The effect of baseline O conditions on the response of prostate cancer cells to hypoxia.

The transcriptional response to hypoxia is largely regulated by the hypoxia-inducible factors (HIFs), which induce the expression of genes involved in glycolysis, angiogenesis, proliferation, and migration. Virtually all cell culture-based hypoxia experiments have used near-atmospheric (18% O) oxygen levels as the baseline for comparison with hypoxia. However, this is hyperoxic compared with mammalian tissue microenvironments, where oxygen levels range from 2% to 9% O (physioxia). Thus, these experiments actually compare hyperoxia to hypoxia. To determine how the baseline O level affects the subsequent response to hypoxia, we cultured PC-3 prostate cancer cells in either 18% or 5% O for 2 wk before exposing them to hypoxia (∼1.1% pericellular O) for 12-48 h. RNA-seq revealed that the transcriptional response to hypoxia was dependent on the baseline O level. Cells grown in 18% O before hypoxia exposure showed an enhanced induction of HIF targets, particularly genes involved in glucose metabolism, compared with cells grown in physioxia before hypoxia. Consistent with this, hypoxia significantly increased glucose consumption and metabolic activity only in cells previously cultured in 18% O, but not in cells preadapted to 5% O. Transcriptomic analyses also indicated effects on cell proliferation and motility, which were followed up by functional assays. Although unaffected by hypoxia, both proliferation and migration rates were greater in cells cultured in 5% O versus 18% O. We conclude that an inappropriately hyperoxic starting condition affects the transcriptional and metabolic responses of PC-3 cells to hypoxia, which may compromise experiments on cancer metabolism in vitro. Although human cell culture models have been instrumental to our understanding of the mechanisms involved in the cellular response to hypoxia, in virtually all experiments, cells are routinely cultured in near-atmospheric (∼18% O) oxygen levels, which are hyperoxic relative to physiological conditions in vivo. Here, we show for the first time that cells cultured in physiological O levels (5% O) respond differently to subsequent hypoxia than cells grown at 18%.