Skip to main content
NCBI home page
Proceedings of the National Academy of Sciences of the United States of America logoLink to Proceedings of the National Academy of Sciences of the United States of America
. 2005 Jul 11;102(29):10212–10217. doi: 10.1073/pnas.0504520102

Activation of hypoxia-inducible factors in hyperoxia through prolyl 4-hydroxylase blockade in cells and explants of primate lung

Tiina M Asikainen *,†, Barbara K Schneider *, Nahid S Waleh ‡, Ronald I Clyman ‡, Wen-Bin Ho §, Lee A Flippin §, Volkmar Günzler §,†, Carl W White *,†,
PMCID: PMC1177409  PMID: 16009933

Abstract

Preterm neonates with respiratory distress syndrome (RDS) often develop a chronic form of lung disease called bronchopulmonary dysplasia (BPD), characterized by decreased alveolar and vascular development. Ventilator treatment with supraphysiological O2 concentrations (hyperoxia) contribute to the development of BPD. Hyperoxia down-regulates and hypoxia up-regulates many angiogenic factors in the developing lung. We investigated whether angiogenic responses could be augmented through enhancement of hypoxia-inducible factors 1α and 2α (HIF-1α and -2α, respectively) via blockade of prolyl hydroxylase domain-containing proteins (HIF-PHDs) in human microvascular endothelial cells from developing and adult lung, in epithelial A549 cells, and in fetal baboon explants in relative or absolute hyperoxia. PHD inhibitor (FG-4095) and positive control dimethyloxaloylglycine (DMOG), selective and nonselective HIF-PHD inhibitors, respectively, enhanced HIF-1α and -2α, vascular endothelial growth factor (VEGF), and platelet-endothelial cell adhesion molecule 1 expression in vitro in 95% and 21% O2. Furthermore, VEGF receptor fms-like tyrosine kinase 1 (Flt-1) was elevated, whereas kinase insert domain-containing receptor/fetal liver kinase 1 (KDR) was diminished in endothelial, but not epithelial, cells. Intracellular Flt-1 and KDR locations were unchanged by PHD blockade. Like VEGF, FG-4095 and DMOG increased angiogenesis in vitro, both in 95% and 21% O2, an effect that could be blocked through either Flt-1 or KDR. Notably, FG-4095 was effective in stimulating HIFs and VEGF also in fetal baboon lung explants. FG-4095 or DMOG treatment appeared to stimulate the feedback loop promoting HIF degradation in that PHD-2 and/or -3, but not PHD-1, were enhanced. Through actions characterized above, FG-4095 could have desirable effects in enhancing lung growth in BPD.

Keywords: angiogenesis, prematurity, bronchopulmonary dysplasia, alveolization

Hypoxia-inducible factors 1α and 2α (HIF-1α and -2α, respectively) are transcription factors principally responsible for hypoxia response. Both α-subunits have basic helix–loop–helix and Per-Arnt-Sim domains and dimerize with HIF-1β (1). The active HIF complex then binds to hypoxia response elements of target genes enhancing their transcription. HIF-1α and -2α share great structural similarities in their DNA-binding and dimerization domains but differ in their transactivation domains, suggesting they may also diverge in the selection of their target genes. Indeed, HIF-1α and -2α have distinct actions depending on cell lineage (2).

HIF degradation pathways have been identified previously (3, 4). The stability of HIF-1α is regulated principally through hydroxylation of proline residues by prolyl hydroxylase domain-containing proteins (PHDs) 1, 2, and 3 (mammalian homologues of EGLN -2, -1, and -3, respectively) (5, 6). Unlike other mammalian prolyl hydroxylases, the HIF-PHDs have negligible activity against procollagen (5–7). The HIF-PHDs are strictly O2 dependent and also require α-ketoglutarate, iron, and ascorbate for activity (5–7). In addition to PHDs, HIF-1α is regulated through hydroxylation of an asparagine residue by factor inhibiting HIF (8) and acetylation of a lysine residue by acetyl transferase 1 (9). After these modifications, the product of VHL tumor suppressor gene recognizes and binds HIF-1α resulting in recruitment of an E3 ubiquitin–protein ligase complex, causing ubiquitination and proteosomal destruction of HIF-1α (10).

Hypoxia is an imbalance between delivery and demand of O2 in tissues, i.e., a pathological state, such as in stroke, infarction, or cancer. However, during fetal development, hypoxia (3–5% O2) is physiological. Furthermore, hypoxia enhances vascular growth and fetal lung branching morphogenesis (11, 12). Transition from fetal to ambient (21%) O2 concentration creates relative oxidative stress for any newborn, but especially preterm neonates whose lung development is interrupted. Although 100% O2 is no longer routinely used to treat respiratory distress syndrome (RDS), inspired O2 concentrations usually are well above 21%. With lower O2 concentrations, the pathological features of bronchopulmonary dysplasia (BPD, also called chronic lung disease of the newborn) also have changed from severe fibrosis to a form of pulmonary hypoplasia with reduced numbers of small blood vessels and alveoli (13). Interestingly, blocking of vascular endothelial growth factor (VEGF) or inhibition of kinase insert domain-containing receptor (KDR) during fetal lung development results in inhibition of angiogenesis and alveolar hypoplasia (14), resembling the pathological features of BPD. Moreover, altered expression of angiogenic factors likely plays a role in the development of BPD. For example, preterm human and/or baboon neonates with RDS or BPD appear deficient in VEGF, platelet-endothelial cell adhesion molecule 1 (PECAM-1), fms-like tyrosine kinase 1 (Flt-1), and angiopoietin 1 receptor Tie-2 (15–17). Many, if not all, of these blood vessel growth- and differentiation-promoting proteins are targets of HIFs (18). Moreover, expression of some of these factors, such as VEGF and its receptors, is increased by hypoxia (19) and diminished by hyperoxia (20).

We have shown that HIF-1α and -2α proteins can be increased by pharmacological inhibition of PHDs in developing and adult pulmonary cells in vitro in ambient O2 concentrations, with concomitant increases in VEGF (21). It is unknown whether this approach could be effective at higher O2 concentrations, which are commonly required in the treatment of severe lung disease, such as RDS and evolving BPD. Therefore, the aims of this study were to investigate whether HIF actions could be stimulated through PHD inhibition in hyperoxia resulting in enhanced expression of angiogenic proteins and angiogenesis in human lung cells relevant to the development of BPD. In addition, the potential efficacy of the PHD inhibitor FG-4095 (previously called PHI-1) in vivo was evaluated by using a fetal baboon lung explant model.

Materials and Methods

Materials for this study were purchased from: Primary human developing lung microvascular endothelial cells (HLMVE-F); endothelial cell medium and supplements (ScienCell, San Diego); primary adult HLMVE (HLMVE-A); endothelial cell basal medium and supplements including rhVEGF-A165 (Cambrex BioScience, Walkersville, MD); alveolar epithelium-like (A549) cells (American Type Culture Collection), and F12K media (GIBCO); Matrigel and antibodies against HIF-1α and GM130 (BD Biosciences, San Diego); antibodies against HIF-2α; PHD-1, -2, and -3 (Novus Biologicals, Littleton, CO); angiopoietin 1 and 4, Tie-2, Flt-1, and KDR (R & D Systems); Flt-1 (Santa Cruz Biotechnology); PECAM-1 (Dako); β-actin (Sigma); 3-[(2,4-dimethylpyrrol-5-yl)methylidene]-indolin-2-one (SU5416) and 4-[(4′-chloro-2′-fluoro)phenylamino]-6,7-dimethoxyquinazoline (CB676475) (Calbiochem); donkey normal serum and FITC- or Texas red-conjugated secondary antibodies (Jackson ImmunoResearch); and Prolong Gold-DAPI (Molecular Probes). FG-4095 was obtained from Fibro-Gen and dimethyloxaloylglycine (DMOG) from C. Pugh (University of Oxford, Oxford, U.K.).

Cell Culture. In preliminary experiments, we found that the VEGF ELISA method did not detect any VEGF present in FBS, but that addition of human VEGF as a supplement to cells obscured detection of endogenously produced VEGF. Therefore, cells were exposed in media containing 5% or 10% serum and all other supplements except VEGF. Cells were studied at 80% confluence. Hyperoxic exposures were performed at sea level atmospheric pressure.

Lung Explant Model. Lung explants from fetal baboons (125 and 140 d, corresponding to 27 and 32 human gestational weeks; term is 186 d) were processed as described (16, 22) with modifications and cultured for 24 h in either 21% or 3% O2. For hypoxic exposures, a special hypoxic working station (Bactron 1.5 Anaerobic chamber, Sheldon Manufacturing, Cornelius, OR) was used to enable harvesting in 3% O2.

PHD Inhibitors. The effective concentrations and toxicity profile of FG-4095 (5–500 μM), a selective HIF-PHD inhibitor, for the cell types in this study have been characterized previously (21). DMOG (1 mM), a more nonselective PHD inhibitor, was used as a positive control for comparison to FG-4095 at a standard concentration (1 mM) reported in literature (3). PHD inhibitors were diluted in PBS.

Western Blotting and ELISA. Detection of HIF-1α and -2α and PHD-1, -2, and -3 proteins by Western blotting and of VEGF, Flt-1, and KDR by ELISA was performed as described (21). SU5416, a KDR inhibitor (IC50 1 μM), and CB676475, a Flt-1 (IC50 2 μM) and KDR (IC50 100 nM) inhibitor, were diluted in DMSO (0.1%) and used at concentrations of 10 and 7 μM, respectively.

Immunolocalization of Flt-1 and KDR. Following FG-4095 or DMOG treatment for 24 h, HLMVE-A cells were fixed with 4% paraformaldehyde. After permeabilization and blocking, cells were incubated with primary antibodies against Flt-1 or KDR and GM130 detecting cis-Golgi for 60 min. In some experiments, two different Flt-1 antibodies were used with identical findings. Negative controls included normal goat and mouse IgG at the same concentrations as the primary antibodies. FITC- or Texas red-conjugated secondary antibodies were then applied for 60 min. After mounting with Prolong Gold-DAPI, slides were viewed by using a Zeiss Axiovert 200 M fluorescent microscope and digital images recorded by using slidebook software (Intelligent Imaging Innovations, Denver).

VEGF Real-Time PCR. Total RNA was isolated and VEGF mRNA from HLMVE-A and A549 cells was quantified as described (23).

Angiogenesis Assay. HLMVE-F or -A cells in complete media without exogenous VEGF were seeded in wells coated with growth factor-reduced Matrigel. Exogenous VEGF (20 ng/ml) was used as a positive control. In pilot experiments, maximal branching was found at 7 h, after which the cells regressed and eventually died. After exposure, cells were fixed with 4% paraformaldehyde. Digital images (final magnification ×6.3) were recorded by using a Nikon eclipse ε 600 microscope and the branch points counted. In some experiments, neutralizing antibodies against Flt-1 (4 μg/ml) and KDR (2.5 μg/ml) were used.

Statistical Analyses. Results were compared by Mann–Whitney nonparametric test, Student's t test, or ANOVA combined with Fisher's protected least significant difference for post-hoc comparisons (statview 4.51, Abacus Concepts, Berkeley, CA). P values of ≤0.05 were considered statistically significant.

Results

HIF-1α and -2α Proteins. Basal HIF expression was negligible in 21% O2 and was unaltered by exposure to 95% O2 in HLMVE-A and A549 cells. Levels of both HIF-1α and -2α were enhanced in response to FG-4095 (125 μM) treatment for 4 h (not shown) and 24 h in both 21% and 95% O2 (Fig. 1). HIF-1α and -2α increased 3- to 6-fold by FG-4095 and 3- to 10-fold by DMOG (1 mM) in HLMVE-A cells (Fig. 1). In A549 cells, FG-4095 (250 μM) caused a 2- to 3-fold increase (P ≤ 0.01, n = 3) in HIF-1α and -2α at both O2 concentrations. In contrast to HLMVE-A cells, FG-4095 was more effective than DMOG (P ≤ 0.05, n = 3) in stabilizing HIFs in A549 cells in hyperoxia (not shown). Effect of PHD blockade in HLMVE-F cells in ambient O2 has been characterized by us (21), and only selected key experiments were conducted using these cells in the current study.

Fig. 1.

Fig. 1.

Open in a new tab

Effect of FG-4095 (125 μM) or DMOG (1 mM) on HIF-1α (A) and -2α (B) proteins in 21% and 95% O2 in HLMVE-A cells (24 h). Data are shown as means (SD, n = 3). *, P ≤ 0.01 vs. control in 21% and 95% O2; #, P ≤ 0.01 FG-4095 vs. DMOG in 21% and 95% O2.

VEGF mRNA and Protein. To investigate whether HIF stimulation through FG-4095 treatment would cause transcriptional activation of an important target gene controlling angiogenesis, VEGF mRNA was measured. In HLMVE-A cells, hyperoxia per se increased VEGF mRNA 2-fold (Fig. 2A). FG-4095 treatment (125 μM) caused a 5- to 10-fold increase and DMOG (1 mM) a 14- to 37-fold increase in VEGF mRNA in 21% or 95% O2 (Fig. 2 A). Both FG-4095 and DMOG were as effective in 95% as in 21% O2. VEGF mRNA was elevated by PHD blockade in A549 cells also, but the magnitude of increase was only 2- to 4-fold (P ≤ 0.05, n = 3) when compared with untreated cells. VEGF protein was undetectable in media and in cytosolic or nuclear extracts of HLMVE-A cells. Hyperoxia had no effect on VEGF protein in A549 cells. In A549 cells, FG-4095 increased VEGF protein by 20–50% (P ≤ 0.05, n = 3–6) and DMOG by 30–90% (P ≤ 0.01, n = 3–6) in media and cell lysates at both O2 concentrations.

Fig. 2.

Fig. 2.

Open in a new tab

Assessment of VEGF (A), Flt-1 (B), and KDR (C) in HLMVE-A cells treated with FG-4095 (125 μM) or DMOG (1 mM) in 21% and 95% O2 (24 h). (A) VEGF mRNA was analyzed by real-time PCR, and the change in gene expression was determined by calculating ΔCT, where the threshold cycle (CT) value of the target gene (VEGF) was subtracted from that of the housekeeping gene malate dehydrogenase (MDH). Each unit of ΔCT represents a 2-fold change in VEGF mRNA. (B) Flt-1 and (C) KDR proteins (pg per mg total cell protein) measured by ELISA. Data are expressed as means (SD, n = 3–6). *, P ≤ 0.05 vs. control in 21% and 95% O2; #, P ≤ 0.01 vs. DMOG in 21% and 95% O2; ⁁, P ≤ 0.001 control in 21% vs. 95% O2.

VEGF Receptors Flt-1 and KDR. Treatment with FG-4095 and DMOG increased protein expression of Flt-1 and decreased that of KDR in HLMVE-A cells exposed to either 21% or 95% O2 (Fig. 2 B and C). In A549 cells, KDR was undetectable and Flt-1 protein unchanged in response to PHD blockade at both O2 concentrations. As described above, despite several-fold increases in VEGF mRNA in response to FG-4095 or DMOG, VEGF protein was undetectable in HLMVE-A cells. To address this issue and to investigate the role of altered Flt-1 and KDR expression due to PHD-blockade, HLMVE-A cells were treated with SU5416 (KDR inhibitor) or CB676475 (both Flt-1 and KDR inhibitor). VEGF remained undetectable in media and lysates of HLMVE-A cells after VEGF receptor blockade. SU5416 treatment prevented, and CB676475 diminished, the FG-4095- or DMOG-induced increase of Flt-1 in media and lysates of HLMVE-A cells (not shown). SU5416 also prevented or decreased the decline in KDR in media and lysates of FG-4095- or DMOG-treated HLMVE-A cells (not shown). Exposure of FG-4095- or DMOG-treated A549 cells to SU5416 or CB676475 did not alter VEGF, Flt-1, or KDR expression (not shown).

Flt-1 and KDR Localization. We next examined whether altered Flt-1 and KDR expression due to PHD blockade caused changes in their intracellular localization. Intense staining for Flt-1 (not shown) and KDR (Fig. 3 A–C) was found in the perinuclear region, and lesser positive staining was also found in the cytosol and extracellular membrane, whereas nuclear staining was weak or absent. Negative controls showed no staining. The intracellular localization of either receptor was unaltered after treatment with FG-4095 or DMOG (Flt-1, not shown; KDR, Fig. 3 A–C). However, KDR staining was reduced by PHD blockade, supporting the results of KDR measurement by ELISA (Figs. 2C and 3 A–C). To elucidate perinuclear staining pattern, double stainings for mitochondria or Golgi (Fig. 3 D–F) were performed. Neither Flt-1 nor KDR localized to mitochondria (not shown). However, perinuclear staining for both receptors localized to cis-Golgi (Flt-1, not shown; KDR, Fig. 3 G–I). The KDR staining pattern was more diffuse than that for Flt-1 (not shown).

Fig. 3.

Fig. 3.

Open in a new tab

Localization of KDR protein in untreated (A, D, and G) and FG-4095-(125 μM) (B, E, and H) or DMOG- (1 mM) (C, F, and I) treated HLMVE-A cells in 21% O2. KDR was localized to cytosol, extracellular membrane, and cis-Golgi. (A–C) KDR stained with FITC; (D–F) cis-Golgi stained with Texas red; and (G–I) merged images of KDR and Golgi. Nuclei were stained with gold-DAPI (blue). All pictures were taken with ×40 magnification, and digital images were saved by using identical color intensity settings. [Bar (A), 10 μm.]

Other Angiogenic Proteins. Expression of several other angiogenic proteins after PHD blockade was assessed next. PECAM-1 was enhanced by 30–50% after FG-4095 or DMOG (P ≤ 0.05 for both, n = 6) treatment for 24 h in 21% and 95% O2. Angiopoietins 1 and 4, important in angiogenesis and permeability, were undetectable by Western blotting of cell lysates or cytosolic extracts of HLMVE-A cells. Expression of angiopoietin receptor Tie-2 remained unchanged after FG-4095 or DMOG treatment. Nitrite, an indirect indicator of nitric oxide production, was below detection limit in the media of HLMVE-A and A549 cells in all groups.

Angiogenesis in Vitro. To assess modulation of angiogenesis by PHD inhibition, in vitro angiogenesis assays using HLMVE-F or -A cells were performed. In both HLMVE-F (Fig. 4 A and B) and -A (not shown) cells, FG-4095 and DMOG increased branching by 25–75% in 21% and 95% O2. In HLMVE-F cells (Fig. 4B), the increase induced by FG-4095 was half that induced by VEGF, whereas in HLMVE-A cells (not shown), FG-4095-induced angiogenesis exceeded that induced by VEGF in hyperoxia. The angiogenic response of HLMVE-F cells in 95% O2 was comparable to that in 21% O2 (Fig. 4B). In further experiments, the roles of Flt-1 and KDR in FG-4095-induced angiogenesis in HLMVE-A cells were evaluated. Untreated cells or cells treated with FG-4095, DMOG, or VEGF were exposed to neutralizing antibodies against Flt-1, KDR, or control IgG (Fig. 4C). In untreated cells, Flt-1 blockade did not alter the angiogenic response, whereas KDR blockade reduced it by 50% (Fig. 4C). In FG-4095-, DMOG-, or VEGF-induced angiogenesis, blockade of Flt-1 or KDR resulted in reduction of angiogenesis by 20–50% (Fig. 4C).

Fig. 4.

Fig. 4.

Open in a new tab

Assessment of in vitro angiogenesis following PHD blockade. (A and B) Angiogenesis assays of HLMVE-F cells after treatment with FG-4095 (125 μM), DMOG (1 mM), or VEGF (20 ng/ml) in 21% and 95% O2 (7 h). (A) a, control; b, FG-4095; c, DMOG; and d, VEGF treatment in 95% O2. Cell branch points were counted in 21% (black bars) and 95% (white bars) O2, and data are presented as percent change from control (means, SD; n = 21–28; magnification, ×6.3). * and #, P ≤ 0.001 vs. control in 21% O2. (C) Role of Flt-1 and KDR in FG-4095- or DMOG-induced angiogenesis. HLMVE-A cells were treated with control IgG (black bars) or neutralizing antibodies against Flt-1 (gray bars) or KDR (white bars). Cell branch points were counted in 21% O2, and data are presented as percent change compared with untreated control (means, SD, n = 22–40). *, P ≤ 0.05 vs. IgG-treated cells; #, P ≤ 0.001 in cells treated with neutralizing Flt-1 antibody vs. IgG treated cells; °, P ≤ 0.001 in cells treated with neutralizing KDR antibody vs. IgG treated cells; +, P ≤ 0.01 vs. control in cells treated with neutralizing Flt-1 antibody; •, P ≤ 0.001 vs. control in cells treated with neutralizing KDR antibody.

Expression of PHDs. The paradoxical induction of PHDs by HIFs could limit the efficacy of HIF-enhancing therapies. Therefore, expression of HIF-PHD enzymes after PHD blockade was measured in HLMVE-A cells exposed to 21% or 95% O2. Protein expression of PHD-2 was increased 6- to 9-fold (P ≤ 0.05, n = 3), whereas that of PHD-1 remained unchanged. PHD-3 was elevated 1.5- to 2-fold by DMOG (P ≤ 0.01, n = 6) but not by FG-4095 at both O2 concentrations. We next studied the time course for PHD-2 induction in HLMVE-A cells after treatment with FG-4095 for 24 h. After removal of FG-4095 from the media, PHD-2 remained elevated for 4–24 h (1.5- to 2-fold, P ≤ 0.05, n = 3) and had returned to baseline by 48 h.

Effect of FG-4095 on Fetal Baboon Lung. To evaluate potential actions of FG-4095 (5–500 μM, 24 h) on preterm lungs in vivo, lung explants from fetal baboons were investigated ex vivo. Both HIF-1α and -2α proteins were enhanced by FG-4095 (125 and 500 μM) in 125 d (Fig. 5 A and B) and 140 d (400–500%, P ≤ 0.05, n = 4–5) fetal lung. The increase in HIF-1α and -2α by FG-4095 was comparable to that gained by exposure to positive controls 3% O2 or DMOG (Fig. 5 A and B).

Fig. 5.

Fig. 5.

Open in a new tab

Evaluation of FG-4095 actions in lung explants of fetal (125 d) baboons ex vivo (24 h). (A) lung HIF-1α protein, (B) lung HIF-2α protein, (C) medium VEGF protein, and (D) lung PHD-2 protein after PHD blockade. Values are means (SD) from three or four separate experiments; *, P ≤ 0.05 vs. control in 21% O2.

VEGF was measured from the explant culture media to evaluate activation of angiogenic markers by PHD-blockade. FG-4095 (125 and 500 μM) increased media VEGF from both 125 d (Fig. 5C) and 140 d (3- to 10-fold, P ≤ 0.05, n = 8–10) explants. Similar VEGF increases were detected after exposure to 3% O2 or DMOG (125 d, Fig. 5C; 140 d, not shown).

Activation of the HIF-degradation pathways by PHD inhibition was assessed next. As in cell experiments, protein for PHD-2, but not for PHD-1 or -3 (not shown), was up-regulated after exposure to FG-4095 (125 and 500 μM) in both 125 d (Fig. 5D) and 140 d (50%, P ≤ 0.05, n = 4–5) lung explants.

Discussion

The goals of this study were to investigate whether HIF proteins could be enhanced in extreme hyperoxia by pharmacological inhibition of HIF-PHDs in lung cells pertinent to development of BPD, and whether that enhancement would cause an angiogenic response that potentially could limit lung hypoplasia in BPD. Unexpectedly, we found that FG-4095 was as effective in augmenting HIFs in 95% as in 21% O2. Notably, HIFs and VEGF were also enhanced in fetal baboon lung explants ex vivo in relative hyperoxia.

HIF expression is principally controlled at the posttranslational level through regulation of stability of α-subunits by PHDs, whose expression (PHD-2 and -3) is induced in hypoxia (24, 25). This induction could limit the HIF response under prolonged hypoxia and facilitate rapid degradation of HIFs upon reoxygenation (25). PHD-2 and/or -3 can also be up-regulated by specific glycolytic and tricarboxylic acid cycle metabolites independently of hypoxia (26). The three PHDs have specific, nonredundant roles (27). For example, PHD-2 has been suggested to be critical in regulating HIFs in normoxia (28). We found that PHD-2 and -3 were up-regulated by FG-4095 and/or DMOG in hyperoxia and normoxia, and that PHD-2 remained elevated for at least 24 h after removal of FG-4095. The finding that PHD-3 was up-regulated by DMOG but not by FG-4095 may reflect different mechanisms of action of these compounds. Specifically, DMOG, but not FG-4095, is a potent inhibitor of factor-inhibiting HIF (FIH) (29). FIH inhibition allows p300 binding, which in turn enhances the activity of the C-terminal transactivation domain of HIF, leading to a stronger HIF response (29). Importantly, the up-regulation of PHDs in vitro or ex vivo did not appear to limit HIF-mediated actions (likely due to continuous presence of inhibitors in our model), because VEGF and PECAM-1, downstream targets of HIFs, were induced, even in 95% O2.

Based on gene inactivation studies, both KDR and Flt-1 receptors are indispensable for fetal vascular development (30, 31). VEGF binding to KDR leads to tyrosine phosphorylation and activation of several signaling pathways mediating mitogenic, chemotactic, migratory, proliferative, survival, and permeability-enhancing effects of VEGF (32). The role of Flt-1 in angiogenesis is less well defined. Some studies suggest that Flt-1 is a decoy receptor that antagonizes KDR (33). Compared with KDR, Flt-1 has 10-fold higher affinity for VEGF but much weaker tyrosine kinase activity. Flt-1 may therefore regulate KDR through ligand trapping (34). Unlike KDR, Flt-1 is devoid of ligand-dependent down-regulation (35). Other studies characterize Flt-1 as a molecular switch for endothelial cell differentiation through limiting KDR-mediated cell proliferation and promoting capillary network formation (36).

Similarly to normoxia (21), PHD inhibition mediates Flt-1 increase and KDR decrease in hyperoxia also. Other investigators have described identical findings in hypoxia (37), but the mechanisms or physiological significance underlying these changes are not well understood. Furthermore, opposing results have been reported, in that in some studies both Flt-1 and/or KDR were up-regulated by hypoxia (38). Hypoxia-initiated events can also up-regulate KDR in a paracrine manner (39). That HIF-PHD blockade did not alter VEGF receptor expression in epithelial cells suggests a paracrine function for VEGF synthesis. Interestingly, Flt-1, but not KDR, has a hypoxia response element in the promotor region (37). This could contribute to the increased Flt-1 in response to HIF stimulation observed in our study. Moreover, because KDR is sensitive to ligand-dependent down-regulation (40), it is possible that the several-fold increase in VEGF may have triggered KDR decline. This is further supported by the finding that blocking KDR–VEGF interaction prevented FG-4095- or DMOG-induced down-regulation of KDR. In addition, KDR blockade diminished Flt-1 up-regulation, suggesting that one function for the FG-4095- or DMOG-induced increase in Flt-1 was to inhibit KDR–VEGF interaction via ligand trapping.

Immunofluorescence was used to determine whether modifications in Flt-1 and KDR expression after PHD blockade were linked to changes in their intracellular localization. Staining for KDR was more diffuse than for Flt-1, but neither receptor's primary location changed after PHD inhibition. The most intense staining for each receptor was in the perinuclear region. Almost no staining was present in nuclei, implying that changes in nuclear extracts determined by ELISA resulted from perinuclear protein. Perinuclear staining for Flt-1, and most of the staining for KDR, was colocalized to cis-Golgi, the protein-producing side of the organelle. Traditionally, the Golgi apparatus has been thought to function primarily in posttranslational modification and storage of proteins. However, recent studies propose a more active role. For example, endothelial NO synthase (eNOS), which has been considered to be primarily an outer membrane protein, has been localized to the Golgi apparatus (41). Importantly, the membrane-bound and Golgi-associated eNOS appear to have different signaling functions, which could provide flexibility for endothelial cell responses to different stimuli (41).

We next determined whether FG-4095 could increase angiogenesis in vitro, even in extreme hyperoxia. Remarkably, both FG-4095 and DMOG increased tube branch points in developing and adult lung endothelial cells in 21% as well as in 95% O2. To elucidate roles for Flt-1 and KDR in FG-4095-induced angiogenesis, specific neutralizing antibodies were used. As in earlier studies (32), blockade of KDR under basal conditions resulted in reduction of branching, whereas blockade of Flt-1 did not. However, in each model of induced angiogenesis (FG-4095, DMOG, or VEGF), both receptors appeared to play a role, i.e., antagonism of either Flt-1 or KDR diminished branching. The finding that Flt-1, in addition to KDR, participated in induced angiogenesis is supported by findings of other investigators. Specifically, Flt-1 antagonism prevented fibroblast growth factor 2-induced angiogenesis (42). In addition, administration of anti-VEGF antibody prevented enhancement of angiogenesis, but this could be restored by placenta growth factor 1, again suggesting Flt-1 involvement (42). Collectively, the data indicate that under unstimulated conditions, branching is mainly regulated through KDR, whereas both Flt-1 and KDR are responsible for the enhanced branching in response to VEGF or stimuli that act through VEGF. This further implies that increased Flt-1 after PHD-blockade serves as a functional stimulus to promote branching of endothelium.

Therapies that drive single angiogenic factors could cause untoward effects. For example, excess VEGF production has been associated with increased neonatal mortality, pulmonary hemorrhage and edema, alveolar remodeling, and inflammation (43, 44). In addition, in developing lung, temporal and spatial expression of VEGF must be meticulously controlled, or it can lead to altered airway branching morphogenesis (45). However, because HIFs govern a plethora of genes involved in angiogenesis, their stimulation might lead to more balanced and physiological angiogenesis than treatment with only one angiogenic factor. This is supported by findings such that mice overexpressing HIF-1α (46) are devoid of an abnormal form of angiogenesis detected in mice overexpressing only VEGF or VEGF combined with angiopoietin 1 (47, 48).

Conclusion

We have shown that HIF-1α and -2α can be stimulated through PHD blockade in extreme hyperoxia as effectively as in normoxia. Treatment with FG-4095 or DMOG causes up-regulation of PHD-2 and/or -3, but this does not prevent HIF-mediated actions as VEGF levels are increased. PHD blockade also augments PECAM-1 expression. Flt-1 and KDR were localized to cis-Golgi, cytosol, and plasma membrane, and their locations were unaltered by PHD blockade. Remarkably, FG-4095 and DMOG increase angiogenesis in vitro, even in severe hyperoxia, in a manner similar to VEGF, an effect that could be blocked through either Flt-1 or KDR. In addition, FG-4095 enhanced HIFs and VEGF in fetal baboon lung explants cultured under relative hyperoxia. Through actions characterized above, FG-4095 could have desirable effects in enhancing lung growth in BPD.

Acknowledgments

We are grateful to L.-Y. Chang, C. Cool, N. Markham, and V. Balasubramaniam (National Jewish Medical and Research Center and University of Colorado Health Sciences Center) for helpful discussions and to V. Winter, J. Coalson, and L. Buchanan (University of Texas Health Sciences Center) for their exceptional work with the preterm primate model. Financial support was obtained from National Institutes of Health Grants U01 HL56263 (to C.W.W.), HL52636, and P51RR13986; the Academy of Finland (T.M.A.); the Foundation for Pediatric Research in Finland (T.M.A.); and the Finnish Antituberculosis Association Foundation (T.M.A.).

Author contributions: T.M.A. and C.W.W. designed research; T.M.A., B.K.S., N.S.W., R.I.C., and C.W.W. performed research; T.M.A., N.S.W., R.I.C., W.-B.H., L.A.F., and V.G. contributed new reagents/analytic tools; T.M.A. analyzed data; and T.M.A. and C.W.W. wrote the paper.

Abbreviations: BPD, bronchopulmonary dysplasia; DMOG, dimethyloxaloylglycine; Flt-1, fms-like tyrosine kinase 1; HIF-1α and -2α, hypoxia-inducible factor 1α and 2α; HLMVE-F/-A, human developing and adult lung microvascular endothelial cells, respectively; KDR, kinase insert domain-containing receptor/fetal liver kinase; PECAM-1, platelet-endothelial cell adhesion molecule 1; PHD, prolyl hydroxylase domain-containing protein; RDS, respiratory distress syndrome; VEGF, vascular endothelial growth factor.

References

  • 1.Wood, S. M., Gleadle, J. M., Pugh, C. W., Hankinson, O. & Ratcliffe, P. J. (1996) J. Biol. Chem. 271, 15117-15123. [DOI] [PubMed] [Google Scholar]
  • 2.Hu, C. J., Wang, L. Y., Chodosh, L. A., Keith, B. & Simon, M. C. (2003) Mol. Cell. Biol. 23, 9361-9374. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Jaakkola, P., Mole, D. R., Tian, Y. M., Wilson, M. I., Gielbert, J., Gaskell, S. J., Kriegsheim, A., Hebestreit, H. F., Mukherji, M., Schofield, C. J., et al. (2001) Science 292, 468-472. [DOI] [PubMed] [Google Scholar]
  • 4.Ivan, M., Kondo, K., Yang, H., Kim, W., Valiando, J., Ohh, M., Salic, A., Asara, J. M., Lane, W. S. & Kaelin, W. G., Jr. (2001) Science 292, 464-468. [DOI] [PubMed] [Google Scholar]
  • 5.Epstein, A. C., Gleadle, J. M., McNeill, L. A., Hewitson, K. S., O'Rourke, J., Mole, D. R., Mukherji, M., Metzen, E., Wilson, M. I., Dhanda, A., et al. (2001) Cell 107, 43-54. [DOI] [PubMed] [Google Scholar]
  • 6.Bruick, R. K. & McKnight, S. L. (2001) Science 294, 1337-1340. [DOI] [PubMed] [Google Scholar]
  • 7.Hirsilä, M., Koivunen, P., Günzler, V., Kivirikko, K. I. & Myllyharju, J. (2003) J. Biol. Chem. 278, 30772-30780. [DOI] [PubMed] [Google Scholar]
  • 8.Lando, D., Peet, D. J., Whelan, D. A., Gorman, J. J. & Whitelaw, M. L. (2002) Science 295, 858-861. [DOI] [PubMed] [Google Scholar]
  • 9.Jeong, J. W., Bae, M. K., Ahn, M. Y., Kim, S. H., Sohn, T. K., Bae, M. H., Yoo, M. A., Song, E. J., Lee, K. J. & Kim, K. W. (2002) Cell 111, 709-720. [DOI] [PubMed] [Google Scholar]
  • 10.Cockman, M. E., Masson, N., Mole, D. R., Jaakkola, P., Chang, G. W., Clifford, S. C., Maher, E. R., Pugh, C. W., Ratcliffe, P. J. & Maxwell, P. H. (2000) J. Biol. Chem. 275, 25733-25741. [DOI] [PubMed] [Google Scholar]
  • 11.Stenmark, K. R. & Gebb, S. A. (2003) Am. J. Respir. Cell Mol. Biol. 28, 133-137. [DOI] [PubMed] [Google Scholar]
  • 12.van Tuyl, M., Liu, J., Wang, J., Kuliszewski, M., Tibboel, D. & Post, M. (2005) Am. J. Physiol. 288, L167-L178. [DOI] [PubMed] [Google Scholar]
  • 13.Jobe, A. H. & Bancalari, E. (2001) Am. J. Respir. Crit. Care Med. 163, 1723-1729. [DOI] [PubMed] [Google Scholar]
  • 14.Jakkula, M., Le Cras, T. D., Gebb, S., Hirth, K. P., Tuder, R. M., Voelkel, N. F. & Abman, S. H. (2000) Am. J. Physiol. 279, L600-L607. [DOI] [PubMed] [Google Scholar]
  • 15.Bhatt, A. J., Pryhuber, G. S., Huyck, H., Watkins, R. H., Metlay, L. A. & Maniscalco, W. M. (2001) Am. J. Respir. Crit. Care Med. 164, 1971-1980. [DOI] [PubMed] [Google Scholar]
  • 16.Coalson, J. J., Winter, V. T., Siler-Khodr, T. & Yoder, B. A. (1999) Am. J. Respir. Crit. Care Med. 160, 1333-1346. [DOI] [PubMed] [Google Scholar]
  • 17.Lassus, P., Turanlahti, M., Heikkilä, P., Andersson, L. C., Nupponen, I., Sarnesto, A. & Andersson, S. (2001) Am. J. Respir. Crit. Care Med. 164, 1981-1987. [DOI] [PubMed] [Google Scholar]
  • 18.Pugh, C. W. & Ratcliffe, P. J. (2003) Nat. Med. 9, 677-684. [DOI] [PubMed] [Google Scholar]
  • 19.Forsythe, J. A., Jiang, B. H., Iyer, N. V., Agani, F., Leung, S. W., Koos, R. D. & Semenza, G. L. (1996) Mol. Cell. Biol. 16, 4604-4613. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Hosford, G. E. & Olson, D. M. (2003) Am. J. Physiol. 285, L161-L168. [DOI] [PubMed] [Google Scholar]
  • 21.Asikainen, T. M., Ahmad, A., Schneider, B. K., Ho, W.-B., Arend, M., Brenner, M., Günzler, V. & White, C. W. (2005) Free Radic. Biol. Med. 38, 1002-1013. [DOI] [PubMed] [Google Scholar]
  • 22.Das, K. C., Guo, X. L. & White, C. W. (1999) Am. J. Physiol. 276, L530-L539. [DOI] [PubMed] [Google Scholar]
  • 23.Waleh, N., Kajino, H., Marrache, A. M., Ginzinger, D., Roman, C., Seidner, S. R., Moss, T. J., Fouron, J. C., Vazquez-Tello, A., Chemtob, S., et al. (2004) Circulation 110, 2326-2332. [DOI] [PubMed] [Google Scholar]
  • 24.D'Angelo, G., Duplan, E., Boyer, N., Vigne, P. & Frelin, C. (2003) J. Biol. Chem. 278, 38183-38187. [DOI] [PubMed] [Google Scholar]
  • 25.Marxsen, J. H., Stengel, P., Doege, K., Heikkinen, P., Jokilehto, T., Wagner, T., Jelkmann, W., Jaakkola, P. & Metzen, E. (2004) Biochem. J. 381, 761-767. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Dalgard, C. L., Lu, H., Mohyeldin, A. & Verma, A. (2004) Biochem. J. 380, 419-424. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Appelhoff, R. J., Tian, Y. M., Raval, R. R., Turley, H., Harris, A. L., Pugh, C. W., Ratcliffe, P. J. & Gleadle, J. M. (2004) J. Biol. Chem. 279, 38458-38465. [DOI] [PubMed] [Google Scholar]
  • 28.Berra, E., Benizri, E., Ginouves, A., Volmat, V., Roux, D. & Pouyssegur, J. (2003) EMBO J. 22, 4082-4090. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Koivunen, P., Hirsilä, M., Günzler, V., Kivirikko, K. I. & Myllyharju, J. (2004) J. Biol. Chem. 279, 9899-9904. [DOI] [PubMed] [Google Scholar]
  • 30.Fong, G. H., Rossant, J., Gertsenstein, M. & Breitman, M. L. (1995) Nature 376, 66-70. [DOI] [PubMed] [Google Scholar]
  • 31.Shalaby, F., Rossant, J., Yamaguchi, T. P., Gertsenstein, M., Wu, X. F., Breitman, M. L. & Schuh, A. C. (1995) Nature 376, 62-66. [DOI] [PubMed] [Google Scholar]
  • 32.Ferrara, N., Gerber, H. P. & LeCouter, J. (2003) Nat. Med. 9, 669-676. [DOI] [PubMed] [Google Scholar]
  • 33.Rahimi, N., Dayanir, V. & Lashkari, K. (2000) J. Biol. Chem. 275, 16986-16992. [DOI] [PubMed] [Google Scholar]
  • 34.Hiratsuka, S., Maru, Y., Okada, A., Seiki, M., Noda, T. & Shibuya, M. (2001) Cancer Res. 61, 1207-1213. [PubMed] [Google Scholar]
  • 35.Meyer, R. D., Singh, A., Majnoun, F., Latz, C., Lashkari, K. & Rahimi, N. (2004) Oncogene 23, 5523-5531. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Bussolati, B., Dunk, C., Grohman, M., Kontos, C. D., Mason, J. & Ahmed, A. (2001) Am. J. Pathol. 159, 993-1008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Gerber, H. P., Condorelli, F., Park, J. & Ferrara, N. (1997) J. Biol. Chem. 272, 23659-23667. [DOI] [PubMed] [Google Scholar]
  • 38.Tuder, R. M., Flook, B. E. & Voelkel, N. F. (1995) J. Clin. Invest. 95, 1798-1807. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Brogi, E., Schatteman, G., Wu, T., Kim, E. A., Varticovski, L., Keyt, B. & Isner, J. M. (1996) J. Clin. Invest. 97, 469-476. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Duval, M., Bedard-Goulet, S., Delisle, C. & Gratton, J. P. (2003) J. Biol. Chem. 278, 20091-20097. [DOI] [PubMed] [Google Scholar]
  • 41.Fulton, D., Babbitt, R., Zoellner, S., Fontana, J., Acevedo, L., McCabe, T. J., Iwakiri, Y. & Sessa, W. C. (2004) J. Biol. Chem. 279, 30349-30357. [DOI] [PubMed] [Google Scholar]
  • 42.Kanda, S., Miyata, Y. & Kanetake, H. (2004) J. Biol. Chem. 279, 4007-4016. [DOI] [PubMed] [Google Scholar]
  • 43.Le Cras, T. D., Spitzmiller, R. E., Albertine, K. H., Greenberg, J. M., Whitsett, J. A. & Akeson, A. L. (2004) Am. J. Physiol. 287, L134-L142. [DOI] [PubMed] [Google Scholar]
  • 44.Kaner, R. J., Ladetto, J. V., Singh, R., Fukuda, N., Matthay, M. A. & Crystal, R. G. (2000) Am. J. Respir. Cell Mol. Biol. 22, 657-664. [DOI] [PubMed] [Google Scholar]
  • 45.Akeson, A. L., Greenberg, J. M., Cameron, J. E., Thompson, F. Y., Brooks, S. K., Wiginton, D. & Whitsett, J. A. (2003) Dev. Biol. 264, 443-455. [DOI] [PubMed] [Google Scholar]
  • 46.Elson, D. A., Thurston, G., Huang, L. E., Ginzinger, D. G., McDonald, D. M., Johnson, R. S. & Arbeit, J. M. (2001) Genes Dev. 15, 2520-2532. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Larcher, F., Murillas, R., Bolontrade, M., Conti, C. J. & Jorcano, J. L. (1998) Oncogene 17, 303-311. [DOI] [PubMed] [Google Scholar]
  • 48.Thurston, G., Suri, C., Smith, K., McClain, J., Sato, T. N., Yancopoulos, G. D. & McDonald, D. M. (1999) Science 286, 2511-2514. [DOI] [PubMed] [Google Scholar]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences

RESOURCES