread expression of HIF-1 was observed in developingand mature kidneys. ARNT2/HIF-2 protein distribution was restrictedto distal segments of developing nephrons and in mature kidneywas confined specifically to thick ascending limb of Henle’sloop. The data presented here suggest that ARNT2/HIF-2 is requiredat high levels during nephrogenesis in distal tubules and laterexclusively in thick ascending limb. Furthermore, Hypoxyprobe-1and lotus lectin co-localization studies showed that developingproximal convoluted tubules were the most severely hypoxic nephronsegment in immature kidney. Because HIF-2 protein was not abundantlyexpressed in this segment, it may not be engaged in mediatingresponses to severe hypoxia. The differential distribution patternsfor HIF-1 and -2 suggest divergent roles during kidney developmentfor these highly related bHLH-PAS proteins.
The basic helix-loop-helix-Per-Arnt-Sim (bHLH-PAS) family oftranscription factors functions as / heterodimeric DNA bindingcomplexes and regulates a remarkably diverse number of biologicevents. These include responses to oxygen deprivation and exposureto toxins, control of circadian rhythms, and regulation of hormonesignaling (1). The bHLH-PAS heterodimers typically consist ofone broadly expressed, readily available member ( subunit) thatheterodimerizes with a partner ( subunit) that has a tightlyrestricted expression profile and is sensitive to environmentalcues. Perhaps the most widely expressed and best understoodprotein in this family is the aryl hydrocarbon receptor nucleartransporter (ARNT), which acts as the subunit for transcriptionalresponses to both toxin exposure and hypoxic stimuli (2–6).To aid in metabolism of the toxin benzo[a]pyrene, ARNT heterodimerizeswith the bHLH-PAS protein aryl hydrocarbon receptor (AHR) andbinds to the xenobiotic response element located in the enhancerregions of the cytochrome P450 gene isoforms Cyp1a1, Cyp1a2,and Cyp1b1 (7). Expression of these proteins purges the systemof the toxin by way of their metabolic activity toward benzo[a]pyrene.Ordinarily, AHR is held quiescent in the cytoplasm by bindingheat shock protein 90 and AHR interaction factor. AHR is releasedby this complex, combines with ARNT, and induces gene expressionupon binding of the toxin (8). AHR therefore acts as the subunitof bHLH-PAS heterodimer because its function is reliant on anenvironmental insult, which in this case is the cellular toxinbenzo[a]pyrene.
Another clearly defined role for ARNT as a subunit of bHLH-PASheterodimers is in the hypoxia responsive pathway, where ARNTis now commonly known as hypoxia-inducible factor-1 (HIF-1).Early studies of erythropoietin (Epo) gene expression led tothe identification of a heterodimeric transcription factor thatstimulates Epo expression in response to low oxygen tension(9). This factor was purified and found to contain ARNT (HIF-1)and a new member of the bHLH-PAS family, named HIF-1. Sincethe discovery of the HIF-1/HIF-1 heterodimer (HIF-1), the biochemicalbehavior of these proteins in response to oxygen deprivationhas been studied extensively. Most evidence indicates that HIF-1protein expression occurs in all cells and that the proteinis stable and is found abundantly in the nucleus regardlessof oxygen tension (10). HIF-1, conversely, is regulated directlyby cellular oxygen availability (11). When intracellular oxygentension is sufficient to satisfy metabolic demands (normoxia),HIF-1 protein undergoes an oxygen-dependent hydroxylation ona highly conserved proline residue (12,13). This hydroxyprolineresidue is vital, because it renders HIF-1 recognizable by thevon Hippel-Lindau ubiquitin ligase complex. Hydroxylated HIF-1becomes polyubiquinated and rapidly destroyed by proteasomes.However, when oxygen tension falls below critical levels (hypoxia),HIF-1 escapes proline hydroxylation, is not recognized by vonHippel-Lindau, and therefore avoids proteasomal degradation.Stabilized HIF-1 then translocates to the nucleus, heterodimerizeswith the readily available HIF-1, and binds the hypoxia-responsiveelement (HRE) located in the promoter/enhancer of HIF targetgenes. HIF-2, initially called endothelial PAS domain protein,has very similar structural and biochemical properties to HIF-1and also heterodimerizes with HIF-1 (14–16). Among thegenes induced by HIF-1 and -2 are those highly involved in kidneydevelopment, including vascular endothelial growth factor (VEGF),VEGF receptor-1 (VEGFR-1) and -2, angiopoietin-2, Tie-2, Epo,and WT-1 (17–23).
Although much has been learned about the subunits of bHLH-PASheterodimers (AHR and HIF-1 and -2), relatively little is knownabout the HIF- counterparts. However, recent studies have suggestedthat bHLH-PAS proteins are more complex than initially thought.For example, five different splice variants of HIF-1 have beenidentified in the rat (24), and additional isoforms have beenshown to contain deletions at exon 5, 6, or 11 or an insertionat exon 20. Other HIF-1 mRNA isoforms contain a variable numberof codons in exon 16 encoding glutamine. Expression profilesand functions of each of the alternately spliced variants haveyet to be described. In addition, in some cell lines, HIF-1protein is apparently sensitive to oxygen concentration andaccumulates in hypoxia, as observed in a human neuroblastomacell culture system (25).
HIF-1 and -2 both are required for proper development, as demonstratedin mice with targeted deletions of these genes (14,26,27). HIF-1null embryos die at embryonic day 9.5 (E9.5) from cardiovascularand neural tube defects, as well as widespread mesenchymal celldeath. HIF-2 knockout mice on the C57Bl/6 or 129/Sv geneticbackground die at E13.5 to E15.5 from bradycardia and lack ofcatecholamine synthesis. Overall vascular development in thesemice, at least at the gross anatomic level, seems normal (14).HIF-2 mutants on the ICR/129Sv background, however, show severevascular defects and die by E13.5 (15).
HIF-1 null embryos die at E11 from abnormal hematopoietic andvascular development (28). A second -class bHLH-PAS protein,ARNT2 (HIF-2), has been identified and found to share 63% homologywith HIF-1 (29,30). HIF-2 has subsequently been shown to formfunctional heterodimers with AHR, HIF-1, and HIF-2 (30–32).In situ hybridization studies in developing mouse show thatHIF-1 is nearly ubiquitously expressed at E11 through postnatalday 1.5 (P1.5), whereas HIF-2 mRNA is expressed most prominentlyin the brain and kidney at these stages (33). The in situ hybridizationsin this earlier study demonstrated that HIF-2 mRNA is expressedstrongly in the outer cortex of developing kidney, althoughcellular identification is not possible because of the low-magnificationfields presented. Unlike HIF-1 mutants, HIF-2 knockout micesurvive until birth and, with the exception of stunted hypothalamusformation, seem to develop normally (34). Cultured neurons fromHIF-2–/– mice show less induction of the hypoxia-induciblegenes VEGF, Glut3, and Pgk compared with HIF-2+/+ neurons, andmobility shift DNA-binding assays of nuclear extracts from hypoxiccultures of wild-type neurons show both HIF-1 and -2 bindingDNA as a complex with HIF-1 (34). These observations thereforesuggest that HIF-2 is capable of participating in hypoxic responses.
Crosses between HIF-1+/– and HIF-2+/– mutants revealedthat progeny lacking any combination of at least two wild-typealleles of either subunit died by E8.5 (34). These experimentsindicate that HIF-1 and HIF-2 have overlapping roles in earlydevelopment and may compensate for loss of one another. AfterE8.5, however, HIF- subunit functions begin to diverge, andeach protein has a unique, indispensable role in embryonic development,which has yet to be clearly defined. This is consistent withthe restricted expression pattern seen for HIF-2 during mousedevelopment when compared with the ubiquitous expression patternfor HIF-1 (33). We previously immunolocalized HIF-1 proteinexpression in newborn mouse kidney and found that every celltype seemed to contain nuclear expression at that age (35).Although the developing kidney is one of two sites of intenseHIF-2 expression, this protein has not been studied in detailduring metanephrogenesis. Here, we examined HIF-2 protein expressionbeginning at the earliest stages of nephron development andcontinuing into maturity and compared expression patterns withthose of HIF-1. Our findings show that HIF-2 is selectivelyexpressed in distal segments of developing nephrons and becomesrestricted to thick ascending limb (TAL) of loop of Henle. Wealso show that, like HIF-1, HIF-2 heterodimerizes with bothHIF-1 and -2, and that HIF-2 protein expression in organ-culturedmetanephroi is not regulated by oxygen levels.
Western Blots and Immunoprecipitations
Wild-type CD-1 mouse kidneys were dissected at 1 d, 7 d, and6 wk of age and disrupted in a Dounce homogenizer in RIPA buffersupplemented with a protease inhibitor cocktail as describedpreviously (35). After lysates were cleared by centrifugationat 4°C for 25 min at 15,000 x g, total protein concentrationswere determined by the colorimetric Bio-Rad DC Protein Assay(Bio-Rad Laboratories, Hercules, CA). Equal amounts of protein(50 to 100 µg) were separated by 5 to 15% SDS-PAGE andtransferred to nitrocellulose. Standard immunoblot techniqueswere then applied with the following antibodies: polyclonalHIF-1 (Novus-Biologicals, Littleton, CO; 1:100) and polyclonalARNT2 (HIF-2; sc-5581; Santa Cruz Biotechnology, Santa Cruz,CA; 1:50). The ECL reagent and ECL Hyperfilm (Amersham PharmaciaBiotech, Piscataway, NJ) were used to visualize the bands. Forloading controls, membranes were also incubated with mouse monoclonalanti–smooth muscle actin antibody (1:100 dilution; Sigma-Aldrich,St. Louis, MO).
For immunoprecipitation experiments, lysates from 1-d-old kidneyswere additionally cleared by incubation with protein A agarose(1 µl/10 µg total protein) for 15 min and then centrifugedagain at 15,000 x g for 15 min. The clarified lysates then underwentimmunoprecipitation with HIF-2 antibodies, using proceduresdescribed previously (35). The immunoprecipitated proteins wereanalyzed on Western blots by standard protocols with monoclonalanti–HIF-1 and -2 antibodies used at a dilution of 1:100.Purified rabbit IgG (Sigma-Aldrich) was also used as a nonspecificcontrol to demonstrate the specificity of the ARNT2 antibodies,and no bands were detected on Western blots probed with anti–HIF-1or -2.
Embryonic Kidney Organ Culture and Western Blots
Timed-pregnant CD-1 mice were killed and embryos were removedat E12. Metanephroi were dissected and cultured on PET membranecell culture inserts (0.4-µm pore; BD Biosciences, SanJose, CA). Organ culture media and growth conditions were asdescribed previously (35). Culture oxygen concentrations wereset at either constant room air (20% oxygen) or 5% oxygen for5 d or at 20% oxygen for 4 d and then reduced to 2% for thefinal 24 h. After culture periods, kidneys were homogenizedin RIPA buffer, protein concentrations were determined, andWestern blots were performed as described above. In additionto HIF-1 and -2, WT-1 (Santa Cruz Biotechnology; 1:50) and cyclo-oxygenase(Cox-2; Chemicon International, Temecula, CA; 1:200) antibodieswere used to probe membranes. In some cases, explants were fixedand processed for immunohistochemistry as described below.
Peroxidase Immunohistochemistry
For immunohistochemical analysis, kidneys were dissected andfrozen in OCT in isopentane in a dry-ice/acetone bath. Cryostatsections were cut at a thickness of 6 µm and air-dried.Sections were fixed for 10 min in ice-cold methanol, washedin PBS, and placed in 3% H2O2 in methanol for 10 min. Blockingof nonspecific protein interactions was achieved by incubatingsections with 10% goat serum in PBS. Slides were then incubatedin primary antibodies: Polyclonal anti–HIF-1 and HIF-2diluted in PBS (1:100 and 1:50, respectively) for 1 h at roomtemperature in a humidified slide chamber. Sections were thenwashed with PBS, incubated sequentially with biotinylated secondaryantibodies (15 µg/ml) and streptavidin-HRP conjugatesfor 30 min each, and color-developed with 3,3'-diaminobenzidinetetrahydrochloride. Rabbit IgG diluted in PBS at the same concentrationas the primary antibodies served as a negative control.
Immunofluorescence Analysis
Sections (6 µm thick) of 3-d-old mouse kidney were fixedin 100% methanol on ice and incubated with polyclonal anti–HIF-2(1:50) and either fluorescein-conjugated Lotus lectin, to labelproximal convoluted tubule (Sigma Chemical Co., St. Louis, MO;1:200), or rhodamine-conjugated Dolichos Biflorus agglutinin(DBA), to label collecting duct (Vector Laboratories, Burlingame,CA; 1:10), for 30 min at room temperature. After three washesin PBS, Alexa Fluor 488–conjugated anti-rabbit antibody(Molecular Probes, Eugene, OR) was applied for 30 min at 1:200.Sections were again washed in PBS and permanently mounted withthe fluorescence preserving reagent Prolong (Molecular Probes).Serial sections of 6-d-old mouse kidney were fixed in the sameway for co-localization of HIF-2 and Tamm-Horsfall protein (THP).HIF-2 immunofluorescence was performed as described above onthe first serial section, and goat anti-THP antibody (CappelLaboratories, Durham, NC) was applied to the second serial sectionfor 30 min at 1:200. Rabbit anti-goat fluorescein was addedas a secondary antibody to the anti-THP–labeled slides,which were also mounted with Prolong. Irrelevant IgG was incubatedwith sections and treated sequentially with fluorescein andrhodamine-conjugated secondary antibodies for controls.
For identifying hypoxic tissues, pimonidazole hydrochloride(Hypoxyprobe-1; Chemicon) was injected intraperitoneally into6-d-old mice (200 mg/kg), and fixation and labeling were carriedout as before (35). In addition to the Hypoxyprobe-1 monoclonalantibody, sections were incubated with Lotus lectin-fluorescein(1:200) for 30 min at room temperature. Sections from saline-injectedmice were also immunolabeled for Hypoxyprobe-1 and served ascontrols.
HIF- Protein Expression
Total mouse kidney protein was analyzed for HIF-1 and -2 byquantitative Western blots. Three stages of development wereexamined: 1 d, 7 d, and 8 wk of age (Figure 1A). HIF-1 proteinwas readily detectable on blots at each age, with a slight increasein expression at day 7. HIF-2 was strongly expressed in newbornkidney and decreased significantly at 7 d of age. In 8-wk-oldkidney, HIF-2 protein was barely detectable by Western blot.
Figure 1. (A) Total renal hypoxia-inducible factor-2 (HIF-2) protein expression decreases with age. Quantitative Western blots for HIF-1 and -2 with kidney lysates from newborn (Nb), 7-d-old, and 8-wk-old mouse kidneys revealed sustained expression of HIF-1 throughout kidney maturation, with a peak at 7 d of age. By contrast, HIF-2 expression was most intense in newborn kidney, was significantly decreased by day 7, and was barely detectable at 8 wk. Anti–smooth muscle actin staining of blots serves as loading controls. (B) HIF-2 heterodimerizes with both HIF-1 and -2. Proteins immunoprecipitated by HIF-2 IgG from 3-d-old mouse kidney were analyzed by Western blot with HIF-1– and -2–specific antibodies. Both subunits were evident by Western blot, demonstrating that HIF-2 forms a complex with both subunits. NS, nonspecific bands detected in both cases by secondary antibody.
To determine whether HIF-2 heterodimerizes with either HIF-1or HIF-2, we conducted immunoprecipitation and Western blotexperiments of newborn kidney lysates. Both HIF-1 and -2 proteinswere observed on Western blots of the HIF- immune complexes,revealing that HIF-1 formed heterodimers with both HIF-1 and-2 in developing kidney (Figure 1B).
Effect of Hypoxia on HIF- Protein Expression in Kidney Organ Culture
E12 kidney explants were cultured for 4 d at 20% oxygen followedby a 24-h exposure to 2% oxygen (Figure 2, left) or for 5 dat 5% oxygen (Figure 2, right), followed by Western blots forHIF1- and -2. In cultures that were exposed to acute or chronichypoxia, no large changes in either HIF-1 or -2 protein expressionwere evident. However, the HIF-inducible genes WT-1 and Cox-2both were induced by hypoxia in organ culture, but WT-1 increasedonly modestly at 5% oxygen (Figure 2).
Figure 2. HIF- and HIF target gene expression in embryonic day 12 (E12) organ cultures that were exposed to hypoxia. E12 mouse metanephroi were dissected and cultured for 4 d in 20% oxygen and then shifted to a hypoxic environment (2% oxygen) for 1 d (left) or were cultured for 5 d in 5% oxygen (right). In both cases, HIF-1 and -2 protein levels remained constant. WT-1 and cyclo-oxengenase-2 (Cox-2), two potential HIF target genes during kidney development, however, showed increased protein levels in response to hypoxia in both cases, although upregulation of WT-1 at 5% oxygen seemed more modest. All blots are representative of two independent experiments.
HIF-1 and -2 Immunolocalization
HIF-1 protein was found to be widely distributed in nuclei ofapparently all cells in both developing and maturing kidneyat every stage examined. In E14 kidney, HIF-1 was expressedubiquitously by mesenchymal cells, ureteric bud epithelium,and all cells in developing nephrons (Figure 3A). During laterstages of kidney development and into maturity, widespread HIF-1protein expression persisted; nuclei within all tubular segments,interstitial cells, and cells within glomeruli all stronglylabeled with anti–HIF-1 antibody (see A in Figures 4 through 7). Sections that were labeled with nonspecific rabbit IgG inplace of HIF-1 antibodies were negative in all cases (Figure 3D).
Figure 3. Immunolocalization of HIF-1 and -2 in E14 kidney. (A) HIF-1 protein expression at E14 was nearly ubiquitous and uniformly intense. Strong nuclear staining was apparent in ureteric bud (UB), developing nephrons (N), and mesenchymal cells. (B) HIF-2 protein expression was much more restricted and variable at E14 compared with HIF-1. Faint nuclear labeling was found in ureteric bud, uninduced mesenchyme in extreme outer cortex, but was particularly prominent in developing nephrons. (C) Higher-power views show that HIF-2 protein was expressed strongly and specifically in the distal segment of developing nephrons (*), including cells that may form the macula densa. Weaker expression was also evident in the visceral (VE) and parietal epithelium (PE), which are destined to form the podocytes (Po) and Bowman’s capsule (BC), respectively (arrows). (D) Nonspecific rabbit IgG incubated with sections in place of primary antibodies served as controls, and no labeling was observed.
Figure 4. Sections of E12 metanephroi maintained for 5 d at room air (20% oxygen; A and C) or 5% oxygen (B and D) labeled for HIF-1 (A and B) and HIF-2 (C and D). Hypoxia did not induce changes in either HIF-1 or -2 abundance or distribution. Note that a nephron segment that contained intense signal for HIF-2 (arrows) seemed similar to that observed in native, E14 kidney (Figure 3).
Figure 5. Expression of HIF-1 and -2 protein in 3-d-old (P3) mouse kidney. (A) HIF-1 protein distribution was widespread at 3 d of age, and most cells showed strong nuclear staining. (B) HIF-2, conversely, remained tightly restricted at day 3. Developing nephrons (N) again showed especially strong nuclear labeling. Portions of a specific developing tubule segment projecting toward the medulla (arrows) also displayed strong nuclear reaction product.
Figure 6. HIF-1 and -2 protein distribution in 7-d-old (P7) mouse kidney. (A) HIF-1 protein localization at day 7 again showed extensive nuclear staining throughout the kidney, including glomeruli (G). (B) At day 7, HIF-2 distribution was restricted to a specific tubular segment. In addition, this was the first age point examined at which specific glomerular labeling was observed. The inset shows podocyte labeling for HIF-2 (P).
Figure 7. HIF-1 and -2 protein expression in 10-d-old mouse kidney. (A) Similar to all other age points, HIF-1 protein distribution at day 10 was nearly ubiquitous. Glomeruli (G) again were consistently positive for expression. (B) HIF-2 expression at day 10 showed relatively little protein in the outer cortex, and only a subset of tubular segments showed labeling. In contrast to day 7, glomeruli now showed little or no staining for HIF-2. (C) Certain medullary tubules of 10-d-old kidney showed intense labeling for HIF-2.
In contrast to HIF-1, the distribution of HIF-2 protein washighly restricted among the time points examined. At E14, HIF-2was seen in nuclei of ureteric bud epithelium and less intenselyin mesenchymal cells of the outer cortex (Figure 3B). In earlynephrons, both the visceral and parietal epithelium were weaklypositive, but developing distal segments showed particularlyintense HIF-2 labeling (Figure 3, B and C).
As an additional test for whether HIF-1 and/or -2 changes inabundance or expression pattern, organ cultures of E12 kidneysmaintained for 5 d at 20 or 5% oxygen were examined. As shownin Figure 4, no differences were observed.
In 3-d-old kidney, developing nephrons of the extreme outercortex showed the same overall expression pattern observed inE14 kidney with especially strong localization to distal segments(Figure 5B). In addition, tubules projecting through the subcorticalregion also showed strong nuclear labeling at this age. Theexpression profile in 5-d-old kidney was indistinguishable fromthat in 3-d-old kidney (data not shown). In 1-wk-old kidney,HIF-2 protein distribution was restricted to certain tubularsegments (Figure 6B), and podocyte nuclei of maturing stageglomeruli were also labeled (Figure 6B, inset). By 10 d of age,restricted tubular expression persisted in the cortex, but glomerularexpression was now absent (Figure 7B). The medulla of 10-d-oldkidney showed intense HIF-2 expression (Figure 7C). Finally,in fully mature, 8-wk-old kidney, HIF-2 expression was diminishedoverall and found only sparsely in the cortex (Figure 8B). Medullaryexpression of HIF-2 was prominent (Figure 8B, inset upper right),but glomeruli showed little expression (Figure 8, inset lowerright).
Figure 8. Immunolocalization of HIF-1 and -2 in 8-wk-old mouse kidney. (A) HIF-1 protein expression persisted in all tubular segments and glomeruli (G), as seen throughout development. (B) HIF-2 protein expression in the cortex was limited, and few positive tubular cells could be found (arrows). Glomeruli (G of inset) were negative. The medulla was the site of the most intensive HIF-2 labeling at this stage (inset top left).
Identification of Tubular Segment Expressing HIF-2
For identifying specifically which tubule segment expressedHIF-2, co-localization studies with the tubule-specific markersLotus lectin (specific for proximal convoluted tubule), DBA(specific for collecting duct), and THP (specific for TAL ofHenle’s loop) were undertaken on sections of 6-d-old mousekidney. In serial sections that were merged digitally, co-localizationwith THP showed that HIF-2 protein was expressed in developingTAL of loop of Henle (Figure 9, A through C). Expression wasrestricted to TAL as no co-localization of HIF-2 was apparentin the same section that dually labeled with proximal tubularmarker Lotus lectin (Figure 9, D through F) or collecting duct–specificDBA (Figure 9, G through I). For controls, slides were incubatedwith nonspecific IgG and both fluorescein- and rhodamine-conjugatedsecondary antibodies (Figure 9J).
Figure 9. Identification of HIF-2–positive tubular segment: Double labeling with tubule-specific markers. (A through C) Tamm-Horsfall’s protein (THP) is expressed intensely in the thick ascending limb (TAL) of Henle’s loop. Serial cryostat sections from 6-d-old kidney were labeled for THP (A) and HIF-2 (B), and the resulting images were overlaid (C). THP and HIF-2 expression patterns overlap in the majority of cells, demonstrating that TAL is the primary site of HIF-2 protein expression. (D through F) The Lotus lectin specifically binds carbohydrates found on proximal tubule epithelial membranes. Cryostat sections from 3-d-old mouse kidney double labeled for fluorescein-conjugated Lotus lectin (D) and HIF-2 showed no overlapping fluorescence, demonstrating that HIF-2 is not expressed by proximal tubules. (G through I) Dolichos Biflorus agglutinin (DBA) binds specifically to collecting duct epithelium. Double labeling of DBA (G) and HIF-2 (H) on the same 3-d-old kidney section revealed no co-localization (I), demonstrating that collecting duct epithelial cells are not sites of HIF-2 expression. (J) Controls with nonspecific IgG treated sequentially with fluorescein- and rhodamine-conjugated secondary antibodies were negative for each set of experiments.
Identification of Hypoxic Nephron Segments
To identify nephron segments that contained hypoxic cells, wedoubly labeled sections with Hypoxoyprobe-1 and tubule-specificlectins. Nearly all proximal tubules (Lotus lectin positive)in 6-d-old kidney were positive for Hypoxyprobe-1 labeling aswell (Figure 10). Because HIF-2 expression was excluded fromproximal tubules (Figure 9F), it is thereby also excluded fromthe most severely hypoxic cells in developing kidney. In contrastto proximal tubules, medullary tissues of 6-d-old kidney werenot labeled with Hypoxyprobe-1 (or Lotus lectin; Figure 11)and therefore were not severely hypoxic at this time.
Figure 10. Identification of hypoxic cells in vivo. Using the hypoxia marker Hypoxyprobe-1, we identified severely hypoxic cells in 6-d-old mouse kidney. By double labeling with the Hypoxyprobe-1 (A) and Lotus lectin, a proximal convoluted tubule marker (B), we saw that proximal tubules were the only extremely hypoxic cells in 6-d-old kidney (merged image C). Some proximal tubules were not positive for Hypoxyprobe-1 (arrow). Glomeruli did not show Hypoxyprobe-1–specific fluorescence. (D) Control, saline-injected mice that were treated with Hypoxyprobe-1 antibodies showed no fluorescence.
Figure 11. Section of medulla from 6-d-old mouse kidney showing only low levels of nonspecific background labeling for Hypoxyprobe-1 (A) or Lotus lectin (B). Images are merged in C.
During nephrogenesis, renal mesenchymal cells and ureteric budepithelium differentiate into numerous, distinct cell typesin a spatially precise sequence. Transcriptional regulationof this intricate process is poorly understood, but proteinsthat belong to the bHLH-PAS family of transcription factorsare highly expressed during kidney development, implicatingthem as candidate regulators of nephron formation (33,35,36).One subunit of these transcriptional complexes, HIF-1 (ARNT1),is expressed ubiquitously in newborn mouse kidney (35). A highlyrelated subunit, HIF-2 (ARNT2), is also expressed at very highlevels in developing kidney and brain (33,36), but the expressionpatterns of HIF-2 had not yet been studied extensively. Ouraims here were to define clearly the expression patterns ofHIF-2 in kidney development and compare directly its distributionwith HIF-1. We found that HIF-1 protein was expressed ubiquitouslyat every age examined, whereas HIF-2 expression was tightlyrestricted. We also sought to determine with which HIF- proteinsHIF-2 heterodimerized and found both HIF-1 and -2 present incomplexes immunoprecipitated with anti–HIF-2 antibodies.Despite that HIF-2 heterodimerized with both HIF- subunits,HIF-2 protein was not found in the most extremely hypoxic cells,which we determined to be proximal tubule epithelium. By contrast,HIF-2 was expressed predominantly in distal segments of developingnephrons and became restricted to TAL of Henle’s loop.In addition, we found that neither HIF- subunit was prominentlyinduced by hypoxia in metanephric organ cultures but that twopotential HIF target genes that are important for kidney development,WT-1 and Cox-2, were upregulated in low oxygen tension.
Northern blot analysis and in situ hybridization experimentshave demonstrated intense mRNA expression of HIF-1 and -2 indeveloping kidney (36). These experiments, however, did notexamine expression beyond P1.5 or specifically address proteinexpression. By quantitative Western blot, we showed abundantHIF-1 protein expression in kidneys at birth, 7 d, and 8 wkof age. Our data are consistent with the previous in situ hybridizationexperiments (36) in that our Western blots showed very intenseexpression of HIF-2 in newborn kidney. We extended the analysisto include maturing kidneys and showed that HIF-2 protein expressiondeclined significantly at P7 and was barely detectable on Westernblots at 8 wk of age. The relative abundance of HIF-2 in developingkidney and decline in maturation therefore suggest a role forthis protein in renal organogenesis.
HIF-1 and -2 are capable of binding a number of bHLH-PAS domainproteins, including HIF-1, HIF-2, and AHR (30–32). Ofparticular interest to kidney development are the heterodimersconsisting of HIF-1 and either HIF- subunit, because many ofthe genes that are known to be induced by these complexes arecrucial for kidney formation, including VEGF, VEGFR-1 and -2,WT-1, angiopoietin-2, Tie-2, and Epo. We showed previously,by immunoprecipitation and Western blotting, that HIF-1 formsheterodimers with both HIF-1 and -2 in developing kidney (35).Here we show that HIF-2, in the same way, complexes with either subunit in 3-d-old mouse kidney. Which genes are targeted byHIF-2–containing complexes has yet to be clearly defined.However, cultured neurons from HIF-2–/– mice showless induction of VEGF in response to hypoxia than HIF-2+/+neurons, suggesting that VEGF may be a target gene for heterodimersthat contain HIF-2 (34).
By applying immunoperoxidase techniques, we compared HIF-1 and-2 expression patterns at various stages of kidney development.At E14, HIF-2 protein was expressed specifically in the uretericbud and at particularly high levels in distal segments of developingnephrons. Visceral and parietal epithelia, which eventuallyform the podocytes and Bowman’s capsule, respectively,also showed positive staining, although at weaker levels. Mesenchymalcells of the outer cortex of E14 kidneys were also weakly positive.At 3 d of age, the expression pattern in the early nephric figuresof the extreme outer cortex persisted. In addition, a specificelongating tubular segment showed prominent nuclear labeling.This segment proved to be developing TAL of Henle’s loop,by co-distribution immunofluorescence analysis with THP on serialsections (37). There was a dramatic change in the expressionpattern observed at day 7, where, in addition to TAL, glomerularpodocytes were now positive for HIF-2. In mature, 8-wk-old kidneys,expression was once again restricted to TAL, with no apparentglomerular labeling. In contrast to HIF-2, strong nuclear HIF-1labeling was observed at every stage examined and in nearlyevery cell. These immunohistochemical expression patterns areconsistent with the quantitative Western blots. HIF-1 expressionwas widespread and abundant at all ages examined, whereas HIF-2was highly expressed in developing kidney in many cell typesbut with age became restricted to TAL, and, quantitatively,lower levels were detected on immunoblots.
We have used the hypoxia marker Hypoxyprobe-1 and immunofluorescencemicroscopy to identify extremely hypoxic cells within newbornmouse kidney (35). Here, we show in 6-d-old kidney that thevast majority of hypoxic cells are Lotus lectin positive andtherefore correspond to proximal tubular epithelium. Surprising,even though HIF-2 co-immunoprecipitated the hypoxia responsiveproteins HIF-1 and -2, HIF-2 did not immunolocalize to proximaltubules. This suggests that HIF-2 may not participate in thehypoxic response pathway in the most intensely hypoxic cells.However, Hypoxyprobe-1 is optimally reactive only in extremelyhypoxic cells (<1% O2) (38). We therefore cannot rule outthat mild hypoxia (2 to 18% O2) may occur in glomeruli or TALand that HIF-2 could mediate hypoxia-induced gene expressionin these cells.
Whether HIF-1 gene expression is increased in hypoxia or notis unclear. In most systems analyzed, HIF-1 protein levels remainconstant regardless of oxygen tension, whereas in certain celltypes, such as human neuroblastoma cells, HIF-1 protein accumulatesunder hypoxic stress (25,39). We show here in E12 metanephricorgan cultures that oxygen tension had no effect on either HIF-1or -2 protein levels or their expression patterns. This evidencesuggests that gene transactivation by HIF in developing kidneymay be totally reliant on subunit stabilization.
As described above, the VEGF gene is a potential target forHIF-2–containing heterodimers. What other genes relevantto kidney development might be induced by HIF-2 transcriptionfactors? A recent study identified a HRE in the WT-1 gene promoterand shown that HIF-1 activates transcription of this gene inhypoxia (23). This study showed by DNA-binding electrophoreticmobility shift assays that the transcription factor that inducedthe HRE in the WT-1 promoter contained HIF-1, but the subunitwas not identified. In early kidney development, WT-1 is expressedin renal mesenchyme and developing nephrons and then becomeslocalized specifically to podocytes (40). Our studies presentedhere show that HIF-2 was expressed in renal mesenchymal cellsand early nephrons as well as in podocytes at day 7. We thereforesuggest that the HIF complex driving WT-1 expression may includeeither HIF-1 or -2. Cox-2 displays an expression pattern remarkablysimilar to that of HIF-2 in mouse kidney. Cox-2 is expressedexclusively in macula densa cells and TAL (41). Our studiesalso demonstrated that Cox-2 expression was induced in hypoxickidney organ cultures, suggesting that Cox-2 might be an additionaltarget of HIF-2, but this speculation demands further inquiry.
Studies of neuronal cell lines overexpressing HIF-2 with SIM-1,which is highly expressed in developing kidney, may give moreinsight into which genes HIF-2 specifically targets (42,43).These earlier reports showed that Neuro-2a cells overexpressingHIF-2 showed strong induction of laminin 2, which is a componentof the laminin-2 trimer (211). This laminin isoform is expressedat low levels by Henle’s loop, distal tubule, Bowman’scapsule, and collecting duct epithelium (44). We observed HIF-2expression in primitive nephrons in cells that eventually formdistal tubule, Bowman’s capsule (parietal epithelium),and collecting duct (ureteric bud) as well as Henle’sloop. Perhaps HIF containing HIF-2 is driving laminin 2 expressionin these cells. Another potential HIF-2 target identified inNeuro-2a cells is the janus kinase 2, which is partially responsiblefor expression of cytokine-stimulated inducible nitric oxidesynthase in kidney epithelial cells (45). Again, these putativeHIF-2 targets in developing kidney need further experimentalconfirmation.
We have hypothesized previously that a delicate balance of HIFprotein expression and stabilization occurs in various celltypes during nephrogenesis so that different HIF heterodimerspredominate in certain cell types and lead to selective transcriptionof different HIF target genes (46). The current study supportsthis hypothesis by showing differential expression patternsfor HIF-1 and -2 during kidney development: HIF-1 distributionwas widespread, but HIF-2 was most abundant initially in distalnephric segments and then became restricted to TAL. The cellularavailability of subunits and competition for binding HIF- maydetermine which HIF target genes will be expressed and thereforehas a vital role in determining cellular phenotype. The events(e.g., microenvironmental oxygen tension) that prompt formationof a particular HIF heterodimer and the precise temporal andspatial occurrence of these events need to be investigated thoroughlyin developing kidney to gain a better understanding of the HIF-mediatedtranscriptional regulation.
Acknowledgments
This study was funded by National Institutes of Health GrantsDK052483 and DK065123.
We thank Kathryn Isom, Eileen Roach, and Pat St. John for technicalhelp.
Crews ST: Control of cell lineage-specific development and transcription by bHLH-PAS proteins. Genes Dev 12: 607–620, 1998[Free Full Text]
Hoffman EC, Reyes H, Chu F, Sander F, Conley LH, Brooks BA, Hankinson O: Cloning of a subunit of the DNA binding form of the Ah (dioxin) receptor. Science 252: 954–958, 1991[Abstract/Free Full Text]
Wang GL, Jiang BH, Rue EA, Semenza GL: Hypoxia-inducible factor 1 is a basic-helix-loop-helix-PAS heterodimer regulated by cellular O2 tension. Proc Natl Acad Sci U S A 92: 5510–5514, 1995[Abstract/Free Full Text]
Jiang BH, Rue EA, Wang GL, Semenza GL: Dimerization, DNA binding, and transactivation properties of hypoxia-inducible factor 1. J Biol Chem 271: 17771–17778, 1996[Abstract/Free Full Text]
Wang GL, Semenza GL: Purification and characterization of HIF-1. J Biol Chem 270: 1230–1237, 1995[Abstract/Free Full Text]
Denison MS, Whitlock JP Jr: Xenobiotic-inducible transcription of cytochrome P450 genes. J Biol Chem 270: 18175–18178, 1995[Free Full Text]
Ma Q, Whitlock JP Jr: A novel cytoplasmic protein that interacts with the Ah receptor, contains tetratricopeptide repeat motifs, and augments the transcriptional response to 2, 3, 7, 8-Tetrachlorodibenzo-p-dioxin. J Biol Chem 272: 8878–8884, 1997[Abstract/Free Full Text]
Semenza GL, Nejfelt MK, Chi SM, Antonarakis SE: Hypoxia-inducible nuclear factors bind to an enhancer element located 3' to the human erythropoietin gene. Proc Natl Acad Sci U S A 88: 5680–5684, 1991[Abstract/Free Full Text]
Semenza GL: HIF-1, O2, and the 3 PHDs: How animal cells signal hypoxia to the nucleus. Cell 107: 1–3, 2001[CrossRef][Medline]
Jiang GH, Semenza GL, Bauer C, Marti HH: HIF-1 levels vary exponentially over a physiologically relevant range of oxygen tension. Am J Physiol 271: C1172–C1180, 1996
Ivan M, Kondo K, Yang H, Kim W, Valiando J, Ohh M, Salic A, Asara JM, Lane WS, Kaelin WG: HIF targeted for VHL-mediated destruction by proline hydroxylation: Implications for O2 sensing. Science 292: 464–468, 2001[Abstract/Free Full Text]
Jaakkola P, Mole DR, Tian YM, Wilson MI, Gielber J, Gaskell SJ, von Kriegsheim A, Hebetreit HF, Mukherji M, Schofield CJ, Maxwell PH, Pugh CW, Ratcliffe PJ: Targeting of HIF-1 to the von Hipppel-Lindau ubiquitylation complex by O2-regulated prolyl hydroxylation. Science 292: 468–472, 2001[Abstract/Free Full Text]
Tian H, Hammer RE, Matsumoto AM, Russell DW, McKnight SL: The hypoxia-responsive transcription factor EPAS1 is essential for catecholamine homeostasis and protection against heart failure during embryonic development. Genes Dev 12: 3320–3324, 1998[Abstract/Free Full Text]
Peng J, Zhang L, Drysdale L Fong GH: The transcription factor EPAS1/hypoxia-inducible factor 2 plays an important role in vascular remodeling. Proc Natl Acad Sci U S A 97: 8386–8391, 2000[Abstract/Free Full Text]
O’Rourke JF, Tian YM, Ratcliffe PJ, Pugh CW: Oxygen-regulated and transactivating domains of endothelial PAS protein 1: Comparison with HIF-1. J Biol Chem 274: 2060–2071, 1999[Abstract/Free Full Text]
Levy AP, Levy NS, Wegner S, Goldberg MA: Transcriptional regulation of the rat vascular endothelial growth factor gene by hypoxia. J Biol Chem 270: 13333–13340, 1995[Abstract/Free Full Text]
Liu Y, Cox SR, Morita T, Kourenbanas S: Hypoxia regulates VEGF expression in endothelial cells: Identification of a 5' enhancer. Circ Res 77: 638–643, 1995[Abstract/Free Full Text]
Kappel A, Ronicke V, Damert A, Flamme I, Risau W, Breir G: Identification of vascular endothelial growth factor (VEGF) receptor-2 (Flk-1) promoter/enhancer sequences sufficient for angioblast and endothelial cell-specific transcription in transgenic mice. Blood 93: 4284–4292, 1999[Abstract/Free Full Text]
Gerber HP, Condorelli F, Park J, Ferrara N: Differential transcriptional regulation of the two vascular endothelial growth factor receptor genes: Flt-1, but not Flk-1/KDR, is upregulated in hypoxia. J Biol Chem 272: 23659–23667, 1997[Abstract/Free Full Text]
Maisonpierre PC, Suri C, Jones PF, Bartunkova S, Weigand SJ, Radziejeuski C, Compton D, McClain J, Aldrich TH, Papadopoulos N, Daly TJ, Davis S, Sato TN, Yancopoulos GD: Angiopoietin-2, a natural antagonist for Tie2 that disrupts in vivo angiogenesis. Science 277: 55–60, 1997[Abstract/Free Full Text]
Oh H, Takagi H, Suzuma K, Otani A, Matsumura M, Hinda Y: Hypoxia and vascular endothelial growth factor selectively upregulate angiopoietin-2 in bovine microvascular endothelial cells. J Biol Chem 274: 15732–15739, 1999[Abstract/Free Full Text]
Wagner KD, Wagner N, Wellmann S, Scheley G, Bondke A, Theres H, Scholz H: Oxygen-regulated expression of the Wilms’ tumor suppressor Wt1 involves hypoxia-inducible factor-1 (HIF-1). FASEB J 17: 1364–1366, 2003[Abstract/Free Full Text]
Korkalainen M, Tuomisto J, Pohjanvirta R: Identification of novel splice variants of ARNT and ARNT2 in the rat. Biochem Biophys Res Commun 303: 1095–1100, 2003[CrossRef][Medline]
Stolze I, Berchner-Pfannschmidt U, Freitag P, Wotzlaw C, Rossler J, Frede S, Acker H, Fandrey J: Hypoxia-inducible erythropoietin gene expression in human neuroblastoma cells. Blood 100: 2623–2628, 2002[Abstract/Free Full Text]
Iyer NV, Kotch LE, Agani F, Leung SW, Laughner E, Wenger RH, Gassman M, Gearhart JD, Lawler AM, Yu AY, Semenza GL: Cellular and developmental control of O2 homeostasis by hypoxia-inducible factor 1. Genes Dev 12: 149–162, 1998[Abstract/Free Full Text]
Ryan HE, Lo J, Johnson RS: HIF-1 is required for solid tumor formation and embryonic vascularization. EMBO J 17: 3003–3015, 1998
Maltepe E, Schmidt JV, Baunoch D, Bradford CA, Simon MC: Abnormal angiogenesis and responses to glucose deprivation in mice lacking the protein ARNT. Nature 386: 403–407, 1997[CrossRef][Medline]
Drutel G, Kathmann M, Heron A, Schwartz JC, Arrang JM: Cloning and selective expression in brain and kidney of ARNT2 homologous to the Ah receptor nuclear translocator (ARNT). Biochem Biophys Res Commun 225: 333–339, 1996[CrossRef][Medline]
Hirose K, Morita M, Ema M, Mimura J, Hamada H, Fujii H, Saijo Y, Gotoh O, Sogawa K, Fujii-Kuriyama Y: cDNA cloning and tissue specific expression of a novel basic-helix-loop-helix/PAS factor (ARNT2) with close sequence similarity to the aryl hydrocarbon receptor nuclear translocator (ARNT). Mol Cell Biol 16: 1706–1713, 1996[Abstract]
Maltepe E, Keith B, Arsham A, Brorson J, Simon M: The role of ARNT2 in tumor angiogenesis and the neural response to hypoxia. Biochem Biophys Res Commun 273: 231–238, 2000[CrossRef][Medline]
Powell WH, Hahn ME: Identification and functional characterization of hypoxia-inducible factor 2 from the estuarine teleost, Fundulus heteroclitus: Interaction of HIF-2 with two ARNT2 splice variants. J Exp Zool 294: 17–29, 2002[CrossRef][Medline]
Aitola MH, Pelto-Huikko MT: Expression of Arnt and Arnt2 mRNA in developing murine tissues. J Histochem Cytochem 51: 41–54, 2003[Abstract/Free Full Text]
Keith B, Adelman DM, Simon MC: Targeted mutation of the murine arylhydrocarbon receptor nuclear translocator 2 (ARNT2) gene reveals partial redundancy with ARNT.Proc Natl Acad Sci U S A 98: 6692–6697, 2001[Abstract/Free Full Text]
Freeburg PB, Robert B, St. John PL, Abrahamson DR: Podocyte expression of hypoxia-inducible factor (HIF)-1 and HIF-2 during glomerular development. J Am Soc Nephrol 14: 927–938, 2003[Abstract/Free Full Text]
Jain S, Maltepe E, Lu MM, Simon C, Bradfield CA: Expression of ARNT, ARNT2, HIF1, HIF-2, and Ah receptor mRNAs in the developing mouse. Mech Dev 73: 117–123, 1998[CrossRef][Medline]
Hoyer JR, Resnick JS, Michael AD, Vernier RL: Ontogeny of Tamm-Horsfall urinary glycoprotein. Lab Invest 30: 757–761, 1974[Medline]
Arteel GE, Thurman RG, Yates JM, Raleigh JA: Evidence that hypoxia markers detect oxygen gradients in liver: Pimonidazole and retrograde perfusion of rat liver. Br J Cancer 72: 889–895, 1995[Medline]
Catron T, Mendiola MA, Smith SM, Born J, Walker MK: Hypoxia regulates avian cardiac Arnt and HIF-1 mRNA expression. Biochem Biophys Res Commun 225: 602–607, 2001
Armstrong JF, Pritchard-Jones K, Bickmore WA, Hastie ND, Bard JB: The expression of the Wilms’ tumour gene, WT1, in the developing mammalian embryo. Mech Dev 40: 85–97, 1993[CrossRef][Medline]
Liu C, Goshu E, Wells A, Ran CM: Identification of downstream targets of SIM1 and ARNT2, a pair of transcription factors essential for neuroendocrine cell differentiation. J Biol Chem 278: 44857–44867, 2003[Abstract/Free Full Text]
Fan CM, Kuwana E, Bulfone A, Fletcher CF, Copeland NG, Jenkins NA, Crews S, Martinez S, Puelles L, Rubenstein JL, Tessier-Lavigne M: Expression of the two murine homologs of Drosophila single-minded suggests possible roles in embryonic patterning and in the pathogenesis of Down syndrome. Mol Cell Neurosci 7: 1–16, 1996[CrossRef][Medline]
Poljakovic M, Nygren JM, Persson K: Signalling pathways regulating inducible nitric oxide synthase expression in human kidney epithelial cells. Eur J Pharmacol 469: 21–28, 2003[CrossRef][Medline]
Freeburg PB, Abrahamson DR: Hypoxia-inducible factors and kidney vascular development. J Am Soc Nephrol 14: 2723–2730, 2003[Abstract/Free Full Text]
Received for publication April 14, 2004.
Accepted for publication July 15, 2004.
This article has been cited by other articles:
R. L. Chevalier and M. S. Forbes Generation and Evolution of Atubular Glomeruli in the Progression of Renal Disorders
J. Am. Soc. Nephrol.,
February 1, 2008;
19(2):
197 - 206.
[Abstract][Full Text][PDF]
K. Shchors, E. Shchors, F. Rostker, E. R. Lawlor, L. Brown-Swigart, and G. I. Evan The Myc-dependent angiogenic switch in tumors is mediated by interleukin 1beta
Genes & Dev.,
September 15, 2006;
20(18):
2527 - 2538.
[Abstract][Full Text][PDF]
E. B. Rankin, J. E. Tomaszewski, and V. H. Haase Renal cyst development in mice with conditional inactivation of the von hippel-lindau tumor suppressor.
Cancer Res.,
March 1, 2006;
66(5):
2576 - 2583.
[Abstract][Full Text][PDF]
and Aryl Hydrocarbon Receptor Nuclear Transporter-2 in Developing Kidney
Top
Abstract
Introduction
/
heterodimeric transcription factors of the basic helix-loop-helix-Per-Arnt-Sim (bHLH-PAS) superfamily and are chiefly responsible for cellular adaptation to oxygen deprivation. HIF function relies on the stabilization of the 
20% oxygen) or 5% oxygen for 5 d or at 20% oxygen for 4 d and then reduced to 2% for the final 24 h. After culture periods, kidneys were homogenized in RIPA buffer, protein concentrations were determined, and Western blots were performed as described above. In addition to HIF-1
1). This laminin isoform is expressed at low levels by Henle’s loop, distal tubule, Bowman’s capsule, and collecting duct epithelium (44). We observed HIF-2
