Linking measured intercellular oxygen concentration to human cell functions

To the Editor The cell responses to a specific pericellular oxygen concentration depend on a number of factors including cell type, duration of the particular oxygen concentration, cell cycle stage, pH and concentration of nutrients and regulatory molecules. The isolated effect of low oxygen, however, seems decisive enough to allow some general statements about cell regulatory responses to various pericellular oxygen concentrations. Quantification of oxygen tension around the cells is therefore an important indicator of the cell phenotype (1). The developing understanding is that responses to hypoxia are orchestrated by a number of overlapping regulatory processes. These processes result in a sequence of changes in the cell phenotype according to changes in the oxygen concentration. Reliable oxygen measurements in both in vitro model systems (2 /5) and in vivo (6 /9) are indicative of which phenotype the cell will exhibit within a given range of oxygen concentrations. The present short survey deals with these relationships. When the intercellular oxygen concentration in tissues is reduced from normoxia (4 /1% O2:/4 /1 kPa and corresponding to 40 000 / 10 000 ppm) to moderate hypoxia (1 /0.1% O2 and kPa, corresponding to 10 000 /1 000 ppm) , respiration, and as a result the ATP production, is largely unaffected (2, 10). For a long time, this observation has stood in the way of realizing that this is not the case for many other cell functions. In fact, there is a striking effect on gene expression due to transcription, mRNA stability, protein synthesis, and post-translational modifications. In the range of 0.1 /1% O2/kPa O2 (1 000 /10 000 ppm O2), the central hypoxia-responsive pathways of the hypoxia-inducible factor (HIF) are functionally very important. These pathways function by binding to hypoxia-responsive elements on promoters of more than 60, but not all, hypoxia-regulated genes (11). The activity of this HIF transcription regulator is itself governed by prolyl hydroxylases. In the presence of a sufficiently high concentration of molecular oxygen, these prolyl hydroxylases convert specific prolyl residues in the HIF-1a subunits into hydroxyproline. This allows capture by the von Hippel-Lindau tumour suppressor E3 ligase complex, leading to ubiquitination and destruction, thus preventing HIF activity at higher oxygen concentrations. Below 1% O2/1 kPa (10 000 ppm O2), oxygenation is too small to induce degradation of the HIF-1a subunit, and HIF-1 accumulates instantaneously and influences specific promoter regions (12, 13). A second mechanism regulating HIF-1 activity is mediated by an asparaginyl hydroxylase, which modulates transactivations by HIF-1. Downstream of HIF, carbonic anhydrase IX (CAIX) is an important example of a factor with a number of other functions. It is furthermore responsive to an upward shifted range of hypoxia (0.1 /5% O2/kPa or 1 000 /50 000 ppm) and has a longer half-life. Moderate hypoxia first results in a general translation inhibition, which later switches to a more specific, cap-dependent translation inhibition. Certain genes, including HIF-1 and VEGF, contain an internal ribosome entry site (IRES) allowing ribosomes to bind directly to the mRNA, thus bypassing the cap-dependent transcription repression (14). In vitro stress proteins, e.g. HSP70 and HSP90, may become expressed, but may also appear as a response to reoxygenation (15) and function as a nuclear transfer factor by HIF-1. The functional goal of these different cell activities can be seen as a restoration of the oxygen supply by HIF-1 mediated upregulation of the growth factor (VEGF), the platelet-derived endothelial cell growth factor vasodilating NO and erythropoietin production (16). From 0.1% O2 (1 000 ppm) down to 0.001% O2 (10 ppm) is the range of severe hypoxia where cell survival is affected after a variable duration of hypoxia . 0.1% O2 is about the oxygen concentration where aerobic glucolysis yields to ATP production by anaerobic glucolysis (2, 10). The low-energy output by anaerobic glucolysis is in normal tissue and during short periods of severe hypoxia partly countered by increased glucose uptake by the cells, but not necessarily so in tumour tissue (17). Protein accumulation is greatly diminished, partly due to reduced overall protein synthesis as a result of phosphorylation of the eukaryotic initiation factor eIF2R and partly due to a slight upregulation of overall protein degradation (18). mRNA splicing may be altered at hypoxia (19). The inhibitor of HIF-1, iPAS, is a splicing variant of HIF-3. At the low end of severe hypoxia, mRNA translation / the most energy-costly process in the cell / is much suppressed, as is ATP production. Respiration is reduced, and there is a limit as to how long cells can stay alive. However, there are two regulatory LETTER TO THE EDITOR