P ERSPECTIVES Indeed, a recent study shows that FOXO activating the oxidative stress response 3. J. H. Kim et al., Nature 434, 921 (2005). transcription factors are acetylated, and that might promote quiescence and thereby 4. M.A. G. Essers et al., Science 308, 1181 (2005). 5. R. H. Medema et al., Nature 404, 782 (2000). deacetylation promotes cell cycle arrest and antagonize cancerous cell proliferation. 6. G. J. Kops et al., Mol. Cell. Biol. 22, 316 (2002). quiescence over programmed cell death. (8). Thus, the oxidative stress response may not 7. B. M.T. Burgering, G. J. Kops, Trends Biochem. Sci. 27, Modifiers of TCF, FOXO, or -catenin that only promote cell survival and longevity, but 352 (2002). β 8. A. Brunet et al., Science 303,2011 (2004);published can shift the balance of their interactions to also may prove useful in the development of online 19 February 2004 (10.1126/science.1094637). favor quiescence would have potential cancer therapies. 9. M.A. Essers et al., EMBO J. 23, 4802 (2004). implications for cancer therapy. It is also 10. G. J. Kops et al., Nature 419, 316 (2002). curious that despite epidemiological evi- References 11. C. Kenyon, Cell 120, 449 (2005). 1. C.Y. Logan, R. Nusse, Annu. Rev. Cell Dev. Biol. 20,781 12. Y. Honda, S. Honda, FASEB J. 13, 1385 (1999). dence indicating that antioxidants lower (2004). cancer risk, these new findings suggest that 2. T. Reya, H. Clevers, Nature 434, 843 (2005). 10.1126/science.1113356 PLANETARY SCIENCE Thermal evolution models of the mar- tian core indicate that core solidification The Interior of Mars would have generated a long-lived (>1 bil- lion years) dynamo (12). This scenario is Yingwei Fei and Constance Bertka not consistent with the observed martian magnetic field history. The short-lived he planetary core is the engine of a Earth (5, 6). Interpretations of the chem- early martian dynamo may have been planet: It drives convection of the man- istry and mineralogy of martian meteorites caused by an initially superheated core after Ttle, shapes the planet’s surface, and—if (7) and of the basaltic rocks at the landing rapid core formation (12) or by mantle the core contains convective molten metal that sites of the Mars Exploration Rover processes such as overturning of a chemi- creates a dynamo—generates a global mag- Mission (8), although model dependent, are cally or mineralogically distinct layer (13), netic field. Based on its mean density and the consistent with this conclusion. resulting in an increased heat flux out of the bulk chemistry of terrestrial planets, Mars is Several lines of independent evidence core. These models require a liquid core to believed to have a suggest that the martian core has been liquid initiate a dynamo. Evidence for a subse- Enhanced online at dense metallic core throughout its history. The first line of evi- quent short-lived (<0.4 billion years) mar- www.sciencemag.org/cgi/ and a silicate mantle. dence comes from the discovery of strongly tian core dynamo around 3.75 billion years content/full/308/5725/1120 However, because no magnetized ancient crust by Mars Global ago (14) would further strengthen the case seismic data exist for Surveyor (9, 10). The magnetization was for a liquid core, because it is difficult to Mars, the density profile of its interior and the acquired more than 4 billion years ago, produce a short-lived dynamo with multi- depth of the core-mantle boundary are not implying a short-lived (~0.5 billion years) ple episodes if the core starts to solidify. known precisely, and it remains unclear early martian core dynamo (11). Such a core The second line of evidence for a liquid whether the martian core is solid or liquid. dynamo may be driven either by composi- martian core comes from measurements of For the past decade, space missions to tional convection (which is set in motion by a the solar tidal deformation of Mars, Mars have provided important constraints on composition gradient) in a liquid outer core obtained by analyzing Mars Global the physics and chemistry of its interior, due to solidification of an inner core, or by Surveyor radio tracking data (2). The meas- although these missions were primarily thermal convection in a fully liquid core due urements indicate that at least the outer part designed to map and understand surface or to high heat flux out of the core (11). of the core is liquid and are also consistent near-surface features of the planet. From a with an entirely liquid core. combined analysis of Mars Global Surveyor 2600 High-pressure experimental tracking data and Mars Pathfinder and Viking Fe melting 2400 Mantle solidus melting data for martian core mate- Lander range and Doppler data, the planet’s rials at martian core pressures pro- 2200 moment of inertia—an important geophysi- vide the third line of evidence for a 2000 Areotherm cal parameter for understanding the planet’s liquid martian core (see the figure). internal density distribution—has been 1800 At core pressures, the iron-nickel- determined at high precision (1, 2). 1600 Melting with 14.2 sulfur system begins to melt at such 1400 weight % S Fe-Ni-S Topography and gravity data collected by eutectic melting a low temperature (~1400 K) that Mars Global Surveyor have constrained the (K) Temperature 1200 any amount of sulfur in the core global average thickness of the martian crust 1000 would lead to at least a liquid outer to between 30 and 80 km (3, 4). 800 Iron-rich silicate mantle Liquid core core for any reasonable thermal Models that combine these data with a 0 5 10 15 20 25 30 35 model (15, 16). Given an estimated range of possible core compositions allow Pressure (GPa) present-day core temperature of the boundaries for mantle density to be 2000 K (12) and a model core com- Evidence for a liquid core. Melting curves of putative defined. The results indicate that the mar- position containing 14.2 weight % martian mantle and core materials are compared with the tian mantle is more iron-rich than that of estimated temperature profile (areotherm) for the mar- sulfur, the martian core is most cer- tian interior. The martian mantle is expected to be solid, tainly liquid (see the figure). Y. Fei is with the Geophysical Laboratory, Carnegie because its temperature is lower than the mantle solidus The size of the martian core is Institution of Washington, Washington, DC 20015, (the temperature at which melting begins). The minimum the least-constrained physical USA. E-mail: [email protected]. C. Bertka is with the melting temperature in the Fe-Ni-S system (the eutectic parameter of the planet, but it has Geophysical Laboratory, Carnegie Institution of melting temperature) at martian core pressures is also important implications for the Washington,Washington, DC 20015, USA and the American Association for the Advancement of shown. Given an estimated core temperature of 2000 K, chemical compositions of the core Science, Washington, DC 20005, USA. E-mail: Mars has an entirely liquid core for a model core composi- and the mantle. Space missions [email protected] tion with 14.2 weight % sulfur. with multiple landers equipped 1120 20 MAY 2005 VOL 308 SCIENCE www.sciencemag.org Published by AAAS P ERSPECTIVES with seismometers are required to pre- have played throughout the history of the 6. C. M. Bertka,Y. Fei, Science 281, 1838 (1998). cisely determine the size of the core. Such planet. However, the forces that have 7. H.Wänke, G. Dreibus, Meteoritics 20, 367 (1985). 8. H.Y. McSween et al., Science 305, 842 (2005). missions would also provide fundamental shaped the planet’s surface are driven in 9. M. H.Acuña et al., Science 284, 790 (1999). information on the structure and density large part by the evolution of its interior. A 10. J. E. P. Connerney et al., Science 284, 794 (1999). profile of the martian interior, which is comprehensive understanding of the 11. D. J. Stevenson, Nature 412, 214 (2001). critical for understanding both the forma- planet’s history requires a greater under- 12. J.-P.Williams, F. Nimmo, Geology 32, 97 (2004). 13. L. T. Elkins-Tanton et al., Lunar Planet. Sci. 34,1479 tion and evolution of Mars. This under- standing of its interior. A mission that (2003). standing is essential for providing a gen- focuses on the martian interior is overdue. 14. R. J. Lillis et al., Lunar Planet. Sci. 36, 1578 (2005). eral context to explore the formation and 15. Y. Fei et al., Am. Mineral. 85, 1830 (2000). evolution of terrestrial planets, including References and Notes 16. Y. Fei, C. M. Bertka, Lunar Planet. Sci. 34, 1829 (2003). 1. W. M. Folkner et al., Science 278, 1749 (1997). 17. We thank F. Nimmo for helpful discussions. Funding our own. 2. C. F.Yoder et al., Science 300, 299 (2003). was provided by the NASA Cosmochemistry program Previous and ongoing missions are pro- 3. M.T. Zuber et al., Science 287, 1788 (2000). and the Carnegie Institution of Washington. 4. M.A.Wieczorek, M.T. Zuber, J. Geophys. Res. 109, viding a wealth of information about the 10.1029/2003JE002153 (2004). martian surface and the role that water may 5. C.M.Bertka,Y.Fei,Earth Planet. Sci. Lett.157,79 (1998). 10.1126/science.1110531 DEVELOPMENTAL BIOLOGY mainly in cardiac and neural tissues, in which cell plasticity is less robust, although Ignoratio Elenchi: Red Herrings not nonexistent. With such diversity of experimental models, only one study has actually exactly duplicated another.