Hypoxia induces protection against etoposide-induced apoptosis: molecular profiling of changes in gene expression and transcription factor activity

Abstract

Background: it is now well established that hypoxia renders tumor cells resistant to radio- but also chemotherapy. However, few elements are currently available as for the mechanisms underlying this protection.

Results: in this study, physiological hypoxia was shown to inhibit apoptosis induced in HepG2 cells by etoposide. Indeed, hypoxia reduced DNA fragmentation, caspase activation and PARP cleavage. The DNA binding activity of 10 transcription factors was followed while the actual transcriptional activity was measured using specific reporter plasmids. Of note is the inhibition of the etoposide-induced activation of p53 under hypoxia. In parallel, data from low density DNA microarrays indicate that the expression of several pro- and anti-apoptotic genes was modified, among which are Bax and Bak whose expression profile paralleled p53 activity. Cluster analysis of data unravels several possible pathways involved in the hypoxia-induced protection against etoposide-induced apoptosis: one of them could be the inhibition of p53 activity under hypoxia since caspase 3 activity parallels Bax and Bak expression profile. Moreover, specific downregulation of HIF-1alpha by RNA interference significantly enhanced apoptosis under hypoxia possibly by preventing the hypoxia mediated decrease in Bak expression without altering Bax expression.

Conclusion: these results are a clear demonstration that hypoxia has a direct protective effect on apoptotic cell death. Moreover, molecular profiling points to putative pathways responsible for tumor growth in challenging environmental conditions and cancer cell resistance to chemotherapeutic agents.

Figure 1

Hypoxia protects HepG2 cells against etoposide-induced apoptosis. HepG2 cells were incubated under normoxic (N) or hypoxic (H) conditions with or without etoposide (e) at 50 μM for 16 hours (A, B, C, D, F) or 40 hours (E). A, the overall caspase activity was assayed by measuring free rhodamine 110 released after cleavage of the caspase substrate DEVD-R110. Results are expressed as means ± 1 SD (n = 3). B, after the incubation, cells were fixed, permeabilized and stained for active caspase 3 using a specific antibody (green). Nuclei were detected with To-Pro-3 (blue). Observation was performed using a confocal microscope with a constant photomultiplier. C, after the incubation, DNA fragmentation was assayed using an ELISA for soluble nucleosomes. Results are expressed as means ± 1 SD (n = 3). D, PARP-1 and cleaved 85 kDa fragment were detected in total cell extracts by western blotting, using a specific mouse anti-PARP-1 antibody. E, after the incubation, LDH was assessed. Results are expressed as means ± 1 SD (n = 3). F, after the incubation, complete medium was added back and cells were incubated for 7 days before being fixed and stained with crystal violet. **, ***: p < 0.01, p < 0.001 vs. normoxia ; §§§: p < 0.001 vs. normoxia+etoposide.

Figure 2

Effect of hypoxia and/or etoposide on DNA binding activity of 8 transcription factors, measured with the TF Chip MAPK microarrays. Cells were incubated under normoxic (N) or hypoxic (H) conditions with or without etoposide (e) at 50 μM for 16 hours. After the incubation, nuclear extracts were performed and hybridized on the arrays as described in Materials and Methods. A, Arrays hybridized with nuclear extracts from each condition. Fluorescence is represented in pseudocolor scale and corresponds to the DNA binding activity. Red circles point out spots for detecting p53 DNA binding. B, each value is the average of the corrected values calculated from three independent experiments and expressed as mean ± 1 SD (n = 3). *, **, ***: p < 0.05, p < 0.01, p < 0.001 vs. normoxia ; §: p < 0.05 vs. normoxia+etoposide.

Figure 3

Effect of hypoxia and/or etoposide on the transcription activity of 6 transcription factors, measured with a reporter assay. Cells were cotransfected with the corresponding reporter plasmid encoding the firefly luciferase and the pCMVβ normalization plasmid before being incubated 16 hours under normoxia (N) or hypoxia (H) in the presence or absence of etoposide (e) at 50 μM. Results are expressed as means of the ratio between firefly luciferase activity and the β-galactosidase activity ± 1 SD (n = 3). Results are expressed in induction levels by comparison with the reference condition, normoxia. *, **, *** : p < 0.05, p < 0.01, p < 0.001 vs. normoxia ; §§, §§§: p < 0.01, p < 0.001 vs. normoxia+etoposide.

Figure 4

Cluster analysis of the gene expression patterns and transcription factor activity profiles. The expression values of up- and down-regulated genes as well as the activity profiles of the transcription factors of interest, measured with the DNA binding array and using a reporter system, were subjected to K-means clustering generating 10 clusters. A. Pseudocolors are given for increased expression/activity (red) or decreased expression/activity (green). B, Act– (highlighted in red) means the transcriptional activity measured by a reporter system for each transcription factor of interest; Db– (highlighted in yellow) is for the DNA binding activity measured with the array. C, D Cluster analysis of the gene expression patterns, the transcription factor activity profiles and the "phenotypic" profiles of the apoptosis level. C, "phenotype" values from figure 1, for caspase activity (caspactivity) and DNA fragmentation (DNAfrag) (highlighted in blue), were transformed in log values of the ratio normalized to the corresponding normoxic control. D, these values were subjected to K-means clustering generating 10 clusters. Act– means the transcriptional activity measured by a reporter system for each transcription factor of interest; Db– is for the DNA binding activity measured with the array.

Figure 5

Effect of HIF-1α silencing on the hypoxia-induced protection against the etoposide-induced apoptosis. A, cells were transfected with 5, 20 or 50 nM HIF-1α siRNA, 50 nM non-targeting siRNA or with the transfection reagent alone (DMF) for 24 hours. Cells were then incubated under normoxia or hypoxia for 6 hours and total cell extracts were analyzed by western blot for HIF-1α protein level. A, B, C, cells were transfected with 50 nM HIF-1α siRNA or non-targeting siRNA for 24 hours. They were then incubated under normoxic (N) or hypoxic (H) conditions with or without etoposide (e, 50 μM) for 16 hours. B, after the incubation, total RNA was extracted, submitted to reverse transcription and then to amplification in the presence of SYBR Green and specific primers. α-tubulin was used as the house keeping gene for data normalization. Data are given in fold-induction. C, caspase 3 activity was assayed. Results are expressed as means ± 1 SD (n = 3). *** p < 0.001 vs. normoxia ; §§§ p < 0.001 vs. normoxia+etoposide ; ΔΔΔ p < 0.001 vs. hypoxia ; °°° p < 0.001 vs. hypoxia+etoposide. D, PARP-1 and cleaved 85 kDa fragment were detected in total cell extracts by western blotting, using a specific mouse anti-PARP-1 antibody.

Figure 6

A, Gene expression profiling in HepG2 cells incubated with or without etoposide under normoxic or hypoxic conditions after HIF-1α siRNA transfection. Cells were transfected with 50 nM HIF-1α siRNA or non-targeting siRNA for 24 hours. They were then incubated under normoxic (N) or hypoxic (H) conditions with or without etoposide (e, 50 μM) for 16 hours before total RNA extraction, reverse transcription and amplification by real-time PCR in the presence of SYBR Green and specific primers. α-tubulin was used as the house keeping gene for data normalization. Data are given in fold-induction as the mean ± 1 SD for experimental triplicates or as the mean for experimental duplicates. ** p < 0.01 vs hypoxia+etoposide.