| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Molecular Pathways |
Authors' Affiliation: Cancer Research UK, Weatherall Institute of Molecular Medicine, University of Oxford, John Radcliffe Hospital, Oxford, United Kingdom
Requests for reprints: Adrian L. Harris, Cancer Research UK, Weatherall Institute of Molecular Medicine, University of Oxford, John Radcliffe Hospital, Oxford OX3 9DS, United Kingdom. Phone: 44-1865226184; Fax: 44-1865226179; E-mail: aharris.lab{at}cancer.org.uk.
Abstract
The lack of oxygen delivery to tumor cells has profound consequences for tumor growth and correlates with poor prognosis. Some tumors contain regions of very severe hypoxia called anoxia, which constitutes a functionally different state to hypoxia. In response to anoxia, mammalian cells induce coordinated cytoprotective programs that are critical for tumor survival: the unfolded protein response and integrated stress response. Therefore, targeting additional components of anoxic pathways, besides the hypoxia-inducible response, may be effective for future anticancer therapies.
and the aryl hydrocarbon receptor nuclear translocator (reviewed in ref. 2). HIF1 is activated by a broad range of hypoxia and switches on the mammalian gene transcription program that protects cells from the damaging effect of oxygen stress (3). HIF1 contributes to these survival measurements by increasing angiogenesis and anaerobic glycolysis, therefore promoting tumor growth (4, 5). Therefore, hypoxic regions in tumors in general correlate with poor prognosis, radiotherapy and chemotherapy resistance, and increased metastatic potential (6). However, in addition to mild hypoxia (0.01-2% O2), some tumors contain regions of severe hypoxia called anoxia (<0.01% O2). During the past 10 years, there has been increasing evidences that anoxia is functionally different state to hypoxia (7).
Unfolded protein response is a main consequence of anoxic exposure. Under anoxia, cells encounter conditions that interfere with the glycosylation and disulfide bonding that are required for normal folding, maturation, and function of proteins that traverse the secretory pathway. In eukaryotes, oxidative protein folding occurs in the endoplasmic reticulum (ER). Formation of disulfide bonds is catalyzed by protein disulfide isomerase and ER oxidoreductin 1 that acts as an electron relay system for oxidative folding. ER oxidoreductin 1 recharges protein disulfide isomerase, whereby its cofactor flavin adenine dinucleotide (FAD) is reduced to FADH2. ER oxidoreductin 1 then uses molecular oxygen as terminal electron acceptor at the end of this relay system, converting FADH2 into FAD and H2O, providing the driving force for protein folding in the ER (8). ER oxidoreductin 1 is induced by HIF, providing a link between the hypoxia and anoxia responses (9). As a result of accumulation of these misfolded proteins, a coordinated, cytoprotective, and HIF1-independent program is induced called the unfolded protein response (UPR; ref. 10).
UPR is an intracellular signaling pathway that relays signals from the lumen of the ER to activate target genes in the nucleus. A class of novel ER transmembrane receptors, including Ire1, pancreatic ER eukaryotic translation initiation factor (eIF) 2 kinase (PERK), and activating transcription factor (ATF) 6, mediates UPR signal transduction (Fig. 1
). A recent study has shown that components of this pathway are critical for tumor survival (11). The cytoprotective components of UPR include a marked decrease in overall protein synthesis, increased expression of ER chaperones, and arrest in the G1 phase of the cell cycle. In addition, UPR allows eukaryotic cells to increase their proteasome degradation capacity to unload themselves of improperly folded proteins. Eventually, if ER homeostasis cannot be reestablished, proapoptotic signaling pathways eliminate those chronically stressed cells (12).
|
In addition, under chronic anoxia disruption of cap-binding complex, eIF4F was recently proposed to be partially responsible for repression in protein synthesis (14). Finally, the majority of stalled transcripts accumulate in discrete foci called stress granules (15). Stress granules are dense aggregations, 100 to 200 nm in size in the cytosol, composed of proteins and RNAs, not surrounded by membrane, and associated with the ER. Stress granules also contain components of preinitiation complex, including RNA-binding proteins TIA1 and TIAR. These therefore may represent protected foci for rapid reinitiation of protein synthesis after removal of stress. Stress granules might also function as a decision point for untranslated mRNAs to head for further storage or degradation.
PERK is a ER stress-sensing kinase and increases tolerance to extreme hypoxia. Phosphorylation of eIF2 under anoxia involves ER-resident kinase PERK, and in the absence of PERK, cells become more sensitive to oxygen stress (16). PERK activity is controlled by ER Hsp70 chaperone bone-inducing protein (BIP) and Hsp40 chaperone P58IPK that block its activity (16). When unfolded proteins accumulate in ER, bone-inducing protein dissociates from PERK by binding these proteins. Unbound PERK oligomerizes and is activated by trans-autophosphorylation. A feedback inhibitory mechanism involves another UPR-activated gene, GADD34, which recruits the protein phosphatase PP1c to eIF2, which dephosphorylates eIF2 and restores translation (17, 18).
ATF4 is up-regulated in response to anoxia. Although translation of most mRNAs under anoxia is attenuated, some transcripts are preferentially translated. One of those is ATF4/cyclic AMP (cAMP)-responsive element binding protein 2 (19) that regulates the integrated stress response (ISR), a signaling pathway initiated by phosphorylation of eIF2, and protects cells against metabolic consequences of ER stress. This basic region-leucine zipper (bZip) transcription factor is primarily induced under anoxia, and its induction does not involve HIF1 or electron transport (20). ATF4 binds to cAMP response sites and can activate the ISR. ATF4 induction involves two short upstream open reading frames in the 5'-untranslated region. In the unstressed cells, a second uncoding open reading frame overlapping with the ATF4 open reading frame has repressive function and prevents downstream translation of ATF4. However, under anoxia, the UPR increases the level of eIF2 phosphorylation, reducing translation initiation at the second uncoding open reading frame and favoring reinitiation at ATF4 over inhibitory uncoding open reading frames (21). Ribosomal scanning is one of the best documented effects regulated by phosphorylation of eIF2; however, some evidence suggests that it also activates translation from some internal ribosomal entry sites and the use of alternative AUG start codons (22).
There is also another posttranslational mechanism of ATF4 regulation by its reduced proteolysis rate under anoxia. Under normoxia, specific phosphorylation of ATF4 at Ser219 promotes interaction with the SCFßTrCP ubiquitin ligase complex that leads to proteasome degradation (23).
Transcriptional induction under anoxia is also mediated by XBP1 and ATF6. Two other main components of ISR are XBP1 and ATF6. Unfolded proteins bind and sequester bone-inducing protein and prevent its interaction with ER luminal domains of Ire1 and ATF6 similarly to PERK. The former kinase is thus activated, and the latter is enabled to translocate.
The transcription factor XBP1 was identified as a target of two transmembrane ER kinases: Ire1a and Ire1b (24). These two kinases are composed of luminal-sensing domain and COOH-terminal catalytic domain with endonuclease activity. The Ire1 catalyses a unique splicing event that removes 26 nucleotides from XBP1 mRNA followed by a religation that alters the XBP1 reading frame (25). The remodeled XBP1 protein possesses the original DNA-binding domain tethered to a new transactivation domain.
The second is the bZIP transcription factor ATF6. ATF6 also encodes a transmembrane protein with ER stress-sensing lumen domain. In response to stress, ATF6 translocates from the ER to the Golgi where it is cleaved by two proteases, SP1 and SP2 (26), liberating as a cytosolic DNA-binding domain that translocates to the nucleus.
ISR as a culmination of cellular adaptation to anoxia. The above three transcription factors are implicated in the coordinated gene expression program called the ISR (27). The contribution of the ISR transcription factors XBP1, ATF6, and ATF4 to UPR target gene expression was determined by microarray analysis (27–29). ISR-dependent genes encode ER chaperones, folding enzymes, and components of ER protein degradation apparatus. They promote amino acid sufficiency and redox homeostasis and therefore support cell survival under anoxia. The ISR may also have protective role against oxidative stress, as improper vascularization of tumors can cause dynamic periods of altered oxygenation sufficient to induce DNA damage. ATF4 can function as a homodimer or heterodimer with members of the CAAT/enhancer binding protein family of transcription factors, such as the apoptosis-inducing protein CHOP, CAAT/enhancer binding protein , ß, and 
, and other bZIP transcription factors induced by anoxia, such as ATF3 and ATF6 (30). Although the ISR promotes several genes that enhance survival under anoxia, the CHOP/GADD153 transcription factor promotes cell cycle arrest and cell death. Prolonged anoxia, but not hypoxia, promotes tumor cell toxicity and apoptotic death independent of HIF1 status (31).
Clinical-translational advances. Substantial work already indicates that the hypoxia response by HIF1 is important for tumor growth in vivo and is a therapeutic target (5). However, the effects of HIF1 inhibition are variable, and sometimes, growth can be stimulated and it does not completely block tumor growth (32). In these experiments, the anoxia response may still be intact. Xenograft studies knocking out various components of the ISR, including ATF4, ATF6, XBP1, and PERK, indicate more profound inhibition of the tumor growth (11, 28, 33). An intriguing suggestion therefore is that anoxia, occurring even in very small numbers of cells implanted in these models, is probably more important than the hypoxia response in survival. Tumors will contain a mixture of hypoxic and anoxic components, and targeting both will need separate strategies. Being able to classify tumors prospectively into anoxic and hypoxic compartments, with some idea of the relevant proportions of these, will be of interest in future therapeutic studies. Although it may be thought that the anoxic cells will not contribute to growth, the results of xenograft studies above indicate that this stress response is critical for tumor growth.
It is likely that specific targeting will be possible without causing more tissue toxicity because knockout mice for ATF4 and some of the other components, such as PERK, are viable (34–36). Inhibitors of PERK kinase or antagonists of ATF4 should be considered for therapeutic development. Particularly, selective PERK kinase inhibitors should be achievable, considering its unusual structure with a stress-sensing ER luminal domain connected by a transmembrane segment to a cytoplasmic effector domain. Because of the interaction of ATFs with cAMP-binding–dependent transcription factors, there are likely to be small molecules already available that modify these pathways.
Recent gene profiling has highlighted a hypoxia profile (37) that is of prognostic significance, and it would be of interest to see if such arrays can be deconvoluted to describe an anoxic as well as a hypoxic component. From clinical studies staining for the expression of ATF4, this looks like only a minority of cells are positive around areas of necrosis, but xenograft work would suggest that these pathways are quite critical early on in tumor development and presumably also in metastasis. It will be important to investigate this in clinical samples and to compare the coexpression of hypoxic and anoxic signaling pathways, their synergy, and contribution to outcome. Ultimately, targeting both components of anoxia and hypoxia responses may be more effective than targeting either alone, and this is likely to be relevant to all modalities of therapy as is hypoxia.
Received 8/24/06; revised 10/30/06; accepted 11/22/06.
References
plays a key role in a HIF-1-mediated pathway to improve disulfide bond formation and VEGF secretion under hypoxia: implication for cancer. Oncogene 2005;24:1011–20.[CrossRef][Medline]. Mol Cell Biol 2002;22:7405–16. to the endoplasmic reticulum and promotes dephosphorylation of the subunit of eukaryotic translation initiation factor 2. Mol Cell Biol 2003;23:1292–303. is required for activation of internal translation initiation during cell differentiation. Eur J Biochem 2002;269:2810–9.[Medline] as a cancer drug target. Mol Cancer Ther 2004;3:647–54. kinase is required for the development of the skeletal system, postnatal growth, and the function and viability of the pancreas. Mol Cell Biol 2002;22:3864–74.This article has been cited by other articles:
|
|
 |
|
  P. J. Belmont, A. Tadimalla, W. J. Chen, J. J. Martindale, D. J. Thuerauf, M. Marcinko, N. Gude, M. A. Sussman, and C. C. Glembotski Coordination of Growth and Endoplasmic Reticulum Stress Signaling by Regulator of Calcineurin 1 (RCAN1), a Novel ATF6-inducible Gene J. Biol. Chem., May 16, 2008; 283(20): 14012 - 14021. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
