HIFing the Brakes: Therapeutic Opportunities for Treatment of Human Malignancies

See allHide authors and affiliations

Science's STKE  30 May 2006:
Vol. 2006, Issue 337, pp. pe25
DOI: 10.1126/stke.3372006pe25


The unfortunate ability of tumor cells to survive and expand in an uncontrolled manner has captivated the attention of clinicians and basic scientists alike. The molecular mechanisms that tumor cells use to grow are the very same pathways used in normal cell growth and differentiation. One important pathway conferring a growth advantage on tumor cells is the epidermal growth factor receptor (EGFR) pathway. Signaling through the EGFR leads to activation of the phosphatidylinositol 3-kinase and Akt pathway and to increased activity of multiple effectors, including hypoxia-inducible factors (HIFs), which are cellular transcription factors involved in environmental stress response. The target genes that HIF members stimulate that are relevant to tumor growth include transcriptional activators and repressors and cytokines and growth factors, as well as their receptors. In this Perspective, findings from several recent studies are discussed in terms of their effect on the signal transducers, target genes, and tumor properties that are ultimately affected during EGFR-stimulated HIF signaling in cancer cells.

How do you stop a runaway freight train? Oncologists and cancer biologists face this dilemma in their daily clinical and research encounters. The development of novel therapeutic agents that slow or stop the growth of cancer remains a desirable, but elusive, goal for both the clinician and the bench scientist. Much attention has focused on influencing intracellular signal transduction pathways that regulate cell proliferation and growth. However, although a given signal transduction pathway may principally regulate cell growth in cell culture, the reality is that cross talk between multiple signal transduction pathways results in the unfortunate ability of cancer cells to usurp control of cell growth in many different ways. Yet hope remains that despite the increased complexity of signaling used by cancer cells in vivo, blockade of a select few components of the signal transduction pathway will set the stage for innovative tumor treatment strategies. Recent studies provide optimism about this statement.

The transcription factor known as hypoxia-inducible factor 1 (HIF-1) is of particular interest to cancer biologists. HIF-1, the prototypical founder of a family of stress-responsive transcription factors, was initially identified by Semenza and colleagues in the early 1990s and is composed of two polypetides: HIF-1α and HIF-1β (13). Two additional HIF α members, the closely related HIF-2α (47) and more distantly related HIF-3α (8), were subsequently identified with the completion of the human genome project. HIF activity is regulated at the posttranscriptional level by protein degradation of HIF α subunits (9, 10). After oxygen-dependent hydroxylation of specific proline residues, HIF α members are targeted for proteasome degradation by complexing with the von Hippel–Lindau (VHL) protein, a component of an E3 ubiquitin ligase complex. During hypoxia, the prolyl hydrolylases are inactive and the HIF α members are not complexed with VHL, thereby allowing for the formation of active HIF complexes.

Besides environmental stresses, growth factors also stimulate HIF signaling (11). Activation of HIF-1 signaling results in the induction of target genes involved in cellular proliferation and growth, cell survival or death, oxygen and nutrient delivery, and anaerobic energy metabolism. Because these are very important features for the growth of solid tumors, activation of HIF by growth factor signal transduction pathways is believed to be supportive, if not causative, in cancer (12). There are now numerous studies showing enhanced expression of the genes encoding HIF-1α and HIF-2α associated with specific malignancies (13). The malignancy in which roles for HIF members are most clearly delineated is renal clear-cell carcinoma (RCC). The majority of sporadic RCCs are associated with loss-of-function mutations in, or silencing of, the gene encoding VHL, an early event in the pathogenesis of RCC. Loss of VHL function results in constitutively high HIF-1α and HIF-2α levels, making RCC tumor biology of interest to HIF and VHL biologists alike (1416). Although HIF-1α was initially implicated as an etiologic agent in the development of RCC in which VHL was inactive (RCC VHL−/−) (1719), HIF-2α rather than HIF-1α is the culprit (20, 21). It should come as no surprise that academic and industrial investigators are intent on developing pharmacological agents that affect signaling at the extremes of HIF signal transduction pathways, both at the proximal part, with therapies affecting growth factor receptor activation, and at the distal part, with identification of key HIF target genes. One growth factor pathway of particular importance for tumor growth and HIF activation is the epidermal growth factor (EGF) pathway. Signaling through the EGF receptor (EGFR) is common in tumor cells, even in the absence of EGF, and ultimately results in the production of factors responsible for tumor cell growth and survival.

How is EGFR signaling in tumors maintained in the absence of EGF? Transforming growth factor–α (TGF-α), a cytokine produced by many tumor cell types, is a ligand for the EGFR (Fig. 1). TGF-α–activated EGFR signaling results in oxygen-independent activation of HIF signaling and subsequent induction of downstream HIF-dependent target genes, including the gene encoding vascular endothelial cell growth factor (VEGF). Worse yet, HIF-2α stimulates TGF-α production in VHL−/− RCC (2022), a result also observed in embryonic cells containing a knockin of HIF-2α into the HIF-1α locus (23), thereby establishing a positive feedback loop for EGFR signaling. For cancer cells without mutations in VHL, HIF-1α overexpression may stimulate TGF-α production by competing with VHL and indirectly elevating HIF-2α levels. Although this type of cross talk between HIF-1α and HIF-2α target genes accompanying HIF overexpression limits the ability of investigators to identify bona fide specific genes regulated by endogenous HIF-1α and HIF-2α, the overexpression studies are nevertheless relevant to RCC and other cancers involving enhanced levels of HIF signaling. Regardless of how TGF-α levels are increased, a bad situation is made worse for the patient when TGF-α–activated EGFR signaling, with its associated induction of HIF-dependent and HIF-independent cellular processes favoring tumor growth and metastasis, is combined with the proangiogenic and prosurvival effects of VEGF signaling, which are also stimulated by EGFR signaling.

Fig. 1.

Signal transduction pathway involving EGFR and HIF activation and induction of HIF target genes leading to cancer progression. Activation of the EGFR by EGF or TGF-α induces stabilization of HIF α proteins (HIF-1α and HIF-2α) and subsequent formation of active HIF-1 and HIF-2 complexes that can bind to and activate HREs in target genes, including those encoding transcriptional repressors (TCF3, ZFHX1A, and ZHX1B), cytokines (TGF-α), and growth factors (VEGF). Activation of the EGFR pathway also leads to an increased ability of Sp1 to induce VEGF gene expression. HIF-1 may also influence VEGF gene expression through protein-protein interactions with Sp1 (dotted line) rather than by direct binding to the HRE. Increased VEGF production leads to increased tumor angiogenesis and thereby increases oxygen delivery to the expanding tumor mass. The transcriptional repressors induced by HIF-1 prevent expression of the gene encoding E-cadherin, thereby favoring epithelial-mesenchymal transitions; whereas induction of TGF-α promotes establishment of a positive feedback loop favoring continued growth and expansion of tumor cells.

Is EGFR-induced VEGF gene expression solely mediated by HIF-1? There is evidence in support of HIF-dependent (24), as well as HIF-independent (25), mechanisms for the increase in VEGF abundance after EGF stimulation. EGFR signaling through the phosphatidylinositol 3-kinase (PI3K)–to–Akt (PI3K/Akt) pathway may stimulate VEGF expression through the transcription factor Sp1 (26). While investigating the mechanisms by which the EGFR tyrosine kinase inhibitors gefitinib (Iressa, AstraZeneca) and erlotinib (Tarceva, Genentech) were effective for treating squamous cell carcinomas, Maity and colleagues found two reasons for a reduction in VEGF abundance (27). The first was a decrease in HIF-1α protein synthesis; the second was modulation of the DNA binding activity of the transcription factor Sp1. Both of these effects were overcome by stimulation of the PI3K/Akt pathway. The authors proposed the existence of two separate signaling arms downstream of Akt, leading to increased VEGF gene expression with a HIF-1–dependent arm affecting HIF-1α protein stability, as well as a HIF-1–independent arm affecting Sp1 phosphorylation and hence DNA binding activity (Fig. 1).

Are there specific consequences of HIF activation that favor tumor metastases? E-cadherin is a protein that when decreased results in the epithelial-mesenchymal transition, an early step in carcinogenesis (28). Semenza and colleagues report that the repression of E-cadherin in VHL−/− RCC is due to a HIF-1α–dependent induction of three transcriptional repressors: TCF3 (also known as E12/E47), ZFHX1A (also known as yEF1 or ZEB1), and ZFHX1B (also known as SIP1 or ZEB2). Knockdown of HIF-1α results in restoration of normal cell-cell adhesion properties, results also obtained using dominant-negative HIF or by restoring VHL to the VHL−/− RCC. Research by Maxwell and colleagues also provides evidence that E-cadherin is repressed in a HIF-dependent manner (29). Overexpression of constitutively active HIF-1α or HIF-2α reduces E-cadherin abundance in VHL-competent cells, suggesting that either HIF α member may regulate E-cadherin abundance in the setting of HIF overexpression (Fig. 1). It is important to recognize that E-cadherin is certainly not the only metastatic factor regulated by HIF-1α during activation of EGFR signaling, as evidenced by the HIF-1–dependent regulation of the chemokine receptor CXCR4 in RCC (30) as well as non–small-cell lung cancer cells (31). Furthermore, VHL-dependent effects on growth factor signaling that are independent of HIF action may also contribute to tumor metastasis. VHL regulates fibroblast growth factor receptor (FGFR) endocytosis by a process involving the tumor suppressor Nm23; the absence of VHL results in elevated cell-surface FGFR levels, as well as increased cell migration, another mechanism that may be responsible for RCC metastasis (32).

The data support a broad role for HIF members in the development of malignancies. They also reveal how cancer cells have hijacked multiple signaling pathways, as well as key transcriptional regulators, so as to favor growth and proliferation, even in an environment depleted of growth factors and nutrients. Nevertheless, monotherapy that inhibits HIF signaling may yet prove effective in the treatment of certain human malignancies. The cross talk between HIF-1 and HIF-2 that occurs with elevated HIF-1α or HIF-2α levels, a possible potentiating factor in cancer progression, may be alleviated with nonselective HIF α antagonists. As for VEGF regulation, HIF may still play a role in Sp1-mediated VEGF gene expression, although this may not be mediated by traditional HIF signaling through the hypoxia-responsive enhancer element (HRE) (33). Nontraditional HIF signaling through protein-protein interactions with coactivators or other transcription factors, as well as indirect regulation by HIF target genes encoding transcriptional regulators such as the repressors identified by Semenza and colleagues, may partly explain the expanding list of target genes that HIF-1 apparently regulates (Fig. 1). One thing is for certain, elucidating the molecular mechanisms regulating cancer cell growth and progression continues to provide the necessary complement of brakes that will finally halt this runaway train.


  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.
View Abstract

Navigate This Article