Encyclopedia of Cancer, 2015
Being oxygen deprived.
Cellular hypoxia is a common stress in normal development and numerous pathological conditions, including cancer. By the time a tumor has grown to a detectable size, poor and disordered angiogenesis, leaky vessels, and high interstitial tumor pressure all result in significant tumor hypoxia. Studies in human tumor xenografts reveal a mean pO2 of <5 mmHg 70–80 mm from a vessel wall. Both noninvasive measurements and direct assessment of tumor oxygenation in patients have demonstrated the presence of profound hypoxia in a marked variety of solid tumors, including melanoma, prostate cancer, head and neck cancer, and cervical cancer, as well as in hematological malignancies including myeloma. In one study, using oxygen electrodes to assess breast cancer oxygenation in 36 patients and controls, the average pO2 was 30 mmHg in tumors as compared to 65 mmHg in normal breast tissue and 67 mmHg in benign fibrocystic breast disease. Areas of severe hypoxia or anoxia (<5.0 mmHg) were noted in >30 % of the measurements. There was no correlation between tumor hypoxia and tumor size or stage.
For over 30 years, hypoxic tumors have been known to be relatively chemoresistant and radioresistant. More recently, tumor hypoxia has been correlated with poor clinical response to radiation in cervical and breast cancer, chemoradiation in patients with head and neck cancer, and chemotherapy in myeloma. Proteins upregulated in hypoxic cells predict a poor prognosis when identified in a resected tumor or, in some cases, in patient serum. Intriguingly, prospective studies of head and neck and cervical cancer patients show that tumor hypoxia correlates with local/regional progression irrespective of whether surgery or radiation is applied as primary treatment. This suggests that while hypoxic tumors are less responsive to cytotoxic therapy, potentially due to drug delivery issues, they are also inherently more aggressive than non-hypoxic tumors.
Phenotypes of hypoxic cells
Several hypotheses could explain why hypoxic tumors may be less responsive to therapy and inherently more aggressive than non-hypoxic tumors (Fig. 1). These hypotheses are based primarily on the phenotypes of hypoxic neoplastic cells, for the most part studied in vitro. One hypothesis of why hypoxic tumors are more aggressive than non-hypoxic tumors is that hypoxia selects against cells with a proapoptotic phenotype. Hypoxia is a potent stimulus for apoptosis in a variety of tumor cell lines. Hypoxia-induced apoptosis can be blocked by BCL-2 overexpression or a variety of other mutations (including, in some cases, p53 mutations). Thus, only hypoxic cells with a strong antiapoptotic advantage survive, resulting in the selection of cells that are less responsive to therapy and a more aggressive tumor.
A second hypothesis for the aggressiveness of hypoxic tumors is that hypoxic cells are more prone to genomic instability, which may lead to inactivation of tumor suppressor genes or, conversely, to activation of oncogenes. Several studies have suggested that hypoxia downregulates several mismatch repair genes and that some hypoxic cells have an increased rate of mutagenesis. Genomic instability appears to be increased in cells that are rendered hypoxic and then reexposed to oxygen, perhaps due to DNA damage from the generation of reactive oxygen species. This hypoxia-reoxygenation pattern is known to occur in stroke and myocardial infarctions (i.e., reperfusion damage) but may also play an important role in the dynamic tissue bed of tumors.
An additional hypothesis of why hypoxic tumors are more aggressive than non-hypoxic tumors is based on in vivo and in vitro data demonstrating that while most non-transformed, differentiated cells do not proliferate when hypoxic, a variety of tumor cells can proliferate in these conditions. While hypoxia-induced growth arrest has been noted in several transformed, neoplastic cell lines, hypoxic proliferation can occur in some cervical cancer, osteosarcoma, and transformed rodent cell lines. p53 status does not affect proliferation of hypoxic cells, but other mutations in the retinoblastoma (Rb) axis allow cells to proliferate when moderately hypoxic (~1 % environmental oxygen); these mutations are well documented in a majority of tumors and include elevated CDK2 activity (due to cyclin E overexpression in breast cancer, increased cyclin D activity in lymphoma, and decreased p27 found in many cancers) and mutations in Rb. Thus, it is conceivable that hypoxic cells which can proliferate have cell cycle checkpoint abnormalities, and these abnormalities render these tumors more aggressive. A lack of differentiation is a hallmark of cancer, and hypoxia has also been demonstrated to regulate differentiation. These effects vary, as do proliferation and apoptosis, depending on the cell type. Some hypoxic cells, including some hematopoietic stem cells, are resistant to pharmacologically induced differentiation. Other cells undergo increased differentiation when hypoxic. Therefore, it is plausible that hypoxia correlates with absence of differentiation, hence aggressive tumorigenesis. A final hypothesis to explain the aggressiveness of hypoxia is that hypoxic tumors, through upregulation of vascular endothelial growth factor (VEGF) and other pro-angiogenic factors, are able to metastasize more rapidly than non-hypoxic tumors. Indeed, studies in cervical cancer have shown more metastatic disease in those patients with resected hypoxic tumors. Hypoxic cells have also been shown to upregulate genes that may play an important role in digesting the extracellular matrix and tumor cell invasion.
Gene regulation in hypoxic cells: implications for tumorigenesis
In an attempt to better understand the biology of hypoxic cells and hypoxic tumors, much work has focused into the delineation of the mechanisms by which cell signaling and gene regulation are altered in hypoxic cells and whether this may be extrapolated to hypoxic tumors. Several signaling events and transcription factors are generated in hypoxic cells, including NFkB, activation of p38, and fluxes in reactive oxygen species. However, the best studied regulator of hypoxia-induced genes is the transcription factor hypoxia-inducible factor 1 (HIF-1), a heterodimer HIF-1α and HIF-1b. HIF-1b may also heterodimerize with the HIF-1α-related proteins HIF-2a and HIF-3a, and while the individual roles of each of these are still being evaluated, HIF-2 may play an important role in tumorigenesis. HIF-1 was first characterized as the transcription factor responsible for upregulating erythropoietin, a growth factor for red blood cells that is induced during periods of anemia (and thus systemic hypoxia).
Under normoxic conditions, HIF-1α is hydroxylated by a prolyl hydroxylase, ubiquitinated, and degraded by the proteasome. In hypoxic cells, prolyl hydroxylases are inactive, and HIF-1α is stabilized. HIF-1α is overexpressed in renal cancer when the E3 ubiquitin ligase responsible for its degradation, the von Hippel-Lindau protein (VHL), is absent. In fact, renal cell cancers are often characterized by increased vasculature and associated with erythrocytosis, both features of HIF-1 target gene activation. HIF-1α is also stabilized by ras, AKT, Her-2 overexpression, and other oncogenic alterations often found in cancer (Fig. 2). Overexpression of HIF-1 is an independent indicator of poor prognosis in many malignancies.
Complete knockout of HIF-1α is embryonic lethal in mice. Original experiments done on embryonic stem cells suggested that apoptosis may be increased in hypoxic cells lacking HIF-1 function. In a mouse model of astrocytomas, HIF-1 was required for growth in a hypoxic environment, and HIF-1 in endothelial cells also seems necessary for solid tumor growth. Although the complete functional significance of HIF-1 in both normal cells and in cancer remains unclear, most consider HIF-1 activity as important in promoting the survival and adaptation of hypoxic tumors. Thus, the interference of HIF-1 activity is being actively pursued as a therapeutic target in cancer therapy. Because of the important, although unclear, role of HIF-1 in cancer, a major challenge in the field is to determine what genes are regulated by the HIF-1 transcription factor. The full range of HIF-1 targets is not known, but knockout studies have predicted anywhere from 100 targets to 2.6 % of the genome. It is clear that HIF-1α transactivates many genes necessary for hypoxic adaptation, including glucose transporters, glycolytic enzymes, and growth factors important for angiogenesis. It has recently been demonstrated that many HIF-1 targets such as VEGF and erythropoietin are protective against apoptotic cell death. Common HIF-1 targets are vital for tumor invasion, metastasis, and angiogenesis, and many of these targets, such as VEGF and LDH, are elevated in many aggressive cancers. One well-recognized result of hypoxia, and HIF-1 upregulation, is an increase in anaerobic (glycolytic) metabolism (glycolytic metabolism). The cancer cell preference to utilize anaerobic metabolism even in the presence of ample oxygen was first noted by Otto Warburg. This property of many malignant cells is an inefficient method of energy production, requiring increased utilization of glucose. In fact, the increased utilization of glucose by hypoxic and non-hypoxic tumors forms the basis for positron emission tomography (PET) scanning in cancer, where tumor cells may be recognized noninvasively through their high uptake of labeled glucose. Recent studies indicate that this increased metabolism of glucose even in normoxic tumors is due to increased activity of HIF-1, which is known to increase the expression of many of the enzymes involved in glycolysis. HIF-1 also inhibits flux through the Krebs cycle and thus suppresses aerobic metabolism directly. In addition to the well-recognized role hypoxia has on transcription, through HIF-1 and other transcription factors, there are other mechanisms that have been described for altering the expression of proteins important in the neoplastic process in hypoxic cells. The mRNA of an undetermined number of transcripts, including VEGF, is stabilized in hypoxic cells, although the mechanism(s) has not been entirely clarified. More recent studies have documented that hypoxia can suppress protein translation through the mTOR pathway and others. Misfolded proteins in the hypoxic cell’s endoplasmic reticulum also activate the unfolded protein response (UPR). The UPR is an efficient mechanism to halt protein translation, while paradoxically increasing the translation of a select number of mRNAs, including the transcription factor ATF-4. The hypoxic suppression of protein translation is thought to be a protective mechanism by which hypoxic cells conserve energy. Intriguingly, recent data suggests that some of these regulatory mechanisms may be altered in hypoxic neoplastic cells, potentially providing cancer cells another mechanism to manipulate the hypoxic environment.
Fig. 1. A model of how tumor hypoxia may affect the aggressiveness of these tumors. Hypoxia may promote deregulated proliferation and apoptosis, genomic instability, and differentiation and thus select for aggressive tumors
Fig. 2. The hypoxia-induced transcription factor, HIF-1, may be highly expressed in both hypoxic and normoxic tumors. This overexpression leads to the upregulation of a variety of genes that allow a tumor to survive and grow
Dang CV, Semenza GL (1999) Oncogenic alterations of metabolism. Trends Biochem Sci 24(2):68–72
Green SL, Giaccia AJ (1998) Tumor hypoxia and the cell cycle: implications for malignant progression and response to therapy. Cancer J Sci Am 4(4):218–223
Semenza GL (2003) Targeting HIF-1 for cancer therapy. Nat Rev Cancer 3(10):721–732