Mitochondrial respiratory function induces endogenous hypoxia

Abstract

Hypoxia influences many key biological functions. In cancer, it is generally believed that hypoxic condition is generated deep inside the tumor because of the lack of oxygen supply. However, consumption of oxygen by cancer should be one of the key means of regulating oxygen concentration to induce hypoxia but has not been well studied. Here, we provide direct evidence of the mitochondrial role in the induction of intracellular hypoxia. We used Acetylacetonatobis [2-(2'-benzothienyl) pyridinato-kN, kC3'] iridium (III) (BTP), a novel oxygen sensor, to detect intracellular hypoxia in living cells via microscopy. The well-differentiated cancer cell lines, LNCaP and MCF-7, showed intracellular hypoxia without exogenous hypoxia in an open environment. This may be caused by high oxygen consumption, low oxygen diffusion in water, and low oxygen incorporation to the cells. In contrast, the poorly-differentiated cancer cell lines: PC-3 and MDAMB231 exhibited intracellular normoxia by low oxygen consumption. The specific complex I inhibitor, rotenone, and the reduction of mitochondrial DNA (mtDNA) content reduced intracellular hypoxia, indicating that intracellular oxygen concentration is regulated by the consumption of oxygen by mitochondria. HIF-1α was activated in endogenously hypoxic LNCaP and the activation was dependent on mitochondrial respiratory function. Intracellular hypoxic status is regulated by glucose by parabolic dose response. The low concentration of glucose (0.045 mg/ml) induced strongest intracellular hypoxia possibly because of the Crabtree effect. Addition of FCS to the media induced intracellular hypoxia in LNCaP, and this effect was partially mimicked by an androgen analog, R1881, and inhibited by the anti-androgen, flutamide. These results indicate that mitochondrial respiratory function determines intracellular hypoxic status and may regulate oxygen-dependent biological functions.

Figure 1. Oxygen concentration surrounding the cells and oxygen consumption rates in various cell lines.

Correlation between oxygen consumption rate and oxygen concentration surrounding the cells after 3 = −23.045x +250.07 with a R² value of 0.8161. Error bars represent standard error.

Figure 2. Oxygen concentrations surrounding the cells with different cell densities.

(A) Time dependent changes in oxygen concentrations surrounding cells at an indicated cell density of LNCaP. (B) Oxygen concentration at 180 minutes from panel A (n = 3, ***P<0.001 when compared to 1×10⁶ cells/ml). Error bars represent standard error.

Figure 3. Intracellular oxygen status of various cell lines.

(A) Detection of intracellular hypoxia by BTP in prostate cancer cell lines LNCaP, C4-2 and PC-3 (upper panels). DIC images show the positions of imaged cells (lower panels). (B) Quantification of BTP phosphorescence in panel A (n = 10, ***P<0.001 when compared to LNCaP). (C) Detection of intracellular hypoxia by BTP in MCF-7 and MDAMB231 (upper panels). DIC images show the positions of imaged cells (lower panels). (D) Quantification of BTP phosphorescence in panel C (n = 10, **P<0.01 when compared to MCF-7). Error bars represent standard error.

Figure 4. Intracellular hypoxia was dependent on mitochondrial respiration.

(A) BTP phosphorescence of cells treated with indicated concentrations of rotenone. (B) Quantification of results in panel A (n = 10, ***P<0.001 when compared to 0 nM control). Error bars represent standard error.

Figure 5. Reduction of mtDNA content reversibly inhibited intracellular hypoxia.

(A) BTP phosphorescence of LNCaP, LNρ0-8, and LNCyb under normoxic conditions. (B) Quantification of BTP phosphorescence in panel A (n = 10, ***P<0.001 when compared to LNCaP). Error bars represent standard error. (C) Western blot showing HIF-1α expression in nuclear extracts from LNCaP, LNρ0-8, and LNCyb cells under normal incubation conditions (left). Nuclear extracts from all three cells lines following treatment with CoCl₂ served as positive controls for the detection of HIF-1α (center). All three cell lines were also exposed to hypoxia (0.2% O₂) for 6 hours (right). PCNA served as a loading control. (D) Densitometric analysis of western blotting results in C. Band intensities were normalized to corresponding PCNA loading control band.

Figure 6. Exogenous hypoxia- and hyperoxia-induced changes in intracellular oxygen concentration.

(A) BTP phosphorescence of LNCaP cells cultured at normoxia (20% O₂), hypoxia (0.2% O₂), or hyperoxia (40% O₂) for 1 hour. (B) Quantification of results in panel A (n = 10, ***P<0.001 when compared to 20% O₂ control). Error bars represent standard error.

Figure 7. Roles of mitochondrial respiratory function on the induction of intracellular hypoxia under exogenous hypoxic condition.

(A) BTP phosphorescence of LNCaP, LNρ0-8, and PC-3 under exogenous normoxia or hypoxia. (B) Quantification of results in panel A (n = 10, ***P<0.001 when compared to LNCaP at 20% O₂, ***P<0.001 when LNρ0-8 at 20% O₂ is compared to LNρ0-8 at 0.2% O₂, ***P<0.001 when PC-3 at 20% O₂ is compared PC-3 at O.2% O₂). Error bars represent standard error.

Figure 8. Glucose regulated oxygen concentration surrounding the cells.

Cells were incubated in the presence of glucose, pyruvate, or hydroxyurea in glucose and pyruvate-free DMEM medium (this media alone served as a control). Oxygen concentration surrounding cells was measured using OxoPlate.

Figure 9. Glucose impacted hypoxia surrounding the cells in a parabolic fashion.

Cells were incubated in the presence of varying concentrations of glucose. Oxygen concentrations surrounding the cells were measured using OxoPlate.

Figure 10. Glucose concentration regulated intracellular hypoxia in a parabolic fashion.

(A) BTP phosphorescence of the cells treated with varying concentrations of glucose. (B) Quantification of results in panel A (n = 10, **P<0.01, ***P<0.001 when compared to 4.5 mg/ml glucose). Error bars represent standard error.

Figure 11. FCS induced hypoxia.

(A) LNCaP cells were incubated with or without FCS. Oxygen concentration surrounding cells was measured using OxoPlate. (B) Final oxygen concentrations surrounding the cells at 180 minutes from panel A (n = 3, **P<0.01 when compare to minus serum). (C) BTP phosphorescence in LNCaP after incubation with or without FCS. (D) Quantification of results in panel C (n = 10, ***P<0.001 when compared to control). Error bars represent standard error.

Figure 12. Androgen played a major role in the modulation of intracellular hypoxia in LNCaP.

(A) LNCaP cells were incubated with indicated conditions and BTP phosphorescence was detected. (B) Quantification of results in panel A (n = 10, *P<0.05, **P<0.01, ***P<0.001 when compared to control, ***P<0.001 when+serum is compared to+R1881– serum, ***P<0.001 when+serum is compared to+flutamide+serum). Error bars represent standard error.