Hypoxic Response Takes Shape

The Faculty of 1000 is aWeb-based literature awareness tool published by BioMed Central. For more information visit www.facultyof1000.com. Cells control their responses to the presence or absence of oxygen by an elegant system of checks and balances. Recent structural studies clarify some high-impact findings in hypoxia research, lending insight into the dynamic nature of HIF-1a , the hypoxia-induced transcription factor. This infamous molecule, when unchecked, turns on genes that enhance tumor growth. Some say these three studies,1-3 two of which are highlighted by the Faculty of 1000,1,2 offer new angiogenesis-promotion and inhibition targets. But, whether drug developers will react may rest on new knowledge about the nature of unfolded proteins. Regardless, the molecular story is solidifying. "We always say, 'Crystal structure makes things crystal clear,'" says Yigong Shi, associate professor of molecular biology at Princeton University and an F1000 faculty member. Shi's former postdoctoral adviser, Nikola Pavletich, captured the crystal structure of HIF-1a bound to the Von Hippel-Lindau protein (pVHL), a tumor suppressor. pVHL ubiquinylates HIF-1a , targeting it for proteasomal degradation under normal cellular oxygen conditions. Their results verify last year's finding that pVHL recognizes a hydroxylated proline residue on HIF-1a .4 This proline modification, previously seen only in structural proteins such as collagen, is almost the sole basis for interaction. Pavletich, Howard Hughes Medical Institute associate investigator and professor, biochemistry and structural biology, Dana Farber Cancer Institute, says, "VHL isn't really recognizing too many other amino acids from HIF-1a . It seems that it only cares about the hydroxy-proline." The four or so residues that interact with hydroxyproline 564 are highly conserved in pVHL orthologs of human, mouse, frog, fly, and worm. Mutations in these regions are often those associated with Von Hippel Lindau disease, from which the gene gets its name. Downstream of the pVHL/HIF-1a interaction, hydroxylation accounts for another oxygen-sensing switch, according to a recent study.5 Two simultaneously published crystal structures verified the switch a little more than a month later.2,3 HIF-1a heterodimerizes with HIF-1ß to turn on genes such as those encoding erythropoietin and vascular endothelial growth factor. It also binds transcriptional coactivators such as CBP or p300. These coactivators have a cysteine/histidine-rich binding domain called CH1 or TAZ1 (transcription activator zinc-finger domain) that binds HIF-1a . The TAZ1 motif interacts with a wide range of important transcription factors, and the two structural studies begin to show why this is so. The HIF-1a C-terminal transactivation domain that interacts with TAZ1 is unfolded prior to interaction, meaning that it coils and folds as it interacts with the transcriptional coactivator. Peter Wright, professor and chairman in the Department of Molecular Biology, Scripps Research Institute, La Jolla, Calif., draws the analogy: "Instead of having two hardened spheres making contact, you have one hardened sphere with a piece of tape wrapped around it. This leads to very high binding affinity with multiple contacts all over the surface." One of these contacts gets disrupted when HIF-1a 's asparagine 803 becomes hydroxylated. This provides a back-up antiangiogenic mechanism that deactivates any HIF-1a protein escaping degradation. The structures from Wright's lab and that of Michael Eck, of the Dana Farber Cancer Institute, resolved asparagine buried deeply in the interface when it was not hydroxylated. (Eck's nearly identical work, which appeared in the same journal as Wright's but in a different section, was not reviewed by F1000.) Some say that the results from all three papers do not offer much promise for lining up angiogenesis-directed drug targets. Pavletich quips, "Enzymes are easy to inhibit; protein-protein interactions are not." Yet he does anticipate targeting the pVHL/HIF-1a interaction, hoping to spur angiogenesis for patients with cardiovascular disease or stroke. "Among protein-protein interactions, this doesn't look that difficult to inhibit because it's localized to a small area." Eck says the case for inhibiting HIF-1a binding to transcriptional coactivators might be more difficult. "The more extensive the interaction, the more difficult to disrupt--this turns out to be a very extensive interaction," he says. Such inhibition could potentially hinder angiogenesis as a treatment for patients with cancer. Wright expresses optimism, however, saying that the unique activity of the unfolded HIF-1a binding domain might be more amenable to inhibition than others believe. "Even a subtle modification, such as an interaction with a tightly bound small molecule, I think, will effectively compete with these interactions. Witness the hydroxylation at the asparagine." Wolfgang Jahnke, a scientific expert at Novartis Pharmaceuticals, disagrees. Jahnke, who has worked with Wright and nominated the paper for F1000 says, "No drugs would bind to an unfolded protein because the entropic costs for structure formation would be way too high." Brendan A. Maher can be contacted at bmaher@the-scientist.com References 1. J.H. Min et al., "Structure of an HIF-1a -pVHL complex: Hydroxyproline recognition in signaling," Science, 296:1886-9, June 7, 2002 2. S.A. Dames et al., "Structural basis for HIF-1a /CBP recognition in the cellular hypoxic response," Proceedings of the National Academy of Sciences (PNAS), 99[8]:5271-6, April 16, 2002. 3. S.J. Freedman et al., "Structural basis for recruitment of CBP/p300 by hypoxia-inducible factor-a," PNAS, 99[8]:5367-72, April 16, 2002. 4. J.M. Perkel, "Seeking a cellular oxygen sensor," The Scientist, 15[10]:15, May 14, 2001. 5. D. Lando et al., "Asparagine hydroxylation of the HIF transactivation domain: A hypoxic switch," Science, 295:858-61, Feb. 1, 2002.

Hypoxic Response Takes Shape

Interested in reading more?

Become a Member of