Tight control of hypoxia-inducible factor-α transient dynamics is essential for cell survival in hypoxia

Abstract

Intracellular signaling involving hypoxia-inducible factor (HIF) controls the adaptive responses to hypoxia. There is a growing body of evidence demonstrating that intracellular signals encode temporal information. Thus, the dynamics of protein levels, as well as protein quantity and/or localization, impacts on cell fate. We hypothesized that such temporal encoding has a role in HIF signaling and cell fate decisions triggered by hypoxic conditions. Using live cell imaging in a controlled oxygen environment, we observed transient 3-h pulses of HIF-1α and -2α expression under continuous hypoxia. We postulated that the well described prolyl hydroxylase (PHD) oxygen sensors and HIF negative feedback regulators could be the origin of the pulsatile HIF dynamics. We used iterative mathematical modeling and experimental analysis to scrutinize which parameter of the PHD feedback could control HIF timing and we probed for the functional redundancy between the three main PHD proteins. We identified PHD2 as the main PHD responsible for HIF peak duration. We then demonstrated that this has important consequences, because the transient nature of the HIF pulse prevents cell death by avoiding transcription of p53-dependent pro-apoptotic genes. We have further shown the importance of considering HIF dynamics for coupling mathematical models by using a described HIF-p53 mathematical model. Our results indicate that the tight control of HIF transient dynamics has important functional consequences on the cross-talk with key signaling pathways controlling cell survival, which is likely to impact on HIF targeting strategies for hypoxia-associated diseases such as tumor progression and ischemia.

Keywords: Cell Death; Hypoxia; Hypoxia-inducible Factor; Imaging; Mathematical Modeling; Negative Feedback Loop; Prolyl Hydroxylase; p53.

FIGURE 1.

Single cell dynamics of HIF-α nuclear levels and HIF-dependent transcription. A, HeLa cells were transfected with HIF-1α-EGFP or EGFP-HIF-2α. 24 h after transfection cells were exposed to hypoxia (1% O₂) for the indicated time points. HIF-EGFP fusion protein levels were assessed by Western blot using an anti-GFP antibody and the bands were quantified by densitometry analysis. B, HeLa cells were transfected with HIF-1α-EGFP together with dsRED-XP expression plasmid for normalization purposes. % of red and green fluorescent cells were measured by flow cytomtery and plotted as a function of time in hypoxia. Nontransfected controls (NT) were used for gating. C, selected representative images of HeLa cells transiently co-transfected with HIF-1α-EGFP and an empty dsRED Express control plasmid (not shown on the picture) to monitor the localization and number of transfected cells. Transfection efficiency was ranged from 30 to 40%. Cells were imaged using time lapse confocal microscopy every 5 min in 20.8% O₂ for 1 h and then switched to 1% O₂ for 20 h. See also supplemental Movies S1 and S2. D and E, nuclear fluorescence levels for HIF-1α (D) and HIF-2α (E) were plotted as a function of time for 4 representative cells. The straight line represents the threshold used for automatic peak detection (see “Experimental Procedures”). All traces for HIF-1α are shown in supplemental Fig. S1A. Some traces are shorter than the entire time course due to either cell death or migration out of the imaging field. F, classification of the observed HIF-α response kinetics. Transient bell shapes curves and multiple peaks were scored using a threshold (see “Experimental Procedures”). The number of cells scored in each category are indicated on the plot. G, duration of the HIF accumulation in transient response. Duration was determined as the time between the point at passing half-maximum fluorescence and returning below this value, the 25th to 75th quintile is indicated on the plot. H, percentage of transfected cells showing an increase of green fluorescence levels in hypoxia and normoxia. I, a stable HeLa cell line expressing the HIF-1α ODD-dependent ODD from residues (amino acids 529–652) fused to EGFP was generated by lentiviral transduction of a HIV-ODD-EGFP-ires-dTomato vector into HeLa cells. The ODD cell line was imaged in normoxia before a switch to 1% O₂ for 20 h. Fluorescence intensities were quantified and plotted as a function of time. Four representative plots are shown, 50 cells were tracked in total, 82% displayed pulsed dynamics. J, HeLa cells were transiently transfected with HIF-1α-EGF, either stably expressing ODD-EGFP or left non-transfected. They were cultured in normoxia or hypoxia (1% O₂) for 8 h prior to cell lysis and mRNA extraction. mRNA levels for PHD2 and -3 were measured by qPCR and normalized to cyclophilin A mRNA.

FIGURE 2.

A, luminescence images of C51 cells stably transfected with HRE-luciferase (PH3SVL cells) in 1% O₂. Luminescence levels were imaged using wide field microcopy in the presence of luciferin in cell culture in 1% O₂ for 80 h. B, mean luminescence levels of single cells were plotted as a function of time. Each color line is a representative cell. C, HeLa cells were transiently transfected with a HRE-luciferase reporter vector. D, luminescence levels in hypoxia were acquired and quantified as in B.

FIGURE 3.

Mathematical modeling of the generic HIF-PHD feedback loop. A, description of the model (see “Experimental Procedures”). HIF-α (x) is produced at rate S and removed due to PHD (y) hydroxylation. The maximal hydroxylation rate, h, is oxygen dependent and saturation of hydroxylation is determined by the parameter γ. PHD (y) is produced through induction by HIF-α at rate k and undergoes basal degradation at rate d. B, HeLa cells were transiently co-transfected with HIF-1α-EGFP and an empty dsRED-Express control plasmid. Cells were imaged every 5 min after a switch from 20 to 1% O₂ using time lapse confocal microscopy. The pictures show a typical example of HIF-1α nuclear accumulation occurring after cell division. This was observed in 50% of the cells showing a HIF-1α increase. C, HIF-1α levels in hypoxia plotted as a function of time, synchronized based on cell cycle. D, single cell data of HIF dynamics (blue line) were fitted computationally using the model (red line). See also supplemental Fig. S1 for more cell fitting. The model PHD profile is in green. The model cells are initially at equilibrium in normoxia (h = 1) and are de-oxygenated into hypoxia (h = 0.14) at t = 0. E, selected representative images of HeLa cells transiently transfected with the PHD2prom-PHD2-EGFP expression plasmid. Cells were imaged using time lapse confocal microscopy every 5 min in 20.8% O₂ for 1 h and then switched to 1% O₂ for 20 h. F, the plots represent the whole cell fluorescence intensity produced from PHD2prom-PHD2-EGFP as a function of time. G, the percentage of responsive cells is calculated from the number of transfected cells showing an increase of the green fluorescence level over time from PHD2prom-PHD2-EGFP in hypoxia and in normoxia.

FIGURE 4.

Single cell dynamics during re-oxygenation and validation of the mathematical model. A, cells were transfected as described in the legend to Fig. 1B. 24 h after transfection, cells were exposed to 1% O₂ for 6 h in the microscope stage and then re-oxygenated to 20.8% O₂. Fluorescent levels were measured 1 h prior and during the re-oxygenation period and plotted as a function of time. B, HeLa cells were cultured in hypoxia for 6 h and then subjected to re-oxygenation. HIF-1α levels were measured by Western blot at the indicated time points. Densitometry analysis of the bands were plotted as a function of time. C, using the model described in the legend to Fig. 2, we fitted the single cell traces obtained experimentally in A. The cells are initially at equilibrium in hypoxia (h = 0.14) and are re-oxygenated back into normoxia (h = 1) at t = 0. D, parameter sensitivity analysis was conducted by varying one parameter at a time, as a perturbation from the artificial cell obtained from the median parameters (S = 2.38 × 10¹ AU min⁻¹, γ = 2.98 × 10² AU, k = 4.71 × 10⁻⁴ min⁻², d = 4.71 × 10⁻⁴ min⁻¹]). Pre-stimulation (normoxic) equilibria have been normalized to 1 to emphasize the qualitative effects of parameter variation. The black curve represents the median cell model output and parameters were individually deviated either way in steps of 25% of the median value varying up to ±75%.

FIGURE 5.

A, measurement of the half-life of PHD1, -2, and -3. HeLa cells were transfected with PHD1, -2, -3-EGFP. 24 h after transfection, cells were treated with cycloheximide (10 μg/μl) and the PHDs levels were monitored for up to 24 h by measuring fluorescence intensity. B, box and whisker plot of the half-life measured in single cells for PHD1, -2, and -3. C, qPCR analysis of PHD2 and PHD3 mRNA induction during a hypoxic time course (1% O₂). Each time point sample was generated in triplicates. The plot is representative of one experiment. The experiment was repeated 4 times. D, four component model run. The model cell is initially at equilibrium in normoxia and is then de-oxygenated into hypoxia at t = 0. E, the parameter sets for the 4-component model are detailed in the table (time units in min).

FIGURE 6.

PHD2 knock-down affects HIF temporal profile. A–C, predictions are based on a 4-component model (Fig. 5A) with PHD1, -2, or -3 removed representing in silico knock-out of PHD1, -2, and -3. The models are initially at equilibrium in normoxia and then de-oxygenated into hypoxia at t = 0. D, Western blot analysis of PHD2 and PHD3 levels in WT and sh-PHD2 HeLa cells. E, Western blot analysis of HIF-1α levels in WT and SH-PHD2 HeLa cells cultured in normoxia submitted to 5 h hypoxia (1% O₂). F, Western blot analysis of HIF-1α levels in sh-PHD2 HeLa cells cultured in 1% O₂ for the indicated time points. G, single cell analysis of HIF-1α-EGFP levels in sh-PHD2 cells exposed to 1% O₂. H, Western blot analysis of PHD3 levels in WT and sh-PHD3 HeLa cells. I, single cell analysis of HIF-1α-EGFP levels in sh-PHD3 cells exposed to 1% O₂.

FIGURE 7.

A, pictures of a representative field of sh-PHD2 cells expressing HIF-1α-EGFP at the indicated times in hypoxia show cell death associated with high and long lasting levels of HIF-1α. B, the percentage of transfected sh-PHD2 or WT HeLa cells dying within the 20 h of an imaging experiment was calculated for control cells transfected with either an empty EGFP plasmid or with HIF-1α-EGFP in normoxia and hypoxia (n = 40 cells/conditions). C and D, HeLa cells (WT, shPHD2 or shPHD3 lines) were imaged simultaneously using a 4-compartment glass bottom dish (Greiner). They were labeled with Annexin V-FITC (green) and PI (red) 10 min before imaging. Cells were imaged for 2 h in normoxia prior to the switch to 1% O₂. Images were recorded every 15 min for 24 h. The number of apoptotic cells (Annexin labeling preceding the PI labeling) was counted out of the total number of cells and plotted (C). A typical field of cells at several time points is shown (D). E–G, WT, HeLa cells, or HeLa cells expressing shPHD2 or shPHD3, were cultured in 1% O₂ for the indicated time points. E, VEGF mRNA levels were assessed by qPCR. F, PUMA mRNA levels were assessed by qPCR. G, Noxa mRNA were assessed by qPCR. E–G, the plots represent the average ± S.D. of triplicate samples from a representative experiment. The experiments were performed 4 times. Results are the mean of three independent experiments ± S.E. ***, **, and * indicate statistical difference with p < 0,001, p < 0.01, and p < 0.05, respectively.

FIGURE 8.

Prediction of p53 dynamics following hypoxic switch. A, the hypoxic switch at t = 0 is represented by an instantaneous switch in HIF from a low normoxic equilibrium level to a high hypoxic equilibrium level. The p53 levels are obtained by solving the HIF-dependent p53-MdM2 feedback model. B and C, the hypoxic switch drives transient HIF dynamics determined by the HIF-PHD model, which is coupled to the p53-MdM2 feedback model. Cells are WT (B) or sh-PHD2 (C).