¿du
An important question, crucial to the design of effective therapies, is whether preventing an adaptive physiological re- sponse, such as cardiac hypertrophy, can be beneficial, especially in the presence of sustained hemodynamic load. For exam- ple, in experimental models of left-ven- tricular pressure overload, inhibition of Gq signaling and cardiac hypertrophy by cardiac-specific overexpression of a spe- cific regulator of G protein signaling, RGS- 4, caused the rapid development of dilated cardiomyopathy and heart fail- ure8. Although targeted disruption of car- diac Gq and G11 block the hypertrophic response9, it is not known whether pro- gression to heart failure would have been prevented in these mice. Similarly, the re- sults of Asakura et al. should be viewed with some caution, considering the long- term benefits and risks of blocking cardiac hypertrophy in response to GPCR activa- tion or pressure overload. For example, it is quite possible that chronic inhibition of HB-EGF processing could actually lead to a more rapid cardiac decompensation by bypassing the adaptive stage of cardiac hypertrophy. Indeed, a monoclonal anti- body, trastuzumab (Herceptin), which blocks the EGFR subtype, EGFR-2, caused increased incidence of unexplained car- diomyopathy in clinical trials for breast cancer10. This raises the question as to whether blocking EGFR signaling by an- tagonizing ADAM12 could actually pre- vent heart failure. However, antagonism of ADAM12 may be more tissue-specific, and may therefore yield different clinical outcomes compared with therapies that directly block EGFR transactivation.
Finally, for this novel HB-EGF shedding pathway to be physiologically relevant, it must be put into context of existing path- ways, which are also known to elicit the
hypertrophic response. For example, stimulation of GPCRs or exposure to pres- sure overload induces the activation of small G proteins of the Ras and Rho fam- ily3. These small G proteins activate mito- gen-activated protein (MAP) kinases and mediate the generation of reactive oxygen species (ROS)11, processes that are neces- sary for the hypertrophic response. In par- ticular, the Rho family of small G proteins is thought to have critical roles in the de- velopment of cardiac hypertrophy by reg- ulating cell morphology and contractile elements.
What then, is the relationship between HB-EGF shedding and small G proteins? In some cellular systems, heterotrimeric G proteins activate small G proteins through the autocrine/paracrine transac- tivation of tyrosine kinase receptors such as EGFR (ref. 12). Thus, it is likely that small G proteins are activated down- stream of HB-EGF shedding. But then, what about the activation of MAP kinases and the generation of intracellular ROS, processes which are also dependent on small G proteins? Do they function up- stream of HB-EGF shedding by activating ADAM12 or some other metalloprotease? Or do they also work downstream of EGFR transactivation? Also, where does the calcium-dependent calcineurin/CaM kinase signaling pathway fit into all of this? Does it have a role in HB-EGF shed- ding or vice versa? Questions such as these remind us of just how little we really understand about the hypertrophic process.
This fascinating study by Asakura et al. provokes more questions than it answers. But the finding that HB-EGF shedding is integral to the hypertrophic process may help sort out some of the complex hierar- chical relationships between the various
signaling pathways that lead to cardiac hypertrophy.
1. Levy, D., Garrison, R.J., Savage, D.D., Kannel, W.B. & Castelli, W.P. Prognostic implications of echocardio- graphically determined left ventricular mass in the Framingham Heart Study. N. Engl. J. Med. 322, 1561-6 (1990).
2. Hunter, J.J. & Chien, K.R. Signaling pathways for car- diac hypertrophy and failure. N. Engl. J. Med. 341, 1276-83 (1999).
3. Asakura, M. et al. Cardiac hypertrophy is inhibited by antagonism of ADAM12 processing of HB-EGF: Metalloprotease inhibitors as a potential new ther- apy for cardiac hypertrophy. Nature Med. 8, 35-40 (2001).
4. Black, R.A. & White, J.M. ADAMs: focus on the pro- tease domain. Curr. Opin. Cell. Biol. 10, 654-659 (1998).
5. Prenzel, N. et al. EGF receptor transactivation by G- protein-coupled receptors requires metalloprotease cleavage of proHB-EGF. Nature 402, 884-888 (1999).
6. Moss, M.L. et al. Cloning of a disintegrin metallopro- tease that processes precursor tumour-necrosis fac- tor-a. Nature 385, 733-736 (1997).
7. Li, Y.Y. et al. Myocardial extracellular matrix remod- eling in transgenic mice overexpressing tumor necrosis factor alpha can be modulated by anti- tumor necrosis factor & therapy. Proc. Natl. Acad. Sci. USA 97, 12746-12751 (2000).
8. Rogers, J.H. et al. RGS4 causes increased mortality and reduced cardiac hypertrophy in response to pressure overload. J. Clin. Invest. 104, 567-576 (1999).
9. Wettschureck, N. et al. Absence of pressure overload induced myocardial hypertrophy after conditional inactivation of Gaq/Ga11 in cardiomyocytes. Nature Med. 7, 1236-12340 (2001).
10. Slamon, D.J. et al. Use of chemotherapy plus a mon- oclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N. Engl. J. Med. 344, 783-792 (2001).
11. Takemoto, M. et al. Statins as antioxidant therapy for preventing cardiac myocyte hypertrophy. J. Clin. Invest. 108, 1429-1437 (2001).
12. Daub, H., Weiss, F.U., Wallasch, C. & Ullrich, A. Role of transactivation of the EGF receptor in signalling by G-protein-coupled receptors. Nature 379, 557-560 (1996).
Cardiovascular Division, Department of Medicine Brigham & Women’s Hospital and Harvard Medical School Cambridge, Massachusetts, USA Email: jliao@rics.bwh.harvard.edu
Tumor-specific mutations in p53: The acid test
Tumor-specific, ‘signature’ mutations of p53 have been identified in several cancers. The molecular basis of these tissue-specific selections has been poorly understood. A recent report sheds light on this issue by identifying the molecular basis of one such signature mutation found in adrenal cortical carcinomas.
M utations of the tumor-suppressor gene encoding p53 (TP53) com- monly occur in many human cancers. Normally, the p53 protein controls cell- cycle arrest in response to stressors such as DNA damage or hypoxia. In cancerous cells, however, this essential ‘brake’ is often inactivated by somatic point muta- tions. Because p53 mutations allow cells to escape growth suppression under
PIERRE HAINAUT
stress, they confer a selective advantage to the cancerous cells1. About 200 families have been identified as carriers of germline p53 mutations, which cause a predisposition to inherited Li-Fraumeni syndrome (LFS), a rare disease whose symptoms include sarcomas, breast can-
cers, brain tumors, childhood adrenal cor- tical carcinomas (ACC) as well as a wide range of early-onset cancers2.
In several cancers, tumor-specific, ‘sig- nature’ p53 somatic mutations have been observed, such as R249S (arginine to ser- ine) in liver cancers found in parts of Asia and Africa. How do such signature muta- tions arise? In the case of liver cancer, we now know that aflatoxin B1-a carcino-
3du
D352
R337H
DBD
DBD
RRRCA/TA/TGYYY
RRRCA/TA/TGYYY
pH 7
pH 8
R337H
D352
R337H
DBD
DBD
RRRCA/TA/TGYYY
RRRCA/TA/TGYYY
gen present in the staple diets of many de- veloping countries-induces mutations at specific codons in p53, including R249. However, why only R249S emerges as a signature mutation is poorly understood, although we can speculate that R249S makes an unique contribution to the pathogenesis of liver cancer and is there- fore selected in a tissue-specific manner1. Understanding the molecular basis of these tissue-specific selections has far- reaching implications, as the functional properties of mutant p53 may determine the response of cancer cells to cytotoxic therapy.
A paper by DiGiammarino et al.3, pub- lished in the January issue of Nature Structural Biology, takes us a step forward in that direction by identifying the nature of the molecular defects underlying a p53 tissue-specific, signature mutation. DiGiammarino et al. analyzed the struc- tural and biochemical properties of a rare p53 mutation discovered earlier by Ribeiro et al.4 in adrenal cortical carcino- mas that occur with much higher fre- quency than normal in children of southern Brazil. Ribeiro et al. examined the cancer history and germline p53 mu- tation status in childhood ACC patients and their families. They found that most of these patients had an identical germline point mutation, encoding an arginine instead of a histidine at codon 337 (R337H). The families did not have a common ancestry, and R337H was not a common polymorphism among southern
Brazilians. Moreover, the wild-type allele was deleted and the mutant p53 protein was highly expressed in the ACCs of R337H carriers, sug- gesting that the inher- ited mutant allele contributed to the de- velopment of ACC. Although these features were consistent with LFS-associated ACC, there was no history of increased incidence of other cancers among family members. Thus, the R337H mutation appeared to be a low-penetrance allele causing an exclu- sive predisposition to ACC. This was the first demonstration of a germline signa- ture mutation, which contributed to can- cer in a tissue-specific manner.
However, when tested for functional properties in various experimental sys- tems, the R337H mutation retained the capacity to regulate p53-responsive genes and to suppress the growth of colonies of cancer cells in vitro. This is unlike com- mon LFS mutants, which show dramatic alterations of these biological proper- ties4,5.
This finding prompted DiGiammarino et al. to hypothesize that R337H has a sub- tle structural defect, causing its dysfunc- tion only under particular conditions and in particular tissues. The p53 protein is made of four distinct domains: an N-ter- minal transactivation domain, a central DNA-binding domain, a tetramerization (tet) domain (where Arginine 337 falls) and a C-terminal regulatory domain. Wild-type p53 preferentially assembles into a tetramer, which is the best configu- ration for high-affinity binding to specific DNA sequences. Although over 80% of the mutations found in cancers fall within the DNA-binding domain (be- tween residues 102 and 292), several so- matic and germline mutations have been observed in the tet domain (www.iarc.fr/p53/). These mutations pre- sumably disrupt tetramer assembly, inac- tivating wild-type p53 function5.
Although R337H behaved normally in in vitro assays, DiGiammarino et al. reasoned that these assays were not sensitive enough to detect all the subtle alterations in p53 behavior.
Therefore, DiGiammarino et al. com- pared the capacity of R337H to form a tetramer with that of wild-type p53. The tet domains normally assemble in a ‘dimer of dimers’ made of a four-helix bundle flanked by anti-parallel ß-strands6 (each p53 monomer contributing one ß- trand and one helix) (Fig. 1). Arginine 337 forms a salt bridge with aspartic acid 352 across the helix-helix interface within dimers. This salt bridge is stable between pH 5 and 9 because the functional groups involved remain charged within this range. DiGiammarino et al. found that the R337H tet domain adopted a quasi- wild-type fold at pH 7 but had a much lower thermal stability than the wild-type tet domain. Furthermore, they found that the assembled R337H tet domain was highly sensitive to small changes in pH. They hypothesized that these pH changes may affect the charge of histidine, thus af- fecting its capacity to form a salt bridge with Asp 352. In fact, they found that the imidazole ring of histidine was deproto- nated in the high range of physiological pH, resulting in the collapse of the salt bridge at pH 8.0 (the measured pKa value of His 337 was 7.7). Thus, the R337H mu- tant functioned normally when His 337 was protonated (that is, at pH below 7.7), but would unfold at physiological tem- perature when the salt bridge was neutral- ized. This unique biochemical property defines R337H as a pH-sensitive, dysfunc- tional p53 mutant.
These results have far-reaching implica- tions for our understanding of what makes a ‘mutant’ p53, and provides con- clusive evidence that the effect of the R337H mutation is highly tissue-specific. The strength of the demonstration lies in the fact that R337H is a germline muta- tion, thus ruling out that its association with ACC results from mutagenic events occurring selectively in adrenal cells. Is pH sensitivity the basis of tissue speci- ficity? And if so why does this mutation only predispose individuals to ACC? Are
3du
there particular conditions in adrenal cells that would destabilize R337H?
DiGiammarino et al. suggest that the an- swer lies in the extensive tissue remodeling that occurs through selective apoptosis in the adrenal gland during pre- and postna- tal development7. A rise in intracellular pH often occurs in apoptotic cells, and the re- sultant loss of p53 function may allow adrenal cells to escape apoptosis and favor the survival of cells at high risk of malig- nant transformation. However, the adrenal gland is not the only organ that undergoes perinatal tissue remodeling, and there is no physiological evidence for a particular pH- dependence in adrenal cells. Another pos- sibility is that R337 may be part of a docking site for interaction of p53 with an- other protein in adrenal cells, and that dis- ruption of this interaction may cause ACC. Creating a transgenic mouse carrying the R337H mutation is an obvious step in try- ing to address these questions. More gener- ally, however, these observations have
caused us to re-assess the definition of pre- disposition syndromes associated with germline p53 mutations. Many families with p53 mutations do not match the orig- inal definition of LFS and several tentative definitions have been proposed for ‘Li-Fraumeni-like’ (LFL) syndromes8. A bet- ter understanding of the molecular basis of tissue specificity may also shed some light on why tumors with frequent somatic p53 mutations-such as lung cancers, cancers of the head and neck or cancers of the esophagus-are under-represented in LFS families. The R337H mutant is likely to rep- resent an extreme case in the spectrum of effects caused by p53 mutations, but it may also represent an Ariadne’s thread in the labyrinth of tissue-specific carcinogenesis.
1. Hainaut, P. & Hollstein, M. The p53 tumor sup- pressor gene: The first ten thousand mutations. Adv. Cancer Res. 77, 81-137 (2000).
2. Malkin, D. et al. Germ line p53 mutations in a fa- milial syndrome of breast cancer, sarcomas, and other neoplasms. Science 250, 1233-1238 (1990).
3. DiGiammarino, E.L. et al. A novel mechanism of tumorigenesis involving pH-dependent destabi- lization of a mutant p53 tetramer. Nature Struct. Biol. 9, 12-16 (2002).
4. Ribeiro, R.C. et al. An inherited p53 mutation that contributes in a tissue-specific manner to pedi- atric adrenal cortical carcinoma. Proc. Natl. Acad. Sci. USA 98, 9330-9335 (2001).
5. Lomax, M.E., Barnes, D.M., Hupp, T.R., Picksley, S.M., & Camplejohn, R.S. Characterization of p53 oligomerization domain mutations isolated from Li-Fraumeni and Li-Fraumeni like family members. Oncogene 17, 643-649 (1998).
6. Jeffrey, P.D., Gorina, S. & Pavletich, N.P. Crystal structure of the tetramerization domain of the p53 tumor suppressor at 1.7 angstroms. Science 267, 1498-1502 (1995).
7. Spencer, S.J., Mesiano, S., Lee, J.Y. & Jaffe, R.B. Proliferation and apoptosis in the human adrenal cortex during the fetal and perinatal periods: im- plications for growth and remodeling. J. Clin. Endocrinol. Metab. 84, 1110-1115 (1999).
8. Birch, J.M. et al. Relative frequency and morphol- ogy of cancers in carriers of germline TP53 muta- tions. Oncogene 20, 4621-4628 (2001).
Molecular Carcinogenesis Group International Agency for Research on Cancer Lyon, France
Email: hainaut@iarc.fr
Virus, cells and videotape
A new study catches viruses in the act of infecting human cells. A labeling technique for tracking the movements of single molecules in real-time has now been used to get a look at the step-by-step path of viral infection. The study may open the way to the develop- ment of more efficient gene therapy vectors and antiviral drugs.
A virus typically makes contact with a cell, gets inside it and is transported into the nucleus where it replicates. Until now, the dif- ferent stages of the infection pathway have been visualized using methods-such as electron microscopy-that cannot be used in liv- ing cells. In the 30 November issue of Science, a team lead by Christoph Bräuchle at the Ludwig-Maximilians University in Munich, Germany report that they have obtained real-time images of single viruses moving inside host cells. The low degree of labeling per virus and the low concentration of viruses per cell ensures nat- ural and physiological conditions to study the infection process, ac- cording to the researchers.
The study used the adeno-associated virus (AAV)-a virus often used in gene therapy studies-as a model system. The researchers tagged individual viruses with one or two molecules of the dye Cy5. They then infected Hela cells with viral particles (10-1,000 per cell) and took snapshots every 40 milliseconds to construct two-dimen- sional projections of single AAV trajectories. The measurement of probably thousands of trajectories yielded a picture of unprece- dented detail of the various stages of virus infection.
There were a few surprises. For one thing, AAV punched through the cell membrane faster than expected (in about 64 milliseconds) and was in the nucleus within about 15 minutes. Although in most cases viruses diffused freely in the cytoplasm, some of them started to move at constant speed and along well-defined pathways once inside the nucleus. All trajectories inside the nucleus were oriented in roughly the same direction. Based on these observations, the researchers sug- gested that there are physical tracks in the nuclear membrane that carry viruses-although such structures have yet to be characterized.
The photo represents a frame from a ‘virus documentary’ with the trajectory of a single AVV particle traveling in the cytoplasm of the cell shown in red. This particular virus zipped around the cytoplasm and bounced off the nuclear membrane (outlined with a yellow cir- cle) a few times, but never managed to get inside the nucleus-or at least not on tape. For a movie of viral infection, visit www.single- virus-tracing.com.
LAURA BONETTA