Growth factor-mediated induction of HDM2 positively regulates hypoxia-inducible factor 1alpha expression

Abstract

The hypoxia-inducible factor 1 (HIF-1) transcriptional complex is regulated by cellular oxygen levels and growth factors. The phosphoinosotide 3-kinase (PI-3K)-Akt/protein kinase B (PKB) pathway has been shown to regulate HIF-1 activity in response to oncogenic signals and growth factors. We assessed whether the HDM2 oncoprotein, a direct target of Akt/PKB, could regulate HIF-1alpha expression and HIF-1 activity under normoxic conditions. We found that growth factor stimulation, overexpression of Akt/PKB, or loss of PTEN resulted in enhanced expression of both HIF-1alpha and HDM2. Growth factor-mediated induction of HIF-1alpha was ablated by transient expression of a dominant negative form of Akt/PKB or by treatment with LY294002. Transient expression of HDM2 led to increased expression of HIF-1alpha. Pulse-chase and cycloheximide experiments revealed that HDM2 did not significantly affect the half-life of HIF-1alpha. Growth factor-induced HIF-1alpha and HDM2 proteins were localized to the nucleus, and induction of both proteins was observed in both p53(+/+) and p53(-/-) HCT116 cells to comparable levels. Importantly, insulin-like growth factor 1-induced HIF-1alpha expression was observed in p53-null mouse embryo fibroblasts (MEFs) but was significantly impaired in p53 Mdm2 double-null MEFs, indicating a requirement for Mdm2 in this process. Finally, we showed that phosphorylation at Ser166 in HDM2 contributed in part to growth factor-mediated induction of HIF-1alpha. Our study has important implications for the role of the PI-3K-Akt/PKB-HDM2 pathway in tumor progression and angiogenesis.

FIG. 1.

HIF-1α expression is regulated by growth factor stimulation and the PI-3K-Akt/PKB signaling pathway. (A) Western blot analysis of HIF-1α and phosphorylated Akt (Akt-P) (Ser473) proteins from HCT116 cells treated with LY294002 (10 μM) for 30 min prior to stimulation with IGF-1 (100 ng/ml) for 6 h. β-Actin expression was used as a load control in all experiments. (B) Inhibition of PI-3K ablates HIF-1 activity. U2OS cells stably expressing pGL-HRE (U2OS-HRE-luc) were transiently transfected with either 5 μg of the pCMV control vector (−) or 5 μg of HIF-1α (pCMVβ-HA-HIF-1α) (+) expression constructs. pCMV-β-Gal (50 ng) was cotransfected in all experiments as an internal transfection control. LY294002 (10 μM) was added for 6 h prior to harvest, and HIF-1 activity was determined by luciferase assay. Experiments were performed in triplicate, with the standard deviation indicated. Activity is expressed as RLU standardized to β-Gal expression. The lower panel shows Western blot analysis of HIF-1α expression. β-Actin expression was used as the load control.

FIG. 2.

Growth factor-mediated induction of HIF-1α requires activation by Akt/PKB. (A) Western blot analysis of HIF-1α and phosphorylated Akt (Akt-P) (Ser473) or total Akt proteins from U2OS cells which were transiently cotransfected with 10 μg of the pCMV vector control (−), HIF-1α expression constructs, and/or 5 μg of Akt expression constructs. Green fluorescent protein (GFP) (1 μg) was used as a transfection control as previously described (4). (B) Western blot analysis of HIF-1α or total Akt proteins from U2OS cells transiently cotransfected with either the pCMV vector control (−), 10 μg of HIF-1α, and/or 5 μg of kinase-dead (KD)-Akt expression constructs. (C) Western blot analysis of HIF-1α and Akt proteins from U2OS cells transiently cotransfected with either 5 μg of the pCMV vector control (−) or 5 μg of KD-Akt expression constructs. Cells were starved for 36 h posttransfection and then stimulated with IGF-1 (100 ng/ml) for 6 h. β-Actin expression was used as the load control in all experiments.

FIG. 3.

HDM2 expression is regulated by the PI-3K-Akt/PKB signaling pathway. (A) Western blot analysis of HIF-1α, HDM2, and phosphorylated Akt (Akt-P) (Ser473) proteins from primary human fibroblasts (MRC-5 cells). Cells were starved overnight and then treated with MG132 (50 μM) for 3 h, followed by incubation with or without IGF-1 (100 ng/ml) for 6 h. (B) Immunofluorescence staining of HIF-1α (upper panels) and HDM2 (lower panels) proteins in HCT116 cells before treatment (−) and after treatment with IGF-1 (+IGF-1). Cells were counterstained with DAPI to localize the nucleus. (C) Western blot analysis of HIF-1α, HDM2, and phosphorylated Akt (Akt-P) (Ser473) proteins from U87MG cells treated or not treated with LY294002 (10 μM) 30 min prior to lysis. β-Actin expression was used as the load control in all experiments.

FIG. 4.

The HDM2 oncoprotein enhances HIF-1α expression. (A) Western blot analysis of HIF-1α and HDM2 proteins from U2OS cells that were transiently cotransfected with either the pCMV vector control (−), 10 μg of HIF-1α, and/or 5 μg of HDM2 expression constructs. (B) Graphical representation of radioactively labeled HIF-1α protein transfected as described in panel A. Transiently transfected U2OS cells were pulse-labeled with methionine-cysteine-free medium containing [³⁵S]methionine/cysteine and chased with unlabeled complete medium for the times indicated. (C) Western blot analysis of HIF-1α and HDM2 proteins from U2OS cells which were transiently cotransfected as described in panel A and then treated with 50 μg of cycloheximide (CHX)/ml for the times indicated prior to harvesting. (D) Western blot analysis of HIF-1α and HDM2 proteins from HCT116 cells, which were transiently transfected with 10 μg of HDM2 expression plasmid, starved for 36 h, stimulated with IGF-1 (100 ng/ml) for 6 h, and then treated with 50 μg of cycloheximide/ml for the times indicated prior to harvesting.

FIG. 5.

HDM2-mediated effects in the presence of p53. (A and B) U2OS cells were transiently cotransfected with the vector control (−), 10 μg of HIF-1α, 5 μg of HDM2, and/or 5 μg of p53-FLAG expression constructs. Western blot analysis was carried out using specific antibodies against HIF-1α, p53, and HDM2. Ub, ubiquitinated p53. (C) Saos-2 HIF-1α TetOn-inducible cells were transiently cotransfected with the pGL-HRE plasmid and 2.5 μg of either the vector control (−), p53-FLAG, or HDM2 expression constructs, followed by doxycycline treatment (6 μg/ml) for 24 h. pCMV-β-Gal (50 ng) was cotransfected in all experiments as an internal transfection control. HIF-1 activity was determined by luciferase assay. Experiments were performed in duplicate, with the standard deviation indicated. Activity is expressed as RLU standardized to β-Gal expression.

FIG. 6.

Induction of HIF-1α protein expression by growth factors is p53 independent. (A) Western blot analysis of HIF-1α, HDM2, and phosphorylated Akt (Akt-P) (Ser473) proteins from p53^−/− HCT116 (left panel) or p53^+/+ HCT116 cells (right panel) starved for 36 h prior to IGF-1 treatment (100 ng/ml) for 6 h.

FIG. 7.

Induction of HIF-1α protein expression by growth factors requires HDM2. Western blot analysis of HIF-1α and phosphorylated Akt proteins from p53^−/− (A) or p53^−/− Mdm2^−/− (B) MEFs starved overnight and then treated with MG132 (50 μM) for 3 h, followed by incubation with or without IGF-1 (100 ng/ml) for 6 h. β-Actin expression was used as the load control. Graphs are representative of two independent experiments and show the relative amounts of HIF-1α induced for each treatment as outlined. HIF-1α levels from Western blot analyses were measured by densitometric analysis and quantified relative to the actin load control.

FIG. 8.

Role of HDM2 phosphorylation in growth factor-mediated induction of HIF-1α. (A) Western blot analysis of HIF-1α and HDM2 proteins from Saos-2 cells (p53-null) after transient cotransfection with the pCMV vector control, 10 μg of HIF-1α, 5 μg of wild-type HDM2 (HDM2), and/or 5 μg of mutant HDM2-S166D expression constructs. (B) Western blot analysis of HIF-1α, HDM2, phosphorylated Akt (Akt-P) (Ser473), and total Akt proteins from HCT116 cells transiently transfected with either 10 μg of the pCMV vector control (−), wild-type (WT) HDM2, or mutant HDM-S166A (S166A) expression constructs. Cells were then starved for 36 h and treated with or without IGF-1 (100 ng/ml) for 6 h prior to harvest. β-Actin expression was used as the load control in all experiments.