Oxygen Sensing: Getting Pumped by Sterols

Science's STKE  21 Jun 2005:
Vol. 2005, Issue 289, pp. pe30
DOI: 10.1126/stke.2892005pe30


Oxygen plays a pivotal role in the maintenance of life for all eukaryotes, with the exception of strict anaerobes. Eukaryotes have developed mechanisms to sense and respond to decreased oxygen levels. How eukaryotes sense oxygen is still not fully understood. What is (or are) the oxygen sensor(s)? This question has vital physiological and pathophysiological implications, because all living aerobic organisms have adaptive mechanisms to maintain oxygen homeostasis. A recent report describes a novel eukaryotic oxygen-sensing mechanism in the fission yeast Schizosaccharomyces pombe, involving the depletion of sterols as a trigger to induce gene expression in response to decreased oxygen levels. It is not yet clear whether this mechanism is involved in the mammalian response to hypoxia, possibly in conjunction with activation of one or both of the hypoxia-inducible factor (HIF-1 or HIF-2) transcription factors.

For most eukaryotes, molecular oxygen is central to survival. Oxygen is the terminal acceptor of electrons in the respiratory chain and is required for oxidative phosphorylation. Oxygen is also required for the synthesis of sterols, heme, and fatty acids. Thus, it is not surprising that many of the genes whose expression is affected by conditions of low oxygen concentration (hypoxia) or oxygen deprivation (anoxia) are involved in cellular bioenergetics, sterol, heme, and fatty acid synthesis (1). Much progress has been made in identifying the transcription factors that regulate the transcription of genes that are induced or repressed during hypoxia or anoxia. However, the fundamental question of how cells sense a decrease in oxygen concentration to regulate gene expression is not fully understood.

Yeast has been used as a model system to examine mechanisms of oxygen sensing in eukaryotes. In a recent report, Hughes et al. describe a novel eukaryotic oxygen-sensing mechanism in the fission yeast Schizosaccharomyces pombe (2). They uncovered a transcription factor, Sre1, that is activated by either sterol or oxygen depletion to induce genes involved in glycolytic metabolism as well as in sterol and heme biosynthesis. This raises the possibility that Sre1 responds to oxygen availability by means of oxygen-dependent sterol synthesis.

Oxygen Sensing in Saccharomyces cerevisiae

To date, most studies of oxygen-regulated gene expression in yeast have been conducted in S. cerevisiae (3). The oxygen-regulated genes in this yeast are classified into two categories: "aerobic genes," which are abundant at room air conditions but are suppressed under anoxic conditions; and "hypoxic genes," which are repressed under room air conditions but are induced slightly by hypoxia and maximally by anoxia (4). Some of the hypoxic genes include ERG11, OLE1, and HEM13, which encode enzymes required for biosynthesis of sterols, unsaturated fatty acids, and heme, respectively.

At present, there are two known mechanisms by which S. cerevisiae transduces decreased oxygen concentrations to the induction of gene expression. The first involves intracellular heme, which activates the transcription factor Hap1 (heme-activated protein 1) to stimulate the transcription of the ROX1 gene. ROX1 is a transcriptional repressor for a subset of hypoxic genes, including ERG11 and HEM13 (59). Heme abundance decreases under anaerobic conditions because heme biosynthesis requires molecular oxygen (10). This results in diminished Hap1 activity and ROX1 gene expression, which leads to derepression of hypoxic genes (Fig. 1A). The second mechanism requires the mitochondrial respiratory chain (Fig. 1A) (11). The mitochondrial respiratory chain is required for the hypoxic activation of the transcription factor Mga2, which stimulates the transcription of subsets of hypoxic genes, such as OLE1, by binding to the low-oxygen response element (LORE) in the promoter of these genes (1214). OLE1 induction does not occur in ρ° strains of yeast, which lack a mitochondrial genome and are respiratory incompetent. In contrast, the induction of heme-Hap1–regulated genes is maintained in ρ° strains (11). The signaling mechanism that couples the respiratory chain to the induction of genes such as OLE1 remains unknown. However, it has been postulated that increases in mitochondria-generated reactive oxygen species (ROS) during the shift from aerobic to anaerobic conditions might serve as a signal to induce gene expression (15).

Fig. 1.

Models for oxygen sensing in yeast. (A) Different oxygen-sensing pathways are involved in modulating the expression of two classes of hypoxic genes in Saccharomyces cerevisiae. The biosynthesis of heme requires oxygen; therefore, during anoxia or 0.2% O2, heme concentration decreases and the transcription of a set of hypoxic genes, including HEM13 and ERG11, is derepressed in a Hap1-dependent manner (left). A second pathway requires the mitochondrial electron transport chain, which activates Mga2. Mga2 in turn activates low-oxygen response element (LORE)–dependent hypoxic genes such as OLE1 (right). (B) To date, one oxygen-sensing pathway has been identified in Schizosaccharomyces pombe. Under anoxia or 0.2% O2, sterol concentration decreases, which in turn increases activation of the transcription factor Sre1. Up-regulation of Sre1 stimulates the transcription of genes required for growth under anaerobic conditions, such as the glycolytic regulator 6-phosphofructo-2-kinase (PFK26).

Oxygen Sensing in Saccharomyces pombe

In contrast to the abundant data on oxygen-regulated gene expression in S. cerevisiae, limited information is available on hypoxic gene expression in the fission yeast S. pombe (3). Hughes and colleagues describe a novel mechanism through which S. pombe induces hypoxic genes (2). The investigators were searching for yeast homologs for the sterol regulatory element binding proteins (SREBPs), mammalian transcription factors that control lipid homeostasis. In mammalian cells depleted of sterols, SREBPs transcriptionally activate a number of enzymes involved in fatty acid, triacylglycerol, cholesterol, and phospholipid synthesis [for a review, see (16, 17)]. SREBPs are synthesized as inactive precursors and are anchored by SREBP cleavage-activating protein (SCAP) to the endoplasmic reticulum (ER). In sterol-depleted cells, SCAP escorts SREBPs from the ER to the Golgi, where the SREBPs are cleaved. This allows the N-terminal segment of the SREBP to be released from the Golgi membrane to enter the nucleus and activate transcription. How sterol depletion is sensed to activate SREBPs remains unknown.

After sequence searches failed to identify a SREBP homolog in S. cerevisiae, the authors turned to the S. pombe database. They identified two genes in S. pombe that were homologous to the human SREBP-1a, SRE1 and SRE2, and a single homolog to human SCAP, SCP-1. They demonstrated that, as with human SREBP, sterol depletion activated the cleavage of Sre1, but not Sre2, in a Scp-1–dependent manner. Microarray data revealed a host of genes required for biosynthesis of sterols and heme that were regulated by Sre1 in response to sterol depletion. Surprisingly, some hypoxic genes were induced by Sre1 during sterol depletion. Cleavage of Sre1 in a Scp-1–dependent manner was observed with 0.2% or 0% O2, and the loss of Sre1 resulted in failure of S. pombe to grow under anaerobic conditions. Because sterol synthesis was reduced by a factor of 5 under hypoxic conditions and by a factor of 30 under anoxic conditions, Hughes et al. proposed that hypoxia or anoxia depletes sterol levels, which leads to cleavage of Sre1 and results in the induction of Sre1-dependent genes involved in adaptation to anaerobic conditions (Fig. 1B).

Hypoxic Gene Expression in Mammalian Cells

It remains to be determined whether SREBPs control gene expression in mammalian cells exposed to hypoxic or anoxic conditions. Many of the genes induced during hypoxia or anoxia in mammalian cells are regulated by the transcription factors HIF-1 or HIF-2 (hypoxia-inducible factor 1 or 2) [for a review, see (18)]. HIFs are heterodimeric transcription factors that consist of HIF-1α or -2α and HIF-1β subunits (19). HIF activity depends on the availability of the α subunit, which is tightly controlled by cellular oxygen tension (20, 21). Under normal oxygen conditions (21% O2), the α subunit is polyubiquitinated and targeted for degradation by an E3 ubiquitin ligase complex that contains the von Hippel–Lindau tumor suppressor protein (pVHL) (22, 23). This degradation is regulated by hydroxylation of specific α-subunit proline residues by a family of prolyl hydroxylases (PHDs), which allows binding of the α subunit to pVHL (2427). Additionally, the α subunit is hydroxylated at an asparginine residue by factor inhibiting HIF-1 (FIH-1) (2831). FIH-1–dependent hydroxylation prevents binding of the coactivators p300 and CBP to the α subunit, thus inhibiting HIF-mediated gene transcription. The rate of hydroxylation decreases as oxygen concentration decreases (26). Consequently, pVHL cannot target the α subunit for degradation, which allows the α subunit to accumulate and dimerize with HIF-1β to target genes. Many HIF target genes have been identified and are involved in angiogenesis, cell survival, glucose metabolism, and invasion. It will be interesting to see whether SREBPs in mammalian cells induce a set of genes distinct from HIF-dependent genes during hypoxic or anoxic conditions and whether sterol levels affect hypoxic or anoxic gene expression.

The study by Hughes et al. also implies that sterol abundance might regulate oxygen sensing in mammalian cells. However, when comparing oxygen-sensing mechanisms in yeast to those in mammals, one must recognize that mammalian cells and yeast induce hypoxic gene transcription at different oxygen concentrations. HIFs begin to be stabilized and activated at around 5% O2, and their activity increases exponentially as oxygen concentration diminishes further (32). Mammalian cells need to sense oxygen and activate hypoxic gene expression at oxygen concentrations well above anoxia, because anoxia induces apoptotic cell death in mammalian cells (33, 34). Hypoxia (0.5% to 5% O2) does not induce cell death but can activate proangiogenic genes, such as vascular endothelial growth factor (VEGF), that enable cells to replenish oxygen levels and prevent them from experiencing anoxia. By contrast, yeast are facultative anaerobes and are capable of surviving by respiration when oxygen is available or by fermentation when oxygen is depleted. Thus, hypoxic gene induction in S. cerevisiae only begins to increase at around 0.2% O2 and is maximal under anoxic conditions (35).

Hughes et al. measured Sre1 activity only at 0.2% O2 or during anoxia. They did not measure Sre1 activity over a full range of oxygen concentrations. Because sterol synthesis is limited by oxygen concentration only at near-anoxic conditions, it is likely that Sre1 activity does not increase until oxygen concentrations drop to about 0.2%. Thus, SREBP in mammalian cells might become activated only under conditions close to anoxia in which sterol synthesis is limited by oxygen availability. This is analogous to activation of the HIFs under anoxic conditions, in which lack of oxygen availability prevents PHDs and FIH-1 from hydroxylating them.

Oxygen Sensing in Mammalian Cells

How hypoxia (0.5 to 5% O2) is detected by an oxygen sensor (or sensors) to activate HIF-dependent gene expression in mammalian cells is not fully understood. Theoretically, the PHDs and FIH-1 could serve as oxygen sensors during hypoxia. Their activities decrease linearly with oxygen concentration from ambient air to anoxia (36, 37). This implies that HIFs would be activated throughout the physiological oxygen range that mammalian cells encounter. However, HIF-1α protein accumulation displays a threshold phenomenon: Protein stabilization does not occur until oxygen levels diminish to 5% (32). Thus, the oxygen dependency of hydroxylase enzymatic activity does not correlate with the oxygen dependency of HIF-1 activation. It is likely that the hydroxylases have sufficient oxygen to carry out hydroxylation of the protein from ambient air to near-anoxic conditions, but that activation of intracellular signaling pathways that are invoked at oxygen levels around 5% begins to prevent the hydroxylases from hydroxylating HIFs.

Numerous intracellular signaling pathways have been implicated in HIF-dependent gene expression (38). Most of these signaling pathways are not directly regulated by molecular oxygen, indicating that an upstream oxygen sensor must activate them. The mitochondrial respiratory chain has been proposed to act as an oxygen sensor that regulates HIF-dependent gene expression during hypoxia through activation of oxidant-dependent intracellular signaling (39, 40). The respiratory chain begins to increase the production of ROS at 5% O2, which are both required and sufficient to activate HIF-1 (Fig. 2). Evidence to support this model comes from the observation that respiration-deficient cells or cells treated with mitochondrial inhibitors fail to generate ROS and activate HIF-1 during hypoxia (3942). The increase in mitochondrial ROS during hypoxia stimulates the p38α mitogen-activated protein kinase (MAPK) signaling pathway to induce HIF-dependent gene expression (43). In contrast, neither a functional electron transport chain nor p38 MAPK are required for anoxic activation of HIF-1 (43, 44). Anoxia directly limits oxygen availability for hydroxylation to occur; thus, the PHDs and FIH-1 are likely to serve as direct oxygen sensors regulating anoxic HIF-dependent gene expression (Fig. 2). It is not clear how mitochondrial ROS-dependent signaling prevents the PHDs or FIH-1 from hydroxylating the HIF-1α protein during hypoxia.

Fig. 2.

Model for oxygen sensing in mammals. Mitochondria have been implicated as potential oxygen sensors that activate HIF-1 during hypoxia by increasing their generation of reactive oxygen species (ROS). ROS can activate signaling pathways that in turn may alter or inhibit the enzymatic activity of hydroxylases and thereby activate HIF-1. During total oxygen deprivation, or anoxia, hydroxylase activity is blocked directly because of the absence of molecular oxygen.

In the past decade, much progress has been made in identifying key transcription factors that control oxygen-regulated gene expression in yeast and mammalian cells. Anoxia regulates gene transcription in yeast because limited oxygen availability prevents heme or sterol biosynthesis, which regulates transcription factors such as Hap1 or Sre1, respectively. In mammalian cells, lack of oxygen during anoxia prevents hydroxylation of HIFs, thus allowing these transcription factors to be activated. In contrast, the mechanisms of oxygen sensing that regulate hypoxic gene expression are less clear. To date, studies in yeast have shown that gene expression is regulated at oxygen concentrations in the range of 0 to 0.2%, thus precluding them as an amenable model with which to dissect mechanisms of oxygen sensing that regulate gene expression in the range of mammalian levels of hypoxia (0.5 to 5% O2). However, if future experiments indicated that Sre1-dependent gene transcription is activated in the hypoxic range of 0.5 to 5% O2 in S. pombe, then this simple genetic model could be used to dissect mechanisms of oxygen sensing during hypoxia. In any event, S. pombe can now be used to genetically dissect how sterol depletion is sensed to activate Sre1-dependent gene expression.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
View Abstract

Navigate This Article