Seeking a Cellular Oxygen Sensor

The fundamental question of how cells sense oxygen has implications for embryogenesis, cancer, stroke, diabetes, and other ischemic diseases.

Jeffrey Perkel
May 13, 2001

The fundamental question of how cells sense oxygen has implications for embryogenesis, cancer, stroke, diabetes, and other ischemic diseases. Clearly, this is important work, and many researchers have taken up the task. Yet, despite the publication of hundreds of papers on this subject, no clear consensus exists regarding what the cellular oxygen sensor is, or even the number of sensing mechanisms there might be. 

The literature presents several possibilities. One theory holds that a heme-containing protein undergoes a conformational change when bound to oxygen, thereby "sensing" oxygen. Two other related hypotheses center around reactive oxygen species (ROS), which are highly unstable, highly reactive superoxides. One ROS theory holds that, as oxygen levels decrease, so do ROS levels. The second theory hypothesizes the opposite, countering that as oxygen levels decrease, ROS levels increase. Each theory has its supporters and detractors, who have published many articles to advance their particular vantage point.1

Recently published data are adding new layers to the discussions. Two new manuscripts, its authors working independently, report that the ability of cells to regulate a factor critical to the hypoxic response depends on a specific post-translational modification.2,3 One interpretation of the data would be that the authors have found a cellular oxygen sensing mechanism, according to one researcher. But some scientists in the field debate the significance of these findings, questioning whether the actual sensory system has been identified. Compounding the argument, evidence exists to support multiple oxygen sensing mechanisms, with one expert suggesting that some or all of the theories might be correct.1

When cells are exposed to low oxygen concentrations (hypoxia), they elevate expressions of genes that help the cells and tissues cope with the hypoxic condition. For example, hypoxic cells upregulate erythropoietin, which stimulates red blood cell production, thereby increasing the blood's oxygen-carrying capacity. Likewise, hypoxic cells increase expression of vascular endothelial growth factor (VEGF), stimulating vascular branching to improve tissue vascularization and oxygenation. These cells also upregulate glycolytic enzymes to help them cope with the transition from an aerobic to an anaerobic environment.

At the molecular level, the hypoxia-inducible factor (HIF) regulates these genes. HIF complexes contain two proteins, HIF-aand HIF-ß. The asubunit is degraded rapidly under normal oxygen concentrations, but it is stabilized under hypoxia. Thus, oxygen-regulated genes are expressed only under hypoxia.1

The product of the von Hippel-Lindau (VHL) tumor suppressor gene pVHL regulates the stability of HIF-aproteins;4 pVHL is a member of a E3-ubiquitin ligase complex that conjugates ubiquitin proteins to HIF-1 and targets it for proteasomal degradation. Last year, researchers in the laboratories of Bill Kaelin at the Dana-Farber Cancer Institute in Boston, and Peter Ratcliffe, Christopher Pugh, Patrick Maxwell, and colleagues at Oxford University in the United Kingdom, observed nearly simultaneously that pVHL and HIF directly interact, and that the pVHL complex ubiquitinates HIF.5,6 Thus, pVHL is the recognition component of the ubiquitin ligase complex that regulates HIF turnover. But just how pVHL "knew" when to target HIF for destruction remained a mystery. Now, two papers published in Science shed some light on this question.5,6

Researcher Mircea Ivan and colleagues in Kaelin's laboratory, as well as Panu Jaakkola and colleagues in the labs of Ratcliffe, Pugh, and Maxwell, simultaneously published evidence that a post-translational modification of HIF-atargets it for degradation.2,3 Both labs show that pVHL only can interact with HIF-asubunits if they are post-translationally modified by a prolyl hydroxylase (PH). This HIF-PH adds a hydroxyl (-OH) group to proline 564, located within the oxygen-dependent degradation domain (ODD) of HIF-1*, in an iron- and oxygen-dependent manner. Kaelin describes this observation as "very satisfying," because it fits neatly with the fact that HIF stability is influenced by both cellular iron availability and oxygen levels. However, neither paper identified nor isolated any particular protein.

Interestingly, while both papers reach the same conclusion, they arrived at that point by asking different questions and adopting different approaches. According to Ratcliffe, this fact "reflects the fact that if two scientists are presented with a particular problem in complete independence, they will address it in different ways." Ratcliffe's group was interested in the oxygen-dependent regulation of erythropoietin. That question led them to HIF, and then to the interaction between pVHL and HIF. In contrast, Kaelin's lab concentrated on the pVHL's function. Thus, they have characterized the members of the pVHL complex, as well as its function. And now, both groups have identified a potentially new regulatory paradigm.

But is this PH activity the cellular oxygen sensor? H. Franklin Bunn, of the Brigham and Women's Hospital in Boston, wrote a commentary accompanying the two research articles.7 He defines the oxygen sensor as the "proximate recorder of oxygen tension within the cell ... the molecule that will pick up changes in the pressure of oxygen in the cell and transduce a signal." The HIF-PH could be the molecular oxygen sensor. Alternatively, the sensor might regulate HIF-PH. Bunn offers a conservative analysis of these data. "I would say they've identified the process by which the HIF transcription factor is post-translationally modified allowing the protein to bind to von Hippel-Lindau protein and to be ... degraded. But that doesn't mean they've isolated the sensor."

These findings may have an impact on other oxygen-sensing theories. According to Bunn, Kaelin's and Ratcliffe's data is difficult to reconcile with one theory of oxygen sensing that argues as oxygen concentration decreases, ROS concentration increases, signaling a cascade that stabilizes HIF. However, one adherent of this theory, Paul Schumacker of the University of Chicago sees no inherent contradiction.8 He notes that although these two papers clearly implicate an iron- and oxygen-dependent PH in HIF regulation, they do not prove that oxygen availability regulates PH activity. That is, although the PH requires oxygen to function, it does not necessarily follow that oxygen concentration directly modulates enzymatic activity. In fact, Schumacker says that regulatory events, such as phosphorylation, often control the activities of many enzyme systems. Moreover, he continues, it is difficult to imagine that an enzyme exists that is regulated so dramatically over so narrow a substrate concentration range.

Numerous mechanisms could regulate HIF-1 activity, both within single cell types, and between cell types. Gregg Semenza is a Johns Hopkins University School of Medicine researcher who characterized and cloned HIF-1aand helped establish its role in embryogenesis.9-12 HIF "is such a critical factor in the cell that I think it is regulated through many different mechanisms, and I don't think that we should [eliminate] all other potential mechanisms on the basis of the identification of this particular [one]," he says.

Schumacker concurs. He suggests, for example, that ROS signals generated by an oxygen sensor could activate a kinase or phosphatase that downregulates HIF-PH function. Moreover, he notes that the PH model clearly does not regulate several well-known hypoxic responses. For example, the release of neurotransmitters by glomus cells in the carotid body is an oxygen-dependent response that does not involve HIF. Cells likely have numerous oxygen-sensing mechanisms, and Schumacker believes that ROS could be the upstream, "universal" signal generated by the oxygen sensor.

The most important follow-up question, therefore, is whether the HIF-PH is the cellular oxygen sensor. The next step would be to identify the specific enzyme that acts upon HIF. The well-characterized PH enzymes, collagen prolyl 4-hydroxylases, do not appear to fit the bill. Other questions include: Do other substrates exist of pVHL in addition to HIF and does prolyl hydroxylation regulate them? Is prolyl hydroxylation used as a signal in other contexts, or is it unique to the pVHL system? And, do other prolyl hydroxylation sites live within HIF?

Semenza, a well-respected researcher in this field, sums up the significance of the Kaelin and Ratcliffe work: "It's a major advance in the field because it seems to indicate that oxygen can directly modulate the activity of HIF-1 [without intervening steps]...[but] I would not really rule anything out at this point."

Jeffrey M. Perkel can be contacted at
1. G.L. Semenza, "Regulation of mammalian O2 homeostasis by hypoxia-inducible factor 1," Annual Review of Cell and Developmental Biology, 15:551-78, 1999.

2. M. Ivan et al., "HIFatargeted for VHL-mediated destruction by proline hydroxylation: Implications for O2 sensing," Science, 292:464-8, April 20, 2001.

3. P. Jaakkola, et al., "Targeting of HIF-ato the von Hippel-Lindau ubiquitylation complex by O2-regulated prolyl hydroxylation," Science, 292:468-72, April 20, 2001.

4. J.F. Wilson, "Understanding the VHL tumor suppressor complex," The Scientist, 15[9]:20, April 30, 2001.

5. M.E. Cockman et al., "Hypoxia inducible factor-alpha binding and ubiquitylation by the von Hippel-Lindau tumor suppressor protein," Journal of Biological Chemistry, 275:25733-41, Aug. 18, 2000.

6. M. Ohh et al., "Ubiquitination of hypoxia-inducible factor requires direct binding to the beta-domain of the von Hippel-Lindau protein," Nature Cell Biology, 2:423-7, July 2, 2000.

7. H. Zhu, H.F. Bunn, "How do cells sense oxygen?" Science, 292:449-51, April 20, 2001.

8. N.S. Chandel et al., "Mitochondrial reactive oxygen species trigger hypoxia-induced transcription," Proceedings of the National Academy of Sciences, 95:11715-20, Sept. 29, 1998.

9. G.L. Semenza, G.L. Wang, "A nuclear factor induced by hypoxia via de novo protein synthesis binds to the human erythropoietin gene enhancer at a site required for transcriptional activation," Molecular and Cellular Biology, 12:5447-54, 1992.

10. G.L. Wang, G.L. Semenza, "Purification and characterization of hypoxia-inducible factor 1," Journal of Biological Chemistry, 270:1230-7, 1995.

11. G.L. Wang et al., "Hypoxia-inducible factor 1 is a basic helix-loop-helix-PAS heterodimer regulated by cellular O2 tension," Proceedings of the National Academy of Sciences, 92:5510-4, 1995.

12. N.V. Iyer et al., "Cellular and developmental control of O2 homeostasis by hypoxia-inducible factor 1*," Genes and Development, 12:149-62, 1998.