Tumor necrosis factor-alpha causes accumulation of a ubiquitinated form of hypoxia inducible factor-1alpha through a nuclear factor-kappaB-dependent pathway

Abstract

Hypoxia-inducible factor-1 (HIF-1) is a regulator of metabolic adaptation to hypoxia. It is now appreciated that HIF-1alpha accumulation is achieved under normoxic conditions by various factors, such as TNF-alpha. Here, it was our intention to gain insight into the signaling mechanisms used by TNF-alpha to stimulate HIF-1alpha. In tubular LLC-PK1 or human embryonic kidney cells, TNF-alpha induced accumulation of HIF-1alpha protein but not HIF-1alpha mRNA. Blocking nuclear factor (NF)-kappaB with sulfasalazine or expression of an IkappaB superrepressor attenuated HIF-1alpha accumulation, whereas transfection of active p50/p65-NF-kappaB subunits mimicked a TNF-alpha response. Experiments with actinomycin D and cycloheximide also pointed to a transcriptional and translational process in facilitating the TNF-alpha response. Interestingly, and in contrast to established hypoxic signaling concepts, TNF-alpha elicited HIF-1alpha accumulation in a ubiquitinated form that still bound the von Hippel-Lindau (pVHL) protein. These data indicate that HIF-1alpha accumulation by TNF-alpha demands the NF-kappaB pathway, preserves ubiquitination of HIF-1alpha, and allows the HIF-1alpha-pVHL interaction.

Figure 1.

HIF-1α accumulation and ubiquitination by TNF-α. (A) LLC-PK₁ cells were stimulated with 500 ng/ml TNF-α for 16 h or remained untreated. Nuclei were visualized by DAPI staining. HIF-1α was detected by an anti-HIF-1α mAb and FITC-labeled anti-mouse secondary antibody. (B) HEK293 cells were cotransfected with 3 μg pCMV-HIS₆-ubiquitin and 0.3 μg pHIF-1α expression plasmids for 16 h. Medium was changed and incubations continued for another 8-h period before stimulation with 500 ng/ml TNF-α for times indicated. HIF-1α and ubiquitinated HIF-1α were determined by Western analysis as described in MATERIALS AND METHODS. Each experiment was performed at least three times, and representative data are shown.

Figure 2.

HIF-1α stabilization and ubiquitination by MG132 and CoCl₂. (A) LLC-PK₁ cells were stimulated with 1 μM MG132 or 100 μM CoCl₂ for 4 h. HIF-1α and DAPI staining were performed as described in Figure 1A. (B) HEK293 cells were cotransfected with 3 μg pCMV-HIS₆-ubiquitin and 0.3 μg pHIF-1α to follow HIF-1α and ubiquitinated HIF-1α accumulation as described in Figure 1B. HEK293 cells were stimulated with 1 μM of the proteasome inhibitor MG132 or 100 μM CoCl₂ for times indicated. Each experiment was performed at least three times, and representative data are shown.

Figure 3.

Interaction between HIF-1α and pVHL. (A) HEK293 cells were cotransfected with 1 μg pCMV-HA-pVHL and 3 μg pHIF-1α expression plasmids. After medium was changed at 16 h, incubations continued for 8 h before stimulation with 500 ng/ml TNF-α or 100 μM CoCl₂ for times indicated. Coimmunoprecipitation was performed with an anti-HA mAb that recognized pVHL- and anti-mouse antibody-coated magnetic beads. Accumulation of HIF-1α in cell lysates (input) and immunoprecipitates (IP) were determined by Western analysis using an anti-HIF-1α mAb as described in MATERIALS AND METHODS. (B) Autoradiography of endogenous [³⁵S]methionine-labeled HIF-1α that coimmunoprecipitated with pVHL. LLC-PK₁ cells were treated for 8 h with 500 ng/ml TNF-α, 1 μM MG132, or 100 μM CoCl₂ or remained unstimulated (C). For details, see MATERIALS AND METHODS. Each experiment was performed at least three times, and representative data are shown.

Figure 4.

Actinomycin D attenuated TNF-α-stimulated HIF-1α accumulation. LLC-PK₁ cells were preincubated for 30 min with the transcription inhibitor actinomycin D. Thereafter, cells were stimulated with (A) 500 ng/ml TNF-α or (B) 3 μM MG132 for 8 h. HIF-1α was determined by Western analysis. For details, see MATERIALS AND METHODS. Each experiment was performed at least three times, and representative data are shown.

Figure 5.

Inhibition of NF-κB attenuated TNF-α-stimulated HIF-1α accumulation. (A) LLC-PK₁ cells were preincubated for 30 min with the NF-κB inhibitor sulfasalazine at the indicated concentrations. (B) Cells were transfected with pCMV-IκBα and pCMVIκBαM expression plasmids for 24 h. Thereafter, cells were stimulated with 500 ng/ml TNF-α for 16 h. HIF-1α was determined by Western analysis. For details, see MATERIALS AND METHODS. Each experiment was performed at least three times, and representative data are shown.

Figure 6.

Cotransfection of p50/p65 NF-κB subunits promoted HIF-1α accumulation. LLC-PK₁ cells were cotransfected with indicated amount of pRSV-NF-κB1/pRSV-RelA expression plasmids. After transfection for 16 h, medium was changed, and incubations continued for 8 h. As the positive control, cells were treated with 500 ng/ml TNF-α for 8 h. HIF-1α was detected by Western analysis as described in MATERIALS AND METHODS. Each experiment was performed at least three times, and representative data are shown.

Figure 7.

HIF-1α mRNA under the impact of TNF-α. LLC-PK₁ cells were transfected with 500 ng pIκBαM (IκB-mutant) or with 500 ng pRSV-NF-κB1/pRSV-RelA expression plasmids as indicated in Figures 4 and 5. Sulfasalazine (500 μM) was preincubated for 30 min. Cells were then stimulated with 500 ng/ml TNF-α for 8 h. Total RNA was isolated, and 1 μg total RNA was used for reverse transcription. PCR products were visualized with ethidium bromide on 2% agarose gels as described in MATERIALS AND METHODS. Each experiment was performed at least three times, and representative data are shown.

Figure 8.

CHX affects TNF-α-induced HIF-1α accumulation. LLCPK₁ cells were stimulated with (A) 500 ng/ml TNF-α for 8 h or (B) 100 μM CoCl₂ for 2 h. Thereafter, the protein translation inhibitor CHX was added (25 μM), and incubations continued for 10-60 min. HIF-1α was detected by Western analysis using the anti-HIF-1α mouse mAb as described in MATERIALS AND METHODS. Each experiment was performed at least three times, and representative data are shown.