CONTENTS
Volume 334 Issue 6058
SPECIAL SECTION
Materials for Grid Energy REVIEWS 928 Electrical Energy Storage for the Grid: A Battery of Choices
INTRODUCTION 921 Electricity Now and When NEWS 922 925
B. Dunn et al.
Saving for a Rainy Day Turning Over a New Leaf
935
Lowering the Temperature of Solid Oxide Fuel Cells E. D. Wachsman and K. T. Lee
Sunlight in Your Tank—Right Away
>> Editorial p. 877, News Focus story p. 896, and Perspective p. 917
EDITORIAL 877 The Energy Research Imperative
Bill Gates >> Materials for Grid Energy section p. 921
LETTERS 899 Race Disparity in Grants: Check the Citations
NEWS & ANALYSIS 883 NSF Creates Fast Track for Out-of-the-Box Proposals 884 Research Projects Could Be Roadkill in Revision of Massive Highway Bill 885 Revolution Brings New Hopes for Libyan Archaeology 886 China Looks to Balance Its Carbon Books An Unsung Carbon Sink
NEWS FOCUS 888 Will Busting Dams Boost Salmon? Out of the Frying Pan? >> Science Podcast
893 896
POLICY FORUM 908 Preparing to Manage Climate Change Financing
H. P. Erickson
S. D. Donner et al.
Race Disparity in Grants: Empirical Solutions Vital
NEWS OF THE WEEK 880 A roundup of the week’s top stories
PERSPECTIVES 910 When More Is More
J. L. Voss
L.-A. Giraldeau >> Report p. 1000
Response
D. K. Ginther et al. 911
Race Disparity in Grants: Oversight at Home
912
Human Locomotor Circuits Conform
S. Grillner >> Report p. 997
Response
F. S. Collins and L. A. Tabak 905
CORRECTIONS AND CLARIFICATIONS
905
TECHNICAL COMMENT ABSTRACTS
914
One Atom Makes All the Difference
S. Ramaswamy >> Brevia p. 940; Report p. 974 915
BOOKS ET AL. 906 The Next Convergence
Antioxidant Strategies to Tolerate Antibiotics
P. Belenky and J. J. Collins >> Reports pp. 982 and 986
M. Spence, reviewed by C. I. Jones
Design with the Other 90%: Cities
C. E. Smith, curator;
Design with the Other 90%: Cities
>> Materials for Grid Energy section p. 921
Understanding Tribal Fates
R. Arthur and J. Diamond
J. L. Sherley
906
Evolutionary Time Travel Dreams of a Lithium Empire
page 888
916
Analyzing Solar Cycles
S. K. Solanki and N. A. Krivova 917
C. E. Smith et al., reviewed by E. M. Sternberg
True Performance Metrics in Electrochemical Energy Storage
Y. Gogotsi and P. Simon >> Materials for Grid Energy section p. 921 919
Retrospective: Steven P. Jobs (1955–2011)
T. J. Misa
CONTENTS continued >>
COVER
DEPARTMENTS
Nighttime view of present-day Chicago, USA. Since the 1893 Chicago World’s Fair introduced alternating current to the public, demand for electrical power has soared. Increased use of power from renewable resources will require new materials to store energy or generate power to ensure proper load balancing, as discussed in the special section beginning on page 921.
875 878 879 1004 1005
This Week in Science Editors’ Choice Science Staff New Products Science Careers
Photo: Jim Richardson/National Geographic/Getty Images
www.sciencemag.org
SCIENCE
VOL 334
Published by AAAS
18 NOVEMBER 2011
869
CONTENTS
BREVIA 940
972
Evidence for Interstitial Carbon in Nitrogenase FeMo Cofactor
T. Spatzal et al. Structural data show that the light atom at the center of the nitrogenase active site cofactor is a carbon. >> Perspective p. 914; Report p. 974
M. Mokkonen et al. Selection of rare-male types in a population can maintain genetic variation that benefits one sex but harms the other. 974
RESEARCH ARTICLE 941
977
The Large, Oxygen-Rich Halos of Star-Forming Galaxies Are a Major Reservoir of Galactic Metals
J. Tumlinson et al. Observations with the Hubble Space Telescope show that halos of ionized gas are common around star-forming galaxies.
952
955
986
962
Ultralight Metallic Microlattices
965
Silica-Like Malleable Materials from Permanent Organic Networks
990
993
997
page 962
Correction of Sickle Cell Disease in Adult Mice by Interference with Fetal Hemoglobin Silencing
Locomotor Primitives in Newborn Babies and Their Development
N. Dominici et al. Mammalian locomotion patterns share common roots. >> Perspective p. 912 1000
C. T. Nelson et al. The role of defects and interfaces on switching in ferroelectric materials is observed with high-resolution microscopy.
Wolbachia Enhance Drosophila Stem Cell Proliferation and Target the Germline Stem Cell Niche
J. Xu et al. Manipulation of a transcriptional repressor promotes expression of protective fetal globin genes.
T. A. Schaedler et al. A route is developed for fabricating extremely low-density, hollow-strut metallic lattices.
Domain Dynamics During Ferroelectric Switching
H2S: A Universal Defense Against Antibiotics in Bacteria
E. M. Fast et al. A bacterial endosymbion up-regulates mitosis of Drosophila germline stem cells and blocks programmed cell death.
D. Montarnal et al. A polymer shows thermoset-like stability while displaying melt processability like that of a thermopolymer.
968
Active Starvation Responses Mediate Antibiotic Tolerance in Biofilms and Nutrient-Limited Bacteria
K. Shatalin et al. Sulfide formation helps to protect various bacteria from antibiotic toxicity. >> Perspective p. 915
Giant Piezoelectricity on Si for Hyperactive MEMS
S. H. Baek et al. High-quality piezoelectric thin films are grown and exhibit superior properties for microelectromechanical systems.
page 906
D. Nguyen et al. During growth arrest, bacteria tolerate the presence of antibiotics, thanks to mechanisms that protect against oxidant stress. >> Science Podcast
A Reservoir of Ionized Gas in the Galactic Halo to Sustain Star Formation in the Milky Way N. Lehner and J. C. Howk Clouds of ionized gas located inside our Galaxy provide a major supply of matter for fueling ongoing star formation.
958
982
The Hidden Mass and Large Spatial Extent of a Post-Starburst Galaxy Outflow
T. M. Tripp et al. A galaxy that has experienced a recent burst of star formation has an extended halo of hot, ionized gas surrounding it.
Structural Basis of Silencing: Sir3 BAH Domain in Complex with a Nucleosome at 3.0 Å Resolution
K.-J. Armache et al. A regulatory protein forms extensive interactions with the nucleosome core particle to create the basis for gene silencing.
REPORTS 948
X-ray Emission Spectroscopy Evidences a Central Carbon in the Nitrogenase Iron-Molybdenum Cofactor K. M. Lancaster et al. A central light atom in a cofactor at the nitrogenase active site is identified as a carbon. >> Perspective p. 914; Brevia p. 940
Crystal Structure of the Eukaryotic 60S Ribosomal Subunit in Complex with Initiation Factor 6
S. Klinge et al. The 3.5 angstrom–resolution structure provides insights into the architecture of the eukaryotic ribosome and its regulation.
Negative Frequency-Dependent Selection of Sexually Antagonistic Alleles in Myodes glareolus
Rational Choice, Context Dependence, and the Value of Information in European Starlings (Sturnus vulgaris)
E. Freidin and A. Kacelnik Context-related information improves serial decision-making but impairs simultaneous choice. >> Perspective p. 910
page 990
CONTENTS continued >>
www.sciencemag.org
SCIENCE
VOL 334
Published by AAAS
18 NOVEMBER 2011
871
CONTENTS
SCIENCEONLINE SCIENCEXPRESS
SCIENCESIGNALING
www.sciencexpress.org
The Structure of the Eukaryotic Ribosome at 3.0 Å Resolution
www.sciencesignaling.org The Signal Transduction Knowledge Environment 15 November issue: http://scim.ag/ss111511
Imaging of Plasmodium Liver Stages to Drive Next-Generation Antimalarial Drug Discovery
R. Merline et al. A component of the extracellular matrix promotes inflammatory responses in sepsis and in tumors.
A. Ben-Shem et al. A close-up view of the ribosome’s 79 proteins and 5500 RNA nucleotides. 10.1126/science.1212642
S. Meister et al. Imidazolopiperazine compounds inhibit liver-stage malaria parasites with one oral dose in mice. 10.1126/science.1211936 >> Science Podcast
Host Proteasomal Degradation Generates Amino Acids Essential for Intracellular Bacterial Growth C. T. D. Price et al. The bacterial pathogen Legionella pneumophila ensures amino acid supplies by promoting degradation of target host proteins. 10.1126/science.1212868
Calibrating the End-Permian Mass Extinction
S. Shen et al. High-precision geochronologic dating constrains probable causes of Earth’s largest mass extinction. 10.1126/science.1213454
The Origin of OB Runaway Stars
RESEARCH ARTICLE: Signaling by the Matrix Proteoglycan Decorin Controls Inflammation and Cancer Through PDCD4 and MicroRNA-21
RESEARCH ARTICLE: Local Application of Neurotrophins Specifies Axons Through Inositol 1,4,5-Trisphosphate, Calcium, and Ca2+/Calmodulin–Dependent Protein Kinases S. Nakamuta et al. Neurotrophins stimulate calcium signaling to promote axon specification.
RESEARCH ARTICLE: The Long-Term Survival Potential of Mature T Lymphocytes Is Programmed During Development in the Thymus
C. Sinclair et al. T cell receptor signaling in thymocytes determines their responsiveness to a survival cytokine later in life.
ST NETWATCH: The Nobel Prize in Physiology or Medicine 2011 This year’s Prize was awarded for breakthroughs in innate and adaptive immunity.
SCIENCETRANSLATIONAL MEDICINE
M. S. Fujii and S. P. Zwart
Most of the unusually fast, young stars in our galaxy are produced in three-body encounters within dense clusters of stars. 10.1126/science.1211927
TECHNICALCOMMENTS Comment on “Global Trends in Wind Speed and Wave Height” F. J. Wentz and L. Ricciardulli Full text at www.sciencemag.org/cgi/content/ full/334/6058/905-b
Response to Comment on “Global Trends in Wind Speed and Wave Height”
I. R. Young et al. Full text at www.sciencemag.org/cgi/content/ full/334/6058/905-c
SCIENCENOW
www.sciencetranslationalmedicine.org Integrating Medicine and Science 16 November issue: http://scim.ag/stm111611
FOCUS: Biomarker-Based Early Cancer Detection—Is It Achievable?
W. D. Hazelton and E. G. Luebeck A new mathematical model evaluates the power of blood-based biomarkers for early cancer detection.
RESEARCH ARTICLE: Mathematical Model Identifies Blood Biomarker–Based Early Cancer Detection Strategies and Limitations
S. S. Hori and S. S. Gambhir Early cancer detection using current clinical bloodbased biomarker assays may not be possible within the first 10 years of tumor growth.
PERSPECTIVE: Targeting Chaperone-Mediated Autophagy in Cancer
www.sciencenow.org Highlights From Our Daily News Coverage
A. Thorburn and J. Debnath Tumors need chaperone-mediated autophagy for growth and metastasis.
Leonardo’s Formula Explains Why Trees Don’t Splinter
RESEARCH ARTICLE: Chaperone-Mediated Autophagy Is Required for Tumor Growth
This geometry rule probably evolved to help prevent wind breakage. http://scim.ag/Leotrees
Not Pulling Your Leg: Tractor Beams May Be Possible
M. Kon et al. Cancer cells depend on chaperone-mediated autophagy for growth revealing a new target for preventing tumorigenesis and inducing tumor regression.
Several different groups come up with similar ideas for laser beams that can pull objects. http://scim.ag/tractor-beam
RESEARCH ARTICLE: Treatment and Prevention of Urinary Tract Infection with Orally Active FimH Inhibitors C. K. Cusumano et al. Optimized mannoside compounds that block pathogenic Escherichia coli entry into bladder epithelium were effective in the treatment and prevention of urinary tract infections in mice.
SCIENCECAREERS
www.sciencecareers.org/career_magazine Free Career Resources for Scientists
Tooling Up: Views on an Interview, Part 2
D. Jensen Ups and downs continue during Scott Jackson’s interview at ABC Technologies. http://scim.ag/TU_InterViews2
Where Are the Neuroscience Jobs?
M. Price Our correspondent reports from the Neuroscience 2011 meeting in Washington, D.C. http://scim.ag/neurocareers2011
A Mycologist Reaps the (Glowing) Fruit of His Labor S. Reed Patrick Hickey’s career has grown in a way as unpredictable as the organisms he cultivates. http://scim.ag/Hickeyprofile
SCIENCEPODCAST
www.sciencemag.org/multimedia/podcast Free Weekly Show On the 18 November Science Podcast: how starved bacteria resist antibiotics, next-generation antimalarial drugs, the science of dam removal, and more.
SCIENCEINSIDER
news.sciencemag.org/scienceinsider Science Policy News and Analysis
SCIENCE (ISSN 0036-8075) is published weekly on Friday, except the last week in December, by the American Association for the Advancement of Science, 1200 New York Avenue, NW, Washington, DC 20005. Periodicals Mail postage (publication No. 484460) paid at Washington, DC, and additional mailing offices. Copyright © 2011 by the American Association for the Advancement of Science. The title SCIENCE is a registered trademark of the AAAS. Domestic individual membership and subscription (51 issues): $149 ($74 allocated to subscription). Domestic institutional subscription (51 issues): $990; Foreign postage extra: Mexico, Caribbean (surface mail) $55; other countries (air assist delivery) $85. First class, airmail, student, and emeritus rates on request. Canadian rates with GST available upon request, GST #1254 88122. Publications Mail Agreement Number 1069624. Printed in the U.S.A. Change of address: Allow 4 weeks, giving old and new addresses and 8-digit account number. Postmaster: Send change of address to AAAS, P.O. Box 96178, Washington, DC 20090–6178. Single-copy sales: $10.00 current issue, $15.00 back issue prepaid includes surface postage; bulk rates on request. Authorization to photocopy material for internal or personal use under circumstances not falling within the fair use provisions of the Copyright Act is granted by AAAS to libraries and other users registered with the Copyright Clearance Center (CCC) Transactional Reporting Service, provided that $25.00 per article is paid directly to CCC, 222 Rosewood Drive, Danvers, MA 01923. The identification code for Science is 0036-8075. Science is indexed in the Reader’s Guide to Periodical Literature and in several specialized indexes.
Ancient Landslide Merged Trout Populations River blockade explains surprising genetic similarities in two groups of fish. http://scim.ag/land-slide
www.sciencemag.org
SCIENCE
VOL 334
Published by AAAS
18 NOVEMBER 2011
873
EDITED BY STELLA HURTLEY
recently experienced a burst of star formation and observed an outflow of ionized gas with large mass and spatial extent. Tumlinson et al. (p. 948) examined the relationship between ionized outflows and the properties of a sample of 42 galaxies. As compared with galaxies having little or no ongoing star formation, galaxies that are still forming stars were more likely to have halos of ionized oxygen around them.
Metal Veil
CREDITS (TOP TO BOTTOM): KATHARINA BOHM AND FELIX VOIGTS HOFFMANN; BAEK ET AL.
The eukaryotic ribosome is much larger and more complex than its structurally well defined bacterial counterpart. Now, Klinge et al. (p. 941, published online 3 November) report the crystal structure of the large (60S) ribosomal subunit in complex with eukaryotic initiation factor 6 (eIF6). The structure reveals a network of interactions between eukaryotic-specific ribosomal proteins and ribosomal RNA expansion segments and uncovers the roles of eukaryotic ribosomal protein elements in the stabilization of the active site. It also elucidates the molecular basis of the interaction with elF6, which is involved in the initiation of protein synthesis and in ribosome maturation.
Gorgeous Galaxies Galaxies grow by accreting gas from their surroundings and by converting that gas into stars. Indeed, it has long been recognized that without a source of gas replenishment, our Galaxy could not sustain its present rate of star formation. Three reports now describe data obtained by the Cosmic Origins Spectrograph and the Space Telescope Imaging Spectrograph onboard the Hubble Space Telescope. Lehner and Howk (p. 955, published online 25 August) observed ionized clouds of gas traveling at high speed against a sample of 28 stars within the Milky Way, which allowed the gas cloud distances to be determined. A fraction of the clouds are inside our Galaxy and massive enough to sustain the Milky Way’s present rate of star formation. As stars evolve, stellar winds and explosions expel matter from galaxies into their immediate surroundings. Tripp et al. (p. 952) probed the gas around a galaxy that has
A number of approaches are available for the fabrication of very-low-density materials— such as aerogels made from silica, which have remarkable strength and insulating properties. Although these materials look like frozen smoke, a 2-gram sample can support a 2.5-kilogram brick. Schaedler et al. (p. 962) devised a method to make very-low-density metallic materials based on a lattice framework. A liquid photomonomer was exposed to collimated ultraviolet light through a patterned mask to produce a 3D lattice material, which was then coated with a thin film of nickel-phosphorous. The polymer was then etched out, yielding a 3D lattice with hollow nickel-phosphorous struts.
Not So Thermoset Synthetic polymers can be broadly divided into two categories: thermoplastics, which can be repeatedly heated or processed into different shapes, and thermosets, which are processed in the liquid state and then chemically or optically cross-linked. Once hardened, most thermoset materials are almost impossible to further process or shape. Montarnal et al. (p. 965) designed a thermoset-like material that could be processed repeatedly at elevated temperatures—and could even be ground up and recycled into a new shape while retaining the mechanical properties of the original material.
Observing Ferroelectric Domain Dynamics Ferroelectric materials possess a spontaneous electric polarization that can be switched through the application of an electric field; this property is useful for making memory chips and radio-frequency identification tags. Nelson et al. (p. 968) studied the switching dynamics of a BiFeO3 film by using high-resolution transmission electron microscopy, and they observed localized
nucleation events at the electrode-oxide interface, domain-wall pinning on point defects, and the formation of metastable ferroelectric states localized to the oxide-oxide interface.
Sexual Antagonism Traits that are beneficial for males can be detrimental for females. For example, high testosterone and aggression may benefit males by promoting dominance, but may be detrimental in females by reducing fecundity. Thus, highly successful males can produce less successful female offspring, and vice versa. How, then, is variation in traits involved in differential fitness maintained? Mokonnen et al. (p. 972) conducted a large-scale experiment to address this question in the bank vole, a species with regular population fluctuations and a promiscuous mating system. Dominant males (with less fecund daughters) were favored in general conditions, but, when males were rare, subordinant males (and their more fecund daughters) contributed substantially to the population. A model derived from the data suggests that a combination of sexual antagonism and frequency-dependent selection can maintain genetic variation, whereas sexual antagonism alone cannot.
Micromechanical Marvels
Microelectromechanical systems, or MEMS, combine the miniaturization of both mechanical and electromechanical elements onto a single platform using the tools of microfabrication. Producing successful MEMS requires materials that are sufficiently stiff and strong, that show an electromechanical response, and that can be fabricated using compatible methods. Baek et al. (p. 958) report the epitaxial growth of a ferroelectric thin film of Pb(Mg1/3Nb2/3)O3-PbTiO3 on a Si substrate that showed a giant piezoelectric response and was used to make miniature cantilevers that could be operated using a low driving voltage.
Wolbachia Finds Its Niche Intracellular bacteria Wolbachia are being used to help control mosquito populations in the fight against certain infectious diseases. However, relatively little is known about the
www.sciencemag.org SCIENCE VOL 334 18 NOVEMBER 2011 Published by AAAS
Downloaded from www.sciencemag.org on November 17, 2011
Large Ribosomal Subunit Structure
Continued on page 876
875
May 14-21, 2012 See the Annular Solar Eclipse over the Grand Canyon, May 20, 2012! Discover Zion and Bryce Canyon national parks, and the amazing Antelope slot canyon. Enjoy a half-day Grand Canyon rafting trip. $3,695 + air
Australia Total Solar Eclipse
November 5-16, 2012 Discover Sydney, the Outback at Ayer’s Rock and Alice Springs, plus Cairns in Queensland for the Total Eclipse November 14th. Explore the Great Barrier Reef! $4,995 plus air; New Guinea $1,995
For a detailed brochure, please call (800) 252-4910
All prices are per person twin share + air
BETCHART EXPEDITIONS Inc.
17050 Montebello Road Cupertino, California 95014
Email:
[email protected] www.betchartexpeditions.com
This Week in Science Continued from page 875
cellular mechanisms responsible for Wolbachia transmission. Now, Fast et al. (p. 990, published online 20 October) describe two cellular events in the host insect that are affected by Wolbachia maternal transmission. When Drosophila mauritiana was infected with Wolbachia, mitotic activity was elevated in germline stem cells, and programmed cell death was reduced in developing egg chambers—both of which tend to increase the reproductive success of infected insects.
Sickle Cell Intervention The damaging effects of sickle cell disease, caused by a point mutation in the gene that encodes a form of hemoglobin expressed in adults, can be lessened by reintroduction of the form of hemoglobin expressed during fetal development. Unfortunately, the fetal globin gene is usually turned off as development progresses. Using mice carrying both endogenous murine globin genes and transgenic human globin genes, Xu et al. (p. 993, published online 13 October) studied the repressor BCL11a, which silences fetal globin gene expression. Without BCL11a, expression of fetal globin persisted beyond its normal phase of development. Most patients, however, are identified after the developmental shutdown of fetal globin expression. In adult mice in which the developmental shift in globin expression had already taken place, inducible repression of BCL11a allowed a resurgence of fetal globin expression and lessened disease burden.
Downloaded from www.sciencemag.org on November 17, 2011
Grand Canyon Annular Eclipse
Arrest and Tolerate When starving bacteria arrest their growth, they can resist killing by nearly all classes of antibiotics. Starvation is also a major cause of drug tolerance in biofilms, a bacterial community structure found in many chronic infections. Nguyen et al. (p. 982; see the Perspective by Belenky and Collins) show that such antibiotic tolerance occurs, not because the targets for antibiotics have become inactive during growth arrest, but because starvation-sensing mechanisms generate protective responses. Bacterial mutants unable to detect nutrient limitation were orders of magnitude more sensitive to antibiotic exposure, less able to establish animal infections, and failed to generate antibioticresistant mutants.
Baby Steps
Since 1965, SB&F has been the authoritative guide to science resources, bringing expert information to make the best decisions when choosing science materials for libraries, classrooms, or institutions. And this is just one of the ways that AAAS is committed to advancing science to support a healthy and prosperous world. Join us. Together we can make a difference. To learn more, visit: aaas.org/plusyou/sbf
When Is Less More? Sometimes too much information can hamper our decision-making ability, resulting in our making suboptimal choices. Freidin and Kacelnik (p. 1000; see the Perspective by Giraldeau) explored this “less is more” effect in starlings and confirmed that contextual information hampered the ability of the animals to choose the “best” food item. This was true, however, only when the birds were presented with multiple choices simultaneously. In contrast, when the birds were presented with a sequence of prey choices, knowing the context of the find improved their ability to make the optimal choice. In nature, starlings forage for invertebrates and are unlikely to encounter many prey items simultaneously. Thus, decision-making has evolved to favor using contextual information to make choices, despite the fact that it fails when choices are simultaneous. 18 NOVEMBER 2011 VOL 334 SCIENCE www.sciencemag.org Published by AAAS
CREDIT: DAO NGUYEN AND TORSTEN BERNHARDT
Science Books & Films
Newborn babies can make irregular stepping motions, although it will take many months, and many falls, before those turn into walking. Even the bipedally mobile toddler does not walk like an adult. Dominici et al. (p. 997; see the Perspective by Grillner) analyzed how the patterns of human walking change with development. Walking was analyzed and categorized to be the result of a discrete set of neuromuscular components. The neonate has some of these components, which are similar to walking patterns in other mammals. The toddler has added two more components onto the basic set. And the adult has refined the components for optimal walking.
EDITORIAL
The Energy Research Imperative
CREDITS: (TOP) BGC3; (RIGHT) ISTOCKPHOTO.COM
AS SOMEONE NOW WORKING FULL TIME IN GLOBAL HEALTH AND DEVELOPMENT, I SEE FIRSTHAND
how the U.S. government’s support for scientific research has improved people’s lives. That support is vital in another area—affordable, clean energy. I believe it is imperative that the government commit to clean energy innovation at a level similar to its research investments in health and defense. In a time of economic crisis, asking policy-makers in Washington, DC, to spend more money might not be the most popular position. But it’s essential to protect America's national interests and ensure that the United States plays a leading role in the fast-growing global clean energy industry. There is really no other choice. Carbon-based fuels are prone to wild price gyrations and are causing the planet to overheat. The United States spends close to $1 billion a day on foreign oil, while countries such as China, Germany, Japan, and Korea are making huge investments in clean energy technologies. The creation of new energy products, services, and jobs is a good thing wherever it occurs, but it would be a serious miscalculation if America missed out on this singular opportunity. The United States is uniquely positioned to lead in energy innovation, with great universities and national laboratories and an abundance of entrepreneurial talent. But the government must lend a hand. Market incentives, alone, will not create enough affordable, clean energy to get the nation to near-zero CO2 emissions, the level of emissions that developed countries must achieve if we are going to keep Earth from getting even hotter.* Moreover, developing major new technologies, where the time frames necessary for true innovation stretch past the normal horizons of patent protection, requires up-front investments that are too large for venture capital and traditional energy companies. History has repeatedly proven that federal investments in research return huge payoffs, with incredible associated benefits for U.S. industries and the economy. Yet over the past three decades, U.S. government investment in energy innovation has dropped by more than 75%. In 2008, the United States spent less on energy R&D as a percentage of gross domestic product than China, France, Japan, or Canada. Last year, I joined with other business leaders in a call to increase federal investment in energy R&D from $5 billion to $16 billion a year.† (Others, including the President’s Council of Advisors on Science and Technology, have also recommended substantial increases.‡) Recently, our group, the American Energy Innovation Council (AEIC), issued a second report outlining ways to ensure that government research dollars are targeted wisely to achieve optimal returns. The report also suggests ways to pay for the increased investment: reducing or eliminating current subsidies to well-established energy industries, diverting a portion of royalties from domestic energy production, collecting a small fee on electricity sales, or imposing a price on carbon. Any combination of these could provide the funds needed to increase energy innovation. Even at almost triple the current level of government investment in energy innovation, the research dollars that the AEIC is advocating would represent a small fraction of the money presently spent on renewable energy subsidies and efficiency grants. Energy transformations take generations. But if the United States begins in earnest today, the nation’s energy challenges can be solved in ways that truly set America on a path of energy independence and that provide affordable energy for everyone, especially the poor. The return on this kind of investment could change—perhaps even save—the world and provide generations to come with a brighter future. – Bill Gates
Downloaded from www.sciencemag.org on November 17, 2011
Bill Gates is cochair of the Bill & Melinda Gates Foundation, chairman of Microsoft, and a member of the American Energy Innovation Council.
10.1126/science.1216290
*www.agu.org/pubs/crossref/2008/2007GL032388.shtml. †www.americanenergyinnovation.org. ‡www.whitehouse.gov/administration/eop/ostp/pcast.
www.sciencemag.org SCIENCE VOL 334 18 NOVEMBER 2011 Published by AAAS
877
EDITORS’CHOICE EDITED BY KRISTEN MUELLER AND JAKE YESTON EDUCATION
Less Is More
Downloaded from www.sciencemag.org on November 17, 2011
Although educators agree that moving away from “cookbook” laboratory units allows for a better emulation of authentic scientific practice, our understanding of how laboratory materials influence student learning with respect to problem-solving remains limited. Jordan et al. investigated whether the removal of laboratory materials during initial discussions resulted in increased student planning and creativity. Junior and senior undergraduates were divided into two groups, one that was given materials to work with and one that was not, and were asked to design two experiments to determine plant transpiration rate using stem cuttings. Students were asked to record time to completion, describe each experiment, provide drawings of the experimental setup, and to describe in a written report what additional information they needed. Students who were given materials listed two “standard” solutions, whereas students without materials listed an additional five novel solutions, had a shorter time to completion, reported more discussion beyond the instructions, and more often mentioned the natural environment. These results suggest that novice students, given a typical laboratory-based experimental task, focus on available materials, supporting the notion that removal of the laboratory materials can result in greater and more collaborative planning that leads to more creative solutions. — MM J. Res. Sci. Teach. 48, 1010 (2011).
reflector on each cantilever across the array to facilitate simple optical readout of any deflection of the cantilevers. The approach offers a relatively cheap and flexible route to designer sensor imaging arrays. — ISO
CELL SIGNALING
Opioid Receptors Up to Scratch
Cell 147, 447 (2011). PHYSICS
Metamaterials to See in THz Certain bands of the electromagnetic spectrum are useful for specific sensing applications, whether chemical detection or thermal imaging. Terahertz frequencies, for instance, can readily penetrate clothing and paper and so are finding use at airports in the form of full-body scanners. These systems are relatively big though; shrink-
878
Opt. Express 19, 21620 (2011). ASTRONOMY
Bigger than Earth, but No Giant
Smaller sensors on the way?
ing both the radiation source and detector size would make them potentially more mobile. Tao et al. have combined a micromechanical cantilever system with a metamaterial-engineered split-ring resonator. The cantilever array is made from a bilayer of two metals, the different thermal properties of which cause a cantilever to move when it absorbs heat. That heat can also be provided in form of absorbed photons. As the split-ring metamaterial can be tuned by design to operate at any desired wavelength, the authors chose its geometry to confer sensitivity to absorption in the microwave and terahertz frequency range. Moreover, they incorporated a
18 NOVEMBER 2011
VOL 334
SCIENCE
Published by AAAS
The microlensing method of detecting planets takes advantage of the fact that the light of a background star gets deflected by the gravitational field of a foreground star with which it is spatially aligned. If the foreground star hosts a planet, light will be deflected in a way that furthermore depends on the planet’s mass and distance from its host star. Using this method to analyze observations from 13 different telescopes around the world and one in space, Muraki et al. detected a planet 10.4 times more massive than Earth. The planet orbits a star 0.56 times as massive as the Sun, at a distance comparable to Jupiter’s distance from the Sun. This puts the planet’s orbit beyond the snow line—the distance from a star beyond which ice can condensate—and thus in the sort of environs where giant planets such as Jupiter and Saturn form. However, the new planet looks more like a failed giant planet; one, in other words, that accreted enough solid material to form the core of a giant planet but never acquired a gaseous envelope because the protoplanetary disk lost its gas before the solid core was massive enough to efficiently attract hydrogen and helium. — MJC
www.sciencemag.org
Astrophys. J. 741, 22 (2011).
CREDITS: (TOP TO BOTTOM) ISTOCKPHOTO.COM; L-3 SECURITY & DETECTION SYSTEMS
Itching and pain are two distinct sensations, and it’s probably not advantageous to mix the two, given that they cause very different responses —scratching in response to the former and withdrawal in response to a painful stimulus. Nevertheless, there seems to be some overlap of the two sensations. For example, a common side effect of injection of opioids into the spinal cord for pain management is opioid-induced itch. Liu et al. report that this effect of the opioid morphine is brought about by heterodimeric receptors composed of the µ-opioid receptor (MOR1D), which binds morphine, and the gastrin-releasing peptide receptor (GRPR), which binds GRP. This hybrid receptor only allows signaling of itch sensation as binding of morphine activates GRPR, but binding of GRP doesn’t activate MOR1D. Specifically targeting this heterodimeric receptor holds promise as a strategy to prevent opioid-induced itch while retaining the analgesic effects of opioids. — LBR
www.sciencemag.org
EDITOR-IN-CHIEF Bruce Alberts EXECUTIVE EDITOR NEWS EDITOR
Monica M. Bradford
Colin Norman
MANAGING EDITOR, RESEARCH JOURNALS Katrina L. Kelner DEPUTY EDITORS R. Brooks Hanson, Barbara R. Jasny,
Editorial: 202-326-6550, FAX 202-289-7562 News: 202-326-6581, FAX 202-371-9227
Bateman House, 82-88 Hills Road Cambridge, UK CB2 1LQ
+44 (0) 1223 326500, FAX +44 (0) 1223 326501 SUBSCRIPTION SERVICES For change of address, missing issues, new
orders and renewals, and payment questions: 866-434-AAAS (2227) or 202-326-6417, FAX 202-842-1065. Mailing addresses: AAAS, P.O. Box 96178, Washington, DC 20090-6178 or AAAS Member Services, 1200 New York Avenue, NW, Washington, DC 20005 I NSTITUTIONAL S ITE L ICENSES please call 202-326-6755 for any questions or information REPRINTS: Author Inquiries 800-635-7181 Commercial Inquiries 803-359-4578 PERMISSIONS 202-326-7074, FAX 202-682-0816 MEMBER BENEFITS AAAS/Barnes&Noble.com bookstore www.aaas.org/bn; AAAS Online Store www.apisource.com/aaas/ code MKB6; AAAS Travels: Betchart Expeditions 800-252-4910; Apple Store www. apple.com/eppstore/aaas; Bank of America MasterCard 1-800-833-6262 priority code FAA3YU; Cold Spring Harbor Laboratory Press Publications www.cshlpress.com/affiliates/aaas.htm; GEICO Auto Insurance www.geico.com/landingpage/go51.htm?logo=17624; Hertz 800-654-2200 CDP#343457; Office Depot https://bsd. officedepot.com/portalLogin.do; Seabury & Smith Life Insurance 800424-9883; Subaru VIP Program 202-326-6417; VIP Moving Services www.vipmayflower.com/domestic/index.html; Other Benefits: AAAS Member Services 202-326-6417 or www.aaasmember.org.
[email protected] (for general editorial queries)
[email protected] (for queries about letters)
[email protected] (for returning manuscript reviews)
[email protected] (for book review queries) Published by the American Association for the Advancement of Science (AAAS), Science serves its readers as a forum for the presentation and discussion of important issues related to the advancement of science, including the presentation of minority or conflicting points of view, rather than by publishing only material on which a consensus has been reached. Accordingly, all articles published in Science—including editorials, news and comment, and book reviews—are signed and reflect the individual views of the authors and not official points of view adopted by AAAS or the institutions with which the authors are affiliated. AAAS was founded in 1848 and incorporated in 1874. Its mission is to advance science, engineering, and innovation throughout the world for the benefit of all people. The goals of the association are to: enhance communication among scientists, engineers, and the public; promote and defend the integrity of science and its use; strengthen support for the science and technology enterprise; provide a voice for science on societal issues; promote the responsible use of science in public policy; strengthen and diversify the science and technology workforce; foster education in science and technology for everyone; increase public engagement with science and technology; and advance international cooperation in science.
INFORMATION FOR AUTHORS
See pages 784 and 785 of the 11 February 2011 issue or access www.sciencemag.org/about/authors
SENIOR EDITORIAL BOARD
A. Paul Alivisatos, Lawrence Berkeley Nat'l. Laboratory Cori Bargmann, The Rockefeller Univ. Ernst Fehr, Univ. of Zurich Richard Losick, Harvard Univ. Michael S. Turner, University of Chicago
BOARD OF REVIEWING EDITORS
Adriano Aguzzi, Univ. Hospital Zürich Takuzo Aida, Univ. of Tokyo Sonia Altizer, Univ. of Georgia Sebastian Amigorena, Institut Curie Angelika Amon, MIT Kathryn Anderson, Memorial Sloan-Kettering Cancer Center Siv G. E. Andersson, Uppsala Univ. Peter Andolfatto, Princeton Univ. Meinrat O. Andreae, Max Planck Inst., Mainz John A. Bargh, Yale Univ. Ben Barres, Stanford Medical School Jordi Bascompte, Estación Biológica de Doñana, CSIC Facundo Batista, London Research Inst. Ray H. Baughman, Univ. of Texas, Dallas David Baum, Univ. of Wisconsin Yasmine Belkaid, NIAID, NIH Philip Benfey, Duke Univ. Stephen J. Benkovic, Penn State Univ. Gregory C. Beroza, Stanford Univ. Peer Bork, EMBL Bernard Bourdon, Ecole Normale Superieure de Lyon Ian Boyd, Univ. of St. Andrews Paul M. Brakefield, Univ. of Cambridge Christian Büchel, Universitätsklinikum Hamburg-Eppendorf Joseph A. Burns, Cornell Univ. William P. Butz, Population Reference Bureau Gyorgy Buzsaki, Rutgers Univ. Mats Carlsson, Univ. of Oslo Mildred Cho, Stanford Univ. David Clapham, Children’s Hospital, Boston David Clary, Univ. of Oxford J. M. Claverie, CNRS, Marseille Jonathan D. Cohen, Princeton Univ. Robert Cook-Deegan, Duke Univ. James Collins, Boston Univ. Alan Cowman, Walter & Eliza Hall Inst. Robert H. Crabtree, Yale Univ. Wolfgang Cramer, Medit. Inst. for Ecology & Paleoecology F. Fleming Crim, Univ. of Wisconsin Jeff L. Dangl, Univ. of North Carolina
FULFILLMENT SYSTEMS AND OPERATIONS (
[email protected]); CUSTOMER
EDITORIAL
Pat Butler; SPECIALISTS LaVonda Crawford, Vicki Linton, April Marshall; MANAGER, DATA ENTRY Mickie Napoleoni; DATA ENTRY SPECIALISTS Tarrika Hill, JJ Regan, Eva Mae Campbell
EDITORS
BUSINESS OPERATIONS
M. Sugden
1200 New York Avenue, NW Washington, DC 20005
EXECUTIVE PUBLISHER Alan I. Leshner PUBLISHER Beth Rosner
Andrew
SENIOR EDITORS/COMMENTARY Lisa D. Chong, Brad Wible; SENIOR Gilbert J. Chin, Pamela J. Hines, Paula A. Kiberstis (Boston), Marc S. Lavine (Toronto), Beverly A. Purnell, L. Bryan Ray, Guy Riddihough, H. Jesse Smith, Phillip D. Szuromi (Tennessee), Valda Vinson, Jake S. Yeston, Laura M. Zahn (San Diego); ASSOCIATE EDITORS Kristen L. Mueller, Jelena Stajic, Sacha Vignieri, Nicholas S. Wigginton; BOOK REVIEW EDITOR Sherman J. Suter; ASSOCIATE LETTERS EDITOR Jennifer Sills; EDITORIAL MANAGER Cara Tate; SENIOR COPY EDITORS Jeffrey E. Cook, Cynthia Howe, Harry Jach, Lauren Kmec, Barbara P. Ordway, Trista Wagoner; COPY EDITOR Chris Filiatreau; SENIOR EDITORIAL COORDINATORS Carolyn Kyle, Beverly Shields; EDITORIAL COORDINATORS Joi S. Granger, Anita Wynn; PUBLICATIONS ASSISTANTS Ramatoulaye Diop, Aneera Dobbins, Jeffrey Hearn, Michael Hicks, Lisa Johnson, Scott Miller, Jerry Richardson, Teresa R. Sakon, Brian White; EDITORIAL ASSISTANT Patricia M. Moore; EXECUTIVE EDITORIAL ASSISTANT Yolanda O'Bannon (San Francisco); EXECUTIVE ASSISTANT Alison Crawford; ADMINISTRATIVE SUPPORT Maryrose Madrid; EDITORIAL FELLOW Melissa R. McCartney E DITORIAL DIRECTOR , WEB AND NEW MEDIA Stewart Wills; RESEARCH ASSOCIATE Corinna Cohn; WEB DEVELOPMENT MANAGER Martyn Green; WEB DEVELOPER Andrew Whitesell; INTERNS Andrew Green, Kerry Klein NEWS DEPUTY NEWS EDITORS Robert Coontz, David Grimm (Online), Eliot Marshall, Jeffrey Mervis, Leslie Roberts, John Travis; CONTRIBUTING EDITORS Elizabeth Culotta, Polly Shulman; NEWS WRITERS Yudhijit Bhattacharjee, Adrian Cho, Jennifer Couzin-Frankel, Carolyn Gramling, Jocelyn Kaiser, Richard A. Kerr, David Malakoff, Greg Miller, Elizabeth Pennisi, Robert F. Service (Pacific NW), Erik Stokstad; WEB DEVELOPER Daniel Berger; INTERN Daniel Strain; CONTRIBUTING CORRESPONDENTS Jon Cohen (San Diego, CA), Daniel Ferber, Ann Gibbons, Sam Kean, Andrew Lawler, Mitch Leslie, Charles C. Mann, Virginia Morell, Gary Taubes; COPY EDITORS Linda B. Felaco, Melvin Gatling, Melissa Raimondi; ADMINISTRATIVE SUPPORT Scherraine Mack; BUREAUS San Diego, CA: 760-942-3252, FAX 760-9424979; Pacific Northwest: 503-963-1940 PRODUCTION DIRECTOR Wendy K. Shank; ASSISTANT MANAGER Rebecca Doshi; SENIOR SPECIALISTS Steve Forrester, Chris Redwood, Anthony Rosen; PREFLIGHT DIRECTOR David M. Tompkins; MANAGER Marcus Spiegler; SPECIALISTS Jason Hillman, Tara Kelly ART DIRECTOR Yael Fitzpatrick; ASSOCIATE ART DIRECTOR Laura Creveling; SENIOR ILLUSTRATORS Chris Bickel, Katharine Sutliff; ILLUSTRATOR Yana Hammond; SENIOR ART ASSOCIATES Holly Bishop, Preston Huey, Nayomi Kevitiyagala, Matthew Twombly; ART ASSOCIATE Kay Engman; PHOTO EDITOR Leslie Blizard
SCIENCE INTERNATIONAL EUROPE (
[email protected]) EDITORIAL: INTERNATIONAL MANAGING EDITOR Andrew M. Sugden; SENIOR EDITOR/COMMENTARY Julia Fahrenkamp-
Uppenbrink; SENIOR EDITORS Caroline Ash, Stella M. Hurtley, Ian S. Osborne, Peter Stern; ASSOCIATE EDITOR Maria Cruz; LOCUM EDITOR Helen Pickersgill; EDITORIAL SUPPORT Samantha Hogg, Alice Whaley; ADMINISTRATIVE SUPPORT John Cannell, Janet Clements, Nicola Morris; NEWS : DEPUTY NEWS EDITOR , U . K . Daniel Clery; NEWSWRITER Sara Reardon; CONTRIBUTING EDITOR, EUROPE Martin Enserink; CONTRIBUTING CORRESPONDENTS Michael Balter (Paris), John Bohannon (Vienna), Gretchen Vogel (Berlin) ASIA Japan Office: Asca Corporation, Tomoko Furusawa, Rustic Bldg. 7F, 77 Tenjin-cho, Shinjuku-ku, Tokyo 162-0808, Japan; +81 3 6802 4616, FAX +81 3 6802 4615,
[email protected]; ASIA NEWS EDITOR Richard Stone (Beijing:
[email protected]); CONTRIBUTING CORRESPONDENTS Dennis Normile [Japan: +81 (0) 3 3391 0630, FAX +81 (0) 3 5936 3531;
[email protected]]; Hao Xin [China:
[email protected]]; Mara Hvistendahl [China: mara@ marahvistendahl.com]; Pallava Bagla [South Asia: +91 (0) 11 2271 2896;
[email protected]]
Tom Daniel, Univ. of Washington Stanislas Dehaene, Collège de France Emmanouil T. Dermitzakis, Univ. of Geneva Medical School Robert Desimone, MIT Claude Desplan, New York Univ. Ap Dijksterhuis, Radboud Univ. of Nijmegen Dennis Discher, Univ. of Pennsylvania Gerald W. Dorn II, Univ. of Cincinnati College of Medicine Jennifer A. Doudna, Univ. of California, Berkeley Julian Downward, Cancer Research UK Bruce Dunn, Univ. of California, Los Angeles Christopher Dye, WHO David Ehrhardt, Carnegie Inst. of Washington Tim Elston, Univ. of North Carolina at Chapel Hill Gerhard Ertl, Fritz-Haber-Institut, Berlin Barry Everitt, Univ. of Cambridge Paul G. Falkowski, Rutgers Univ. Ernst Fehr, Univ. of Zurich Tom Fenchel, Univ. of Copenhagen Alain Fischer, INSERM Anne C. Ferguson-Smith, Univ. of Cambridge Wulfram Gerstner, EPFL Lausanne Karl-Heinz Glassmeier, TU Braunschweig Diane Griffin, Johns Hopkins Bloomberg School of Public Health Elizabeth Grove, Univ. of Chicago Taekjip Ha, Univ. of Illinois at Urbana-Champaign Christian Haass, Ludwig Maximilians Univ. Steven Hahn, Fred Hutchinson Cancer Research Center Gregory J. Hannon, Cold Spring Harbor Lab. Martin Heimann, Max Planck Inst., Jena Isaac Held, NOAA James A. Hendler, Rensselaer Polytechnic Inst. Janet G. Hering, Swiss Fed. Inst. of Aquatic Science & Technology Ray Hilborn, Univ. of Washington Michael E. Himmel, National Renewable Energy Lab. Kai-Uwe Hinrichs, Univ. of Bremen Kei Hirose, Tokyo Inst. of Technology David Hodell, Univ. of Cambridge Ove Hoegh-Guldberg, Univ. of Queensland David Holden, Imperial College Lora Hooper, UT Southwestern Medical Ctr at Dallas Jeffrey A. Hubbell, EPFL Lausanne Steven Jacobsen, Univ. of California, Los Angeles Kai Johnsson, EPFL Lausanne Peter Jonas, Universität Freiburg William Kaelin Jr., Dana-Farber Cancer Inst. Barbara B. Kahn, Harvard Medical School
www.sciencemag.org
SCIENCE
SERVICE SUPERVISOR
Deborah RiveraRandy Yi; MANAGER, FULFILLMENT SYSTEMS Frits Buningh; SYSTEMS ANALYST Nicole Mehmedovich; MANAGER , BUSINESS ANALYSIS Eric Knott; MANAGER , BUSINESS OPERATIONS Jessica Tierney; BUSINESS ANALYSTS Priti Pamnani, Celeste Troxler; Christine Wehrli; RIGHTS AND PERMISSIONS : ADMINISTRATOR Emilie David; ASSOCIATE Elizabeth Sandler; MARKETING DIRECTOR Ian King; MARKETING MANAGERS Allison Pritchard, Alison Chandler, Julianne Wielga, Samantha Smith; MARKETING ASSOCIATES Aimee Aponte, Mary Ellen Crowley; SENIOR MARKETING EXECUTIVE Jennifer Reeves; DIRECTOR, SITE LICENSING Tom Ryan; DIRECTOR, CORPORATE RELATIONS Eileen Bernadette Moran; SENIOR PUBLISHER RELATIONS SPECIALIST Kiki Forsythe; PUBLISHER RELATIONS MANAGER Catherine Holland; PUBLISHER RELATIONS, EASTERN REGION Phillip Smith; PUBLISHER RELATIONS, WESTERN REGION Ryan Rexroth; CUSTOMER RELATIONS MANAGER Iquo Edim; MARKETING MANAGER Christina Schlecht; MARKETING ASSOCIATE Chad Johnson; ELECTRONIC MEDIA: DIRECTOR Lizabeth Harman; ASSISTANT MANAGER Lisa Stanford; PRODUCTION SPECIALISTS Antoinette Hodal, Nichele Johnston, Kimberly Oster; DIRECTOR, WEB AND NEW MEDIA Will Collins; PROJECT MANAGER Trista Snyder; COMPUTER SPECIALISTS Walter Jones, Kai Zhang ADVERTISING DIRECTOR, WORLDWIDE AD SALES Bill Moran COMMERCIAL EDITOR Sean Sanders: 202-326-6430 ASSISTANT COMMERCIAL EDITOR Tianna Hicklin 202-326-6463 Wienhold;
VOL 334
ADMINISTRATION
DIRECTOR
P R O D U C T (
[email protected]);
M I D W E S T Rick Bongiovanni: 330-405-7080, FAX 330-405-7081; EAST COAST / E . CANADA Laurie Faraday: 508-747-9395, FAX 617-507-8189; WEST COAST/W. CANADA Lynne Stickrod: 415-931-9782, FAX 415-5206940; UK/EUROPE/ASIA Roger Goncalves: TEL/FAX +41 43 243 1358; JAPAN ASCA Corporation, Makiko Hara: +81 (0) 3 6802 4616, FAX +81 (0) 3 6802 4615;
[email protected]; CHINA/TAIWAN Ruolei Wu: +86 1367 1015 294
[email protected] WORLDWIDE ASSOCIATE DIRECTOR OF SCIENCE CAREERS Tracy Holmes: +44 (0) 1223 326525, FAX +44 (0) 1223 326532 CLASSIFIED (
[email protected]); U.S.: MIDWEST/WEST COAST/ SOUTH CENTRAL/CANADA Tina Burks: 202-326-6577; EAST COAST/INDUSTRY Elizabeth Early: 202-326-6578; SALES ADMINISTRATOR Marci Gallun; E U R O P E / R OW SA L E S Simone Jux, Alex Palmer; SA L E S A S S I S TANT Sarah Rouquette; JAPAN Makiko Hara +81 (0) 90 98539982, FAX +81 (0)3-6369-4491;
[email protected]; CHINA / TAIWAN Ruolei Wu: +86 1367 1015 294
[email protected]; ADVERTISING SUPPORT MANAGER Karen Foote: 202-326-6740; ADVERTISING PRODUCTION OPERATIONS MANAGER Deborah Tompkins; SENIOR PRODUCTION SPECIALIST / GRAPHIC DESIGNER Amy Hardcastle; PRODUCTION SPECIALIST Yuse Lajiminmuhip; SENIOR TRAFFIC ASSOCIATE Christine Hall; SALES COORDINATOR Shirley Young AAAS B OARD OF D IRECTORS RETIRING PRESIDENT , CHAIR Alice Huang; PRESIDENT Nina Fedoroff; PRESIDENT - ELECT William Press; TREASURER David E. Shaw; CHIEF EXECUTIVE OFFICER Alan I. Leshner; BOARD Nancy Knowlton, Stephen Mayo, Raymond Orbach, Julia M. Phillips, Sue V. Rosser, David D. Sabatini, Inder Verma, Thomas A. Woolsey
Daniel Kahne, Harvard Univ. Bernhard Keimer, Max Planck Inst., Stuttgart Joel Kingsolver, Univ. of North Carolina at Chapel Hill Robert Kingston, Harvard Medical School Alberto R. Kornblihtt, Univ. of Buenos Aires Leonid Kruglyak, Princeton Univ. Mitchell A. Lazar, Univ. of Pennsylvania David Lazer, Harvard Univ. Virginia Lee, Univ. of Pennsylvania Ottoline Leyser, Cambridge Univ. Olle Lindvall, Univ. Hospital, Lund Marcia C. Linn, Univ. of California, Berkeley John Lis, Cornell Univ. Jianguo Liu, Michigan State Univ. Richard Losick, Harvard Univ. Jonathan Losos, Harvard Univ. Ke Lu, Chinese Acad. of Sciences Laura Machesky, CRUK Beatson Inst. for Cancer Research Andrew P. MacKenzie, Univ. of St Andrews Anne Magurran, Univ. of St Andrews Oscar Marin, CSIC & Univ. Miguel Hernández Charles Marshall, Univ. of California, Berkeley Martin M. Matzuk, Baylor College of Medicine Graham Medley, Univ. of Warwick Yasushi Miyashita, Univ. of Tokyo Richard Morris, Univ. of Edinburgh Edvard Moser, Norwegian Univ. of Science and Technology Sean Munro, MRC Lab. of Molecular Biology Thomas Murray, The Hastings Center Naoto Nagaosa, Univ. of Tokyo James Nelson, Stanford Univ. School of Med. Timothy W. Nilsen, Case Western Reserve Univ. Pär Nordlund, Karolinska Inst. Helga Nowotny, European Research Advisory Board Luke O'Neill, Trinity College, Dublin Stuart H. Orkin, Dana-Farber Cancer Inst. Christine Ortiz, MIT Elinor Ostrom, Indiana Univ. Andrew Oswald, Univ. of Warwick Jane Parker, Max-Planck Inst. of Plant Breeding Research Donald R. Paul, Univ. of Texas at Austin P. David Pearson, Univ. of California, Berkeley Reginald M. Penner, Univ. of California, Irvine John H. J. Petrini, Memorial Sloan-Kettering Cancer Center Simon Phillpot, Univ. of Florida Philippe Poulin, CNRS Colin Renfrew, Univ. of Cambridge Trevor Robbins, Univ. of Cambridge Barbara A. Romanowicz, Univ. of California, Berkeley
Published by AAAS
AND
BUSINESS SYSTEMS AND FINANCIAL ANALYSIS DIRECTOR
18 NOVEMBER 2011
Jens Rostrup-Nielsen, Haldor Topsoe Edward M. Rubin, Lawrence Berkeley National Lab Mike Ryan, Univ. of Texas, Austin Shimon Sakaguchi, Kyoto Univ. Miquel Salmeron, Lawrence Berkeley National Lab Jürgen Sandkühler, Medical Univ. of Vienna Randy Seeley, Univ. of Cincinnati Vladimir Shalaev, Purdue Univ. Joseph Silk, Univ. of Oxford Denis Simon, Univ. of Oregon Alison Smith, John Innes Centre Davor Solter, Inst. of Medical Biology, Singapore John Speakman, Univ. of Aberdeen Allan C. Spradling, Carnegie Institution of Washington Jonathan Sprent, Garvan Inst. of Medical Research Elsbeth Stern, ETH Zürich Ira Tabas, Columbia Univ. Yoshiko Takahashi, Nara Inst. of Science and Technology John Thomas, Duke Univ. Herbert Virgin, Washington Univ. Bert Vogelstein, Johns Hopkins Univ. Cynthia Volkert, Univ. of Gottingen Bruce D. Walker, Harvard Medical School Douglas Wallace, Leibniz Inst. of Marine Sciences Ian Walmsley, Univ. of Oxford David A. Wardle, Swedish Univ. of Agric Sciences David Waxman, Fudan Univ. Detlef Weigel, Max Planck Inst., Tübingen Jonathan Weissman, Univ. of California, San Francisco Sue Wessler, Univ. of California, Riverside Ian A. Wilson, The Scripps Res. Inst. Timothy D. Wilson, Univ. of Virginia Jan Zaanen, Leiden Univ. Kenneth Zaret, Univ. of Penn. School of Medicine Mayana Zatz, University of Sao Paolo Jonathan Zehr, Univ. of California, Santa Cruz Huda Zoghbi, Baylor College of Medicine Maria Zuber, MIT
BOOK REVIEW BOARD
John Aldrich, Duke Univ. David Bloom, Harvard Univ. Angela Creager, Princeton Univ. Richard Shweder, Univ. of Chicago Ed Wasserman, DuPont Lewis Wolpert, Univ. College London
879
NEWS OF THE WEEK Berlin 3
AROUND THE WORLD
5 4
Baikonur Cosmodrome, Kazakhstan 1
Phobos-Grunt Still Earth-Bound Russia’s first solar system exploration mission since 1996 remains stranded in Earth orbit this week after the engines that should propel it to Mars failed to fire on 9 November. After a perfect launch, the Phobos-Grunt probe—intended to return samples from Mars’s moon, Phobos—has remained stubbornly silent. Vladimir Popovkin, head of the Russian Federal Space Agency, told reporters Monday that the chances of salvaging the mission were low. Ground controllers don’t know what went wrong with the craft because they haven’t reestablished contact with the onboard flight control computer. But Popovkin said the craft is maintaining its orientation relative to the sun, so its batteries should remain charged. He added that attempts to communicate with the spacecraft would continue for some weeks. “I still have some hope. We still have a few days for
In limbo. Phobos-Grunt, prelaunch.
880
1
reprogramming before the end of the Mars accessibility window for 2011,” says Lev Zelenyi, director of the Institute of Space Research (IKI) in Moscow which developed the mission. Attention is now turning to the possibility of an uncontrolled reentry between the end of November and mid-January. PhobosGrunt carries 10 tons of toxic fuel, but this is expected to burn up on reentry. Further updates will be posted on ScienceInsider. http://scim.ag/ScienceInsider
Menlo Park, California 2
Geron bails out of stem cells Geron, the company that helped pioneer human embryonic stem (hES) cell research, said this week that it is stopping its firstin-the-world clinical trial and pulling out of further stem cell work. The Menlo Park, California-based company will instead concentrate on its anti-cancer therapies. Geron helped to fund the work of James Thomson at the University of Wisconsin, who in 1998 was the first to isolate hES cells (Science, 6 November 1998, p. 1145). The trial, launched last year, was designed to treat eight patients with spinal cord injury using neuronal cells derived from hES cells. Four people have been treated so far, and CEO John Scarlett, who joined Geron 2 months ago, says there have been no adverse effects. The company will continue following the patients, but it will not enroll any new participants. Scarlett said the decision enables Geron to continue operating without raising new money for the next year and a half, when it expects results from a half-dozen phase II trials of two cancer drugs. As part of the stem cell downsizing, the company will cut 66 full-time positions, 38% of its workforce. http://scim.ag/Geron
18 NOVEMBER 2011
VOL 334
SCIENCE
Published by AAAS
Ankara 4
Turkish Academy Members Resign in Protest
Members of the Turkish Academy of Sciences (TÜBA) are making good on their threat to resign in protest of what they see as government intrusion on the autonomy of the organization. At least 60 members—more than one-third of the total—have said they will cancel their membership. The government’s initial plan, announced 27 August, would have expanded TÜBA’s membership from its current 140 members to 300: 100 appointed by Turkish Prime Minister Recep Erdoğan, 100 appointed by the government-run Council of Higher Education, and the rest elected by sitting members. That announcement garnered letters of opposition from numerous scientific academies and societies. On 4 November, the government made what appeared to be a concession, stating
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
2
For more than 3 decades, Germany has been storing its waste at a site called Gorleben in Lower Saxony near the former border between East and West Germany. Germany has spent an estimated €1.6 billion building and testing the Gorleben facility, but the choice has long been controversial. At a meeting 11 November between the federal environment minister and officials from the 16 German Länder (states), leaders agreed to work toward drafting a law to govern a new search. A task force will begin work this month, and should have a draft law ready by next summer. The search will consider sites all over the country, including Gorleben, said environment minister Norbert Röttgen at a press conference. There will be “no taboos.” Leaders in the southern states of BadenWürttemberg and Bavaria, which host a majority of the country’s nuclear power plants, had long argued against a new site selection process, in part because granite and clay deposits there are likely to be on the list of possible candidates. But officials in both states now say they are open to a new search.
CREDITS (TOP TO BOTTOM): © CHRISTIAN RUMMEL/ISTOCKPHOTO.COM; OLEG URUSOVZ/ITAR-TASS/NEWSCOM
Germany to Look for New Nuclear Waste Site
3
NEWS THEY SAID IT “Even though Rick Perry’s life was not being threatened, his brain was responding as if there was a lion in the audience about to pounce on him.”
Airlifting Rhinos to a New Home
CREDITS (TOP TO BOTTOM): GREEN RENAISSANCE/WWF; TILBURG UNIVERSITY; DOE
It was a daring, but photogenic, plan: To help save the species, 19 southern black rhinos were wafted by helicopter out of their habitat in South Africa’s Eastern Cape last week. The airlift was part of the World Wildlife Fund’s (WWF’s) Black Rhino Range Expansion Project, which works with landowners to add habitat—particularly important for black rhinos, which are less social than white rhinos and require more space. Many rhino species are in trouble: The western black rhinoceros was declared officially extinct in the wild by the International Union for Conservation of Nature on 10 November. But the southern black rhinoceros is still hanging on, with 4880 animals in the wild, including 1915 in South Africa. Since 2003, the WWF project has transported nearly 120 animals. Because these 19 rhinos lived in a region without roads, helicopters airlifted the sedated animals out by the ankles on a 10-minute trip across the treetops, where trucks were waiting to carry them 1500 kilometers to their new home in Limpopo Province.
that it would appoint a small committee to appoint the 100 academy members. But this is just “a cosmetic change,” says Erol Gelenbe, a computer scientist at Imperial College London and one of the resigning TÜBA members. Those who have resigned plan to form their own organization, called Science Academy Society, which will retain TÜBA’s infrastructure and will be open to all its members. http://scim.ag/TurkishAcad Washington, DC 5
First 2012 Spending Bill Backs Science Programs
The first 2012 budget bill contains surprisingly good news for the U.S. scientific community. At press time, Congress was expected to approve legislation that gives the National Science Foundation (NSF) a 2.5% increase, preserves funding for NASA’s James Webb Space Telescope, and provides enough money for the National Oceanic and Atmospheric Administration
to continue its polar-orbiting environmental satellites program. The spending bill (http://scim.ag/ fy2012approp) is linked to an extension of current spending levels for all federal agencies, a necessary step to avoid a government shutdown. But for the science agencies included in this $182 billion slice of the overall U.S. budget, the bipartisan support for basic research was very gratifying. NSF’s increase, for example, comes after the GOPled House of Representatives approved a flat budget and the Democratic Senate applied a 2.5% cut. But rather than splitting the difference, the conferees from each body added $173 million to NSF’s pot. The rising cost of the $8.7 billion Webb telescope will come out of other parts of NASA’s $17.8 billion budget, which is some $650 million less than this year, including some science programs. And essentially all of NOAA’s $306 million increase was given to the Joint Polar Satellite System, a two-satellite system scheduled to have its first launch in 2016.
www.sciencemag.org
SCIENCE
VOL 334
Published by AAAS
Downloaded from www.sciencemag.org on November 17, 2011
—David Diamond, a behavioral neuroscientist at the University of South Florida in Tampa, explaining to The Washington Post how the Republican presidential candidate struggled during a debate on 9 November to remember the name of a department he wanted to eliminate: the Department of Energy.
NEWSMAKERS
Disgraced Dutch Psychologist Returns Doctoral Degree After a devastating report accusing him of fraud in dozens of papers, Dutch social psychologist Diederik Stapel has given up his doctor’s title. The University of Amsterdam, where Stapel Stapel worked from 1993 until 1999, issued a short statement 10 November saying that he had voluntarily relinquished the degree and returned his diploma. The panel investigating Stapel’s misconduct had concluded that it was impossible to determine whether his 1997 thesis was based on fraud, in part because the data had been destroyed. But it recommended that the university investigate whether Stapel could be stripped of his title, “on the grounds of exceptional academically unworthy conduct.”
DOE Science Boss Steps Down Steven Koonin is leaving his job as undersecretary for science in the Department of Energy (DOE) after what others say was an unhappy stint in a poorly defined position. Koonin was nomi- Koonin nally responsible for scientific activities throughout DOE, but in practice had no control over budgets—not even for DOE’s $4.8 billion Office of Science. “Here was a guy who had no budget authority, and that’s a tough position,” says Michael Lubell, a lobbyist with the American Physical >> Society in Washington, D.C.
18 NOVEMBER 2011
881
NEWS OF THE WEEK
It started as a joke. Michael Schreiber, editor-in-chief of EPL, wanted to dispel the notion that his journal was only for European science. During a video interview with physicsworld.com last May to commemorate EPL’s 25th anniversary, Schreiber noted that the journal takes papers from India, China, and Brazil. “If a scientist would submit an excellent paper from the ISS, the International Space Station, I’m sure we would accept it too,” he added. “It was said, more or less, in fun,” Schreiber says now. But in the back of his mind, he adds, he was thinking about a talk he’d just heard by plasma physicist Hubertus Thomas at the Max Planck Institute for Extraterrestrial Physics in Garching, Germany, who presented data collected from the ISS. Schreiber wrote to Thomas, sending him a link to the video interview and posing a question: Would Thomas and his team of German and Russian scientists consider submitting their research to EPL—but from the ISS? After obtaining permission from the Russian government, the team agreed. On 27 October, Russian cosmonaut Sergey Alexandrovich Volkov submitted a manuscript (published online 11 November) about complex plasmas in microgravity conditions from the ISS; his address on the paper is listed as “International Space Station (present address).” That Volkov, who performed many of the plasma experiments, happened to still be on the ISS was a lucky break for the journal, Schreiber says: The cosmonaut was originally scheduled to return to Earth 2 months ago but troubles with the Soyuz rocket delayed the launch that would have brought him home. Having submitted the first paper from space, Volkov will soon be on his way home: A Soyuz rocket bound for the space station launched suc- Floating text. The submitted manuscript in the ISS (top). Cosmonaut Sergey Volkov (bottom at right). cessfully 14 November.
>>NEWSMAKERS
Koonin acknowledges the issue but says that he was effective since taking the post in May 2009. “In terms of what actually gets done, I think I influenced things quite a bit,” he says. For example, Koonin helped build bridges between researchers doing largescale computer simulations in DOE’s nuclear weapons programs and scientists in other parts of DOE who could turn such simulations to other problems. He also headed DOE’s first Quadrennial Technology Review, released 27 September, which aims to explain and sharpen DOE’s mission. Koonin’s next stop is the Institute for Defense Analyses’ Science and Technology Policy Institute in Washington, D.C. However, he hopes to have a post at a university lined up by the next academic year.
882
FINDINGS
When the Bough Doesn’t Break Leonardo da Vinci observed that when a tree’s trunk splits into two branches, the total cross section of those secondary branches will equal the cross section of the trunk. Now, a new study suggests that trees may grow this way to resist wind damage. Christophe Eloy, a physicist at the University of California, San Diego, who is also affiliated with University of Provence in France, found the wind connection by modeling a tree as beams anchored at one end and assembled into a fractal, a shape that can be split into parts, each of which is a smaller version of the larger structure. So when a branch split into smaller branches, it split into the same number of branches, at
18 NOVEMBER 2011
VOL 334
SCIENCE
Published by AAAS
90 million Number of children who contract seasonal influenza annually around the world, according to a study published online 10 November in The Lancet.
approximately the same angles and orientations. Most natural trees grow in a fairly fractal fashion. Eloy modeled the force of wind blowing on a tree’s leaves as a force pressing on the unanchored end of the beam. When he plugged that wind force equation into his model and assumed that the probability of a branch breaking due to wind stress is constant, he came up with Leonardo’s rule, Eloy reveals in a paper soon to be published in Physical Review Letters. http://scim.ag/Leotrees
Biggest of Mass Extinctions Looking Even Worse The most detailed look yet at events at the end of the Permian Period 252 million years ago confirms a link between the biggest and baddest of eruptions and the most massive of extinctions. Paleontologist Shu-zhong Shen of Nanjing Institute of Geology and Palaeontology in China and his co-authors intensively sampled the fossil records at nine sites across South China and dated 300 mineral grains using the radioactive decay of uranium to lead. The extinction in the sea took no longer than 200,000 years and likely less than a geologically fleeting 100,000 years, the researchers report in a paper published online this week in Science. A similarly abrupt extinction struck rainy forests at exactly the same time. And the new, highly precise age for the extinction puts it within a mere tens of thousands of years of the great volcanic outpourings that formed the Siberian Traps, says geochronologist Paul Renne of the Berkeley Geochronology Center in California. Somehow, the nasty spewings of the eruptions— greenhouse gases and acid-generating sulfur among them—must have done in most of Permian life. http://scim.ag/Permianext
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
First Paper Submitted From Space
BY THE NUMBERS $850,000 Estimated price of a meteorite, found by a Missouri farmer in 2006, now identified as a rare pallasite.
CREDITS (TOP TO BOTTOM): M. SCHREIBER, EPL 96 (DECEMBER 2011) © EPLA, 2011, (PHOTO) RSC-ENERGIA; NASA
Random Sample
NEWS & ANALYSIS
U.S. RESEARCH FUNDING
CREDIT: PETER AND MARIA HOEY, WWW.PETERHOEY.COM
NSF Creates Fast Track for Out-of-the-Box Proposals Nothing strikes the National Science Foundation (NSF) closer to the bone than criticism that its vaunted peer-review system is too conservative. For years, the community and Congress have pressured NSF to lay more bets on research that is risky but that could rock our world if it succeeds. NSF officials have always insisted that everything the basicresearch foundation supports is “potentially transformative” and that there’s no need for a special program. Until now. Last week, NSF Director Subra Suresh unveiled an initiative (www. nsf.gov/pubs/2012/nsf12011/nsf12011.jsp) that aims to roll the dice on a relative handful of researchers with unorthodox ideas about how to tackle complex problems. At $24 million, the Creative Research Awards for Transformative Interdisciplinary Ventures (CREATIV) will take up only a tiny portion of NSF’s $5.5 billion research portfolio. But NSF hopes it will send a big signal to the U.S. research community. “It’s a new way of doing business for NSF,” says Richard Behnke, who co-chaired an internal NSF committee that designed the new program. The biggest difference between proposals submitted to CREATIV and those processed in the traditional manner is that the former won’t be judged by external review panels. Instead, researchers need only win over NSF program managers. In fact, researchers must receive prior, written approval from at least two program managers before even submitting a proposal. But once that hurdle has been cleared, researchers should expect an answer within 2 or 3 months. That’s more than twice as fast as the usual turnaround time. Researchers from U.S. institutions can submit proposals in any area that NSF now funds, and there are no priority topics. The maximum grant is $1 million over 5 years, and proposals will be accepted starting 1 December and processed on a first-come, first-served basis. In an NSF-sponsored webcast (www. tvworldwide.com/events/nsf/111109/), Suresh said that not every research idea is right for CREATIV. The most distinguishing characteristic is what he called “interdisciplinary proposals [without] a
recognizable home” within the foundation. NSF officials say they have no idea how many researchers will submit CREATIV proposals. But Samuel Rankin, head of the Coalition for NSF Funding and associate executive director of the American Mathematical Society, thinks that it will be popular and that it meets a need. “Given the conservative nature of review panels and reviewers,” he says, “I think having a [separate] mechanism that requires transformative research can be helpful in acquiring and evaluating such proposals.”
NSF officials acknowledge that they are stepping into a long-running controversy by launching CREATIV. “There are quite diverse opinions on how significant the problem is and what needs to be done,” notes Thomas Russell, the internal panel’s second co-chair. Russell and Behnke are part of a group of a dozen NSF senior managers who began meeting last spring to discuss how to flesh out an initiative in NSF’s current budget request called INSPIRE (Integrated NSF Support Promoting Interdisciplinary Research and Education). Starting with $12 million in 2012, Suresh hopes that INSPIRE will eventually grow to be a $120-million-a-year effort by 2016. CREATIV is the first of what Suresh hopes
www.sciencemag.org
SCIENCE
VOL 334
Published by AAAS
will be several activities under the INSPIRE umbrella. “It was seen as an effective and relatively easy-to-implement first step,” Russell says. Another idea on the table is greater support for research larger in scope than what is typically carried out by individual investigators but smaller than what is done at a center. Pending approval by Congress, CREATIV will actually have access to $24 million— $12 million from the director’s office and a matching amount from NSF’s six directorates. That amount might support 40 to 50 awards, Behnke says. Although CREATIV is a pilot program, NSF expects it to continue in 2013 and beyond. Some scientists worry that CREATIV could favor researchers with a track record of NSF support, those who are familiar with its inner workings and who enjoy a close relationship with their program manager. Suresh says that’s not his intent and that the
Downloaded from www.sciencemag.org on November 17, 2011
NEWS & ANALYSIS
need to get approval to apply from more than one NSF staffer will prevent such bias. “But we’ll be monitoring it closely to see how it evolves,” he adds. There are also concerns that the new program could undermine NSF’s traditional merit-review process. In response, Behnke noted that the program, even at its projected maximum size, would represent less than 2% of NSF’s overall research budget. “For the great majority of proposals, we will continue the traditional merit-review process,” he said. “The gold standard remains in place.” Amen to that, Rankin says. “As the CREATIV program grows, I hope that NSF balances it with the regular review process.”
18 NOVEMBER 2011
–JEFFREY MERVIS
883
Research Projects Could Be Roadkill In Revision of Massive Highway Bill When engineers rebuilding a beach near Lewes, Delaware, in 2004 began finding bits of Colonial-era pottery mixed in with the sand, archaeologists quickly realized they had found a historic shipwreck. Historians believe it is the British vessel Severn, sunk by a storm in 1774, and the 45,000 artifacts discovered to date are “world class,” says David Clarke, the state archaeologist for the Delaware Department of Transportation. “Nowhere else do we have this kind of stuff from this period.” Delaware received $200,000 to help finance the underwater dig from a federal program that funds transportation-related projects in the 50 states. Since 1992, the program has provided more than $51 million to archaeological projects like the Severn dig and more than $100 million to environmental restoration projects and studies aimed at preventing deadly collisions with wildlife. But that funding stream is now threatened, as Congress debates the value of the Transportation Enhancements Program (TEP). Last week, a U.S. Senate committee made the first move, significantly modifying the program as part of a plan to reauthorize the government’s $50-billion-a-year surface transport funding system. But TEP faces serious opposition in the House of Representatives from lawmakers who consider it to be wasteful and unnecessary spending. Many archaeologists “are watching the process with bated breath,” Clarke says. Under current law, states must spend 10% of certain federal transport funds on enhancement projects. Those can include 12 types of activities, including building sidewalks and bike trails, beautifying roadways, and saving
884
historic sites. States can also fund “archeological planning and research” and “environmental mitigation,” including projects to prevent polluted storm-water runoff and preserve wetlands. In practice, about 2% of overall transport budgets go to enhancements, an amount that has totaled about $12 billion since 1992. About 75% of those funds have gone to walking and biking programs, with environmental and archaeology projects getting less than 1% each. Still, TEP funding has been important to those sparsely funded fields. Archaeologist Clarke, for instance, says “we couldn’t have done the Lewes project without it.” The existing highway law expired in 2009. Since then, Congress has passed short-term extensions, but efforts to complete a longterm deal have been complicated by a steep decline in gas tax revenues, the law’s main source of money. Many states, meanwhile, complain that the law is too complicated and onerous. Earlier this year, House and Senate leaders pledged to do a major rewrite, sparking debate over TEP’s future. Last month, for example, Republican senators John McCain (AZ), Rand Paul (KY), and Tom Coburn (OK) each offered unsuccessful amendments to pending legislation to end or scale back the program. One “wasteful” project they repeatedly highlighted was a multimillion-dollar “turtle tunnel” that allows reptiles and other animals to pass under Highway 27 near Lake Jackson, Florida. Those attacks “were completely misguided,” argues ecologist Matt Aresco of the Nokuse Plantation, a reserve in Bruce, Florida. As a graduate student, Aresco documented that cars were killing tens of thou-
18 NOVEMBER 2011
VOL 334
SCIENCE
Published by AAAS
www.sciencemag.org
CREDIT: BOB HALLINEN/ANCHORAGE DAILY NEWS/MCT/GETTY IMAGES
U.S. CONGRESS
sands of reptiles and amphibians trying to cross Highway 27 and spent a decade spearheading efforts to build fences and passages to end the carnage. Highway officials, he notes, backed the idea because of safety concerns: “You get hundreds of turtles on the road, and people are swerving all over and slamming on the brakes.” In Idaho, Nicholas Sharp, a biologist with the Wildlife Conservation Society, is using TEP funds to study the threat to both wildlife and humans posed by the moose and elk populations crossing the busy Highway 20. “When you hit a moose, it usually doesn’t end up good for the moose or the driver,” he says. In last week’s action, the Senate Committee on Environment and Public Works stopped short of ending TEP in its bill, dubbed Moving Ahead for Progress in the 21st Century. Instead, the panel approved a plan to consolidate TEP with several other programs as part of a bid to simplify funding and give states a greater say in spending. Under the legislation, the TEP pot could shrink by 25% and be divided more ways. The program could be in for an even rougher ride in the House. Representative John Mica (R–FL), chair of the Transportation and Infrastructure Committee, has said that enhancements won’t be included in the reauthorization bill he expects to release later this year. And House Majority Leader Eric Cantor (R–VA) has promised that states “will not be required to spend a specific amount of funding on specific types of projects, such as transportation museums or landscaping.” An outline of the House bill also calls for deep cuts in overall spending. Such potential outcomes “have a lot of people a little afraid of what is going to happen,” says Julie Schablitsky, chief archaeologist at the Maryland Department of Transportation’s State Highway Administration. Her state has used TEP funds for some high-profile projects, including $2 million to help raise and study a warship from the War of 1812. Although many states don’t tap TEP funds for archaeology, she predicts that states with a commitment to preservation will continue to fund archaeology and other enhancement projects even if Congress scales back the program: “Some understand the value of cultural and natural resource projects and will continue to invest.” The current extension of the federal surface transportation bill expires next spring. That’s also when TEP’s fate is likely to be determined. –DAVID MALAKOFF
Downloaded from www.sciencemag.org on November 17, 2011
Highway havoc. Efforts to reduce deadly collisions between wildlife and cars could be curtailed.
NEWS&ANALYSIS
SCIENTIFIC COMMUNITY
CREDIT: TOBY SAVAGE
They haven’t packed their bags or booked their flights, but now that Libya’s civil war is over, many foreign archaeologists are eager to get back to the country and pick up their studies some 9 months after beating a hasty, sometimes panicked retreat. They’ll land in a country whose revolution has claimed tens of thousands of lives but whose abundant, little-known cultural heritage remains virtually unscathed. Many of those who have conducted studies in Libya hope that from the turmoil will arise a new, democratic nation that invests more of its oil wealth in research and takes a keener interest in its archaeological treasures. “Libya has relied heavily on outside scientific expertise,” says David Mattingly, an archaeologist at the University of Leicester, U.K., who has worked in the country for 3 decades. “What it needs to be thinking most about now is a large-scale investment in its own capacity in archaeology.” Indeed, Libyan scientists are hoping for a broad scientific revival. After the international sanctions against their country were lifted in 2004, Muammar Gaddafi promised to turn it into a beacon of science; plans included an African center for disease surveillance and research and new investments in astronomy, said to be one of Gaddafi’s passions (Science, 8 April 2005, p. 182). “In the end, that was mostly talk and propaganda,” says Bashir Elmejrab, a Libyan geophysicist working for Shell. “Nothing really happened.” What’s needed, he adds, is an ambitious reform and modernization of the education and research systems. As for archaeology, Mattingly and colleagues tend to burst out in superlatives when they describe Libya’s riches, which include f ive UNESCO-designated World Heritage Sites: the Greek city of Cyrene; Leptis Magna, the Roman city where emperor Septimius Severus was born; a Phoenician trading post called Sabratha; Ghadamès, an
ancient oasis town; and a vast collection of rock paintings near the Algerian border. On top of that, there are massive hidden treasures, says Mattingly, who runs a U.K. project in the Sahara called Desert Migrations, spanning everything from the northward movement of early hominins to ancient desert civilizations and 19th century trade routes. “The evidence is everywhere you look, and it’s of extraordinary value,” he says. “It’s the envy of the rest of the world.” When the war broke out, many worried about destruction or looting at the heritage sites or Libya’s museums. But so far, the damage appears to be miraculously limited, according to a report by a four-person mission from Blue Shield and the International Military Cultural Resources Working Group—two groups aimed at protecting heritage sites—that toured Libya in late September. The team found some minor bullet damage at the amphitheatre in Sabratha and heard reports about four amphorae stolen from the museum in Apollonia, but they uncovered little else of concern. That conf irms what archaeologist Vincent Michel of the University of Poitiers in France has heard. Not a single item appears to be missing from the country’s most important museum in Tripoli, says Michel, who plans to return to Libya for a few days in December to prepare further excavations in April. What may have saved Libya’s treasures, Michel says, is that unlike Egypt or Syria, it never had much tourism. As a result, “nobody would even know who to sell stolen or pillaged antiquities to,” he says. The transitional government’s plan to tap Libya’s tourist potential worries Michel, who says it should be accompanied by a plan to protect archaeological sites and finds. Mattingly has his own concerns. Drought has preserved exquisite artifacts in the
www.sciencemag.org
SCIENCE
VOL 334
Published by AAAS
Hidden secrets. Members of the Desert Migrations Project at work at a Garamantian castle in southern Libya.
Sahara extraordinarily well from weathering, he says, but they’re vulnerable. Already, construction and oil exploration projects start without much consideration for what may lie beneath, he says, and even tourists’ four-wheel drive vehicles may crush evidence hidden in the desert. Just how Libyan archaeology will evolve depends to a large degree on what happens to the country’s Department of Antiquities. Its director, Salah Agab—whom Science was unable to contact—was suspended from his job, Michel says, like many other highranking officials, but he has since been reinstated. Mattingly hopes that Agab will keep his post, calling him a “wonderful guy and genuinely someone of vision.” But the department needs to be rebuilt and strengthened, Mattingly adds. Scientists from other disciplines who have worked in Libya are eager to return as well. “If things settle down really quickly, I could be back next year,” says University of Chicago paleontologist Paul Sereno, who visited Libya to finalize a study agreement 10 days before the uprising started. With Libyan partners, Sereno hopes to search for dinosaur fossils in a mountain ridge extending along the coast. “It’s really terra incognita,” he says, “and as the Earth gets smaller, it’s great to have a place that is that unknown.” “Gaddafi cared a lot about security but not really about science and education,” says Salem Sharata, who teaches geology at the University of Az Zawiyah. Yet Sharata senses that Libya is finally moving in the right direction. But it will continue to need help, he says. “I hope you guys won’t leave us alone.” –MARTIN ENSERINK
With reporting by David Malakoff.
18 NOVEMBER 2011
885
Downloaded from www.sciencemag.org on November 17, 2011
Revolution Brings New Hopes For Libyan Archaeology
NEWS&ANALYSIS Heavy on the carbs. China emits so much CO2 that tallying errors can throw off global emissions estimates.
BEIJING—Like other members of China’s del-
egation to the Copenhagen Summit in 2009, geophysicist Ding Zhongli returned home with fire in his heart. The talks, dogged by controversy, failed to find a solution to how to equitably rein in carbon dioxide (CO2) emissions. Some observers laid the blame on rising powers like China, Brazil, and India, which rejected binding reductions. Ding and his compatriots chafed at what they saw as an injustice. “Before 1990, developed countries emitted so much CO2. They should do something about it,” Ding says. “We need to allocate emission rights in a fair way.” An equitable solution is proving elusive, and with the Kyoto Protocol set to expire in
2012, time is running out. As nations grope for a consensus, China is pressing ahead on its own to sharply reduce energy intensity by shuttering inefficient coal-fired power plants and capping energy use (Science, 8 February 2008, p. 730). Last week, the State Council approved a plan to promote low-carbon energy and slash CO2 emissions by 17% per unit of GDP by 2015. But these efforts mask major uncertainties in China’s carbon balance sheet: just how much CO2 the country emits and how much its landscape absorbs. Ding, a vice president of the Chinese Academy of Sciences (CAS), is leading an ambitious new program that aims to answer these questions. In a letter to Premier Wen
An Unsung Carbon Sink
Limestone cowboy. As a proxy for CO2 absorption, Cao Jianhua measures dissolved ions in a karst formation in southwest China.
886
GUILIN, CHINA—Cao Jianhua bounds up the craggy karst, leaving his panting colleagues far behind. Halfway up the 250-meter-high limestone outcrop, he meticulously arranges vials and water droppers the way a surgeon lays out scalpels. At this site near Old Dragon Spring, water is gradually dissolving calcite, a reaction that consumes carbon dioxide (CO2) and spits out what Cao, a soil scientist here at the International Research Center on Karst, intends to measure: calcium and bicarbonate ions. “We’re working backwards to figure out how much CO2 has been taken out of the air,” he says. The answer could have global implications. Carbonate karst formations cover roughly 15% of Earth’s land surface, including broad swaths of southwestern China. Limestone degradation could be a substantial inorganic carbon sink, says George Veni, executive director of the National
18 NOVEMBER 2011
VOL 334
SCIENCE
Published by AAAS
Cave and Karst Research Institute in Carlsbad, New Mexico. The Guilin team, adds Nico Goldscheider, a hydrogeologist at the Karlsruhe Institute of Geology in Germany, “is doing pioneering work to understand and quantify the role of karst processes as a global carbon sink.” As China embarks on a campaign to trace the flow of carbon on land, sea, and air (see main text), inorganic sinks are now increasingly understood as a critical part of the equation. A few years ago, for instance, scientists discovered that alkaline soils in China’s Gubantonggut Desert and the U.S.’s Mojave Desert absorb CO2 (Science, 13 June 2008, p. 1409). Because almost a third of Earth’s land surface is desert or semiarid, “the total carbon absorption in deserts should be significant in the global sense,” says Li Yan, a plant ecophysiologist at Xinjiang Institute of Ecology and Geography in Urumqi. “If you think of the global carbon cycle like a bank account, we’re trying to keep track of all the deposits and withdrawals impacting the level of CO2 in the atmosphere,” says geologist Chris
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
China Looks to Balance Its Carbon Books
CREDITS (TOP TO BOTTOM): STEPHEN SHAVER/UPI/NEWSCOM; C. LARSON
C L I M AT E C H A N G E
Jiabao on the heels of the Copenhagen talks, Ding proposed that China invest heavily in climate science and emissions monitoring. The result is a 5-year, $125 million initiative just getting under way that will compile a CO 2 emissions inventory, explore carbon sequestration and how China can adapt to climate change, and probe how the climate responds to rising greenhouse gas levels. The initiative involves hundreds of researchers at two dozen CAS institutes and four other government bodies, including the powerful National Development and Reform Commission (NDRC), which oversees economic planning. It wasn’t easy getting the project off the ground. Ding met resistance from other agencies when he sought to build broad support for the initiative earlier this year. In part that’s because Ding has openly challenged aspects of the scientific consensus on climate change, such as whether current agreed targets—such as capping atmospheric CO2 concentrations at 450 parts per million and limiting global temperature rise to 2°C on average—are meaningful. Ding believes the climate may not be as sensitive to rising CO2 levels as models predict. “If you look at the data in the last 200 years, there are so many uncertainties,” he says. In a March 2010 interview with China’s CCTV, Ding was blunter, comparing the U.N. Intergovernmental Panel on Climate
Change’s (IPCC’s) use of computer modeling to forecast rising temperature to a fortune teller gazing into a crystal ball. Some Chinese scientists question Ding’s motives. “Doubting IPCC’s findings in an effort to reduce the pressure on China to curb carbon emissions will not help China’s economic development,” says one Chinese IPCC delegate. But Ding insists that the new initiative is a purely scientific undertaking that will demonstrate China’s commitment to serious climate research. After months of internal discussions, China’s research community has largely rallied around the initiative, which has evolved into something broader and more ambitious than Ding first envisioned. One major new goal is to strengthen climate modeling by using China’s homegrown supercomputers and satellite imagery to investigate how greenhouse gas emissions influence climate change. Scientists will also take a stab at a comprehensive tally of China’s CO2 emissions. The project will calculate emissions by economic sector, relying on researchers to collect energy-use figures for categories such as construction, agriculture, and mining. The carbon numbers will be compiled in a national database, Ding says, and fed to NDRC, which reports greenhouse gas emissions to the U.N. Framework Convention on Climate Change secretariat. A better understanding of China’s emissions would be welcome, says Gregg Marland, an environmental scientist at Appalachian
State University in Boone, North Carolina, pressure to curb emissions, Ding’s initiative and a lead author of IPCC’s 2006 national carries political implications. Pang Zhonghe, emission inventory guidelines. In a 2008 anal- an energy expert at CAS’s Institute of Geolysis, Marland and colleagues concluded that ogy and Geophysics who shares Ding’s belief estimates of Chinese CO2 emissions may be that the link between CO2 emissions and risoff by as much as 20%, compared with a max- ing temperatures requires more proof, says the imum 5% error for U.S. data. Because China project may help guide China’s climate policy. emits more CO2 than any For example, if the findother nation, any large ings show that China has error would skew global underestimated the stortotals, Marland says. age capacity of its carbon “Emissions from China sinks, a revision would are so huge that it really allow climate negotiators matters,” he says. to argue that a larger porDing hopes the initiation of China’s emissions tive will also illuminate are being offset. how much CO 2 ChiThe initiative may na’s landscape absorbs indeed give negotiators and identify sequestraheftier scientific clout. tion sites. China’s car- Lightning rod. Project director Ding At future climate change bon sinks, including Zhongli has questioned some aspects of the summits, Chinese delpreviously underval- international consensus on climate change. egates would be able to ued ones such as desbring higher-quality data erts and karst formations (see sidebar), are to the table—a contribution that will “increase believed to suck in 260 million tons of CO2 their bargaining power,” argues physicist Feng a year—roughly equivalent to the U.S. total. An, executive director of the Innovation CenChina hopes to boost its overall carbon sink ter for Energy and Transportation in Beijing to 416 million tons by 2020 through conser- and Los Angeles, California. vation and reforestation. To firm up these figDing is fully aware that some in the interures, CAS researchers will fan out to collect national community may regard the project atmospheric data at thousands of sites. Other with suspicion. However, he says, “the most teams will scrutinize China’s six major refor- important thing is to gather enough data.” estation projects to track CO2 uptake by for- That’s one point, at least, on which there is litests, a major natural sink. tle debate. –LI JIAO AND RICHARD STONE At a time when China faces mounting Li Jiao is a writer in Beijing. With reporting by Ni Wei.
Groves, director of the Hoffman Environmental Research Institute at Western Kentucky University in Bowling Green. “Right now, the numbers don’t all add up: There’s a missing carbon sink scientists are hunting for.”
Sweeting of the University of Oxford, to come for lecturing stints. Yuan and Guilin proved irresistible. “If you want to learn about karst systems,” says Groves, who first met Yuan in 1995 and now collaborates with Cao’s team, “this is one of the top places in the world to be.” But karst was a domain for a small band of specialists until researchers began to realize that limestone could be a significant CO2 sink. To probe this phenomenon more deeply, Groves says, the carbon study aims to “finetune how you make detailed estimates” of CO2 absorption rates in karst landscapes. It’s complicated, he says, because factors such as rainfall patterns, land use, limestone composition, and acidic pollution all affect CO2 absorption. Taking a stab for one region, Cao’s group estimates that karst in southern China’s Pearl River Basin—about 17% of China’s surface karst— absorbs about 2 million tons of CO2 a year, the team reported in the April issue of Chinese Science Bulletin. To put that number in perspective, the U.S. Geological Survey estimates that
CREDIT: R. STONE/SCIENCE
The allure of limestone
When Yuan Daoxian came to this region in 1959 to pioneer the study of China’s karst formations, villagers warned him that the peaks jutting like jade hairpins from verdant rice paddies were haunted. That didn’t deter the hydrogeologist, who recorded, among other things, how weather alters geology—more rainfall equals more limestone dissolution—and how the region’s aquifers are relatively inaccessible. In 1978, Yuan helped found the Institute of Karst Geology, which houses the international research center. He’s the “godfather” of karst in China, Groves says. After China began to open up in 1978, Yuan, now 78, says he recognized “a gap between the science of karst in China and in Western nations.” He wooed experts, including the late Marjorie
www.sciencemag.org
SCIENCE
VOL 334
Published by AAAS
Downloaded from www.sciencemag.org on November 17, 2011
NEWS&ANALYSIS
volcanoes belch about 200 million tons of carbon a year. Human activity releases roughly 24 billion tons of CO2 annually. Hints are emerging about how to enlarge the limestone sink. Early results indicate that CO2 absorption is more efficient in karst covered by soil and vegetation; among other mechanisms, soil cover accelerates dissolution by keeping limestone in contact with water longer. Reducing soil erosion, “a major problem in karst areas,” Goldscheider says, may augment CO2 absorption. Casing out this underappreciated carbon sink means digging deep as well as reaching high. The day after Cao’s climb, institute colleague Zhang Cheng led a team into Pan Long Cave. Rivulets seeping through cracks in the limestone glistened in the halos of flashlights: a promising spot to measure calcium and bicarbonate. “This is where we could set our instruments,” Zhang said, the geologist’s deep voice echoing in the chamber. In the hunt for missing CO2, no limestone will be left unturned. –CHRISTINA LARSON Christina Larson is a writer in Beijing.
18 NOVEMBER 2011
887
Will Busting Dams Boost Salmon? ELWHA RIVER, WASHINGTON—“It’s like an alien abduction,” says George Pess, a fisheries biologist with the National Oceanic and Atmospheric Administration (NOAA) Fisheries in Seattle, Washington, as he walks past a pair of colleagues preparing to snatch their quarry. One of those colleagues, Martin Liermann, wades calf deep in a side channel of the Elwha River in Washington’s Olympic
Peninsula with a giant battery pack strapped to his back. Holding a 2-meter-long yellow pole connected to the pack, he dunks the business end of it in the water and pulls a red trigger. An electric jolt stuns a handful of steelhead fingerlings, a jolt that Liermann avoids thanks to the rubber chest waders and neoprene boots he’s wearing. Liermann’s partner Holly Coe scoops
up a few fingerlings and plunks them into a white 5-gallon bucket. When they’ve gathered a couple of dozen or so, they shuffle over to what looks like a makeshift surgical unit strewn amid the rocks on the river’s edge. “Have you puked these guys over here?” Liermann asks Sarah Morley, a fisheries biologist with NOAA Fisheries. “Just one,” Morley replies, as she plucks a fish out of the
Glines Canyon Dam on Its Way Down Between September and November 2011
888
18 NOVEMBER 2011
VOL 334 SCIENCE www.sciencemag.org Published by AAAS
CREDITS: NATIONAL PARK SERVICE (6)
As dams fall on the Elwha River in Washington state and other rivers around the country, scientists are relishing rare opportunities to watch natural ecosystems restore themselves
Downloaded from www.sciencemag.org on November 17, 2011
NEWSFOCUS
NEWSFOCUS
CREDITS: NATIONAL PARK SERVICE (6)
Online
www.sciencemag.org SCIENCE VOL 334 Published by AAAS
time, the focus has been on tearing out small, obsolete structures in the Midwest and East. But that pattern has begun to shift, says Martin Doyle, a river restoration expert at Duke University in Durham, North Carolina. “Bigger dams are starting to come out, out west,” Doyle says, in part as an effort to rebuild endangered salmon and steelhead runs. Even longtime proponents of dam building, such as officials at the U.S. Bureau of Reclamation, now support removals on a case-by-case basis. At a recent ceremony commemorating the demolition of the Elwha dams, Bureau Commissioner Michael Connor was enthusiastic. “This is not only an historic moment, but it’s going to lead to historic moments elsewhere across the country,” Connor was quoted in the Los Angeles Times. All these dam removals have created a unique opportunity for scientists to study how quickly rivers revert to their old ways once obstructions come down. “We’re stoked,” says Mike McHenry, a habitat biologist with the Lower Elwha Klallam Tribe in Port Angeles, Washington, who has worked on river-restoration projects on the Elwha since 1989. “The hope here is we’ll learn something we can apply in other places.”
Downloaded from www.sciencemag.org on November 17, 2011
Morley’s team also surveys aquatic invertebrates and algae to get a full sense of what types of food the salmon have available to them. Others collect tissue samples from the fish for genetic studies; track numbers of bears, otters, and other mammals in the area; sample isotopes of nitrogen and phosphorus nutrients in the water to determine the percentage that comes from the ocean, brought No more. The Elwha Dam (above), built in 1913, is now largely in by spawning salmon; and gone (previous page). All that remains is the original power house even cut out the ear bones of dead salmon that have already and much of the sediment that was lodged behind the dam. spawned to look for telltale bucket, quickly measures and weighs it, and signs of whether they came from hatcheries carefully wedges a syringe needle in the fish’s or were wild stock. mouth and pushes the plunger. Water shoots This intense interest came to a head as into the fish’s stomach and forces the con- engineers were set to begin tearing out two tents out onto a fine mesh sieve. “He had a massive dams that have blocked the free flow full belly,” she says as she spots a small slug, of water and salmon on the river for nearly a few mayflies, a couple of leeches, and some 100 years. The dams—Glines Canyon and green bits she can’t identify. Morley carefully the Elwha—are some 64 and 33 meters high, spoons the contents into a sample jar for later respectively, the largest dams ever removed lab analysis and returns the fish to the bucket, in the United States. It will take as long as after which it will be returned to the Elwha, 3 years to tear them out completely. But bullhungry but otherwise unharmed. dozers and excavators have already made There’s good reason to be careful with the quick progress. Water now flows around steelhead. They’re one of three fish the carved-out right side of species in the Elwha listed as either the lower Elwha Dam and over threatened or endangered under the the top of four notches hacked Endangered Species Act. (Chinook sciencemag.org into the face of Glines Canyon Podcast interview salmon and bull trout are the othDam. Today, the river is close to with author ers.) Morley and Liermann’s work Robert F. Service. flowing freely for the first time on a bright day here in September is in 98 years. one of nearly a dozen related research projThe Elwha isn’t alone. After a century ects designed to document the current state of furious dam construction in the United of the river. Just downstream, Pess and three States, the era of dam building has ebbed. colleagues sort through a small wedge of the Over the past 3 decades, increasing concerns streambed measuring and weighing rocks about safety and the high cost of upkeep have and other pieces of sediment. Upstream, pushed the number of removals up from an John McMillan, a NOAA Fisheries contrac- average of about 10 per year to six times that tor in a black dry suit, blue mask, and snorkel, number today, with a total of nearly 1000 sloshes his way down the river, occasion- dams removed, according to numbers kept ally plunging in to count the number of adult by American Rivers, a river advocacy group salmon he sees. based in Washington, D.C. For much of this BEFORE
Tipping the scales There are plenty of possibilities. According to the U.S. Army Corps of Engineers, the United States has more than 80,000 dams 3 meters tall or taller, most of them privately operated. River blockages of all sizes likely top 2 million, according to another survey. Many of these aging structures have fallen into disrepair, are too expensive to fix, or have been made obsolete as nearby mills have closed and large coal and natural-gas power plants have come online. “The need for these dams just doesn’t exist anymore,” Doyle says. “If you have a dam that’s no longer generating revenue, you are basically sitting on a liability.” Privately held dams are licensed by the Federal Energy Regulatory Commission (FERC). As licenses run out, owners can apply for 35to 50-year extensions. In most cases, however,
18 NOVEMBER 2011
889
NEWSFOCUS
Stockholm Environment Institute (SEI) in Davis, California, calls the assessment “important” but notes that individual rivers will be affected not only by temperatures and stream flows but also by how humans manage water flows through dams and other diversions. Purkey, along with colleagues at SEI, the University of California, Davis, and the National Center for Atmospheric Research in Boulder, Colorado, recently ran a similar climate-change analysis on the Butte Creek Watershed in California. Their study took into account not only hydrology, CO2 emissions scenarios, and six different climate models but also two different scenarios describing how dam operators managed water flows throughout the year. In a business-as-usual scenario in which operators didn’t change water flows during summer temperature spikes, spring Chinook salmon runs became extinct by about 2070. But when water was allowed to flow through the dams in the hot summer months when the fish are most vulnerable, the additional cold water in the rivers allowed at least small runs to survive the century in some, although not all, scenarios. One potential bit of good news for the salmon, Stanford says, is that rivers along the North Slope of Alaska and other arctic regions that typically freeze for much of the year are likely to become new free-flowing spawning habitats as temperatures warm. The North Pacific, however, would provide far smaller feeding ground for salmon than the current region between the northwestern United States and Japan in which the fish now spend the bulk of their lives. Ocean acidity at high latitudes may pose an even bigger threat. Fossil fuel burning emits carbon dioxide into the atmosphere; the gas is absorbed by the oceans and quickly converted to carbonic acid. This reaction releases hydrogen ions into the water, reducing its pH—in other words, making the water acidic. Since preindustrial times, the overall pH of the oceans has declined by 0.1 pH unit. That may not sound like much, but the pH scale,
the requirements for relicenslater, to provide electricity for ing dams are now far more strinsawmills in nearby Port Angegent than when the dams went les. By the 1990s, most of those in, meaning that complying with Good news, bad news. Dam removals are likely to increase salmon popula- mills had closed. The dams still environmental and safety stan- tions, but climate change and ocean acidification could do the opposite. provided power—but a total of dards can cost more money than only 19 megawatts, far below the the dams generate. So it’s often cheaper to tear salmon and steelhead runs in the basin are 500 to 1000 megawatts of a typical gas- or out the dams than to bring them up to current now listed as either endangered or threatened. coal-fired plant. When it came time for FERC standards. In recent years, Pennsylvania and To counter that decline, fisheries man- relicensing, it was clear the costs of repairWisconsin have led the way in dam removals, agers now mandate that most dams provide ing the dams and adding fish passages didn’t each state tearing out hundreds. fish passage by building a parallel “ladder” pencil out. Still, the company that owned the Finding the proper balance between the or waterway. For dams that historically had dams, James River Corp., didn’t have the usefulness of dams and their environmen- no such passage, that is typically an expen- money to remove them. tal costs has become particularly acute in the sive proposition; in many cases it has now In 1992, the U.S. Congress stepped in and Northwest, where communities are struggling tipped the scales in favor of dam removal. In authorized money to tear out the dams. But to safeguard salmon populations that have 2008 and 2009, for example, that equation congressional opponents, including former been declining for decades. In the Columbia led Portland General Electric to rip out two Washington State Senator Slade Gordon, River Basin, which at 673,000 square kilo- power-producing dams on the Sandy River blocked dam-removal funding proposals meters stretches across most of the North- in suburban Portland. Dams have also come for more than a decade. Among Gordon’s western U.S. states and into British Columbia, out along the Rogue River in southern Ore- concerns were that if dam removal sucdams have blocked an estimated 55% of his- gon, and the Wind and White Salmon rivers ceeded in the Elwha, there would be a clamor toric salmon spawning habitat, says Michelle in southwest Washington. to remove dams throughout the west— McClure of NOAA Fisheries. Those dams, particularly those on the lower Snake River together with loss of habitat from develop- Unique opportunity in Idaho, which have long drawn the ire of ment, as well as competition from hatchery- Much the same calculation was at work on salmon advocates but have been a boon bred salmon, have dropped wild salmon runs the Elwha. The Elwha Dam was installed to farmers looking to barge their goods to to roughly 10% of their historic numbers; 13 in 1913, and Glines Canyon Dam 14 years coastal ports. “There was concern that when
890
18 NOVEMBER 2011
VOL 334 SCIENCE www.sciencemag.org Published by AAAS
Downloaded from www.sciencemag.org on November 17, 2011
Despite the continuing rise in dam removals, not all long-term trends are good news for salmon. Climate change and ocean acidification, according to recent studies, threaten to wipe out possible gains and could push many salmon runs to extinction by the end of this century. A new climate-modeling effort called the Riverscape Analysis Project finds that climate change will affect rivers throughout the Pacific Rim. The effort, led by Jack Stanford and John Kimball at the University of Montana’s Flathead Lake Biological Station in Polson, starts with a standard hydrologic model for how temperature and precipitation changes affect river temperatures and stream flows. The researchers incorporated information from satellite and space shuttle images on the physical characteristics of 1500 rivers, as well as the extent of existing salmon runs, and examined how expected air temperature changes given by scenarios from the Intergovernmental Panel on Climate Change would likely affect river temperatures, flows, and fish populations. Stanford says that when he and his colleagues input past climate data, their models could explain 70% of the observed variation in river flows and temperatures. Not perfect, but by a modeler’s standards, “that’s pretty good,” Stanford says. When they looked forward to the end of this century, they found that the model flagged dozens of rivers at serious risk of losing their salmon. The most vulnerable tend to be in California and the interior regions of the Columbia Basin in Oregon, Washington, and Idaho. The results are preliminary, Stanford emphasizes, as he and his team continue to refine their models. But the general picture is clear, Stanford says: “Salmon will experience a warming [of water temperatures] across their entire range. In some places, it will get so warm that salmon will be eliminated.” David Purkey, a hydrologist at
CREDIT: LEN KANNAPELL
Out of the Frying Pan?
CREDITS (TOP TO BOTTOM): RICHARD FEELY ET AL., SCIENCE 320 (13 JUNE 2008); LEN KANNAPELL
like the Richter scale for earthquakes, is logarithmic: Over this period, seawater acidity has increased by 30%. According to emission scenarios from the Intergovernmental Panel on Climate Change, pH values will drop an additional 0.2 to 0.3 units by 2100, potentially doubling current acidity levels. Added hydrogen ions reduce seawater concentrations of carbonate ions, which oysters and other organisms use to build their shells. Lab and field studies suggest the drop in carbonate ions is already affecting oyster larvae and other organisms. The trend is particularly bad news for juvenile salmon: One of their primary food sources in the Pacific is pea-sized snails called pteropods. Even without added CO2 from human civilization, pteropods and other shelled critters in some coastal regions periodically experience high acidity levels. It happens when populations of algae, plankton, and other ocean organisms bloom and die, falling to the ocean floor. There they are gobbled up by bacteria that churn out carbon dioxide, making deep waters highly acidic. When surface winds push the top layer of water out to the open ocean, the deep corrosive waters from underneath surge upward, explains Richard Feely, a chemical oceanographer with the National Oceanic and Atmospheric Administration’s Pacific Marine Environmental Laboratory in Seattle. At the American Fisheries Society meeting in Seattle in September, Feely reported that he and colleagues had just completed a cruise throughout the North Pacific and down the West Coast of the United States. According to the team’s latest surveys, ocean acidification from human causes already accounts for between 29% and 49% of the corrosive waters present along the West Coast of the United States through the North Pacific. The human contribution to this high acidity could rise to as much as 80% by 2050, Feely estimates. In a more-acidic ocean, pteropods and other organisms probably won’t be able to build their shells. So the ability of salmon to thrive in the ocean could hinge on their ability to find food sources different from those they’ve relied on throughout the course of evolution. –R.F.S.
one dam falls, they are all going to fall,” says Patrick Crain, a fisheries biologist with the National Park Service. Ultimately, however, the Elwha was just too good a restoration opportunity to pass up. The Elwha dams sit close to the river’s mouth where it flows out into the Strait of Juan de Fuca, near where Puget Sound meets the Pacific Ocean. Upstream, 83% of the 112-kilometer-long river lies within the protected wilderness of Olympic National Park. Before the dams were built, an estimated 390,000 salmon from 10 different species returned to the river to spawn each year— so many that early white settlers along the river complained that the slapping of fish was so loud during spawning periods that they couldn’t sleep. Now that the dams are coming out, the Elwha has become not just the nation’s second largest ecological restoration site after the Florida Everglades, but a rare opportunity to study how natural processes return to normal and how ecosystems respond. “It’s really a unique opportunity, because the habitat is in such good shape,” Pess says. That opportunity sent Pess and colleagues from several federal, state, and
Caustic. Highly acidic ocean waters (red) periodically well up to shallow
depths near the west coast of North America. As CO2 concentrations in seawater increase, ocean acidification is expected to spread.
says Peter Kiffney, a biologist with NOAA Fisheries in Seattle. As they began to collect data on the Elwha a few years ago, it became clear that the river was dramatically altered from its free-flowing past. Not only are the fish counts drastically reduced, so too are the numbers of eagles, bears, and other animals that used to depend on the robust fish runs for survival. Over the decades, fine gravel in which salmon prefer to lay their eggs has Near death. A female Chinook salmon is scarred and bat- largely washed out because it wasn’t being replaced, leaving mostly rocks tered from her final journey up the Elwha. softball-sized and larger. The dams tribal organizations fishing for money to even blocked plant seeds from washing downtrack the Elwha before and after the dams stream, fragmenting some plant communities came down. Despite the removal project’s in the watershed. $325 million price tag, little of it was dediSo how long will it take for the river to cated to monitoring efforts. In the end, that recover? “It will be a great experiment,” Pess was something of a benefit, Pess and oth- says. Initially, the changes may not bode ers say, because it forced researchers from well for the salmon. The biggest concern, numerous organizations to build a broad Pess says, is that 18 million cubic meters of network of collaborators to track the recov- silt and other sediment is trapped behind the ery. “We have very little information about dams, enough to fill 200,000 dump trucks; it how the ecosystem works. That’s why so will move downstream. Engineers are taking much effort is going into this project,” the dams down slowly to ensure that it doesn’t
www.sciencemag.org SCIENCE VOL 334 Published by AAAS
18 NOVEMBER 2011
891
Downloaded from www.sciencemag.org on November 17, 2011
NEWSFOCUS
NEWSFOCUS
892
ters of prime salmon-rearing habitat. About 1 million fish a year still returned to the lower river. Some 30 years later, fish ladders were built to reopen access to the upper river. In the 40 years after they went in, salmon populations in the river doubled. At the American Fisheries Society meeting in September in Seattle, Rory Saunders of NOAA Fisheries reported that after two dams were recently removed from the Sedgeunkedunk Stream in Maine, Atlantic salmon, sea lamprey, and alewife all showed improved runs within 2 to 3 years. “This was much faster than any of us expected,” Saunders says. Still, Kiffney cautions that although salmon and other fish can return quickly, the recovery of the broader ecosystem can be far slower. For much of the past decade since fish ladders were installed in 2003, Kiffney has tracked the rise in salmon spawning up the Cedar River near Seattle. “After the fish ladders went in, whoosh, salmon numbers started climbing,” Kiffney says. But the physical mass of all the salmon in the runs is still only 0.006 kilograms of fish per square meter of river—too low to make a broader impact. “In experiments we’ve done, we see ecosystem changes at 0.6 kg/m2. So we need a 100-fold increase,” Kiffney says. Some estimates suggest such a return to historical salmon runs could be in store for the Elwha, says William Robert Irvin, president of American Rivers. But Pess and others are more cautious. Self-sustaining populations of different salmon species are likely to take 10 to 30 years to become established, Pess says. “There will be recovery. But it won’t be the same as it was historically.” As the fish runs begin to improve, the first to spot the changes could well be Kent Mayer, a fish biologist with the Washington Depart-
18 NOVEMBER 2011
U.S. Dam Removals by Year 60 50 40 30 20 10 0
1991
2001
2006
2010
Busted. Dam removals continue to rise (above),
particularly in the East (top). But teardowns are increasing in the West in hopes of saving salmon runs.
ment of Fish and Wildlife. Last year, Mayer built a fish weir—essentially a temporary trap for large adult fish—2 kilometers below the Elwha Dam. The weir, which spans the full width of the river, is the largest such structure in the United States outside of Alaska. It enables Mayer and his colleagues to not only count fish moving upstream and downstream but also document their size and weight and take tiny tissue samples for genetic profiling before releasing them. Last month, on the day Pess, Morley, and their colleagues were weighing rocks and puking fingerlings, Mayer caught and released the first bright red sockeye salmon known to have spawned in the Elwha in modern memory—hopefully a harbinger that the long-extinct salmon run could soon return to the river. “There are days when I love my job,” Mayer says standing waist deep in the fish trap. “This is one.”
VOL 334 SCIENCE www.sciencemag.org Published by AAAS
1996
–ROBERT F. SERVICE
MAP AND GRAPH SOURCE: AMERICAN RIVERS
all wash out at once. Some of the sediment will actually be trucked out. Botanists will then race to plant native species in the banks of remaining sediments, in an effort to stabilize them—a particular worry during spring high water flows. High sediment levels can not only harm fish directly but can also suffocate eggs waiting to hatch. Such concerns prompted recovery officials to approve the construction of a new hatchery for steelhead and chum, coho, and pink salmon in the lower Elwha. But that hatchery itself has become highly controversial, as some researchers worry that hatchery-raised salmon will outcompete natives and further imperil wild fish runs. “I think the hatcheries [on the Elwha] are unneeded,” says Jack Stanford, an ecosystem scientist with the University of Montana’s Flathead Lake Biological Station in Polson. Stanford notes that every year, highly productive salmon streams such as the Copper River in Alaska carry sediment loads equivalent to what the Elwha will now face after the dams come out. Stanford has plenty of allies. On dedication day for the demolition of the Elwha and Glines Canyon dams, opponents filed a notice of intent to file a lawsuit to block the new hatchery. Even with sediment problems and hatchery competition, giving fish access to new river territory can produce a quick reaction. “The response of fish can be pretty immediate,” Pess says. “Salmon are very opportunistic creatures. If given the opportunity to utilize a habitat, they will use it.” Pess points to one such example on the Fraser River in British Columbia, Canada. A landslide in 1913 triggered by railroad builders trying to blast a route for a track through Fraser Canyon cut off 80% of 1400 kilome-
Downloaded from www.sciencemag.org on November 17, 2011
Dam Removals by State 2001–2010
NEWSFOCUS
Bigger, better. An experiment evolving big-
E VO L U T I O N A RY B I O LO G Y
Evolutionary Time Travel
CREDITS: WILLIAM RATLIFF/UNIVERSITY OF MINNESOTA (2)
With clever and challenging lab experiments, researchers are forcing species to become multicellular, develop new energy sources, and start having sex In December 2009, evolutionary biologist Michael Travisano was debating with his future postdoc William Ratcliff what to do next in their lab at the University of Minnesota (UMN), Twin Cities. They had just seen a talk on slime molds that delved into what it meant to be a multicellular organism. Under certain conditions, some of these single-cell amoebas can coalesce into masses of millions of cells that act in a coordinated fashion, as if a whole organism. Inspired by this, Travisano and Ratcliff began musing about how one of evolution’s apparently major leaps up the ladder, the jump to multicellularity, takes place. A typical evolutionary biologist might tackle this challenge by comparing fossils or genomes of related unicellular and multicellular species, but the duo had a more daring idea. They decided to try to force the evolution of yeast, normally a single-celled creature, into a multicellular one. “I wouldn’t have wagered a large sum of money that this would have worked,” Ratcliff recalls. “But if it [did] work, it would be the coolest thing we could think of.” Researchers have long deliberately bred animals, plants, and microbial species for specific purposes—leaner meat, droughtresistant plants, chemical-producing bacteria, and so forth—but what Ratcliff and Travisano wanted to do was probe how evolution itself happens, by forcing it to occur, under controlled conditions, as scientists
watch. Despite Ratcliff ’s reservations, they succeeded in producing multicellularity, at least in a limited form, in just 60 days. This type of research, known as experimental evolution, has existed almost since Darwin put forth his theories. The approach has risen in popularity over the past few decades, in large part thanks to the pioneering work of Richard Lenski. An evolutionary biologist at Michigan State University in East Lansing, Lenski has for several decades now conducted an ongoing study in which 12 populations of the bacterium Escherichia coli live, and evolve, in flasks with limited supplies of glucose for energy. In an extraordinarily long-term effort, Lenski and his lab members have followed more than 50,000 generations of E. coli and in so doing gleaned insights into the pace and reproducibility of microbial evolution (Science, 25 June 1999, p. 2108). Lenski’s work “really highlighted the power” of experimental evolution to other biologists, says Nick Colegrave, an evolutionary biologist at the University of Edinburgh in the United Kingdom. At this summer’s 13th Congress of the European Society for Evolutionary Biology in Tübingen, Germany, it was clear that the field of experimental evolution has itself evolved. Biologists today conduct controlled evolution studies with everything from viruses to fish. And as the multicellularity experiment conducted by Travisano and Ratcliff indicates, many are trying to
www.sciencemag.org SCIENCE VOL 334 Published by AAAS
Downloaded from www.sciencemag.org on November 17, 2011
ger yeast yielded tubes with more settled yeast through time (above, left to right) and resulted in multicellular organisms with dead cells (red) helping reproducing yeast fragment.
address complex questions such as how evolution fashions major changes in a creature’s lifestyle. One team at the Germany meeting reported tackling how sex evolves, for example, whereas another examined how an alga deals with losing access to light, its main source of energy. “Evolution is expanding from a strictly comparative and observational science to an experimental one,” says Graham Bell, an evolutionary biologist at McGill University in Montreal, Canada. Even the field’s pioneer admired the ambitious work presented in Tübingen. “As the field grows, people are thinking about more and more specific hypotheses and complex scenarios,” Lenski says. Dark science For Bell, the complex scenario was to try to get a plant to grow in the dark. The cells of plants, which survive by photosynthesis, are structured around harnessing light and converting it to chemical energy. What if there were no light? The plant might have to shift from depending on its photosynthetic machinery to another source of energy, perhaps the mitochondria that power most nonphotosynthetic eukaryotes. A few parasitic plants and at least one protist have made such a switch, and Bell wanted to see if he could drive that change in the lab. It “constitutes a new way for life” for a plant, he explains. Bell couldn’t test a typical plant—none grows and reproduces fast enough to make such an experiment in evolution feasible. So he turned to a photosynthetic microbe belonging to the plant kingdom, the singlecelled green alga called Chlamydomonas, often studied as a model system in cell and molecular biology. Bell knew this organism already had some ability to feed off acetate in a pinch and wondered if it could build on that to thrive in the dark. After all, he points out, “you can’t evolve something from nothing.”
18 NOVEMBER 2011
893
Dark days. Very few
algae survived growing in the dark (right), but those that did evolved a variety of colors and shapes (above).
of algal “survivors” are thriving in the darkness that would be lethal to the ancestors. The alga lines vary significantly in appearance. Some form clusters that are circular; others are ragged. Some are green, whereas others are yellow or white. Most can still grow in the light, but a few can’t and rely solely on acetate, Bell said: “You have a whole range of solutions to growing in the dark.” Bell and his colleagues are now looking at whether these transformed algae can mate with their original lines, or whether more traits than morphology and metabolism have changed through time. He has a list of genes involved with acetate processing that he will check for enabling mutations. At some point, Bell says he might put the acetate users back into the light to see if they can come to depend on photosynthesis again. The genes needed for photosynthesis have likely degraded, and he wants to know
894
whether the algae can fix the broken genes or whether other genes will be brought into play. With this work, we “gain insights that would otherwise be impossible,” Lenski says. Showing sex is good Aneil Agrawal, an evolutionary biologist at the University of Toronto in Canada, and his postdoctoral fellow Lutz Becks are also exploring how a species can go back and forth between certain lifestyles. They’re addressing the puzzle of sex. The evolutionary quandary is that requiring two individuals, typically a male and a female, to generate offspring makes reproduction much less efficient than asexual cloning, wherein each individual reproduces. Thus one would expect that even if sexual reproduction evolves, it shouldn’t persist should asexual individuals subsequently arise, unless the sexual mixing of genomes provided some significant advantage. Yet in most species, sex predominates. By the 1990s, researchers had put forth more than 20 theories to explain this puzzle. Some experimental work in yeast suggested sex was advantageous in changing environments. And a study in viruses, which swap DNA in a way similar to sex, indicated that sex thrives because it weeds harmful genetic mutations out of a population (Science, 28 November 1997, p. 1562). But for the most part, sex’s purported evolutionary advantages have gone untested. Three years ago, Agrawal and Becks decided to look more deeply at the longand short-term benefits of sex through a series of lab studies, some involving experimental evolution. Researchers had proposed that recombining genes through sex could lead in the short run to fitter offspring. In the long run, the many possible genetic combinations produced by sex mean that there will likely be more genetic variation in the population and greater ability to adapt rapidly—an advantage that could favor the maintenance of sex. The duo’s studies, which center on rotifers, microscopic animals found in lakes and ponds, are showing that it is easy for sex to develop in certain populations but difficult for it to persist. So-called Bdelloid roti-
18 NOVEMBER 2011
fers are famously asexual, but other rotifer species, including Brachionus calyciflorus, the one Agrawal and Becks study, sometimes resort to sexual reproduction but only in crowded conditions. (The buildup of a rotifer-secreted chemical induces this sexual behavior in crowds.) Becks and Agrawal began by testing descendants of wild-caught B. calyciflorus rotifer for reproductive fitness, counting the number of eggs offspring produced in conditions that favored either sexual or asexual reproduction. The asexual populations (in uncrowded conditions) produced more than twice as many offspring as the crowded sexual populations, confirming a large fitness cost for sex, they reported in the March Journal of Evolutionary Biology. The sexual populations were also not more variable in their fitness, suggesting there was no long-term potential benefit to sex per se. “It showed the big problem we have explaining sex,” Becks says. Next, Becks and Agrawal used experimental evolution methods, tweaking the rotifiers’ diet to look at the effect of environment on the balance between sexual and asexual reproduction. They fed one batch of 10,000 rotifers nitrogen-rich algae, a similarly sized batch got less-nutritious algae, and a third batch, broken into subgroups of about 5000 rotifers, were regularly exposed to both. Becks would weekly transfer 1% or 10% of each subgroup from the one kind of alga food source to the other, simulating migration between two environments. At the beginning of the experiment and after 6 and 12 weeks—45 and 90 generations— he tested the rotifers’ propensity for sex by exposing a subset from each batch to the sex-stimulating chemical and looking at the eggs produced. (Asexually produced eggs appear solid under the microscope, whereas sexually produced ones seem partially void.) In this way, Becks determined what proportion of the rotifers were able to switch to sexual reproduction. Rotifers maintained with a consistent food source—whether of high or low quality— produced half as many sexual eggs as rotifers regularly switching between the two kinds of food, Becks and Agrawal reported online 13 October 2010 in Nature. And when they followed the rotifers a month longer, they found propensity for sex increased in the rotifers with a mix of foods but declined in the batches where rotifers experienced a constant food environment. In a constant environment, sexual reproduction eventually disappeared altogether, Becks later found.
VOL 334 SCIENCE www.sciencemag.org Published by AAAS
CREDITS: GRAHAM BELL
“It was a big challenge,” Lenski says. But “experimental evolution offers a way to see the relevant processes in action.” Bell set up 2880 cultures of Chlamydomonas on acetate-laden media and left them in a corner of his lab in constant darkness. Every other month he transferred 5% of the algae-media mix in each culture into a new dish of media. Unless the alga was increasing its population 20-fold monthly, a sign of strong growth, it would eventually be diluted out of existence by the periodic transfers. About 90% of the cultures stopped growing within a year, but a few hundred alga lines kept pace, he reported at the meeting. After about 12 months, he started transferring these algae each month, then every other week, and finally weekly, such that only the fastest growers would survive. “You have to be fairly patient,” Bell says. Five years into the experiment, 241 lines
Downloaded from www.sciencemag.org on November 17, 2011
NEWSFOCUS
NEWSFOCUS
CREDIT: LUTZ BECKS/UNIVERSITY OF COLOGNE
Yeast beasts Just as the evolution of sex represented a major transition for life, so did the leap to multicellularity. When Travisano and Ratcliff began to consider what experimental conditions would encourage the yeast Saccharomyces cerevisiae to go multicellular, they focused on size. One rather obvious hallmark of multicellular creatures is that they are bigger than unicellular organisms. On its own, largeness could offer many evolutionary
advantages: a greater ability to access nutrients or avoid being eaten by small predators, for example. Still, finding the right selective force to make yeast go big was a challenge. Working with their UMN colleagues R. Ford Denison and Mark Borrello, Travisano and Ratcliff first tried exposing cultures of yeast to detergent, thinking that some might form a multicellular entity in which the outer cells would shield the inner ones from the detergent’s destructive power. But nothing survived. Next, they turned to gravity as the selective force, letting tubes of yeast in solution, after being shaken to distribute the microbes evenly, sit quiet for 45 minutes. The researchers then transferred the bottom 1% of the tube’s contents to new cultures. Bigger yeast, which would be more likely to settle out, should have a better chance of surviving
ultimately discovered that the yeast in these snowflakes make much less of the enzyme that enables normal separation. The snowflakes also began reproducing like a multicellular organism. Individual cells in the snowflakes would divide but not detach, enlarging the snowflake. And once the snowflake reached a certain size, it would fragment, releasing a daughter snowflake. At first the fragments were about equal size, but as more and more generations went by, a pattern developed: Snowflakes broke into a smaller and much larger part. That’s presumably “so it could produce more offspring by allocating fewer resources for each one,” Ratcliff said. The newly multicellular yeast also evolved a division of labor that facilitated its uneven fragmentation. At first all the cells in the snowflake divided. But after hundreds of generations, some cells in each snowflake stopped dividing and eventually died. These cells in effect were sacrificed for the benefit of the multicellular organism, becoming sites where the snowflakes fragmented, Ratcliff reported. “It’s the beginning of specialization,” Lenski says. Some scientists question whether the new complex yeast are true multicellular organisms, but others nonetheless praise the experiments. “The origin of multicellularity with the subsequent evolution of specialized cell lineages represents one of the most important transitions in the history of life,” David Reznick, an evolutionary biologist at the University of California, Riverside, says. Sex mania. In typically asexual rotifers, the ratio of sexually “I would have never dreamed that it derived eggs (darker) increases in novel environments but could be possible to study it from an decreases after conditions stabilize. experimental perspective.” By doing so, Ratcliff says, he’s these transfers. “It turns out that 45 minutes gained a better understanding of this tranis pretty lenient,” Ratcliff says. Almost all the sition. “The constraints on the evolution yeast settled out, so there was little selection of multicellularity may be lower than we for larger size. thought,” Ratcliff concludes. As a result, he To speed up the process, they instead gen- now believes multicellularity arose more tly spun the tubes for 10 seconds in a centri- often than researchers have realized—most fuge before transferring the bottom 1%. Two current estimates suggest it has emerged weeks later, ever-bigger pellets of yeast were about 20 times in various lineages—but then settling to the bottom of the tubes in two of subsequently faded away. the 10 cultures, Travisano reported at the The attempt to transform yeast into a meeting. Spherical clusters of cells loosely multicellular species also impressed Lenski. resembling snowflakes eventually dominated “That’s some of the neatest work going on all 10 setups. Tests showed that these weren’t right now,” he says. “It makes you think about simply individual cells that managed to stick what is a major transition” in evolution of life. together, as happens when yeast in brewing Indeed, Colegrave says, “experimental beer aggregate. Instead, these clusters arose evolution is an approach which can in prinbecause dividing yeast cells had lost the abil- ciple be used to address any of the big quesity to separate completely. The researchers tions in evolution.” –ELIZABETH PENNISI
www.sciencemag.org SCIENCE VOL 334 Published by AAAS
18 NOVEMBER 2011
Downloaded from www.sciencemag.org on November 17, 2011
Becks and Agrawal wanted to see whether the pattern observed in the simulated migration—more sex in a more diverse environment—held true if the rotifers simply were confronted with a completely novel environment. Over the course of a week, Becks gradually added low-quality food to batches of B. calyciflorus rotifers that were used to feeding on high-quality food, transforming their diet. In the first few days of the experiment, as he was changing the food out, the rotifer population started to crash, reaching a low in 12 days. At that time, sexual reproduction began to increase and continued to do so for about 3 weeks, Becks reported at the evolution meeting. Then the trend reversed: The population kept growing but became increasingly dominated by asexually reproducing rotifers, a trend that continued through the final 10th week of the experiment. “We saw this increase [in sex], and then it went down again,” he said. They repeated the experiment for a period of a month and got the same results. In contrast, in control batches kept on a consistent diet, sexual reproduction waned from week one. According to Becks, this experiment and the one with the mix of two food environments suggest that new challenges favor sex but only until a way of coping with that challenge has developed. “In the end, sex was only beneficial during the time that they adapted to their environments,” concluded Becks, who is now at the Max Planck Institute for Evolutionary Biology in Plön, Germany. Once the right adaptations had evolved, sex was no longer favored. “What’s remarkable is that they’ve developed a rapidly evolving empirical system that allows them to watch the evolution of sex in real time,” says Sarah Otto, an evolutionary biologist at the University of British Columbia, Vancouver, in Canada. “By tracking changes in the frequency of sex, the rotifer system promises to allow us to tease apart the mechanisms that promote the loss or maintenance of sex.” Lenski adds that “the work allowed them to challenge and support a classic model for the evolution of sex.”
895
NEWSFOCUS
N AT U R A L R E S O U R C E S
Downloaded from www.sciencemag.org on November 17, 2011
Dreams of a Lithium Empire SALAR DE UYUNI, BOLIVIA—Once a giant prehistoric lake, the empty expanse of Bolivia’s majestic Salar de Uyuni seems frozen in time. But beneath the surface of this 10,000-square-kilometer salt flat lies a dynamic zone of unknown depth partially comprised of mineral-laden brine. The geology has fascinated Guillaume Roelants, a Belgian nuclear engineer, for 30 years. Now it has the world’s attention, too: The salty liquid holds the planet’s largest reserve of lithium, the key ingredient for lithium-ion batteries. Bolivia wants to extract lithium to power the world’s future electric vehicles, and Roelants leads this nation’s quest. “This is not the first time Bolivia has a resource the rest of the world covets,” explains Roelants, 59, a nationalized Bolivian who’s lived here since 1981. Spanish colonizers drained the country’s vast silver wealth; Bolivia’s tin enriched Europe’s coffers in the 19th century. Recently, the Andean nation’s natural gas reserves, the second largest in South America, have been developed largely for export. Today, Mitsubishi, LG, Lithium Corporation of America, and more have all tried to make inroads here. But Evo Morales—the first indigenous president—has decided Bolivia will develop the industry on its own: state control, with Roelants at the scientific helm. Critics, from Japan to the United States, say Bolivia is squandering its good fortune: The country ought to bring in foreign experts and capital to master Uyuni’s complicated brine and environment. Roelants shrugs, black leather Harley-Davidson jacket dangling on a finger over his shoulder, concluding: “We will prove them wrong.”
896
Self-sufficient The elevated evaporation ponds rise from the vast expanse like a mirage. At first, they look like toys, staged on an endless white carpet. A salar this big eliminates depth perception. Only at a few hundred meters do the ski-masked workers seem life-size. Roelants is standing in faded black jeans and a pilling checkered sweater. He’s frustrated because it’s 1 p.m. on Saturday and he’s only just arrived. The father of three spends every weekend on the flat, normally arriving on Saturdays by 6:30 a.m. after a 10-hour overnight journey by bus on unpaved roads from La Paz. Delays on his first attempt flying into the newly completed Uyuni Airport have cost him half his working day. “I won’t be doing that anymore,” he says, conceding that his budget can’t accommodate the $175 weekly airfare anyway. Despite his annoyance, Roelants is upbeat as division heads of the Bolivian State Mining Corp. approach with worksite progress reports. Approximately 10 solar evaporation ponds covering 12 hectares lie in front of him, arranged in stages. Mineral-laden brine is pumped in at the front end; as it moves downstream, solar evaporation causes minerals to crystallize in fractions, isolating the unwanted ions in the brine—including sulfates, magnesium, and potassium. The final ponds will soon yield lithium chloride, to be converted into marketable lithium carbonate, the white powder base material for lithiumion batteries. Wind blows Roelants’s thinning white hair vertical as he explains that this is a personal dream reborn. In 1989, he persuaded the Belgian government to finance a lithium-
18 NOVEMBER 2011
extraction pilot project, “at that time, for use in ceramics and aluminum production,” Roelants recalls. But the Bolivian government wasn’t interested, so Roelants moved on: He established Bolivia’s borax mining industry and spent more than a decade working near the Salar de Uyuni as an adviser to local communities on agricultural production. After the 2005 presidential election, Roelants says, “representatives from these communities asked me to accompany them to a meeting with President Morales regarding regional economic projects.” The president listened to Roelants’s ideas, and after reviewing a feasibility study in 2008, committed $5.7 million in state funds for a lithium carbonate pilot plant. Roelants was appointed as head. “I stayed in Bolivia to be challenged by working in an impoverished country, with the people, rather than for multinational companies,” he says, “so this was a great opportunity.” Indeed, the possibilities are grand. The Salar de Uyuni is a bowl-shaped depression of alternating brine and clay layers. The deepest well, at 220 meters, does not touch rock, so Uyuni’s total yield is impossible to determine. The basin is considered to hold at least 100 million tons of lithium, and the southeastern portion boasts some of the world’s highest concentrations, reaching 4000 ppm of lithium. The U.S. Geological Survey estimates that Bolivia’s reserves are more than half of the world’s total. But Uyuni presents special challenges. Salt flats are like snowflakes: No two brines are alike, and evaporative conditions vary. So though the majority of the
VOL 334 SCIENCE www.sciencemag.org Published by AAAS
CREDIT: © DIEGO CASADAMON/SODAPIX/CORBIS
Bolivia is betting that a former nuclear engineer, Guillaume Roelants, will develop a new extraction process for the world’s largest lithium reserve
CREDIT: NOAH FRIEDMAN-RUDOVSKY
Liquid wealth. Bolivia’s lithium riches lie in the Salar de Uyuni, a vast salt reservoir south of La Paz.
world’s lithium carbonate is produced via solar evaporation from salt flat brine (it’s also possible to mine lithium from rock), Roelants’s research team had to develop a unique process. “There is no blueprint for this work,” says Ihor Kunasz, former chief geologist for U.S.-based Foote Mineral Co. and one of the world’s experts in brines. Each pond must maintain specific mineral concentration levels while maximizing evaporation rates. The world’s largest producer of lithium carbonate, SQM in Santiago, says it considered Uyuni but decided instead on what’s now their flagship operation in Chile’s Atacama salt flat. “Uyuni is problematic because of its brine’s magnesium levels and the evaporative conditions,” says Eduardo Morales, general manager of the Sociedad Chilena de Litio. Magnesium, he explains, is found in almost all salt flats and is difficult to separate because of its chemical similarity to lithium. The higher the magnesium-to-lithium ratio, the more complicated and costly lithium carbonate production will be. Uyuni’s ratio, Roelants admits, is 20:1; Atacama’s is 6:1. Evaporation rates at Uyuni, which is relatively cool at 3600 meters above sea level, are good but not great. Rain slows the process, too. “These projects are extremely complex especially with the challenges presented in Uyuni,” says geologist R. Keith Evans, who has worked in the industry for 30 years and is arguably the world’s best known consultant on lithium. “Bolivia should be looking for every expertise out there, but instead they decided to go it alone.” Roelants may be the nation’s lithium chief, but he has no formal training with the metal. Nor do the 20 meteorologists, chemists, and geologists that form Roelants’s R&D team. They are, though, all Bolivian and all graduates of Bolivian public universities. The government stresses that it does not shun foreigners, but encourages outside involvement through a scientific advisory committee that offers Bolivia voluntary assistance. It’s comprised of national and international scientists, lithium experts, and automobile industry representatives. “I have felt very welcomed,” says Waldo Rojas, a Chilean hydraulic engineer with 20 years’ experience, who’s been helping here with pond design. But Evans says that without industry big names on the committee, outside involvement means little. Brine expert Kunasz agrees: “Bolivia made a politically
driven decision to exclude many in the industry, and it hurts them scientifically.” State administration means technical limitations, too. Uyuni’s critical meteorological conditions are tracked on a clunky Dell laptop that sits in a wooden shack next to the ponds. The project’s main laboratory is housed in a decrepit building on the slopes of La Paz. “We have had to make do with what we have,” says Jose Bustillos, the Uyuni team’s lead electrochemistry researcher. But, even so, after researchers scanned methods used at other salt flats, he says, they developed an extraction process tailored to Uyuni’s conditions that solved the magnesium problem in 2009. “It was a feeling like none other,” Bustillos says, glowing at the memory of holding Uyuni flat’s first
at Universidad Mayor de San Andrés, the country’s largest university, says Bolivia is being swept by “a scientific revolution,” noting that the number of doctoral programs went from less than a dozen in 2005 to more than 200 currently. “Our national lithium project is helping to encourage this trend,” he adds, and the young are interested. In the sparkling new “international level” laboratory accompanying the pilot plant, chemist Jon Alvarez, 30, looks over the printout of the 240 pond samples that the team of seven analyzes daily. He’s like the rest of the R&D team: average age between 30 and 35. “This is a long-term project, and we are preparing to carry it into the future,” Alvarez says, adding that he would have had no interest in working for a foreign lithium company.
By the bootstraps. Guillaume Roelants, a Belgian-born engineer, was chosen by Bolivia’s president, Evo Morales, to create and lead an autonomous national lithium-extraction industry.
grams of lithium carbonate. The team will not reveal details until international patents are secured. There have been delays. Uyuni should already be producing 12,000 tons of potash per year, a byproduct to be sold to Brazil as fertilizer. Lithium carbonate production was set to be 500 tons per year by 2010. Instead, the potash plant is still being completed, and the final evaporation pond that should be yielding lithium chloride is empty. Building for the future As the setting sun casts its rosy hue over the flat, Roelants arrives at the newly constructed pilot plant offices. Two professors from La Paz’s metallurgy institute are visiting to learn about Uyuni’s process, a regular occurrence for Roelants. Gustavo Calderón Valle, who recently stepped down as head of doctoral programs
www.sciencemag.org SCIENCE VOL 334 Published by AAAS
Alvarez is right about the time range. The global lithium carbonate market is currently saturated, and the Atacama alone could supply the world’s demand for 10 years. But if in 25 or 50 years, electric vehicles do outnumber gasoline-powered transport, the world will most likely need Bolivia’s lithium. Also, the country’s goals go beyond selling lithium carbonate. The government wants to develop the downstream chemicals to manufacture the lithium-ion batteries themselves. “This is about transforming the country from exporter of raw materials into one that industrializes its natural resources,” Roelants says. It’s another, still-far-off collective dream for the scientist, his colleagues, and an impoverished nation.
–JEAN FRIEDMAN-RUDOVSKY
Jean Friedman-Rudovsky is a print and radio journalist based in La Paz, Bolivia.
18 NOVEMBER 2011
897
Downloaded from www.sciencemag.org on November 17, 2011
NEWSFOCUS
COMMENTARY Locomotion circuits
A lull in solar activity?
912
916
LETTERS I BOOKS I POLICY FORUM I EDUCATION FORUM I PERSPECTIVES
LETTERS edited by Jennifer Sills
THE REPORT “RACE, ETHNICITY, AND NIH RESEARCH AWARDS” (D. K. GINTHER ET AL., 19 AUGUST, p. 1015), documenting a lower grant funding rate for black than for white applicants, has generated widespread alarm that there may be a hidden racial bias in the grant review process. However, the study is flawed. Ginther et al. looked at two groups—those with fewer and greater than 84 citations—and found that blacks in both groups were substantially unsuccessful with grant funding. This range does not seem relevant to competitive scientists. A meaningful cutoff for a typical assistant professor, approaching tenure and the renewal of his first grant, would be closer to 1000 citations, not 84. The study also overlooks the fact that citations accumulate over time. Ginther et al. should have used an algorithm that normalizes citations to age. I believe they would have found that scientists who are highly successful, as defined by age-normalized citations or h index, would be equally successful in grant funding, with no disparity for race and ethnicity. HAROLD P. ERICKSON Departments of Cell Biology, Biochemistry, and Biomedical Engineering, Duke University, Durham, NC 27710, USA. E-mail: harold.
[email protected]
CREDIT: DON BAYLEY/ISTOCKPHOTO.COM
Race Disparity in Grants: Empirical Solutions Vital THERE IS A PRESSING MORAL IMPERATIVE TO eliminate racial disparity such as the success gap in R01 awards for black researchers reported by D. K. Ginther et al. (“Race, ethnicity, and NIH research awards,” Reports, 19 August, p. 1015). However, it was troubling to read the reactionary and unscientific proposals offered in response to the finding (“NIH uncovers racial disparity in grant awards,” J. Kaiser, News & Analysis,19 August, p. 925). These include NIH “brainstorm” panels to increase success and even the chilling notion of having “reviewers and staff undergo tests to learn about implicit biases,” presumably with the use of questionably valid “implicit association” tests (1). The achievement gap in black R01 funding is a correlational result, and mechanistic information is required
to determine which, if any, of our intended manipulations might reduce the disparity. Unlike laboratory approaches for identifying causal pathways, disparity-reduction policies represent social experiments with tremendously important consequences, the effects of which could take decades to identify. As scientists, we should not give in to the temptation to value our knee-jerk reactions more than our empirical inclinations. We also cannot ignore morally complicated evidence. Indeed, much of the racial disparity reported could be attributed to black R01 applicants having half the citation count and one-fifth as many last-authored publications as white applicants from similarly ranked institutions. Coupled with the finding that R01s were awarded to highly ranked applications irrespective of race, this suggests that R01 disparity is due to lower research success among black applicants rather than to any problems with NIH review. One pos-
www.sciencemag.org SCIENCE VOL 334 Published by AAAS
sible explanation for black applicants being less successful yet nonetheless having faculty positions at equally competitive institutions is affirmative action hiring policies. Although these policies can have tremendous short-term benefits for individuals, long-term negative consequences can result if race is emphasized over qualification. Such longterm negative consequences have already been suggested for black competitiveness and disparity in the legal profession (2). Ginther et al. acknowledge (by discussing the idea of “cumulative advantage”) that racial disparity is a deep societal problem. Although it is comforting to think that we can undo this endemic injustice just by increasing access to R01 grants or faculty positions, current evidence suggests that these efforts at best whittle down the tip of the proverbial iceberg, and at worst exacerbate racial inequalities in research success (2).
Downloaded from www.sciencemag.org on November 17, 2011
Race Disparity in Grants: Check the Citations
JOEL L. VOSS
Beckman Institute for Advanced Science and Technology, University of Illinois Urbana-Champaign, Urbana, IL 61801, USA. E-mail:
[email protected]
References
1. H. Blanton et al., J. Appl. Psychol. 94, 583 (2009). 2. R. H. Sander, Stanford Law Rev. 57, 367 (2003).
Response
ERICKSON FEELS THAT OUR CUTOFF OF 84 citations was unrealistic and may have affected our results that NIH awarded significantly fewer R01 grants to black researchers than to white researchers. We do not believe that this was a factor. Our data included about 300 early-career individuals (defined as having obtained their Ph.D. after 1994) that had ~1000 citations. However, these are the top 1% of individuals with citations in the data. A recent evaluation of the NIH K program, targeted at early-career biomedical researchers, indicates that awardees publish about 10 publications in the 5 years following the award, each of which garners about 15 citations, yielding an average of 150 citations per person (1). Citation patterns vary widely across academic disciplines, and
18 NOVEMBER 2011
899
LETTERS National Institutes of Health, we cannot respond directly to the past and proposed NIH policies. We are in favor of the NIH approach to experiment with interventions in the NIH review process in order to determine what works and what does not before fully implementing these programs. We are confident that such an evidence-based approach will address the concerns raised by Voss regarding the need to have better evidence before putting policies into place.
DONNA K. GINTHER,1* WALTER T. SCHAFFER,2 LAUREL L. HAAK,3 RAYNARD KINGTON4
Department of Economics and Center for Science, Technology, and Economic Policy, Institute for Policy and Social Research, University of Kansas, Lawrence, KS 66045, USA. 2 National Institutes of Health, Bethesda, MD 20892, USA. 3 Discovery Logic/Thomson Reuters, Rockville, MD 20850, USA. 4President, Grinnell College, Grinnell, IA 50112, USA. 1
*To whom correspondence should be addressed. E-mail:
[email protected]
References
1. NIH Individual Mentored Career Development Awards Program Evaluation (http://grants.nih.gov/training/K_ Awards_Evaluation_FinalReport_20110901.pdf). 2. R. H. Sander, Stanford Law Rev. 57, 367 (2003). 3. D. K. Ginther et al., “Diversity in academic biomedicine: An evaluation of education and career outcomes with implications for policy” (Social Science Research Network, Rochester, NY, 2009). 4. National Research Council, Research Training in the Biomedical, Behavioral, and Clinical Research Sciences (National Academies Press, Washington, DC, 2011). 5. T. B. Hoffer, K. Grigorian, E. Hedberg, “Postdoc participation of science, engineering, and health doctorate recipients” (National Science Foundation, Washington, DC, 2008); www.nsf.gov/statistics/infbrief/nsf08307/. 6. R. B. Freeman et al., Science 294, 2293 (2001).
Race Disparity in Grants: Oversight at Home NIH DIRECTOR FRANCIS COLLINS HAS PROMised to lead the NIH in identifying the cause(s) of and redressing the recently reported disparity in NIH funding associated with race (“Weaving a richer tapestry in biomedical science,” L. A. Tabak and F. S. Collins, Policy Forum, 19 August, p. 940; “Race, ethnicity, and NIH research awards,” D. K. Ginther et al., Reports, 19 August, p. 1015). The limited public discussion on the possible underlying factors has focused on the NIH review process. Although this is an obvious place to continue the investigation, the explanation may lie elsewhere. Investigators in the National Cancer Institute have reported that a major responsible factor cited by unfunded and underfunded minority investigators was “barriers at the home institution” (1). These included “inadequate research infrastructure, training, and development,” “barriers to development as independent investigators,” “inadequate mentoring,”
www.sciencemag.org SCIENCE VOL 334
18 NOVEMBER 2011 Published by AAAS
Downloaded from www.sciencemag.org on November 17, 2011
this may account for the difference between Erickson’s experience and what we observed. We believe that age-normalizing citations would not change the results we reported for early-career investigators, but we are investigating Erickson’s suggestion fully. Voss suggests that race preferences in higher education provide a potential explanation for the differences in NIH R01 awards found in our paper. It could be that affirmative action in graduate school admissions plays a role in the funding differences that we observe. This is not a hypothesis that we can examine with our data. We do not have any information about the admissions process and cannot examine the effect of affirmative action on outcomes in the manner of Sander (2), who showed that black law students admitted under affirmative action had worse student and early-career outcomes— lower grades and lower chances of passing the bar exam—than white students. That said, there is substantial evidence that affirmative action does not explain the results of our study. First, in our companion study, we examined race and ethnicity differences in obtaining a tenure-track job and receiving tenure in academic biomedicine. We found that blacks and whites were equally likely to receive tenure at higher education institutions that are research intensive (3). Second, biomedical scientific careers require a lengthy training process— an average of 5.5 years in graduate school and 2 years of postdoctoral study (4, 5). As a result, those NIH grant applicants who are a bad match for research careers will have most likely been weeded out earlier in their careers. The average age of applicants in the sample was 48 years old. Clearly, these applicants have progressed beyond the early career stage. Third, the pay in biomedical research is much lower than other occupations requiring graduate education (6), which would suggest that the small number of black scientists who are doing research are positively selected to the field (they are willing to sacrifice a higher salary to be scientists). If there is adverse selection of black scientists into research (for example, they could not get into medical school and chose biomedical research instead), we would again expect that the lengthy training process would weed them out before they rose to the level of submitting an R01 application. On balance, we think there is a case to be made for positive selection of black scientists—that they are the best of the best— as opposed to being bad matches resulting from affirmative action. Although we have close ties to the
and “lack of institutional support.” Although the report focused on barriers to NIH funding for minority investigators pursuing primarily cancer health disparities research, many of the identified obstacles apply to biomedical research funding in general. To address these issues, Director Collins must be prepared to extend NIH policies to providing better oversight of the manner in which minority investigators are treated in their home institutions. For example, although NIH requires the writing of minority recruitment plans by its grantee institutions, it currently neither evaluates how nor even whether such plans are implemented. NIH also does not require outcome measures for whether minority scientists are fairly supported and promoted for successful research careers. Such stark omissions in NIH funding policies could certainly perpetuate or engender unfair racially biased attitudes and practices that would not have been uncovered by Ginther et al.’s study. JAMES L. SHERLEY Adult Stem Cell Technology Center, Programs in Regenerative Biology and Cancer Biology, Boston Biomedical Research Institute, Watertown, MA 02472, USA. E-mail:
[email protected]
Reference
1. V. L. Shavers et al., J. Natl. Med. Assoc. 97, 1063 (2005).
AAAS is here –
bringing educational infrastructure to the developing world. AAAS is helping the Rwandan government rebuild its educational infrastructure as a way to help drive economic growth and development. By providing materials such as the Project 2061 Atlas of Science Literacy, lesson plans from Science NetLinks, and access to Science digital libraries, AAAS is helping the people of Rwanda work toward a future built around science and technology. As a AAAS member your dues support these efforts. If you’re not yet a AAAS member, join us. Together we can make a difference.
To learn more, visit aaas.org/plusyou/rwanda
Response
SHERLEY IS CONCERNED THAT NIH DOES NOT provide sufficient oversight of the minority recruitment plans at institutions. We agree that this is an important point. NIH recognizes a unique and compelling need to promote diversity in the biomedical, behavioral, clinical, and social sciences workforce. Therefore, NIH requires recruitment plans to enhance diversity, including underrepresented minorities, for institutional training grants at the pre- and postdoctoral levels (1). The plans on all NRSA training grants are rigorously reviewed, and if they are deficient, the grants are not funded until corrective action is taken on the part of the grantee. Awarded training grants that are subsequently submitted for renewal are reviewed for the recruitment plan’s results. If the plans are judged ineffective, this assessment affects its likelihood of being funded again. FRANCIS S. COLLINS AND LAWRENCE A. TABAK*
National Institutes of Health, 1 Center Drive, Bethesda, MD 20892, USA. *To whom correspondence should be addressed. E-mail:
[email protected]
Reference
1. Part I Section 8.7 of the SF424 (R&R) Application Guide (http://grants.nih.gov/grants/funding/424/SF424_RR_ Guide_General_Adobe_VerB.pdf).
Published by AAAS
Downloaded from www.sciencemag.org on November 17, 2011
LETTERS
Science Translational Medicine Integrating Medicine and Science
The new journal from the publisher of Science stands at the forefront of the unprecedented and vital collaboration between basic scientists and clinical researchers. • Cardiovascular Disease • Neuroscience/Neurology/ Psychiatry • Infectious Diseases • Cancer • Health Policy • Bioengineering • Chemical Genomics/ Drug Discovery • Other Interdisciplinary Approaches to Medicine
Submit your research at www.submit2scitranslmed.org
CORRECTIONS AND CLARIFICATIONS News & Analysis: “Immunology prize overshadowed by untimely death of awardee,” by J. Travis (7 October, p. 31). The story incorrectly referred to a dendritic cell vaccine designed and studied by scientists at Baylor College of Medicine in Texas. In fact, the research team is at the Baylor Institute for Immunology Research, which is part of Baylor University Medical Center at Dallas, an unrelated institution. Reports: “Productivity is a poor predictor of plant species richness” by P. B. Adler et al. (23 September, p. 1750). Scott L. Collins’s affiliation was listed incorrectly. He is at the Department of Biology, MSC03-2020, University of New Mexico, Albuquerque, NM 87131, USA. He is not affiliated with The University of Queensland in Australia. News Focus: “Taking stock of the biodefense boom” by J. Kaiser (2 September, p. 1214). The article discussed $60 billion in federal funding for biodefense preparedness since 9/11, including $19 billion for biodefense research. The text and Funding Boom graph on p. 1215 should have explained that these figures include only civilian spending. Also, the $19 billion is for research on countermeasures, not all biodefense-related research.
TECHNICAL COMMENT ABSTRACTS
Comment on “Global Trends in Wind Speed and Wave Height” Frank J. Wentz and Lucrezia Ricciardulli Young et al. (Reports, 22 April 2011, p. 451) reported trends in global mean wind speed much larger than found by other investigators. Their report fails to reference these other investigations and does not discuss the consequences that such large wind trends would have on global evaporation and precipitation. The difference between their altimeter and buoy trends suggests a relatively large trend error. Full text at www.sciencemag.org/cgi/content/full/334/ 6058/905-b
Response to Comment on “Global Trends in Wind Speed and Wave Height” Ian R. Young, Alexander V. Babanin, Stefan Zieger
Chief Scientific Adviser
Elias A. Zerhouni, M.D. Former Director, National Institutes of Health
We acknowledge that the special sensor microwave/ imager studies identified by Wentz and Ricciardulli were overlooked. These studies report wind speed trends 1.4 to 2.4 times smaller than our altimeter data. However, the reported altimeter wind speed trends are consistent with limited buoy data and exhibit scatter consistent with the calculated error statistics. Full text at www.sciencemag.org/cgi/content/full/334/ 6058/905-c
Letters to the Editor Letters (~300 words) discuss material published in Science in the past 3 months or matters of general interest. Letters are not acknowledged upon receipt. Whether published in full or in part, Letters are subject to editing for clarity and space. Letters submitted, published, or posted elsewhere, in print or online, will be disqualified. To submit a Letter, go to www.submit2science.org.
ScienceTranslationalMedicine.org Published by AAAS
Downloaded from www.sciencemag.org on November 17, 2011
Call for Papers
LETTERS
Comment on “Global Trends in Wind Speed and Wave Height” Frank J. Wentz* and Lucrezia Ricciardulli Young et al. (Reports, 22 April 2011, p. 451) reported trends in global mean wind speed much larger than found by other investigators. Their report fails to reference these other investigations and does not discuss the consequences that such large wind trends would have on global evaporation and precipitation. The difference between their altimeter and buoy trends suggests a relatively large trend error. urface wind is an important driver of oceanic and atmospheric circulation. Even small wind trends can have a large impact on atmospheric and ocean dynamics, on air-sea fluxes, and on the hydrological cycle (1–3). An excellent review of global wind trends is given by Tokinaga and Xie (4). Young et al. (5) appear to be unaware of these other investigations. Their statement that “the radar altimeter provides by far the longest-duration record” of wind speed suggests that they are also unaware of the multitude of satellite microwave (MW) radiometers and scatterometers that have been launched since 1987, all of which provide highly accurate ocean measurements of wind speed. For example, one series of MW radiometers, the SSM/Is (special sensor microwave/ imagers), has been in continuous operation without interruption for 25 years, which is longer than the continuous operation of altimeters. Furthermore, the spatial sampling of the MW radiometers and scatterometers is greatly superior to the altimeters because they have swath widths of 1000 to 1400 km, as compared with the altimeter swath width of about 5 km. The paper has no references to any of the wind results coming from these other satellite sensors. More important, the reported wind trends in the paper are 2.5 to 5 times higher than those reported by other investigators (2, 4). Using SSM/I winds, Wentz et al. (2) report a 1987 to 2006 global trend over the oceans of 0.08 m s−1 decade−1 (1.0% decade−1) and also provide a wind trend map for this period. By adjusting shipbased anemometer readings to agree with wave observations, Tokinaga and Xie (4) estimated the 1988 to 2008 global wind trend to be 0.084 m s−1 decade−1 (1.1% decade−1). Their paper reviewed wind trends reported by other investigators and from several reanalysis data sets. Excluding an uncorrected ship-based data set, which was known to be spurious, the global wind trend estimates that exceed the 99% confidence level ranged
S
from 0.9% to 1.8% decade−1. Their estimate of the SSM/I wind trend in the ship-sampled regions for 1988 to 2008 is 0.134 m s−1 decade−1 (1.7% decade−1), which is 0.05 m s−1 decade−1 higher than the ship value. This difference is within the error bar reported by (2). They also found a very high correlation of their monthly winds with the SSM/I retrievals, exceeding 0.98 in most areas. In contrast to these previous investigations, Young et al. (5) report a global trend over the oceans of 2.5% to 5% per decade for the 1991 to 2008 period. Perhaps the strongest argument against such high wind trends is the effect this would have on global evaporation. About 86% of the world’s evaporation comes from the oceans, and evaporation is directly proportional to wind speed. Evaporation also depends on the surface relative humidity (RH) (6), but current climate modeling predicts a relatively stable RH, and no substantial RH trend has been observed (7). If RH remains constant, a 2.5% (5%) decade−1 increase in winds
would result in a 5% (10%) increase in evaporation and precipitation over 20 years. This is an enormous increase and would certainly have been observed by precipitation-measuring satellites. It has not been (2, 8). To further put this into perspective, both climate models and satellite observations agree that the total water in the atmosphere increases at the Clausius-Clapeyron (CC) rate of 6.5%/K as the climate warms (2). However, most climate models predict a muted response of evaporation and precipitation to warming, on the order of 2.5%/K. This muted response is due to radiative constraints on the surface energy budget (3, 9, 10). Changes in forcing mechanisms like cloud cover and type or aerosols can increase this muted response, but the CC rate of 6.5%/K is a reasonable upper bound. During the past two decades, the surface and lower troposphere warmed at a rate near 0.2K decade−1 (2). Hence, the expected increase in evaporation is between 1% (radiative cooling constraint) and 2.6% (upper CC limit). The increase of 5% to 10% in evaporation implied by the large wind trends (5) is far above even the upper limit. Most climate models do not predict a significant increase of global surface winds in response to anthropogenic forcing. Rather, a robust prediction of the models seems to be a weakening of the tropical circulation (11). The observational data also suggest a weakening of the trade winds over the past decades (12–14). Young et al. (5) wind trends are positive almost everywhere. Young et al. (5) give no error estimates on their results, but an error can be inferred from their table 1, which compares altimeter-derived trends with buoy trends. Our Fig. 1 gives the root mean square (RMS) difference and correlation of
Fig. 1. Plot of altimeter versus buoy wind trends (m s−1 decade−1) for the 12 selected buoys, as given in table 1 of Young et al. (5). The figure includes the correlation coefficient and the RMS difference for buoy minus altimeter wind trends. For a mean global wind of 7.5 m/s, the RMS corresponds to 3.1% decade−1.
Downloaded from www.sciencemag.org on November 17, 2011
TECHNICAL COMMENT
Remote Sensing Systems, 444 10th Street, Santa Rosa, CA 95401, USA. *To whom correspondence should be addressed. E-mail:
[email protected]
www.sciencemag.org
SCIENCE
VOL 334
18 NOVEMBER 2011
905-b
TECHNICAL COMMENT
905-b
difficulty in constructing precision time series from multiple satellites, we find it hard to place much credence on the claim that high winds have increased by 15% over the past 20 years. A recent study (15) using wide-swath satellite radiometer wind retrievals show the opposite effect: a decrease in the frequency of tropical high-wind events. In summary, we question the validity of global wind trends at the 2.5 to 5% decade−1 level. Such a large increase disagrees with wind retrievals from wide-swath satellite radiometers and scatterometers over the past 25 years. Furthermore, the associated increase in global evaporation and precipitation that would have occurred is unreasonably high. Finally, the difference of 3.1% decade−1 between the altimeter wind trends and those from the collocated buoys is probably indicative of the inherent error in the wind trends reported in this paper.
18 NOVEMBER 2011
VOL 334
SCIENCE
References 1. L. Yu, J. Clim. 20, 5376 (2007). 2. F. J. Wentz, L. Ricciardulli, K. Hilburn, C. Mears, Science 317, 233 (2007). 3. I. M. Held, B. J. Soden, J. Clim. 19, 5686 (2006). 4. H. Tokinaga, S.-P. Xie, J. Clim. 24, 267 (2011). 5. I. R. Young, S. Zieger, A. V. Babanin, Science 332, 451 (2011). 6. T. Schneider, P. A. O'Gorman, X. Levine, Rev. Geophys. 48, RG3001 (2010). 7. A. Dai, J. Clim. 19, 3589 (2006). 8. R. F. Adler et al., J. Geophys. Res. 113 (D22), D22104 (2008). 9. M. R. Allen, W. J. Ingram, Nature 419, 224 (2002). 10. G. L. Stephens, T. D. Ellis, J. Clim. 21, 6141 (2008). 11. G. A. Vecchi, B. J. Soden, J. Clim. 20, 4316 (2007). 12. S. K. Gulev, V. Grigorieva, Geophys. Res. Lett. 31, L24302 (2004). 13. G. A. Vecchi et al., Nature 441, 73 (2006). 14. H. Tokinaga, S.-P. Xie, Nat. Geosci. 4, 222 (2011). 15. G. Gastineau, B. Soden, Geophys. Res. Lett. 38, L09706 (2011). 24 June 2011; accepted 19 October 2011 10.1126/science.1210317
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
the mean altimeter and buoy trends. The RMS difference is 3.1% decade−1, which is of the same order as the estimated mean trend. Young et al. (5) present a trend map of the NCEP/NCAR (National Centers for Environmental Prediction/National Center for Atmospheric Research) reanalysis winds and say that it is “qualitatively consistent” with the altimeter but do not provide any statistics. When we look at the two maps [Figure 1 and fig. S6 in (5)], we see very different patterns, such as large basinwide areas in the Atlantic and Indian Ocean where NCEP/NCAR is showing negative trends and the altimeter is showing positive trends. Thus far, our comments have been related to the mean trends found by Young et al. (5), not their 90th and 99th percentile results. Considering the very small sample size, the altimeter versus buoy error for the mean results, the decrease in radar sensitivity at high winds, and the inherent
Response to Comment on “Global Trends in Wind Speed and Wave Height” Ian R. Young,* Alexander V. Babanin, Stefan Zieger We acknowledge that the special sensor microwave/imager (SSM/I) studies identified by Wentz and Ricciardulli were overlooked. These studies report wind speed trends 1.4 to 2.4 times smaller than our altimeter data. However, the reported altimeter wind speed trends are consistent with limited buoy data and exhibit scatter consistent with the calculated error statistics. e would like to thank Wentz and Ricciardulli (1) for bringing to our attention the Wentz et al. (2) and Tokinaga and Xie (3) results, which add a valuable additional data set to the discussion of wind speed trends. Let us first consider the relative magnitudes of the trends reported. Because differences in spatial variation make such comparisons difficult, the global average trend in mean wind speed has been used as the reference. Wentz et al. (2), in a paper that concentrates on evaporation and precipitation, report a global mean trend of 0.08 ms−1 decade−1 (1987 to 2006) using special sensor microwave/imager (SSM/I) data. Tokinaga and Xie (3)—in a paper that appeared when our paper, Young et al. (4), was in press—report 0.134 ms−1 decade−1 (1988 to 2008) for SSM/I data. When calculated for the altimeter data in (4), the global mean trend is 0.192 ms−1 decade−1 (1991 to 2008). The National Centers for Environmental Prediction (NCEP) model data reported in (4) yields 0.108 ms−1 decade−1 (1991 to 2008), or 0.150 ms−1 decade−1 if the Indian Ocean is excluded. In summary, the altimeter data set in (4) produces trends 1.4 to 2.4 times as large as the reported SSM/I data, rather than the 2.5 to 5 times stated by Wentz and Ricciardulli. Figure 14 of Tokinaga and Xie (3) does suggest that the trend may be stronger in recent years, a result supported by the present altimeter data. This may account for some of these differences, noting that (4) uses data from more recent years. In conducting such long-term trend assessments, an essential requirement is that there is a consistently validated and calibrated data set over the period. Changes in satellite orbit, instrumentation, and instrumental drift can easily result in influences as large as any trends that may exist. In the case of the altimeter database, a consistent multiplatform validation and calibration exists over the full period of the study (5). For these reasons,
W
we believe that the database of Zieger et al. (5) provides a high-quality database for such studies. The SSM/I data set reported by Wentz et al. (2) has also been the subject of calibration and validation studies (6). Numerous studies of radar altimeter winds (5) indicate that the root mean square (RMS) error for such measurements is between 0.8 m/s and 1.5 m/s, depending on the retrieval algorithm used. Wentz (6) reports an SSM/I RMS error of 0.9 m/s. Hence, these calibrations indicate comparable performance from the instruments. We do not believe that one can infer that one instrument is superior, based on these comparisons with in situ data. Both instruments have limitations in the recovery of the wind speed. The SSM/I retrieval process yields the wind stress rather than the wind speed. Numerous studies have considered the variability of the drag coefficient Cd as a function of wind speed (7–11). As reported by Babanin and Makin (11), Cd can vary by an order of magnitude for a given value of wind speed. Similarly, the altimeter transfer function is a nonlinear function of
radar cross section and, hence, as indicated by (4), the measurement of wind speed is less accurate than wave height. Wentz and Ricciardulli (1) state, “Young et al. give no error estimates on their results.” This statement overlooks the extensive analysis of statistical variability in the supporting online material for (4). Notably, this analysis recognizes that the statistical variability of the trend estimate depends on both the accuracy of the measured data and the chosen trend extraction method. Because (1) and (4) use different trend extraction methods, this may also account for some of the differences. Wentz and Ricciardulli plot the data from table 1 of (4) and state that the comparison between altimeter- and buoy-derived values of trend indicates large errors in the altimeter estimates. This figure is reproduced in Fig. 1. The 1:1 correlation line is shown along with the 95% confidence limits reported in (4) (i.e., T0.264 ms−1decade−1). It should be noted that because the RMS error for SSM/I data is similar to that of the altimeter data, the 95% confidence limits on SSM/I estimates will be comparable in magnitude. Figure 1 shows that the data scatter is consistent with the confidence limits, and the data scatters about the 1:1 line with 9 of 12 points within the 95% confidence limits. Because the data set is small and the scatter significant, we do not claim that this validates the altimeter data, only that the results are comparable. If the altimeter values in this figure were reduced by a factor of 2, to be consistent with the SSM/I results, all the data points would fall below the 1:1 line. That is, the SSM/I data, as reported, would be significantly smaller than the buoy data. Hence, we conclude that the altimeter results are consistent with the in situ buoy data. Fig. 1. Scatter plot of trends in mean wind speed from table 1 of Young et al. (4). Dashed lines show 95% confidence limits (T0.264 ms−1 decade−1) as reported in (4).
Downloaded from www.sciencemag.org on November 17, 2011
TECHNICAL COMMENT
Swinburne University of Technology, Melbourne, Victoria, 3122, Australia. *To whom correspondence should be addressed. E-mail:
[email protected]
www.sciencemag.org
SCIENCE
VOL 334
18 NOVEMBER 2011
905-c
TECHNICAL COMMENT
905-c
statement ignores the very detailed analysis undertaken by us in (4), considering the validity of high–wind speed altimeter measurements, the impact of sampling size, and variation over time. This was a comprehensive analysis, which should not be discarded without basis. In summary, the results of Young et al. (4) indicate trends 1.4 to 2.4 times as large as SSM/I data, rather than 2.5 to 5 times as stated by Wentz and Ricciardulli. These trends are consistent with the buoy data reported in (4), with the scatter also consistent with the error statistics in (4). The variability in these results, and indeed between the SSM/I data as reported by different authors, indicates that further study is required. The combined use of altimeter, SSM/I, and scatterometer (as longer data sets become available) will provide an invaluable data set.
18 NOVEMBER 2011
VOL 334
SCIENCE
References 1. F. J. Wentz, L. Ricciardulli, Science 334, 905 (2011); www.sciencemag.org/cgi/content/full/334/6058/905-b. 2. F. J. Wentz, L. Ricciardulli, K. Hilburn, C. Mears, Science 317, 233 (2007). 3. H. Tokinaga, S.-P. Xie, J. Clim. 24, 267 (2011). 4. I. R. Young, S. Zieger, A. V. Babanin, Science 332, 451 (2011). 5. S. Zieger, J. Vinoth, I. R. Young, J. Atmos. Ocean. Technol. 26, 2549 (2009). 6. F. J. Wentz, J. Geophys. Res. 102 (C4), 8703 (1997). 7. T. E. Nordeng, J. Geophys. Res. 96 (C4), 7167 (1991). 8. S. D. Smith et al., Boundary-Layer Meteorol. 60, 109 (1992). 9. I. R. Young, Wind Generated Ocean Waves (Elsevier, Amsterdam, 1999). 10. V. K. Makin, V. N. Kudryavtsev, C. Mastenbroek, Boundary-Layer Meteorol. 73, 159 (1995). 11. A. V. Babanin, V. K. Makin, J. Geophys. Res. 113 (C2), C02015 (2008). 19 July 2011; accepted 19 October 2011 10.1126/science.1210548
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
Wentz and Ricciardulli indicate that the reported trends are inconsistent with our knowledge of evaporation. Because the difference between the trends in (4) and the SSM/I results is smaller than assumed in (1), the strength of this criticism obviously is reduced. However, we should point out that evaporation is a complex physical process depending on other parameters—including sea surface temperature, ocean mixing, and humidity—in addition to wind speed. Therefore, we do not believe it prudent to discard the altimeter-derived trends, based on the inferred impact on evaporation. In terms of 90th and 99th percentile results, Wentz and Ricciardulli state that based on the challenges of sampling extreme events accurately, they “find it hard to place much credence on the claim that high winds have increased...” This
ECONOMICS
To Close the Gap Charles I. Jones
I
f the time since modern humans first either facilitate or hinder growth in coming appeared, some 200 thousand years ago, decades: globalization, financial investment, were magically compressed into a sin- the Washington consensus, climate change, gle day, the era of sustained growth in liv- urbanization, politics, exchange rates volaing standards would occupy only the most tility, information technology, and the finanrecent minute and a half. Prior to this emer- cial crisis. (The only notable absence is a disgence (around 1800), living standards were cussion of problems related to government low and did not vary enormously throughout debt, on display in Europe today and associthe world—perhaps by a factor of 2 accord- ated with rising entitlement spending in the ing to numbers compiled by the late econo- United States in the future.) Economics does mist Angus Maddison (1). The emergence not always offer precise, unequivocal preof modern economic growth in some parts scriptions regarding each of these topics, but of the world but not others has caused living having Spence as a brilliant guide, intimate standards to diverge dramatically. By the end with both the research and policy frontiers, is of the 20th century, the ratio a perfect way to get up to speed. of per capita gross domestic It is impossible to summarize The Next Convergence product (GDP) in the richsuch a wide-ranging book in a brief The Future of Economic est countries exceeded that review. Instead, let me discuss a Growth in a Multispeed in the poorest by more than few of the highlights. First, Spence World a factor of 50. emphasizes that economic growth by Michael Spence Michael Spence takes is tied fundamentally to innovaFarrar, Straus and Giroux, this history as his point tion and knowledge. Openness to New York, 2011. 312 pp. of departure in The Next international trade and globaliza$27, £17.99. Convergence: The Future tion is one of the strongest correISBN 9780374159757. of Economic Growth in a lates of economic growth in the Multispeed World. Looking data because it facilitates access to toward the future, he foresees a convergence knowledge. Imagine the standard of living in between the economies of the world’s rich any single country if it had never encountered and poor countries. That is, he believes that ideas invented elsewhere in the world. the current dynamics of economic growth, Second, Spence provides a careful, accompanied by the right policies and institu- nuanced discussion of the role of government tions, can lead to a new period of substantial in economic growth. Quoting the famous catch-up by the poor countries. development economist Arthur Lewis, “govSpence, a highly regarded economist ernments may fail either because they do at New York University’s business school, too little, or because they do too much” (3). shared the 2001 Nobel Prize in economics Some kind of economic freedom is central to for work involving the microeconomics of growth. Spence notes that Chinese economic imperfect information—how markets func- growth was spurred when Deng Xiaoping and tion when some participants know more than other reformers allowed the market mechaothers. There is a certain distance between the nism to work in agriculture, permitting farmauthor’s prize-winning research in microthe- ers to sell the excess of their production over ory and the applied macroeconomic topics of their planned quotas on open markets. Howconcern in the book. That distance is closed ever, democracy is evidently not a necesby a combination of Spence’s economic train- sary ingredient for economic success, as the ing, his decade as the dean of Stanford Uni- example of China also indicates. Instead, what versity’s business school, and his most recent appears to be essential is a government that service as the chair of the independent Com- creates an environment in which investment mission on Growth and Development (2). by the private sector is profitable. The book provides a fascinating tour Lastly, Spence highlights a well-known of the global economic issues that may association between urbanization and economic growth. He describes China as needing to build a new Los Angeles every year to The reviewer is at the Graduate School of Business, Stanford accommodate its annual flow of 15 million University, 655 Knight Way, Stanford, CA 94305–4800, USA. E-mail:
[email protected] people from the countryside. Yet despite three
906
decades of such movements, more than 50 percent of China’s population remains rural. The number in India is even higher, around 70 percent, and Spence suggests that India will likely experience massive urbanization in the next two decades. Echoing a point that Paul Romer has explored creatively in recent years, urban infrastructure appears to be a crucial ingredient to sustained growth (4). In concluding, Spence acknowledges the uphill battle that economic progress faces. “Getting there from where we are now will be difficult. Political, business, and academic elites have lost credibility with the populace in many countries. We have been wrong about important characteristics of the economy that affect people’s lives, and relatively insensitive to distributional issues.” But engaging, intellectual treatments like this one will surely make progress easier. References
1. A. Maddison, “Statistics on world population, GDP, and per capita GDP, 1–2008 AD” (March 2010); available at www.ggdc.net/MADDISON/oriindex.htm. 2. www.growthcommission.org. 3. W. A. Lewis, The Theory of Economic Growth (Allen and Unwin, London, 1955). 4. P. Romer, “For richer, for poorer.” Prospect (London), 167 (January 2010); www.prospectmagazine. co.uk/2010/01/for-richer-for-poorer/. 10.1126/science.1214409
ENGINEERING
Low Costs and Considerable Gains Esther M. Sternberg
F
ormerly subsistence economies are developing rapidly while highly developed nations are experiencing widespread unemployment and poverty. Design with the Other 90%: Cities provides hope that both extremes can come together to help lift the underprivileged out of poverty through creative design. Organized by the Smithsonian Institution’s Cooper-Hewitt National Design Museum under the direction of Cynthia Smith (the museum’s curator of socially responsible design), the exhibition is at the United Nations headquarters— a most appropriate venue for spotlighting design solutions that have been instituted by more than 60 projects in 23 countries around the globe. As one of the designers notes, humans need food, water, and shelter to live. The show points the way to cost-effective The reviewer is the author of Healing Spaces: The Science of Place and Well-Being. Web site: www.esthersternberg.com
18 NOVEMBER 2011 VOL 334 SCIENCE www.sciencemag.org Published by AAAS
Downloaded from www.sciencemag.org on November 17, 2011
BOOKS ET AL.
and clever solutions to provide these necessities and more to improve the quality of life of those who live in informal settlements in emerging and developing economies. Sustainability and ease of use are central to many of the designs. Although perhaps not their primary goal, several designs also improve human health, interrupting the downward-spiraling cycle that can occur when illnesses sap energy and motivation. Most of the projects emphasize sustainability—using cheap and recycled materials for new and useful applications. However, they do not ignore our increasingly digitized world in which even slum dwellers are connected to the Internet and have mobile phones. To power such devices where electrical outlets are sparse or don’t exist, some designers tapped human feet (a bicycle phone charger in Arusha, Tanzania), whereas others turned to the Sun (the digital drum in Kampala, Uganda). As the exhibition’s title suggests, the designers worked closely with the people for whom they were designing to identify problems that were impeding progress. Urban Mining (23 students from MAS Urban Design, ETH Zurich, led by Rainer Hehl) determined that one issue preventing people from constructing their own homes in Heliópolis, São Paolo, Brazil, was the weight of concrete. That forced people to use sophisticated and expensive equipment to construct a home. In response, the designers mixed cement with Styrofoam or exploded clay and expanded glass to create a material that is 40% lighter than standard concrete. This allows people to build their own homes on site using light-weight, precast molds. A potential unexpected health benefit of this solution might be a reduction in construction-related injuries, including arthritis, back pain, and tendonitis. Other
Colorful coating. In Praça Cantão’s Favela Painting project, Dutch artists Jeroen Koolhaas and Dre Urhahn teamed with youth from the Santa Marta community and the Coral Paint Company.
projects turned to simple, low-tech means Besides increasing the risks of water-borne to aid in transporting people or goods. For illnesses, the lack of running water forces the example, Design With Africa’s Bicycle Mod- women of the campamento to haul gallons of ules recycled bicycle parts water per day to do launto create carts to transport dry, cook, and shower. Design with the Other 90%: Cities patients to hospitals or As a result, many villagCynthia E. Smith, curator goods to markets. ers experience debiliCooper-Hewitt National Design Museum, In several projects, tating joint, back, and Smithsonian Institution. At the United Nations Headquarters, the health benefits were neck pains. Using foot New York. Through 9 January 2012. explicitly considered pumps and recycled plashttp://designother90.org/cities/home during the design stage. tic materials, the project Some of these were put developed a gravity-fed Design with the Other 90%: Cities together as public health water system (with storefforts to improve toilet age tanks and indoor and by Cynthia E. Smith et al. facilities, purify water, or outdoor faucets) and a Cooper-Hewitt National Design more effectively deliver more efficient kitchen Museum, New York, 2011. 234 pp. Paper, $29.95. ISBN 9780910503839. needed medications. Othsink. As in Caracas’s ers included improved stairs project, these soluhealth among multiple goals. For example, the tions not only provide clean and accessible Integral Urban Project found that the steep and water but also reduce the aches related to muddy slopes and the lack of running water carrying water and provide important social contributed to the unhealthy living conditions spaces where villagers can congregate. in a barrio of Caracas, Venezuela. Working Some of the design solutions featured in with the barrio’s inhabitants, the designers the exhibition specifically address emotional came up with a low-tech solution, stairs. The well-being, which is as essential to health as project built concrete staircases that incorpo- one’s physical condition. The Favela Paintrate public services, including gravity-drained ing Project of Praça Cantão in Rio de Janeiro waste-disposal pipes running downhill and used low-tech, low-cost paint to improve pipes for conveying pump-driven potable the visual appeal of the Santa Marta comuwater uphill. It also strategically placed social nidade. Engaging the community in painting and green spaces on stairway landings, pro- their dwellings, the project coated the homes viding places for people to congregate and with brightly colored geometric designs. In a thus addressing another feature important to similar project, artist and activist JR covered people’s health, social support. homes on the slopes of Kibera (an informal Problems stemming from the absence settlement in Nairobi) with enormous phoof running water were also the focus of the tos of villagers’ faces. In a particularly creSafe Agua project by a group of designers ative twist, the houses at the top of the hill, (including my daughter) and students from just below a railroad line, showed faces only the Designmatters Program at the Art Center below the eyes. The designer then mantled College of Design (Pasadena, California) freight-train cars with the faces’ upper porworking with the Chilean nongovernmental tions. This produced a larger than life-sized organization Un Techo para Mi País and inhab- game: villagers looking up the slope can see itants of Santiago’s Campamento San José. a series of amusing faces created when trains complete the pictures as they pass by. The sometimes playful and always creative design solutions showcased in Design with the Other 90%: Cities are addressing problems that help keep people stuck in poverty. The collaborative nature of the projects, involving design teams and the communities for which they are designing, provides hope for a world in which rich and poor nations join in creative solutions to improve urban environments. For the first time in history, the majority of Earth’s population lives in cities, and projections suggest that by 2030 two billion people will be living in informal settlements. In such a world, collaborations and the solutions they can produce are not luxuries—they are imperatives.
Downloaded from www.sciencemag.org on November 17, 2011
CREDIT: PHOTO © HAAS&HAHN/COURTESY COOPER-HEWITT NATIONAL DESIGN MUSEUM
BOOKS ET AL.
10.1126/science.1215740
www.sciencemag.org SCIENCE VOL 334 18 NOVEMBER 2011 Published by AAAS
907
POLICYFORUM ENVIRONMENT AND DEVELOPMENT
Preparing to Manage Climate Change Financing
Lessons from the failures and successes of international development should guide investment in developing-world responses to climate change.
A
t the 2010 Cancun Conference of the Parties to the United Nations Framework Convention on Climate Change (UNFCCC), the international community agreed in principle to one of the largest development programs in history. The developed nations pledged to mobilize U.S.$100 billion per year by the year 2020 to “address the needs of developing countries” in responding to climate change (1). The funds, which may apply to adaptation and mitigation, are proposed to flow through multiple channels, including existing development banks, official development assistance, bilateral programs, international private investment flows (e.g., carbon markets), and other public and private mechanisms. Recommendations provided by a transitional committee for the management and operation of the proposed climate change financing will be considered by the parties to the UNFCCC at the upcoming conference in Durban, South Africa (2). At the center of this climate finance system will be a new international Green Climate Fund (GCF), which is in charge of the initial U.S.$30 billion in fast-track financing, to be raised by 2012, and a “significant” yet undetermined fraction of the eventual U.S.$100 billion per year (1). In designing the GCF, the UNFCCC must heed the lessons of past international development failures and successes to build the capacity of the international institutions and recipient countries to mobilize and manage climate funds and to set a good precedent for other institutions involved in climate change financing. Rapid Expansion, Radical Alteration
If realized, this new annual financing could radically alter international lending. It is equivalent to roughly 80% of all existing annual official development assistance from major donors, more than double the standard annual lending by the World Bank, and on the same scale as the gap in 2010 annual financDepartment of Geography, University of British Columbia, Vancouver, BC V6T 1Z2, Canada. 2Liu Institute for Global Issues, University of British Columbia, Vancouver, BC V6T 1Z2, Canada. 3Institute for Resources, Environment and Sustainability, University of British Columbia, Vancouver, BC V6T 1Z2, Canada. 1
*Author for correspondence:
[email protected]
908
ing deemed necessary to meet the UN Millennium Development Goals (see the graph). The annual GCF financing is roughly equivalent to the cost of the entire 4-year Marshall Plan, the U.S. program to rebuild the European economy after World War II, which is often invoked as a model for addressing climate change. This financial support is critical. Developing nations are expected to face an enormous human and economic toll from climate change, despite being the least responsible for greenhouse gas emissions (3). Additional financial support is needed to help countries shift to less carbon-intensive development paths and adapt to future impacts of climate change. By some estimates, the cost of adaptation alone in the developing world could be upwards of U.S.$100 billion per year (4). However, the climate change financing pledge made in Cancun also represents a large expansion of an international aid system already fraught with problems. The willingness of donor nations to meet the longterm funding commitment, given political and economic constraints, may depend on successful administration of the initial fasttrack financing by the GCF and faith in the long-term management plan. Securing New and Additional Funding
Upon the insistence of developing countries, the Cancun agreement specifically states that the proposed climate change financing must be “new” and “additional” to existing development aid. There is a long history of shifting aid money or relabeling aid projects in response to new aid priorities. For example, a shift in U.S. foreign aid to countries deemed vital to security caused a relative decline in aid to the neediest countries (5). The “new” and “additional” terminology is included in response to concerns that funding will come at the expense of other aid priorities and the label “climate change” will be applied to existing aid proposals. Careful consideration of what qualifies as new and additional climate financing is critical to ensuring that donors meet the actual policy goal of assisting the developing world in mitigation and adaptation, rather than the narrower political goal of demonstrating par-
ticipation in climate financing. The specific definition of “new” and “additional” climate change financing is as yet unresolved by UNFCCC negotiations (2). Of the 20 nations that have pledged contributions to the U.S.$30 billion fast-track climate change financing intended for the 2010–2012 period, only four are thought to be providing entirely “new” or “additional” investments (6). The inefficiencies of the Clean Development Mechanism (CDM) under the Kyoto Protocol serve as a cautionary tale of loopholes related to “additionality” of actions in international climate agreements. The creation of an international system for purchasing emission credits may have inspired companies in India and China to produce the coolant HFC-23, a powerful greenhouse gas, and then sell “credits” for destroying the HFC23 before it was emitted to the atmosphere (7). These actions thus had no net additional impact on greenhouse gas emissions (7). A solution to the additionality challenge lies in the regulatory framework under development to ensure that GCF funds are directed efficiently to climate change activities. To be effective, this framework must be adaptive: The rules and definitions need to be subject to regular assessments, as is the case with the Montreal Protocol on Substances that Deplete the Ozone Layer (8). The CDM experience demonstrates that failure to regularly review and adapt the regulatory framework can lead to projects that are not additional and do not serve the policy goals. Regular, ideally annual, assessments will help close loopholes that permit project relabeling or climate funding coming at the expense of other development aid. Avoiding Waste and Misappropriation
Introducing a large amount of aid into a program or country in a short period of time often leads to waste and misappropriation by both donors and recipients. For example, the unprecedented scale of donations to international relief agencies after the 2004 Indian Ocean tsunami overwhelmed local capacity and led to those agencies wasting funds on publicity stunts aimed at showing the international audience that the agencies were “taking action” (9). In recipient countries with poor
18 NOVEMBER 2011 VOL 334 SCIENCE www.sciencemag.org Published by AAAS
Downloaded from www.sciencemag.org on November 17, 2011
Simon D. Donner,1* Milind Kandlikar,2,3 Hisham Zerriffi2
POLICYFORUM
Ensuring Results on the Ground
The history of development aid suggests that preconceived notions of the methods of aid delivery that are most effective heavily influence aid programs (15). A number of economists have become skeptical about the ability of “top-down,” large-scale aid efforts to
Conclusion
The $100 billion per year pledge was a major victory in the UNFCCC negotiations. Now it is up to those tasked with designing the Green Climate Fund, and the other actors in this emerging climate change finance regime,
16. A. V. Banerjee, E. Duflo, Poor Economics (Public Affairs, New York, 2011). 17. J. Cohen, P. Dupas, Q. J. Econ. 125, 1 (2010). 18. B. K. Sovacool, Clim. Policy 11, 1177 (2011).
Supporting Online Material
www.sciencemag.org/cgi/content/full/334/6058/908/DC1
www.sciencemag.org SCIENCE VOL 334 18 NOVEMBER 2011 Published by AAAS
Downloaded from www.sciencemag.org on November 17, 2011
Annual expenditures and $125 governance, aid can even estimated needs for variset a country back by coous major programs and $105 opting the best public ser$100 international aid. The year(s) vants to fulfill donor needs of the programs and aid are rather than meet the other shown in parentheses. The demands of their jobs (10). values for actual expenditures (gray shading) and estimated Furthermore, a substantial needs (dark shading) were all fraction of aid money is $39 Billions of dollars per year (2010) $37 adjusted to 2010 U.S. dollars recycled back to the donor $27 using the U.S. Bureau of Labor nation through rules and Statistics Consumer Price norms surrounding hiring Index inflation calculator. See $3 $1 of consultants and procurethe supporting online material Marshall Global Bill and World Foreign aid Millennium Investment UNFCCC ment of equipment from for details. Plan Environment Melinda Gates Bank budget of Development needs for commitment the donor nation. This “tied (1948– Facility Foundation lending developed Goals universal aid” represents over 60% of 1951) lending lending (2008) economies financing modern (2009) (2009) (2008) gap (2003) energy (2009) all aid provided by countries that are members of the Organisation for Economic Co-operation and make a difference to the lives of the poor to develop mechanisms to use the money Development (OECD) (11). (16). Mechanisms must be created such that wisely. Effective administration in the near One tool for minimizing waste, misap- the GCF, a body conceived under a UN con- term will help build the public and political propriation, and tied aid is for the GCF to vention by government representatives, is not confidence in the international aid system use a third-party auditor, similar to the U.S. biased toward top-down projects. necessary to ensure that climate change fundGovernment Accountability Office (GAO) A scientific approach to decision-making, ing is sustained through political cycles in or Canada’s Auditor General, to evaluate all based on specific and measurable outcomes donor countries. spending programs. Estimates suggest that of individual projects, can help increase aid References and Notes every dollar spent by the GAO on monitoring effectiveness and avoid a bias toward pre1. UNFCCC, Outcome of the Work of the Ad Hoc WorkU.S. government spending results in savings conceived notions of aid delivery. In some ing Group on Long-Term Cooperative Action Under the of $114 (12). There is no specific mention of cases, rigorous, randomized control trials Convention (2011); http://unfccc.int/meetings/cop_16/ items/5571.php. a third-party auditor in UNFCCC materials, can test specific hypotheses about aid initia2. UNFCCC, Report of the Transitional Committee for the although the transitional committee for the tives and policies (16). For example, recent Design of the Green Climate Fund (2011); http://unfccc.int/ design of the GCF has proposed using an experiments in Kenya found that providing cancun_agreements/green_climate_fund/items/6038.php. “independent evaluation unit” (2). free malaria bednets could save more lives 3. A. Agarwal, S. Narain, Global Warming in an Unequal World: A Case of Environmental Colonialism (Centre for To ensure transparency and avoid con- than cost-sharing programs, thus overturning Science and Environment, New Delhi, 1991). flict of interest, the agency that serves as the previously held assumptions about how to 4. M. Parry, Assessing the Costs of Adaptation to Climate GCF trustee cannot also be the auditor and finance such basic interventions (17). Change (International Institute for Environment and Development, London, 2009). evaluation unit. Internal evaluation departA portion of the GCF should be used to 5. R. K. Fleck, C. Kilby, J. Dev. Econ. 91, 185 (2010). ments, common with aid agencies and devel- create a research and scientific analysis pro6. World Resources Institute, Summary of Developed Counopment banks, rarely find failure, even in the gram like Innovations for Poverty Action or try ‘Fast-Start’ Climate Finance Pledges (2011); www. face of strong evidence (13). If the World the Poverty Action Lab, both international wri.org/publication/summary-of-developed-country-faststart-climate-finance-pledges. Bank, the trustee of the fast-track financing, nonprofit organizations that use randomized 7. M. Wara, Nature 445, 595 (2007). becomes the permanent GCF trustee, then evaluations to test which programs are most 8. D. Kaniaru, The Montreal Protocol (Cameron May, Lonauditing should be conducted by third par- effective. Proactive research and analysis can don, 2007). 9. J. Stirrat, Anthropol. Today 22, 11 (2006). ties rather than by the World Bank’s existing provide data for the GCF Board to consider in J. Svensson, in Reinventing Foreign Aid, W. Easterly, Ed. internal auditor, the Independent Evaluation funding decisions. This approach is one way 10. (MIT Press, 2008), pp. 311–332. Group. The auditor should be a decentralized to ensure that community-based and lower- 11. OECD, 2005 Development Co-operation Report, Vol. 7, No. 1 (OECD, Paris, 2006). and loose network of independent evaluators technology measures for adapting to climate rather than a bureaucracy that could develop change, like mangrove planting to reduce 12. GAO, Citizens’ Report (2009); www.gao.gov/products/ GAO-09-2SP. a clientelist relationship with the funding coastal erosion in low-lying islands, are fairly 13. R. Levine, W. Easterly, Testimony before the U.S. Senate agency (13). For example, since its inception evaluated alongside large infrastructure projForeign Relations Committee, Hearing on Multilateral Development Banks (U.S. Government Printing Office, in 2004, the U.S.-funded Millennium Chal- ects traditionally favored by the major develDC, 2006). lenge Corporation (MCC) has hired indepen- opment banks, like construction of sea walls 14. Washington, C. Droggitis, W. D. Savedoff, Measuring Impact: Lessons dent researchers to conduct studies assessing (18). It can also set an important precedent from the MCC for the Broader Impact Evaluation Community (Center for Global Development, Washington, 2011). the impact of its spending programs (14). for the other institutions expected to contrib15. W. Easterly, J. Econ. Lit. 47, 373 (2009). ute to climate change financing.
10.1126/science.1211886
909
PERSPECTIVES PSYCHOLOGY
Whether contextual information is helpful in making decisions depends on the situation.
When More Is More
A
man students answered that they had never heard of San Antonio and thus assumed that the city they recognized was larger. Despite having less information, they made the correct decision more often: Less is more. To find out whether this effect holds more generally in decision-making, Freidin and Kacelnik first ask how more information affects the ability of starlings to perform in a simultaneous-choice situation (see the figure, panel A) like the one that was given to the college students. Then, because starlings normally face sequential options in their natural foraging activity (see the figure, panel B), the authors provide them with a sequential choice and ask how more inforB
A matter of choice. In a study of starling decision-making, Freidin and Kacelnik show that the value of contextual information depends on the choice situation. (A) When starlings face a simultaneous choice between two prey items, only information about the value of each item is relevant to choice; the context is irrelevant. (B) When a starling faces sequential choice because it encounters items one at a time, its choice is between this one and what it expects to find next: contextual information.
choices than having more. On page 1000 of this issue, Freidin and Kacelnik (5) challenge this idea by showing that when starlings are faced with food choices, they show a less-ismore effect only when choice is simultaneous and not when it is sequential. An example of the less-is-more effect is provided by a study in which U.S. college students were asked whether San Diego or San Antonio was the larger city. Only two-thirds got it right. Yet, when students in Germany were asked the same question, 100% got it right (4). When asked how they did it, GerDépartement des Sciences Biologiques, UQAM, Case Postale 8888, Succursale Centre-ville, Montréal, QC, Canada H3C 3P8. E-mail:
[email protected]
910
mation affects their performance. Behavioral ecologists often adopt an approach in which decision-making processes are assumed to have evolved through natural selection to yield optimal (rational) choices. This optimality approach typically disregards decision mechanisms. It focuses only on behavioral outcomes and assumes that the underlying cognitive processes can always generate optimal behavior. This position, known as the behavioral gambit, has recently been criticized (6), and arguments have been made for focusing instead on the evolution of decision mechanisms (7). Kacelnik is one of the few behavioral ecologists who directs his research questions at these cognitive decision mechanisms.
In their study, Freidin and Kacelnik placed starlings in automated operant devices and trained them to peck at colored keys that provided the same food reward after different short delays. When the starlings are presented with pairs of different delay options, they perform rationally, choosing the shorter delay. But when they are trained beforehand to expect choices to occur as combinations, such as A versus B or C versus D—that is, they are provided with contextual information—their rationality breaks down when an unexpected pair occurs (such as B versus C): They show a less-is-more effect, erroneously choosing the longer delay option more often. The cleverness of the experimental design lies in using the same starlings, the same operant devices, and the same delay options to show that when the trained birds have to choose between options such as “take A versus some probability of getting another choice”—the contextual information—then knowing which option will likely come up next improves choice. Because context information caused a less-is-more effect in simultaneous choice but improved decision-making in the sequential choice, Freidin and Kacelnik argue that the starlings’ decision mechanism must have evolved to perform in a sequential-choice situation: The sequential choice, they say, corresponds to a more natural foraging choice, the type a starling faces when foraging for prey hidden in the soil. As a bird probes the soil with its bill, it can sometimes detect a target and then needs to decide whether to pursue and eat this item or move on, expecting something better. This sequential choice is typical of optimal prey models in behavioral ecology (8) and requires information about what to expect in the future to correctly reject the current option. Not surprisingly, when starlings in the operant device had a choice between taking a known option and calling up another, they performed more rationally when they were provided with contextual information about the likelihood of the next option: a more-is-more effect. By showing that the less-is-more effect disappears in sequential choice, Freidin and Kacelnik argue that the effect may simply be a by-product of an unnatural simultaneous-choice situation. They may be correct,
18 NOVEMBER 2011 VOL 334 SCIENCE www.sciencemag.org Published by AAAS
CREDIT: P. HUEY/SCIENCE
E
very day humans face choices, whether shopping for coffee beans or voting for a representative. Economists usually assume that we are rational agents who consistently choose options with the greatest utility. Yet, there are numerous departures from rational choice. People often justify continued investment in an unprofitable option because they have already invested in it (1), or switch their preference for one of two products when a third, less attractive one is introduced (2). Similarly, people may prefer an option based on past experience rather than its physical properties (3). Another departure from rational choice is the “less-is-more” effect (4), where having less information leads to better
Downloaded from www.sciencemag.org on November 17, 2011
Luc-Alain Giraldeau
PERSPECTIVES of their companions (9–11) when choosing among options. Perhaps the behavior of others provides readily accessible contextual information that would further improve the starlings’ performance in sequential decisions. Future research on decision-making should no longer ignore the social information component. References
1. I. Simonson, A. Tversky, J. Mark. Res. 29, 281 (1992). 2. M. Bateson, S. D. Healy, T. A. Hurly, Proc. Biol. Sci. 270, 1271 (2003). 3. L. Pompilio, A. Kacelnik, Proc. Natl. Acad. Sci. U.S.A. 107, 508 (2010).
4. D. G. Goldstein, G. Gigerenzer, Psychol. Rev. 109, 75 (2002). 5. E. Freidin, A. Kacelnik, Science 334, 1000 (2011). 6. L.-A. Giraldeau, F. Dubois, Adv. Stud. Behav. 38, 59 (2008). 7. J. M. McNamara, A. I. Houston, Trends Ecol. Evol. 24, 670 (2009). 8. D. W. Stephens, J. R. Krebs, Foraging Theory (Princeton Univ. Press, Princeton, NJ, 1986). 9. J. J. Templeton, L.-A. Giraldeau, Behav. Ecol. 6, 65 (1995). 10. E. Danchin, L.-A. Giraldeau, T. J. Valone, R. H. Wagner, Science 305, 487 (2004). 11. G. Rieucau, L.-A. Giraldeau, Philos. Trans. R. Soc. B 366, 949 (2011). 10.1126/science.1214777
Downloaded from www.sciencemag.org on November 17, 2011
but their birds’ decision-making was still far from perfect, even if they did better using contextual information in the sequentialchoice trials. This type of qualitative support, but quantitative failure, has been the hallmark of almost every test of optimal foraging models in behavioral ecology ( 8). Thus, using Freidin and Kacelnik’s own logic, some important natural component of a starling’s decision process may have been overlooked in the study. Many social animals, including starlings, use information gained by observing the successes and misfortunes
GEOGRAPHY
Understanding Tribal Fates Ronan Arthur and Jared Diamond
CREDIT: ERIK FOLTZ/ISTOCKPHOTO.COM
O
n 29 August 1911, an exhausted, emaciated, terrif ied Indian was found in California’s foothills. Later known as Ishi, he was the last survivor of the Yahi tribe, and the last “wild” stonetool-using hunter/gatherer in the lower 48 states of the United States (1). This centennial year of Ishi’s emergence is an occasion to reflect on the varied fates of the hundreds of Native American tribes that existed in North America in 1492 C.E. (2–4). Many tribes decreased in numbers, disappeared, or lost their homeland, language, or cultural identity. The Navajos are a striking exception, which may illuminate the range of outcomes observed when indigenous peoples encounter modern influences elsewhere in the world. Today, far more Native Americans in the United States speak Navajo than any other native language (5). When Europeans arrived, there were only a few thousand Navajos (they call themselves Diné); now there are over 300,000, half of whom speak Navajo (5–9). Three geographic features of the Navajo homeland—isolation, environment, and resources—together with cultural features of the Navajos themselves, may help to explain the Navajos’ demographic and linguistic explosion (2–9). The Navajo homeland (see the photo) is among the most remote from Euro-American population centers in the lower 48 states. It has not been transected by a major wagon trail, railroad, freeway, or Euro-AmeriGeography Department, University of California, Los Angeles, CA 90095–1524, USA.
can city. Hence Navajos are less exposed to Euro-American people, culture, diseases, and languages; they spend much of their lives on their large reservation. Their homeland’s remoteness meant that parts of it remained difficult for Spanish, Mexican, or U.S. armies to control. American pressure on Navajos therefore developed relatively late in American history, at a time when it was becoming less acceptable to exterminate or permanently expel Native Americans, as had happened to many other tribes.
Geographic and cultural factors help to explain the population explosion of Navajos among Native Americans.
However, many peoples elsewhere in the world, such as New Guinea and Amazonia, are even more isolated but have not exploded in numbers as have the Navajos. This suggests that the Navajos live in optimally intermediate isolation. They were near enough to Spanish and American settlements to acquire livestock and metal tools, but far enough to remain more economically independent than did the closely related Western Apache, who lived nearer to the settlements.
Geographic good luck. Optimally intermediate isolation, an environment suited to the Navajo life-style but not to that of white farmers, and resources not considered valuable until the 20th century may have contributed to the population explosion of the Navajos.
www.sciencemag.org SCIENCE VOL 334 18 NOVEMBER 2011 Published by AAAS
911
PERSPECTIVES
912
ball, lobby Congress, and farm and herd while still living dispersed in their traditional houses (hogans), speaking Navajo, and considering themselves Navajo. Navajo adoptions have also been of other peoples, as recorded historically: Navajos incorporated individuals, clans, and spouses from virtually all neighboring peoples, including Puebloan and Apache groups, Utes, Paiutes, and Spanish/Mexicans. Why have the Navajos been so flexible while maintaining their identity? Part of the reason may be their comparatively recent long-distance migration and arrival in the Southwest, which means that their connection to their land and life-style is less ancient than that of, for example, the Puebloans. Navajo subsistence in their remote new location adapted to a flexible mixed economy based on variable proportions of hunting/ gathering, farming, herding, trading, and raiding. Their residential patterns involve small, mobile, socially fluid groups, rather than the large fixed settlements of Puebloans. This mix of factors may have contributed to the Navajo population explosion, compared to populations of other tribes that share only some elements of this mix. Such questions about differences between Navajos, Ishi’s Yahi tribe, and other Native Americans are not unique to North American sociology. Readers familiar with New Guinea, Africa, and Amazonia will think of similar questions and similar ranges of outcomes among peoples in those areas.
For example, the contrasting current conditions of three New Guinea tribes—the very remote and nonexpanding Fayu, the moderately remote and expanding Simbu, and the Dani constrained by Indonesian immigration—may provide another illustration of optimally intermediate isolation (13). References and Notes
1. T. Kroeber, Ishi in Two Worlds: A Biography of the Last Wild Indian in North America (Univ. of California Press, Berkeley, CA, 1961). 2. Harvard Project on American Indian Economic Development, The State of the Native Nations: Conditions Under U.S. Policies of Self-Determination (Oxford Univ. Press, New York, 2008). 3. D. Champagne, Notes from the Center of Turtle Island (Alta Mira, Landham, MD, 2010). 4. M. Jorgensen, Ed., Rebuilding Native Nations (Univ. of Arizona Press, Tucson, 2007). 5. M. P. Lewis, Ed., Ethnologue: Languages of the World (SIL International, Dallas, ed. 16, 2009). 6. P. Iverson, Diné: A History of the Navajos (Univ. of New Mexico Press, Albuquerque, 2002). 7. R. McPherson, The Northern Navajo Frontier 1860–1900: Expansion Through Adversity (Utah State Univ. Press, Logan, UT, 2001). 8. G. Bailey, R. Bailey, A History of the Navajos: The Reservation Years (Univ. of New Mexico Press, Albuquerque, 1986). 9. R. Towner, The Archaeology of Navajo Origins (Univ. of Utah Press, Salt Lake City, 1996). 10. E. Rosser, Environ. Law 40, 437 (2011). 11. M. Warburton, R. M. Begay, Ethnohistory 52, 533 (2005). 12. M. Nielsen, J. Zion, Eds., Navajo Nation Peacemaking (Univ. of Arizona Press, Tucson, 2005). 13. S. Kuegler, Jägerin und Gejagte (Droemer Verlag, München, 2009). 14. We thank D. Champagne, S. Cornell, G. Gumerman, P. Iverson, J. Kalt, C. Redman, R. Towner, S. Travis, and M. Warburton for helpful discussions. 10.1126/science.1213787
NEUROSCIENCE
Human Locomotor Circuits Conform Sten Grillner Similar patterns of neural activity that drive locomotion in humans, mammals, and birds suggest a conserved mechanism over vertebrate evolution.
W
e tend to consider our species in a different league from that of our relatives in the animal kingdom. The last few decades have been humbling in that respect. Not only is our genome 99% similar to that of apes, but we share many genes with fruit flies and yeast cells. Admittedly, humans excel in cognitive abilities and in skilled movements like writing or talking. Nobel Institute of Neurophysiology, Department of Neuroscience, Karolinska Institute, SE 17177 Stockholm, Sweden. E-mail:
[email protected]
But when it comes to the basic motor repertoire such as walking, posture, or orienting movements, the situation is different; cheetahs, for instance, run faster, as do bipedal ostriches. Animals like horses and antelopes locomote shortly after birth, whereas humans can walk without support after about 1 year after birth. This has in many circles been taken to indicate that the neural control of human bipedal walking is fundamentally different from that of quadruped mammals, and requires the development of new neural circuits during the year following birth.
18 NOVEMBER 2011 VOL 334 SCIENCE www.sciencemag.org Published by AAAS
Downloaded from www.sciencemag.org on November 17, 2011
A second Navajo geographic advantage was their environment: varied, much of it dry and rugged, it was useful for the Navajos’ life-style of dispersed families practicing a mixed economy of hunting, herding, and farming, but unattractive to Euro-Americans looking for lush farmland. Cherokees and many other tribes were unlucky to occupy prime farmland coveted by whites, who expelled them. When the U.S. army finally did expel the Navajos in the 1860s, many remained hidden in canyons. The survivors were eventually permitted to return to their homeland, and then gradually to acquire surrounding land unattractive to white farmers, so that the Navajos now occupy the largest reservation in the lower 48 states. A by-product of their remote location and dispersed life-style was that smallpox and other European epidemic diseases reached the Navajos less often, spread among them with more difficulty, and hence caused lower mortality than among many other tribes such as the nearby Hopis, who lived in small, densely populated pueblos. The remaining Navajo geographic advantage lay in their homeland’s commercially exploitable resources (2, 8, 10). Navajos were fortunate to lack gold, the early discovery and recognized value of which led to the expulsion of the Cherokees and Black Hills Sioux and the massacre of California foothill Indians. Instead, the main Navajo resources are uranium, vanadium, oil, natural gas, and coal, which were not discovered on Navajo land or considered valuable until 1922 and later. By then, Navajos could assert rights to their resources (although deals offered to them were often unfair) (2, 8, 10). As for cultural explanations, the mainstream view of archaeologists and linguists is that Navajos and Apaches originated as Canadian Athabaskan hunter/gatherers, who migrated 1600 km south to reach the U.S. Southwest between about 1000 and 1500 C.E. (9, 11). Among those Athabaskan groups, Navajos were the ones who devoted themselves most to farming, which they learned from Pueblo Indians in the Southwest. Many accounts of Navajos stress their cultural flexibility and selective borrowing, whether from Puebloans, Paiutes, Spaniards, or Anglo-Americans (2, 4, 6, 8–12). While Navajos adopted some new practices, they retained much original Navajo culture and identity. Hence, they increased in numbers and preserved their distinctiveness more than did either conservative tribes that resisted cultural borrowing, or tribes that adopted so wholesale that they lost their identity. Navajos drive trucks, play basket-
CREDIT: B. STRAUCH/SCIENCE
PERSPECTIVES required a different design of the maternal pelvis. The f indings of Dominici et al. and others (3, 11) Brainstem suggest that the human loco(command center) motor circuits are similar to those that control walking in mammals and birds. Because mammals and birds evolved from a common reptilian Spinal cord ancestor, the walking motor (central pattern generators) pattern in humans has at least a reptilian origin—and indeed there are similarities with the limb movements of turtles Four basic phases and perhaps even salamanof motor-neuronal ders. The overall control sysactivity tem for locomotion is, however, even older, and dates back to the oldest vertebrate group that emerged some Limbs (muscle activation) 560 million years ago, at a stage when limbs had not yet evolved (2, 3, 12). Mammals are relative newcomers, having emerged only some 130 Locomotion million years ago. Although human locomotion is controlled by a neural structure similar to that of other walking vertebrates, A common locomotion control system. Central pattern generators each species, from humans in the vertebrate spinal cord produce motor patterns with appropriate to cheetahs and antelopes, timing. Each generator is affected by sensory feedback from the moving has developed specific adaplimb and is activated from the brainstem locomotor command regions. tations to the biomechanical constraints and specialception to walking is linearly related to adult izations. Humans and hominids have, for brain weight (11). Elephants take the longest instance, a particular adaptation related to time to start walking as estimated from the upright gait: They walk with a heel-strike time of conception, followed by humans, well in front of the body, in contrast to most whereas mice represent the other extreme. mammals, which touch the ground with their This correlation may seem surprising, but toes first or with the entire foot. it appears that the brain must reach a cerReferences tain degree of maturity before the ability to 1. N. Dominici et al., Science 334, 997 (2011). walk can be manifested. This may relate to 2. S. Grillner, Science 228, 143 (1985). the concomitant development of the neces3. S. Grillner, Neuron 52, 751 (2006). 4. O. Kiehn, Curr. Opin. Neurobiol. 21, 100 (2011). sary ability to stand and maintain balance 5. S. Harkema et al., Lancet 377, 1938 (2011). and body orientation during walking, which 6. S. J. Harkema, M. Schmidt-Read, D. Lorenz, V. R. Edgerrequires the interaction of many different ton, A. L. Behrman, Arch. Phys. Med. Rehabil. 10.1016/ senses. Moreover, visuomotor coordination j.apmr.2011.01.024 (2011). 7. A. Wernig, S. Müller, A. Nanassy, E. Cagol, Eur. J. Neurois required to steer movements and make sci. 7, 823 (1995). them goal-directed. It appears to take a lon8. H. Forssberg, S. Grillner, J. Halbertsma, Acta Physiol. ger time to develop these neural circuits in Scand. 108, 269 (1980). 9. S. Rossignol, R. Dubuc, J. P. Gossard, Physiol. Rev. 86, 89 order to integrate this information in larger (2006). brains comprising a greater number of neu10. R. G. Lovely, R. J. Gregor, R. R. Roy, V. R. Edgerton, Exp. rons and synapses than in smaller brains— Neurol. 92, 421 (1986). and apparently the ability to walk can only 11. M. Garwicz, M. Christensson, E. Psouni, Proc. Natl. Acad. Sci. U.S.A. 106, 21889 (2009). be manifested at a certain stage of develop12. M. Stephenson-Jones, E. Samuelsson, J. Ericsson, B. Robment. If the human baby were born 1 year ertson, S. Grillner, Curr. Biol. 21, 1081 (2011). and 9 months after conception, it might start walking right away, but this would have 10.1126/science.1214778
www.sciencemag.org SCIENCE VOL 334 18 NOVEMBER 2011 Published by AAAS
Downloaded from www.sciencemag.org on November 17, 2011
However, on page 997 of this issue, Dominici et al. (1) show that the locomotor system of humans and that of other mammals and birds evolved from similar circuitry. Dominici et al. compared the electromyographic activity (electrical activity produced by skeletal muscles) of around 20 different muscles during walking in toddlers (ages 11 to 14 months), preschoolers (ages 22 to 48 months), and adults (ages 25 to 40 years) and the “reflex walking” of the newborn (an infant will attempt to take steps when held upright and moved along a surface with the feet in contact). The authors conclude that the motor pattern of the newborn can be reduced to two phases (flexion and extension in each limb and alternation between the two limbs). The more mature pattern of the toddler requires four phases corresponding not only to flexion and extension but also to transition phases (such as “toe off,” when the toes push off the ground before the flexion phase starts). The pattern in toddlers becomes distinct in preschoolers and adults. They show in addition that the relations between the four different phases in humans are virtually identical to those of various quadruped mammals (Rhesus monkey, cat, rat, mouse) and also birds (guinea fowl), which are bipedal (1) (see the figure). This demonstrates that the locomotor pattern of these different vertebrates is generated by similar neural circuits. Because the complete motor pattern can be generated during walking in spinal animals (2, 3), it follows that the pattern-generating circuits are located at the spinal level in humans as well. The two-phase motor pattern of reflex walking in the infant has also been found in the newborn rodent, where “fictive” locomotion can be generated by its isolated spinal cord (4). This indicates the need to study the spinal cord also at a more mature level of animal development when analyzing the microcircuits underlying walking. A corollary observation is that some patients with partial spinal cord injury can regain some locomotor capacity through treadmill training (5–7). The rationale for this approach is that by activating the sensory receptors that normally signal the progress of the locomotor movements, one can reactivate the spinal locomotor circuits (8– 10). This depends on the presence of spinal locomotor circuits both in animals and humans. Why should it take the human baby a year to start walking, whereas many animals stand or walk on four or two legs minutes after birth (or hatching)? A study of 24 mammalian species, representing 11 different orders, showed that the time from con-
913
PERSPECTIVES BIOCHEMISTRY
One Atom Makes All the Difference
The interstitial atom in the nitrogenase FeMo cofactor cluster is carbon, not nitrogen as previously surmised.
S. Ramaswamy
914
18 NOVEMBER 2011 VOL 334 SCIENCE www.sciencemag.org Published by AAAS
Downloaded from www.sciencemag.org on November 17, 2011
M
ore than 100 years ago, Mar- of the FeMo cofactor cluster (8): the pres- FeMo cofactor cluster and that the interstitin Beijerinck showed that pure ence of an interstitial atom in the middle of tial atom may not play a role in the catalytic cultures of the Azobacter bacte- the cluster (see the figure, center panel). The mechanism (11). rium could fix atmospheric nitrogen (N2). atom was tentatively identified as nitrogen. To help elucidate the role of the cluster, it The enzymes responsible, the nitrogenases, This surprise finding threw up new chal- is crucial to know the identity of the interstibreak down the very strong triple bond in N2 lenges. In parallel, a number of research tial atom. This is what the two papers in this to form ammonia (NH3); at least four types groups studied the nif genes, which are issue achieve: They identify the interstitial of nitrogenases have been isolated from Azo- involved in the biosynthesis of the cluster atom as a carbon atom (see the figure, right bacter and other bacteria (1). It has long (9). Yet, to date, it remains unclear how the panel). been hoped that molecular understanding of proposed interstitial atom is incorporated Spatzal et al. report a higher-resoluthe nitrogenase enzymatic mechanism may into the cluster. tion (1.0 Å) crystal structure. They used lead to an enzymatic process to replace the This lack of understanding did not deter 13C-labeled protein and carried out pulsed high-pressure, high-temperature Haber pro- chemists from trying to use the FeMo cofac- electron paramagnetic resonance studies to cess used commercially to convert nitrogen tor cluster as a model to design catalysts for confirm the observation. The success of the to ammonia. Yet despite intensive reported work is due to the subresearch, it remains unclear how stantial effort to improve crystal1992 2002 2011 nitrogenases efficiently catalyze lographic resolution, as well as the energy-intensive conversion optimization of the production of 13 of nitrogen to ammonia under C-labeled protein. In an indeambient conditions. Two papers pendent experiment, Lancaster in this issue, by Lancaster et al. et al. used x-ray emission specon page 974 (2) and by Spatzal troscopy to uniquely characterize et al. on page 940 (3), resolve a the central atom in the cluster as N C key structural detail that should a carbon atom. The common conhelp to elucidate the mechanism clusion reached by the two groups of the most widely studied type using different techniques proof nitrogenase, the molybdenum vides a strong argument that the (Mo)–dependent nitrogenase. interstitial atom is a carbon atom. When the structures of the two As long as the identification of nitrogenase subunits were first the interstitial atom was inconclureported in 1992 (4, 5), the expecsive, the possibility that it could tation was that knowledge of the be nitrogen sparked speculation Filling the gap. The structure of the FeMo cofactor cluster based on data from structure would quickly translate 1992 (5), 2002 (7), and Lancaster et al. and Spatzal et al. The positions of the that it might be the site of catalto elucidation of the mechanism heavy atoms sulfur, iron, and molybdenum are essentially unchanged over the ysis. Earlier spectoscopic studies (6). The catalytic component of years, but the interstitial atom not seen in 1992 and proposed to be nitrogen did suggest that the central atom the nitrogenase system contains in 2002 has now been identified as carbon. Atom colors: oxygen, red; nitrogen, was not nitrogen (12), but the two metal clusters, the P clus- blue; sulfur, yellow; molybdenum, green; iron, orange; carbon, turquoise; inter- identity of the central atom—as ter and the FeMo cofactor clus- stitial carbon, black. carbon—was not conclusive until ter. The P cluster is involved in now. The availability of the comelectron transfer to the FeMo cofactor clus- the reduction of nitrogen to ammonia. In plete atomic-resolution structure of the cluster, which is the site of substrate binding and 2003, Yandulov and Schrock (10) synthe- ter and the protein will provide a strong platcatalysis. The nitrogenase structures (4, 5) sized a Mo cluster that looks very different form for the elucidation of the mechanism revealed a lot about the P cluster, but their from that in nitrogenase and showed that this of this enzyme. Whether the central carbon resolution (2.7 Å) was insufficient to fully cluster could catalyze the reduction reac- atom in fact plays a role in catalysis is curresolve the FeMo cofactor cluster (see the tion at atmospheric pressure. The proposed rently unknown. Different binding sites for figure, left panel). The structures thus could catalytic mechanism used only biologically N2 have been proposed, but no direct evinot clarify all the details of the catalytic accessible oxidation states of molybdenum. dence exists for how N2 binds to the cluster. mechanism. This allowed the authors to suggest a reason- The new data will aid computational work A decade later, a higher-resolution (1.16 able mechanistic model for how the N2 binds that in turn can guide the design of experiÅ) structure (7) revealed the “inner secrets” to the Mo atom in the model compound— ments to unravel the mechanistic details. and possibly in nitrogenase. In an accomThe work of Lancaster et al. and Spatzal panying commentary, Leigh argued that the et al. is important beyond the field of nitrogeInstitute for Stem Cell Biology and Regenerative Medicine, results provide support for a catalytic mech- nase. It clearly demonstrates how higher resBangalore, Karnataka 560065, India. E-mail: ramas@ncbs. anism based on N2 binding to the Mo of the olution and improved and optimized methres.in
PERSPECTIVES ods can provide more precise information on biological systems. It also clearly shows how multiple experimental techniques provide complementary information, allowing inferences to be drawn that could not be derived from results based on a single technique. The new results give hope that by the end of this decade, most of the mysteries of nitrogenase will be unveiled to enable the design of
benign catalysts for the conversion of atmospheric N2 to biologically accessible forms of nitrogen. References 1. L. C. Seefeldt, B. M. Hoffman, D. R. Dean, Annu. Rev. Biochem. 78, 701 (2009). 2. K. M. Lancaster et al., Science 334, 974 (2011). 3. T. Spatzal et al, Science 334, 940 (2011). 4. M. M. Georgiadis et al., Science 257, 1653 (1992).
5. 6. 7. 8. 9.
J. Kim, D. C. Rees, Science 257, 1677 (1992). W. H. Orme-Johnson, Science 257, 1639 (1992). O. Einsle et al., Science 297, 1696 (2002). B. E. Smith, Science 297, 1654 (2002). Y. Hu, A. W. Fay, C. C. Lee, J. Yoshizawa, M. W. Ribbe, Biochemistry 47, 3973 (2008). 10. D. V. Yandulov, R. R. Schrock, Science 301, 76 (2003). 11. G. J. Leigh, Science 301, 55 (2003). 12. T.-C. Yang et al., J. Am. Chem. Soc. 127, 12804 (2005). 10.1126/science.1215283
MICROBIOLOGY
Bacteria use two convergent strategies to combat toxic reactive oxygen species produced in response to antibiotic treatment.
Downloaded from www.sciencemag.org on November 17, 2011
Antioxidant Strategies to Tolerate Antibiotics Peter Belenky1 and James J. Collins1,2,3
CREDIT: Y. HAMMOND/SCIENCE
I
n living organisms, aerobic metabolism produces toxic reactive oxygen species (ROS) (1). Life can thus be seen as a balance between metabolic rate and a cell’s ability to detoxify ROS. This understanding has led to intense public interest and increased consumption of dietary antioxidants. Although the effectiveness of antioxidant supplements is not yet established, there is no doubt that eukaryotic and prokaryotic cells have developed efficient endogenous antioxidant mechanisms (1, 2). On pages 982 and 986 of this issue, Nguyen et al. (3) and Shatalin et al. (4) describe two such mechanisms that confer antibiotic tolerance in bacteria. It has been proposed that bactericidal antibiotics can induce cellular death through a common oxidative damage mechanism that relies on the production of ROS (see the figure). Through their various primary targets, antibiotics can activate cellular respiration, which leads to the formation of superoxide and the release of iron from iron-sulfur clusters (5–7). Free iron then activates a chemical reaction (the Fenton reaction) to produce ROS in the form of hydroxyl radicals (OH•). These radicals can cause cellular death by damaging proteins, lipids, and DNA (1, 5), or can cause mutations leading to the development of antibiotic resistance (8). Bacteria respond to ROS by up-regulating antioxidant enzymes, including superoxide dismutase (SOD) and catalase (1). They also produce small antioxidant molecules such as Howard Hughes Medical Institute, Department of Biomedical Engineering and Center for BioDynamics, Boston University, Boston, MA 02215, USA. 2Boston University School of Medicine, Boston, MA 02118, USA. 3Wyss Institute for Biologically Inspired Engineering, Harvard University, Boston, MA 02115, USA. E-mail:
[email protected] 1
ascorbic acid and glutathione (2). Further, some bacteria generate nitric oxide (NO), which can induce antibiotic tolerance by blocking the Fenton reaction and stimulating antioxidant enzyme action (9). Nguyen et al. and H2S Shatalin et al. present two convergent strategies used by bacteria to SOD and combat ROS that is proROS catalase duced as a result of antiNutrient biotic treatment. Nguyen stress et al. describe an antioxidant mechanism by Pro-oxidant HAQs which the starvation-signaling stringent response in Pseudomonas aeruginosa and Escherichia coli leads to antibiotic Bacterial antioxidant strategies for antibiotic tolerance. A proposed tolerance in response to mechanism of antibiotic-induced cellular death involves increasing the pronutrient limitation. The duction of reactive oxygen species (ROS), whereas two overlapping tolerstringent response modu- ance mechanisms block ROS formation. SOD, superoxide dismutase; HAQs, lates the transcription of 4-hydroxy-2-alkylquinolines. bacterial genes, diverting resources from growth to nutrient synthesis to grow in aggregates and may thus have limited promote survival until nutrient conditions in access to nutrients. The study by Nguyen et the environment improve. Nguyen et al. found al. provides insights into why such bacteria that mutant bacteria deficient in the stringent can be so difficult to eradicate. response exhibited tolerance to a wide range Shatalin et al. examined the role of endogof antibiotics (including ofloxacin, merope- enously produced hydrogen sulfide (H2S) gas nem, colistin, and gentamicin) by increasing in bacteria. They found that Gram-negative antioxidant enzyme production and blocking and Gram-positive bacteria could be sensithe production of pro-oxidant molecules, thus tized to a wide array of antibiotics by deletreducing toxic OH•. These mutant bacteria ing or inhibiting enzymes that produce H2S, also were more susceptible to ofloxacin in a indicating that the gas confers antibiotic tolmouse infection model. In a biofilm, bacteria erance. H2S elevated the antioxidant capacity
www.sciencemag.org SCIENCE VOL 334 18 NOVEMBER 2011 Published by AAAS
915
PERSPECTIVES The benefit of oxidative stress hormesis has been demonstrated in yeast, worms, and flies (10), and it is likely, as shown by Nguyen et al., Shatalin et al., and related work on lowlevel antibiotic stress (11, 12), that a similar mechanism functions in bacteria. The treatment of bacterial infections is becoming more difficult because of a decline in the current arsenal of useful antibiotics, the development of antibiotic resistance, and the slow rate of new drug development (13). This situation is further aggravated by biofilms and other tolerant bacteria that underlie chronic and recurrent infections. This is particularly problematic with implantable devices such as prosthetic hips, which often require surgical removal to eliminate the infection. Potentiation of currently available antibiotics presents a cost-effective option to overcome these challenges. Both Nguyen et al. and Shatalin et al. identify critical aspects of bacterial biology that could be commandeered as part of new potentiation strategies. For example, each study indicates that it may
ASTRONOMY
be worthwhile to target bacterial antioxidant enzymes and associated pathways as a means to enhance the killing efficacy of bactericidal antibiotics. This could have a great impact on clinical practice and patient outcomes. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.
References
J. A. Imlay, Annu. Rev. Biochem. 77, 755 (2008). E. Cabiscol, J. Tamarit, J. Ros, Int. Microbiol. 3, 3 (2000). D. Nguyen et al., Science 334, 982 (2011). K. Shatalin, E. Shatalina, A. Mironov, E. Nudler, Science 334, 986 (2011). M. A. Kohanski, D. J. Dwyer, B. Hayete, C. A. Lawrence, J. J. Collins, Cell 130, 797 (2007). D. J. Dwyer, M. A. Kohanski, B. Hayete, J. J. Collins, Mol. Syst. Biol. 3, 91 (2007). M. A. Kohanski, D. J. Dwyer, J. Wierzbowski, G. Cottarel, J. J. Collins, Cell 135, 679 (2008). M. A. Kohanski, M. A. DePristo, J. J. Collins, Mol. Cell 37, 311 (2010). I. Gusarov, K. Shatalin, M. Starodubtseva, E. Nudler, Science 325, 1380 (2009). M. Ristow, K. Zarse, Exp. Gerontol. 45, 410 (2010). T. Dörr, K. Lewis, M. Vuli , PLoS Genet. 5, e1000760 (2009). E. A. Debbia, S. Roveta, A. M. Schito, L. Gualco, A. Marchese, Microb. Drug Resist. 7, 335 (2001). G. D. Wright, Adv. Drug Deliv. Rev. 57, 1451 (2005). 10.1126/science.1214823
Does the recent longer-than-usual minimum in sunspot activity indicate that we are heading for an extended period of solar inactivity?
Analyzing Solar Cycles Sami K. Solanki,1, 2 and Natalie A. Krivova1
S
ince observational records began about 300 years ago, and very likely for millions of years before that, the Sun has displayed cyclically varying magnetic activity (1). Approximately every 11 years, a maximum of activity is reached, with a large number of sunspots (see the figure, panel A) present on the solar surface, strong x-ray emission from the corona, and a peak in the number of flares and coronal mass ejections. The latter cause mid- and low-latitude aurorae, disrupt radio communications, perturb navigation systems and radars, produce electric power outages, and can pose radiation hazards for astronauts and aircraft crew. Solar cycle activity maxima are separated by minima during which only a few or no sunspots are present on the solar surface and other indicators of solar activity are equally muted (see the figure, panel B). Minima have lasted typically 2 to 3 years in the 20th century. Consequently, as solar activity decreased to nearminimum levels in 2005–2006, most solar astronomers expected that the Sun would be Max-Planck-Institut für Sonnensystemforschung, 37191 Katlenburg-Lindau, Germany. 2School of Space Research, Kyung Hee University, Yongin, Gyeonggi 446-701, Korea. E-mail:
[email protected] 1
916
bubbling with activity again by 2007 or 2008. However, the Sun did not restart displaying appreciable activity until 2010. Also, the rise in activity has been slow relative to most other cycles during the last century. Surprised by this unexpectedly long minimum, the solar physics community reacted in various ways. Interpretations ranged from a lull before the storm, with the next cycle to be particularly strong, to the beginning of a grand minimum, a multidecadal episode of almost nonexistent solar activity. Such a prolonged period of quiescence last occurred in the 17th century, when almost no sunspots were visible for around 60 years—the so-called Maunder minimum (2). Which, if any, of these scenarios is correct? In particular, are we heading for a grand minimum? Predictions of solar activity have been notoriously wayward in the past, with similar scatter of predicted behavior also true for the maximum of cycle 23, as little as 5 or 6 years before it was reached (1, 3). The best record is produced by empirical methods relying on precursors, but even they give reasonably accurate predictions of its maximum only after a cycle is well under way. To estimate the future of solar magnetic
activity beyond the next cycle, we must therefore take guidance from its past. During the past 70 years or so, the Sun has been in a grand maximum, a period of strong activity cycles, which by chance coincided with the space age and the great variety of data that it has provided. In the 19th century, the cycle minima were similarly long and quiet as the one we have just left. Also, the slow start of the present cycle—cycle 24—suggests (according to a rule named after the Swiss solar physicist Max Waldmeier) that it will be relatively weak, peaking at a yearly averaged sunspot number value of 60 to 100 (1, 3), compared with 120 in cycle 23 and even larger values in four of the five cycles before that. There is similarity between the present cycle and the beginning of solar cycle 14 (see the figure, panel C). Cycle 14, the weakest cycle of the 20th century, peaked in 1905 at a yearly averaged sunspot number of 63.5. The sunspot number averaged over the first 9 months of 2011 is 45.5 (solid orange circle in panel B). Although this is low relative to the past nine cycles, it still exceeds 20, the amplitude of the two last cycles preceding the Maunder minimum (4). This speaks against, but does not rule out, a grand min-
18 NOVEMBER 2011 VOL 334 SCIENCE www.sciencemag.org Published by AAAS
Downloaded from www.sciencemag.org on November 17, 2011
of bacteria by suppressing the Fenton reaction and stimulating SOD and catalase production. Interestingly, the authors also show that H2S can act as a diffusible protective agent in bacterial populations. Further, cells deficient in H2S produced increased amounts of NO, and the two gases can act synergistically to induce antibiotic tolerance, demonstrating some redundancy in these protective mechanisms. The antibiotic tolerance mechanisms presented in these two studies have several strong similarities. The most obvious common aspect is that the stringent response and H2S both induce tolerance by elevating the production of antioxidant enzymes. These effects can be explained, in part, by considering that nutrient limitation and the production of toxic H2S are forms of cellular stress. One possibility is that these mechanisms may act as low-level stress conditions that activate antioxidant responses, priming bacterial cells to counteract the more lethal oxidative stress induced by antibiotics—thus confirming the adage, “that which does not kill you only makes you stronger.”
PERSPECTIVES
Sunspot number
Grand maximum
B
200 150
Cycle 0
100
0 1600
Sunspot number
10
15
20
Maunder minimum
50
250 200 150 100 50 0
5
1700
1800
1900
Year
2000
C Cycle 19 Cycle 14
Cycle 24 0
2
4
6
8
Years from cycle start
10
12
Solar (in)activity. (A) High-resolution image of a sunspot. (B) Yearly averaged Zurich (orange) and group (blue) sunspot number (4, 10, 11). Before around 1880, group sunspot number is thought to be a more robust representation of actual levels of activity. The Zurich number (also called the Wolf number) was introduced in the 1840s by Rudolf Wolf as an objective measure of the number of sunspots. The group sunspot number is a latter-day improvement, but is not yet officially available for cycle 23. The solid orange circle marks the average over the first 9 months of 2011. (C) Monthly averaged Zurich sunspot number for cycles 14 (green), 19 (blue), and 24 (red). Cycle 19 is the strongest on record.
imum starting within the next two decades. A statistical analysis of earlier grand maxima and minima may provide a bigger-picture view of longer-term behavior of solar activity. As these occurred well before the invention of the telescope, we rely on indirect indicators such as the cosmogenic isotopes 14C and 10Be, produced when cosmic ray particles collide with constituents of our atmosphere. Modeling allows solar activity to be reconstructed back to the beginning of the Holocene period, about 11,000 years ago. The records reconstructed in this manner (5–7) reveal a rich array of grand minima and maxima. A statistical analysis of the grand maxima shows that they were in general shorter than the one that just ended (5, 6, 8). Its demise was (statistically) overdue. What happens after a grand maximum is over? 10Be data indicate that the probability of a grand minimum occurring within 40 years of the end of a grand maximum is 8%, rising to 40 to 50% within 200 years (9). Similar results are found from the compilation of 27 grand minima and 19 maxima since 9500 B.C.E. based on 14C data (6). However, there is no guarantee that the Sun will gradually slide into a grand minimum after the justended grand maximum. Half the grand maxima in (6) were followed by one or more subsequent grand maxima before a grand minimum finally occurred. In addition, the mean time between the end of a grand maximum and the beginning of the next grand minimum was 318 years. This average interval is also only slightly shorter than the 349 years that passed on
Prediction of solar activity has not been reliable, because of the nonlinearity of the solar dynamo producing the magnetic field that is responsible for solar activity. On long time scales, our best bet is to consider the statistical evidence gleaned from previous grand minima and maxima. But these also give a mixed message. A grand minimum might be just around the corner and could hit us in the next 30 years, although with a probability below 10%. It is not even clear in which direction solar activity will develop in the longer term. Thus, the next grand extremum is just as likely to be a maximum as a minimum.
average between the end of a grand minimum and the start of the next one. The Maunder minimum ended approximately 300 years ago, which is longer than the majority of such intervals (the median time between grand minima is 240 years), but still short relative to the 1420 years that passed between the two grand minima that occurred between 3000 and 5000 years ago.
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.
References and Notes
D. H. Hathaway, Liv. Rev. Sol. Phys. 7, 1 (2010). J. A. Eddy, Science 192, 1189 (1976). K. Petrovay, Liv. Rev. Sol. Phys. 7, 6 (2010). J. M. Vaquero, M. C. Gallego, I. G. Usoskin, G. A. Kovaltsov, Astrophys. J. 731, L24 (2011). S. K. Solanki, I. G. Usoskin, B. Kromer, M. Schüssler, J. Beer, Nature 431, 1084 (2004). I. G. Usoskin, S. K. Solanki, G. A. Kovaltsov, Astron. Astrophys. 471, 301 (2007). F. Steinhilber, J. A. Abreu, J. Beer, Astrophys. Space Sci. Trans. 4, 1 (2008). J. A. Abreu, J. Beer, F. Steinhilber, S. M. Tobias, N. O. Weiss, Geophys. Res. Lett. 35, L20109 (2008). M. Lockwood, Proc. R. Soc. London Ser. A 466, 303 (2009). Solar Influences Data Analysis Center, http://sidc.oma.be. D. V. Hoyt, K. H. Schatten, Sol. Phys. 179, 189 (1998). We thank J. Hirzberger for providing panel A, and I. Usoskin for valuable comments. Supported by Korean Ministry of Education, Science and Technology WCU grant R31-10016. 10.1126/science.1212555
MATERIALS SCIENCE
True Performance Metrics in Electrochemical Energy Storage Y. Gogotsi1 and P. Simon2 Exceptional performance claims for electrodes used in batteries and electrochemical capacitors often fail to hold up when all device components are included.
A
dramatic expansion of research in the area of electrochemical energy storage (EES) during the past decade has been driven by the demand for EES in handheld electronic devices, transportation, and storage of renewable energy for the power grid (1–3). However, the outstanding properties reported for new electrode materials may Department of Materials Science and Engineering and A. J. Drexel Nanotechnology Institute, Drexel University, Philadelphia, PA 19104, USA. 2Université Paul Sabatier– Toulouse III, CIRIMAT UMR-CNRS 5085, 118 Route de Narbonne, 31062 Toulouse, France. E-mail: gogotsi@drexel. edu,
[email protected] 1
not necessarily be applicable to performance of electrochemical capacitors (ECs). These devices, also called supercapacitors or ultracapacitors (4), store charge with ions from solution at charged porous electrodes. Unlike batteries, which store large amounts of energy but deliver it slowly, ECs can deliver energy faster (develop high power), but only for a short time. However, recent work has claimed energy densities for ECs approaching (5) or even exceeding that of batteries. We show that even when some metrics seem to support these claims, actual device performance may be rather mediocre. We will focus here
www.sciencemag.org SCIENCE VOL 334 18 NOVEMBER 2011 Published by AAAS
917
Downloaded from www.sciencemag.org on November 17, 2011
A
PERSPECTIVES 105
918
Carbon material (10 mg/cm2)
Power density (Wh/liter)
Power density (Wh/kg)
which is a typical value for commercial ECs with activated carDevice with 1 mg bon. If a 2-µm film of the same Maximum of carbon per cm2 reported Carbon material material is used in the device, a values Device with 10 mg 3 (density 0.3 g/cm ) of carbon per cm2 much greater drop occurs, which is why “paper batteries” or thinCarbon 104 104 film ECs cannot be used for storCarbon ing large amounts of energy. Device with a 120-µm film Divided Divided (4 mg of carbon per cm2) Collector The gravimetric energy denby ~4 by ~5 Carbon sity is almost irrelevant compared Divided Divided by ~100 by ~12 to areal or volumetric energy for Collector 103 103 0.2 1 10 100 0.1 1 10 50 microdevices and thin-film ECs, because the weight of the active Energy density (Wh/kg) Energy density (Wh/liter) A tale of two plots. One way to compare electrical energy storage devices is to use Ragone plots (10), which show both power material used in a micrometer-thin density (speed of charge and discharge) and energy density (storage capacity). These plots for the same electrochemical capaci- film on a chip or a nanotube coattors are on a gravimetric (per weight) basis in (A) and on a volumetric basis in (B). The plots show that excellent properties of ing on a smart fabric is negligible. carbon materials will not translate to medium- and large-scale devices if thin-film and/or low-density electrodes are used (10). These systems may show a very high gravimetric power density on ECs, but these considerations also apply to accounts for about 30% of the total mass of and discharge rates, but those characteristics lithium (Li)−ion batteries. the packaged commercial EC, a factor of 3 to will not scale up linearly with the thickness Typically, the performance of both bat- 4 is frequently used to extrapolate the energy of the electrode (7), i.e., the devices cannot teries and ECs is presented by using Ragone or power of the device from the performance be scaled up to power an electric car. Ragone plots (see the figure) that show the relation of the material. Thus, the energy density of 20 plots are only one measure of a device; they do between energy density (how far an electric Wh/kg of carbon will translate to about 5 Wh/ not show other important properties, such as car can go on a single charge) and power den- kg of packaged cell. the device’s cycle lifetime, energy efficiency, sity (how fast the car can go). A commercial However, this extrapolation is only valid self-discharge, temperature range of operaEC can harvest or release more energy than for electrodes with thicknesses and densities tion, or cost. They may also provide misleada typical Li-ion battery can deliver on time similar to those of commercial electrodes (100 ing information for flow and semisolid batterframes of less than 10 s, and it can be used to 200 µm or about 10 mg/cm2 of carbon film). ies (3, 9), where energy and power densities for an almost unlimited number of charge and An electrode of the same carbon material are decoupled. discharge cycles (4). A near-term application that is 10 times thinner or lighter will reduce By presenting energy and power densities will be storing energy for car starter motors to energy density by three- to fourfold (from 5 in a consistent manner, we can facilitate introallow engine shut-offs when stopped (6) and down to 1.5 Wh/kg based on cell weight, see duction of new materials and find solutions harvesting braking energy. panel A), with only a slight increase in power for EES challenges the world faces. National Increasing the energy density of ECs usu- density. Our ability to predict the performance and international testing facilities should be ally comes at the cost of losses in cyclabil- of a 200-µm-thick electrode by testing a 1-µm created for benchmarking electrodes and ity (5) or power, which are the most impor- film (7) or a small amount of material in a cav- devices similar to the facilities that exist for tant properties of ECs and without which they ity microelectrode (8) is still very poor. Exper- benchmarking photovoltaics. Clear rules for become mediocre batteries. A major effort has imental data show that there may be an addi- reporting the performance of new materials been directed toward increasing the energy tional drop in the capacitance by a factor of for EES devices would help scientists who are density of ECs by either increasing the capac- 2 to 3 when the thickness of the nanoporous not experts in the field, as well as engineers, itance of the material, C, or the operation carbon electrode increases (7). investors, and the general public, who rely on voltage window, V, or both, since the energy Much of this uncertainty stems from the data published by the scientists, to assess stored is proportional to CV 2. Some recent reporting gravimetric, rather than volumet- competing claims. publications on graphene and nanotube-based ric, energy and power densities of materials and Notes materials have used Ragone plots to argue that and devices. Many nanomaterials, such as References 1. J. R. Miller, P. Simon, Science 321, 651 (2008). supercapacitors can achieve the energy den- nanotubes or graphene, have a low packing 2. M. Armand, J.-M. Tarascon, Nature 451, 652 (2008). 3. Z. Yang et al., Chem. Rev. 111, 3577 (2011). sity of batteries. Those claims are summarized density (<0.5 g/cm3), which leads to empty 4. J. R. Miller, A. F. Burke, Interface 17, 53 (2008). in the gray area in the upper right corner of space in the electrode that will be flooded by 5. C. Liu et al., Nano Lett. 10, 4863 (2010). panel A in the figure. electrolyte, thereby increasing the weight of 6. www.greencarcongress.com/2010/10/conti-20101014. Reporting the energy and power densi- the device without adding capacitance. An html 7. J. Chmiola et al., Science 328, 480 (2010). ties per weight of active material alone on a extreme case would be the use of a carbon 8. C. Portet et al., Electrochim. Acta 53, 7675 (2008). Ragone plot (panel A) may not give a realis- aerogel with 90% porosity. The volumetric 9. M. Duduta et al., Adv. Energy Mater. 1, 5111 (2011). tic picture of the performance that the assem- energy of such an electrode will be 20% that 10. Details for the figure plots are available at Science Online. bled device could reach because the weight of of a carbon electrode with just 50% porosity. the other device components also needs to be If we consider a low-density graphene 11. We thank the Partner University Fund for funding our collaborative efforts. taken into account. ECs are similar to Li-ion electrode (0.3 g/cm3) with an extremely high batteries in that they contain current collec- gravimetric energy density of 85 Wh/kg (gray Supporting Online Material tors, electrolyte, separator, binder, connec- area in panel A of the figure), its volumetric www.sciencemag.org/cgi/content/full/334/6058/917/DC1 tors, and packaging, in addition to carbon- density will be 25.5 Wh/liter for the electrode SOM Text based electrodes. Because the carbon weight and ~5 Wh/liter for the device (panel B), 10.1126/science.1213003 A
B
Device with a 2-µm film (0.07 mg of carbon per cm2)
18 NOVEMBER 2011 VOL 334 SCIENCE www.sciencemag.org Published by AAAS
Downloaded from www.sciencemag.org on November 17, 2011
105
PERSPECTIVES RETROSPECTIVE
Steven P. Jobs (1955–2011)
A brash, innovative, and intuitive thinker spurred radical technological changes that reshaped the cultural and economic landscape.
CREDIT: JEFF CHIU/AP PHOTO
L
ast month, the world lost a technological leader, widely hailed for forging a new age of personal computers and digital products, forever changing the economic and social landscape of our time. A master of new media and new technology, the career of Apple cofounder and computer visionary Steve Jobs was closely followed around the world. Eventually, after several years of serious illness in the public eye, mortality called. His death occasioned an overflow of public grief and deep sadness tinged with sincere affection. Clearly, an era has passed. Eighty years earlier to the month, the world lost another technological genius who brought entertainment to the masses, created several new industries, and transformed the economy—Thomas Edison. The business journalist Randall Stross viewed the two men’s many similarities, including modest education, quick judgment, strong temper, and success with the entertainment industry, to be less impressive than the fundamental contrast between Edison’s failure as a businessman (he lost a fortune in iron ore) and Jobs’s success as a “beloved billionaire” and accessible “mortal” (1). Perhaps Jobs was more like the financially successful entrepreneurs and industrialists Henry Ford or Andrew Carnegie? Jobs himself offered a clue in his choice of Walter Isaacson as biographer, whose best-selling books profiled Benjamin Franklin and Albert Einstein. Both Edison and Jobs were indeed visionaries, but one thing is for sure, despite public belief to the contrary: Edison most certainly did not invent the light bulb, and Jobs did not invent the personal computer. Might our public fascination with such innovators offer clues about American history and culture? Edison and Jobs each thrived during a historical moment of radical and unsettling technological change, Edison’s involving the pervasive use of electricity, and Jobs’s, digital computing. Edison became a personified “wizard” just when American business was reorganizing itself into unfathomably large productive units— Edison’s electricity companies were merged into General Electric in 1892. He seemCharles Babbage Institute, University of Minnesota, Minneapolis, MN 55455, USA. E-mail:
[email protected]
ingly provided a human face for the new age. The tendency to personify computers with Steve Jobs hints at some unease. Today’s pervasive personal computing provides instant entertainment, communication, and information but also brings worrisome shifts in the globalized economy and concerns over security, privacy, and intimacy. Historically it’s not quite clear which personal computer Jobs supposedly invented. The original Apple II, created in a Silicon Valley garage, made Jobs and Apple cofounder Steve Wozniak very wealthy. More important, the Apple Macintosh launched a personal-computing paradigm. Yet the Macintosh project was not Jobs’s creation (2). In 1979, Jef Raskin hatched the concept of an inexpensive yet sophisticated information appliance (the Macintosh, named after his favorite apple). The graphical user interface, with mouse, windows, and icons, owed much to Raskin’s familiarity with Xerox PARC’s Alto computer. Raskin recruited several geniuses (including Burrell Smith and Bill Atkinson) to redesign the expensive PARC device. Jobs took over the project in 1981, added Andy Hertzfeld, and brought Raskin’s vision to market. A year after the Mac’s resonant 1984 launch, Jobs, then a distressingly dysfunctional manager with a massive block of stock, was fired from Apple. The mass-market microcomputer was just one of the historical trends in the 1980s. In 1981, the year IBM introduced its eponymous Personal Computer, Virginia Rometty joined the company with an undergraduate degree in computer science. Computing then was distinctly hospitable to women. In the mid-1980s, women constituted 38% of the white-collar computing workforce in the United States. Eleven women were among the 40 or so who famously signed the inside case of the Mac (3). Rometty, who was recently named as IBM’s CEO, experienced computing at a unique historical moment. So did Carly Fiorina, later CEO of Hewlett-Packard; Meg Whitman, recently named CEO of Hewlett-Packard; and Ursula Burns, who joined Xerox as a summer engineering intern in 1980 and is now the com-
pany’s CEO. But last year, women accounted for just 11% of undergraduate computer science degrees (down from 37% in the mid1980s), and workforce figures have slid as well (4). Programming personal computers has apparently become a male enclave. Will there be a next generation of women executives in information technology? Another historical question concerns charismatic leadership. Many such leaders—Edison, mathematician and early computer scientist John von Neumann, supercomputer pioneer Seymour Cray—personally inspired innovative teams but failed to create organizations capable of sustaining innovation in their absence. Apple, briefly rivaling Exxon-Mobil as the largest capitalized company in the world, must be nervous. Apple’s stunning success with mobile media and personal computing owes much to Jobs. It is difficult to know if Apple will achieve a centennial of its founding, as IBM did this summer. One key will be whether Apple’s famed designer Jony Ive and logistics maven and CEO Tim Cook can consistently realize Jobs’s aspiration (as one Apple insider phrased it) “to do the greatest thing possible, or even a little greater” (5). References
1. R. Stross, “The wizard and the mortal: Two sides of genius,” New York Times, 8 October 2011, tinyurl. com/3llosr3. 2. A. Hertzfeld, Revolution in the Valley (O’Reilly Media, Sebastopol, CA, 2005), p. 272. 3. A. Hertzfeld,”Signing party,” tinyurl.com/3q263vo (1982). 4. T. Misa, Ed., Gender Codes (Wiley, Hoboken, NJ, 2010), pp. 4–6; p. 21, note 2. 5. W. Isaacson, Steve Jobs (Simon and Schuster, New York, 2011), p. 123.
www.sciencemag.org SCIENCE VOL 334 18 NOVEMBER 2011 Published by AAAS
Downloaded from www.sciencemag.org on November 17, 2011
Thomas J. Misa
10.1126/science.1216019
919
SPECIALSECTION
INTRODUCTION
Downloaded from www.sciencemag.org on November 17, 2011
Electricity Now and When WHETHER IT STARTED WITH FALLING WATER OR WITH THE BURNING OR RADIOACTIVE
CREDIT: FOTOSEARCH
decomposition of fuels, creating and delivering electrical power used to be a straightforward process of trying to balance generation, distribution, and demand at a reasonable cost to end users. Peak power requirements have grown, as has the size of the fluctuations between daily maximum and minimum requirements. Very little capacity exists for storing electricity, but an increased reliance on renewable sources, especially solar and wind power, will require better solutions to electricity storage to cope with their intermittent nature. Dunn et al. (p. 928) review the present situation with regard to electrical energy storage, which is now dominated by sodium-sulfur (Na/S) and sodium– metal chloride (Na/MeCl2) batteries that operate with high-temperature electrolytes. Redox flow and lithium batteries are emerging options, and they also discuss the “rolling storage” of electricity in battery-powered vehicles. In a related Perspective (p. 917), Gogotsi and Simon demonstrate a need for a better way to assess and compare the properties of electrochemical capacitors and lithium ion batteries, because current metrics do not necessarily reflect device performance. If there were efficient conversion methods, electrical energy could be stored as a fuel rather than directly as stored charge. This is often discussed in terms of a hydrogen economy, but that is by no means the only fuel of interest. Solid-oxide fuel cells, which operate at high temperatures, could allow distributed electrical generation from natural gas or regenerated fuels created from excess electrical power, or allow supplementation of the grid during peak power periods. Wachsman and Lee (p. 935) discuss developments that should allow lower operating temperatures and costs for these sources, which could widen their adoption as both stationary and mobile sources. Two News stories describe aspects of better ways to harvest solar power. Cartlidge (p. 922) describes efforts to improve thermal storage, a technology that enables solar plants to continue generating electricity after dark. Service (p. 925) discusses recent progress in artificial photosynthesis to create hydrogen and hydrocarbon fuels, which could be used either for transportation or for centralized electricity generation. A growing population and the push toward renewable and less polluting resources are driving the construction of a wider range of methods for electricity generation and a much more complicated electricity grid. In many developed countries, a reliable supply of electrical power is taken for granted, but in many developing countries, regular and widespread outages can be the norm. The research outlined in these pieces points to some of the ideas being considered to ensure that the lights can stay on.
Materials for Grid Energy CONTENTS News 922
Saving for a Rainy Day
925
Turning Over a New Leaf
Sunlight in Your Tank—Right Away
Reviews 928
Electrical Energy Storage for the Grid: A Battery of Choices B. Dunn et al.
935
Lowering the Temperature of Solid Oxide Fuel Cells E. D. Wachsman and K. T. Lee
See also Editorial p. 877; News Focus story p. 896; and Perspective p. 917
– MARC LAVINE, PHILLIP SZUROMI, ROBERT COONTZ
www.sciencemag.org SCIENCE VOL 334 18 NOVEMBER 2011 Published by AAAS
921
922
ral gas (about $0.06). Increasing output by being able to generate electricity after sunset will reap economies of scale, says Yogi Goswami, a chemical engineer at the University of South Florida in Tampa. But capital costs must also be reduced. Cheaper, simpler mirrors will be essential, says Fabrizio Fabrizi of Italy’s national agency for energy research, ENEA, and improved storage technology also has an important role to play. “There is a need to force down the cost of investment,” Fabrizi says, estimating that improvements in storage “could reduce that cost by up to 25%.” Oil, salt, and steam The standard storage material for plants like Andasol, a mixture of 60% sodium nitrate and 40% potassium nitrate, already does its job very well. It is stable at temperatures up to 600°C. Its high specific heat capacity and a high density enable it to store a lot of energy in very little space. Its low vis-
18 NOVEMBER 2011 VOL 334 SCIENCE www.sciencemag.org Published by AAAS
Downloaded from www.sciencemag.org on November 17, 2011
particularly those that use photovoltaic panels, which convert sunlight directly into electricity. In principle, batteries could store such renewable energy, but they remain very expensive. Instead, Andasol hoards its raw product: heat. Its “batteries”—three pairs of innocuous-looking metal tanks containing molten salt—hold enough energy to generate electricity for about 7.5 hours, allowing the plant to provide almost round-theclock electricity during the summer. Indeed, Spain’s national grid operator has classified Andasol as a “predictable” source of energy, allowing the percentage of locally generated electricity that is provided by renewable sources to increase. Experts say storage systems could help concentrating solar power (CSP) clear another major hurdle: cost. Its price per kWh is currently about $0.17—slightly more expensive than that of photovoltaics ($0.16), and nearly three times the cost of natu-
CREDIT: PAUL LANGROCK/SOLAR MILLENNIUM AG
The Andasol complex at the foot of the Sierra Nevada mountains in Granada, southern Spain, is one of the world’s largest solar power stations. Its 600,000 parabolic mirrors, lined up in hundreds of rows over an area of several square kilometers, concentrate the sun’s rays to provide heat that creates steam for electricity generation. The plant produces some 150 megawatts, enough to meet the needs of about half a million people. What sets Andasol apart, however, is not how much energy it delivers—it’s how much it holds back. The plant is designed to store part of the solar energy it collects so that it can produce electricity after the sun sets or disappears behind a cloud. The f ickleness of sunlight, like the unsteadiness of wind, poses a major obstacle for renewable energy. Grid operators can take up the slack with fossil fuel or nuclear plants, but this need to compensate limits the contribution of wind and solar plants—
SPECIALSECTION Canned heat. The Andasol power station in Spain
uses tanks of molten salt to store solar energy so that it can continue generating electricity when the sun isn’t shining.
1
2
3 5
CREDITS (TOP TO BOTTOM): SOLAR MILLENNIUM AG; ADAPTED FROM SOLAR MILLENNIUM AG
1. Solar field; 2. Storage; 3. Heat exchange; 4. Steam turbine and generator; 5. Condenser
cosity when molten makes it easy to pump through pipes. And its ingredients are cheap and abundant. The challenge is to make better use of its virtues. Andasol produces electricity in two stages. First, the parabolic mirrors concentrate the sun’s energy along the length of a pipe fixed just above their surface. Then synthetic oil flowing through the pipe heats up and travels to a heat exchanger, where it generates steam that turns a turbine. En route, some of the oil takes a detour through a separate heat exchanger. Molten salt being pumped from a “cold” tank at a temperature of about 290°C takes heat from the oil, flows into a “hot” tank at about 390°C, and sits until it is needed. Later, the salt is pumped back to the cold tank; as it passes back through the heat exchanger, it reheats oil returning from the steam generator, giving it enough thermal energy for another round. In this “two-tank indirect storage” system, the oil serves as a heat
transfer fluid (HTF). The salt provides the heat that keeps the electricity flowing, but it’s the HTF that raises steam. The obvious way to improve on this approach is to cut out the middleman. Instead of using one material to absorb the sun’s heat and a second material to store it, why not use one material to do both? Such a “direct storage” approach would eliminate one heat exchanger. With the right material—one that remained stable at higher temperatures—it would also raise the temperature of the hot tank, making electricity production more efficient, and would increase the amount of heat stored in a given volume. Direct storage has already supplied electricity to the grid in California. The Solar Energy Generating Systems (SEGS) I power plant in the Mojave Desert—one of nine sister plants that together make up the world’s largest operating solar power station—used hot mineral oil to meet
www.sciencemag.org SCIENCE VOL 334 18 NOVEMBER 2011 Published by AAAS
Downloaded from www.sciencemag.org on November 17, 2011
4
demand on winter evenings starting in 1985, but in 1999 the storage system was damaged by a fire and was not replaced. Installing storage systems in SEGS II through IX would have been prohibitively expensive because the HTF used in those plants, a synthetic oil called Therminol, would need special pressurized tanks to keep it liquid. (Most of the plants use gas boilers as backup when the sun doesn’t shine.) Most current research projects on direct storage rely on molten salt—the same mixture used at Andasol. One is the Italian energy agency ENEA’s Archimede project, which switched on in July 2010. Archimede is a 5-megawatt pilot plant incorporated into a combined-cycle gas power station close to the Sicilian city of Syracuse. The plant’s hot tank is maintained at 550°C—160° higher than Andasol’s. As a result, Fabrizi says, more heat is lost in transit from the mirrors to the steam generator, but higher generating efficiencies and savings on storage considerably outweigh the loss. “Using the same molten salt as Andasol, we can store the same amount of energy but using about 40% less salt,” he points out. “That’s a very dramatic reduction of cost.” One significant challenge for Archimede arises from the salt’s melting point, 240°C. If the molten salt cools to that temperature, it will freeze solid and block the pipes—a problem that could be extremely costly and time-consuming to resolve. To keep the pipes hot, Fabrizi says, the plant continuously circulates the salt through them and turns on electric heaters if necessary. If salts with lower melting points were available, Fabrizi notes, Archimede’s operators could use less energy to keep the pipes hot and—if circulation were to stop for some reason— would have more time to intervene before the fluid froze. Researchers at Sandia National Laboratories in Albuquerque, New Mexico, are working on the melting-point problem. David Gill, a mechanical engineer at Sandia, says that he and colleagues have found several mixtures that freeze below 100°C, but reaching such low temperatures cheaply is a challenge. Two of the salt mixtures— a four-component salt that freezes below 80°C, and a five-component salt that freezes closer to 70°C—should be “economically
923
Towers and blocks Gill and other researchers hope that new designs will keep direct-storage CSP plants from freezing up. Plants like the 20-megawatt Gemasolar plant in Seville, Spain, which opened in June, attack the prob- Let it freeze? lem by doing away with parabolic mirrors In a still more radical departure, some entirely. Instead, hundreds or thousands of researchers hope to achieve much higher small reflectors known as heliostats direct storage capacities by harnessing the heat the sun’s radiation to the top of a tall tower associated with a material’s change of phase. and onto a single receiving module through Instead of raising the temperature of already which the HTF flows. Such “power tower” molten salt, their scheme uses solar energy designs achieve high operating temperatures to change salt to a liquid. Hot HTF passes and thus very high efficiencies. Like Archimede, Gemasolar Light work. Power towers uses molten salt as both HTF and may prove cheaper than storage medium. But the much parabolic mirrors. shorter length of its tubing minimizes both heat losses and the chance that the salt will freeze. Gill says the combination of high efficiency and low losses is attracting increasing investment to power towers in the United States. “Parabolic troughs have a long track record and so are generally seen as a less risky investment,” he says. “But putting molten salt in is seen by some investors as putting the risk back in.” Another approach to thermal storage scraps flowing molten salt in favor of solid materials that just sit still. In a “passive” storage system developed by researchers at the German Aerospace Center (DLR) in Stuttgart, HTF from the mirror array passes through solid salt, melting it; later, cool through pipes embedded in concrete or cast- HTF absorbs energy, so refreezing the storable ceramic materials. The scientists have age material into a solid. Because the latent found that the ceramics offer superior heat heat associated with a material’s change of capacities and thermal conductivities but phase is much greater than the “sensible” are too expensive and impractical. “On the heat required to raise its temperature, much one side you want good thermophysical less storage material would be needed than properties, and on the other side the mate- in conventional molten-salt storage. rial has to be durable, workable, and cheap,” Again, nitrate salts are likely to be the DLR’s Doerte Laing says. “Concrete repre- storage medium of choice. But because they sents a good compromise between the two.” are relatively poor conductors of heat, sevSince 2008, the German group has been eral research groups are working on designs making detailed thermal tests of a 20-cubic- that cause the salts to absorb or lose heat meter concrete block at a test facility at the more effectively. University of Stuttgart. The tests have shown To increase the surface area of heat transgood heat transfer between the concrete and fer, Goswami and colleagues at the Uniembedded pipes while at the same time versity of South Florida seal the salt inside
924
small capsules with diameters of a few centimeters and let the HTF flow around them. The team is currently working on optimizing the size of the capsules and developing an industrial-scale method for carrying out the encapsulation process. Laing and coworkers, meanwhile, route the HTF through the salt via pipes outfitted with aluminum fins to speed heat transfer. In tests carried out with the utility company Endesa at the Litoral power plant in southern Spain, the German group has shown that this phase-change-based technology could provide storage for so-called direct steam generation. This technology, which uses the water from a solar power plant as the HTF, does away with the expensive oils and the steam generator that normally feeds the turbine. It also reaches higher temperatures than oil-based HTFs do—up to 550°C with superheated steam. The method is potentially cheap and efficient, but it requires that heat be transferred to and from the storage medium at a near constant temperature, which would require an enormous volume of salt in the case of sensible heat storage. The phase-change approach would be far better suited, Laing says. She adds that her group has shown that the approach is technically feasible and is working to reduce costs for industrial-scale application. Other researchers are working on approaches that include adding nanoparticles to a molten salt or an ionic liquid to increase the material’s specific heat capacity, and studying ways to store both hot and cold salt in the same tank. Whichever storage technology—or combination of technologies—eventually pans out, it will be critical to the future of CSP. The current global capacity of solar thermal power plants is minuscule: just over 1 gigawatt, about the power output of one large fossil-fuel or nuclear power station. Another 15 gigawatts are currently in development or under construction in the United States, Spain, North Africa, China, India, and elsewhere, according to the International Energy Agency. If CSP is to become a major player in the future, the ability to store some of the sun’s energy and then use it to generate electricity when needed will be essential.
18 NOVEMBER 2011 VOL 334 SCIENCE www.sciencemag.org Published by AAAS
–EDWIN CARTLIDGE
Downloaded from www.sciencemag.org on November 17, 2011
avoiding undue stress inside the material as a result of thermal expansion. The researchers say a storage system with the same energy capacity as the molten-salt tanks at the Andasol plant would require some 250 concrete blocks, each 200 cubic meters and weighing 400 tons, spread out over an area equal to about five football pitches.
CREDIT: GEMASOLAR PLANT, OWNED BY TORRESOL ENERGY © TORRESOL ENERGY
feasible,” Gill says, although the cost is hard to gauge because both contain the economically volatile element lithium.
The next time you groan when it’s time to mow your lawn, take a second first to marvel at a blade of grass. Plants are so commonplace that it’s easy to take their wizardry for granted. When they absorb sunlight, they immediately squirrel away almost all of that energy by using it to knit together a chemical fuel they use later to grow and multiply. It sounds so simple. Yet it’s anything but. Modern society runs on fossil fuels precisely because researchers have never managed to duplicate the chemical mastery of a fescue. Now, with the side effects of our massivescale use of fossil fuels piling up (climate change, acidified oceans, oil spills, and so on), researchers around the globe are struggling to play catch-up with biology in hopes of harnessing the sun’s energy to synthesize gasoline or other fuels that are the bedrock of modern society. Humans, of course, already have ways to capture solar energy. Today’s photovoltaic solar cells typically trap 10% to 20% of the energy in sunlight and convert it to electricity, and PV prices continue to drop. But because electricity is difficult to store on a large scale, the effort to store sunlight’s energy in chemical fuels has risen to one of the grand challenges of the 21st century. “You’re talking about turning the energy world on its head. Today we turn hydrocarbon fuels into electricity. But in the future, we need to find a way to turn electricity [from sunlight] into fuels,” says Daniel DuBois, a chemist at the Pacific Northwest National Laboratory in Richland, Washington.
The problem is daunting. Energy production is the world’s largest enterprise. Today the world consumes power at an average rate of 17.75 trillion watts, or 17.75 terawatts, 85% of which starts out as fossil fuels, coal, oil, and natural gas. Thanks to rising populations and incomes, by 2050 the world’s demand for power is expected to at least double. To keep fossil fuels from stepping in to fill that need, with potentially devastating side effects, any new solar fuels technology will have to provide power just as cheaply, and it must have the potential to work on an equally massive scale. Enter artificial photosynthesis. Researchers around the globe are working to combine materials that capture sunlight with catalysts that can harness solar energy to synthesize fuels. This dream has been pursued for
The splits. An artificial leaf harnesses energy in
sunlight to split water into oxygen and hydrogen.
decades. But recent strides are adding new zip to the field. “In the last 5 to 10 years, there has been amazing progress,” DuBois says. Anthony Rappé, a chemist at Colorado State University, Fort Collins, agrees. However, he adds, “the bottom line is we’re not there yet.” Molecular shuffle To get there, most artificial photosynthesis researchers look to natural photosynthesis for inspiration. During photosynthesis, plants absorb sunlight, water, and CO2. Then they use two protein complexes—called photosystem I and II—to split water and synthesize fuel. First, in photosystem II, energy in sunlight splits two water molecules into four hydrogen ions (H+), four electrons, and a molecule of oxygen (O2). The O2 wafts away as waste; the protons and electrons are sent to photosystem I and used to energize the coenzyme NADP to NADPH, which in turn is used to help synthesize sugars—a key series of metabolic steps. Of course, artif icial photosynthesis researchers aim to make fuel not for plants but for planes, trains, and automobiles. So after splitting water into H+, electrons, and oxygen molecules, most make very different use of those ingredients. Some researchers are working to combine the protons and electrons with carbon dioxide (CO2) to make methane gas and other hydrocarbon fuels (see sidebar, p. 927). But most are working on what they believe is a simpler approach: combining the pieces they get from splitting pairs of water molecules into molecules of O2 and hydrogen gas (H2). That H2 can then either be burned in
www.sciencemag.org SCIENCE VOL 334 18 NOVEMBER 2011 Published by AAAS
Downloaded from www.sciencemag.org on November 17, 2011
CREDITS: (TOP MAIN) FOTOSEARCH; (TOP SUPERIMPOSED AND BOTTOM) KIMBERLY SUNG/SUN CATALYTIX
SPECIALSECTION
925
Box step. Natural photosynthesis depends on a molecular cube (left) made
from manganese atoms (purple), oxygens (red), and a calcium atom (yellow).
tively charged electron vacancy called a hole. The holes are shuttled over to a compound called the oxygen-evolving complex, which grabs two oxygen atoms, holds them close together, and rips out an electron from each to fill the holes. The electron-deficient oxygens regain their stability by combining to form O2. In an artificial system, the electrons and protons liberated by water splitting then must migrate to a second catalyst, which combines them into two molecules of H2. A successful artificial photosynthesis system must therefore meet several demands. It must absorb photons, use the energy to create energized electrons and holes, and steer those charges to two different catalysts to generate H2 and O2. It also has to be fast, cheap, and rugged. “This is a much more stringent set of requirements than [those for] photovoltaics,” says John Turner, a water-splitting expert at the National Renewable Energy Laboratory (NREL) in Golden, Colorado.
926
ple is the quest for H2-forming catalysts. Natural photosynthesis carries out the reaction using enzymes called hydrogenases, which are built from the abundant elements iron and nickel. The enzymes have evolved until they can knit roughly 9000 pairs of hydrogen atoms into molecular H2 every second. Many early water-splitting systems performed the same reaction even faster using pure platinum as the catalyst. But platinum is too rare and expensive to be broadly useful. In recent years, researchers have synthesized numerous compounds aimed at mimicking the core complex of hydrogenases. All work more slowly (if at all), however, largely because they lack parts of the natural protein around the core that optimizes the core’s activity. In 2008, Thomas Rauchfuss, a chemist at the University of Illinois, Urbana-
The catalyst splits water molecules (blue), generating molecular oxygen. Synthetic versions (center and right) have similar cube-shaped cores.
Unfortunately, these systems, too, had drawbacks. The metals in the best lightabsorbing molecular dyes are too rare to be viable as a large-scale technology. To get enough ruthenium to power the world with water splitting, “we would need to harvest 1% of the Earth’s total continental crust to a depth of 1 kilometer,” Rappé says. Scale-up is problematic with the semiconductor system as well. Although Turner’s devices convert 12% of sunlight to hydrogen, the materials would cost as much as $50,000 per square meter, according to an estimate by Harry Gray, a chemist at the California Institute of Technology (Caltech) in Pasadena. To be viable on a large scale, “we need to build something this good for $100 per square meter,” Gray says. Wanted: the perfect catalyst So more recently, much of the work in the water-splitting field has begun to shift to trying to make light collectors and catalysts from abundant and cheap materials. A prime exam-
Champaign, and colleagues devised catalysts with molecular arms that act like a bucket brigade to ferry protons to the catalytic core. In the 12 August 2011 issue of Science (p. 863), DuBois and his colleagues described how they had refined this strategy further by creating a nickel-based catalyst that stitches 106,000 H2 molecules together every second (http://scim. ag/_DuBois). The new H2 makers still aren’t ideal. They work only at high speed when researchers apply an electrical voltage of more than 1 volt to their system, a sizable energetic penalty. So DuBois’s team is now working to tweak the catalysts to work at a lower added voltage. In a paper published online in Science on 29 September (http://scim.ag/Nocera), Dan Nocera, a chemist at the Massachusetts Institute of Technology in Cambridge, reported that he and his colleagues had come up with another H2 catalyst that works with an extra voltage of only 35 thousandths of a volt (millivolts). It, too, is made from rela-
18 NOVEMBER 2011 VOL 334 SCIENCE www.sciencemag.org Published by AAAS
Downloaded from www.sciencemag.org on November 17, 2011
In 1972, Japanese researchers took on the challenge by using particles of titanium dioxide to split water. The method was impractical for commercial use because TiO2, which absorbs only ultraviolet light, could make no use of 95% of the solar spectrum. But the demonstration inspired numerous other water-splitting systems. One setup uses molecular dyes made with ruthenium and other rare metals to absorb a variety of wavelengths of light and pass the charges to metal catalysts. Another, developed by Turner’s NREL team, absorbed light with semiconductor wafers made from gallium arsenide (GaAs) and gallium indium phosphide (GaInP). A platinum electrode served as the catalyst to split water and generate O2, while the semiconductor acted as the electrode to produce H2.
CREDITS (LEFT TO RIGHT): Y. UMENA, K. KAWAKAMI, J.-R. SHEN AND N. KAMIYA; NOCERA LAB AT MIT; DAVID ROBINSON
an engine or run through a fuel cell, where the water-splitting reaction runs in reverse: combining two H2s with O2 from the air to generate water and electricity. Although plants split water with seeming ease, it’s not a simple task, and it requires electrons to perform an intricate quantummechanical dance. Quantum mechanics dictates that electrons can exist only at discrete energy levels—or “bands.” In semiconductors, for example, electrons can sit in either a lower energy state known as the valence band, where they are closely bound to the atom on which they sit, or a more freewheeling energized state in the conduction band. Molecules like chlorophyll in plants act like tiny semiconducting proteins. When they absorb sunlight, they kick an electron from the valence to the conduction band, leaving behind a posi-
SPECIALSECTION Sunlight in Your Tank—Right Away Using sunlight to split water and generate hydrogen doesn’t make the most useful chemical fuel. To use hydrogen on a large scale, societies would have to develop a new infrastructure to store, transport, and distribute the energy carrier. With that limitation in mind, some researchers are looking to use artificial photosynthesis to generate hydrocarbon fuels like those we already burn. Their goal is essentially to run combustion in reverse, starting with carbon dioxide (CO2) and water and using the energy in sunlight to knit the chemical bonds needed to make hydrocarbons, such as gaseous methane and liquid methanol. “That’s a technology that’s going to come,” says Harry Gray, a chemist at the California Institute of Technology in Pasadena. “But it is hard.” The difficulty is that CO2 is a very stable molecule. In converting CO2 to hydrocarbons, the first step is to strip off one of the oxygen atoms, leaving behind a molecule of carbon monoxide (CO), a more reactive combination of carbon and oxygen. CO can then be combined with molecular hydrogen and converted into liquid hydrocarbons using an industrial process known as Fischer-Tropsch synthesis. That first step of converting CO2 to CO is the energy hog. A minimum of 1.33 electron volts (eV) of energy must be applied to carry out the reaction. Over the past few decades, researchers have developed numerous catalysts that carry out the process. But virtually all of them require adding a lot of extra energy, typically another 1.5 eV. As a result, it would take far more energy to synthesize a hydrocarbon fuel than the fuel’s molecules could store in their chemical bonds. On 29 September, however, researchers led by Richard Masel of Dioxide Materials in Champaign, Illinois, and Paul Kenis of the University of Illinois, Urbana-Champaign, reported online in Science (http://scim.ag/_Masel) that they’ve come up with a less energy-intensive way to convert CO2 to CO. By adding a type of solvent called an ionic liquid to the CO2 in their setup, they reduced the added energy needed for splitting CO2 by 90%. Ionic liquids are liquid salts that are adept at stabilizing negatively charged compounds. Adding a negative charge is the first step required to convert CO2 to CO; the Illinois researchers suspect the increased stability reduces the voltage needed to do the job. The Illinois catalysts are slow, and so far the researchers have not powered them with electrical charges from a solar cell. But other labs are taking an approach that looks more like full-fledged artificial photosynthesis. At Lawrence Berkeley National Laboratory in California, for example, chemist Heinz Frei and his colleagues reported in 2005 that for the first time they had used energy from visible light to convert CO2 to CO using a porous catalyst made from silica and impregnated with zinc and copper. Frei’s team has used related catalysts to split water to generate molecular hydrogen. Now the group is working to put the two pieces together to combine light-generated CO and H2 to make methanol, one of the simplest hydrocarbons. It’s not ExxonMobil yet. But with further developments, the technology could lead to fuels made basically from air, water, and sunlight. –R.F.S.
low-energy green photons and reemitting it as lower numbers of higher energy blue photons. They are now working on using this upconversion process to make use of more of the solar spectrum to split water. Researchers led by Steve Cronin of the University of Southern California in Los Angeles are adding metal nanoparticles to conventional solar absorbers as another way to convert low-energy photons to electrical charges that can then be harnessed to improve the efficiency of watersplitting setups. And Gray’s group at Caltech has teamed up with students at 17 other universities to create a “solar army” that has already made progress in finding new watersplitting catalysts. These and other advances will need to continue if artificial photosynthesis ever hopes
to contend with fossil fuels. With today’s low natural gas prices, companies can use a mature technology called steam reforming to convert natural gas to hydrogen at a cost of about $1 to $1.50 per kilogram of H2 generated, which contains about the same amount of energy as a gallon of gasoline. Yet a recent analysis by Turner and his colleagues showed that, even if researchers could create an artificial photosynthesis system that cost $200 per square meter for the equipment and was 25% efficient at converting sunlight to H2, the H2 would still cost $2.55 per kilogram. That’s not saying artificial photosynthesis isn’t worth pursuing—only that fossil fuels are the leading energy source for a reason, and they won’t be easy to dethrone.
www.sciencemag.org SCIENCE VOL 334 18 NOVEMBER 2011 Published by AAAS
Downloaded from www.sciencemag.org on November 17, 2011
tively cheap metals: molybdenum, nickel, and zinc. But Nocera’s catalyst is slower than DuBois’s, so the race is on to marry the best attributes of each. Balancing speed and extra energy input has been an even tougher problem with the catalysts needed for other reactions in water splitting, which grabs oxygen atoms from two water molecules and links them together as O2. In 2008, Nocera and his team made headlines when they unveiled a cobalt-phosphate (Co-Pi) catalyst that works at 300 millivolts applied potential over the minimum 1.23 electron volts required to link two oxygen atoms. The group followed that up with a nickelborate compound that does much the same thing. And in the 29 September online paper, the researchers described a triple-layer silicon wafer lined with their Co-Pi catalyst on one face and with their H2 catalyst on the other. The silicon absorbed sunlight and passed charges to the two catalysts, which then split water. “I love the triple junction. It’s pretty sexy,” says Felix Castellano, a chemist at Bowling Green University in Ohio. Turner cautions that the overall efficiency of the device—it converts just 5% of the energy in sunlight to hydrogen—is still too low, and the extra voltage input required is still too high, to be commercially useful. Nocera counters that this initial system was built using amorphous silicon wafers as the sunlight absorbers. Such wafers are only 8% efficient in converting light to electrical charges. An artificial leaf based on crystalline silicon solar cells, which are 20% efficient, could convert sunlight to chemical energy with an efficiency of 12%, he says. But Nocera’s team has yet to demonstrate such a device. Other related catalysts are also entering the picture. Charles Dismukes, a chemist at Rutgers University in Piscataway, New Jersey, and colleagues reported last year that they had made a series of O2-forming catalysts using lithium, manganese, and oxygen. And earlier this year, Dismukes’s team reported in the Journal of the American Chemical Society that they had created another oxygen-forming complex with cobalt and oxygen. What’s unique about all these new oxygen formers is that they share almost an identical cubic molecular structure, which is also at the heart of the natural O2-forming complex in photosystem II. “There is only one blueprint from biology that can be copied,” Dismukes says. Many other advances are also making their way out of the lab. Castellano and colleagues have recently created a family of cheap polymers capable of absorbing the energy from
–ROBERT F. SERVICE
927
Electrical Energy Storage for the Grid: A Battery of Choices Bruce Dunn,1 Haresh Kamath,2 Jean-Marie Tarascon3,4 The increasing interest in energy storage for the grid can be attributed to multiple factors, including the capital costs of managing peak demands, the investments needed for grid reliability, and the integration of renewable energy sources. Although existing energy storage is dominated by pumped hydroelectric, there is the recognition that battery systems can offer a number of high-value opportunities, provided that lower costs can be obtained. The battery systems reviewed here include sodium-sulfur batteries that are commercially available for grid applications, redox-flow batteries that offer low cost, and lithium-ion batteries whose development for commercial electronics and electric vehicles is being applied to grid storage.
1 Department of Materials Science and Engineering and California NanoSystems Institute, University of California, Los Angeles, Los Angeles, CA 90095, USA. 2Electric Power Research Institute (EPRI), Palo Alto, CA 94304, USA. 3Université de Picardie Jules Verne, Laboratoire de Réactivité Chimie des Solides, Amiens 80039, France. 4Collège de France, Paris 75231, France.
928
UPS Power quality
T & D grid support Load shifting
Bulk power management
Hours
Pumped hydro Flow batteries: Zn-Cl, Zn-Br Vanadium redox New chemistries NaS battery
Compressed air Energy storage
Advanced lead-acid battery
High-energy supercapacitors Minutes
Discharge time at rated power
T
pumped hydroelectric storage. This is far below the energy storage levels in Europe (10%) and Japan (15%), where more favorable economics and policies are in place (2). Energy storage technologies available for large-scale applications can be divided into four types: mechanical, electrical, chemical, and electrochemical (3). Pumped hydroelectric systems account for 99% of a worldwide storage capacity of 127,000 MW of discharge power. Compressed air storage is a distant second at 440 MW. The characteristics for several of these EES systems in terms of power rating, which identifies potential applications, and duration of discharge are illustrated in Fig. 1. Potential grid applications range from frequency regulation and load following, for which short response times are needed,
NaNiCl2 battery
Li-ion battery Lead-acid battery NiCd NiMH
Seconds
he August 2003 blackout in the Northeast and the recent September 2011 power failure that extended from Southern California to Mexico and Arizona are two of the more widely publicized examples in which power outages affected many millions of consumers. From a broader perspective, such power outage events underscore the complex set of issues associated with the generation and use of electricity: the reliability of the grid, the use of fossil fuels and related carbon emissions, the development of electric vehicles to decrease dependence on foreign oil, and the increased deployment of renewable energy resources. Underlying these considerations is the recognition that inexpensive and reliable energy is vital for economic development. Moreover, most of these issues are international in scope, with the additional caveat that worldwide demand for electricity is projected to double by 2050. Electrical energy storage (EES) cannot possibly address all of these matters. However, energy storage does offer a well-established approach for improving grid reliability and utilization. Whereas transmission and distribution systems are responsible for moving electricity over distances to end users, the EES systems involve a time dimension, providing electricity when it is needed. A recent study identified a number of high-value applications for energy storage, ranging from the integration of renewable energy sources to power quality and reliability (1). Despite the anticipated benefits and needs, there are relatively few storage installations in operation in the United States. Only ~2.5% of the total electric power delivered in the United States uses energy storage, most of which is limited to
High-power flywheels High-power supercapacitors 1 kW
10 kW
100 kW
1 MW
10 MW
100 MW
1 GW
System power ratings, module size
Fig. 1. Comparison of discharge time and power rating for various EES technologies. The comparisons are of a general nature because several of the technologies have broader power ratings and longer discharge times than illustrated (1). [Courtesy of EPRI]
18 NOVEMBER 2011
VOL 334
SCIENCE
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
to peak shaving and load shifting, both of which can lead to improvements in grid reliability, stability, and cost (4). The electric power profile shown in fig. S1 indicates how storage can integrate renewable resources and be used to accommodate peak loads. Load shifting represents one of the more tantalizing opportunities for EES because of the benefit in storing energy when excess power is generated and releasing it at times of greater demand. The technical requirements, however, are quite rigorous. As indicated in Fig. 1, there are several energy storage technologies that are based on batteries. In general, electrochemical energy storage possesses a number of desirable features, including pollution-free operation, high round-trip efficiency, flexible power and energy characteristics to meet different grid functions, long cycle life, and low maintenance. Batteries represent an excellent energy storage technology for the integration of renewable resources. Their compact size makes them well suited for use at distributed locations, and they can provide frequency control to reduce variations in local solar output and to mitigate output fluctuations at wind farms. Although high cost limits market penetration, the modularity and scalability of different battery systems provide the promise of a drop in costs in the coming years. Today, sodium/sulfur (Na/S) battery technology is commercially available for grid applications, with some 200 installations worldwide, accounting for 315 MW of discharge power capacity. Moreover, there are emerging opportunities for other battery systems because of potential low cost (redox-flow) and enhanced performance [lithium (Li)–ion]. In this Review,
REVIEW
SPECIALSECTION we present some of the overarching issues facing the integration of energy storage into the grid and assess some of the key battery technologies for energy storage, identify their challenges, and provide perspectives on future directions.
peak and baseload generation (fig. S1), allowing loads at any time to be serviced by the lowest cost energy resources (6). Storage solutions based on the technologies we have today are so expensive that historically it has been far more cost-effective to expand generation as well as transmission and distribution to serve the peak load and provide sufficient operating margin to meet consumer demands for reliability. In those cases in which storage is used, pumped hydroelectric plants are generally involved. These plants are composed of lowcost materials (dirt, concrete, and water) that have a lifetime of over 40 years, minimal maintenance costs, and relatively high round-trip ef-
energy (7). Batteries, regardless of their chemistry— aqueous, nonaqueous, Li or Na-based—store energy within the electrode structure through charge transfer reactions. By comparison, fuel cells, which are not rechargeable, store energy in the reactants that are externally fed to the cells. Both of these differ from redox-flow cells, which store energy in the redox species that are continuously circulating through the cells. Supercapacitors offer yet a different energy storage mechanism, via a capacitive process arising from an electrochemical double layer at the electrode-electrolyte interface. Each mechanism has different strategies that can be used to improve the power and energy densities of the EES approach.
www.sciencemag.org
SCIENCE
VOL 334
18 NOVEMBER 2011
Downloaded from www.sciencemag.org on November 17, 2011
Utility Perspective on Energy Storage EES has often been described as the “Holy Grail” of the electric utility industry. This phrase evokes the eagerness of utilities and other stakeholders to achieve cost-effective storage options, which could potentially cure many of the ills faced by the electric power enterprise. However, the phrase Holy Grail also suggests that the search for energy storage will be long, difficult, and perilous. We are unlikely to find, at least in the near term, a single technology Table 1. Energy storage for utility transmission and distribution grid support. The megawatt- and kilowatt-scale energy that can repeatedly and efficiently storage systems listed here have potential impact in several areas, including transmission and distribution substation store large quantities of electric grid support, peak shaving, capital deferral, reliability, and frequency regulation (1). [Courtesy of EPRI] energy at low cost. On the other Capacity Power Duration % Efficiency Total cost Cost hand, a portfolio approach that is Technology option Maturity (MWh) (MW) (hours) (total cycles) ($/kW) ($/kWh) based on using a combination of Demo 250 50 5 (>10,000) 1950–2150 390–430 technologies may be the most ef- CAES (aboveground) fective means to introduce and inAdvanced Demo 3.2–48 1–12 3.2–4 75–90 2000–4600 625–1150 tegrate energy storage. Pb-acid (4500) The usefulness of EES stems Commercial 7.2 1 7.2 75 3200–4000 445–555 from the operational character- Na/S (4500) istics of the grid as a supply chain Demo 5–50 1–10 5 60–65 1670–2015 340–1350 of a commodity, electric power. Zn/Br flow (>10,000) At present, the electric power inDemo 4–40 1–10 4 65–70 3000–3310 750–830 frastructure functions largely as V redox (>10,000) a just-in-time inventory system R&D 4 1 4 75 1200–1600 300–400 in which a majority of energy is Fe/Cr flow (>10000) generated and then transmitted R&D 5.4 1 5.4 75 1750–1900 325–350 to the user as it is consumed. Zn/air (4500) Without the ability to store energy, Demo 4–24 1–10 2–4 90–94 1800–4100 900–1700 there must be sufficient generation Li-ion (4500) capacity on the grid to handle peak demand requirements, despite the likelihood that much of Although not discussed here, capacitive enerthat capacity sits idle daily as well as for large ficiency (between 65 and 75%). Although there portions of the year (fig. S2). Correspondingly, are obvious limitations because of geographical gy storage offers some promising opportunities for the transmission and distribution system must considerations, pumped hydro will be the bench- grid-scale applications (Fig. 1). Supercapacitors provide higher power and longer cycle life than also be sized to handle peak power transfer re- mark for grid-scale storage for years to come. In the near term, utilities are aware of the that of batteries and are receiving renewed attenquirements, even if only a fraction of that power transfer capacity is used during most of the rising need for EES solutions but are skeptical of tion as researchers try to better understand fundayear. Operationally, electrical power generation the technologies that have been proposed to date. mental interfacial processes and improve energy must be continuously ramped up and down to Even in cases in which technology has substan- density (8). The technology is of interest for power ensure that the delicate balance between supply tial merit, the absence of cost-effective products quality applications, such as alleviating short-term and demand is maintained. The up and down with a track record of safe and reliable operation disruptions of a few minutes until a generator, cycling reduces power plant efficiency, resulting has made the industry skittish about their use. fuel cell, or battery can be placed in service. Bein higher fuel consumption and higher emissions Table 1 lists some of the current maturity levels cause the lifetime costs for supercapacitors can per kilowatt-hour (kWh) produced. This proce- for various energy storage technologies, their be attractive (6), there is the prospect that this dure also causes more wear on the equipment operational characteristics, and cost estimates. If technology will be used in conjunction with batsuccessful, the outcomes from these projects may teries so as to provide future grid storage solutions. and reduces the lifetime of power plants (5). A battery is composed of several electrochemBy decoupling generation and load, grid en- alleviate industry concerns of matters such as perergy storage would simplify the balancing act formance, cycle life, economics, and risks. Another ical cells that are connected in series and/or in between electricity supply and demand, and on promising development is that the industry has parallel in order to provide the required voltage and capacity, respectively. Each cell is composed overall grid power flow. EES systems have po- begun working to establish standards and targets. of a positive and a negative electrode, which are tential applications throughout the grid, from where the redox reactions take place. The elecbulk energy storage to distributed energy func- Electrochemical Energy Storage tions (1). The availability of energy storage Electrochemical energy storage approaches can trodes are separated by an electrolyte, usually a would help to eliminate the distinction between be distinguished by the mechanisms used to store solution containing dissociated salts so as to
929
930
300 200
Sodiummetal chloride
Nickelmetal hydride
Li metalpolymer
Sodium-sulfur V redox flow
0
50
100
150
200
Specific energy (W h kg-1)
Fig. 2. Gravimetric power and energy densities for different rechargeable batteries. Most of these systems are currently being investigated for grid storage applications.
A e
e
Cathode
Anode
-
+
Electrolyte
e
e
e
e
e e e e e
Cu current collector
Al current collector Graphene structure
Li+
Solvent molecule
LiMO2 layer structure
Fig. 3. Schematic of a LIB. The negative electrode is a graphitic carbon that holds Li in its layers, whereas the positive electrode is a Li-intercalation compound—usually an oxide because of its higher potential— that often is characterized by a layered structure. Both electrodes are able to reversibly insert and remove Li ions from their respective structures. On charging, Li ions are removed or deintercalated from the layered oxide compound and intercalated into the graphite layers. The process is reversed on discharge. The electrodes are separated by a nonaqueous electrolyte that transports Li ions between the electrodes. [Derived from (4)]
18 NOVEMBER 2011
VOL 334
SCIENCE
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
100
Lead acid Nickelcadmium
0
Lithium Ion Batteries The Li-ion battery (LIB) technology commercially introduced by Sony in the early 1990s is based on the use of Li-intercalation compounds. Li ions migrate across the electrolyte located between the two host structures, which serve as the positive and negative electrodes (Fig. 3). Liion batteries outperform, by at least a factor of 2.5, competing technologies [nickel (Ni)–metal hydride, Ni-cadmium (Cd), and lead (Pb)–acid)] in terms of delivered energy while providing high specific power (Fig. 2). The overwhelming appeal of Li-electrochemistry lies in its low molecular weight; small ionic radius, which is beneficial for diffusion; and low redox potential [E°(Li+/Li) = −3.04 V vs standard hydrogen electrode (SHE)] (11). The latter enables high-output voltages and therefore high-energy densities. Such attractive properties, coupled with its long cycle life and rate capability, have enabled Li-ion technology to capture the portable electronics market and make in-roads in the power tools equipment field. LIBs are also regarded as the battery of choice for powering the next generation of hybrid electric vehicles (HEVs) as well as plug-in hybrids (PHEVs), provided that improvements can be achieved in terms of performance, cost, and safety (12). Because long-term stability, high-energy density, safety, and low cost are common to developing batteries for both automotive and grid applications, considerable synergy should exist between the two areas, although there will be certain differences. Figures of merit for electric vehicle applications call for a reduction in the price per kilowatt-hour by a factor of 2 and a doubling of the present energy density. The realization of such goals will be beneficial for grid storage systems, although with probably more emphasis on cost and less on energy density. Other differences between the two technologies include safety, which is easier to achieve in stationary situations than in mobile ones, whereas long cycle life is a key factor for grid applications. LIBs for vehicles require versatility in their energy and power capabilities in order to meet the needs of the various types of electric vehicles and the associated performance requirements, whereas LIBs for the grid are likely to be modular. A number of advances have been made in the LIB field by controlling particle size in addition to composition, structure, and morphology in order to design better electrodes and electrolyte components (13). Decreasing electrochemically active materials to sub-micrometer and smaller
Li ion Li(TM)O2-C
Li ion LiFePO4-C
Specific power (W kg-1)
enable ion transfer between the two electrodes. Once these electrodes are connected externally, the chemical reactions proceed in tandem at both electrodes, liberating electrons and providing the current to be tapped by the user (9, 10). The energy storage properties for most of the common rechargeable batteries are shown in Fig. 2, with additional details provided in table S1.
SPECIALSECTION ly promising and suggests that the performance of organic electrodes could become comparable in gravimetric energy density, life cycle, and power rate to today’s best inorganic electrodes, with the distinct advantage of providing a botanic alternative to the mineral approach currently in practice. At the research level, there is interest in rechargeable LIB systems that have significantly higher energy densities (22, 23). Although the Li-O2 system has been available for many years as a primary battery, the prospect of developing it into a reversible (secondary) battery has become tantalizing because of a projected three- to fourfold increase in gravimetric energy density as compared with the current Li-ion technology (24).
little doubt that rechargeable Li-air cells either for electric vehicles or grid storage applications still have a long research and development path. The prospect of developing Li-ion technology for both transportation and stationary storage raises the issue of whether the demand for lithium will affect the existing world reserves. Na is an attractive alternative because its intercalation chemistry is similar to that of Li, there are ample reserves, and its cost is low. These advantages are partially offset by the gravimetric energy density penalty for using Na, which is both heavier and less electropositive than Li. The development of room-temperature Na-ion cells that are costeffective, sustainable, and environmentally benign will require a new generation of Na-intercalation
e-
Downloaded from www.sciencemag.org on November 17, 2011
sizes combined with carbon-coating approaches to achieve core-shell morphologies has led to new directions in electrode materials (14). Reaction mechanisms and materials systems that were previously discarded are being reconsidered for the next generation of LIBs. Moving from bulk materials to nanosize particles has enabled (i) the ability to use new Li-reaction mechanisms, in which conversion-reaction electrodes show enormous capacity gains (15); (ii) the use of negative electrodes based on alloy reactions—Tin (Sn)–based LIB technologies have already reached the marketplace (such as NEXELION), and Sibased ones are emerging (16); (iii) the identification of poorly conducting polyanionic compounds or fluorine-based compounds that exhibit excellent electrochemical performance (17); and (iv) the transformation of the poorly conducting lithium iron phosphate (LiFePO4) insertion electrode into perhaps the most valued electrode material for electric vehicle applications (18). LIBs based on LiFePO4 are extremely attractive because of safety and cost. The former arises from the fact that the operating voltage of the LiFePO4 system is compatible with the thermodynamic stability of the electrolyte, whereas the latter is based on the use of abundant and low-cost constituents. In addition to being an attractive LIB for the electric vehicle market, LiFePO4-based batteries are being evaluated in stationary energy storage demonstration projects (1). A substantial segment of the battery materials community is moving toward developing electrode materials on the basis of abundance and availability of the relevant chemicals. Materials centered on sustainable 3d metal redox elements such as manganese (Mn) [lithium-manganese oxide (LiMn2O4)], Fe (LiFePO4, Li2FeSiO4) and titanium (Ti) (TiO2, Li4Ti5O12), and made via eco-efficient processes, are receiving increased attention (19). In addition, there is resurging interest in low-temperature–solution chemistry routes in which hydro(solvo)thermal, ionothermal, and bio-mineralization processes are used to prepare electrode materials at temperatures >500°C lower than traditional powder synthesis (20). Life cycle costs represent another important consideration. A foreseeable strategy for battery processing will involve the use of electro-active organic electrode materials synthesized from “green chemistry” concepts through low-cost processes free of toxic solvents; this will also enlist the use of natural organic sources [carbon dioxide (CO2)–harvesting entities] as precursors, which will be biodegradable and easily destroyed by combustion (providing CO2) so that the battery assembly/recovery processes will have a minimal CO2 footprint. Proof of this concept was demonstrated with the development of renewable organic electrodes belonging to the family of oxocarbons (Li2C6O6) or carboxylates (Li2C8H4O4) and the assembly of the first eco-compatible LIB laboratory prototype (21). This work is extreme-
e-
-
+ O2
Li+
Discharge O2 +
Li
Lithium
Electrolyte Li2O2
Catalyst
Composite electrode Carbon
Fig. 4. The Li-air cell uses Li as the anode and a cathode consisting of a porous conductive composite, usually carbon and a catalyst, that is flooded with electrolyte. Oxygen from the atmosphere dissolves in the electrolyte and is reduced. On discharge, Li ions pass through the electrolyte and react with the reduced oxygen. The process is reversed on charging. Either aqueous or nonaqueous electrolytes can be used. For the former, a Li-ion–conducting solid electrolyte separates the metallic Li from the aqueous electrolyte. However, the volumetric energy density may not be much greater than that of Li-ion batteries (25). The renewed interest in this system can be traced to the rechargeable behavior demonstrated in a nonaqueous Li-O2 system (Fig. 4) (26). Although there has been considerable progress in the past 5 years in the area of electrode materials and architectures (27, 28), a number of fundamental problems still need to be addressed, and it is difficult to anticipate which of the advanced Li-O2 aqueous and/or Li-O2 nonaqueous systems will be able to achieve capabilities beyond today’s Li-ion batteries (29). Thus, there is
www.sciencemag.org
SCIENCE
VOL 334
compounds (30). The knowledge gained from developing Li-ion insertion electrodes should be applicable here. Thus, the demonstration of a viable Na-ion technology for stationary energy storage should come well before that of Li-air technology because of the accumulated experience with Li-ion technology and high-temperature Na battery technologies. Sodium-Sulfur and Sodium-Metal Halide Batteries High-temperature Na-based battery technologies can be traced back to the 1960s, when researchers
18 NOVEMBER 2011
931
932
Beta alumina tube
Beta alumina
-
V
+ Sulfur electrode
Sulfur container Molten Na
Discharge +
Na
Na+
Charge
Protection layer
Discharge
Fig. 5. Schematic of the Na/S battery. The central Na design has molten Na (negative electrode) contained within a Na b″-alumina solid electrolyte tube with molten S (positive electrode) surrounding the tube. The S electrode includes carbon in order to provide sufficient electronic conduction to carry out the electrochemical reactions. The magnified cross section of the cell shows the direction of Na+ transport through the b″-alumina electrolyte. On discharge, Na combines with the S to form Na polysulfides. These reactions are reversed during charge, and Na returns to the interior of the tube. in which “blocks” of closely packed Al-O are separated by “conduction planes” (35). The latter are loosely packed layers that contain the mobile Na+ along with O2– ions that bridge adjacent blocks. Ion motion occurs in two-dimensional honeycomb-like pathways around the bridging oxygen. The polycrystalline b″-alumina tubes used in the Na/S and Na/MeCl2 batteries do not exhibit the anisotropic transport properties of single crystals because the fine-grained, randomly oriented microstructures effectively eliminate the anisotropy. Nonetheless, there are grain boundary and tortuosity effects so that the conductivity of single-crystal Na b″-alumina at 300°C, ~1 S cm−1, is three to five times greater than the corresponding polycrystalline material (32). A recent study suggests that tortuosity effects can be diminished because Na b″-alumina electrolytes in a planar configuration exhibit higher ionic conductivity than that of tubular materials (36). From inception, development for both systems targeted stationary energy storage and electric vehicles. As a result, the technologies share a number of common features (and challenges), even though specific designs differ somewhat. In both cases, the b″-alumina ceramic tubes are acknowledged to be the key element for determining battery operation and cost. Considerable development effort has gone into establishing large-scale manufacturing processes for automating the fabrication of high-quality ceramics with appropriate mechanical and electrical properties (37). Fracture of the ceramic is
18 NOVEMBER 2011
VOL 334
SCIENCE
a vital concern because it leads to cell failure, whereas poor control of the ceramic microstructure results in interfacial reactions with the reactants. Large-scale production of b″-alumina has been established, but production yields and costs are major concerns (38). Other critical battery components are seals, which must not only be hermetic in the 300 to 350°C range but also withstand the vapor and/or actual contact with the highly reactive molten electrode materials. Recent activities in this area have involved the development of glass-ceramic sealing materials whose thermal expansion coefficient matches that of a- and b-alumina components (39). There is also the issue of identifying a low-cost material for containing the molten positive electrode. The corrosion problem is particularly difficult for Na/S batteries because both S and polysulfides are highly corrosive. The deposition of corrosionresistant coatings such as carbides onto inexpensive substrates has proven successful (40). Na/S battery technology has been commercialized in Japan since 2002, where it is largely used in utility-based load-leveling and peakshaving applications. Among the advantages identified for stationary storage are its relatively small footprint (a result of high energy density), high coulombic efficiency, cycling flexibility, and low maintenance requirements (41). The production of megawatt-size energy storage batteries has involved considerable effort on such interrelated issues as electrical networking, cell reliability, thermal management, and safety (42).
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
at Ford discovered that a common ceramic refractory, sodium b-alumina (NaAl11O17), exhibited extremely high ionic conductivity for Na ions (31). At 300°C, the ionic conductivity for NaAl11O17 approaches that of the aqueous electrolyte, H2SO4, suggesting the possibility of using NaAl11O17 as a solid electrolyte in a hightemperature electrochemical cell. Although solids with high ionic conductivity had been known previously, none had b-alumina’s combination of chemical and thermal stability and low electronic conductivity. The recognition that inorganic materials with high vacancy concentrations could exhibit “fast ion conduction”—many orders of magnitude greater than traditional alkali halides—led to the development of the field known as solid-state ionics. The two high-temperature Na batteries, Na/S and Na-metal chloride (Na/MeCl2), are based on using b-alumina as a Na+-conducting membrane between two liquid electrodes (32). The batteries operate at temperatures of 270 to 350°C so as to take advantage of the increased conductivity of the b-alumina at elevated temperatures and ensure that the active electrode materials are molten. During discharge in the Na/S battery, Na is oxidized at the solid electrolyte interface, and the resulting Na+ migrates through the electrolyte to react with S that is reduced at the positive electrode, forming Na2S5 (Fig. 5). Initially, a two-phase liquid is formed because Na2S5 is immiscible with S at these temperatures. Over half of the discharge occurs in the two-phase region, where the open-circuit voltage is 2.08 V (33). During charge, the Na polysulfides are oxidized, and when the Na content falls below Na2S5, the two phase-region of Na2S5 and S reappears. In this case, the formation of S must be managed appropriately, or else the S can deposit on or near the electrolyte, increasing cell resistance and limiting the amount of charging. Early in its development in the 1980s, the Na/MeCl2 battery was nicknamed the ZEBRA battery partially because of its scientific origins in South Africa, although its acronym stands for Zero-Emission Battery Research Activities. The positive electrode in this battery is a semisolid combination of an electrochemically active metal chloride such as NiCl2 and a molten secondary electrolyte, NaAlCl4, which conducts Na+. During discharge, metallic Na is oxidized at the solid electrolyte interface. Na+ ions are transported through the b-alumina electrolyte to the cathode via the molten NaAlCl4. The solid metal chloride is converted into NaCl and the parent metal (Ni in the case of NiCl2). The open-circuit voltage is 2.58 V (34). On charge, the Ni is oxidized, and the charge capacity is determined by the amount of NaCl available in the cathode. Both batteries are based on the ion transport properties of the b-alumina family of materials. The high ionic conductivity of these materials is the result of an unusual structure
SPECIALSECTION
Redox-Flow Batteries Redox-flow batteries also have their origins in the 1960s, with the development of the zinc/chlorine (Zn/Cl) hydrate battery. As a general description, a redox-flow cell uses two circulating soluble redox couples as electroactive species that are oxidized and reduced to store or deliver energy (44). By comparison, batteries rely on internal solid electrodes to store energy. The flow-cell assembly (Fig. 6) has an ionselective membrane separating the positive and negative redox species, which are contained in separate storage tanks. During operation, redox-active ions undergo oxidation or reduction reactions when they are in contact or close proximity to the current collector; the membrane allows the transport of non-reaction ions (such as H + and Na+ ) to maintain electroneutrality and electrolyte balance. Since the 1970s, numerous types of redox flow battery systems have been investigated (45). A partial list includes iron/chromium, vanadium/ bromine, bromine/polysulfide, zinc-cerium, zinc/ bromine (Zn/Br), and all-vanadium. The allvanadium (1.26 V) and Zn/Br (1.85 V) systems are the most advanced and have reached the demonstration stage for stationary energy storage. Interest in the all-vanadium system is based on having a single cationic element so that the cross-over of vanadium ions through the membrane upon long-term cycling is less deleterious than with other chemistries (46). Redox-flow batteries possess a number of advantages (47). The simplicity of the electrode reactions contrasts with those of many conventional batteries that involve, for example, phase transformations, electrolyte degradation, or electrode morphology changes. Perhaps their most
attractive feature is that power and energy are uncoupled, a characteristic that many other electrochemical energy storage approaches do not have (48, 49). This gives considerable design flexibility for stationary energy storage applications. The capacity can be increased by simply increasing either the size of the reservoirs holding the reactants or increasing the concentration of the electrolyte. In addition, the power of the system can be tuned by either (i) modifying the numbers of cells in the stacks, (ii) using bipolar electrodes, or (iii) connecting stacks in either parallel or series configurations. This provides modularity and flexible operation to the system. Despite the apparent advantages for redoxflow batteries, application of this technology to
Ionselective membrane
Electrode Electrolyte tank
H
H
4+
+
V /V
O 2 -H
H+
O V
+
H+
+
3+
H H+
+
+ 4
2-
Electrolyte tank
+
O2- V4+
Catholyte
issues associated with the lack of appropriate membranes for controlling long-term ion cross-over effects. Designing better membranes is necessary, but whether such membranes can be of low cost is far from certain. Another important issue with redox-flow systems is that the currently used redox couples, even with enhanced solubility, are limited to concentrations of about 8 M. This feature is largely responsible for the fact that redox-flow systems do not surpass 25 Wh kg−1 (Fig. 2). The identification of lower-cost redox couples with high solubility would seem to be an essential development in order for this technology to succeed. Researchers recognize that redox-flow approaches represent potentially new directions for increasing energy density. The semisolid Li battery
V +
O 2-
H+
+
H
H
2-
O
+
V2
+
H
5+
H+
2-
O
e
Downloaded from www.sciencemag.org on November 17, 2011
To provide appropriate voltages, energy, and power, cells are assembled in series-parallel configurations to form modules, and the modules themselves are connected in series-parallel arrangements to form batteries. This networking approach is designed to minimize the effect of individual cell failures. Modules are thermally insulated and equipped with auxiliary heaters in order to maintain a minimum operating temperature. Thermal management is especially challenging. The internal temperature of a module increases on discharge because of joule heating and exothermic cell reactions, whereas during charge, there is a gradual cooling largely because of the cell endothermic reaction (41). The Na/MeCl2 batteries were developed almost exclusively for electric vehicles. At the time of their development, the technology seemed to offer certain advantages over Na/S in terms of tolerance to overcharge and overdischarge, the ability to assemble cells in the discharged state, a safe low-resistance failure mode, and potentially easier solutions for corrosion and sealing (42). Only recently have these batteries been directed at potential utility applications (43).
V5
+
O2-
H
+
H H
+
H
O 2-
e
-
Anolyte
V2+/ V3+
+
+
V2
+
e Pump
e
Pump
Power source-load
Fig. 6. Schematic of the various components for a redox-flow battery. The cell consists of two electrolyte flow compartments separated by an ion-selective membrane. The electrolyte solutions, which are pumped continuously from external tanks, contain soluble redox couples. The energy in redox-flow batteries is stored in the electrolyte, which is charged or discharged accordingly. In practice, individual cells are arranged in stacks by using bipolar electrodes. The power of the system is determined by the number of cells in the stack, whereas the energy is determined by the concentration and volume of electrolyte. In the vanadium redox-flow battery shown here, the V(II)/V (III) redox couple circulates through the negative compartment (anolyte), whereas the V (IV)/V(V) redox couple circulates through the positive compartment (catholyte). [Derived from (38)] stationary energy storage is still uncertain. One principal reason is that redox-flow systems have been limited to relatively few field trials. In contrast, other battery technologies have benefited from extensive experience in the development of products for portable electronics and automotive applications. A related disadvantage of flow batteries is the system requirements of pumps, sensors, reservoirs, and flow management (48, 49). From a technical standpoint, there are reliability
www.sciencemag.org
SCIENCE
VOL 334
demonstrated by Massachusetts Institute of Technology researchers uses electrode materials identical to those found in the LIB, but now the electrode materials are conducting inks (for example, suspensions of LiCoO2 and of Li4Ti5O12 powders in nonaqueous electrolyte solutions) rather than solids (50). The inks circulate separately on either side of a membrane that regulates the Li-ion transport between positive and negative electrodes. Both half cells and full cells have been demonstrated.
18 NOVEMBER 2011
933
Future Directions There are two related questions that need to be addressed: What are the expectations for EES in the future, and what role will batteries play in this future? The first part is becoming clearer as the value of energy storage becomes increasingly evident. A recent EPRI study identified a number of high-value opportunities for energy storage, including wholesale energy services, integration of renewables, commercial and industrial power quality and reliability, transportable systems for transmission and distribution grid support and energy management (1). Moreover, some of these benefits are complementary, further improving the economics of energy storage. The success of these applications of energy storage will depend on how well storage technologies can meet key expectations. The most important of these are low installed cost, high durability and reliability, long life, and high round-trip efficiency. The installed cost comprises the materials costs, production costs, and installation costs for the system. In the future, the preferred energy storage technologies will be composed of low-cost, easily acquired materials that are developed into
934
products through a relatively simple manufacturing process and installed with few special requirements. Operations and maintenance costs are also important; these costs are often tied to the durability and lifetime of the energy storage solution, for which the lifetimes of most assets are measured in decades. Last, a premium will be placed on energy-efficient systems that do not lose energy through self-discharge or parasitic losses. With so many potential financial considerations, it is not surprising that cost is given as the reason that energy storage is not widely used on the grid. The battery systems reviewed here satisfy several, but not all, of the energy storage criteria mentioned above. Na/S is commercially viable, and if this emerging technology follows patterns similar to others, costs can be expected to decrease as more production and operational experience is gained. The technology, which is more than 30 years old, needs to integrate some of the scientific advances that have taken place in the design of materials, creating new electrode architectures and identifying new chemistries to provide safe operation. Lowering the Na/S operating temperature is one topic that will affect the technology. Moreover, these advances will benefit Na-ion technology, which is of growing interest because of its promise as a low-cost approach for grid storage applications. Redox-flow batteries possess several promising attributes for energy storage, with low cost being one of the important drivers for this technology. A number of demonstration projects, ranging in size from 5 to 50 MWh and using a variety of different chemistries, are under way (48). The outcomes from these projects over the next 2 to 4 years will have a substantial influence on the future of this technology. The recent developments involving Li-redox flow and alkali-redox flow batteries stand as great opportunities that leverage existing knowledge of Li-ion batteries with the advantages of redox-flow systems. Energy storage systems based on Li-ion batteries are expected to take a different route than either Na/S or redox-flow batteries. The development of Li-ion batteries for commercial electronics and automotive applications enabled this technology to address reliability, cycle life, safety, and other factors that are equally as important for stationary energy storage. The research environment for developing new low-cost materials is well established, and recent efforts directed at low-temperature processing and renewable organic electrodes provide the basis for future advances in the field. However, it is the volume production anticipated for the electric vehicle market that can lead to improvements in manufacturing process and provide an economy of scale that will bring about the lower costs required to make this battery technology viable for EES. Another interesting scenario is the prospect of recovering Li-ion batteries used in automotive industries and to give them a “second life” in large-scale energy storage applications.
18 NOVEMBER 2011
VOL 334
SCIENCE
Note added in proof: Na/S batteries were responsible for a fire that occurred at a power plant in Joso City (Ibaraki Prefecture) on 21 September 2011 (www.ngk.co.jp/english/news/2011/1028_01. html). Although the cause of the fire is still under investigation, this event underscores the fact that safety issues for Na/S batteries have not been completely resolved. References and Notes 1. “Electrical energy storage technology options” (Report 1020676, Electric Power Research Institute, Palo Alto, CA, December 2010). 2. EPRI-DOE Handbook of Energy Storage for Transmission and Distribution Applications (1001834, EPRI, Palo Alto, CA, and the U.S. Department of Energy, Washington, DC, 2003). 3. G. L. Soloveichik, Annu. Rev. Chem. Biomol. Eng. 2, 503 (2011). 4. “Basic research needs for electrical energy storage” (Office of Basic Energy Sciences, U.S. Department of Energy, Washington, DC, July 2007). 5. “Power Generation from Coal: Measuring and Reporting Efficiency Performance and CO2 Emissions” (International Energy Agency, October 2010). 6. “Energy Storage for the Electricity Grid: Benefits and Market Potential Assessment” (Report SAND 2010-0815, Sandia National Laboratories, Albuquerque, NM, February 2010). 7. M. Winter, R. J. Brodd, Chem. Rev. 104, 4245 (2004). 8. P. Simon, Y. Gogotsi, Nat. Mater. 7, 845 (2008). 9. J. B. Goodenough, Y. Kim, Chem. Mater. 22, 587 (2010). 10. J. M. Tarascon, M. Armand, Nature 414, 359 (2001). 11. V. Etacheri, R. Marom, R. Elazari, G. Salitra, D. Aurbach, Energy Environ. Sci. 4, 3243 (2011). 12. M. Armand, J. M. Tarascon, Nature 451, 652 (2008). 13. A. S. Aricò, P. Bruce, B. Scrosati, J. M. Tarascon, W. van Schalkwijk, Nat. Mater. 4, 366 (2005). 14. H. Li, Z. X. Wang, L. Q. Chen, X. J. Huang, Adv. Mater. (Deerfield Beach Fla.) 21, 4593 (2009). 15. P. Poizot, S. Laruelle, S. Grugeon, L. Dupont, J. M. Tarascon, Nature 407, 496 (2000). 16. M. N. Obrovac, L. Christensen, Electrochem. Solid-State Lett. 7, A93 (2004). 17. P. Barpanda et al., Inorg. Chem. 49, 7401 (2010). 18. A. Yamada, S. C. Chung, K. Hinokuma, J. Electrochem. Soc. 148, A224 (2001). 19. J. M. Tarascon, Philos. Trans. R. Soc. A 368, 3227 (2010). 20. J. M. Tarascon et al., Chem. Mater. 22, 724 (2010). 21. H. Chen et al., J. Am. Chem. Soc. 131, 8984 (2009). 22. G. Girishkumar, B. McCloskey, A. C. Luntz, S. Swanson, W. Wilcke, J. Phys. Chem. Lett. 1, 2193 (2010). 23. K. T. Lee et al., Adv. Energy Mater. 1, 34 (2011). 24. K. M. Abraham, Z. Jiang, J. Electrochem. Soc. 143, 1 (1996). 25. J. P. Zheng, R. Y. Liang, M. Hendrickson, E. J. Plichta, J. Electrochem. Soc. 155, A432 (2008). 26. T. Ogasawara, A. Débart, M. Holzapfel, P. Novák, P. G. Bruce, J. Am. Chem. Soc. 128, 1390 (2006). 27. A. Débart, A. J. Paterson, J. Bao, P. G. Bruce, Angew. Chem. Int. Ed. 47, 4521 (2008). 28. Y. C. Lu et al., J. Am. Chem. Soc. 132, 12170 (2010). 29. G. Girishkumar, B. McCloskey, A. C. Luntz, S. Swanson, W. Wilcke, J. Phys. Chem. Lett. 1, 2193 (2010). 30. J. F. Whitacre, A. Tevar, S. Sharma, Electrochem. Commun. 12, 463 (2010). 31. Y. F. Y. Yao, J. T. Kummer, J. Inorg. Nucl. Chem. 29, 2453 (1967). 32. X. C. Lu, G. G. Xia, J. P. Lemmon, Z. G. Yang, J. Power Sources 195, 2431 (2010). 33. J. L. Sudworth, A. R. Tilley, The Sodium Sulfur Battery (Chapman & Hall, London, 1985). 34. C. H. Dustmann, J. Power Sources 127, 85 (2004). 35. G. C. Farrington, J. L. Briant, Science 204, 1371 (1979). 36. D. La Rosa et al., ChemSusChem 3, 1390 (2010). 37. J. L. Sudworth et al., MRS Bull. 25, 22 (2000). 38. Z. Yang et al., Chem. Rev. 111, 3577 (2011). 39. S. F. Song et al., J. Solid State Electrochem. 14, 1735 (2010). 40. B. Dunn, M. W. Breiter, D. S. Park, J. Appl. Electrochem. 11, 103 (1981).
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
The novel feature here is the use of redox-active materials in suspension so as to circumvent the problem of the relatively low solubility of the metal ion redox couples in aqueous solution. The flowable inks will be in the 10 to 40 M range, which is at least 5 times higher than traditional redox flow systems. Combining the higher materials concentration with the feasibility of achieving 4-V working systems is likely to lead to considerable improvement in energy density, perhaps without substantially affecting power density. Another Li-ion–based flow system was demonstrated recently by Goodenough and colleagues. In this design, an aqueous cathode operating in a flow-through mode was separated from a metallic Li anode by a Li-ion–conducting solid electrolyte and an organic liquid electrolyte (51). This redox-flow system used an aqueous cathode containing 0.1 M K3Fe(CN)6 and demonstrated highly efficient energy storage at 3.4 V. The design strategy presented here offers some noteworthy advances: (i) Li+ ion transport in solution is enhanced as compared with that in a solid insertion cathode and (ii) the absence of structural changes during charge/discharge is beneficial for longterm cycling. The first laboratory prototypes were limited by low solubility of the metal-ion redox couple in the aqueous solvent and the poor mobility of Li+ in the solid electrolyte. It is expected that the performance of the rechargeable alkaliion cathode flow battery will improve substantially through the use of a better solid electrolyte and the possibility of using cathode inks. But perhaps the more important point illustrated in these studies is that redox-flow concepts adapt to other chemistries and hold considerable promise for improving battery performance and especially energy density.
SPECIALSECTION 47. M . Skyllas-Kazacos et al., J. Electrochem. Soc. 158, R55 (2011). 48. D. H. Doughty, P. C. Butler, A. A. Akhil, N. H. Clark, J. D. Boyes, Electrochem. Soc. Interface 19, 49 (2010). 49. T. Nguyen, R. F. Savinell, Electrochem. Soc. Interface 19, 54 (2010). 50. M. Duduta et al., Adv. Energy Mater. 1, 511 (2011). 51. Y. H. Lu, J. B. Goodenough, J. Mater. Chem. 21, 10113 (2011). Acknowledgments: Support (B.D.) is from the Center for Molecularly Engineered Energy Materials, an Energy Frontier Research Center funded by the U.S. Department of Energy (DOE) Office of Basic Energy Sciences (DE-SC001342)
REVIEW
Lowering the Temperature of Solid Oxide Fuel Cells Eric D. Wachsman* and Kang Taek Lee Fuel cells are uniquely capable of overcoming combustion efficiency limitations (e.g., the Carnot cycle). However, the linking of fuel cells (an energy conversion device) and hydrogen (an energy carrier) has emphasized investment in proton-exchange membrane fuel cells as part of a larger hydrogen economy and thus relegated fuel cells to a future technology. In contrast, solid oxide fuel cells are capable of operating on conventional fuels (as well as hydrogen) today. The main issue for solid oxide fuel cells is high operating temperature (about 800°C) and the resulting materials and cost limitations and operating complexities (e.g., thermal cycling). Recent solid oxide fuel cells results have demonstrated extremely high power densities of about 2 watts per square centimeter at 650°C along with flexible fueling, thus enabling higher efficiency within the current fuel infrastructure. Newly developed, high-conductivity electrolytes and nanostructured electrode designs provide a path for further performance improvement at much lower temperatures, down to ~350°C, thus providing opportunity to transform the way we convert and store energy. uel cells are the most efficient means to directly convert stored chemical energy to usable electrical energy (an electrochemical reaction). Although the more common protonexchange membrane fuel cells (PEMFCs) require hydrogen fueling, because they are based on proton conducting electrolytes, solid oxide fuel cells (SOFCs) can oxidize essentially any fuel, from hydrogen to hydrocarbons to even carbon, because the electrolyte transports an oxygen ion. An SOFC consists of three major components: two porous electrodes (cathode and anode) separated by a solid oxygen ion (O2–) conducting electrolyte (Fig. 1A). At the cathode, O2 (from air) is reduced and the resulting O2– ions are transported through the electrolyte lattice to the anode where they react with gaseous fuel, yielding heat, H2O, and (in the case of hydrocarbon fuels) CO2, and releasing e– to the external circuit. Multiple cells are combined in series via interconnects that provide both electrical contacts and gas channels between individual cells. The resulting “stacks” are then arranged in series and parallel configurations to provide desired voltage and power outputs from portable power and
F
University of Maryland Energy Research Center, College Park, MD 20742, USA. *To whom correspondence should be addressed. E-mail:
[email protected]
transportation applications, to distributed generation and large-scale power generation, in both civilian and military sectors (Fig. 1B). Among the technologies available to convert hydrocarbon-based resources (which include not only fossil fuels but also, potentially, biomass and municipal solid waste) to electricity, SOFCs are unique in their potential efficiency. For stand-alone applications, SOFC chemical to electrical efficiency is 45 to 65%, based on the lower heating value (LHV) of the fuel (1), which is twice that of an internal combustion (IC) engine’s ability to convert chemical energy to mechanical work (2). In a combined cycle, there are numerous combined heat and power (CHP) applications using SOFC systems, which have the potential to achieve efficiencies of >85% LHV (3). Unfortunately, government policy, the popular press, and many scientific publications have focused on fuel cells as part of a broader hydrogen economy, thereby relegating fuel cells to a “future energy” solution due to the need for a required overhaul of our current hydrocarbonfueling infrastructure. Although this may be true for PEMFCs, SOFCs have the advantage of fuel flexibility that allows them to be used on our existing hydrocarbon fuel infrastructure (4) while simultaneously providing efficiency gains (and corresponding CO2 emission reductions).
www.sciencemag.org
SCIENCE
VOL 334
and from the DOE Office of Electricity, Energy Storage Systems Program. The authors greatly appreciate the insightful comments provided by G. Farrington and A. Shukla. We also thank E. Lan and L. Smith for their assistance with the manuscript.
Supporting Online Material www.sciencemag.org/cgi/content/full/334/6058/928/DC1 Figs. S1 to S3 Table S1 References (52–62) 10.1126/science.1212741
Why Reduce SOFC Operating Temperature? The key technical issue that has limited the development and deployment of this transformative technology is its high operating temperature, resulting in higher systems costs and performance degradation rates, as well as slow start-up and shutdown cycles, the latter dramatically limiting applicability in portable power and transportation markets. Over the past decade, considerable progress has been achieved in bringing the temperature down to an intermediate temperature (IT) range of 650 to 800°C so that metallic interconnects could be used to reduce cost. Low-temperature (LT) SOFCs (≤650°C) can further reduce system cost due to wider material choices for interconnects and compressive nonglass/ceramic seals, as well as reduced balance of plant (BOP) costs. Moreover, below 600°C, both radiative heat transfer (Stefan-Boltzmann) and sintering rates exponentially drop off, thus reducing insulation costs and primary performance degradation mechanisms, respectively. At even lower temperatures (≤350°C), cheap stamped stainless steel interconnects, elastomeric/ polymeric seals (e.g., Kapton), and off-the-shelf BOP are possible. In addition, rapid start-up and repeated thermal cycling, from ambient to operating temperature, becomes possible. These are critical parameters for portable power and transportation applications, and it was because of PEMFCs’ lower operating temperature (~100°C) that they were chosen for these applications over SOFCs, even though PEMFCs require hydrogen fueling. Another reason to reduce operating temperature is maximum theoretical efficiency. In contrast to the Carnot cycle temperature dependence of IC engines, theoretical fuel cell efficiency increases with decreasing temperature [fig. S1 and supporting online material text (SOM text)]. For example, the maximum theoretical efficiency of an SOFC using CO as a fuel increases from 63% at 900°C to 81% at 350°C. At first glance, this would imply that PEMFCs are more efficient than SOFCs because of their lower operating temperature. However, this ignores two important contributors to overall system efficiency. The first is that the vast majority of all H2 produced today comes from hydrocarbon resources (typically CH4), thus requiring additional external processes [e.g., steam reforming or catalytic partial oxidation (CPOX), water
18 NOVEMBER 2011
Downloaded from www.sciencemag.org on November 17, 2011
41. A. Bito, “Overview of the sodium-sulfur battery for the IEEE Stationary Battery Committee,” paper presented at the IEEE Power Engineering Society General Meeting, San Francisco, CA, 12 to 16 June 2005. 42. J. W. Braithwaite, W. L. Auxer, in Handbook of Batteries, D. Linden, T. B. Reddy, Eds. (McGraw-Hill, 2004), chap. 40. 43. www.geenergystorage.com/ 44. C. P. de Leon et al., J. Power Sources 160, 716 (2006). 45. M. Bartolozzi, J. Power Sources 27, 219 (1989). 46. M. Skyllas-Kazacos, M. Rychcik, R. G. Robins, A. G. Fane, M. A. Green, J. Electrochem. Soc. 133, 1057 (1986).
935
What Are the Technological Issues for LT-SOFCs and Where Are We Today? Overall efficiency depends on thermodynamics (attained voltage relative to the theoretical open circuit potential and fuel use) and kinetics (polarization losses) during operation (fig. S2 and SOM
text). Addressing the increasing polarization losses at lower temperatures (associated with electrolyte conduction and electrode reaction kinetics) is the key issue and has been the focus of many groups over the past couple of decades (4, 6). The entire SOFC material set is predicated by the selection of the electrolyte, in terms of chemical and thermomechanical stability with the electrolyte. The vast majority of SOFCs use a zirconia-based electrolyte, typically yttria-stabilized zirconia (YSZ), because of its superior stability. Although a good oxygen-ion conductor, it is far from having the highest conductivity (Fig. 2); thus, the SOFC community has transitioned from electrolyte-supported cells to electrode-supported cells to reduce the electrolyte’s ohmic polarization. These, typically anode-supported cells, allow for significantly thinner electrolytes (7) and have allowed the community to reduce operating temperatures to the IT range.
Electronic conductor Triple phase boundary Ionic conductor
A Single SOFC
Stack cell
With a typical open-circuit potential (OCP) of 1 V, a targeted power density of 1 W/cm2 requires a total cell area-specific resistance (ASR) of less than ~0.25 W-cm2 (based on simple linear currentvoltage behavior). Thus, assuming that 60% of the total cell ASR is attributed to the electrolyte (0.15 W-cm2), an operating temperature of 950°C is necessary to achieve this targeted ASR with ~150-mm YSZ, and to operate at 500°C would require the electrolyte to have a thickness less than 1 mm (8). Therefore, a variety of deposition technologies have been employed to fabricate thin-film electrolytes (9–11). For example, the Prinz group recently reported fabrication of SOFCs with a 100-nm electrolyte (a bilayered structure of 50-nm YSZ and 50-nm gadolinia-doped ceria (GDC)), achieving a peak power density of ~400 mW/cm2 at 400°C (12). Thus, demonstrating extremely small polarization loss at low temperatures (albeit with Pt electrodes) is possible.
Cathode
e-
O2-
Electrolyte 1 Electrolyte 2 Anode
O2
e-
O2-
Cathode O2-
Interconnect
Electrolyte
eeO2-
Electrolyte 1 2O2- Electrolyte O
Electrolyte 2
eO2-
Anode e-
SOFC Fuel channel Interconnect Air channel
e-
5 cells
e-
H2O
CO H2
CO2
1 module = 10 stacks
10 cm
10
1 kw
Fig. 1. LT-SOFCs across the energy conversion spectrum, from portable power and transport to stationary applications, in both civilian and military sectors. (A) Schematic diagrams of structure of high-performance LT-SOFCs from low magnification (stack) to high magnification (nano/micro-structured electrodes). Functionally graded bismuth oxide (Electrolyte 1) / ceria (Electrolyte 2) bilayered electrolytes effectively reduce ohmic polarization at lower temperatures. Carefully controlled nanostructured electrodes by infiltration provide highly extended reaction sites compensating exponentially reduced oxygen reaction
936
1 stack = 50 cells
CH4
10 cm
cm 10
100 W
O
Electronic conductor Ionic conductor Triple phase boundary
10 cm
Power density = 2 W/cm2
O2-
O ee-
2-
cm
10 x 10 cell ( 200 W)
B
O2-
2-
18 NOVEMBER 2011
VOL 334
10 kw
100 kw
kinetics at cathode and allow use of hydrocarbon fuels at anode at reduced temperatures. (B) Estimation of power output with LT-SOFCs from a single cell to a module (upper) and schematic diagram of power requirements according to various applications (lower). On the basis of demonstrated high power density (~2 W/cm2 at 650°C) of the state-of-the-art LT-SOFC, a 10 cm by 10 cm planar cell corresponds to ~200 W power output. A stack of 50 planar cells with interconnects (10 cm by 10 cm by 10 cm) can provide 10 kW, and a module consisting of 10 stacks can provide 100 kW. SCIENCE
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
gas shift (WGS), and membrane separation or preferential oxidation (PROX)], each step having a thermodynamic penalty that decreases overall system efficiency (5). Moreover, they cannot take advantage of the higher theoretical fuel cell efficiency of CO over H2 at lower temperature (fig. S1) because CO is a poison for PEMFCs versus a fuel for SOFCs. The second reason that lower temperature does not necessarily result in higher system efficiency is that all of the major cell polarization losses are thermally activated. Thus, the difference between attained efficiency and theoretical efficiency increases as temperature is lowered.
SPECIALSECTION oxides are particularly attractive because of their superior ionic conductivity at lower tempera700 650 600 550 500 450 400 350 tures (Fig. 2). For example, at 500°C, the ASRs 6 of 10-mm-thick YSZ, GDC, and erbia-stabilized bismuth oxide (ESB) are 1.259, 0.143, and 0.037 DWSB 5 W-cm2, respectively (14). Thus, at the same thickness and temperature, doped ceria and stabilizedbismuth oxide can reduce ohmic losses by 1 to 4 ESB 2 orders of magnitude, respectively, compared with YSZ. 3 SNDC Unfortunately, higher conductivity comes at the expense of lower thermodynamic stability, 2 GDC with CeO2 electrolytes becoming electronically conductive and Bi2O3 electrolytes decomposing 1 to metallic Bi under the reducing fuel environYSZ ment (15, 16). The electronic leakage current 0 with CeO2 electrolytes results in a reduced OCP (4), which is a decrease in efficiency (SOM text). -1 To overcome this issue, we proposed a function1.0 1.1 1.2 1.3 1.4 1.5 1.6 ally graded ceria/bismuth-oxide bilayered elec1000/T (K-1) trolyte (Fig. 3A), where the GDC layer on the anode (fuel) side protects the ESB layer from Fig. 2. Comparison of ionic conductivity of various solid oxide decomposing while the ESB layer on the cathelectrolytes. Stabilized bismuth oxides (ESB-Er0.4Bi1.6O3 and DWSBode (oxidant) side blocks the leakage current Dy0.08W0.04Bi0.88O1.56) show superior ionic conductivity compared through the GDC layer because of its high transwith that of doped ceria (GDC-Gd0.1Ce0.9O1.95 and SNDCference number (ratio of ionic to total conductivSm0.075Nd0.075Ce0.85O2-d) and stabilized zirconia (YSZ-Y0.16Zr0.92O2.08). ity). Using this synergistic structure, we demonstrated the ability to obtain near-theoretical OCP with two highly A L L L L conductive electrolytes that by themselves would not have been suffiPO ciently stable for SOFC applications 2 B air 1.2 4.0 Decomposition (17). Moreover, the bilayer electroPO 2 lyte was stable for 1400 hours of DWSB/SNDC PO 2 3.5 (projected) testing (17) and showed no indicafuel Interfacial 1.0 PO profile tion of interfacial phase formation or PO 2 2 3.0 thermal mismatch (18). LGDC/LESB < optimal LGDC/LESB > optimal With thin highly conductive elec0.8 ESB decomposes ESB is stable trolytes, electrode polarization losses 2.5 dominate as temperature is reduced. C 1.00 ESB/GDC ESB (4m) For example, with an anode-supported 0.6 2.0 GDC (48m) ~10-mm-thick GDC electrolyte (under GDC/ESB(48:4) 0.95 wet H2/dry air conditions) the non1.5 ESB (6m) 0.4 ohmic electrode ASR (0.036 W-cm2) 0.90 GDC/ESB(16:6) GDC (16m) was only ~41% of the total cell ASR 1.0 0.85 at 650°C but increased to ~73% (0.48 GDC 0.2 ESB (5m) W-cm2) at 450°C (19). Moreover, the 0.5 0.80 GDC/ESB GDC (10m) thermally activated kinetics of the (10:5) oxygen reduction reaction (ORR) GDC 0.75 0.0 0.0 (single layer) result in cathode polarization being 0 1 2 3 4 5 6 7 8 GDC (10m) the primary loss mechanism at low 0.70 Current density (A/cm2) 0 500 550 600 650 temperatures. Temperature ( C) We recently integrated the concepts above into an anode-supported Fig. 3. (A) Schematic of ceria/bismuth oxide bilayer concept demonstrating the effect of relative thickness on interfacial cell composed of a thin, dense GDC oxygen partial pressure and ESB stability. (B) Current-voltage behavior (left y axis) and power density (right y axis) for (~10 mm)/ESB(~4 mm) bilayered SOFCs with GDC single-layer (solid blue line) and ESB/GDC bilayer (solid red line) electrolytes at 650°C using 90 standard electrolyte with a newly developed cubic centimeter per minute of 3% wet H2 (anode side)/dry air (cathode side). With ESB/GDC bilayer electrolyte, a power high-performance bismuth ruthenatedensity of ~2 W/cm2 at 650°C was achieved because of higher OCP and reduced cathodic polarization. Assuming higher bismuth oxide (BRO7-ESB) comOCP (~1 V) by controlling total thickness and thickness ratio of more conductive DWSB/SNDC bilayer electrolyte, the posite cathode and demonstrated an projected maximum power density (dotted red lines) is ~3.5 W/cm2 under the same conditions. (C) Effect of total thickness exceptionally high power density of 2 and thickness ratio of bilayered electrolyte on OCP. OCP increases as total thickness and ESB/GDC thickness ratio increase ~2 W/cm at 650°C (20). This is one of the highest reported power densities and as temperature decreases, indicating the potential to achieve theoretical OCP at these temperatures (SOM text). Ln [ T (S.cm-1.K)]
GDC
ESB
Potential (V)
ESB
Power density (W/cm2)
Open-circuit potential (V)
GDC
Downloaded from www.sciencemag.org on November 17, 2011
However, this was done with semiconductor processing (e.g., sputtering) on a Si wafer, and based on the reported active area (240 mm by 240 mm), the peak power output per individual SOFC is only ~0.23 mW. Although these methods are suitable for micro-SOFCs, it is unlikely that they are scalable and cost-effective for mass production of large-scale (kW to MW) SOFCs. Rather, from a practical standpoint for large-scale manufacturing, conventional multilayer thick-film ceramic processing (e.g., tape casting) is more appropriate. These processes imply a minimum thickness of ~10 mm, which for YSZ limits operating temperature to ≥700°C. Therefore, low-temperature (LT) SOFCs are only possible with higher conductivity electrolytes. Various alternative electrolytes have been investigated (13), among which aliovalent-doped ceria and isovalent-cation–stabilized bismuth
T (°C)
www.sciencemag.org
SCIENCE
VOL 334
18 NOVEMBER 2011
937
Outlook: Toward Further Performance Increase at Lower Temperatures The high power density LT-SOFCs (~650°C) described above are already suitable for numerous stationary applications. However, significant increases in power density and reductions in temperature are readily achievable by optimizing the bilayer thicknesses to increase OCP, incorporating even greater conductivity electrolytes, and engineering infiltrated nanostructured catalytically active electrodes.
938
18 NOVEMBER 2011
SCIENCE
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
Specific power (W/kg)
VOL 334
Specific energy (Wh/kg)
How Do LT-SOFCs Compare with Competing Technologies? For stationary applications, Bloom Energy is arguably the current commercial leader in terms of deployed SOFC units. Their zirconia-based SOFCs are reported to deliver power densities of ~0.2 W/cm2 at ~900°C (21, 22). Our current LT-SOFC power densities (at the cell level) are higher by a factor of 10 at ~250°C lower temperature, indicating the potential for much higher energy efficiency with considerable cost reduction. For portable and transportation applications, volumetric and gravimetric power densities are key performance metrics. The total thickness of our LT-SOFC is 0.5 mm, and the expected interconnect thickness is 1.5 mm. Thus, based on areal power density of 2 W/cm2, the stack volumetric power density and the gravimetric power density are ~10 W/cm3 and ~3 kW/kg (fig. S3 and SOM text), respectively, exceeding that of an IC engine (Fig. 4 A). Moreover, with liquid hydrocarbon fueling, SOFCs and IC engines have essentially the same specific energy, that of the fuel (~1 kWh/kg) (23). Thus, because our LT-SOFC has essentially the same power and energy density as an IC engine (Fig. 4B), it could potentially transform the automotive sector as, for example, a range extender for plug-in hybrid electric vehicles (PHEVs) operating on conventional fuels. The corresponding 10-kW stack would only be a small cube of 10 cm per edge (Fig. 1B). However, it must be noted that other groups have also achieved ~2 W/cm2 power densities, albeit at 800°C with YSZ-based cells (7, 24), and these laboratory-scale button-cell results do not directly translate to full-scale stack performance because of numerous parasitic losses such as cellinterconnect interfacial resistances, thermal gradients, and higher fuel use. Nevertheless, this demonstrates the potential of this technology if these parasitic losses can be addressed.
Optimize the electrolyte layers to increase Moreover, considering the twelvefold higher OCP without increasing ASR. Although the OCP conductivity of DWSB compared with SNDC was increased with addition of the bilayer in at 650°C, increasing the DWSB thickness from the ~2 W/cm2 cells (20), the full theoretical 7.5 to 22.5 mm adds less than 9% to the ohmic value was not achieved because neither the ASR, while significantly increasing total bilayer total nor the relative thickness was optimized. thickness to 36 mm, relative (22.5:13.5 ratio) As the thickness of mixed ionic and electronic bismuth oxide thickness, and as a result OCP (e.g., conducting (MIEC) electrolytes, such as GDC, Fig. 3C). The effect of increasing OCP to 1 V is decreased, the electronic leakage current in- without increasing total polarization would have creases, resulting in lower OCP. As such, there a significant effect on maximum power density, as is an optimum thickness in terms of tradeoff between reducing ASR with thinner electrolytes and inCombustion engines 10000 A PEM fuel cell creasing OCP with thicker electroDMFC SOFC Photovoltaic cells lytes (25). For bilayer electrolytes, Electro magnetic generator Thermo electric generator OCP further depends on relative 1000 thickness of the constituent layers, Combustion engines increasing with relative ESB thickDirect methanol ness (26). Recently, we investigated 100 fuel cell the effect of ESB/GDC bilayer thickPEM fuel cell nesses on OCP in anode-supported Electro cells, achieving near-theoretical OCP magnetic 10 (~0.95 V) at 500°C by modifying generator total thickness and thickness ratio Photovoltaic cells (Fig. 3C) (SOM text). These reThermo electric generator sults show that higher OCP, and 0.01 0.1 1 10 thus efficiency (SOM text), can be Power density (W/cm3) achieved with a thicker electrolyte and greater relative bismuth oxide B 1000 IC thickness. However, to negate any SOFC engine Fuel ohmic ASR increase with thickness, 100 h cells EV goal we would use even more conductive Li-ion 100 PHEV goal electrolytes. Ni-MH Lead-acid Based on two decades of research 10 h HEV goal on the fundamentals of ion conduc10 Capacitors tion (27), we developed the highest reported conductivity solid oxide elec1h 36 s 0.1 h 3.6 s trolyte with a co-doped stabilized 1 0 1 2 3 10 10 10 104 10 bismuth oxide [Dy0.08W0.04Bi0.88O1.56 Specific power (W/kg) (DWSB)], an increase by a factor Acceleration of 4 over ESB at 500°C (14). In fact, 2 at 350°C, the ASR of 10-mm-thick Fig. 4. (A) Comparison of specific power of the present ~2W/cm 2 DWSB is only 0.6 W-cm , sufficiently SOFC at 650°C compared with various energy conversion devices low for SOFC operation at this tem- as a function of power density (23). (B) Ragone plot (specific perature. Using this approach and energy versus specific power) for various energy devices (40) insight from molecular dynamic sim- compared with the present SOFC. ulation studies by Andersson et al. (28), we subsequently developed a higher conductivity co- projected for DWSB/SNDC in Fig. 3B. Howdoped ceria electrolyte [Sm0.075Nd0.075Ce0.85O2-d ever, increasing the OCP by blocking the par(SNDC)] with a ~30% increase over GDC at allel electronic current would increase the cell 550°C (29). The conductivity of these newer ASR by the small amount of electronic current electrolytes is compared with ESB, GDC, and that was blocked. As temperature decreases, ceria-based electroYSZ in Fig. 2. For 650°C operation, we can use the factor lytes have a wider electrolytic domain (the region 1.9 higher conductivity of DWSB (versus ESB) where ionic conductivity dominates over elec(14) and the factor 1.4 higher conductivity of tronic), and bismuth oxide–based electrolytes SNDC (versus GDC) (29) to increase their re- have higher thermodynamic stability under respective thicknesses with no impact on electro- ducing conditions. Thus, at lower temperatures, lyte ASR at that temperature. The ~2W/cm2 obtaining theoretical OCP using DWSB/SNDC SOFC (20) had a ~14-mm-thick ESB/GDC bilayers can be achieved with both thinner total (4:10 ratio) electrolyte. Increasing this to ~21-mm- electrolyte and higher relative bismuth oxide thick DWSB/SNDC (7.5:13.5 ratio) would sig- thicknesses, further reducing ohmic resistance as nificantly increase OCP with no change in ASR. temperature decreases. In fact, the Bi2O3 decomRange
for LT-SOFCs, twice that of an identical cell with a single (~10 mm) layer GDC electrolyte (Fig. 3B), and is a result of both an OCP increase and a dramatic decrease, ~ 40%, of the cathodic ASR. However, the electrode and electrolyte microstructures have not yet been fully optimized; thus, substantial performance improvement is envisioned, as discussed below.
SPECIALSECTION composition and microstructure of electrodes for LT operation. Key for market penetration: Fuel flexibility and thermal cycling. How best to use SOFC fuel flexibility depends on desired fuel choice and operating temperature for a particular application. For stationary distributed generation applications, the ability to internally reform natural gas with conventional Ni-YSZ cermet anodes at ≥700°C has been well demonstrated. Unfortunately, at lower temperatures Ni-YSZ anodes experience performance degradation due to carbon coking and sulfur poisoning as well as Ni oxidation to NiO during thermal cycling (35). However, we use CeO2-based anodes that have been demonstrated to increase both coking and sulfur tolerance as well as the ability to operate directly on hydrocarbon fuels [as an addition to Ni-YSZ anodes (36) and as a Cu-CeO-YSZ composite (37)]. All ceramic anodes are also being developed [e.g., La0.4Sr0.6Ti1-xMnxO3 (38) and Sr2Mg1-xMnxMoO6-d (39)] because they do not undergo metal/metal-oxide phase transition (e.g., Ni/NiO) during thermal cycling. They also exhibit enhanced coking and sulfur tolerance, but to date have lower performance due to insufficient electronic conductivity and/or low electrocatalytic hydrocarbon oxidation activity. Regardless, as temperature is reduced, the tendency toward coking can be compensated by a higher degree of external reforming. The DWSB/SNDC bilayer electrolyte makes SOFC operation down to ~350°C feasible if appropriate electrodes are developed. Although these temperatures would require a thermally integrated external fuel reformer, the overall system efficiency should still be higher than PEMFCs using hydrocarbons as the source of H2. Concluding Remarks SOFCs have tremendous potential for numerous applications, from stationary to mobile power, with high system efficiencies. Depending on application requirements, such as power density, fuel choice, thermal cycling, and system costs, operating temperatures can range from 650°C down to 350°C, the latter allowing for use of simple stamped stainless steel interconnects and elastomeric sealants as well as relatively rapid startup conditions for portable/transportation applications. It is evident that this technology has not fully matured and that major advances are still possible. Nevertheless, LT-SOFC should be a technology of choice for these applications as long as we are in a hydrocarbon-based energy infrastructure. References and Notes 1. EG&G Technical Services, Ed., Fuel Cell Handbook (U.S. Department of Energy, Office of Fossil Energy, Washington, DC, ed. 7, 2004). 2. R. v. Basshuysen, F. Schafer, Internal Combustion Engine Handbook: Basics, Components, Systems, and Perspectives (Society of Automotive Engineers, Warrendale, PA, 2004), pp. 22–24.
www.sciencemag.org
SCIENCE
VOL 334
3. National Renewable Energy Laboratory, 1–10 kW Stationary Combined Heat and Power Systems Status and Technical Potential (NREL/BK-6A10-48265, 2010; www.nrel.gov/docs/fy10osti/48265.pdf). 4. B. C. H. Steele, A. Heinzel, Nature 414, 345 (2001). 5. C. S. Song, Catal. Today 77, 17 (2002). 6. M. Mogensen, S. Skaarup, Solid State Ion. 86-88, 1151 (1996). 7. S. de Souza, S. J. Visco, L. C. DeJonghe, Solid State Ion. 98, 57 (1997). 8. B. C. H. Steele, Solid State Ion. 75, 157 (1995). 9. J. Will, A. Mitterdorfer, C. Kleinlogel, D. Perednis, L. J. Gauckler, Solid State Ion. 131, 79 (2000). 10. L. G. Coocia et al., Appl. Surf. Sci. 96-98, 795 (1996). 11. E. Gourba et al., Ionics 9, 15 (2003). 12. H. Huang et al., J. Electrochem. Soc. 154, B20 (2007). 13. J. C. Boivin, G. Mairesse, Chem. Mater. 10, 2870 (1998). 14. D. W. Jung, K. L. Duncan, E. D. Wachsman, Acta Mater. 58, 355 (2010). 15. K. Eguchi, T. Setoguchi, T. Inoue, H. Arai, Solid State Ion. 52, 165 (1992). 16. T. Takahashi, T. Esaka, H. Iwahara, J. Appl. Electrochem. 7, 299 (1977). 17. E. D. Wachsman, P. Jayaweera, N. Jiang, D. M. Lowe, B. G. Pound, J. Electrochem. Soc. 144, 233 (1997). 18. J. Y. Park, H. Yoon, E. D. Wachsman, J. Am. Ceram. Soc. 88, 2402 (2005). 19. J. S. Ahn, H. Yoon, K. T. Lee, M. A. Camaratta, E. D. Wachsman, Fuel Cells (Weinh.) 9, 643 (2009). 20. J. S. Ahn et al., Electrochem. Commun. 11, 1504 (2009). 21. P. Wray, Am. Ceram. Soc. Bull. 89, 25 (2010). 22. A. Maschietto, K. Kahler, www.mobilemag.com/2010/02/ 25/bloom-energy-server-unveiled-bloom-box-not-for-thehome-just-yet/ (2010). 23. S. F. J. Flipsen, J. Power Sources 162, 927 (2006). 24. A. V. Virkar, J. Chen, C. W. Tanner, J. W. Kim, Solid State Ion. 131, 189 (2000). 25. K. L. Duncan, K. T. Lee, E. D. Wachsman, J. Power Sources 196, 2445 (2011). 26. J. Y. Park, E. D. Wachsman, Ionics 12, 15 (2006). 27. E. D. Wachsman, J. Eur. Ceram. Soc. 24, 1281 (2004). 28. D. A. Andersson, S. I. Simak, N. V. Skorodumova, I. A. Abrikosov, B. Johansson, Proc. Natl. Acad. Sci. U.S.A. 103, 3518 (2006). 29. S. Omar, E. D. Wachsman, J. C. Nino, Appl. Phys. Lett. 91, 144106 (2007). 30. T. Z. Sholklapper, H. Kurokawa, C. P. Jacobson, S. J. Visco, L. C. De Jonghe, Nano Lett. 7, 2136 (2007). 31. M. J. Zhi, N. Mariani, R. Gemmen, K. Gerdes, N. Q. Wu, Ener. Environ. Sci. 4, 417 (2011). 32. J. M. Vohs, R. J. Gorte, Adv. Mater. (Deerfield Beach Fla.) 21, 943 (2009). 33. J. R. Smith et al., Solid State Ion. 180, 90 (2009). 34. E. N. Armstrong, K. L. Duncan, D. J. Oh, J. F. Weaver, E. D. Wachsman, J. Electrochem. Soc. 158, B492 (2011). 35. C. W. Sun, U. Stimming, J. Power Sources 171, 247 (2007). 36. Z. L. Zhan, S. A. Barnett, Science 308, 844 (2005). 37. H. P. He, R. J. Gorte, J. M. Vohs, Electrochem. Solid-State Lett. 8, A279 (2005). 38. Q. X. Fu, F. Tietz, D. Stover, J. Electrochem. Soc. 153, D74 (2006). 39. Y. H. Huang, R. I. Dass, Z. L. Xing, J. B. Goodenough, Science 312, 254 (2006). 40. V. Srinivasan, Batteries for Vehicular Applications (Lawrence Berkeley National Laboratory, Berkeley, CA, 2008); http://bestar.lbl.gov/venkat/files/batteries-forvehicles.pdf.
Downloaded from www.sciencemag.org on November 17, 2011
position PO2 decreases from 10−11.9 atm at 650°C to 10−22.1 atm at 350°C. The latter is comparable to typical anode-fuel PO2’s and, as such, it is possible that DWSB could be used as a single layer at 350°C to take advantage of both its low ASR and unity transference number (thus obtaining theoretical OCP). Optimize electrode microstructure to compensate for thermal activation. Exponentially decreasing area-specific electrode reaction rates (activation polarization) with decreasing temperature can be compensated by shifting the effective particle diameter of the catalytic phase from the micro (10−6) to the nano (10−9) regime, dramatically increasing three-dimensional triple phase boundary (TPB) density [(10−6/10−9)3 = 109], and thus proportionally reducing activation polarization. Moreover, it is the reduced operating temperature that makes these nanostructured electrodes stable against coarsening, the primary performance degradation mechanism. However, this particle size reduction must be done without negatively impacting percolation of the ionic/electronic and gas phase conduction paths that contribute to the electrode’s ohmic and concentration polarizations, respectively. Therefore, nanostructured cathodes have been fabricated by infiltration of precursor solutions into porous ionic-electronic conducting scaffolds (30). For example, recent work by Zhi et al. demonstrated that infiltrated La0.8Sr0.2MnO3 (LSM) in nanofiber YSZ scaffolds effectively decreased cathodic polarization by 70 to ~90% compared with a conventionally mixed LSM-YSZ cathode (31). Moreover, infiltration has been demonstrated to result in low polarization and stability at temperatures below 600°C (32). To reduce the temperature further requires a multifaceted, multidisciplinary approach to deconvolute the multiple mechanistic contributions to electrode polarization, including catalytic, solidstate, and pore transport contributions. By combining focused ion beam and scanning electron microscopy to quantify the cathode microstructure (in terms of tortuosity and porosity for gas diffusion, solid-phase surface area for gas adsorption/surface diffusion, and TPBs for the charge transfer reaction) with electrochemical impedance spectroscopy (EIS), we have been able to obtain direct logarithmic relationships between charge-transfer resistance and TPB length in typical random porous electrode structures (33). Using heterogeneous catalysis techniques (e.g., 18 O-exchange), we have obtained kinetic rate constants and mechanistic results to demonstrate that cathode materials like LSM have facile dissociative adsorption of O2 and are rate limited by the lattice incorporation step, whereas La0.6Sr0.4Co0.2Fe0.8O3 (LSCF) has rapid incorporation and is limited by oxygen surface coverage (34). These kinetic mechanistic results combined with the microstructure-polarization results provide the ability to rationally design the
Supporting Online Material www.sciencemag.org/cgi/content/full/334/6058/935/DC1 SOM Text Figs. S1 to S3 Table S1 References (41–46) 10.1126/science.1204090
18 NOVEMBER 2011
939
Evidence for Interstitial Carbon in Nitrogenase FeMo Cofactor Thomas Spatzal,1 Müge Aksoyoglu,2 Limei Zhang,3 Susana L. A. Andrade,1,4 Erik Schleicher,2 Stefan Weber,2 Douglas C. Rees,3 Oliver Einsle1,4* he enzyme nitrogenase is the only known already noticeable, and this in part biased our biological system able to break the triple analysis of the earlier structure toward nitrogen. bond of dinitrogen to yield bioavailable In Fig. 1B, plotted values do not represent inammonium (1). Its active site, the FeMo cofactor, tegrated ED but rather the average ED within the is a [Mo:7Fe:9S:X]:homocitrate cluster, the largest sphere of given radius. Alternatively, we have and most complex biological metal center known plotted the average of EDs for each exact atom to date. The exact mode and position of N2 bind- position versus its B factor (Fig. 1C and fig. S2). ing to the FeMo cofactor is unknown. The iden- Again, the different light atoms group into distification of a light atom X (C, N, or O) at its tinct areas of the plot, with the interstitial atoms center in a 1.16 Å resolution crystal structure close to the center of the carbon distribution. (2) gave rise to contradictory mechanistic proposals Although the r0 plot indicated the atom to be a that require clarification. Our analysis of the carbon both with the 1.16 Å and with the 1.0 Å diffraction data indicated that, although an un- resolution data, the ED/radius plot was far more ambiguous assignment could not be made, X was ambiguous at the lower resolution and did not allow most plausibly a nitrogen species. Through fur- us to distinguish between C and N for the central ther optimization of protein isolation and crystal- atom (3). The seemingly modest improvement in lization, we have obtained an improved structural resolution obtained with the new data set corremodel at 1.0 Å resolution (Fig. 1A) (3). In the sponds to an increase of the data-parameter ratio 1.16 Å resolution structure, the central atom was (with an anisotropic Uij temperature factor model) obscured by the geometry of the FeMo cofactor. of 50%, from 4.3 (1.16 Å) to 6.6 (1.0 Å), underOur suggestion of a nitrogen species was based on integrating electron density (ED) at the cofactor center using a probe radius of 1.4 Å, the approximate van der Waals radius of a candidate atom. We now varied this probe radius on a very fine ED grid and performed a statistical analysis for all light atoms in the structure. A plot of ED versus probe radius for the average C, N, and O atoms shows that these can be distinguished (Fig. 1B), with the distinction clearer for smaller radii. The corresponding curves for the two central atoms in the two copies of the FeMo cofactor Fig. 1. Carbon in the center of the FeMo cofactor. (A) The [Mo:7Fe:9S:C]: in the asymmetric unit homocitrate FeMo cofactor. (B) Average ED in a sphere of given radius for overlay perfectly with all carbon (black), nitrogen (blue), and oxygen (red) atoms in the structure. the curve obtained for The two central atoms in the asymmetric unit (green) closely follow the trace all other carbons. A slight for carbon. (C) ED at the atomic positions (r0) versus residual B factors deviation to higher val- shows C, N, and O to occupy distinct areas. The central atoms (green) fall ues was seen at a radius within the carbon area. (D) X-band three-pulse ESEEM frequency domain of 1.4 Å, indicating that spectra of wild-type (WT) (green), U-15N–labeled (blue), and U-13C–labeled the influence of the sur- (black) nitrogenase. (Inset) 13C resonances recorded with two different t valrounding Fe atoms is ues. FT, Fourier transform; g, g-tensor.
T
940
18 NOVEMBER 2011
VOL 334
SCIENCE
lining that the improved resolution was essential for assigning the identity of the central atom. To complement the diffraction data, we produced Azotobacter vinelandii MoFe protein labeled with 13 C or 15N, respectively, for resonance spectroscopy. The isolation procedure (3) was optimized to yield complete incorporation of the isotopes, far exceeding the labeling ratio of ~5% reported earlier (4). Inspection of wild-type and two uniformly isotopelabeled (U-15N and U-13C) nitrogenases (3) by electron spin echo envelope modulation (ESEEM), a powerful electron paramagnetic resonance technique for detection of weak hyperfine couplings (hfcs) in paramagnetic moieties such as the clusters of nitrogenase (3), revealed for the U-13C–labeled (nuclear spin quantum number = 1/2) sample an additional spectral pattern, centered at the free 13 C Larmor frequency (3.7 MHz, Fig. 1D) with a splitting of 2.5 MHz, that is not detected in the other samples. Two types of resonances can be discriminated, one originating from 13C atoms that are very weakly coupled to the paramagnetic FeMo cofactor (Fig. 1D, [I]) and the other from a more strongly coupled 13C hfc with significant unpaired electron spin density (Fig. 1D, [II]). The latter is expected for a carbon nearby or within the FeMo cofactor. This observation is consistent with X being C, because the crystal structure does not reveal any carbons within the first coordination sphere of the cluster that could account for such a large hfc. References and Notes 1. D. C. Rees et al., Philos. Trans. R. Soc. Lond. A 363, 971 (2005). 2. O. Einsle et al., Science 297, 1696 (2002). 3. Materials and methods are available as supporting material on Science Online. 4. D. Lukoyanov et al., Inorg. Chem. 46, 11437 (2007). Acknowledgments: We thank the staff at Swiss Light Source, Villigen, Switzerland, and S. M. Consiglio for assistance with the calculations. The work was supported by Deutsche Forschungsgemeinschaft (grants Ei-520/7 to O.E., An-676/1 to S.L.A.A., and IRTG 1478), Natural Sciences and Engineering Research Council of Canada postdoctoral fellowship to L.Z., NIH (grant GM45162 to D.C.R.), and HHMI (to D.C.R.). The structural model and structure factors have been deposited with the Protein Data Bank (accession code 3U7Q).
Supporting Online Material www.sciencemag.org/cgi/content/full/334/6058/940/DC1 Materials and Methods Figs. S1 and S2 Table S1 References (5–8) 15 September 2011; accepted 19 October 2011 10.1126/science.1214025 1 Institut für Organische Chemie und Biochemie, Albert-LudwigsUniversität Freiburg, 79104 Freiburg, Germany. 2Institut für Physikalische Chemie, Albert-Ludwigs-Universität Freiburg, 79104 Freiburg, Germany. 3Howard Hughes Medical Institute (HHMI), California Institute of Technology, Division of Chemistry and Chemical Engineering, Pasadena, CA 91125, USA. 4BIOSS Centre for Biological Signalling Studies, Albert-Ludwigs-Universität Freiburg, 79104 Freiburg, Germany.
*To whom correspondence should be addressed. E-mail:
[email protected]
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
BREVIA
RESEARCH ARTICLE
Sebastian Klinge,* Felix Voigts-Hoffmann,* Marc Leibundgut, Sofia Arpagaus, Nenad Ban† Protein synthesis in all organisms is catalyzed by ribosomes. In comparison to their prokaryotic counterparts, eukaryotic ribosomes are considerably larger and are subject to more complex regulation. The large ribosomal subunit (60S) catalyzes peptide bond formation and contains the nascent polypeptide exit tunnel. We present the structure of the 60S ribosomal subunit from Tetrahymena thermophila in complex with eukaryotic initiation factor 6 (eIF6), cocrystallized with the antibiotic cycloheximide (a eukaryotic-specific inhibitor of protein synthesis), at a resolution of 3.5 angstroms. The structure illustrates the complex functional architecture of the eukaryotic 60S subunit, which comprises an intricate network of interactions between eukaryotic-specific ribosomal protein features and RNA expansion segments. It reveals the roles of eukaryotic ribosomal protein elements in the stabilization of the active site and the extent of eukaryotic-specific differences in other functional regions of the subunit. Furthermore, it elucidates the molecular basis of the interaction with eIF6 and provides a structural framework for further studies of ribosome-associated diseases and the role of the 60S subunit in the initiation of protein synthesis. n all domains of life, protein synthesis is catalyzed by ribosomes. These giant ribonucleoprotein particles consist of two subunits with distinct functions. Information contained in mRNA is decoded by the small ribosomal subunit, whereas peptide bond formation is mediated by the RNA component of the large ribosomal subunit (1). Ribosome-associated factors bind to the large ribosomal subunit and interact with nascent polypeptides emerging from the ribosomal tunnel (2). In addition, the large subunit is a target for antibiotics that interfere with the peptidyl transferase reaction and with the progression of the nascent polypeptide chain through the tunnel (3). Over the past decade, our knowledge of prokaryotic translation has advanced with the publication of the crystal structures of the small (30S) and large (50S) ribosomal subunits as well as the complete (70S) prokaryotic ribosome (1, 4–6). In recent years, structures of prokaryotic ribosomes in complex with protein factors involved in translation have provided us with molecular snapshots of various stages of protein synthesis at atomic resolution (1, 7–9). However, our molecular understanding of protein synthesis in eukaryotes remains incomplete. Eukaryotic ribosomes are considerably larger than their bacterial counterparts. Both eukaryotic ribosomal subunits contain numerous RNA expansion segments, which coevolved with many additional eukaryotic-specific ribosomal protein elements. As a consequence, the eukaryotic 60S
I
Institute of Molecular Biology and Biophysics, ETH Zürich, 8092 Zürich, Switzerland. *These authors contributed equally to this work. †To whom correspondence should be addressed. E-mail:
[email protected]
subunit in yeast or T. thermophila has a total molecular weight of about 2 million daltons, whereas that of the 50S subunit in Escherichia coli is 1.3 million daltons. The increased level of structural complexity of eukaryotic ribosomes reflects functional differences between prokaryotes and eukaryotes. First, ribosome biogenesis in eukaryotic cells is elaborate. It takes place in different cellular compartments and involves about 200 trans-acting proteins in the processing and modification of ribosomal RNA (rRNA), the import of ribosomal proteins into the nucleus, and the export of preribosomal subunits into the cytoplasm (10). Second, the regulation of protein synthesis is much more complex in eukaryotes and at the level of initiation mostly involves the 40S ribosomal subunit. Recently, cryoelectron microscopy (cryo-EM) reconstructions of yeast and wheat germ 80S ribosomes at 5.5 to 6.1 Å resolution and a partial interpretation of crystallographic data from the 80S yeast ribosome at 4.15 Å resolution have provided us with the positions and topology of eukaryotic RNA expansion segments and several proteins for which homologous structures are known (11–13). However, in these studies it was not possible to correctly assign and build many eukaryotic proteins. The recent crystal structure of the T. thermophila small ribosomal subunit (40S) in complex with eIF1 represents the first complete atomic model of the eukaryotic 40S ribosomal subunit (14), whereas a corresponding structure of the large ribosomal subunit (60S) has so far remained elusive. The 60S ribosomal subunit is subject to several regulatory processes during initiation. Binding of eukaryotic initiation factor 6 (eIF6) to the large ribosomal subunit inhibits subunit joining and thus
www.sciencemag.org
SCIENCE
VOL 334
18 NOVEMBER 2011
Downloaded from www.sciencemag.org on November 17, 2011
Crystal Structure of the Eukaryotic 60S Ribosomal Subunit in Complex with Initiation Factor 6
prevents translation initiation (15). The crystal structures of isolated eIF6 and its archaeal homolog aIF6 in isolation have highlighted its pentameric shape (16). The mechanism by which eIF6 prevents 80S complex formation is most likely steric hindrance; however, the exact interaction between eIF6 (or aIF6) and the large ribosomal subunit is currently unclear, as chemical probing and lowresolution EM data have provided conflicting evidence (17, 18). In addition to its function as an anti-association factor, eIF6 has also been implicated in 60S maturation (19, 20). Crystallization and structure determination. We present the crystal structure of the T. thermophila 60S subunit in complex with eIF6 (Fig. 1). Cocrystallization with the antibiotic cycloheximide yielded crystals diffracting to 3.5 Å resolution and permitted visualization of its binding site. Bulky and ordered side chains are clearly visible, and the registry for all well-ordered regions of the structure is unambiguous. We have also included in the model some flexible solvent-exposed loops of proteins for which the sequence register is less reliable because it was established by extrapolation from the last identifiable residue. Detailed crystallization and structure determination procedures are provided in table S1 and (21). Table S2 lists all proteins included in this structure, their UNIPROTcodes, and homologs from yeast, bacteria, and archaea. Throughout this article the human nomenclature is used for all ribosomal proteins; where the yeast nomenclature differs, the name of the yeast homolog is included in parentheses (Fig. 1, C and D, and Fig. 2) (22). A PYMOL script that displays two Protein Data Bank (PDB) files corresponding to one complete 60S subunit and labels all proteins according to the human, yeast, and E. coli nomenclatures is available (21). Structure of the 26S and 5.8S rRNAs. The T. thermophila 60S ribosomal subunit contains three rRNA molecules (5S, 5.8S, and 26S rRNA). As a result of differences in pre-rRNA processing in eukaryotes, the 5.8S rRNA occupies the same region as the 5′ end of the bacterial 23S rRNA (figs. S1 and S2) (23). The core structure of the 5.8S and 26S rRNA is highly homologous to the 23S rRNA seen in archaea and bacteria or the 5.8S and 25S rRNA in yeast. However, despite the similarities in overall topology, the expansion segments (ESs) cannot be readily superimposed because of differences in their sequences (fig. S3) (13). With the exception of ES31 and ES41, RNA expansion segments do not occupy the ribosomal subunit interface (Fig. 1, A and B). Two clusters of RNA expansion segments serve as binding platforms for eukaryotic-specific proteins or eukaryotic-specific extensions of conserved ribosomal proteins. The first cluster is composed of ES7 and ES39 and, to a lesser extent, ES9 and ES12 (fig. S4A). ES7 is the largest expansion segment of the 60S ribosomal subunit and can be divided into three parts (helices ES7A, ES7B, and ES7C). ES7A is evolutionarily variable in length and
941
bridges the distance between ribosomal proteins RPL28 and RPL13 (Fig. 3, A and B). In most eukaryotes ES7A is stabilized by RPL28, whereas in budding yeast RPL28 is missing and ES7A is considerably shorter (11, 12). ES7B is positioned orthogonally with respect to ES7A, and the two helices thus define boundaries around the 5S rRNA and the juxtaposed ES9 and ES12 (Fig. 1A and Fig. 3A). ES7C protrudes from between ES7A and ES7B and is involved in several RNA-protein interactions (Fig. 1A and Fig. 3, A to D). ES39 is composed of two helices linked by an internal loop, which facilitates a sharp bend between the helices while also packing against the surface of the 60S ribosomal subunit. The opened loop structure also serves as a binding platform for the eukaryotic-specific protein RPL6 (Fig. 3, C and D).
The second cluster is formed by ES5, ES19, ES31, ES20, and ES26, with outer boundaries of this region defined by helices H18 and H58 (fig. S4B). An internal loop between H18 and ES5 facilitates a sharp bend, which allows ES5 to be sandwiched between H18 and ES4 (the 5′ end of the 26S rRNA and the 3′ end of the 5.8S rRNA). ES19 is surrounded from one side by ES4 and ES3 of the 5.8S rRNA and from the other side by ES31, which contains a three-way junction, as well as ES20 and ES26, which together with ES31 create a binding platform for RPL27 (fig. S5, A and B). Ribosomal proteins. The structure of the 60S ribosomal subunit contains 42 proteins of which 16 are present in all domains of life, 20 are shared between eukarya and archaea, and 6 are eukaryotic-specific (Fig. 2). Most ribosomal proteins contain eukaryotic-specific extensions, which
are critical for establishing an intricate proteinRNA network (Fig. 3). Some of these extensions approach the conserved core regions of the ribosome, such as the active site and the exit tunnel (Fig. 4). The eukaryotic-specific proteins RPL22, RPL29, RPL36, and RPL28, as well as the eukaryotic/archaeal proteins RPL13, RPL34, and RPL38, are not homologous to other ribosomal proteins of known structure and hence expand the range of protein folds found in ribosomes (Fig. 2). In comparison to the prokaryotic 50S ribosomal subunits, the eukaryotic 60S subunit contains a vast network of additional protein-protein and protein-RNA interactions. These are particularly striking around ES7 and ES39, which form a central nexus on the back of the large ribosomal subunit (Fig. 1, A and C, and Fig. 3). The structure of the 60S subunit also reveals some additional
Fig. 1. Architecture of the 60S ribosomal subunit. (A and B) Views of the solvent-exposed (A) and 40S binding side of the 60S subunit (B) with color-coded RNA expansion segments. (C and D) Views of the solvent-exposed (C) and 40S binding side of the 60S subunit (D) with color-coded ribosomal proteins shown as ribbons.
942
18 NOVEMBER 2011
VOL 334
SCIENCE
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
RESEARCH ARTICLE
RESEARCH ARTICLE specific extension of RPL13A (L16). Positioned on top of ES39 is RPL6, which acts as a cornerstone by contacting several RNA expansion segments (ES7A, ES7B, ES7C, and ES39) (Fig. 3, C and D). The interaction between RPL6 and ES39 is further stabilized by C-terminal helical extensions from RPL14 and RPL13A (L16) (Fig. 2 and Fig. 3, C and D). In the vicinity of ES39, three SH3 domains are used for both protein-protein and proteinRNA interactions (Fig. 3, C and D). Although other SH3 domain–containing proteins are scattered throughout the 60S subunit, these three proteins form an archaeal/eukaryotic-specific cluster. Between the two RNA clusters, long terminal extensions of RPL13 and RPL36 mediate contacts among the tip of ES7A, ES9, and the base of ES5 (fig. S5C). Mutations of RPL36, RPL5, RPL11, and RPL35A (L33) have been associated with Diamond-Blackfan anemia, a congenital red blood cell aplasia, which can also be caused by mutations in 40S ribosomal proteins (24). During ribosome biogenesis, the 5S rRNA is assembled into the pre-60S subunit together with its flanking proteins RPL5 and RPL11 (fig. S5D). Therefore, mutations in these proteins most likely result in ribosome assembly defects (25, 26). In analogy to the first RNA cluster, an extensive protein network exists in the vicinity of
Fig. 2. Evolutionary representation of ribosomal proteins of the 60S subunit. 60S large ribosomal subunit proteins are colored according to conservation. Protein cores found in all kingdoms are depicted in light blue, proteins with www.sciencemag.org
SCIENCE
the second expansion segment area around ES31, ES20, and ES26 (fig. S4B). Anchored between helices H55 and H58, the archaeal/eukaryoticspecific protein RPL34 adopts an extended topology and harbors a zinc finger fold (figs. S4B, S5, and S6). Whereas its N terminus is deeply buried within the ribosomal core, its C-terminal helical half projects toward the surface of the 60S subunit, where it interacts with RNA expansion segments ES20 and ES31 as well as with the SH3 domain–containing, eukaryotic-specific ribosomal protein RPL27 (fig. S5, A and B). Positioned on top of this protein-RNA surface, RPL27 is involved in the interaction with two kink-turn binding proteins, RPL7A (L8, homologous to archaeal L7ae) and RPL30. Both RPL7A (L8) and RPL30 have additional functions apart from their ribosomal association. Previous studies have shown how yeast RPL30 recognizes its own pre-mRNA to prevent splicing (27, 28). Furthermore, archaeal L7ae is incorporated into the H/ACA ribonuclear protein family of pseudouridine synthase complexes, which use guide RNAs for targeted pseudo-uridylation (27–30) (fig. S7). RPL38 and RPL22 are localized in the vicinity of RPL27 and the exit tunnel (Fig. 1). RPL38 has recently been shown to have a role in the selective recruitment of the Hox gene mRNAs to the 80S ribosomes in mice (31), whereas RPL22
Downloaded from www.sciencemag.org on November 17, 2011
architectural features not previously observed in bacteria or archaea. Long helices of eukaryoticspecific extensions of RPL4 and RPL7 extend above the surface of the subunit and form cranelike structures, bridging numerous eukaryoticspecific RNA expansion segments and proteins (Fig. 3, A and B). These features have not been observed in the 70S ribosome, where tertiary protein-RNA interactions are mostly mediated by positively charged extensions of ribosomal proteins that weave between the rRNA. Additionally, several proteins of the 60S subunit form a network of interactions by forming interprotein b sheets (Fig. 3B). In this regard, the eukaryoticspecific extension of RPL21 is used as a central mediator with two functions: First, it links the bridging helices by forming secondary structure elements with both RPL7 and RPL18A (L20). Second, together with eukaryotic-specific protein RPL29, it sandwiches ES12 and thus anchors it on the surface of the 60S subunit (Fig. 3, A and B). RPL18A (L20) also has a second function in the stabilization of ES39 and its associated archaeal/eukaryotic ribosomal proteins, three of which [RPL6, RPL13A (L16), and RPL14] contain Src homology 3 (SH3) domains (Fig. 3, C and D). It contacts ES39 and also forms a cradle, in which RPL14 is oriented toward ES39 and a eukaryotic-
archaeal homologs are in gold, and proteins or protein extensions unique to eukaryotes are in red. Positions of N and C termini are indicated. Zinc ions are shown as green spheres. VOL 334
18 NOVEMBER 2011
943
can interact with noncoding RNAs of the EpsteinBarr virus (fig. S8) (21, 32). With the crystal structures of the 40S and 60S ribosomal subunit from T. thermophila, we can now visualize the evolutionary expansion of the ribosome from bacteria to eukaryotes (fig. S9) (14). This is particularly evident for the coevolved areas around ES6 of the 40S subunit and the RNAprotein cluster in the vicinity of ES31 of the large subunit, where helical protein extensions emerge from the 60S subunit and play a role in intersubunit contacts (fig. S9) (13). Hence, long a-helical extensions emerge as mediators of long-range interactions within the 60S subunit as well as intersubunit contacts with the 40S subunit (Fig. 3 and fig. S9). The peptidyl transferase center, the exit tunnel, and antibiotic binding sites. The 26S rRNA in the vicinity of the peptidyl transferase center is highly conserved throughout all kingdoms of life. Like the 23S rRNA in the 50S subunit, the 26S rRNA contains all regions essential for catalysis
and substrate binding including the A-, P-, and Esites, the 8 o’clock helix (helix 93), and the Ploop (5, 6). As a result, the rRNA structure of the active site can be readily superimposed with its bacterial counterpart from Thermus thermophilus (Fig. 4A). At the level of proteins, however, important differences can be observed with respect to their possible roles in the stabilization of the rRNA (Fig. 4, A and B). Eukaryotic-specific protein RPL29 extends from the surface of the 60S subunit to within 17 Å of the active-site adenosine A2808 (A2820 in yeast) (Fig. 4, A and B). Although genetic studies in yeast and mice have shown that RPL29 is not essential, deletion of this protein leads to reduced growth, presumably as a result of protein synthesis defects in both organisms (31, 33, 34). Interestingly, several other proteins (RPL4, RPL10, and RPL21) that are also located in close proximity to the active site are additionally interconnected on the surface of the 60S subunit by eukaryotic-specific extensions (Fig. 4B).
Fig. 3. Architectural features of 60S ribosomal proteins. (A) Overview of ribosomal proteins in the vicinity of ES7 and ES39. Conserved proteins are shown in white; eukaryotic-specific proteins or protein extensions are colorcoded. (B) Schematic representation of long-range and quaternary inter-
944
18 NOVEMBER 2011
VOL 334
Prokaryotic and eukaryotic ribosomes are targets for numerous antibiotics. Cycloheximide and other glutarimide antibiotics are inhibitors of protein biosynthesis with high specificity for large eukaryotic ribosomal subunits (35, 36). This family of molecules is thought to inhibit translation elongation by interfering with deacetylated tRNA in the exit site of the ribosome, but the precise position of their binding site is currently limited by a lack of structural information (35, 37). Difference electron density maps calculated with our structure reveal that cycloheximide binds in a tight pocket on the 60S subunit (Fig. 4, C and D, and fig. S10) that was previously identified as the binding site for nucleotides C75 and A76 of E-site tRNA in the archaeal ribosome (38). Although the shape and size of the difference density agree very well with the bilobal structure of cycloheximide, we have not modeled it into the density because we cannot unambiguously assign the orientation of the molecule at 3.5 Å (Fig. 4, C and D). Our data are in
actions of ribosomal proteins illustrated in (A). (C) Detailed view of ribosomal proteins interacting with ES39. (D) Schematic representation of the architecture of SH3 domain–containing proteins in the vicinity of ES39. SH3 domains are shown as large cylinders. SCIENCE
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
RESEARCH ARTICLE
Downloaded from www.sciencemag.org on November 17, 2011
RESEARCH ARTICLE
Fig. 4. Architecture of the peptidyl transferase center, the binding site of cycloheximide, and the exit tunnel of the 60S subunit. (A) Superposition of the T. thermophila 26S RNA active-site region (light blue) with T. thermophilus 23S rRNA (PDB code 2WDL, gray). Highlighted RNA elements are the activesite adenosine A2808 (bacterial A2451, red), the 8 o’clock helix (yellow), the P-loop (pink), and the T. thermophilus P-site tRNA (green). The N terminus of RPL29 is displayed as a purple sphere. (B) Architecture of ribosomal proteins in close vicinity to the active site. Eukaryotic-specific extensions of RPL4 (blue), RPL21 (yellow), RPL10 (green), and RPL3 (orange) are colored in a lighter shade. The locations of the peptidyl transferase center (PTC; red), RPL8 (L2) (pink), and RPL29 (purple) are indicated. (C and D) Fobs – Fcalc difference Fourier map (green, contoured at 3.5s) showing the binding site of cycloheximide at the tRNA E-site of the 60S subunit. Shown in (C) is a view from the inside of the subunit. The difference density reveals that the binding site of cycloheximide superimposes with A76 of the E-site tRNA (purple), based on superposition with the archaeal 50S subunit in complex with E-site www.sciencemag.org
SCIENCE
tRNA mimic (38). The mutations in RPL27A (L28) (magenta) and RPL36A (L42) (marine) result in cycloheximide resistance. The rRNA is colored gray, with the proximal base C2754 (C2765 in yeast) in orange. Shown in (D) is a view toward the tRNA E-site [opposite direction relative to (C)]. The difference density reveals that the shape and the size of the density matches the molecular structure of cycloheximide shown in the inset. (E) Clipped view of the T. thermophila 60S ribosomal subunit showing the polypeptide exit tunnel with rRNA and proteins as gray and light blue surfaces, respectively. RPL4 (marine) and its eukaryotic-specific extensions at the surface and in the exit tunnel (red) are shown as ribbons. Superimposed elements include an aminoacylated P-site tRNA (green, PDB codes 2WDL and 2WDK) and the macrolide antibiotic erythromycin (yellow, PDB code 1YI2). The position of the P-site aminoacyl moiety of the tRNA is shown as a green sphere. (F) Conservation of RNA and protein elements around the exit tunnel. Eukaryoticspecific RNA and protein elements are color-coded; conserved regions are in gray. VOL 334
18 NOVEMBER 2011
945
agreement with the observation that point mutations in RPL36A (L42) and RPL27A (L28) in Chlamydomonas reinhardtii and budding yeast can result in resistance to cycloheximide (39, 40). Similarly, RNA footprinting analysis showed that a single cytosine (C2754; C2765 in yeast) is in the vicinity of the cycloheximide binding site (35). Therefore, despite their very different chemical structures, the mechanism of cycloheximide inhibition of eukaryotic protein synthesis may be related to the mechanism of mycalamide A and 13-deoxytedanolide, natural products with potent antitumor activities that sterically hinder tRNA binding to the exit site of the large subunit (41, 42). Several antibiotics target the exit tunnel of prokaryotic ribosomes (43). The interior of the 60S ribosomal exit tunnel is mostly conserved, with some notable differences relative to the prokaryotic ribosomes. The site of constriction in the exit tunnel 20 Å below the peptidyl transferase center contains a highly conserved eukaryoticspecific insertion in the distal loop of RPL4, which interacts with helix 23 of the 26S rRNA (Fig. 4E and fig. S11). Hence, changes in the shape of the exit tunnel may result in reduced accessibility of the exit tunnel for larger macrolides. Interestingly, the recombinant insertion of six amino acids in the analogous region of bacterial L4, although too far away to directly contact the antibiotic, was shown to result in resistance to erythromycin and negatively affected SecMmediated stalling of protein synthesis (44, 45). Another critical difference in the rRNA structure in this region of the ribosomal tunnel is the presence of G at position 2395 of the 26S rRNA (G2400 in yeast), which corresponds to A2058 in E. coli and G2099 in H. marismortui. The A2058G mutation in E. coli ribosomes prevents
erythromycin binding, whereas the G2099A mutation in H. marismortui promotes it, implying that erythromycin binding depends on an A in this position (fig. S12) (46). The region of the 60S ribosomal subunit at the tunnel exit is mostly conserved, with eukaryotic features restricted to the second tier of proteins surrounding the exit (Fig. 4F). Eukaryotic elements that could affect the binding of factors interacting with emerging nascent chains are the small ES24 of the 26S rRNA and the eukaryoticspecific protein RPL22 (Fig. 4F). Eukaryotic ES27 has been seen in EM reconstructions in two conformations, one of which positions it directly above the tunnel exit (11, 12). We observe weak electron density, insufficient to build an atomic model, corresponding to ES27 in the conformation above the tunnel in one of the four 60S molecules. In agreement with the observed flexibility, it has been shown that ES27 swings away from the tunnel exit area toward the L1 protuberance of the 60S subunit when the eukaryotic ribosome interacts with the protein-conducting channel (47). In general, the surface surrounding the tunnel exit is very flat, with no bulky eukaryotic features added in this region; this may be attributable to the requirement of the ribosome to approach the flat surface of the membrane during synthesis of membrane proteins (fig. S13). RPL40—a ribosomal ubiquitin fusion protein. RPL40 is a zinc finger protein expressed as a ubiquitin fusion (figs. S6 and S14) (48). The structure of the 60S subunit reveals that RPL40 is positioned in close proximity to the sarcin-ricin loop and the elongation factor binding site (Fig. 5, A and B). The position of the unprocessed, N-terminal ubiquitin domain would sterically prevent elongation factor binding to the large ri-
Fig. 5. Positions of ribosomal ubiquitin fusion proteins. (A) Model of the T. thermophila 80S ribosome from a superposition of 40S (PDB code 2XZM) and 60S ribosomal subunits with EF-Tu–bound 70S (PDB codes 2WRN and 2WRO). EF-Tu is shown in yellow, bound tRNA in green, 40S rRNA and proteins in orange, and 60S rRNAs and proteins in light blue. RPS27A (S31)
946
18 NOVEMBER 2011
VOL 334
bosomal subunit. This is analogous to the role of the N-terminal ubiquitin domain of RPS27A (S31), which extends into the A site of the decoding center on the small ribosomal subunit and would prevent tRNA binding (14). It was previously shown that the cleavage of ubiquitin from RPS27A (S31) is essential for 40S maturation (48). The observation that the ubiquitin domain of RPL40 also occupies a functionally important region of the 60S subunit may suggest a mechanism by which fused ubiquitin domains could prevent immature ribosomal subunits from entering the translational cycle until they are processed. However, the exact timing of this processing step is currently unclear. A superposition of the small and large ribosomal subunits from T. thermophila with the crystal structure of a prokaryotic ribosome with bound elongation factor Tu (EF-Tu) and tRNA (7) demonstrates that both N-terminal ubiquitin fusions are ideally positioned to block an important functional center in each of the ribosomal subunits (Fig. 5A). Eukaryotic initiation factor 6. eIF6 is conserved from archaea to eukarya and has been implicated in the inhibition of 80S complex formation as well as 60S maturation (15, 49). Studies in yeast demonstrated that the highly conserved Shwachman-Bodian-Diamond syndrome (SBDS) protein and the guanosine triphosphatase (GTPase) Efl1 are involved in the removal of eIF6 from late preribosomal 60S subunits (19). SBDS acts as an adaptor by physically coupling the GTP hydrolysis of Efl1 with the release of eIF6 (50). Because all three proteins act in the same pathway, the deletion of SBDS in yeast and the resulting localization of eIF6 to cytoplasmic pre-60S particles can be suppressed by a set of eIF6 mutants. Most mutations result in reduced affinity of the factor for the large
and RPL40 are displayed as red ribbons, with N termini of the processed proteins highlighted as red spheres to indicate the positions of the ubiquitin moieties. (B) Detailed view of RPL40 (red cartoon) with bound zinc (light green) in close proximity to the sarcin-ricin loop (SRL; pink spheres), with surrounding RNA and proteins in blue. SCIENCE
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
RESEARCH ARTICLE
RESEARCH ARTICLE protein family (51). The C terminus of RPL23 mediates the interaction with the major binding surface of eIF6, which is centrally positioned above it (Fig. 6B). Both hydrophobic and hydrophilic residues on the surface of eIF6 are involved in the interaction with RPL23, with a buried surface area of 870 Å2 between the two proteins. With the crystal structure of the 60S ribosomal subunit in complex with eIF6, we can now rationalize previous biochemical data because the electron density for the contact region is of good quality, with many side-chain features clearly visible (fig. S15). Residues critical for the association of eIF6 with the 60S subunit fall into three classes: hydrogen-bonding residues, hydrophobic residues,
and residues that most likely contribute to the structural integrity of eIF6 (19). The arrangement of residues on the surface of RPL23 in close proximity to eIF6 mirrors the positions of residues on the surface of eIF6, which have been shown to be critical for the interaction with the 60S subunit (Fig. 6, C and D) (19). Conclusion. The crystal structure of the large eukaryotic ribosomal subunit provides a complete atomic description of this central assembly in eukaryotic cells. It reveals a high degree of conservation of the active site and small but important differences in the features of the ribosomal tunnel. Clustered in different regions of the 60S subunit, RNA expansion segments and eukaryotic
Downloaded from www.sciencemag.org on November 17, 2011
ribosomal subunit and have been mapped to the putative 60S binding face of eIF6 (19). In agreement with earlier low-resolution cryoEM studies, we find eIF6 bound to RPL23 and in close proximity to the sarcin-ricin loop, where it would prevent binding of the 40S subunit (Fig. 6, A and B) (18). These results are in contrast to chemical probing results with Sulfolobus solfataricus 50S and aIF6 that suggested a contact between aIF6 and helix 69, which in our structure is positioned 75 Å away from eIF6 (17). As a member of the pentein protein family, eIF6 has a five-fold pseudosymmetry and two large flat surfaces, the larger of which commonly serves as a scaffold for active-site residues in other members of the
Fig. 6. Eukaryotic initiation factor 6 bound to the 60S subunit. (A) Top view of eIF6 (orange surface) on the T. thermophila 60S large ribosomal subunit. (B) Side view of eIF6 bound to the C terminus of RPL23 (blue) in proximity to the sarcin-ricin loop (SRL; pink surface) and RPL24 (light green). N and C termini of RPL23 are indicated. (C) Surface view of RPL23 (blue) and the sarcin-ricin loop (pink) with residues within 3.7 Å of eIF6 highlighted as www.sciencemag.org
SCIENCE
white spheres. (D) Surface view of eIF6 (orange) with color-coded mutations known to interfere with 60S binding. Hydrophilic residues (red), hydrophobic residues (light blue), and residues required for the structural integrity of eIF6 (pale yellow) are indicated. Single-letter abbreviations for residues: A, Ala; C, Cys; D, Asp; G, Gly; I, Ile; K, Lys; L, Leu; N, Asn; P, Pro; Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; Y, Tyr. VOL 334
18 NOVEMBER 2011
947
References and Notes 1. T. M. Schmeing, V. Ramakrishnan, Nature 461, 1234 (2009). 2. G. Kramer, D. Boehringer, N. Ban, B. Bukau, Nat. Struct. Mol. Biol. 16, 589 (2009). 3. D. Bulkley, C. A. Innis, G. Blaha, T. A. Steitz, Proc. Natl. Acad. Sci. U.S.A. 107, 17158 (2010). 4. B. T. Wimberly et al., Nature 407, 327 (2000). 5. N. Ban, P. Nissen, J. Hansen, P. B. Moore, T. A. Steitz, Science 289, 905 (2000). 6. M. Selmer et al., Science 313, 1935 (2006). 7. T. M. Schmeing et al., Science 326, 688 (2009). 8. S. Petry et al., Cell 123, 1255 (2005). 9. R. Bingel-Erlenmeyer et al., Nature 452, 108 (2008). 10. V. G. Panse, A. W. Johnson, Trends Biochem. Sci. 35, 260 (2010). 11. J. P. Armache et al., Proc. Natl. Acad. Sci. U.S.A. 107, 19748 (2010). 12. J. P. Armache et al., Proc. Natl. Acad. Sci. U.S.A. 107, 19754 (2010). 13. A. Ben-Shem, L. Jenner, G. Yusupova, M. Yusupov, Science 330, 1203 (2010). 14. J. Rabl, M. Leibundgut, S. F. Ataide, A. Haag, N. Ban, Science 331, 730 (2011). 15. M. Ceci et al., Nature 426, 579 (2003). 16. C. M. Groft, R. Beckmann, A. Sali, S. K. Burley, Nat. Struct. Biol. 7, 1156 (2000).
17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45.
D. Benelli et al., Nucleic Acids Res. 37, 256 (2009). M. Gartmann et al., J. Biol. Chem. 285, 14848 (2010). T. F. Menne et al., Nat. Genet. 39, 486 (2007). K. Y. Lo et al., Mol. Cell 39, 196 (2010). See supporting material on Science Online. A. Nakao, M. Yoshihama, N. Kenmochi, Nucleic Acids Res. 32 (database issue), D168 (2004). B. S. Strunk, K. Karbstein, RNA 15, 2083 (2009). H. T. Gazda et al., Am. J. Hum. Genet. 83, 769 (2008). J. B. Moore 4th, J. E. Farrar, R. J. Arceci, J. M. Liu, S. R. Ellis, Haematologica 95, 57 (2010). J. Zhang et al., Genes Dev. 21, 2580 (2007). J. A. Chao, J. R. Williamson, Structure 12, 1165 (2004). S. Macías, M. Bragulat, D. F. Tardiff, J. Vilardell, Mol. Cell 30, 732 (2008). M. Halic, T. Becker, J. Frank, C. M. Spahn, R. Beckmann, Nat. Struct. Mol. Biol. 12, 467 (2005). L. Li, K. Ye, Nature 443, 302 (2006). N. Kondrashov et al., Cell 145, 383 (2011). J. L. Houmani, C. I. Davis, I. K. Ruf, J. Virol. 83, 9844 (2009). M. L. DeLabre, J. Kessl, S. Karamanou, B. L. Trumpower, Biochim. Biophys. Acta 1574, 255 (2002). C. B. Kirn-Safran et al., Dev. Dyn. 236, 447 (2007). T. Schneider-Poetsch et al., Nat. Chem. Biol. 6, 209 (2010). H. M. Fried, J. R. Warner, Nucleic Acids Res. 10, 3133 (1982). T. V. Pestova, C. U. Hellen, Genes Dev. 17, 181 (2003). T. M. Schmeing, P. B. Moore, T. A. Steitz, RNA 9, 1345 (2003). D. R. Stevens, A. Atteia, L. G. Franzén, S. Purton, Mol. Gen. Genet. 264, 790 (2001). N. F. Käufer, H. M. Fried, W. F. Schwindinger, M. Jasin, J. R. Warner, Nucleic Acids Res. 11, 3123 (1983). G. Gürel, G. Blaha, T. A. Steitz, P. B. Moore, Antimicrob. Agents Chemother. 53, 5010 (2009). S. J. Schroeder, G. Blaha, P. B. Moore, Antimicrob. Agents Chemother. 51, 4462 (2007). A. Yonath, Annu. Rev. Biochem. 74, 649 (2005). S. Zaman, M. Fitzpatrick, L. Lindahl, J. Zengel, Mol. Microbiol. 66, 1039 (2007). M. G. Lawrence, L. Lindahl, J. M. Zengel, J. Bacteriol. 190, 5862 (2008).
46. D. Tu, G. Blaha, P. B. Moore, T. A. Steitz, Cell 121, 257 (2005). 47. R. Beckmann et al., Cell 107, 361 (2001). 48. T. Lacombe et al., Mol. Microbiol. 72, 69 (2009). 49. A. W. Johnson, S. R. Ellis, Genes Dev. 25, 898 (2011). 50. A. J. Finch et al., Genes Dev. 25, 917 (2011). 51. B. Hartzoulakis et al., Bioorg. Med. Chem. Lett. 17, 3953 (2007). Acknowledgments: All data were collected at the Swiss Light Source (SLS, Paul Scherrer Institut, Villigen). We thank T. Tomizaki, M. Müller, V. Olieric, G. Pompidor, and A. Pauluhn for their outstanding support at the SLS; J. Rabl for advice on cell growth and ribosome purification as well as a clone of eIF6; T. Maier for advice on data collection; T. Bucher for the preparation of crystals; J. Erzberger and T. Bucher for critically reading the manuscript; and all members of the Ban laboratory for suggestions and discussions. Supported by the Swiss National Science Foundation (SNSF), the National Center of Excellence in Research (NCCR) Structural Biology program of the SNSF, and European Research Council grant 250071 under the European Community’s Seventh Framework Programme (N.B.) and by EMBO and Human Frontier Science Program fellowships (S.K.). Coordinates and structure factors have been deposited in the Protein Data Bank (accession codes for molecule 1: 4A1E and 4A18; molecule 2, 4A17 and 4A19; molecule 3, 4A1A and 4A1B; molecule 4, 4A1C and 4A1D). ETH Zürich has filed a patent application to use the crystals and the coordinates of the 60S ribosomal subunit for developing compounds that can interfere with eukaryotic translation.
Supporting Online Material www.sciencemag.org/cgi/content/full/science.1211204/DC1 Materials and Methods SOM Text Figs. S1 to S21 Tables S1 and S2 References (52–72) 14 July 2011; accepted 5 October 2011 Published online 3 November 2011; 10.1126/science.1211204
REPORTS The Large, Oxygen-Rich Halos of Star-Forming Galaxies Are a Major Reservoir of Galactic Metals J. Tumlinson,1* C. Thom,1 J. K. Werk,2 J. X. Prochaska,2 T. M. Tripp,3 D. H. Weinberg,4 M. S. Peeples,5 J. M. O’Meara,6 B. D. Oppenheimer,7 J. D. Meiring,3 N. S. Katz,3 R. Davé,8 A. B. Ford,8 K. R. Sembach1 The circumgalactic medium (CGM) is fed by galaxy outflows and accretion of intergalactic gas, but its mass, heavy element enrichment, and relation to galaxy properties are poorly constrained by observations. In a survey of the outskirts of 42 galaxies with the Cosmic Origins Spectrograph onboard the Hubble Space Telescope, we detected ubiquitous, large (150-kiloparsec) halos of ionized oxygen surrounding star-forming galaxies; we found much less ionized oxygen around galaxies with little or no star formation. This ionized CGM contains a substantial mass of heavy elements and gas, perhaps far exceeding the reservoirs of gas in the galaxies themselves. Our data indicate that it is a basic component of nearly all star-forming galaxies that is removed or transformed during the quenching of star formation and the transition to passive evolution.
G 948
alaxies grow by accreting gas from the intergalactic medium (IGM) and converting it to stars. Stellar winds and explo-
sions release gas enriched with heavy elements [or metals (1)], some of which is ejected in galactic-scale outflows (2). The circumgalactic
18 NOVEMBER 2011
VOL 334
SCIENCE
medium (CGM)—loosely defined as gas surrounding galaxies within their own halos of dark matter (out to 100 to 300 kpc)—lies at the nexus of accretion and outflow, but the structure of the CGM and its relation to galaxy properties are still uncertain. Galactic outflows are observed at both low (2–4) and high (5–7) redshift, but it is unclear how far they propagate, what level of heavy-element enrichment they possess, and whether the gas escapes the halo or eventually returns to fuel later star formation. Models of 1
Space Telescope Science Institute, Baltimore, MD 21218, USA. University of California Observatories–Lick Observatory, Santa Cruz, CA 95064, USA. 3Department of Astronomy, University of Massachusetts, Amherst, MA 01003, USA. 4Department of Astronomy, Ohio State University, Columbus, OH 43210, USA. 5 Department of Physics and Astronomy, University of California, Los Angeles, CA 90095, USA. 6Department of Chemistry and Physics, Saint Michael’s College, Colchester, VT 05439, USA. 7Leiden Observatory, Leiden University, NL-2300 RA Leiden, Netherlands. 8Steward Observatory, University of Arizona, Tucson, AZ 85721, USA. 2
*To whom correspondence should be addressed. E-mail:
[email protected]
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
ribosomal protein elements have coevolved to form architectural complexes that mediate longdistance tertiary interactions. The structure also offers insights into the regulatory mechanisms of the eukaryotic ribosome, the maturation of the large ribosomal subunit, principles of antibiotic specificity and of the mechanism of translation inhibition by cycloheximide, and the structural basis of extraribosomal roles of various eukaryoticspecific ribosomal proteins. As such, it provides a starting point for future biochemical, genetic, and structural studies of protein synthesis in eukaryotes.
REPORTS Sun. The QSO sightlines probe projected radial distances to the galaxies (i.e., impact parameters) of R = 14 to 155 kpc. We used the COS data to measure the O VI column densities (NOVI in cm−2), line profiles, and velocities with respect to the target galaxies (Fig. 1) (21). We measured the precise redshift, star formation rate (SFR in M◉ year−1), and metallicity for each of our sample galaxies by means of low-resolution spectroscopy from the Keck Observatory Low-Resolution Imaging Spectrograph (LRIS) and the Las Campanas Observatory Magellan Echellette (MagE) spectrograph (21, 22). Our systematic sampling of galaxy properties allows us to investigate the connection between galaxies themselves and the CGM. The O VI detections extend to R = 150 kpc away from the targeted galaxies, but the whole sample shows no obvious trend with radius R (Fig. 2). The strong clustering of detections within T200 km s−1 of the
Downloaded from www.sciencemag.org on November 17, 2011
the Hubble Space Telescope to directly map the CGM by absorption-line spectroscopy, in which a diffuse gas is detected by its absorption of light from a background source. Our background sources are ultraviolet-bright quasi-stellar objects (QSOs), which are the luminous active nuclei of galaxies lying far behind the galaxies of interest. We focus on the ultraviolet 1032, 1038 Å doublet of O VI (O+5), the most accessible tracer of hot and/or highly ionized gas at redshift z < 0.5. O VI has been used to trace missing baryons in the IGM (13–16), the association of metals with galaxies (17–19), and coronal gas in the Milky Way halo (20). The high sensitivity of COS enables a QSO absorption-line survey of halos around galaxies with a predetermined set of properties. We have selected 42 sample galaxies (tables S1 and S2) that span redshifts zgal = 0.10 to 0.36 and stellar masses [log(M*/M◉)] = 9.5 to 11.5, where M* is the galaxy stellar mass and M◉ is the mass of the
3
galaxy evolution require efficient outflows to explain observed galaxy masses and chemical abundances and to account for metals observed in the more diffuse IGM (8, 9). The CGM may also reflect the theoretically predicted transition from filamentary streams of cold gas that feed low-mass galaxies to hot, quasi-static envelopes that surround high-mass galaxies (10, 11). Both outflow and accretion through the CGM may be intimately connected to the observed dichotomy between blue, star-forming, disk-dominated galaxies and red, passively evolving, elliptical galaxies with little or no star formation (12). However, the low density of the CGM makes it extremely difficult to probe directly; thus, models of its structure and influences are typically constrained indirectly by its effects on the visible portions of galaxies, not usually by observations of the gas itself. We have undertaken a large program with the new Cosmic Origins Spectrograph (COS) aboard
Observed Wavelength (Å) Fig. 1. An illustration of our sampling technique and data. (A) An SDSS composite image of the field around the QSO J1016+4706 with two targeted galaxies, labeled G1 and G2, which are both in the star-forming subsample. (B) The complete COS count-rate spectrum (counts s−1) versus observed wavelength. This QSO lies at redshift zQSO = 0.822 and has an observed far-ultraviolet www.sciencemag.org
SCIENCE
magnitude of 18.1. (C and D) The redshifted O VI 1032, 1038 Å doublet for galaxies G1 (C) and G2 (D). (E and F) The full sample showing the locations of all sightlines in position angle and impact parameter R with respect to the targeted galaxies, for the star-forming (E) and passively evolving (F) subsamples. The circles mark R = 50, 100, and 150 kpc. VOL 334
18 NOVEMBER 2011
949
REPORTS B
log (Mhalo /M )
Impact parameter [kpc] Fig. 2. O VI association with galaxies. (A) O VI column density, NOVI, versus R for the star-forming (blue) and passive (red) subsamples. Solid and open symbols mark O VI detections and 3s upper limits, respectively. The detections in the star-forming galaxies maintain log NOVI ≈ 14.5 to R ≈ 150 kpc, the outer limit of our survey. (B) Component centroid velocities with
A
B
log (sSFR [yr -1])
-9
log (NOVI [cm-2])
Fig. 3. O VI correlation with galaxy properties. (A) O VI column density versus sSFR (≡ M*/SFR). Star-forming galaxies are divided from passively evolving galaxies by sSFR ≈ 10−11 year−1; our detection limit is sSFR ≈ 5 × 10−12 year−1. (B) The galaxy color-magnitude diagram (sSFR versus M*) for SDSS+GALEX galaxies from (23).
respect to galaxy systemic redshift for O VI detections, versus inferred darkmatter halo mass. The range bars mark the full range of O VI absorption for each system. The inset shows a histogram of the component velocities. The dashed lines mark the mass-dependent escape velocity at R = 50, 100, and 150 kpc from outside to inside.
-10
-11
-12
-13
Star-forming galaxies Passive galaxies 9
sSFR [yr -1]
galaxy systemic velocities indicates a close physical and/or gravitational association. CGM gas as traced by O VI reflects the underlying bimodality of the general galaxy population (12, 23). We found a correlation of NOVI with specific star formation rate sSFR (≡ SFR/M*) (Fig. 3). For the 30 galaxies with sSFR ≥ 10−11 year−1, there were 27 detections with a typical column density log NOVI = 14.5 (24) and a high covering fraction fhit ≈ 0.8 to 1 maintained all the way out to R = 150 kpc (Fig. 2). For the 12 galaxies in the passive subsample (sSFR ≤ 10−11 year−1), there were only four detections with lower typical NOVI than the star-forming subsample (25). Accounting for the upper limits in NOVI and sSFR, we can reject the null hypothesis that there is no correlation between NOVI and sSFR at >99.9% confidence for the whole sample and >98% for each of the 50-kpc annuli shown in Fig. 1 (21). This effect remained even when we controlled for stellar mass: A Kolmogorov-Smirnov test over log M* > 10.5, where the star-forming
950
and passive subsamples overlap, rejects at >99% confidence the null hypothesis that they draw from the same parent distribution of NOVI (fig. S2). We therefore conclude that the basic dichotomy between star-forming (“blue-cloud”) and passive (“red-sequence”) galaxies is strongly reflected in their gaseous halos, and that the CGM out to at least 150 kpc either directly influences or is directly affected by star formation. O VI is a fragile ionization state that never exceeds a fraction fOVI = 0.2 of the total oxygen for the physical conditions of halo gas and is frequently much less abundant (Fig. 4). Our observations imply a typical CGM oxygen mass MO, for star-forming galaxies, of 0:2 M O ¼ 5pR2 〈N OVI 〉mO fhit fOVI 7 0:2 M⊙ ð1Þ ¼ 1:2 10 fOVI where we have taken a typical mean column density 〈NOVI〉 = 1014.5 cm−2 and R = 150 kpc, and
18 NOVEMBER 2011
VOL 334
10
11
log (M /M )
SCIENCE
the hit rate correction fhit computed separately in three 50-kpc annuli (Figs. 1 and 2). This mass of oxygen is strictly a lower limit because we have scaled to the maximum fOVI = 0.2 (Fig. 4). The corresponding total mass of circumgalactic gas is Z⊙ MO Mgas ¼ 177 Z Z⊙ 0:2 M⊙ ð2Þ ¼ 2 109 fOVI Z where Z is the gas metallicity, and the solar oxygen abundance is nO/nH = 5 × 10−4 (26). Even for the most conservative ionization correction ( fOVI = 0.2), the OVI-traced CGM contains a mass of metals and gas that is substantial relative to other reservoirs of interstellar and circumgalactic gas. If our sample galaxies lie on the mean trend of gas fraction for low-z galaxies (27), they have interstellar medium (ISM) gas masses of MISM = 5 × 109 to 10 × 109 M◉ and 7 7 contain M O ISM = 2 × 10 to 10 × 10 M◉ of
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
A
REPORTS B
Fig. 4. CGM oxygen masses compared to galactic reservoirs. (A) The curves and the axis labels at right show the fraction of gas-phase oxygen in the O VI ionization state fOVI as a function of temperature, for three overdensities relative to the cosmic mean, r/r. All values of r/ r ≥ 1000 track the black curve on which collisional ionization dominates, whereas for lower values, photoionization by the extragalactic background can increase fOVI at low T. For gas that traces dark matter, r/ r = 1000 is typical at R ≈ 100 kpc; r/ r = 50 to 100 for the outskirts of the halo. The pale green band shows the expected oxygen mass of the galaxies’ ISM if they lie on the standard relation between MISM and M* and follow the mass-metallicity relation (MZR). The green dashed line shows the oxygen mass produced by 3 × 109 M◉ of star formation. The yellow band shows the expected oxygen mass for the extreme assumption that the typical host darkmatter halos (2 × 1011 to 1012 M◉) have the universal baryon fraction and solar metallicity. (B) The CGM oxygen masses compared with the interstellar oxygen mass as a function of M*. Points with range bars show the CGM oxygen mass MO implied by Eq. 1 for fOVI = 0.2, calculated separately for star-forming (blue) and passive (red) galaxies according to the hit rates in four bins of stellar mass. The purple curves O show the calculated MISM for typical star-forming galaxies in the SDSS, accounting for the mean MZR in the central curve and its uncertainties in the shaded region. The data points increase their mass in inverse proportion to fOVI. oxygen, taking into account the observed correlation between galaxy stellar mass and ISM metallicity (Fig. 4) (21). The minimum CGM oxygen mass is thus 10 to 70% of the ISM oxygen (Fig. 4 and fig. S4). The covering fractions and column densities we find for star-forming galaxies are insensitive to M*, whereas the ISM metal masses decline steeply with M* according to the mass-metallicity relation. Thus, the ratio of CGM metals to ISM metals appears to increase for lower-mass galaxies (assuming constant fOVI), perhaps indicating that metals more easily escape from their shallower gravitational potentials. The implied total mass of circumgalactic gas Mgas is more uncertain because it can strictly take on any metallicity; for a fiducial solar metallicity, Eq. 2 implies a total CGM mass comparable to MISM and several times the total mass inferred for Milky Way “high-velocity clouds” (28, 29) or for lowionization (Mg II) gas surrounding low-redshift galaxies to R = 100 kpc (30). For the densities typically expected at radii R ≈ 100 kpc, fOVI exceeds 0.1 only over a narrow temperature range 105.4−5.6 K, and it exceeds 0.02 only over 105.2−5.7 K (Fig. 4). Either a large fraction of CGM gas lies in this finely tuned temperature range—a condition that is difficult to maintain because gas cooling rates peak at T ≈ 105.5 K—or the CGM oxygen and gas masses are much larger than the minimum values we have quoted above. Lower-density photoionized gas can achieve high fOVI ≈ 0.1 over a wider temperature range, but at these low densities it is hard
to produce a 1014.5 cm−2 column density within the confines of a galactic halo, especially if the metallicity is low (fig. S5). Thus, fOVI = 0.02 and Z = 0.1Z◉ are plausible conditions for the O VI– traced gas, but it is unlikely that both conditions hold simultaneously. However, if either condition holds, the CGM detected here could represent an important contribution to the cosmic budgets of metals and baryons. In either case, Mgas is comparable to the total ~3 × 1010 M◉ inside R = 300 kpc inferred from H I measurements at low redshift (19) and to the ~4 × 1010 M◉ inferred for the CGM surrounding rapidly star-forming galaxies at z ≈ 2 to 3 (31). By generalizing our typical MO to all star-forming galaxies with M* > 109.5 M◉, we estimate that the halos of such galaxies contain 15% × (0.02/fOVI) of the oxygen in the universe and 2% × (0.02/fOVI) × (Z◉/Z) of the baryons in the universe. The metals detected out to R ≈ 150 kpc must have been produced in galaxies, after which they were likely transported into the CGM in some form of outflow. However, these outflows need not be active at the time of observation; indeed, the large masses imply long time scales. Because 1 M◉ of star formation eventually returns 0.014 M◉ of oxygen to the ISM (32), at least 8.6 × 108 M◉ of star formation is required to yield the detected oxygen mass. This is equivalent to ~3 × 108 years of star formation at the median SFR = 3 M◉ year−1 of our star-forming sample, in the unlikely event that all oxygen produced is expelled to the CGM, and longer in inverse proportion to the fraction of
www.sciencemag.org
SCIENCE
VOL 334
metals retained in the ISM. Thus, the detected oxygen could be the cumulative effect of steady enrichment over the preceding several billion years, the product of sporadic flows driven by rapid starbursts and an active nucleus (33), or the fossil remains of outflows from as early as z ≈ 1.5 to 3 (7, 31). Although the exact origin of the mass-metallicity relation of galaxies is not yet known, models that explain it in terms of preferential loss of metals imply that a substantial fraction of the metals produced by star formation must be ejected from the galaxy rather than retained in the ISM (28). The CGM detected here could be a major reservoir of this ejected material, with important consequences for models of galactic chemical evolution. The O VI we observe arises in bulk flows of gas over 100 to 400 km s−1, but the relative velocities are usually below halo escape speeds (Fig. 2), even when we take projection effects into account (fig. S1). Thus, much of the material driven into the halo by star formation could eventually be reacquired by the galaxy in “recycled winds,” which may be an important source of fuel for ongoing star formation (34). It is unlikely that the detected gas is predominantly fresh material accreting from the IGM because models of “cold mode” accretion predict very low metallicity and low covering fractions fhit ≈ 10 to 20% (35, 36), and “hot mode” accretion typically involves gas at temperatures T > 106 K with undetectably low fOVI. The passive galaxies in our sample once formed stars; thus, it follows that they would once have possessed halos of ionized, metalenriched gas visible in O VI. The relative paucity of O VI around these galaxies implies that this material was transformed by processes that plausibly accompany the quenching of star formation (37), such as tidal stripping in group environments, reaccretion onto the galaxy in ionized form, or heating or cooling to a temperature at which O VI is too rare to detect. Our findings present a quantitative challenge for theoretical models of galaxy growth and feedback, which must explain both the ubiquitous presence of massive, metal-enriched ionized halos around starforming galaxies and the fate of these metals after star formation ends.
Downloaded from www.sciencemag.org on November 17, 2011
A
References and Notes 1. In astronomical usage, metals are those elements heavier than hydrogen and helium; they are formed only by stellar nucleosynthesis. 2. S. Veilleux, G. Cecil, J. Bland-Hawthorn, Annu. Rev. Astron. Astrophys. 43, 769 (2005). 3. M. D. Lehnert, T. M. Heckman, Astrophys. J. 462, 651 (1996). 4. C. L. Martin, Astrophys. J. 621, 227 (2005). 5. D. S. Rupke, S. Veilleux, D. B. Sanders, Astrophys. J. Suppl. Ser. 160, 115 (2005). 6. A. E. Shapley, C. C. Steidel, M. Pettini, K. L. Adelberger, Astrophys. J. 588, 65 (2003). 7. B. J. Weiner et al., Astrophys. J. 692, 187 (2009). 8. V. Springel, L. Hernquist, Mon. Not. R. Astron. Soc. 339, 312 (2003). 9. B. D. Oppenheimer, R. Davé, Mon. Not. R. Astron. Soc. 373, 1265 (2006).
18 NOVEMBER 2011
951
REPORTS
25.
26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38.
(14, 15, 38) and is higher than the mean value (14.0) measured by the Far Ultraviolet Spectroscopic Explorer through the halo of the Milky Way (20), which would belong in our star-forming sample. O VI emission is seen in elliptical galaxies (39), but this gas is most likely associated with the ISM and not the CGM. M. Asplund, N. Grevesse, A. J. Sauval, P. Scott, Annu. Rev. Astron. Astrophys. 47, 481 (2009). M. S. Peeples, F. Shankar, Mon. Not. R. Astron. Soc. 417, 2962 (2011). M. E. Putman, Astrophys. J. 645, 1164 (2006). N. Lehner, J. C. Howk, Science 334, 955 (2011); 10.1126/science.1209069. H.-W. Chen et al., Astrophys. J. 714, 1521 (2010). C. C. Steidel et al., Astrophys. J. 717, 289 (2010). D. Thomas, L. Greggio, R. Bender, Mon. Not. R. Astron. Soc. 296, 119 (1998). T. M. Tripp et al., Science 334, 952 (2011). B. D. Oppenheimer et al., Mon. Not. R. Astron. Soc. 406, 2325 (2010). K. R. Stewart et al., Astrophys. J. 735, L1 (2011). M. Fumagalli et al., http://arxiv.org/abs/1103.2130 (2011). J. M. Gabor, R. Davé, K. Finlator, B. D. Oppenheimer, Mon. Not. R. Astron. Soc. 407, 749 (2010). T. M. Tripp et al., Astrophys. J. Suppl. Ser. 177, 39 (2008).
The Hidden Mass and Large Spatial Extent of a Post-Starburst Galaxy Outflow Todd M. Tripp,1* Joseph D. Meiring,1 J. Xavier Prochaska,2 Christopher N. A. Willmer,3 J. Christopher Howk,4 Jessica K. Werk,2 Edward B. Jenkins,5 David V. Bowen,5 Nicolas Lehner,4 Kenneth R. Sembach,6 Christopher Thom,6 Jason Tumlinson6 Outflowing winds of multiphase plasma have been proposed to regulate the buildup of galaxies, but key aspects of these outflows have not been probed with observations. By using ultraviolet absorption spectroscopy, we show that “warm-hot” plasma at 105.5 kelvin contains 10 to 150 times more mass than the cold gas in a post-starburst galaxy wind. This wind extends to distances > 68 kiloparsecs, and at least some portion of it will escape. Moreover, the kinematical correlation of the cold and warm-hot phases indicates that the warm-hot plasma is related to the interaction of the cold matter with a hotter (unseen) phase at >>106 kelvin. Such multiphase winds can remove substantial masses and alter the evolution of post-starburst galaxies. alaxies do not evolve in isolation. They interact with other galaxies and, more subtly, with the gas in their immediate environments. Mergers of comparable-mass, gas-rich galaxies trigger star-formation bursts by driving matter into galaxy centers, but theory predicts that such starbursts are short-lived: The central gas is rapidly driven away by escaping galactic winds powered by massive stars and supernova explosions or by a central supermassive black hole (1). Such feedback mechanisms could transform gas-rich spiral galaxies into post-starburst
G
1 Department of Astronomy, University of Massachusetts, Amherst, MA 01003, USA. 2University of California Observatories/ Lick Observatory, University of California, Santa Cruz, CA 95064, USA. 3Steward Observatory, University of Arizona, Tucson, AZ 85721, USA. 4Department of Physics, University of Notre Dame, Notre Dame, IN 46556, USA. 5Princeton University Observatory, Princeton, NJ 08544, USA. 6Space Telescope Science Institute, Baltimore, MD 21218, USA.
*To whom correspondence should be addressed. E-mail:
[email protected]
952
galaxies (2) and eventually into elliptical-type galaxies with little or no star formation (3). Mergers are not required to propel galaxy evolution, however. Even relatively secluded galaxies accrete matter from the intergalactic medium (IGM), form stars, and drive matter outflows into their halos or out of the galaxies entirely (4, 5). In either case, the competing processes of gas inflows and outflows are expected to regulate galaxy evolution. Outflows are evident in some nearby objects (6–9) and are ubiquitous in some types of galaxies (10–15); their speeds can exceed the escape velocity. Nevertheless, their broader impact on galaxy evolution is poorly understood. First, their full spatial extent is unknown. Previous studies (6, 9, 16–22) have revealed flows with spatial extents ranging from a few parsecs up to ~20 kiloparsecs (kpc). However, because of their low densities, outer regions of outflows may not have been detected with previously used techniques, and thus the flows could be much larger. Second,
18 NOVEMBER 2011
VOL 334
SCIENCE
39. J. N. Bregman, E. D. Miller, A. E. Athey, J. A. Irwin, Astrophys. J. 635, 1031 (2005). Acknowledgments: We thank the anonymous reviewers for constructive comments. This work is based on observations made for program GO11598 with the NASA/ESA Hubble Space Telescope, obtained at the Space Telescope Science Institute, operated by AURA under NASA contract NAS 5-26555, and at the W. M. Keck Observatory, operated as a scientific partnership of the California Institute of Technology, the University of California, and NASA. The Observatory was made possible by the generous financial support of the W. M. Keck Foundation. The Hubble data are available from the MAST archive at http://archive.stsci.edu. M.S.P. was supported by the Southern California Center for Galaxy Evolution, a multicampus research program funded by the UC Office of Research.
Supporting Online Material www.sciencemag.org/cgi/content/full/334/6058/948/DC1 SOM Text Figs. S1 to S5 Tables S1 and S2 References (40–62) 15 June 2011; accepted 27 September 2011 10.1126/science.1209840
the total column density and mass of the outflows are poorly constrained. Previous outflow observations were often limited to low-resolution spectra of only one or two ions (e.g., Na I or Mg II) or relied on composite spectra that cannot yield precise column densities. Without any constraints on hydrogen (the vast bulk of the mass) or other elements and ions, these studies were forced to make highly uncertain assumptions to correct for ionization, elemental abundances, and depletion of species by dust. Lastly, galactic winds contain multiple phases with a broad range of physical conditions (6), and wind gas in the key temperature range between 105 to 106 K (where radiative cooling is maximized) is too cool to be observed in x-rays; detection of this so-called “warm-hot” phase requires observations in the ultraviolet (UV). To study the more extended gas around galaxies, including regions affected by outflows, we used the Cosmic Origins Spectrograph (COS) on the Hubble Space Telescope (HST) to obtain high-resolution spectra of the quasi-stellar object (QSO) PG1206+459 (at redshift zQSO = 1.1625). By exploiting absorption lines imprinted on the QSO spectrum by foreground gaseous material, we can detect the low-density outer gaseous envelopes of galaxies, regions inaccessible to other techniques. We focus on far-ultraviolet (FUV) absorption lines at rest wavelengths lrest < 912 Å. This FUV wavelength range is rich in diagnostic transitions (23), including the Ne VIII 770.409, 780.324 Å doublet, a robust probe of warm-hot gas, as well as banks of adjacent ionization stages. The sight line to PG1206+459 pierces an absorption system, at redshift zabs = 0.927, that provides insights about galactic outflows. This absorber has been studied before (24), but previous observations did not cover Ne VIII and could not provide accurate constraints on H I in the individual absorption components.
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
10. D. Keres, N. Katz, D. H. Weinberg, R. Davé, Mon. Not. R. Astron. Soc. 363, 2 (2005). 11. A. Dekel, Y. Birnboim, Mon. Not. R. Astron. Soc. 368, 2 (2006). 12. G. Kauffmann et al., Mon. Not. R. Astron. Soc. 341, 33 (2003). 13. T. M. Tripp, B. D. Savage, E. B. Jenkins, Astrophys. J. 534, L1 (2000). 14. C. W. Danforth, J. M. Shull, Astrophys. J. 679, 194 (2008). 15. C. Thom, H.-W. Chen, Astrophys. J. 683, 22 (2008). 16. J. N. Bregman, Annu. Rev. Astron. Astrophys. 45, 221 (2007). 17. J. T. Stocke et al., Astrophys. J. 641, 217 (2006). 18. H.-W. Chen, J. S. Mulchaey, Astrophys. J. 701, 1219 (2009). 19. J. X. Prochaska, B. Weiner, H.-W. Chen, J. S. Mulchaey, K. L. Cooksey, http://arxiv.org/abs/1103.1891 (2011). 20. K. R. Sembach et al., Astrophys. J. Suppl. Ser. 146, 165 (2003). 21. See supporting material on Science Online. 22. J. K. Werk et al., http://arxiv.org/abs/1108.3852 (2011). 23. D. Schiminovich et al., Astrophys. J. Suppl. Ser. 173, 315 (2007). 24. The typical log NOVI = 14.5 to 15.0 for star-forming galaxies resembles the high end of the column-density distribution seen in blind surveys of intergalactic clouds
REPORTS er Lyman series lines are not saturated), which enables accurate H I column density [N(H I)] measurement (Fig. 1). A wide variety of metals and
Fig. 1. (Top) Small portion of the Keck HIRES spectrum of PG1206+459 (24). Tick marks at top indicate components detected at various velocities in the Mg II 2803.53 Å transition. A velocity scale in the rest frame of the affiliated galaxy 177_9 is inset at bottom. Gray indicates a feature not due to Mg II 2803.53 Å. (Bottom) Small portion of the ultraviolet spectrum of PG1206+459 recorded with the COS on HST that shows H I Lyman series absorption lines (marked with ticks and labels) at the redshift of the Mg II complex in the top graph, including H I Lyz through Lys (highest lines are marked but not labeled).
Fig. 2. Continuum-normalized absorption profiles (black lines) of various species detected in the LL /Mg II absorber shown in Fig. 1, plotted in velocity with respect to the galaxy 177_9 redshift (i.e., v = 0 km s−1 at z = 0.927). Labels below each absorption profile indicate the species and rest wavelength. We fitted nine components to the COS and Space Telescope Imaging Spectrograph data (24). Component centroids are indicated by gray lines, and the Voigt-profile fits are overplotted with red lines (25). Yellow lines indicate contaminating features from other redshifts or transitions. The two graphs at lower left compare apparent column density profiles (39) of the N V and Ne VIII doublets. www.sciencemag.org
SCIENCE
VOL 334
H I lines were detected in at least nine components (25) spanning a large velocity range from −317 to +1131 km s−1 (Figs. 1 and 2). The Ne VIII doublet was unambiguously detected (Fig. 2) with a total N(Ne VIII) = 1014.9 cm−2 (25), which is ~10 times higher than any previous N(Ne VIII) measurements in intervening absorbers (26, 27). The component at +1131 km s−1 exceeds vescape of any individual galaxy, and the other components have very similar properties to the +1131 km s−1 component (25), suggesting a common origin. Whether the other components have v > vescape depends on the (unknown) potential well, but allowing for projection effects and noting that the gas is already far from the affiliated galaxy (see below), several of the other components could also be escaping. Combined with detection of Ne VIII, the detections of banks of adjacent ions (N II, N III, N IV, N V; O III, O IV; S III, S IV, S V) place tight constraints on physical conditions of the gas. Notably, the velocity centroids and profile shapes of lower and higher ionization stages are quite similar (Fig. 3). This strong Ne VIII/LL absorber is affiliated with a galaxy near the QSO sight line (24, 25). This galaxy, which we refer to as 177_9, is the type of galaxy expected to drive a galactic superwind (Fig. 4). Like post-starburst (11) and ultraluminous infrared galaxies (28), galaxy 177_9 is very luminous and blue (29); based on the characteristic magnitude (M*) of the z ~ 1 luminosity function from the Deep Evolutionary Exploratory Probe 2 (DEEP2) (30), the galaxy luminosity L = 1.8 L*. The Multiple Mirror Telescope (MMT) spectrum in Fig. 4 is also similar to those of the post-starburst galaxies in (11), with higher Balmer series absorption lines, [O II] emission and [Ne V] emission indicative of an active galactic nucleus (AGN) (25). Most importantly, the galaxy has a large impact parameter from the QSO sight line, r = 68 kpc (31), which implies that the gaseous envelope of 177_9 has a large spatial extent. The component-to-component similarity of the absorption lines (Fig. 3) suggests a related origin. To further investigate the nature of this absorber, we used photoionization models (32) to derive ionization corrections and elemental abundances (25). These models indicate that the individual components have high abundances ranging from ~0.5 to 3 times those in the Sun (table S2). Such high abundances (or metallicities) favor an origin in outflowing ejecta enriched by nucleosynthesis products from stars; at the large impact parameter of 177_9, corotating outer-disk or halo gas or tidal debris from a low-mass satellite galaxy would be expected to have much lower metallicity. Tidal debris from a massive galaxy could have high metallicity, but we are currently aware of only one luminous galaxy near the sight line at the absorber redshift (33); another luminous galaxy interacting with 177_9 is not evident. The absorber could also be intragroup gas, but somehow it must have been metal-enriched, so some type of galactic outflow is implicated in any case.
18 NOVEMBER 2011
Downloaded from www.sciencemag.org on November 17, 2011
This absorber is illustrated in Figs. 1 to 3, including the COS data (25). The absorber is a “partial” Lyman limit (LL) system (i.e., the high-
953
Fig. 3. Comparison of apparent column density profiles (39) of the LL absorber affiliated with galaxy 177_9. In each graph, the C II 687.05 Å profile (black histogram) is compared to another species (colored circles) as labeled at upper left; the comparison species profile is also scaled by the factor in parentheses after the species label. Gray lines indicate regions contaminated by unrelated absorption. As in Fig. 2, v = 0 km s−1 at z = 0.927.
The photoionization models also constrain the total hydrogen column (i.e., H I and H II), and, combined with spatial extent ≥ 68 kpc, this allows mass estimates. By using fiducial thinshell models (25), we find that the mass of cool, photoionized gas in individual components ranges from 0.6 × 108 to 14 × 108 solar masses (M◉). However, photoionization fails (sometimes by orders of magnitude) to produce enough S V, N V, and Ne VIII; these species must arise in hot gas at temperature T > 105 K. By using equilibrium and nonequilibrium collision ionization models (25), we find that the warm-hot gas contains much more mass than the cold gas, with individual components harboring 10 × 108 to 400 × 108 M◉ in hot material. These are rough estimates with many uncertainties. For example, if the absorption arises in thin filaments analogous to those seen in starburst galaxies (6) or AGN bubbles (34), the cold-gas mass could reduce to ~106 M◉ per component. However, as in the thin-shell models, the warm-hot gas could harbor 10 to 150 times more mass in such filaments (25). In either case (shells or filaments), given the similarity of the cold and warm-hot absorption lines (Fig. 3), the Ne VIII–bearing plasma must be a transitional phase that links the colder and hotter material and thus provides insights on the outflow physics. The Ne VIII/N V phase is not photoionized, so it must be generated through interaction of the cold gas with a hotter ambient medium analogous to x-ray–emitting regions seen in nearby galaxies. How this occurs is an open question; the absorbers could be material cooling from the hot phase down to the cool gas, or the cool clouds could have a hotter and more-ionized surface that is evaporating. Low-density plasma in the T = 105 to 106 K range has been effectively hidden from most
Fig. 4. Montage of observations of the galaxy at zgal = 0.927 that drives a large-scale outflow of metal-enriched plasma. (Top left) The galaxy, and the background QSO that reveals the outflow via absorption spectroscopy, is shown in a multicolor image obtained with the Large Binocular Telescope. This galaxy, which we refer to as 177_9, is the red object 8.63 arc sec south of the bright QSO PG1206 +459 (zQSO = 1.1625) at a position angle of 177° (N through E) from the QSO. At the galaxy redshift, the angular separation from the QSO sight line corresponds to an impact parameter of 68 kpc. (Top right) The large red circle indicates the rest-frame U-B color and absolute B magnitude of 177_9 compared to all galaxies from the DEEP2 survey (gray scale) (30) and DEEP2 galaxies within T0.05 of z(177_9) (cyan points). The small purple circles show poststarburst galaxies from (11). (Bottom) An MMT optical spectrum of 177_9 (upper trace) with its 1s uncertainty (lower trace). The strong feature at ≈ 7600 Å is partially due to telluric absorption.
954
18 NOVEMBER 2011
VOL 334
SCIENCE
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
REPORTS
REPORTS
References and Notes 1. P. F. Hopkins et al., Astrophys. J. Suppl. Ser. 163, 1 (2006). 2. A. I. Zabludoff et al., Astrophys. J. 466, 104 (1996). 3. G. F. Snyder, T. J. Cox, C. C. Hayward, L. Hernquist, P. Jonsson, Astrophys. J. 741, 77 (2011). 4. D. Kereš, N. Katz, R. Davé, M. Fardal, D. H. Weinberg, Mon. Not. R. Astron. Soc. 396, 2332 (2009). 5. B. D. Oppenheimer et al., Mon. Not. R. Astron. Soc. 406, 2325 (2010). 6. S. Veilleux, G. Cecil, J. Bland-Hawthorn, Annu. Rev. Astron. Astrophys. 43, 769 (2005). 7. T. M. Heckman, M. D. Lehnert, D. K. Strickland, L. Armus, Astrophys. J. Suppl. Ser. 129, 493 (2000).
8. D. S. Rupke, S. Veilleux, D. B. Sanders, Astrophys. J. Suppl. Ser. 160, 87 (2005). 9. C. L. Martin, Astrophys. J. 647, 222 (2006). 10. M. Pettini et al., Astrophys. J. 554, 981 (2001). 11. C. A. Tremonti, J. Moustakas, A. M. Diamond-Stanic, Astrophys. J. 663, L77 (2007). 12. C. C. Steidel et al., Astrophys. J. 717, 289 (2010). 13. K. H. R. Rubin et al., Astrophys. J. 719, 1503 (2010). 14. F. Hamann, G. Ferland, Annu. Rev. Astron. Astrophys. 37, 487 (1999). 15. J. P. Grimes et al., Astrophys. J. Suppl. Ser. 181, 272 (2009). 16. K. H. R. Rubin, J. X. Prochaska, D. C. Koo, A. C. Phillips, B. J. Weiner, Astrophys. J. 712, 574 (2010). 17. M. Moe, N. Arav, M. A. Bautista, K. T. Korista, Astrophys. J. 706, 525 (2009). 18. J. P. Dunn et al., Astrophys. J. 709, 611 (2010). 19. D. Edmonds et al., Astrophys. J. 739, 7 (2011). 20. F. Hamann et al., Mon. Not. R. Astron. Soc. 410, 1957 (2011). 21. G. A. Kriss et al., Astron. Astrophys. 534, 41 (2011). 22. D. M. Capellupo, F. Hamann, J. C. Shields, P. Rodríguez Hidalgo, T. Barlow, Mon. Not. R. Astron. Soc. 413, 908 (2011). 23. D. A. Verner, D. Tytler, P. D. Barthel, Astrophys. J. 430, 186 (1994). 24. J. Ding, J. C. Charlton, C. W. Churchill, C. Palma, Astrophys. J. 590, 746 (2003). 25. See further information in supporting material on Science Online. 26. B. D. Savage, N. Lehner, B. P. Wakker, K. R. Sembach, T. M. Tripp, Astrophys. J. 626, 776 (2005). 27. A. Narayanan et al., Astrophys. J. 730, 15 (2011). 28. Y. Chen, J. D. Lowenthal, M. S. Yun, Astrophys. J. 712, 1385 (2010). 29. The galaxy appears to be red in Fig. 4 because of its redshift; in the rest frame of the galaxy, it has a very ultraviolet-blue (U-B) color. 30. C. N. A. Willmer et al., Astrophys. J. 647, 853 (2006). 31. For distance calculations, we assume a cold dark matter cosmology with Hubble constant H0 = 70 km s−1 Mpc−1 and dimensionless density parameters Ωm = 0.30, ΩL = 0.70. 32. G. J. Ferland et al., Publ. Astron. Soc. Pac. 110, 761 (1998). 33. As discussed in the supporting online material, the yellow galaxy northwest of the QSO (Fig. 4) does not have a spectroscopic redshift but is likely to have z << 0.927. 34. A. C. Fabian et al., Nature 454, 968 (2008). 35. T. M. Tripp et al., Astrophys. J. Suppl. Ser. 177, 39 (2008).
A Reservoir of Ionized Gas in the Galactic Halo to Sustain Star Formation in the Milky Way Nicolas Lehner* and J. Christopher Howk Without a source of new gas, our Galaxy would exhaust its supply of gas through the formation of stars. Ionized gas clouds observed at high velocity may be a reservoir of such gas, but their distances are key for placing them in the galactic halo and unraveling their role. We have used the Hubble Space Telescope to blindly search for ionized high-velocity clouds (iHVCs) in the foreground of galactic stars. We show that iHVCs with 90 ≤ |vLSR| ≲ 170 kilometers per second (where vLSR is the velocity in the local standard of rest frame) are within one galactic radius of the Sun and have enough mass to maintain star formation, whereas iHVCs with |vLSR| ≳ 170 kilometers per second are at larger distances. These may be the next wave of infalling material. he time scale for gas consumption via star formation in spiral galaxies is far shorter than a Hubble time (13.8 billion years), requiring an ongoing replenishment of the gas-
T
eous fuel in the disks of galaxies for continued star formation. Analytical models and hydrodynamical simulations have emphasized the importance of cold-stream accretion as a means for
www.sciencemag.org
SCIENCE
VOL 334
36. C. Tremonti, A. M. Diamond-Stanic, J. Moustakas, in Galaxy Evolution: Emerging Insights and Future Challenges, S. Jogee, I. Marinova, L. Hao, G. Blanc, Eds, (Astronomical Society of the Pacific Conference Series, San Francisco, 2009), vol. 419, pp. 369–376. 37. A. L. Coil et al., http://arXiv.org/abs/1104.0681 (2011). 38. J. Tumlinson et al., Science 334, 948 (2011). 39. B. D. Savage, K. R. Sembach, Astrophys. J. 379, 245 (1991). Acknowledgments: This study has its basis in observations made with the NASA/European Space Agency Hubble Space Telescope (HST); the MMT, a joint facility operated by the Smithsonian Astrophysical Observatory and the University of Arizona; and the Large Binocular Telescope, an international collaboration among institutions in the United States, Italy, and Germany. Support for HST program number 11741 was provided by NASA through a grant from the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Incorporated, under NASA contract NAS5-26555. Additional support was provided by NASA grant NNX08AJ44G. The DEEP2 survey was supported by NSF grants AST 95-29098, 00-711098, 05-07483, 08-08133, 00-71048, 05-07428, and 08-07630. Funding for the Sloan Digital Sky Survey has been provided by the Alfred P. Sloan Foundation, the Participating Institutions, NASA, NSF, the U.S. Department of Energy Office of Science, the Japanese Monbukagakusho, and the Max Planck Society. We thank C. Churchill for providing the archival Keck data and the referees for review comments that significantly improved this paper. We are also grateful to the Hawaiian people for graciously allowing us to conduct observations from Mauna Kea, a revered place in the culture of Hawaii. The HST data in this paper are available from the Multimission Archive at the Space Telescope Science Institute (MAST) at http://archive.stsci.edu.
Downloaded from www.sciencemag.org on November 17, 2011
outflow studies. In principle, the O VI 1032,1038 Å doublet can reveal such gas, but it is unclear whether the O VI arises in photoionized 104 K gas, hotter material at ~105.5 K, or both (35). The Ne VIII doublet avoids this ambiguity, and we have found that this warm-hot matter is a substantial component in the mass inventory of a galactic wind. Moreover, this wind has a large spatial extent, and the mass carried away by the outflow will affect the evolution of the galaxy. Whereas earlier studies of poststarburst outflows focused on Mg II and could not precisely constrain the metallicity, hydrogen column, and mass, these studies do indicate that post-starburst outflows are common: 22/35 of the post-starbursts in (36) showed outflowing Mg II absorption with maximum (radial) velocities of 500 to 2400 km s−1, similar to the absorption near 177_9 (Fig. 1), and 77 and 100% of the post-starburst and AGN galaxies, respectively, in (37) drive outflows but with lower maximum velocities, which may be due to selection of wind-driving galaxies in a later evolutionary stage. With existing COS data, the effects of large-scale outflows on galaxy evolution can be studied with the techniques presented here but with larger samples (38), with which it will be possible to statistically track how outflows affect galaxies.
Supporting Online Material www.sciencemag.org/cgi/content/full/334/6058/952/DC1 Materials and Methods SOM Text Figs. S1 to S5 Tables S1 and S2 References (40–54) 15 June 2011; accepted 26 October 2011 10.1126/science.1209850
metal-poor gas (metallicity that is less than 10% of that of the Sun, or Z ≲ 0.1 Z◉) to flow onto galaxies along dense intergalactic filaments (1). However, galaxies may also exchange mass with the local intergalactic medium (IGM) through outflows driven by galactic “feedback,” galactic winds powered by massive stars and their death and from massive black holes. Some of this material may return to the central galaxy as recycled infalling matter—the galactic fountain mechanism (2, 3). The circumgalactic medium about a galaxy is thus a complicated blend of outflowing metal-rich and infalling metal-poor gas. The relative importance of these processes is poorly constrained observationally. Here, we demonstrate that ionized gas in the local galactic halo provides a major supply of matter for fueling ongoing star formation. Department of Physics, University of Notre Dame, 225 Nieuwland Science Hall, Notre Dame, IN 46556, USA. *To whom correspondence should be addressed. E-mail:
[email protected]
18 NOVEMBER 2011
955
REPORTS
956
tocentric distance of <17.7 kpc and subsolar metallicity; it probably traces gas that is being accreted by the Milky Way, possibly associated with the well-known large H I complex C (18, 19). These studies show that some iHVCs probe gas flows in and out of the galactic disk, with some originating from the Milky Way and others having an extragalactic origin. Here, we generalize the result to understand these iHVCs in the context of the Milky Way evolution with a survey of the gas in the foreground of 28 distant galactic halo stars with known distances. These stars were observed with the Cosmic Origins Spectrograph (COS) and SpaceTelescope Imaging Spectrograph (STIS) on board the Hubble Space Telescope (HST), and a large majority (23 of 28) of these were obtained through our HST Cycle 17 program 11592 [supporting online material (SOM) text]. The main criterion for assembling this stellar sample is that these ultraviolet (UV) bright stars are at height |z| ≳ 3 kpc from the galactic plane. This minimum height was adopted in view of absence or scarcity of iHVCs at smaller z (20) and other works on their predominantly neutral counterparts (5). We systematically searched for high-velocity interstellar metal-line absorption in the COS and STIS UV spectra of these stars (SOM text). We noted HVC detections only if high-velocity absorption was seen in multiple ions or transitions (at a minimum two) (Fig. 1). We measured the apparent optical depth-weighted mean velocity 〈v〉 = ∫v ta(v)dv /∫ta(v)dv (table S1). Most of the HVCs seen in absorption against the stars do not have H I 21-cm emission at a level of ≳1018.5 cm−2 according to the Leiden/Argentine/Bonn (LAB) survey (21). The column densities of Si II and O I imply large ionization fractions in several HVCs of our sample (Fig. 1 and SOM text) (22, 17), demonstrating that N(H II) >> N(H I). The iHVC detection rate in our stellar sample is 50% (14 of 28). Although a sightline may have more than one high-velocity absorption component, only one HVC for a given sightline is counted for estimating the covering factor. Defining a sample with a more uniform sensitivity Wl ≥ 15 mÅ near Si IIl1526 and with no stellar contamination (that is, with no observed stellar photospheric absorption at |vLSR| ≥ 90 km; see flag Q = 1 in table S1 and SOM text for more details) gives essentially the same detection rate with 47% (9 of 19). The average distance of the stars in the latter sample in which the iHVCs are detected is 11.5 T 4.1 kpc; the average absolute z-height is 7.3 T 3.0 kpc; for the whole sample, these are 〈d〉 = 11.6 T 6.9 kpc and 〈|z|〉 = 6.5 T 3.2 kpc. Our sky coverage is larger at b > 0° than at b < 0° (Fig. 2); the difference of detection rates between the northern (60%, 12 of 20) and southern (29%, 2 of 7) sky may be due to statistical fluctuations in the smaller southern sample. With the better sampled northern galactic sky, there is some evidence for an increase in the detection rate to 73% (11 of 15) for sightlines with z ≳ 4 kpc, implying that most of the iHVCs
18 NOVEMBER 2011
VOL 334
SCIENCE
could be situated at 4 ≲ z ≲ 9 kpc. There is no iHVC absorption at |vLSR| ≳ 170 km s−1 (VHVCs) toward the stars (Fig. 2 and SOM text), even though these are observed along the path to AGNs (Fig. 2 and SOM text). Thus, there is no evidence for VHVCs in the galactic halo at |z| ≲ 10 kpc. In order to understand the implications of the iHVC detection rate toward stars, we need to compare it with a distance-independent measure of the iHVC covering fraction. Earlier studies
Downloaded from www.sciencemag.org on November 17, 2011
Inflow and outflow in the Milky Way halo can be studied via the so-called high-velocity clouds (HVCs), clouds moving in the local standard of rest (LSR) frame at |vLSR| ≥ 90 km s−1 (4). Determining the distance (d ) of these HVCs is critical for associating HVCs with flows occurring near the Milky Way rather than the IGM of the Local Group and for quantifying their basic physical properties because several of these directly scale with the distance (for example, the mass M º d2). Major progress has been made in the past decade for some of the large, predominantly neutral HVC complexes, placing them 4 to 13 kpc from the sun (excluding here the Magellanic Stream) (5–8). Approximately 37% of the galactic sky is covered by neutral atomic hydrogen (H I) HVCs with column densities N(H I) ≥ 1017.9 cm−2 (9), but the infall rate of the largest H I HVC complex within 10 kpc (complex C), ∼0.14 M◉ year−1 (where M◉ is the mass of the Sun) (8), is far too modest for replenishing the 0.6 to 1.45 M◉ year−1 consumed by Milky Way star formation (10) (we use throughout the following spectroscopic notation: if X is a given atom, then X I refers to the neutral atom X, X II refers to the singly ionized atom X, X IV refers to the triply ionized atom X...). This is not entirely surprising, given that recent models show that inflowing gas should be predominantly ionized in view of the small amounts of neutral gas available for inflow at any epoch (11). In ionized gas, N(H I) becomes small relative to N(H II), and the H I emission becomes extremely difficult to impossible to detect. Low H I content HVCs are, however, routinely found in absorption in the spectra of cosmologically distant objects, such as active galactic nuclei (AGNs), with a detection rate of ∼60 to 80% (12–15). Their total (neutral and ionized) hydrogen column density is shown to be quite large [as large as N(H I) in predominantly neutral HVCs]. These are therefore ionized HVCs (iHVCs)— that is, N(H II) >> N(H I) (12, 16). Given their large covering factor, the iHVCs may represent the long-sought supply of gas needed for continued Milky Way star formation. However, the iHVCs have been mostly detected against AGNs: They may reside within the Galaxy, the Local Group, or the IGM. Thus, as for their larger H I column density counterparts, direct distance constraints are required for determining their masses and for characterizing their role in the evolution of the Milky Way. To determine the distances of iHVCs can be directly undertaken by observing the gas in the foreground of stars at known distances from the sun. Recently, on the basis of observations of high-velocity interstellar absorption in the ultraviolet spectra of two galactic stars, two of these iHVCs were found within 8 to 15 kpc from the sun (17, 18). One of them was found toward the inner Galaxy and has supersolar metallicity, probing gas that has been ejected from and is raining back onto the Milky Way disk (17). Another was observed at l ∼ 103° with a galac-
Fig. 1. Example of COS continuum normalized absorption profiles of various metal-lines and (Top) LAB H I emission line profile toward PG0914+001, a star at d = 16 kpc and z = +8.4 kpc. There are at least two iHVCs seen in absorption at +100 and +170 km s−1, as indicated by the dotted lines and shaded regions. The ∼0 km s−1 absorption and H I emission are from the Milky Way disk. We derive for the iHVC component [O I/S II] ≡ log[N(O I)/N(S II)] − log(AO/AS)◉ ≅ −2.34, where (AO/AS)◉ is the relative solar abundance of oxygen to sulfur, implying H II >> H I. If the iHVC has a solar abundance, N(H II) > 1019.6 cm−2 (based on S II), and N(H I) ≅ 1017.30 cm−2 (based on O I). These column densities would increase if the iHVC metallicity is subsolar. We find very subsolar [Fe II/S II] < −1.5 and [Si II/S II] ≅ −1.4 ratios, indicating strong ionization and dust depletion effects; the latter could suggest a galactic origin. Examples of negative-velocity iHVCs toward stars can be found in (17, 18).
www.sciencemag.org
REPORTS slow through their interaction with the galactic gas as they approach the plane, as predicted by some models of HVCs (23–26). In the latter scenario, some of the VHVCs could be at large distances from the Milky Way or other galaxies (such as Andromeda) and be a component of the multiphase local intergalactic medium, which may permeate the Local Group of galaxies (27, 28). A majority of the so-called O VI HVCs have accompanying C III and H I HVC absorption at similar high velocities, demonstrating their multiphase nature (13). The velocity sky-distribution of the O VI, Si III, and H I HVCs are also alike (12–14), and the sky-distribution of the iHVCs seen toward the AGNs and stars is moreover remarkably similar considering our better sampled north galactic sky (Fig. 2). These and the fact that the H I HVCs and iHVCs are now known to be at similar distances strongly suggest that they are all related, probing separate phases in which the H I HVCs may be the densest regions of the more diffuse ionized complexes. Having demonstrated that the iHVCs are in the local galactic halo, we can reliably estimate their mass and assess their importance for future star formation in the Milky Way. The mass of these iHVCs can be estimated MiHVC ≅ 1.3mH (4pd2) fc NH II ≈ 9.6 × 10−12 (d/12 kpc)2 ( fc/0.5)NHII M◉ (where mH is the mass of hydrogen and the factor 1.3 accounts for the additional mass of helium). The ionized gas probed by Si II, Si III, and Si IV has 〈NH II〉 ≈ 6 × 1018 (Z/0.2Z◉)−1 cm−2 (16), whereas the O VI phase has 〈NH II 〉 ≈ 4.5 × 1018 (Z/0.2Z◉)−1 cm−2 (12, 13, 16), where the metallicity of 0.2Z◉ is representative of complex C and other large H I complexes (5, 19). Because O VI and Si III are not expected to exist in the same gas phase, the total H II column density is a sum of the phases traced by these ions, NH II ≈ 1.1 × 1019 (Z/0.2Z◉)−1
cm−2. These assumptions imply a total mass MiHVC ≈ 1.1 × 108 (d/12 kpc)2 ( fc/0.5)(Z/0.2Z◉)−1 M◉. We estimate infall times of 80 to 130 million years, assuming infall velocity of about 90 to 150 km s−1, which implies a mass infall rate for the ionized gas of ∼0.8 to 1.4 M◉ year−1 (d = 12 kpc, fc = 0.5, and Z = 0.2Z◉). Although this value may be reduced somewhat because there must be a mixture of outflows and inflows, large H I complexes have subsolar abundances (5), suggesting a substantial fraction of the circumgalactic neutral and ionized gas is infalling. Some of the outflowing gas must also be recycled via a galactic fountain because the iHVCs do not reach the Milky Way escape velocity. We did not bracket the distances of the iHVCs, but their distances are unlikely to be much smaller than 3 kpc because otherwise they would already have been detected serendipitously in the UV spectra of more nearby stars (20). If all of the iHVCs were at d = 5 kpc, then the infall rate would still be significant with ∼0.4 to 0.6 M◉ year−1. Although the infall rate depends on several parameters that are not well constrained, our distance estimates allow us to unambiguously place the iHVCs at 90 ≤ |vLSR| ≤ 170 km s−1 in the Milky Way’s halo. Assuming an average metallicity of 0.2Z◉, the estimated infall rate is a factor of 6 to 10 larger than that of the large H I complex C (6, 8). (Even with an unrealistically solar metallicity value for all the iHVCs, their infall rate would still be important.) The presentday total star formation rate of the Milky Way is 0.6 to 1.45 M◉ year−1 (10), and chemical evolution models require a present-day infall rate of only 0.45 M◉ year−1 (29)—a value that could be somewhat reduced according to stellar mass loss models (30). The iHVCs are thus sufficient to sustain star formation in the Milky Way.
Downloaded from www.sciencemag.org on November 17, 2011
determined the covering factors of these iHVCs toward AGNs (12–14, 16), but they concentrated on a single species (O VI or Si III), which differs from our search method. We combined observations from (14) and (15) to assemble an AGN sample using the same search criteria adopted for our stellar survey (using the same metal ions) (SOM text). Our AGN sample is summarized in table S2, and its sky distribution is shown on Fig. 2. It has a similar sensitivity and size as the stellar sample. The detection rate is 67% (16 of 24) for the iHVCs and 42% (11 of 26) for the VHVCs (excluding the four Magellanic Stream sightlines). For the HVC sample, six sightlines with H I LAB emission at 90 ≤ |vLSR| ≤ 170 km s−1 were excluded because of strong selection effects in the AGN sample: Many of these AGNs were indeed initially targeted to study known H I HVCs [including these sightlines yields a covering factor ( fc) = 73%]. After removing those, the AGN sample may still be biased and overestimate the true covering factors because a given AGN could have been specifically targeted for studying an HVC or could have been favored over other AGNs because of previously known HVCs near the line of sight. The covering factors of the iHVCs are therefore fc = 50% (14 of 28) and ≤67% (16 of 24) for the stellar and AGN samples, respectively. Hence, a majority (if not all) of the iHVCs seen toward AGNs are within 〈d 〉 ≅ 12 T 4 kpc and 〈|z|〉 ≅ 6 T 3 kpc, implying that the iHVCs at 90 ≲ |vLSR| ≤ 170 km s−1 mostly trace flows of ionized gas in the Milky Way halo. On the other hand, VHVCs must then lie beyond d ≳ 10 to 20 kpc (|z| ≳ 6 to 10 kpc) because they are not detected toward any star. The VHVCs could be associated with the outer reaches of the Milky Way or gas in the Local Group. In the former scenario, this would imply that iHVCs Fig. 2. Aitoff projection galactic (longitude, latitude) map of the survey directions for the stellar sample (circles) and extragalactic sample (square). A solid circle/square indicates an HVC in the foreground of the star/AGN, whereas an open circle/ square implies no HVC along the stellar/AGN sightline. The velocity value is color-coded following the horizontal color bar. The positive and negative numbers indicate the z-height (in kpc) of the stars. The squares with an “X” are HVCs in which H I 21-cm LAB emission is present at similar velocities seen in absorption. An “S” near a square indicates a sightline that passes through the Magellanic Stream.
www.sciencemag.org
SCIENCE
VOL 334
18 NOVEMBER 2011
957
We have implicitly assumed the HVCs survive their fall onto the galactic disk. Hydrodynamical simulations of HVCs support this idea (25, 26). In these models, the HVCs lose most of their H I content within 10 kpc of the disk and continue to fall toward the disk as warm ionized matter. This is consistent with the iHVCs covering more galactic sky and tracing a larger mass reservoir than do the predominantly neutral HVCs. Because the iHVCs are still overdense relative to the halo medium, they can continue to sink toward the galactic plane, where they decelerate and feed the warm ionized medium (Reynolds layer) 1 to 2 kpc from the galactic disk. In the Reynolds layer, the clouds have low velocities and cannot be identified as HVCs anymore, but as low- or intermediate-velocity clouds, which is consistent with the observed absence of iHVCs at low z-height. In this scenario, VHVCs are also not expected to be seen near the galactic disk because the HVC velocities decrease with decreasing z-height due to drag. This is consistent with the lack of VHVCs at |z| ≲ 10 to 20 kpc and the observed velocity sky distribution (23). References and Notes 1. D. Keres, L. Hernquist, Astrophys. J. 700, L1 (2009). 2. F. Fraternali, J. J. Binney, Mon. Not. R. Astron. Soc. 386, 935 (2008).
3. B. Oppenheimer et al., Mon. Not. R. Astron. Soc. 406, 2325 (2010). 4. B. P. Wakker, H. van Woerden, Annu. Rev. Astron. Astrophys. 35, 217 (1997). 5. B. P. Wakker, Astrophys. J. Suppl. Ser. 136, 463 (2001). 6. B. P. Wakker et al., Astrophys. J. 670, L113 (2007). 7. B. P. Wakker et al., Astrophys. J. 672, 298 (2008). 8. C. Thom et al., Astrophys. J. 684, 364 (2008). 9. E. M. Murphy, F. J. Lockman, B. D. Savage, Astrophys. J. 447, 642 (1995). 10. T. P. Robitaille, B. A. Whitney, Astrophys. J. 710, L11 (2010). 11. A. Bauermeister, L. Blitz, C.-P. Ma, Astrophys. J. 717, 323 (2010). 12. K. R. Sembach et al., Astrophys. J. Suppl. Ser. 146, 165 (2003). 13. A. J. Fox, B. D. Savage, B. P. Wakker, Astrophys. J. Suppl. Ser. 165, 229 (2006). 14. J. A. Collins, M. Shull, M. L. Giroux, Astrophys. J. 705, 962 (2009). 15. P. Richter, J. C. Charlton, A. P. M. Fangano, N. B. Bekhti, J. R. Masiero, Astrophys. J. 695, 1631 (2009). 16. J. M. Shull, J. R. Jones, C. W. Danforth, J. A. Collins, Astrophys. J. 699, 754 (2009). 17. W. F. Zech, N. Lehner, J. C. Howk, W. V. D. Dixon, T. M. Brown, Astrophys. J. 679, 460 (2008). 18. N. Lehner, J. C. Howk, Astrophys. J. 709, L138 (2010). 19. T. M. Tripp, L. Song, Astrophys. J., http://arxiv.org/abs/ 1101.1107. 20. J. Zsargó, K. R. Sembach, J. C. Howk, B. D. Savage, Astrophys. J. 586, 1019 (2003). 21. P. M. W. Kalberla et al., Astron. Astrophys. 440, 775 (2005). 22. N. Lehner, F. P. Keenan, K. R. Sembach, Mon. Not. R. Astron. Soc. 323, 904 (2001).
Giant Piezoelectricity on Si for Hyperactive MEMS S. H. Baek,1 J. Park,2 D. M. Kim,1 V. A. Aksyuk,3 R. R. Das,1 S. D. Bu,1 D. A. Felker,4 J. Lettieri,5 V. Vaithyanathan,5 S. S. N. Bharadwaja,5 N. Bassiri-Gharb,5 Y. B. Chen,6 H. P. Sun,6 C. M. Folkman,1 H. W. Jang,1 D. J. Kreft,2 S. K. Streiffer,7 R. Ramesh,8 X. Q. Pan,6 S. Trolier-McKinstry,5 D. G. Schlom,5,9,10 M. S. Rzchowski,3 R. H. Blick,2 C. B. Eom1* Microelectromechanical systems (MEMS) incorporating active piezoelectric layers offer integrated actuation, sensing, and transduction. The broad implementation of such active MEMS has long been constrained by the inability to integrate materials with giant piezoelectric response, such as Pb(Mg1/3Nb2/3)O3-PbTiO3 (PMN-PT). We synthesized high-quality PMN-PT epitaxial thin films on vicinal (001) Si wafers with the use of an epitaxial (001) SrTiO3 template layer with superior piezoelectric coefficients (e31,f = –27 T 3 coulombs per square meter) and figures of merit for piezoelectric energy-harvesting systems. We have incorporated these heterostructures into microcantilevers that are actuated with extremely low drive voltage due to thin-film piezoelectric properties that rival bulk PMN-PT single crystals. These epitaxial heterostructures exhibit very large electromechanical coupling for ultrasound medical imaging, microfluidic control, mechanical sensing, and energy harvesting. ilicon is the gold standard for microelectronic devices as well as for micro-electromechanical systems (MEMS), which are electrically driven mechanical devices ranging in size from a micrometer to a few millimeters. However, the main drawback in the world of MEMS is that Si is a passive material; that is, it takes metallic electrodes to capacitively displace MEMS. On the other hand, active piezoelectric materials such as lead zirconate titanate (PZT) enable mechanical displacement (1–3). Piezoelectric materials
S
958
transform electrical energy into mechanical energy and vice versa through their linear electromechanical coupling effect. The highest-performing piezoelectric MEMS heterostructures, including energy-harvesting devices and actuator structures, have been fabricated with PZT piezoelectric layers. A. K. Sharma et al. demonstrated the integration of epitaxial PZT thin films on Si substrates with the use of a SrTiO3/MgO/TiN buffer layer (4). Subsequent work further enhanced PZT thin-film piezoelectric properties and demon-
18 NOVEMBER 2011
VOL 334
SCIENCE
23. R. A. Benjamin, L. Danly, Astrophys. J. 481, 764 (1997). 24. J. E. G. Peek, M. E. Putman, C. F. McKee, C. Heiles, S. Stanimirović, Astrophys. J. 656, 907 (2007). 25. F. Heitsch, M. E. Putman, Astrophys. J. 698, 1485 (2009). 26. F. Marinacci et al., Mon. Not. R. Astron. Soc. 415, 1534 (2011). 27. L. Blitz, D. N. Spergel, P. J. Teuben, D. Hartmann, W. B. Burton, Astrophys. J. 514, 818 (1999). 28. F. Nicastro et al., Nature 421, 719 (2003). 29. C. Chiappini, F. Matteucci, D. Romano, Astrophys. J. 554, 1044 (2001). 30. S. N. Leitner, A. V. Kravtsov, Astrophys. J. 734, 48 (2011). Acknowledgments: This work was based on observations made with the NASA/European Space Agency Hubble Space Telescope, obtained at the Space Telescope Science Institute (STScI), which is operated by the Association of Universities for Research in Astronomy, under NASA contract NAS5-26555. We greatly appreciate funding support from NASA grant HST-GO-11592.01-A from STScI. We are grateful to P. Chayer for computing several stellar spectra for us. The data reported in this paper are tabulated in the SOM and archived at the Multimission Archive at STScI (MAST, http://archive.stsci.edu/).
Supporting Online Material www.sciencemag.org/cgi/content/full/science.1209069/DC1 SOM Text Tables S1 to S3 References (31–41) 31 May 2011; accepted 17 August 2011 Published online 25 August 2011; 10.1126/science.1209069
strated high-functionality active piezoelectric devices incorporating PZT on Si (3–6). Relaxor ferroelectrics with engineered domain states have dramatically enhanced piezoresponse over PZT. Pb(Mg1/3Nb2/3)O3-PbTiO3 (PMN-PT), one of the lead-based relaxor ferroelectrics, exhibits strain levels and piezoelectric coefficients that can be 5 to 10 times those of bulk PZT ceramics and has a large electromechanical coupling coefficient of k33 ~ 0.9 (7, 8). Performance of piezoelectric MEMS could be enhanced dramatically by incorporating these materials. This requires the integration of PMNPT films with Si, while preserving their enhanced piezoelectric properties. The piezoelectric properties of ferroelectrics depend on both intrinsic (stoi1 Department of Materials Science and Engineering, University of Wisconsin, Madison, WI 53706, USA. 2Department of Electrical and Computer Engineering, University of Wisconsin, Madison, WI 53706, USA. 3Center for Nanoscale Science and Technology, National Institute of Standards and Technology, Gaithersburg, MD 20899, USA. 4Department of Physics, University of Wisconsin, Madison, WI 53706, USA. 5Department of Materials Science and Engineering, Penn State University, University Park, PA 16802, USA. 6Department of Materials Science and Engineering, University of Michigan, Ann Arbor, MI 48109, USA. 7Center for Nanoscale Materials, Argonne National Laboratory, Argonne, IL 60439, USA. 8Department of Materials Science and Engineering, University of California, Berkeley, CA 94720, USA. 9Department of Materials Science and Engineering, Cornell University, Ithaca, NY 14853–1501, USA. 10Kavli Institute at Cornell for Nanoscale Science, Ithaca, NY 14853, USA.
*To whom correspondence should be addressed. E-mail:
[email protected]
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
REPORTS
REPORTS
Intensity (cps)
001 60
30 40 2θ (degree)
002
440
Perovskite
Perovskite
20
C 104
002 SrRuO3
B
002
111
101
001
311 222
400 331
Pyrochlore
002 SrRuO3
A
20
D
104 103 102 101 100
202 Si
0
60
30 40 2θ (degree)
101 PMN -PT
103 102 101 100
90
180
270
φ (degree)
360
E
PMN-PT PMN-PT
SrRuO3 SrTiO3
SrRuO3
100 nm
Si
2 nm
Fig. 1. Phase-pure, (001)pc oriented epitaxial PMN-PT thin film on a SrTiO3-buffered Si substrate. X-ray diffraction pattern measured with a 2D area detector of PMN-PT heterostructure on (001) Si substrates with (A) zero miscut (T0.1°) and (B) 4° miscut along [110]. The pyrochlore phase was identified as Pb2Nb2O7. Pseudocubic notations are used for both PMN-PT and SrRuO3 peak indexing. (C) f scan of the 101pc PMN-PT and 202 Si diffraction peaks. cps, counts per second. (D) Bright-field cross-sectional TEM image near the interface between PMN-PT and SrRuO3. (Inset) SAED image of PMN-PT along the [100] zone axis. (E) High-resolution TEM image of PMN-PT and SrRuO3 interface.
0 -20 -40 -60
-600 -400 -200
0
200
400
600
0.3 2000 0.2
1000
0
0.1
-100
Electrical field (kV/cm)
0
100
Transverse piezoelectric coefficient -e31,f (C/m2)
20
Relative dielectric permitivity
40
C 3000
Loss tangent
Polarization (µC/cm2)
B
60
30 25 20
50 40
15
30
10
20
5
10
0
0
Energy Harvesting Figure of Merit e 231,f /ε oε r (GPa)
A
an epitaxial SrTiO3 buffer layer and (ii) a miscut Si substrate. The latter approach works well for growing phase-pure thin films of oxide materials with volatile constituents (10). We believe that the effect of high density of steps on the surface of miscut substrates is to maintain film stoichiometry by effectively incorporating volatile constituents (such as PbO) into the film, suppressing formation of pyrochlore, which is lead-deficient. The (001) epitaxial SrTiO3 layer on Si, deposited by reactive molecular-beam epitaxy, enables us to grow a single-crystal (001)pc PMN-PT layer on the epitaxial (001)pc SrRuO3 bottom electrode (11) using the control provided by heteroepitaxy. The conventional method has shown that the structural quality of the SrTiO3 was degraded as the orientation of the silicon wafers deviated from exact (001) due to the high density of steps and the increased reactivity of exposed Si step edges (12). These growth challenges were overcome by growing high-quality SrTiO3 on miscut Si. We used a four-circle x-ray diffractometer with both a twodimensional (2D) area detector and a four-bounce monochromator to study the phase purity, crystal structure, and epitaxial arrangements of 3.5-mmthick PMN-PT/SrRuO3/SrTiO3/Si. Figure 1, A and B show the q-2q scans of 3.5-mm-thick PMNPT films grown under the same conditions on exact (T0.1°) and 4° miscut Si substrates, respectively. The PMN-PT film on exact Si exhibits a large volume of pyrochlore phase as well as polycrystalline perovskite phase. In contrast, the PMN-PT film on 4° miscut Si shows a dramatic improvement in terms of both phase purity and epitaxy of perovskite phase. It has no detectable pyrochlore phase. Azimuthal f scans of this phase-pure PMNPT film show in-plane epitaxy with a cube-oncube epitaxial relation, [100]pc PMN-PT//[100]pc SrRuO3//[100] SrTiO3//[110] Si (Fig. 1C). The full width at half maximum (FWHM) of the 002pc w scan and the 101pc f scan is 0.26° and 0.6°, respectively. Commercial PMN-PT bulk single crystals have a FWHM of 0.14° and 0.27°,
Downloaded from www.sciencemag.org on November 17, 2011
appearance of a pyrochlore phase; a multistep columbite process was developed to bypass the pyrochlore phase in bulk PMN-PTceramics, where magnesium and niobium oxides are reacted with PbO (9). However, such a method is not applicable to thin-film synthesis. Thus, thin-film growth conditions must be carefully controlled to achieve (001)pc oriented single-crystal PMN-PT films. To stabilize the perovskite phase and achieve epitaxial growth of the desired orientation, we have implemented two approaches: (i) the use of
chiometry, orientation) and processing-related (phase purity, defect density) factors. In many ferroelectric solid solutions, the physical properties are maximized at the morphotropic phase boundary, which occurs at 33% PT in the PMNPT solid solution system (8). In rhombohedral or morphotropic phase boundary PMN-PT, (001)pc oriented single crystals (pc, pseudocubic) provide large, nonhysteretic piezoelectric coefficients compared with (111)pc oriented crystals (7). One of the major processing difficulties is the
0
Electrical field (kV/cm)
Fig. 2. Dielectric and piezoelectric properties of the PMN-PT film. (A) Polarization versus electric field measurement on a 1-mm PMN-PT thin film on Si. (B) Dielectric permittivity versus electric field measurement at 1 kHz with a 30 mVrms (rms, root mean square) oscillating voltage. (C) Comparison of figure of merit for micromachined actuators (e31,f ) and energy harvesters (e 231,f /eoe33) of a PMN-PT film from this work with www.sciencemag.org
SCIENCE
other reported values: from left, AlN (3), Bi0.5Na0.5TiO3-Bi0.5K0.5TiO3-BaTiO3 (BNT-BKT-BT) (29), K0.5Na0.5NbO3 (KNN) (30), 0.65Pb(Mg0.33Nb0.67)0.35PbTiO3 (PMN-35PT) (3), random Pb(Zr,Ti)O3 (Random PZT) (3), (001) 0.3Pb(Ni0.33Nb0.67)O3-0.7Pb(Zr0.45Ti0.55)O3 (001 PNN-PZT) (31), (001) Pb(Zr0.52Ti0.48)O3 (001 PZT) (3), and gradient-free (001) Pb(Zr0.52Ti0.48)O3 (gradient-free PZT) (32). VOL 334
18 NOVEMBER 2011
959
respectively. Transmission electron microscopy (TEM) analysis also confirms epitaxial growth of the PMN-PT heterostructure on Si. Figure 1D is a bright-field TEM image with low magnification. The inset shows the selected-area electron diffraction (SAED) pattern taken from the PMN-PT layer along the [010]pc zone axis. The high-resolution TEM image (Fig. 1E) exhibits an atomically sharp interface between the SrRuO3 and PMN-PT layers; the epitaxial match between the layers is clear. Also, chemical composition analysis by wavelength dispersive spectroscopy shows that the PMN-PT films are stoichiometric with 0.67PMN-0.33PT composition within experimental error. These analyses confirm that PMN-PT thin films on miscut Si in this work are stoichiometric, phase-pure, (001)pc oriented single crystals. The polarization–versus–electric field (P-E) hysteresis loops were measured for 1-mm-thick PMN-PT films on Si (Fig. 2A). We observe that the P-E loops are negatively imprinted, indicating the existence of a built-in bias of magnitude –38 kV/cm. As shown in Fig. 2B, the relative dielectric permittivity and loss tangent versus electric field measurement exhibits the same negative shift. This has a number of important consequences. First, the built-in bias increases the magnitude of the remanent polarization (Pr). Typically, a compressive in-plane stress results in a counter-
A
B
from the asymmetry of the top and bottom electrodes (as shown in fig. S4) (17). The effective transverse piezoelectric coefficients e31,f were obtained via a modified wafer flexure method (18). The induced charge was measured while subjecting the 1-mm-thick PMNPT films to a cyclic ac strain to obtain e31,f. The e31, f coefficient is the material’s figure of merit for the majority of micromachined piezoelectric charge-based sensors as well as actuators (3), enabling performance comparisons between systems of different detailed geometries. The substantial self-polarization translated into high e31, f coefficients (from –12 to –22 T 3 C/m2) for the asdeposited PMN-PT films. After short poling steps, the highest e31, f measured was –27 T 3 C/m2. Figure 2C compares these PMN-PT e31, f values to a wide range of previous work, demonstrating that these values are the highest reported for any piezoelectric thin films. This translates directly to a reduction in the required driving voltage for piezoelectric MEMS actuators. As an example, replacing the randomly oriented PZT actuators in the radio frequency MEMS switches reported in (15) with these PMN-PT films would decrease the voltage required to minimize contact resistance from ≈8 to 10 V to ≤2.5 V. This, in turn, enables actuators that can be driven by low-voltage complementary metal-oxide semiconductor circuits with high actuation authority. We have confirmed
Camera
LED Illuminator White-light source
Zoom lens Beamsplitter Objective Reference surface Beamsplitter
VAC VDC
5µm
DUT
D
C 0.4
Length: 34µm
2.8V
0.8
1.0
Tip displacement (µm)
0.0
-0.2
-0.4
Cantilever clamping point
0.6
0.4
0.2
∆Z (µm)
0.8
0.2
Height (µm)
Fig. 3. Fabrication and characterization of PMN-PT microcantilevers. (A) SEM image of a PMN-PT cantilever. (B) Schematic illustration of the measurement set-up. Both dc and ac fields can be applied between the top and bottom electrodes. LED, light-emitting diode; DUT, device under test. (C) PMN-PT cantilever profile as a function of dc voltage. (D) Tip displacement of the 34-mm-long cantilever versus applied dc voltage. (Inset) Linear scale of tip displacement. The dotted red line denotes the simulation result for a PMN-PT cantilever with the same geometry. The blue dotted line represents the modeled result for an electrostatic cantilever with comparable geometry. DZ, tip displacement.
clockwise rotation of the P-E loops of perovskite ferroelectrics, increasing the measured remanent polarization, whereas in-plane tensile stresses act in the opposite way (13). Because of the difference of thermal expansion coefficients between PMN-PT and Si, PMN-PT films on silicon are typically under a tensile strain, decreasing their Pr values. Here, the preferred polarization direction (imprint) displaces the somewhat tilted hysteresis loop, increasing the positive remanent polarization to ≈19 mC/cm2, which in turn increases the piezoelectric response at zero applied field. Second, the built-in bias stabilizes the polarization of PMN-PT in a certain direction (downward for these films). As a result, piezoelectric devices built from these films should show substantially less aging in the piezoelectric response (14) and will be more robust against depolarization due to voltage or temperature excursions (15). This mitigates one of the known problems with undoped PMN-PT single crystals; i.e., that the low coercive field limits the reverse drive voltages that can be applied. The high levels of imprint achieved here exceed those reported for Mn- and Zr-doped PMN-PT crystals (16). Third, the built-in bias decreases the permittivity at zero field, as shown in Fig. 2B. This increases the figure of merit for sensors operating in a voltage-sensing mode, as well as for piezoelectric energy-harvesting systems. One contribution to the built-in bias arises
Modeled PMN-PT cantilever
0.6 0.4 0.2 0.0 0.0
1.5
3.0
Voltage (V)
Actual PMN-PT cantilever
-0.6 0.0
0V -0.8 0
0.02
0.04
Modeled electrostatic cantilever 100
Length (mm)
960
18 NOVEMBER 2011
VOL 334
SCIENCE
www.sciencemag.org
101 Voltage (V)
102
Downloaded from www.sciencemag.org on November 17, 2011
REPORTS
REPORTS
e2
merit of energy-harvesting devices, e031;e33f , where e33 is the relative dielectric permittivity and e0 is the permittivity of free space. The efficiency of a piezoelectric energy-harvesting device is very generally related to both its piezoelectric and elastic responses (19). But in piezoelectric thinfilm energy-harvesting devices, in which passive layers dominate the elastic response (20–22), the above figure of merit describes the performance of PMN-PT relative to other piezoelectric materials (23, 24). Figure 2C compares our films’ figure of merit with that of PZT and other piezoelectric materials, demonstrating the advantages of these PMNPT films for energy-harvesting applications. We have confirmed this figure-of-merit comparison with comparative simulations of end-loaded PMNPT and PZT cantilevers using a self-consistent data set (shown in table S4) (17). The high figure of merit observed here is a consequence of both the large piezoelectric response and the imprint in the hysteresis loop, which reduces the permittivity at zero applied field. Note that in cases where the elastic energy is stored primarily in the piezo electric, a more useful figure of merit is
e231; f ð1 − vÞ2 e0 e33 Y ,
where Y and v are Young’s modulus and Poisson’s ratio, respectively. Given the comparable Young’s modulus and Poisson’s ratio between PMN-PT (115 GPa) (25) and PZT (~100 to 150 GPa) (26, 27), PMN-PT still shows at least a twofoldhigher figure of merit than the best-reported PZT. It is also important to verify that these large piezoelectric coefficients are not degraded by the microfabrication process. To confirm this, we have fabricated PMN-PT cantilevers, a prototypical structure for electromechanical devices using the 31 mode. We then compared the resulting behavior with a model based on parameters from single-crystal PMN-PT. Figure 3A shows the scanning electron microscopy (SEM) image of a cantilever with Pt (60 nm)/PMN-PT (270 nm)/SrRuO3 (100 nm)/SrTiO3 (13 nm) fabricated by conventional microfabrication techniques (shown in fig. S1) (17). The motion of the PMN-PT cantilevers was characterized by white light interferometry while applying dc or ac voltages across the top and bottom electrodes, as shown in Fig. 3B. These measurements were performed at room temperature and ambient pressure. Figure 3C shows the measured profile of a 34-mm-long cantilever as a function of applied voltage. Note that all the cantilevers are bent downward after release from the Si substrate. This is attributed to the relaxation of the epitaxial strain between PMN-PT (apc = 0.402 nm) and SrRuO3 (apc = 0.393 nm). When an electric field is applied parallel to the remanent polarization (i.e., a positive voltage is applied on the top elec-
trode), the PMN-PT contracts laterally. The vertical displacement arises from bending of the cantilever due to the structural asymmetry between the top (60-nm Pt) and bottom (≈110-nm SrRuO3/SrTiO3) layers. Figure 3D shows the tip displacement of the cantilever as a function of the applied voltage. The tip moves 0.375 T 0.005 mm/V, as shown in the inset of Fig. 3D. Finite element simulation was performed using the material parameters reported for bulk single-crystal 0.67 PMN-0.33PT from (25) (fig. S3) [see also (17)]. The experimental and modeled results are consistent, as shown by black and red lines in Fig. 3D, respectively. Note that the actual displacement of the cantilever also depends on geometry; for example, the thickness of the passive layer beneath the piezoelectric layer (28). Here, we emphasize that high piezoelectric coefficients are still manifested in the devices made from the PMN-PT films. A commonly used technique for actuating MEMS devices is electrostatics, which is straightforward to implement as it requires only reasonable conductivity of the device structural materials (2). A major drawback of electrostatic actuation, however, is the required high control voltage, which places a serious demand on the drive electronics, often preventing scaling down and denser integration of the electronic drivers for actuator arrays. Piezoelectric actuation considerably lowers the driving voltage while achieving the same displacement. The simulation shows that an electrostatic cantilever with comparable geometry to our PMN-PT cantilever requires much higher voltages for similar displacement, as shown in Fig. 3D. The strong piezoelectric activity of these PMN-PT thin films will dramatically enhance the freedom of designing small electromechanical devices with better performance. These thin films can provide a wide range of piezoelectric device applications such as ultrasound medical imaging, microfluidic control, piezotronics, and energy harvesting. The geometries fabricated and discussed here take advantage of the 31 piezoresponse. Devices with PMN-PT thin films exploiting the 33 response (20) would also benefit from the enhanced piezoresponse of PMN-PT over PZTand additionally from the higher coupling factor in the 33 mode (21). Although technically challenging for thin films, shear-mode–based structures would also benefit. Silicon on insulator substrates will enable more complex device structures with precisely controlled passive-layer thicknesses to control stiffness and displacement. Beyond electromechanical devices, epitaxial heterostructures with giant piezoelectricity will open a new avenue to tune and modulate multifunctional properties such as ferroelectric, ferromagnetic, superconducting, and multiferroic materials by dynamic strain control. References and Notes 1. K. Uchino, Piezoelectric Actuators and Ultrasonic Motors (Kluwer Academic, Boston, 1996). 2. J. Judy, Smart Mater. Struct. 10, 1115 (2001).
www.sciencemag.org
SCIENCE
VOL 334
3. S. Trolier-McKinstry, P. Muralt, J. Electroceram. 12, 7 (2004). 4. A. K. Sharma et al., Appl. Phys. Lett. 76, 1458 (2000). 5. M. D. Nguyen, H. N. Vu, D. H. A. Blank, G. Rijnders, Adv. Nat. Sci.: Nanosci. Nanotechnol. 2, 015005 (2011). 6. D. Isarakorn et al., J. Micromech. Microeng. 20, 055008 (2010). 7. S. E. Park, T. R. Shrout, J. Appl. Phys. 82, 1804 (1997). 8. S.-E. Park, T. R. Shrout, IEEE Trans. Ultrason. Ferroelectr. Freq. Control 44, 1140 (1997). 9. S. L. Swartz, T. R. Shrout, Mater. Res. Bull. 17, 1245 (1982). 10. S. Bu et al., Appl. Phys. Lett. 79, 3482 (2001). 11. C. B. Eom et al., Science 258, 1766 (1992). 12. Y. Liang et al., J. Appl. Phys. 96, 3413 (2004). 13. Z. Zhang, J. H. Park, S. Trolier-McKinstry, MRS Proc. Ferroelectric Thin Films VIII 596, 73 (2000). 14. R. G. Polcawich, S. Trolier-McKinstry, J. Mater. Res. 15, 2505 (2000). 15. R. G. Polcawich et al., IEEE Trans. Microwwave Theory Tec. 55, 2642 (2007). 16. S. J. Zhang, S. M. Lee, D. H. Kim, H. Y. Lee, T. R. Shrout, Appl. Phys. Lett. 93, 122908 (2008). 17. Supporting material is available on Science Online. 18. J. F. Shepard Jr., P. J. Moses, S. Trolier-McKinstry, Sens. Actuators 71, 133 (1998). 19. S. Roundy, J. Intell. Mater. Syst. Struct. 16, 809 (2005). 20. N. E. Dutoit, B. L. Wardle, S. G. Kim, Integr. Ferroelectr. 71, 121 (2005). 21. K. A. Cook-Chennault, N. Thambi, A. M. Sastry, Smart Mater. Struct. 17, 043001 (2008). 22. R. Elfrink et al., J. Micromech. Microeng. 19, 094005 (2009). 23. T. M. Kamel et al., J. Micromech. Microeng. 20, 105023 (2010). 24. M. A. Dubois, P. Muralt, IEEE Trans. Ultrason. Ferroelectr. Freq. Control 45, 1169 (1998). 25. R. Zhang, B. Jiang, W. Cao, J. Appl. Phys. 90, 3471 (2001). 26. P. Delobelle, E. Fribourg-Blanc, D. Remiens, Thin Solid Films 515, 1385 (2006). 27. T. H. Fang, S. R. Jian, D. S. Chuu, J. Phys. Condens. Matter 15, 5253 (2003). 28. P. Muralt et al., IEEE Trans. Ultrason. Ferroelectr. Freq. Control 52, 2276 (2005). 29. M. Abazari, A. Safari, S. S. N. Bharadwaja, S. TrolierMcKinstry, Appl. Phys. Lett. 96, 082903 (2010). 30. K. Shibata, F. Oka, A. Ohishi, T. Mishima, I. Kanno, Appl. Phys. Exp. 1, 011501 (2008). 31. F. Griggio, S. Trolier-McKinstry, J. Appl. Phys. 107, 024105 (2010). 32. F. Calame, P. Muralt, Appl. Phys. Lett. 90, 062907 (2007). Acknowledgments: This work was supported by the NSF under grant no. ECCS-0708759 and a David Lucile Packard Fellowship (C.B.E.). Work at Penn State was supported by a National Security Science and Engineering Faculty Fellowship. The work was partly supported by a Multidisciplinary University Research Initiative through the Air Force Office for Scientific Research (AFOSR) under grant no. FA9550-08-1-0337 (R.H.B.). Work at the Univ. of Michigan was supported by the U.S. Department of Energy (DOE) under award DE-FG02-07ER46416 and by NSF under awards DMR-0907191 and DMR-0723032 (TEM instrument). Work at Argonne National Laboratory and use of the Center for Nanoscale Materials was supported by the DOE, Office of Science, Office of Basic Energy Sciences, under contract no. DE-AC0206CH11357. Work at Cornell was supported by AFOSR through award no. FA9550-10-1-0524. We thank N. Valanoor for helpful discussions.
Downloaded from www.sciencemag.org on November 17, 2011
this figure-of-merit analysis with comparative numerical modeling of PZT and PMN-PT actuators of similar geometries (see table S3) (17). In addition to the large e31, f values, the PMNPT films also provide large values for a figure of
Supporting Online Material www.sciencemag.org/cgi/content/full/334/6058/958/DC1 SOM Text Figs. S1 to S4 Tables. S1 to S4 References (33–50) 19 April 2011; accepted 6 October 2011 10.1126/science.1207186
18 NOVEMBER 2011
961
Ultralight Metallic Microlattices T. A. Schaedler,1* A. J. Jacobsen,1 A. Torrents,2 A. E. Sorensen,1 J. Lian,3 J. R. Greer,3 L. Valdevit,2 W. B. Carter1 Ultralight (<10 milligrams per cubic centimeter) cellular materials are desirable for thermal insulation; battery electrodes; catalyst supports; and acoustic, vibration, or shock energy damping. We present ultralight materials based on periodic hollow-tube microlattices. These materials are fabricated by starting with a template formed by self-propagating photopolymer waveguide prototyping, coating the template by electroless nickel plating, and subsequently etching away the template. The resulting metallic microlattices exhibit densities r ≥ 0.9 milligram per cubic centimeter, complete recovery after compression exceeding 50% strain, and energy absorption similar to elastomers. Young’s modulus E scales with density as E ~ r2, in contrast to the E ~ r3 scaling observed for ultralight aerogels and carbon nanotube foams with stochastic architecture. We attribute these properties to structural hierarchy at the nanometer, micrometer, and millimeter scales. he effective properties of low-density materials are defined both by their cellular architecture (i.e., the spatial configuration of voids and solid) and the properties of the solid constituent (e.g., stiffness, strength, etc.). In the ultralight regime below 10 mg/cm3, very few materials currently exist: silica aerogels [density r ≥ 1 mg/cm3 (1, 2)], carbon nanotube aerogels [r ≥ 4 mg/cm3 (3)], metallic foams [r ≥ 10 mg/cm3 (4, 5)], and polymer foams [r ≥ 8 mg/cm3 (6, 7)]. These materials have a wide range of applications, such as thermal insulation, shock or vibration damping, acoustic absorption, and current collectors in battery electrodes and catalyst supports (8). All of the ultralow-density materials mentioned above have random cellular architectures. This random cell structure results in some beneficial properties (e.g., high specific surface area and restriction of gas flow), but generally the inefficient distribution of the constituent results in specific properties (e.g., stiffness, strength, energy absorption, and conductivity) far below those of the bulk material (8). As an example, Young’s modulus, E, of ultralight stochastic materials scales poorly with density, generally following E ~ r3 (9), in contrast to the well-known E ~ r2 relationship for random open-cell foams with higher relative densities (8). In large-scale structures, it has been shown that introducing order and hierarchy can substantially improve material utilization and resultant properties. For instance, the Eiffel Tower possesses a relative density similar to that of lowdensity aerogels (10) but is clearly structurally robust. The size difference between the smallest and largest structural features will determine the degree of hierarchy that can be achieved. In this paper, we present a method for creating ordered hollow-tube lattice materials with a minimum scale of ~100 nm. Coupled with control over mmto cm-scale structural features, this enables us to bring the benefits of order and hierarchy down
T
to the materials level. The result is an ultralightweight cellular material with efficient material utilization, a Young’s modulus that follows E ~ r2, and the ability to recover from >50% compression while demonstrating large energy absorption upon cyclic loading. The base architecture of our metallic microlattices consists of a periodic array of hollow tubes that connect at nodes, forming an octahedral unit cell without any lattice members in the basal plane. Figure 1 illustrates how the microlattice architecture can be distilled into three levels of hierarchy at three distinct length scales: unit cell (~mm to cm), hollow tube lattice member (~mm to mm), and hollow tube wall (~nm to mm). Each architectural element can be controlled independently, providing exceptional control over the design and properties of the resulting micro-
1 HRL Laboratories Limited Liability Company, Malibu, CA 90265, USA. 2Department of Mechanical and Aerospace Engineering, University of California, Irvine, CA 92697, USA. 3Division of Engineering and Applied Sciences, California Institute of Technology, Pasadena, CA 91125, USA.
Fig. 1. Design, processing, and cellular architecture of ultralight microlattices. (A) Polymer microlattice templates are fabricated from a three-dimensional array of self-propagating photopolymer waveguides. (B) The open-cellular templates are electroless plated with a conformal Ni-P thin film followed by etch removal of the template. (C) Image of the lightest Ni-P microlattice fabricated with this approach: 0.9 mg/cm3. (D and E) Images of two as-fabricated microlattices along with a breakdown of the relevant architectural elements.
*To whom correspondence should be addressed. E-mail:
[email protected]
962
lattice. The architecture determines the relative density of the lattice, with the absolute density dictated by the film material. The fabrication process begins with solid microlattice templates fabricated by using a selfpropagating photopolymer waveguide technique. In this method, a thiol-ene liquid photomonomer is exposed to collimated ultraviolet (UV) light through a patterned mask, producing an interconnected three-dimensional photopolymer lattice (11). A wide array of different architectures with unit cell dimensions ranging from 0.1 to >10 mm can be made by altering the mask pattern and the angle of the incident light (12, 13). Here, we focus on architectures with 1- to 4-mm lattice member length L, 100- to 500-mm lattice member diameter D, 100- to 500-nm wall thickness t, and 60° inclination angle q, similar to the microlattices depicted in Fig. 1. Conformal nickelphosphorous thin films were deposited on the polymer lattices by electroless plating, and the polymer was subsequently etched out (table S1). The autocatalytic electroless nickel-plating reaction enables deposition of thin films with controlled thickness on complex shapes and inside pores without noticeable mass transport limitations. The ultralight microlattice essentially translates the deposited nanoscale thin film in three dimensions to form a macroscopic material where the base structural elements are hollow tubes. By controlling the reaction time, a 100-nm-thick uniform conformal coating can be achieved, resulting in a cellular material with a density of 0.9 mg/cm3 (Fig. 1). The density is calculated by using the weight of the solid structure but not including the
18 NOVEMBER 2011
VOL 334
SCIENCE
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
REPORTS
REPORTS Characterization of the base constituent by transmission electron microscopy (TEM) revealed that as-deposited electroless nickel thin films have an average grain size of ~7 nm, which is consistent with literature reports (14, 15). Energy-dispersive x-ray spectroscopy confirmed that the film composition is 7% phosphorous and 93% nickel by weight. Because the films were not annealed after deposition, they remained as a supersaturated solid solution of phosphorous in a crystalline
face-centered cubic nickel lattice with no Ni3P precipitates present (14). The 7-nm grain size renders electroless nickel thin films harder and more brittle than typical nano- and microcrystalline nickel. A hardness of 6 GPa and a modulus of 210 GPa were measured by nanoindentation and hollow tube compression experiments, respectively (16). Compression experiments on the as-formed microlattices showed a nearly complete recovery
Downloaded from www.sciencemag.org on November 17, 2011
weight of air in the pores, adhering to the standard practice for cellular materials. The density of air at ambient conditions, 1.2 mg/cm3, multiplied by its volume fraction would need to be added to express the density of the solid-air composite. This method to form microlattices allows significantly more control than typical methods for forming other ultralightweight materials, such as foams and aerogels, where nominally random processes govern porosity formation.
Fig. 2. Nickel microlattices exhibit recoverable deformation. (A) Before deformation. (B) 15% compression. (C) 50% compression. (D) Full recovery after removal of load. (E) Optical image of unit cell unloaded. (F) Example of node buckling under compression. (G) SEM image of node before testing. (H) SEM image of node after six compression cycles at 50% strain. (The compression test is shown in movie S1.)
Fig. 3. Multicycle compression test results of nickel microlattices. (A) Stress-strain curves of a microlattice with 14 mg/cm3 exhibiting recoverable deformation (compare with Fig. 2). (B) History of compressive modulus, yield stress, maximum stress, and energy loss coefficient during the first six compression cycles shown in (A). (C) Stress-strain curve of a microlattice with a density of 1.0 mg/cm3. (D) Stress-strain curves of a microlattice with 43 mg/cm3 showing deformation more typical for metallic cellular materials. (E) SEM micrograph of postnanoindentation mark in 500-nm-thick electroless nickel film demonstrating brittle behavior.
www.sciencemag.org
SCIENCE
VOL 334
18 NOVEMBER 2011
963
from strains exceeding 50% (movie S1). Figure 2, A to D, provides images of a 14 mg/cm3 microlattice sample (L = 1050 mm, D = 150 mm, t = 500 nm) during compression testing, and Fig. 3A conveys the corresponding stress-strain curve measured at a prescribed displacement rate of 10 mm/s. In this experiment, the sample was not constrained by face sheets or attached to any compression platens. Because of a small taper in lattice strut diameter, the deformation typically initiates at a particular surface. Upon first compression, the lattice exhibits a compressive modulus of 529 kPa, with deviations from linear elastic behavior starting at ~10 kPa. The stress decreases slightly after the peak, which is associated with buckling and node fracture events, and a broad plateau is subsequently established in the stress-strain curve as buckling and localized node fracture events spread through the lattice. Figure 2C shows the microlattice at 50% compression. Upon unloading, the stress drops rapidly but does not approach zero until the platen is close to its original position. After removing the load, the microlattice recovers to 98% of its original height and resumes its original shape (Figs. 2D and 3A). The stress-strain behavior corresponding to the first cycle is never repeated during subsequent testing. Rather, during a second compression, the peak stress is absent and the “pseudohardening” behavior changes, but the stress level achieved at 50% strain is only 10% lower than that after the first cycle. Consecutive compression cycles exhibit stress-strain curves nearly identical to those of the second compression. Stiffness and strength diminish with cycle number but remain almost constant after the third cycle (Fig. 3B). The microlattice also shows significant hysteresis during compression experiments. For the first cycle, we estimate the work done in compression to be 4.6 mJ/cm3 and the energy dissipation to be 3.5 mJ/cm3, yielding an energy loss coefficient (Du/u) of 0.77. This large energy dissipation results from extensive node microcracking and thus is limited to the first cycle. After three cycles, a nearly constant energy loss coefficient of ~ 0.4 is calculated (Fig. 3B). From this estimate, we extract a loss coefficient (tan d) of ~ 0.16 (17), an order of magnitude higher than for typical nickel foams (18). Figure 3C shows the stress-strain response of a 1.0 mg/cm3 sample with larger unit cells (L = 4 mm, D = 500 mm, t = 120 nm), illustrating that different microlattice architectures in the ultralow-density regime result in similar behavior [although this sample underwent an additional freeze-drying process step that caused some damage (table S1)]. Increasing the density and wall thickness eventually led to a compression behavior more typical for metallic cellular materials. Figure 3D shows the stress-strain curve for a 43-mg/cm3 sample (L = 1050 mm, D = 150 mm, t = 1400 nm), for which recovery upon unloading from 50% strain is essentially absent. Optical examination of the ultralight microlattices during deformation suggests that defor-
964
mation initiates by local buckling at the nodes (Fig. 2, E and F). A closer inspection of the microlattices by scanning electron microscopy (SEM) shows that cracks and wrinkles are introduced primarily at the nodes during compression (Fig. 2, G and H). This damage is responsible for the 1 to 2% residual strain observed after the first compression cycle, as well as for the drop in yield strength and modulus during subsequent compression cycles. Once stable “relief cracks” form at the nodes, the bulk microlattice material can undergo large effective compressive strains through extensive rotations about remnant node ligaments, but with negligible strain in the solid nickel-phosphorous material, thus requiring no further fracture or plastic deformation. This property results in the reversible compressive behavior shown in Figs. 2 and 3. The extremely small wall thickness–to–diameter ratio is essential for this deformation mechanism. Increasing this aspect ratio leads to excessive fracture and loss of recoverability (Fig. 3D). The rise in stress at strains of ~40% (Fig. 3A) is a result of increased interaction between lattice members after localized compression at the nodes and should not be confused with densification, which in these samples does not occur until strains exceed 90% (fig. S2). Similar stress-strain curves as presented in Fig. 3A are typical for viscoelastic polymer foams (19) and carbon nanotube forests (20) but not for metal-based materials, implying a nonconventional loss mechanism is present. Two energy-loss mechanisms could possibly explain the energy dissipation during compression cycles: (i) structural damping because of snapping events (e.g., kinking or local buckling of the trusses) and (ii) mechanical or Coulomb friction between contacting members (or a combination of both). This mechanical behavior is especially unexpected considering the relatively brittle nature of the constituent electroless nickel thin film, as evidenced by the formation of cracks near a residual indenta-
18 NOVEMBER 2011
tion mark (Fig. 3E) and the rapid collapse upon single hollow truss member compressions (16). The brittle nature of the film arises from its nanograined microstructure (~7 nm) that hinders plastic deformation by dislocation motion (21). However, at the bulk scale microlattices exhibit completely different properties, because the cellular architecture effectively transforms this brittle thin film into a ductile, superelastic lattice by enabling sufficient freedom for deformation and tolerance to local strains through formation of stable relief cracks. Plotting the relative compressive modulus, E/Es, of various fabricated microlattices versus their relative density, r/rs, shows that the modulus scales with (r/rs)2 (Fig. 4). This scaling law indicates bending-dominated mechanical behavior similar to open-cell stochastic foams (8). In contrast, other materials with densities < 10 mg/cm3, such as aerogels and carbon nanotube (CNT) foams, exhibit a steeper scaling of E/Es ~ (r/rs)3 (Fig. 4) (22, 23), because of inefficient load transfer between ligaments. (Incidentally, this also affects the structural stability of a self-supporting cellular material, imposing a lower bound on density.) We notice that, although the relative modulus of topologically designed periodic lattice materials [such as octet-truss lattices (24)] typically scales linearly with relative density (25, 26), we do not expect the same scaling for the microlattices described here because of the absence of struts in the basal plane (26) and the ultrathin-walled hollow nodes, both of which facilitate localized bending deformation (Fig. 2F). Nonetheless, the marked improvement in mechanical efficiency over any other existing ultralightweight material, achieved by controlling both dimensions and periodicity of the architecture, enabled a selfsupporting cellular material with a relative density an order of magnitude lower than previously realized. Further, by transforming a brittle Ni-P thin film into a three-dimensional cellular material and designing a hierarchical cellular architecture at three Fig. 4. Relative compressive modulus (defined as the measured Young’s modulus, E, divided by the Young’s modulus of the constituent solid, Es) of selected cellular materials at low relative density.
VOL 334
SCIENCE
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
REPORTS
REPORTS
References and Notes 1. Guinness Book of World Records, Least Dense Solid (2003), www.guinnessworldrecords.com/records-1/leastdense-solid/. 2. T. M. Tillotson, L. W. Hrubesh, J. Non-Cryst. Solids 145, 44 (1992). 3. J. Zou et al., ACS Nano 4, 7293 (2010). 4. A. Verdooren, H. M. Chan, J. L. Grenestedt, M. P. Harmer, H. S. Caram, J. Am. Ceram. Soc. 89, 3101 (2006). 5. B. C. Tappan et al., J. Am. Chem. Soc. 128, 6589 (2006). 6. M. Chanda, S. K. Roy, Plastics Technology Handbook (CRC Press, Boca Raton, FL, 2007). 7. B. A. S. F. Corporation, Materials Safety Data Sheet for Basotect V3012 (2007), www.basf.co.kr/02_products/ 01_thermoplastics/spe/document/MSDS-Basotect% 20V3012.pdf. 8. L. J. Gibson, M. F. Ashby, Cellular Solids: Structure and Properties (Cambridge Univ. Press, Cambridge, 1997). 9. H.-S. Ma, J.-H. Prévost, R. Jullien, G. W. Scherer, J. Non-Cryst. Solids 285, 216 (2001).
10. R. S. Lakes, Nature 361, 511 (1993). 11. A. J. Jacobsen, W. Barvosa-Carter, S. Nutt, Acta Mater. 55, 6724 (2007). 12. A. J. Jacobsen, W. Barvosa-Carter, S. Nutt, Adv. Mater. 19, 3892 (2007). 13. A. J. Jacobsen, W. Barvosa-Carter, S. Nutt, Acta Mater. 56, 2540 (2008). 14. S. H. Park, D. N. Lee, J. Mater. Sci. 23, 1643 (1988). 15. S. Y. Chang, Y. S. Lee, H. L. Hsiao, T. K. Chang, Metall. Mater. Trans. A 37A, 2939 (2006). 16. J. Lian et al., Nano Lett. 11, 4118 (2011). 17. G. F. Lee, B. Hartmann, J. Sound Vibrat. 211, 265 (1998). 18. M. F. Ashby et al., Metal Foams: A Design Guide (Butterworth-Heinemann, Burlington, MA, 2000), p. 43. 19. N. J. Mills, Cell. Polym. 5, 293 (2006). 20. A. Cao, P. L. Dickrell, W. G. Sawyer, M. N. Ghasemi-Nejhad, P. M. Ajayan, Science 310, 1307 (2005). 21. J. R. Trelewicz, C. A. Schuh, Acta Mater. 55, 5948 (2007). 22. J. Gross, T. Schlief, J. Fricke, Mater. Sci. Eng. A 168, 235 (1993). 23. M. A. Worsley, S. O. Kucheyev, J. H. Satcher Jr., A. V. Hamza, T. F. Baumann, Appl. Phys. Lett. 94, 073115 (2009). 24. V. S. Deshpande, N. A. Fleck, M. F. Ashby, J. Mech. Phys. Solids 49, 1747 (2001). 25. V. S. Deshpande, M. F. Ashby, N. A. Fleck, Acta Mater. 49, 1035 (2001). 26. V. S. Deshpande, N. A. Fleck, Int. J. Solids Struct. 38, 6275 (2001).
Silica-Like Malleable Materials from Permanent Organic Networks Damien Montarnal, Mathieu Capelot, François Tournilhac, Ludwik Leibler* Permanently cross-linked materials have outstanding mechanical properties and solvent resistance, but they cannot be processed and reshaped once synthesized. Non–cross-linked polymers and those with reversible cross-links are processable, but they are soluble. We designed epoxy networks that can rearrange their topology by exchange reactions without depolymerization and showed that they are insoluble and processable. Unlike organic compounds and polymers whose viscosity varies abruptly near the glass transition, these networks show Arrhenius-like gradual viscosity variations like those of vitreous silica. Like silica, the materials can be wrought and welded to make complex objects by local heating without the use of molds. The concept of a glass made by reversible topology freezing in epoxy networks can be readily scaled up for applications and generalized to other chemistries. hermoset polymers such as Bakelite must be polymerized in a mold having the shape of the desired object because once the reaction is completed, the polymer cannot be reshaped or reprocessed by heat or with solvent. In contrast, thermoplastics, when heated, can flow, which permits extrusion, injection, and molding of objects. Depending on the chemical nature of the plastic, during cooling, solidification occurs by crystallization or by glass transition. During vitrification, as the temperature is lowered below the glass transition, the viscosity abruptly in-
T
Matière Molle et Chimie, UMR 7167 CNRS-ESPCI, Ecole Supérieure de Physique et Chimie Industrielles, 10 rue Vauquelin, 75005 Paris, France. *To whom correspondence should be addressed. E-mail:
[email protected]
creases in a narrow temperature range, and the material becomes so viscous that it behaves essentially like a solid with an elastic modulus of about 109 to 1010 Pa (1). Nevertheless, compared to processable plastics, cross-linked polymers have superior dimensional stability; have high-temperature mechanical, thermal, and environmental resistance; and are irreplaceable in many demanding applications, such as in the aircraft industry. High-performance coatings, adhesives, rubbers, light-emitting diode lenses, and solar cell encapsulants are made of permanently cross-linked polymer networks as well. Making covalent links reversible could provide a way to combine processability, reparability, and high performance (2–6). Networks with bonds able to break and reform (7–9) or to exchange pairs of atoms (10) can relax stresses and flow.
www.sciencemag.org
SCIENCE
VOL 334
27. L. J. Gibson, M. F. Ashby, Proc. R. Soc. London Ser. A 382, 43 (1982). 28. Hexcel Corporation, HexWeb Honeycomb Attributes and Properties (1999); datasheet available at www.hexcel. com/Resources/DataSheets/Brochure-Data-Sheets/ Honeycomb_Attributes_and_Properties.pdf. 29. M. Moner-Girona, A. Roig, E. Molins, E. Martinez, J. Esteve, Appl. Phys. Lett. 75, 653 (1999). Acknowledgments: The authors gratefully acknowledge the financial support by Defense Advanced Research Projects Agency under the Materials with Controlled Microstructural Architecture program managed by J. Goldwasser (contract no. W91CRB-10-0305) and thank J. W. Hutchinson and C. S. Roper for useful discussions. A patent application regarding the structure and formation process of the ultralight microlattices has been submitted to the U.S. Patent and Trademark Office. The polymer waveguide process has been patented (U.S. Patent 7,382,959, U.S. Patent 7,653,279, and U.S. Patent 8,017,193), but the template can be fabricated in other ways.
Downloaded from www.sciencemag.org on November 17, 2011
distinct length scales, we demonstrated the emergence of entirely different mechanical properties. In addition to possible applications for an ultralight material with high energy absorption and recoverability, we anticipate that these results will help reshape our understanding of the interaction between material properties and structural architecture.
Supporting Online Material www.sciencemag.org/cgi/content/full/334/6058/962/DC1 Materials and Methods Fig. S1 Table S1 References Movie S1 25 July 2011; accepted 12 October 2011 10.1126/science.1211649
The challenge is to allow rapid reversible reactions at high temperatures or by a convenient stimulus and to fix the network at service conditions. In this context, cleavage or exchange reactions by addition-fragmentation in the presence of radicals offer interesting possibilities (5, 11–14). Scott et al. demonstrated photoinduced plasticity in cross-linked polymers (11). Similarly, reparability and self-healing can be induced either thermally (13) or photochemically (15, 16) in radical systems. However, these systems undergo unavoidable termination reactions that limit reversibility of the networks. In parallel, a completely different concept based on chemical equilibrium between bond breaking and reforming without irreversible side reactions has been developed (17–20). In these systems, heating has two effects: It displaces the equilibrium toward depolymerization and it accelerates the bond breaking and reforming rate (8, 9). The advantage of such reversible links is that both above-mentioned effects act together to bring fluidity and thus processability (5, 17–19). They are, however, detrimental to the network integrity and performance. Chen et al. have shown that to avoid flow and creep at service temperatures, one can rely, as in thermoplastics, on glass transition to quench the system (17). Unfortunately, the systems based on chemical equilibrium between bond breaking and reforming are sensitive to solvents because in the presence of a solvent, the chemical equilibrium is displaced toward network depolymerization and dissolution (19). We sought to show that reversible networks can flow while maintaining their integrity and insolubility at high temperature. The idea is to
18 NOVEMBER 2011
965
rely solely on exchange reactions, without the need of depolymerization-polymerization equilibria or termination reactions (Fig. 1A). The key is to design the chemistry so that at high temperature, exchange reactions enable stress relaxation and malleability and upon cooling, the exchanges become so slow that the topology of the network is essentially fixed and the system behaves like a soft solid. The reversible freezing of the topology controlled by exchange reaction kinetics will thus exhibit features of the glass transition such as cooling- and heating-rate dependence or physical aging. To demonstrate the concept, we used the wellestablished transesterification reaction, which proceeds by association of all partners into an intermediate state before separation into a new partnership (21, 22). We synthesized networks by classical epoxy chemistry: reaction of the diglycidyl ether of bisphenol A (DGEBA) and a mixture of fatty dicarboxylic and tricarboxylic acids. We chose the epoxy/COOH 1:1 stoichiometry to have both –OH and ester groups in the final material and checked by infrared spectroscopy the complete conversion of epoxy groups (fig. S1). Examples of transesterification reactions allowed in this system are illustrated in Fig. 1B. The trans-
esterification kinetics can be controlled conveniently by a large variety of catalysts. Guided by a study of transterification kinetics of model molecules, we chose to work with zinc acetate [Zn(ac)2] (figs. S2 to S5). At room temperature, the cross-linked network behaves like an elastomer (figs. S6 to S9). It has a modulus of about 4 MPa and elongation and stress at break of about 180% and 9 MPa, respectively. Infrared spectroscopy indicated that the reaction is complete and that the number of ester links does not vary when the samples are heated (fig. S10). The permanence of the network was confirmed by dissolution experiments. They showed that samples swell, but do not dissolve, in good solvents even after immersion at high temperature and for a long time. Figure 2A shows swelling data for trichlorobenzene. Rheology and birefringence studies indicated that even though the network is insoluble, it is able to completely relax stresses at high temperatures and to flow (Fig. 2B). The inset in Fig. 2B presents the temperature variation of viscosity, which follows the simple Arrhenius law with an activation energy of ~80 kJ/mol K. Notably, the stress relaxation times are about equal to exchange reaction times measured for model mol-
Fig. 1. Topological rearrangements via exchange reactions preserving the network integrity. (A) Schematic view of a network with exchange processes that preserve the total number of links and average functionality of cross-links. The middle image illustrates that the exchange does not require depolymerization in the intermediate step. (B) Exchange process via transesterification in hydroxy-ester networks.
Fig. 2. Flow and insolubility properties of an epoxy network with 5 mol% Zn(ac)2 catalyst. (A) Swelling during immersion in trichlorobenzene. Temperature and time of immersion are indicated on the histogram. (B) Normalized stress relaxation at different temperatures. The inset shows the temperature variation of zero-shear viscosity. (C) A cross-linked sample broken into pieces is reprocessed in an injection machine to recover its initial
966
18 NOVEMBER 2011
VOL 334
ecules, and so is the activation energy (fig. S5). At 100°C, the measured relaxation time is ~58 hours. Extrapolated to 40°C the value would be ~1 year, and at room temperature ~6 years. Broken or ground samples, despite being permanently cross-linked well beyond the gel point, can be reprocessed by injection molding (Fig. 2C). By adapting the mold temperature and dwell time, molding without shrinkage can be achieved. The gradual Arrhenius-like variation of viscosity enables manufacturing techniques usually limited to a few inorganic glasses. Objects of complex shapes can be easily made without resort to a mold by local heating, deformation, relaxation of residual stresses, and welding if necessary. An elastomer fusilli made by twisting a cross-linked ribbon is shown in Fig. 2D. Because viscosity does not decrease abruptly with temperature, a precise control of temperature is not necessary, and tools such as a hot air blower are sufficient (movie S1). The concept of exchangeable links can also be applied to design materials that are hard at room temperature and malleable but insoluble at elevated temperatures. Widely used resins made by epoxy-anhydride reactions possess hydroxy groups and ester links (23), and thus it is possible to take advantage of transesterification exchanges merely by adding an appropriate catalyst to classical formulations. We synthesized networks by reaction of DGEBA with glutaric anhydride with epoxy/acyl 1:1 in the presence of 5 or 10 mol% zinc acetyl acetonate [Zn(acac)2]. We verified by infrared spectroscopy and swelling experiments that the network does not depolymerize even after a long time at high temperatures (fig. S11). For example, in trichlorobenzene at 180°C, once the swelling equilibrium is reached, the swelling stays constant even after 16 hours of immersion. The material behaves like a classical hard epoxy resin with the glass transition at ~80°C (fig. S12), modulus of ~1.8 GPa (fig. S13), and stress at break of ~55 MPa at room temperature (Fig. 3A). However, in contrast to classical epoxyanhydride resins, transesterification reactions and resulting topology rearrangements allow the network to flow. For example, at 200°C, for the network containing 10 mol% catalyst in an elongational creep experiment, after the transient
aspect and properties. No shrinkage is observed after demolding. (D) A fusillishaped elastomer made by local heating from a cross-linked ribbon of length 10 cm is reversibly deformed by a weight of 1.4 kg. SCIENCE
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
REPORTS
REPORTS
regime the deformation varies linearly with time. The viscosity estimated from the slope is found to be ~1.2 × 1010 Pa s. Measurements at different temperatures show an Arrhenius dependence with activation energy of ~88 kJ/mol K. The cross-linked material that has been ground into a fine powder can be reprocessed and reshaped by compression molding at high temperature. Three minutes of molding at 240°C suffice to produce a recycled object having essentially the same mechanical properties and insolubility as the original one (Fig. 3A). Complex shapes can be wrought at high temperature without using molds. A helical fusilli-like hard epoxy object (Fig. 3B) can be made from a ribbon by successive twists and stress relaxation or by slow torsion at high temperature. Another, more conventional, method of recycling could consist of depolymerizing networks by breaking ester links through hydrolysis or alcoholysis at high temperature and pressure. Fundamentally, at high temperatures, a network with exchangeable links behaves like a viscoelastic fluid. Yet, it differs qualitatively from polymer melts whose flow properties are mainly controlled by monomer friction. Indeed, even well above the glass transition temperature, when the monomer friction is low, the exchange reaction time can be very slow and become commensurable with the experimental time scale. In such a case, the material properties become dependent on thermal history. Thus, we anticipate, for example, that during a cooling ramp, there is a temperature at which network topology rearrangements become too sluggish to be effective. Below that temperature, the cross-links and the network topology appear to be quenched. Only on further cooling will the local monomer motions become frozen, and a classical glass transition to a hard glass will take place. Both elastomer and hard
Fig. 4. Organic strong glass-former. (A) Temperature dependence of thermal expansion of networks with 0.1 mol % (purple) and 5 mol % Zn(acac)2 (black) for various heating rates. Traces are shifted for clarity. (B) Angell fragility plot (26) showing viscosity as a function of inverse temperature normalized to 1 at the glass transition temperature (Tg) for epoxy/anhydride with 5% (red squares) and 10% (black squares) Zn(acac)2; for epoxy/acid with 5% (green squares) and 10% (blue squares) Zn(ac)2; and for silica (27), polystyrene (28), and terphenyl (29). glass are liquids quenched in a metastable, outof-equilibrium state [compare (24)]. Dilatometry experiments provide a classical tool (25) to reveal glass transitions and their thermal history dependence. Cross-linked networks are known to exhibit a lower expansion coefficient than the corresponding non–crosslinked polymers. For the sample with 0.1 mol% Zn(acac)2 catalyst, the linear expansion coefficient remains constant from 50° to 250°C, as expected for a permanently cross-linked network (Fig. 4A). When more catalyst is present (5 mol%), the exchange reactions are faster and an increase in expansion coefficient is observed at a temperature of ~165°C and heating rate of 5 K/min. Notably, the transition is continuous and heating-rate dependent, as expected for a glass transition. The topology freezing transition is well separated from the glass transition, which is visible at lower temperature for both samples with and without a catalyst. Both the topology freezing (liquidelastomer) and classical glass transition shift to higher temperatures when the heating rate is increased. Conventionally, the liquid-to-glass transition temperature is the point at which the viscosity becomes higher than 1012 Pa s (25, 26). For the epoxy-acid networks, viscosity studies give a transition temperature of ~68°, 57°, and 53°C for samples with 1, 5, and 10 mol% Zn catalyst, respectively. The rate of change of viscosity evaluated at the glass transition temperature (the “fragility”) gives a measure of the broadness of the glass transition (26). Silica and a few other inorganic compounds, such as P2O5, show a very broad Arrhenius-like variation and are therefore called “strong” glass formers. All organic and polymer liquids are “fragile”; they show a more rapid increase of viscosity upon cooling than predicted by the Arrhenius equation (Fig. 4B). By contrast, our networks behave like silica. Because the activation energies for monomer
www.sciencemag.org
SCIENCE
VOL 334
Downloaded from www.sciencemag.org on November 17, 2011
Fig. 3. Mechanical properties, malleability, and recyclability of epoxy-anhydride network with 10 mol% Zn(acac)2 catalyst. (A) Tensile test for samples as synthesized (solid lines) and after grinding into a powder and remolding (dashed lines). (B) Fusilli-shaped hard thermoset made by local heating from a cross-linked ribbon of length 10 cm is practically not deformed by a weight of 1.4 kg.
friction and for exchange reactions are different, this topology freezing transition can occur above the classical glass transition temperature. Macroscopically, the topology freezing transition manifests itself like a glass transition except that below the transition, the material behaves like an elastomer and not like a hard glass. The exchange reactions follow the Arrhenius law, and therefore the stress relaxation time and the viscosity also vary as predicted by the Arrhenius equation. The system can be termed a strong glass-former or strong organic liquid as opposed to strong inorganic liquids like silica and to organic glass-formers or polymers that are fragile liquids [compare (26)]. We have designed and realized covalently crosslinked organic networks that behave like silica. The underlying concept is to allow for reversible exchange reactions by transesterification that rearrange the network topology while keeping constant the total number of links and the average functionality of cross-links. The chemistry is versatile, relies on readily available ingredients, and does not require any special equipment. The production of malleable, reparable, recyclable and yet insoluble epoxy networks described here could potentially affect many industries that rely on elastomers, thermosetting polymers, and composites. The experimental and theoretical studies of networks with reversibly exchangeable links and a controllable number of defects could yield insights into the physics of glasses, while the control of glassy dynamics and glass transition with the catalyst is an unusual twist. References and Notes 1. J. Ferry, Viscoelastic Properties of Polymers (Wiley, New York, 1980). 2. S. J. Rowan, S. J. Cantrill, G. R. L. Cousins, J. K. M. Sanders, J. F. Stoddart, Angew. Chem. Int. Ed. 41, 898 (2002). 3. J.-M. Lehn, Prog. Polym. Sci. 30, 814 (2005). 4. T. Maeda, H. Otsuka, A. Takahara, Prog. Polym. Sci. 34, 581 (2009).
18 NOVEMBER 2011
967
REPORTS 16. Y. Amamoto, J. Kamada, H. Otsuka, A. Takahara, K. Matyjaszewski, Angew. Chem. Int. Ed. 50, 1660 (2011). 17. X. X. Chen et al., Science 295, 1698 (2002). 18. B. J. Adzima, H. A. Aguirre, C. J. Kloxin, T. F. Scott, C. N. Bowman, Macromolecules 41, 9112 (2008). 19. Y. Zhang, A. A. Broekhuis, F. Picchioni, Macromolecules 42, 1906 (2009). 20. P. Reutenauer, E. Buhler, P. J. Boul, S. J. Candau, J. M. Lehn, Chemistry 15, 1893 (2009). 21. R. V. Kudryavtsev, D. N. Kursanov, Zhurnal Obshchei Khimii 27, 1686 (1957). 22. J. Otera, Chem. Rev. 93, 1449 (1993). 23. C. A. May, Ed., Epoxy Resins: Chemistry and Technology (Dekker, New York, 1988). 24. R. T. Deam, S. F. Edwards, Philos. Trans. R. Soc. A 280, 317 (1976). 25. J. C. Dyre, Rev. Mod. Phys. 78, 953 (2006). 26. C. A. Angell, Science 267, 1924 (1995). 27. G. Urbain, Y. Bottinga, P. Richet, Geochim. Cosmochim. Acta 46, 1061 (1982). 28. D. J. Plazek, V. M. O’Rourke, J. Polym. Sci. A2 Polym. Phys. 9, 209 (1971).
Domain Dynamics During Ferroelectric Switching Christopher T. Nelson,1 Peng Gao,1 Jacob R. Jokisaari,1 Colin Heikes,2 Carolina Adamo,2 Alexander Melville,2 Seung-Hyub Baek,3 Chad M. Folkman,3 Benjamin Winchester,4 Yijia Gu,4 Yuanming Liu,5 Kui Zhang,1 Enge Wang,6 Jiangyu Li,5 Long-Qing Chen,4 Chang-Beom Eom,3 Darrell G. Schlom,2,7 Xiaoqing Pan1* The utility of ferroelectric materials stems from the ability to nucleate and move polarized domains using an electric field. To understand the mechanisms of polarization switching, structural characterization at the nanoscale is required. We used aberration-corrected transmission electron microscopy to follow the kinetics and dynamics of ferroelectric switching at millisecond temporal and subangstrom spatial resolution in an epitaxial bilayer of an antiferromagnetic ferroelectric (BiFeO3) on a ferromagnetic electrode (La0.7Sr0.3MnO3). We observed localized nucleation events at the electrode interface, domain wall pinning on point defects, and the formation of ferroelectric domains localized to the ferroelectric and ferromagnetic interface. These results show how defects and interfaces impede full ferroelectric switching of a thin film. erroelectric materials have numerous applications, including high-density and nonvolatile memories (1–3) and a broad range of electronic, optical, and acoustic devices (2). The utility of ferroelectrics is derived from a reversible transition between equivalent polar orientation states under an applied electric field and from that transition’s coupling to other material properties including strain (4, 5), magnetic order (6), and surface charge (7). Most of these applications require low-dimensional geometries such as thin films and deterministic control of
F
1 Department of Materials Science and Engineering, University of Michigan, Ann Arbor, MI 48109, USA. 2Department of Materials Science and Engineering, Cornell University, Ithaca, NY 14853, USA. 3Department of Materials Science and Engineering, University of Wisconsin–Madison, Madison, WI 53706, USA. 4Department of Materials Science and Engineering, Penn State University, University Park, PA 16802, USA. 5Department of Mechanical Engineering, University of Washington, Seattle, WA 98195, USA. 6School of Physics, Peking University, Beijing 100871, China. 7Kavli Institute at Cornell for Nanoscale Science, Ithaca, NY 14853, USA.
*To whom correspondence should be addressed. E-mail:
[email protected]
968
the local polarization state at the interface (8). Thus, it is critical to understand the process of polarization switching in epitaxial thin films. In this work, we studied the domain nucleation and evolution during switching of a ferroelectric BiFeO3 thin film, using in situ structural characterization by transmission electron microscopy (TEM). Polarization switching is induced by an applied electric field oriented along the film normal between a surface probe and a planar bottom electrode, the same geometry used for surface probe characterization depicted schematically in Fig. 1. Ferroelectric switching occurs through a process of inhomogenous nucleation and anisotropic growth of favorably oriented domains (9). In this geometry, it is typically modeled by a single nucleation event occurring at the probe contact with the film surface, the maximum of the applied field, followed by rapid propagation of the nucleated domain across the film and slow lateral growth (see simulation in fig. S1 and movie S1). The creep-type lateral expansion stage is well known for switching in thin films from surface probe measurements (10) and mod-
18 NOVEMBER 2011
VOL 334
SCIENCE
29. D. J. Plazek, C. A. Bero, I. C. Chay, J. Non-Cryst. Solids 172–174, 181 (1994). Acknowledgments: We gratefully acknowledge helpful discussions, with H. A. H. Meijer and A. J. Ryan on polymer processing, with K. Matyjaszewski on chemical reactions and catalysis, and with F. Krzakala and A. Maggs on glass transition. We are indebted to L. Breucker and S. Abadie for help with experiments. We acknowledge funding from ESPCI, CNRS and Arkema. The authors are declared to be inventors on three patents filed by CNRS related to the work presented here: L. Leibler, D. Montarnal, F. Tournilhac, M. Capelot, FR10.54213 (2010); FR11.50888 (2011); and FR11.50546 (2011).
Supporting Online Material www.sciencemag.org/cgi/content/full/334/6058/965/DC1 Materials and Methods Figs. S1 to S13 Table S1 Movie S1 15 August 2011; accepted 5 October 2011 10.1126/science.1212648
eling (11); however, experimental observations of the initial normal-axis growth process are absent except at much larger length scales (12, 13). In this work, we followed domain nucleation and evolution in cross-section and directly observed inhomogeneous nucleation events, domain wall pinning, and the formation of ferroelectric domains localized at interfaces. Local switching of a 100-nm (001)P oriented BiFeO3 ferroelectric film was performed by applying an electrical bias between an etched tungsten surface probe and a 20-nm La0.7Sr0.3MnO3 buffer electrode. This bilayer was grown on closely lattice-matched single-crystal (110)O TbScO3 substrates (compressive strain <0.14%) to avoid instabilities from epitaxial strain (14) and misfit dislocations (15) or flexoelectric effects from strain gradients (16) (O indicates orthorhombic indices and P indicates pseudocubic, where [110]O || [001]P). The bias between the surface probe and buffer electrode results in an inhomogenous out-of-plane electric field in the BiFeO3 film, promoting a transition among the eight possible <111>P polarization directions between the four “up” and four “down” orientation states. Three types of switching are possible, classified by the angular rotation of the polarization vector during swiching: 71°, 109°, or 180°. We find that the preferred switching path is the 71° rotation of the spontaneous polarization by the reversal of only the polarization component parallel to the applied field, in this case the film normal (z axis), in agreement with other studies (17–19). Piezoresponse force microscopy (PFM) indicates that the asgrown BiFeO3 films are predominantly upwardpoled. Figure 1A shows an out-of-plane PFM phase image from a 100-nm BiFeO3 film, which was locally switched by a positive probe bias of 20.6 V with a 2-s dwell time, producing a downward-poled domain ~400 nm in diameter. Diffraction contrast TEM was used to resolve the evolution of the domain structure along the depth of the film in situ. A cross-section of the 100-nm BiFeO3 film measured by PFM was
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
5. C. J. Kloxin, T. F. Scott, B. J. Adzima, C. N. Bowman, Macromolecules 43, 2643 (2010). 6. R. J. Wojtecki, M. A. Meador, S. J. Rowan, Nat. Mater. 10, 14 (2011). 7. M. S. Green, A. V. Tobolsky, J. Chem. Phys. 14, 80 (1946). 8. M. Rubinstein, A. N. Semenov, Macromolecules 31, 1386 (1998). 9. F. Tanaka, Polymer Physics: Applications to Molecular Association and Thermoreversible Gelation (Cambridge Univ. Press, Cambridge, 2011). 10. L. Leibler, M. Rubinstein, R. H. Colby, J. Phys. II 3, 1581 (1993). 11. T. F. Scott, A. D. Schneider, W. D. Cook, C. N. Bowman, Science 308, 1615 (2005). 12. Y. Higaki, H. Otsuka, A. Takahara, Macromolecules 39, 2121 (2006). 13. R. Nicolaÿ, J. Kamada, A. van Wassen, K. Matyjaszewski, Macromolecules 43, 4355 (2010). 14. H. Y. Park, C. J. Kloxin, T. F. Scott, C. N. Bowman, Macromolecules 43, 10188 (2010). 15. B. Ghosh, M. W. Urban, Science 323, 1458 (2009).
REPORTS
Fig. 1. Thin-film ferroelectric switching by a surface probe. (A) An out-of-plane PFM phase image shows the reversed downward-poled domain formed by application of a 20.6-V bias to a probe on the surface of the 100-nm BiFeO3 film. (B) A thin cross-sectional TEM image of the same film before and after switching by a 4-V bias. Switching occurs by 71° rotation of the polarization beneath the tungsten tip.
A SPM SPM Tip tip
1)
2)
0 Volts
g
3)
0.9 V
2.2 V
4)
100 nm
P
P
[1 11 ]
[1 11 ]
2.2 V
5)
6)
4V
8)
0V
7)
0V
------------
P[111]
nucleation site -300 nm
D 0 MV/m
60nm
0 nm
C
-70 300 nm
MV/m
10
nucleation site(s) -300 nm
0 nm
-40 300 nm
Domain Width (nm) 600 300 0
B
-5
0
Bias (V)
5
10
Fig. 2. (A) A chronological TEM dark-field image series formed by using the reflection g = 040P depicts the evolution of a P[111]P domain from a single-domain P[111]P film. Nucleation occurs at the La0.7Sr0.3MnO3 electrode interface at 0.9 V (images 1 to 3), producing a metastable stationary domain. Additional bias increases the size and number of these domains up to 2.2 V (image 4). At 2.2 V, the domains merge and propagate forward just short of the surface (image 5), where they remain pinned along a (001)P plane even after the voltage is nearly doubled (image 6). The lateral extent of the domain is reduced slightly with the removal of the bias (image 7) and remains unchanged after the tip is removed (image 8). (B) Normal component of the electric field from the surface probe. (C) Normal component of the electric field, including top and bottom Schottky junctions. (D) A ferroelectric hysteresis loop formed by the primary 71° switching as determined by the area of a switched domain. www.sciencemag.org
SCIENCE
VOL 334
responding to the same 71° switching path observed in the full film. The polarization direction was determined from the cation positions, using high-resolution scanning TEM (STEM) in highangle annular dark field (HAADF) mode (20, 21) (fig. S3). The nucleation of ferroelectric domains occurs when the applied field exceeds a critical nucleation threshold (the coercive field), which is subject to regional variation from defects (22). The interface is the expected nucleation site because of its high free energy due to the broken symmetry, strain, electric fields, charge, and altered chemical structure, which decrease the nucleation barrier (16). The high field concentration formed at the tip of the surface probe is generally assumed to dominate over any built-in electric fields, such as from Schottky junctions, and to initiate switching at the free surface, where the applied field is highest. However, a chronological series of TEM micrographs in Fig. 2 of 71° P[111]P to P[111 ]P switching during a slow ramp of the dc probe bias from 0 to 4 V clearly shows multiple nucleation events occurring exclusively at the bottom interface (movie S2). The out-of-plane electric field, EZ, for this tip geometry at 2-V bias calculated by finite element analysis (Fig. 2B), is concentrated at the free surface. Hall measurements performed on nonbuffered 100-nm BiFeO3 films showed an n-type conductivity. Inclusion of built-in electric fields from Schottky barriers of 0.85 and 0.77 V from junctions of n-type BiFeO3, with La0.7Sr0.3MnO3 and tungsten, respectively (23), produces the total EZ field distributions shown in Fig. 2C. There is a broad band along the La0.7Sr0.3Mn03 interface where P[111 ]P nucleation is likely to occur because of a strong negative field (pointed toward the substrate), which is in agreement with TEM observations. The electrical properties of the Schottky junctions could not be probed because conduction was bulk-limited (fig. S4); however, they are a common feature of planar ferroelectric oxides (16, 24) and are known to determine nucleation sites between symmetric planar electrodes in BiFeO3 films (18). The relative strengths and distributions of the built-in fields vary, given assumptions about the Schottky barrier heights, the depletion width, and the electrode geometry. Regardless, the built-in fields shown in Fig. 2 are of the same order of magnitude as the tip field and cannot be neglected. The triangular domains that nucleated at the La0.7Sr0.3MnO3 interface were metastable, increasing (compare Fig. 2A, images 3 and 4) or decreasing their fixed size with the applied bias. Multiple nucleation events occured before a critical voltage of 2.2 V, whereupon they coalesced and expanded into a single domain that reached its full forward extent within a single 30-ms data sample. The triangular shape of these domains served to minimize depolarizing fields as the domain walls oriented closer to charge-neutral (011)P planes. During this low field stage of thermodynamically limited switching, nuclei extended
18 NOVEMBER 2011
Downloaded from www.sciencemag.org on November 17, 2011
1B, as well as TEM micrographs of the domain structure before and after switching by a 4 V dc bias, which exceeds the threshold switching voltage (Vc) for this region. A ~270 nm-wide P[111 ]P domain formed from the initial P[111]P state, cor-
thinned to electron transparency, ~60 nm in thickness (fig. S2), and biased between the tungsten probe in contact with the film surface and the grounded La0.7Sr0.3MnO3 bottom electrode. The sample geometry is shown schematically in Fig.
969
REPORTS
A
der the probe have a positive voltage offset, indicating a possible positive built-in field (Fig. 2D). The reduction of this offset during cycling (fig. S5) further suggests that it may be the result of polar defects “imprinted” to the original P[111]P polarization (25), which disassociate during cycling. A contribution to the offset is also expected because of dissimilar electrode thermal histories and work functions [4.8 eV for La0.7Sr0.3MnO3 (26) versus ~4.5 eV for polycrystalline tungsten
B g
1) -10 V
g
0V
P
2) 9 V
P
[1 11 ]
180° 1]
P [11
] 11
P [1
[1 11 ]
C θ
180°
1]
P [11
3) 12 V P
[1 11 ]
180°
4) 0 V
] 11
BiFeO 3 C
La 0.7Sr 0.3MnO 3 electrode
Fig. 4. Polarization mapping of coincident 71° and 180° switching. A region of a P[111]P BiFeO3 film is switched in the vicinity of a 180° P[111]P interfacial domain. (A) The interfacial domain appears as a dark band on the right half of the dark-field TEM image, g = 044P, as shown in the corresponding diagram. (B) A P[111]P domain formed by subsequent switching of this region, with the left side pinned mid-film producing a negatively charged domain wall. Polarization maps from the highlighted regions show there is (C) a single P[111]P polarization at the pinning boundary and that (D) the P[111]P domain remains at the interface.
970
18 NOVEMBER 2011
VOL 334
SCIENCE
www.sciencemag.org
P [1
4 nm
Downloaded from www.sciencemag.org on November 17, 2011
[1 11 ]
100 nm 71°
Fig. 3. Near-interface switching. (A) A chronological TEM image series of the P[111]P BiFeO3 cross-section as the voltage increases from –10 V to +12 V (images 1 to 3) and back to zero (image 4) formed from g = 040P diffraction of the vertical planes and (B) g = 04̅4P diffraction of inclined planes. 180orotated P[111]P interfacial domains and a 71o-rotated P[111]P primary domain appear sequentially, shown by the accompanied schematics. (C) A high-resolution HAADF image shows the polarization distribution at the edge of an interfacial domain, where the color and its intensity correspond to polarization angle and magnitude, respectively. The inset HAADF image shows an example of a point defect located at the domain wall.
The metastable behavior indicates an intrinsic thermodynamic instability of the nuclei such as from positive built-in fields or an accumulation of strain, which destabilize the domain away from the interface. We do not see this behavior in similar nonferroelastic 180° switching of tetragonal PbZr0.2Ti0.8O3 films, which suggests that strain from ferroelastic 71° switching is a contributing factor. Hysteresis loops measuring the switched domain area during cyclic switching un-
P
~200 nm laterally across the interface, over three times the contact area of the tip. Such delocalization of switching under a localized field may be an important consideration for surface probe measurements. More fundamentally, this process is a departure from the kinetically limited models used to describe ferroelectric switching (24). This is important for electrical characterization, because the switching transient does not necessarily correspond to the actual nucleation bias.
REPORTS domain wall, because the domain wall charging is absent (Fig. 4B). Although the vertical extent of the interfacial domains depends on the voltage history, the lateral edges are fixed. HAADF images of the inclined 180° domain walls show a high density of point defects, such as the heavy-element substitution in the Fe site shown in the insert in Fig. 3C, as well as defects in the oxygen site (fig. S7). The interaction of the domain walls with the point defects is likely to be strain-mediated because the domain walls are neutrally charged. This is not the case for the negatively charged horizontal domain walls created by pinning of the 71° domain, such as the shoulder on the left sides of Figs. 3A and 4B and the entire switched region in Fig. 2A. The pinning behavior points to ordered planes of charged defects as the active pinning site. Vacancies and defect dipoles are common in perovskite films (25, 28), especially those grown by molecular beam epitaxy (29). Furthermore, the pinning occurs at fixed inhomogeneously distributed sites, which are repeatable for any given region of the film (fig. S8 and movie S6). We could not see any structural defects associated with the pinning sites, any detectable thickness variation in a thickness map of the pinned region (fig. S1), or any systematic steps along the (001)P planes in a topographic atomic force microscopy scan of the ion-etched surface of a prepared TEM sample. The pinning barrier can be overcome with the application of a sufficiently large bias or by repositioning the probe (movie S7), confirming that the film is still ferroelectric and the switched domain is thermodynamically stable beyond the pinning site. Although the oxygen-scattering crosssection is too small for the oxygen site to be seen directly in the HAADF images, these results strongly support the theory of domain pinning by ordered planes of oxygen vacancies (30). The observed ferroelectric switching, although it may manifest in some applications as a 71° single-domain switching process, is found to be more complex than is conventionally assumed. In particular, reversible switching at the interfaces is impeded, which is of particular consequence for heterostructures dependent on the ferroelectric interface state, such as the BiFeO3/La0.7Sr0.3MnO3 film stack studied here. Film orientations that favor single strain-free switching paths such as (001)P tetragonal or (111)P rhombohedral ferroelectrics can avoid some of this complexity, but this is not possible for heterostructures such as this one, which require ferroelastic switching (31). We have also found that substantial contributions by built-in electric fields lead to a large delocalization of switching from the tip, especially the formation of interfacial domains. The energetics favoring the nucleation and growth of 71° domains despite the presence of favorably oriented yet oddly immobile preexisting 180° domains is unknown. Future studies on the interface structure and its effect on switching or the inclusion of additional electrostatic contributions such as from inhomogeneous space charges, depolarizing fields
www.sciencemag.org
SCIENCE
VOL 334
from finite screening, and flexoelectric effects may help elucidate this behavior. References and Notes 1. 2. 3. 4. 5.
J. F. Scott, C. A. Paz de Araujo, Science 246, 1400 (1989). J. F. Scott, Science 315, 954 (2007). V. Garcia et al., Nature 460, 81 (2009). K. Aizu, J. Phys. Soc. Jpn. 27, 387 (1969). F. Kubel, H. Schmid, Acta Crystallogr. Sect. B Struct. Commun. 46, 698 (1990). 6. Y. H. Chu et al., Nat. Mater. 7, 478 (2008). 7. P. Maksymovych et al., Science 324, 1421 (2009). 8. S. M. Wu et al., Nat. Mater. 9, 756 (2010). 9. W. J. Merz, Phys. Rev. 95, 690 (1954). 10. T. Tybell, P. Paruch, T. Giamarchi, J. M. Triscone, Phys. Rev. Lett. 89, 097601 (2002). 11. Y. H. Shin, I. Grinberg, I. W. Chen, A. M. Rappe, Nature 449, 881 (2007). 12. A. Kuroda, S. Kurimura, Y. Uesu, Appl. Phys. Lett. 69, 1565 (1996). 13. V. Gopalan, T. E. Mitchell, J. Appl. Phys. 83, 941 (1998). 14. M. P. Cruz et al., Phys. Rev. Lett. 99, 217601 (2007). 15. M. W. Chu et al., Nat. Mater. 3, 87 (2004). 16. A. K. Tagantsev, G. Gerra, J. Appl. Phys. 100, 051607 (2006). 17. D. Pantel et al., J. Appl. Phys. 107, 084111 (2010). 18. N. Balke et al., Adv. Funct. Mater. 20, 3466 (2010). 19. S. H. Baek et al., Nat. Mater. 9, 309 (2010). 20. C. L. Jia et al., Nat. Mater. 6, 64 (2007). 21. C. T. Nelson et al., Nano Lett. 11, 828 (2011). 22. S. Jesse et al., Nat. Mater. 7, 209 (2008). 23. S. J. Clark, J. Robertson, Appl. Phys. Lett. 90, 132903 (2007). 24. M. Dawber, K. M. Rabe, J. F. Scott, Rev. Mod. Phys. 77, 1083 (2005). 25. C. M. Folkman et al., Appl. Phys. Lett. 96, 052903 (2010). 26. M. P. de Jong, V. A. Dediu, C. Taliani, W. R. Salaneck, J. Appl. Phys. 94, 7292 (2003). 27. E. W. Müller, J. Appl. Phys. 26, 732 (1955). 28. G. L. Yuan, A. Uedono, Appl. Phys. Lett. 94, 132905 (2009). 29. J. Zhang et al., J. Vac. Sci. Technol. B 27, 2012 (2009). 30. J. F. Scott, M. Dawber, Appl. Phys. Lett. 76, 3801 (2000). 31. T. Zhao et al., Nat. Mater. 5, 823 (2006). Acknowledgments: This work was supported by the U.S. Department of Energy (DOE), Office of Basic Energy Sciences, Division of Materials Sciences and Engineering, under award DE-FG02-07ER46416 and partially by NSF under awards DMR-0907191 (P.G. and K.Z.), DMR-0820404 (J.R.J.), and DMR-0723032 (aberration-corrected TEM instrument); at Cornell University by Army Research Office (ARO) grant W911NF-08-2-0032; at Penn State University by the U.S. Department of Energy under award DE-FG02-07ER46417; at the University of Wisconsin–Madison by ARO grant W911NF-10-1-0362; and at the University of Washington by ARO grant W911NF-07-1-0410. The authors also acknowledge the National Center for Electron Microscopy at Lawrence Berkeley National Laboratory for their support under the DOE grant DE-AC02-05CH11231 for user facilities. The project was conceived and directed by X.Q.P.; C.T.N., D.G.S., and X.Q.P. wrote the manuscript; TEM experiments were performed and analyzed by C.T.N. and P.G. under the guidance of X.Q.P.; BiFeO3 and La0.7Sr0.3MnO3 films were fabricated using molecular beam epitaxy by C.H., C.A., and A.M. under the guidance of D.G.S.; PFM experiments were performed by J.R.J. and Y.M.L. under the supervision of J.Y.L.; Hall measurements were performed by K.Z. under the supervision of X.Q.P.; sputtered films were grown by C.M.F. and S.H.B. under the supervision of C.B.E.; phase-field simulations were carried out by B.W. and Y.J.G. under the supervision of L.Q.C.; and E.G.W. participated in the modeling and analysis of interfacial properties.
Downloaded from www.sciencemag.org on November 17, 2011
(27)], but these favor a dominance of the negativefield Schottky junction, which would produce a negative shift. An alternate contribution to this offset may stem from the positive-voltage branch switching by a nucleation event, and the subsequent negative-voltage branch backswitching by the shrinking of the switched domain (fig. S6 and movie S3). The critical field for nucleation and the threshold field for the onset of domain wall creep need not be the same. A 71° polarization reversal accounts for the majority of the area beneath the probe. However, this “primary” domain often does not extend the full width between the two electrodes. Despite the high-energy domain wall that is created, the forward-propagating domain is often pinned midway through the film, leaving the top layer unswitched as shown in Fig. 2A. Furthermore, a thin layer of film along the La0.7Sr0.3MnO3 buffer electrode layer undergoes an independent switching process from the primary domain. A chronological TEM image series in Fig. 3A shows the domain structure evolution as the film is biased from –10 to 20 V and back to zero. At a small positive bias, long horizontal domains confined to the interface appear (Fig. 3A, image 2, and movie S4) and increase in height with increasing bias. Ultimately, a primary 71° switched P[111 ]P domain forms directly beneath the tip (Fig. 3A, image 3). Both the interfacial and primary domains persist after the bias is removed (Fig. 3A, image 4). Dark-field images formed from reflections of the (01l)P planes (Fig. 3B) indicate that the interfacial domains (invisible in the image) do not form the same ferroelastic twin with the unswitched region as the primary domain (bright area). Polarization mapping from HAADF images confirms that the interfacial domains undergo a nonferroelastic 180° polarization rotation (Fig. 3C). The switching thresholds of the interfacial domains have a large negative-voltage offset as compared to the primary domain. They have a lower nucleation bias and form several hundred nanometers away from the tip. In contrast, they require much larger negative voltages to erase (movie S5) because of the need to overcome the negative field in the Schottky junction where they are located. The lateral edges of the interfacial domains are charge-neutral (011)P 180° domain walls. However, the nominally (001)P horizontal domain wall is unfavorable because of its negative charge from the tail-to-tail polarization vectors. Its graded appearance indicates that it is not a sharp (001)P boundary, but is inclined or corrugated to reduce the depolarizing field. Despite the energy cost of this domain wall, the interfacial domains exhibit long-term stability (>6 months and ongoing). Although the interfacial domains and primary switched domain do not overlap in Fig. 3A, Fig. 4 shows a region with an interfacial domain directly beneath the tip before switching (Fig. 4A). The subsequent formation of the primary 71° domain does not consume or otherwise alter the interfacial domain except to create a sharper (001)P
Supporting Online Material www.sciencemag.org/cgi/content/full/334/6058/968/DC1 Materials and Methods Figs. S1 to S8 References (32–34) Movies S1 to S7 14 April 2011; accepted 12 October 2011 10.1126/science.1206980
18 NOVEMBER 2011
971
Negative Frequency-Dependent Selection of Sexually Antagonistic Alleles in Myodes glareolus Mikael Mokkonen,1* Hanna Kokko,2 Esa Koskela,3 Jussi Lehtonen,2,4 Tapio Mappes,1 Henna Martiskainen,3 Suzanne C. Mills5 Sexually antagonistic genetic variation, where optimal values of traits are sex-dependent, is known to slow the loss of genetic variance associated with directional selection on fitness-related traits. However, sexual antagonism alone is not sufficient to maintain variation indefinitely. Selection of rare forms within the sexes can help to conserve genotypic diversity. We combined theoretical models and a field experiment with Myodes glareolus to show that negative frequency-dependent selection on male dominance maintains variation in sexually antagonistic alleles. In our experiment, high-dominance male bank voles were found to have low-fecundity sisters, and vice versa. These results show that investigations of sexually antagonistic traits should take into account the effects of social interactions on the interplay between ecology and evolution, and that investigations of genetic variation should not be conducted solely under laboratory conditions. he direction of evolutionary change generally cannot be predicted without taking into account interactions between conspecifics, be they competitors or mates (1, 2). Additionally, the fitness effects of particular genotypes can often be highly sex-specific (3). Alleles beneficial to a son’s reproduction can be detrimental for a daughter’s success, or vice versa, resulting in different optimal trait values (optima) between the sexes (4–8). The presence of these sexually antagonistic (SA) alleles in a variety of organisms is now widely acknowledged (9–11), but less is known about how variation in such alleles is maintained. All else being equal, SA selection does not decrease genetic variation as rapidly as selection that favors the same alleles in both males and females (6, 12). However, SA selection is not as beneficial for genetic diversity as true negative frequencydependent maintenance of alternative alleles, whereby rare genotypes are favored in reproduction (13). The process of variation depletion may end in intralocus sexual conflict that remains unresolved, such that the traits of males, females, or both do not match their optima [leading to a cost of SA alleles, the gender load (14)]. It may also end in resolved conflict if gene expression evolves to become more sex-specific (15). In the former case sexual antagonism is preserved, in the latter case it disappears, but vari-
T
1 Centre of Excellence in Evolutionary Research, Department of Biological and Environmental Science, University of Jyväskylä, 40014 Jyväskylä, Finland. 2Centre of Excellence in Biological Interactions, Division of Evolution, Ecology and Genetics, Research School of Biology, Australian National University, Canberra, ACT 0200, Australia. 3Department of Biological and Environmental Science, University of Jyväskylä, 40014 Jyväskylä, Finland. 4Department of Biological and Environmental Sciences, University of Helsinki, 00014 Helsinki, Finland. 5Laboratoire d’Excellence “CORAIL,” USR 3278 CRIOBE, CNRS-EPHE, CBETM de l’Université de Perpignan, 66860 Perpignan Cedex, France.
*To whom correspondence should be addressed. E-mail:
[email protected]
972
ation as such is lost in both. Because gender load can exist even if all individuals possess the same genotype—that is, the one that is the best compromise between male and female fitness— variation is not necessarily maintained by gender load per se. Sexual antagonism, however, becomes potentially powerful for variance maintenance once it combines with sex linkage (16), maternal effects (17), or assortative mating (18). Here, we focused on the role of frequency-dependent selection as an alternative for making sexual antagonism successful at maintaining genetic variation. We tested whether frequency-dependent selection on dominance in males and/or females can maintain SA variation in field populations of a common European mammal, the bank vole (Myodes glareolus). The reproductive effort of bank voles is negatively frequency-dependent in the field (19), attesting to the direct influence of an individual’s neighbors on the population dynamics and life history evolution in this species. Previous experiments with this species have also
shown that testosterone is under sexual and SA selection and is the most important determinant of dominance in male-male competition (20–22). This hormone also affects a variety of evolutionarily important processes such as spermatogenesis (23), immune function (24, 25), and secondary sex trait growth (26). A haploid model with intralocus sexual conflict has shown that genetic variation can be maintained if the antagonism is sufficiently “balanced”—that is, if the relative fitness differences within males and females are of similar magnitude (27). This can lead to cyclic dynamics of morph frequency (28) or a protected polymorphism (29) where allele frequencies are roughly invariant over time. The maintenance of polymorphisms can be facilitated if the SA alleles are sex-linked (16), or if the fitness costs of SA alleles are nearly neutral when averaged between sexes and over their lifetimes (30). If we assume additional negative effects for each competitive type as their frequency becomes stronger, maintaining SA genetic variation becomes considerably easier. Our integrated theoretical-empirical study included a large-scale field experiment, replicated over 2 years with a total of 31 populations, that exposed bank voles to terrestrial and avian predators as well as naturally occurring weather conditions and food resources (31). We first artificially selected groups in the laboratory (lines) to create behaviorally dominant males with sisters of low fecundity, and vice versa (31). In a series of malemale competition trials, two males of opposing behavioral dominance competed with each other to mate with a wild female in estrus. The highdominance males overwhelmingly outcompeted the low-dominance males in mating success trials (c2 = 59.71, N = 168, P < 0.001) (Fig. 1A), and they had a significantly higher plasma testosterone level (generalized linear mixed model; line: F1,161 = 6.18, P = 0.014) (Fig. 1B). We then assigned males and females to large outdoor field enclosures by manipulating the reproductive quality (high or low behavioral dominance for males, high or
Fig. 1. Experimental data for males artificially selected for high and low dominance. (A) Relative mating success in laboratory male-male competition for a female in estrus. (B) Mean plasma testosterone level. (C) Mean frequency dependence of reproductive success in the field. Post hoc testing showed that the number of offspring sired by rare dominant males was significantly greater than the number sired by common dominant males (z = 2.97, P = 0.015). No other pairwise groupings differed significantly from each other (P > 0.1). Mf (black bars), high-dominance males with low-fecundity sisters; mF (white bars), lowdominance males with high-fecundity sisters; R, rare; C, common. Error bars in (B) and (C) denote SEM.
18 NOVEMBER 2011
VOL 334
SCIENCE
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
REPORTS
REPORTS B
C
1
1
1
0.9
0.9
0.9
0.8
0.8
0.8
0.7 -1
1 0.5 -0.5
0
0 0.5
D(male)
1 -1
0.7 -1
-0.5
D(female)
1 0.5 -0.5
0
0 0.5
D(female)
SCIENCE
0
0 0.5
D(male)
1 -1
-0.5
D(female)
genetic dominance in males and females; g and t determine the shape of the frequency dependence curve [see (31) for details on the model and parameter estimates]. Parameter values used: a1 = 80, a3 = 20, b1 = 4.53, b3 = 5.32, g = 5, t = 0.5, and (A) ϕ1 = ϕ2 = ϕ3 = 0, (B) ϕ1 = 0.569, ϕ3 = 0.262, (C) ϕ1 = 0.569, ϕ2 = ϕ3 = 0. The appropriate values for a2 and b2 [and ϕ2 for the scenario in (B)], corresponding to different dominance values Dmale and Dfemale, were calculated as a2 = [Dmale(a1 – a3) + a1 + a3]/2, b2 = [Dfemale(b3 – b1) + b3 + b1]/2, and ϕ2 = [Dmale(ϕ1 – ϕ3) + ϕ1 + ϕ3]/2 according to Falconer’s notation (33).
cannot be sired by another (females being a limiting resource). This can lead to the maintenance of genetic variability even if explicit frequency dependence is present in only one male type. We ran our model with explicit frequency dependence on one or both male types (in the former case, frequency dependence is implicit for subordinate males), as well as with no frequency dependence, using parameters extracted from the experimental results. As we do not know the exact genetic system controlling the antagonistic traits in the study system, we ran the model using the entire range of genetic dominance parameters from full recessiveness to full dominance. In the absence of frequency dependence, genetic variation was maintained only when genetic dominance for both female and male traits was high (in other words, A is the dominant allele for males; a is dominant for females) (Fig. 2A). Frequency dependence of the magnitude we found for voles makes the maintenance of genetic variation much easier; it is maintained for a large proportion of the parameter range describing genetic dominance (Fig. 2B). It is maintained to a very similar degree even if frequency dependence arises only implicitly for low-dominance males (Fig. 2C). Our proposed mechanism thus appears reasonably robust, and future work could fruitfully combine these findings with other mechanisms known to promote maintenance of variation under SA selection (16–18). Our results also suggest that studies of this type should not be restricted to laboratory environments. Seminatural conditions led to a sharp difference in the success of rare versus common behaviorally dominant males (Fig. 1C)—an effect that would have remained undetectable had our study been confined to the mating trials conducted in the laboratory. The difference in optimal food and housing conditions in the lab and the limitations of the field environment can have implications for physiological and behavioral measures (32) and fitness consequences alike. Thus, our theoretical and empirical results
www.sciencemag.org
1 0.5 -0.5
-0.5
D(male)
Fig. 2. The outcome of the model, expressed as the frequency of the A allele after convergence for three different frequency dependence scenarios. Allele frequencies smaller than 1 indicate that variation is maintained. (A) No explicit frequency dependence. (B) Explicit negative frequency dependence of SA alleles in all male genotypes. (C) Explicit negative frequency dependence only in the behaviorally dominant AA males. AA, Aa, and aa males have relative mating success values of a1, a2, and a3, and maximum frequency dependence values of ϕ1, ϕ2, and ϕ3, respectively; AA, Aa, and aa females have fecundity values of b1, b2, and b3. Dmale and Dfemale represent low fecundity for females) and frequency (rare or common) of male and female voles in these populations (fig. S1). The reproductive success of males became negatively frequency-dependent in the field: Dominance was costly for males when it was the common tactic in the population (zero-inflated negative binomial count submodel; frequency, Z = –2.61, P = 0.009; line, Z = –1.01, P = 0.311; frequency × line, Z = 2.00, P = 0.046) (Fig. 1C). In contrast, the fitness of sisters demonstrated a sexually antagonistic effect without evidence for frequency dependence. The high-fecundity sisters of lowdominance males had significantly larger litter sizes than the sisters of high-dominance males (i.e., females selected for low fecundity) [5.32 T 0.24 versus 4.53 T 0.27 (SEM)], although, in a clear deviation from the male pattern, the reproductive output of these females did not depend on their frequency in the population (generalized linear model quasi-Poisson; frequency, t = 0.87, P = 0.389; line, t = 2.06, P = 0.043). Combined with previous research on the selection of testosterone (20, 22, 25), our results suggest that there is potential for the sexes to experience antagonistic selection constraints in adaptation as a result of the physiological and behavioral components of their respective life history strategies. As shown in Fig. 1C, the male mating advantage is negatively frequency-dependent, which helps to explain why fixation of any allele is not predicted to occur. In the next phase of our study, we extracted parameter estimates from our field data for use in our model (31), which expands previous work (27) to include diploid genetics and frequency dependence effects within males. Mating propensities of each male type either are constant or change with frequency; the latter choice creates implicit frequency dependence for the mating success of the other male type even if the mating propensity of this type was not set to depend on frequency, because an offspring sired by one male
1 -1
0.7 -1
VOL 334
Downloaded from www.sciencemag.org on November 17, 2011
A
together show that selection favoring rare male morphs can maintain genetic variation in sexually antagonistic traits, while also indicating that ecological and social environments are important in defining the trait optima for males and females. References and Notes 1. 2. 3. 4.
5. 6. 7. 8. 9. 10. 11. 12. 13. 14.
15. 16. 17. 18. 19. 20. 21. 22. 23.
24. 25. 26. 27. 28. 29.
P. Bijma, J. Evol. Biol. 23, 194 (2010). H. Kokko, A. López-Sepulcre, Ecol. Lett. 10, 773 (2007). A. Y. K. Albert, S. P. Otto, Science 310, 119 (2005). G. A. Parker, in Sexual Selection and Reproductive Competition in Insects, M. S. Blum, N. A. Blum, Eds. (Academic Press, London, 1979), pp. 123–166. W. R. Rice, Science 256, 1436 (1992). K. Foerster et al., Nature 447, 1107 (2007). K. M. Fedorka, T. A. Mousseau, Nature 429, 65 (2004). M. Delcourt, M. W. Blows, H. D. Rundle, Proc. Biol. Sci. 276, 2009 (2009). T. Connallon, R. M. Cox, R. Calsbeek, Evolution 64, 1671 (2010). R. M. Cox, R. Calsbeek, Am. Nat. 173, 176 (2009). G. Arnqvist, L. Rowe, Sexual Conflict (Princeton Univ. Press, Princeton, NJ, 2005). M. D. Hall, S. P. Lailvaux, M. W. Blows, R. C. Brooks, Evolution 64, 1697 (2010). F. J. Ayala, C. A. Campbell, Annu. Rev. Ecol. Syst. 5, 115 (1974). S. Bedhomme, A. K. Chippindale, in Sex, Size and Gender Roles: Evolutionary Studies of Sexual Size Dimorphism, D. J. Fairbairn, W. U. Blankenhorn, T. Székely, Eds. (Oxford Univ. Press, New York, 2007), pp. 185–194. G. S. van Doorn, Ann. N.Y. Acad. Sci. 1168, 52 (2009). W. R. Rice, Evolution 38, 735 (1984). M. M. Patten, D. Haig, Biol. Lett. 5, 667 (2009). G. Arnqvist, Evolution 65, 2111 (2011). T. Mappes et al., PLoS One 3, e1687 (2008). S. C. Mills et al., Am. Nat. 173, 475 (2009). S. C. Mills, A. Grapputo, E. Koskela, T. Mappes, Proc. Biol. Sci. 274, 143 (2007). M. Mokkonen, E. Koskela, T. Mappes, S. C. Mills, J. Anim. Ecol. 10.1111/j.1365-2656.2011.01903.x (2011). T. R. Birkhead, D. J. Hosken, S. Pitnick, Eds., Sperm Biology: An Evolutionary Perspective (Academic Press, Oxford, 2009). I. Folstad, A. J. Karter, Am. Nat. 139, 603 (1992). E. Schroderus et al., Am. Nat. 176, E90 (2010). M. Zuk, T. Johnsen, T. Maclarty, Proc. Biol. Sci. 260, 205 (1995). H. Kokko, R. Brooks, Ann. Zool. Fenn. 40, 207 (2003). B. Sinervo, R. Calsbeek, Annu. Rev. Ecol. Syst. 37, 581 (2006). E. I. Svensson, J. Abbott, R. Härdling, Am. Nat. 165, 567 (2005).
18 NOVEMBER 2011
973
REPORTS University (H.K. and J.L.); the Finnish Cultural Foundation and Emil Aaltonen Foundation ( J.L.); and the Centre of Excellence in Evolutionary Research, University of Jyväskylä. We thank the staff of the Experimental Animal Unit and Konnevesi Research Station, University of Jyväskylä; R. Närä and H. Pietiläinen for logistical support; T. Laaksonen, V. Lummaa, and three anonymous reviewers for comments; and C. Soulsbury for statistical advice. The authors declare no conflicts of interest. All co-authors designed this study; M.M., E.K., T.M. and H.M. collected and analyzed the empirical data; J.L. analyzed theoretical results with input from all authors; and M.M. led the preparation of the manuscript with input from
X-ray Emission Spectroscopy Evidences a Central Carbon in the Nitrogenase Iron-Molybdenum Cofactor Kyle M. Lancaster,1 Michael Roemelt,2 Patrick Ettenhuber,2 Yilin Hu,3 Markus W. Ribbe,3* Frank Neese,2,4* Uwe Bergmann,5* Serena DeBeer1,4* Nitrogenase is a complex enzyme that catalyzes the reduction of dinitrogen to ammonia. Despite insight from structural and biochemical studies, its structure and mechanism await full characterization. An iron-molybdenum cofactor (FeMoco) is thought to be the site of dinitrogen reduction, but the identity of a central atom in this cofactor remains unknown. Fe Kb x-ray emission spectroscopy (XES) of intact nitrogenase MoFe protein, isolated FeMoco, and the FeMoco-deficient ∆nifB protein indicates that among the candidate atoms oxygen, nitrogen, and carbon, it is carbon that best fits the XES data. The experimental XES is supported by computational efforts, which show that oxidation and spin states do not affect the assignment of the central atom to C4–. Identification of the central atom will drive further studies on its role in catalysis. itrogenase (N2ase), found in symbiotic and free-living diazotrophs, catalyzes the reduction of dinitrogen (N2) to ammonia (NH3) using eight electrons, eight protons, and 16 MgATPs (ATP, adenosine triphosphate) (1). Industrially, the same reaction is performed by the Haber-Bosch process that produces more than 100 million tons of NH3 each year, thereby accounting for ~1.4% of global energy consumption. Understanding how nature activates the strongest homodinuclear bond in chemistry, the triple bond of N2, is the key for the future design of molecular catalysts. The high-resolution crystal structure of N2ase determined by Einsle et al. (2) showed that the active site of the molybdenum-iron (MoFe) protein component of N2ase binds a complex cluster consisting of seven iron ions, one molybde-
N
num ion, and nine sulfides (Fig. 1A); this cluster is referred to as the iron-molybdenum cofactor (FeMoco) and is thought to be the site of dinitrogen activation. For each FeMoco (of which there are two in the a2b2 tetrameric MoFe protein) there is an additional cluster that consists of eight irons and seven sulfides (Fig. 1B); this
all co-authors. Authors after the first author are listed in alphabetical order. Data have been deposited in the Dryad Repository (doi:10.5061/dryad.6m0f6870).
Supporting Online Material www.sciencemag.org/cgi/content/full/334/6058/972/DC1 Materials and Methods Figs. S1 to S3 Table S1 References (34–41) 20 May 2011; accepted 29 September 2011 10.1126/science.1208708
cluster is referred to as the P cluster. The P clusters serve as electron-transfer sites. Several reaction intermediates in nitrogenase catalysis have recently been observed (3, 4). However, despite the progress in the experimental and theoretical analysis of the FeMoco (4–7), neither the reaction that occurs at the FeMoco nor the structure of FeMoco has been fully clarified. In 2002, Einsle et al. identified a light atom in the center of FeMoco that could be attributed to a single, fully ionized C, N, or O atom (2). No consensus has since emerged concerning the nature of this key atom. Study of FeMoco by electron paramagnetic resonance and related techniques is complicated by complex spin-couplings between the open-shell ions, which are not fully understood. Mössbauer spectroscopy suffers from spectral crowding, and neither nuclear resonance vibrational spectroscopy nor extended x-ray absorption fine structure are sufficiently conclusive (8). Herein, we report iron Kb valence-to-core (V2C) x-ray emission spectroscopy (XES) of N2ase and demonstrate that these data provide a signature for the presence and identity of the central atom. Ka and Kb XES monitor the emission of photons after ionization of a metal 1s electron. The Kb1,3 emission line (~7040 to 7070 eV) corresponds to an electric dipole allowed 3p → 1s transition. To higher emission energies, Fig. 1. The FeMoco (A) and P cluster (B) of nitrogenase (adapted from the Protein Data Bank: identification number 1MIN). Orange, Fe; yellow, S; light blue, Mo; black, C4 –, N3-, or O2–; dark blue, nitrogen; gray, carbon. For clarity, the homocitrate and histidine ligands to the Mo have been omitted.
1 Department of Chemistry and Chemical Biology, Cornell University, Ithaca, NY 14853, USA. 2Institut für Physikalische und Theoretische Chemie, Universität Bonn, D-53115 Bonn, Germany. 3 Department of Molecular Biology and Biochemistry, University of California, Irvine, CA 92697, USA. 4Max-Planck-Institut für Bioanorganische Chemie, Stiftstrasse 34-36, D-45470 Mülheim an der Ruhr, Germany. 5Linac Coherent Light Source, SLAC National Accelerator Laboratory, Menlo Park, CA 94025, USA.
*To whom correspondence should be addressed: mribbe@uci. edu (M.W.R.);
[email protected] (F.N.); bergmann@ slac.stanford.edu (U.B.);
[email protected] (S.D.)
974
18 NOVEMBER 2011
VOL 334
SCIENCE
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
30. A. K. Chippindale, J. R. Gibson, W. R. Rice, Proc. Natl. Acad. Sci. U.S.A. 98, 1671 (2001). 31. See supporting material on Science Online. 32. R. M. Calisi, G. E. Bentley, Horm. Behav. 56, 1 (2009). 33. D. S. Falconer, T. F. C. MacKay, Introduction to Quantitative Genetics (Pearson Prentice Hall, London, ed. 4, 1996). Acknowledgments: Supported by Academy of Finland grants 115961, 119200, and 218107 (E.K.), 132190 (T.M.), and 103508 and 108566 (S.C.M.); the Vanamo Biological Society and Ehrnrooth Foundation (M.M.); the Australian Research Council and Australian National
Downloaded from www.sciencemag.org on November 17, 2011
REPORTS
Fig. 2. (A) Normalized V2C XES spectra of isolated FeMoco (red) and a representative fit to the data (black dashed line). (B) Comparison of the normalized V2C XES data for FeMoco (red), the MoFe protein (gray), and the ∆nifB MoFe protein (black). (Inset) V2C satellite region for Fe2O3 (red), Fe3N (blue), and MoFe protein (gray). Table 1. V2C XES fit parameters. n/a, not applicable. ∆nifB MoFe protein
Kb′′ peak 1 Kb′′ peak 2 Total Kb′′ integrated intensity Total V2C integrated intensity
FeMoco
E (eV)
Integrated intensity
E (eV)
7098.8 n/a n/a
0.30 n/a 0.30
7098.8 7100.2 n/a
7.73
valence-electron transitions into the metal 1s core hole are observed (referred to as the Kb2,5/Kb′′ or V2C region). These transitions have previously been assigned as ligand np → metal 1s (Kb2,5, ~7102 to 7112 eV) and ligand ns → metal 1s (Kb′′ or “satellite,” ~7090 to 7102 eV) transitions (9). V2C XES studies of Cr and Mn complexes have shown that the Kb′′ features provide a signature for the identity of the directly coordinating ligands, because energies of the observed features depend primarily on the ligand 2s ionization energies (10, 11). We recently developed an experimental and theoretical protocol for the analysis of V2C XES spectra and applied it to mono(12–14) and multinuclear (15) iron complexes. Of particular relevance is a study of a six-iron cluster with a central m6-C4– (15). These data show a feature at 7099 eV that is attributed to a transition originating from the m6-C4– 2s orbital. Computationally, this feature is predicted to shift to 7094 eV for a m6-N3– and 7088 eV for a m6-O2–. These trends closely parallel previous observations for infinite lattice complexes and mononuclear molecular complexes, thus highlighting the general applicability of this method (11, 12, 16). Figure 2A presents the normalized V2C XES data of the isolated FeMoco of N2ase, together with a representative fit to the data. Based on previous studies, the features observed with
Integrated intensity 0.18 1.60 1.78 10.45
MoFe protein E (eV)
Integrated intensity
7098.8 7100.5 n/a
0.30 0.58 0.88 7.61
maxima at ~7108 and ~7100 eV are assigned to ligand np and ns contributions, respectively (12). To assess the contribution of the sulfur ligands relative to the interstitial atom X, data were also obtained for the ∆nifB MoFe protein (Fig. 2B). This mutant contains only the two P clusters (17, 18). Based on their structural similarity, it can be assumed that the P cluster and the FeMoco have similar sulfur contributions to their XES spectra; this assumption is also supported computationally (see below). Data were obtained for the intact MoFe protein (containing both clusters) (Fig. 2B). The MoFe protein spectra map well onto an average of the spectra of the P cluster (represented by the ∆nifB MoFe protein) and the isolated FeMoco (fig. S1). Comparison between the data of the isolated FeMoco and that of the P clusters in the ∆nifB MoFe protein allowed us to assess the relative contributions of these clusters to the spectra. The V2C XES data of the P clusters showed only a weak satellite at 7098.8 with 0.30 T 0.03 units of integrated intensity. In contrast, isolated FeMoco exhibited a well-resolved satellite feature to higher energy (7100.2 eV) with an approximately sixfold increase in the integrated intensity of the satellite feature (1.78 T 0.18 units). To better understand the origin of these satellite features, we also compared the data for ∆nifB MoFe protein to the XES data for a [Fe4S4(SPh)4]2– model com-
www.sciencemag.org
SCIENCE
VOL 334
plex (19). Both the P clusters and the Fe4S4 cubane have very similar XES spectra (figs. S2 and S3). Thus, the weak 7098.8-eV feature must be attributed to a S 3s → Fe 1s transition. Under the plausible assumption that the S 3s contributions to the P cluster and FeMoco V2C XES spectra are similar, we can model the satellite region with two features: one fixed at 7098.8 eV (corresponding to the S 3s contributions) and a second to higher energy (7100.2 eV) with increased intensity (1.6 units) (Fig. 2A and Table 1). The higher-energy feature is attributed to the presence of the interstitial light atom. Comparison of the energy of the 7100.2-eV satellite feature in FeMoco to the O 2s → Fe 1s (~7092 eV) and N 2s → 1s (~7095 eV) transitions observed in Fe2O3 and Fe3N, respectively (Fig. 2B, inset), indicates that this feature arises from a ligand with 5- and 8-eV lower ionization potential than O or N, respectively. Therefore, this comparison argues against either N or O 2s contributions and strongly supports a C 2s → Fe 1s assignment. Fe V2C XES spectra can be predicted surprisingly well within a simple scheme based on density functional theory (12). To complement the experimental data, we performed detailed calculations on the FeMoco, the P cluster, and the [Fe4S4(SPh)4]2– model complex. The FeMoco was modeled by a structure containing 152 atoms obtained from the high-resolution crystal structure of Einsle et al. (2), which incorporates the key structural and electronic features of the system. The oxidation state of this iron-sulfur cluster in the resting state of the enzyme has not been determined unambiguously. Thus, we performed calculations for the two oxidation states that have been shown to be most likely (20–22). Although these two states differ by two units of charge, the differences in the calculated V2C XES transition energies and intensities are much smaller than the experimental resolution (fig. S6). The broken symmetry approach was used to approximate the
18 NOVEMBER 2011
975
Fig. 3. (A) Comparison of the calculated V2C XES spectra of FeMoco with an interstitial C4 – (black), N3– (blue), and O2– (red) and of the spectra of the P clusters (gray). (B) Calculated V2C XES spectra of FeMoco with an interstitial C4 – (black) and the P clusters (gray). (C) Experimental difference spectrum of FeMoco with the P clusters (gray), as well as calculated difference spectra of the P clusters with FeMoco containing interstitial C4 – (black), N3– (blue), and O2– (red).
effects of magnetic coupling of the various spin centers in the FeMoco calculations (23). However, calculations reveal that the ligand-to-metal crossover region of the predicted V2C XES spectra is largely unaffected by magnetic coupling (fig. S7). This finding is understandable considering the large linewidth of the experimental spectra and the rather subtle differences in orbital energies arising from different magnetic coupling schemes. More importantly, the predicted V2C spectra were highly sensitive to the identity of the interstitial ion. Figure 3A presents the calculated spectra of the FeMoco, assuming interstitial O2–, N3–, and C 4– ions together with the calculated spectrum for the P cluster. As expected, all four spectra exhibit a relatively strong feature at ~7099.3 eV, corresponding to transitions from S 3s orbitals to the Fe 1s orbitals. The only exception is FeMoco with a central C4– (Fig. 3B), where the maximum is slightly shifted to higher energies due to contributions from C4–-related transitions in the same region. Hence, our presented data, along with analogous calculations on [Fe4S4(SPh)4]2– (fig. S8 and S9), support the aforementioned assumption that the S peak in the V2C region appears at the same position of the spectrum for all measured species. Subtraction of the calculated P-cluster spectrum from the calculated spectrum of the three FeMoco species yields the contributions from the respective interstitial ions (Fig. 3C). Analysis of the difference spectra reveals that the interstitial
976
ions give rise to two features in the V2C spectrum associated with transitions from the ligand 2s and 2p orbitals, respectively. These features occur at 7096.1 and 7105.1 eV for N3– and at 7091.0 and 7104.0 eV for O2–. When a C4 – ion is placed in the center of the FeMoco, the two features are observed at 7100.2 and 7107.9 eV (Fig. 3B). For N3– and C4 –, the higher-energy feature is not distinguishable from the large peak at ~7107 eV that is dominated by transitions originating from the S 3p orbitals. Taken together, the experimental and theoretical results support assignment of the interstitial species as a C4 –. The calculated position of the C4 – 2s → Fe 1s peak matches the experimentally determined position at 7100.2 eV. Both N3– and O2– are unlikely, as their respective calculated spectra show strong features at 7096.1 eV (N3– 2s) and 7091.0 eV (O2– 2s). In addition, the measured spectra do not exhibit any features at lower energies than the S 3s peak, whereas such features have been observed experimentally in other N3– and O2– systems (as shown in the inset of Fig. 2B). The assignment is further supported by our previous studies that have shown that features with a calculated intensity of more than 10 to 15 units of intensity are experimentally observable (12). These studies also showed that the integrated intensities of experimental and calculated V2C agree strongly, with a 19% error for crystallographic structures. Even considering this error, the calculated low-energy features related to the N3– (31 units of intensity) and O2– (26 units of in-
18 NOVEMBER 2011
VOL 334
SCIENCE
tensity) ions considerably exceed this threshold. In addition, several other studies on O2–and N3– have shown features at the corresponding energy offsets (9–12). This finding raises interesting questions about both the role of the central atom and the possible pathways for biosynthesis of such an organometallic cluster. References and Notes 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.
Y. L. Hu, M. W. Ribbe, Acc. Chem. Res. 43, 475 (2010). O. Einsle et al., Science 297, 1696 (2002). B. M. Barney et al., Biochemistry 48, 9094 (2009). B. M. Hoffman, D. R. Dean, L. C. Seefeldt, Acc. Chem. Res. 42, 609 (2009). D. Lukoyanov et al., Inorg. Chem. 46, 11437 (2007). F. Neese, Angew. Chem. Int. Ed. 45, 196 (2005). T. V. Harris, R. K. Szilagyi, Inorg. Chem. 50, 4811 (2011). Y. M. Xiao et al., J. Am. Chem. Soc. 128, 7608 (2006). P. Glatzel, U. Bergmann, Coord. Chem. Rev. 249, 65 (2005). G. Smolentsev et al., J. Am. Chem. Soc. 131, 13161 (2009). S. G. Eeckhout et al., J. Anal. At. Spectrom. 24, 215 (2009). N. Lee, T. Petrenko, U. Bergmann, F. Neese, S. DeBeer, J. Am. Chem. Soc. 132, 9715 (2010). C. J. Pollock, S. DeBeer, J. Am. Chem. Soc. 133, 5594 (2011). K. M. Lancaster, K. D. Finkelstein, S. DeBeer, Inorg. Chem. 50, 6767 (2011). M. U. Delgado-Jaime et al., Inorg. Chem. 10.1021/ic201173j (2011). U. Bergmann, C. R. Horne, T. J. Collins, J. M. Workman, S. P. Cramer, Chem. Phys. Lett. 302, 119 (1999). B. Schmid et al., Science 296, 352 (2002). R. M. Allen, R. Chatterjee, P. W. Ludden, V. K. Shah, J. Biol. Chem. 270, 26890 (1995). B. A. Averill, T. Herskovitz, R. H. Holm, J. A. Ibers, J. Am. Chem. Soc. 95, 3523 (1973). H. I. Lee, B. J. Hales, B. M. Hoffman, J. Am. Chem. Soc. 119, 11395 (1997).
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
REPORTS
REPORTS the SFB 813; M.W.R. thanks the NIH for funding (grant R01-GM 67626). Portions of this research were carried out at the Stanford Synchrotron Radiation Lightsource (SSRL), a U.S. Department of Energy (DOE), Basic Energy Sciences user facility. The SSRL Structural Molecular Biology program is supported by DOE, Biological and Environmental Research, and NIH, National Center for Research Resources, Biomedical Technology Program.
Structural Basis of Silencing: Sir3 BAH Domain in Complex with a Nucleosome at 3.0 Å Resolution Karim-Jean Armache,1,2 Joseph D. Garlick,1,2 Daniele Canzio,3,4 Geeta J. Narlikar,3 Robert E. Kingston1,2* Gene silencing is essential for regulating cell fate in eukaryotes. Altered chromatin architectures contribute to maintaining the silenced state in a variety of species. The silent information regulator (Sir) proteins regulate mating type in Saccharomyces cerevisiae. One of these proteins, Sir3, interacts directly with the nucleosome to help generate silenced domains. We determined the crystal structure of a complex of the yeast Sir3 BAH (bromo-associated homology) domain and the nucleosome core particle at 3.0 angstrom resolution. We see multiple molecular interactions between the protein surfaces of the nucleosome and the BAH domain that explain numerous genetic mutations. These interactions are accompanied by structural rearrangements in both the nucleosome and the BAH domain. The structure explains how covalent modifications on H4K16 and H3K79 regulate formation of a silencing complex that contains the nucleosome as a central component. ukaryotic cells normally carry the complete set of genes needed to specify every cell type. Establishment of a specific cell fate requires the silencing of genes whose expression would disrupt that fate. Several diverse families of protein complexes maintain silencing; however, the mechanisms involved are similar in Saccharomyces cerevisiae and in multicellular eukaryotes (1). Regulation of mating type loci in S. cerevisiae serves as a paradigm for silencing. Yeast growing as haploids can adopt two mating types, a and a. The genes that are expressed at the MAT loci determine cell fate, whereas genes specifying the opposite fate can be found at the silent HMLa or HMRa loci (1, 2). The silent information regulator (Sir) proteins are essential for silencing of HMLa and HMRa, as well as telomeres and the ribosomal DNA (rDNA) loci (1, 2). The Sir proteins create domains of silenced chromatin. A long-standing hypothesis is that these proteins form specific repressive architectures that involve the basic unit of chromatin, the nucleosome. In support of this hypothesis, the SIR complex or Sir3 alone can compact nucleo-
E
1 Department of Molecular Biology, Massachusetts General Hospital, Boston, MA 02114, USA. 2Department of Genetics, Harvard Medical School, Boston, MA 02115, USA. 3Department of Biochemistry and Biophysics, University of California, San Francisco, San Francisco, CA 94158, USA. 4Chemistry and Chemical Biology Graduate Program, University of California, San Francisco, San Francisco, CA 94158, USA.
*To whom correspondence should be addressed. E-mail:
[email protected]
somal arrays in vitro (3–5). The involvement of nucleosomes in the mechanism of silencing was first indicated by the observation that yeast could not silence HMLa and HMRa when they contained a mutated form of histone H4 with a deletion of the N-terminal tail (6). Subsequently, specific point mutations that affected silencing were found in the N-terminal tails and in the globular portions of core histones (7–14), and deacetylation of histone H4 was identified as a hallmark of silenced regions (15). Reporter gene expression, restriction enzyme accessibility, and micrococcal nuclease susceptibility were used to show that domains of silenced chromatin created by the SIR complex are several kb in length (16–21). Several aspects of the extensive body of work on Sir3 interactions with nucleosomes are especially relevant to the structural work described here. Silencing requires deacetylation of histone H4 lysine 16 (H4K16); we describe the atomic contacts in the Sir3 binding pocket for H4K16. We also describe contacts with H3K79, whose methylation has the potential to modulate silencing. Many of the mutations in histones that affect silencing lie in the LRS (loss of rDNA silencing) (11, 12) domain of the nucleosome core, and we describe numerous contacts between that region and Sir3. Mutations that affect silencing have been found both at the N terminus and at the C-terminal part of Sir3 (22). Most of these mutations are clustered in the bromo-associated homology (BAH) domain that is found in the N terminus of Sir3 (23–26). Here, we used a muta-
www.sciencemag.org
SCIENCE
VOL 334
Supporting Online Material www.sciencemag.org/cgi/content/full/334/6058/974/DC1 SOM Text Figs. S1 to S11 Tables S1 and S2 References (24–35) 1 April 2011; accepted 8 September 2011 10.1126/science.1206445
tion in Sir3 (D205N) that confers increased binding to nucleosomes in vitro. Expression of the BAH D205N domain fused to LexA partially restores silencing of mating type loci in a sir3 null background. This domain is able, therefore, to combine with Sir2 and Sir4 to cause partial silencing when it is attached to an ectopic dimerization domain (27). We report the crystal structure of the complex of the hypermorphic D205N Sir3 BAH domain (BAHSir3) and the nucleosome core particle (NCP) at 3.0 Å resolution. Details of complex reconstitution, crystallization, data collection, and refinement can be found in the supporting online material (28). The BAH domain interacts extensively with each of the four core histones and, consequently, the solvent-accessible surface area buried between BAHSir3 and the nucleosome is large (1750 Å2, probe radius 1.4 Å). The structure shows a pseudo-two-fold symmetry, similar to that seen with the RCC1-nucleosome complex (29), in that BAHSir3 interacts in a similar manner with each of the two opposite faces of the nucleosome (Fig. 1). We observed 30 residues of BAHSir3 making contacts predominately with the core histones rather than nucleosomal DNA, suggesting that this protein-protein interface is critical to silencing. Interactions with the core histones are mediated through five regions on the surface of BAHSir3. These regions map well to contacts inferred from genetic screens (see Figs. 1D and 2B for a summary). The BAH domain interacts with the H4 tail, which becomes folded upon binding, and the regions of histones H3 and H4 that make up the LRS domain. In addition, BAHSir3 contacts histone H2B at a position adjacent to the LRS surface and the H2A/H2B acidic patch. Of the histone residues contacted by BAHSir3 only one residue (H4V21) varies between the Xenopus laevis histones used here and yeast histones (Fig. 4B and fig. S3). Both of the histone residues that can be covalently modified and participate in the regulation of silencing (7–9, 30) (H3K79 and H4K16) are ordered in the structure (Fig. 1B and below). Interactions between BAHSir3 and the nucleosome are established through flexible regions, which fold upon interaction (Fig. 2 and Fig. 1C). The structures of both the BAH domain and the NCP alone were determined previously (27, 31, 32), allowing comparison to the structure of the complex described here. One striking transition that accompanies assembly of the complex is folding and ordering of the histone H4 tail through extended interactions with loops 2 and 4 of BAHSir3 (Fig.
18 NOVEMBER 2011
Downloaded from www.sciencemag.org on November 17, 2011
21. A. E. True, M. J. Nelson, R. A. Venters, W. H. Ormejohnson, B. M. Hoffman, J. Am. Chem. Soc. 110, 1935 (1988). 22. S. J. Yoo, H. C. Angove, V. Papaefthymiou, B. K. Burgess, E. Munck, J. Am. Chem. Soc. 122, 4926 (2000). 23. L. Noodleman, J. Chem. Phys. 74, 5737 (1981). Acknowledgments: S.D. thanks Cornell Univ. for financial support and the Alfred P. Sloan Foundation for a fellowship; F.N. acknowledges financial support from the Univ. of Bonn, the Max Planck Society, and
977
REPORTS mutations that affect silencing have been found in core histones (6, 10–14, 23, 25, 33, 34), and the structure provides an atomic description for 14 of these residues (Fig. 1D; red depicts physical contacts; green, genetic contacts; yellow, overlap). Many of these mutations map to complementary electrostatic interactions in the interface between histones and BAHSir3. In several instances, mutations that increase silencing increase the attractive charge in the interface between histones and the BAH domain, emphasizing the importance of this type of interaction to the creation of a silenced chromatin state. The extensive correlation between mutations and molecular contacts indicates that the crystal structure reflects contacts important to biological function. We present the details of these contacts, and how they might
explain both the genetic analysis and the role for covalent modification of histones in silencing, by starting with the H4 tail region and then moving through the body of the nucleosome to the acidic patch in histone H2A and H2B. The demonstration that the N terminus of histone H4 is critical for silencing in yeast was one of the initial findings indicating the importance of nucleosomes in transcriptional regulation. Deletions and mutations of the N terminus of H4 (4 to 29) relieve silencing at HMLa and HMRa but do not impact growth of yeast (6). The charge of H4 residues 16 to 19 (a basic patch) was shown to be essential for silencing because mutations that sustain the positive charge maintained repression, whereas mutations to glycine or glutamine abolished repression (7–9).
Downloaded from www.sciencemag.org on November 17, 2011
1C). Residues in flexible loops 1 and 3 of BAHSir3 are completely disordered in a free BAH domain structure but become ordered and partially ordered, respectively, upon binding the core region of the nucleosome. Additionally, the N terminus of BAHSir3, which is in the vicinity of nucleosomal DNA (Fig. 1C), changes conformation upon binding the nucleosome. We conclude that BAHSir3 forms contacts with a large area of the histone octamer and that regions of the nucleosome and BAHSir3 become ordered upon this interaction. Mutagenesis of the BAH domain of Sir3 has identified 40 amino acid residues that affect silencing (Fig. 2B) (23–26). BAHSir3 contains at least 28 residues that form interactions (less than 4.1 Å distance) with a nucleosome. Of these, 17 were identified in genetic screens. Similarly, at least 30
Fig. 1. (A) General overview of the structure. Two different views of the complex; front view and view rotated by 90° around the y axis. The structure is color coded (BAH domain is depicted in orange, H2A in yellow, H2B in light pink, H3 in blue, H4 in green, and DNA in light gray). (B) Histone H4K16 and histone H3K79. Both residues that are critical in the regulation of SIR complex mediated silencing are shown. Histone H4K16 is depicted in green, histone H3K79 in blue, and the BAH domain surface in orange. (C) Folding transitions in the complex. Both the nucleosome and the BAH domain are
978
18 NOVEMBER 2011
VOL 334
depicted in gray, and regions that get folded upon interaction are shown in red. (D) Correlation between structural and genetic contacts. Open-book view of the complex. The NCP surface is shown on the bottom and the BAH domain on top. Surfaces colored in red represent physical contacts as seen in the structure. Surfaces colored in green represent residues both in the NCP and the BAH domains where mutation has been shown to impact silencing. Yellow surfaces represent the overlay of physical (structure-derived) and genetic contacts. SCIENCE
www.sciencemag.org
REPORTS H4 N terminal tail, the majority are with K16 and H18. Acetylation of K16 could potentially disrupt most of the electrostatic contacts in this pocket (Fig. 3, D and E) and is therefore expected to decrease the affinity of Sir3 for the nucleosome, concordant with previous studies, which infer a 1000-fold impact of acetylation (41). The LRS domain in the body of the nucleosome has been shown to be critically important for Sir3-dependent silencing at telomeres and at mating type loci (11, 12). A systematic mutagenesis study demonstrated that residues 72 to 83 of histone H3 and 78 to 81 of histone H4 are important for silencing (25). The BAH domain makes extensive interactions with a surface of the nucleosome body that includes portions of histone H3, H4, and H2B and that extends from the base of the H4 tail to an H2A region (Fig. 2A). This surface is composed of helix a1 and loop L1 of histone H3, helix a2 and loop L2 of histone H4, and helices a3 and aC of histone H2B (Fig. 4A). The LRS interacting region of BAHSir3 consists of loop 3, which becomes folded in the structure, as well as strands B6 and B8 and helix A8 (Fig. 4A). There are five LRS residues (Q76, D77, F78, K79, and T80) in helix a1 and loop L1 of H3 that contact loop 3 and strands B6 and B8 of BAHSir3 (Fig. 4C). All five of these H3 residues were identified in the slr screen (25) (Fig. 4B). BAH residues contacting histone H3 are located on both the sides of loop 3 and in strands B6 and B8. Most of the residues in the BAH domain
that interact with H3 in the LRS region have been identified as regulating silencing in genetic screens (Figs. 2B and 4C). Many additional contacts are seen between BAHSir3 and other amino acids in the LRS (Fig. 4). The strong correlation between the genetics and the physical interactions support the importance of the contacts between the BAH domain and the LRS surface in generating silencing. We were interested in understanding how the structural contacts made by D205N might lead to a hypermorphic phenotype. We see a potential hydrogen bond between the H3D77 side chain carbonyl and the BAH N205 side chain amide (Fig. 4C). In wild-type (WT) BAHSir3, the interaction between D205 and D77 would be a repulsive interaction, thereby explaining why the affinity of BAHSir3 is increased by mutation to a neutral amino acid that can create hydrogen bonding in BAH D205N. Interestingly, mutations in H3D77 have also been shown to affect silencing (25). Mutations D77N and D77G would either increase binding to BAH D205 or remove repulsion, respectively, creating interactions similar to those seen in BAH D205N with the WT histone (Fig. 4C). Repulsive interactions have been proposed to limit binding affinity of WT Sir3 to the nucleosome, and this appears to be an important aspect of regulation. The BAH D205N mutation, which has increased binding affinity, causes increased telomeric silencing in some mutant backgrounds (9, 23, 25, 43, 44) but instead
Downloaded from www.sciencemag.org on November 17, 2011
The histone H4 tail becomes ordered through residue G13 due to stabilizing interactions with BAHSir3. The H4 tail region interacts with loops 2 and 4, strand B5, and helix A1 of BAHSir3 (Fig. 3A). Each of these structural features contains residues whose mutation generates a silencing phenotype (Fig. 2). Additionally, one residue in BAHSir3 located between strands B7 and B8 participates in this interaction (Fig. 3B). Binding interactions are largely electrostatic between the positively charged histone H4 tail and the negatively charged surface of the BAH domain (Fig. 3C). Sixteen residues in BAHSir3 interact with H4 tail residues 13 to 23, primarily through their side chains (Fig. 3, B and C, and Fig. 4B) (28). An essential role for H4K16 in silencing has been demonstrated by mutational analyses, by chromatin immunoprecipitation and coimmunoprecipitation studies, and by biochemical studies showing that acetylation of this residue disrupts Sir3 binding (5, 9, 35–42). A negatively charged binding pocket of BAHSir3 accommodates the side chains of H4K16 and H4H18 (Fig. 3D). Specificity for H4K16 in the unmodified state is achieved primarily by hydrogen bonding and electrostatic interactions between the e-amino group of H4K16 and several polar or negatively charged side chains of BAHSir3 (Fig. 3E). Five of the BAH residues involved in contacts with K16 and H18 were identified in genetic screens. Of the potential electrostatic contacts that the BAH domain makes with histone residues 13 to 23 of the
Fig. 2. Overview of interactions in the complex. (A) Same view as in Fig. 1A (front). BAH domain is depicted in orange, H2A in yellow, H2B in light pink, H3 in blue, H4 in green, and DNA in light gray. Secondary structure elements are also depicted here. Secondary structure was assigned using KSDSSP. (B) Primary and secondary structure of D205N BAH. Residues with black shading were mapped previously in genetic screens. Spheres above the sequence show histone interactions (within 4.1 Å) that are ordered and visible in the electron density. Colors of spheres represent which histone interacts with this residue of the BAH domain. Bars above the secondary structure indicate which histone is interacting with this particular region of the BAH domain. There are no spheres over the residues in loop1 (residues 17 to 37) because this region is poorly ordered. www.sciencemag.org
SCIENCE
VOL 334
18 NOVEMBER 2011
979
causes decreased silencing in a WT background (25), perhaps due to increased affinity impairing function. Methylation of H3K79 by Dot1p has been implicated in regulating silencing (30, 45, 46). This methylation event, which occurs in the LRS region of the body of the nucleosome, has been shown to decrease binding by Sir3 in vitro (41) and has been proposed to modulate silencing in vivo by preventing localization of Sir3 to nonsilenced regions (30). H3K79 could potentially form three hydrogen bonds with BAHSir3, one to the side chain of E84 and two to the side chain of E140. H3K79 conformation is further stabilized by van der Waals interactions with BAH W86 and H4E74 (Fig. 4C). Methylation of H3K79 would increase the cationic radius and the hydrophobicity of this residue. Progressive methylation would decrease the potential of H3K79 to form hydrogen
bonds, and trimethylation would ablate hydrogen bonding. This could potentially result in a decreased affinity of BAHSir3 for the nucleosome. It is remarkable that at least 16 H4 and H2B residues in the LRS and adjacent regions have the potential to interact with only five residues of BAHSir3. Mutation of four of these amino acids (T78, L79, N80, and K202) was shown to affect silencing in genetic screens, indicating the importance of this interface (Figs. 2B and 4D). In a manner similar to reciprocal mutations in BAH D205 and H3D77, the LRS mutations can be suppressed by a gain-of-function mutation BAH L79I, also identified in the slr screen. This mutation has the potential to increase van der Waals contacts with the BAH domain, elucidating a possible molecular mechanism for this genetic observation. The acidic surface of histones H2A and H2B is a nucleosome interaction surface for proteins
Fig. 3. Overview of H4 tail interactions. (A) General view of BAH structural elements that interact with histone H4. The BAH surface depicted here in orange interacts with the H4 tail in green. The interaction surface is in between two domains of BAH, the helical H domain and the b sheet, and loops 2 and 4 play a crucial role in this interaction. (B) Detailed view of the H4 tail interface. Same view as Fig. 3A. All H4 tail residues (13 to 23) are shown as sticks, whereas in BAH only residues that make contacts are depicted as sticks. Magenta dashes connect residues forming potential hydrogen bonds (≤3.5 Å). There are six possible hydrogen bonds in this interface; K16 forms one, H18 two, the R23 side chain two, and the L22 main chain carbonyl forms one. Other H4 residues that could participate in these polar interactions are K20 (with E182) and R23 (E178 HB and S212). G13, A15, V21, and R19 all make van der Waals interactions with numerous BAH residues (K97, F94, V62, T63, E95, L91, P179, T180, S212, and E178). Side-chain density for the majority of the H4 tail residues is visible (28), the exceptions being side chains of R17 and R19, which are apparently more flexible and display weak side-chain density. (C) Charge complementarity of the interface. Basic histone H4 tail interaction with a negatively charged BAH domain surface. APBS-calculated electrostatics (–5kT to 5kT). Red surface represents negative and blue positive charge, respectively. (D) Close-up view of H4K16 and H4H18 binding in the charged pocket. (E) K16 binding pocket in BAH. Detailed view of K16 and
980
18 NOVEMBER 2011
VOL 334
such as herpesvirus LANA (47) and RCC1 (29). A crystal-packing interaction between a basic region of histone H4 tail and the acidic patch on adjacent nucleosome is observed in the crystal lattice of the Xenopus NCP (32). The BAH domain apparently also makes contacts here, as evidenced by electron density adjacent to the acidic patch; this density is poor and not continuous, but can only be accounted for by residues 17 to 37 of BAHSir3 (Fig. 4E). This region of the BAH domain is disordered in the apo structure (27). The density could be roughly modeled to locate the positively charged region of residues 28 to 34 of BAHSir3 as being close to the acidic residues of H2A and H2B. Mutations of these residues in the BAH domain were shown to affect silencing (24, 25). It is possible that in the context of the full protein, this interaction is stabilized and important for the overall affinity of Sir3 to nucleosome.
H18 side-chain interactions. The K16 e-amino group interacts with polar or negatively charged side chains of the BAH domain (D60, Y69, E95, and S67). K16 appears to form a hydrogen bond with S67 (3.1 Å) and potentially a weak electrostatic interaction with the Y90 main chain carbonyl. Methyl groups of V62 and T63 could stabilize the alkyl chain of K16. Side-chain carbonyls of E137 and E95 and the main-chain carbonyl of P179 can form hydrogen bonds and an electrostatic interaction with the imidazole moiety of H18, respectively. H18 is additionally coordinated through van der Waals contacts. SCIENCE
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
REPORTS
REPORTS Adjacent nucleosomes in the crystal lattice are bridged by dimerization of the BAH domain (Fig. 4F and fig. S4). Interestingly, this dimer interface was also seen in the asymmetric unit of the apo BAH domain crystal lattice (27). To assess whether dimerization is solely a crystalpacking phenomenon, we used sedimentation velocity analytical ultracentrifugation to determine whether the BAH domain dimerizes in solution. Analysis of the weight-average sedimentation coefficient for the BAH domain shows the presence of a weak self-association process, with a dimerization constant of ~2 mM (fig. S5). This weak interaction is expected to be insufficient by itself to promote compacted structures, but might contribute in the context of the fulllength Sir3 protein, which has additional selfassociation interfaces, and linked nucleosomes, which would increase the effective relative con-
Fig. 4. Interactions of BAH domain with a NCP body. (A) General view of interactions in this region. Folded loop 3 and the b strands that interact with regions of histones H3, H4, and H2B are shown. (B) Sequence alignment of regions of Xenopus and yeast histones (color coded the same as structure) that interact with BAH domain. Shaded residues were described in previous genetic screens. Orange spheres above the sequence depict which residues interact with the BAH domain. The region of histone H4 that is disordered in the structure is depicted in gray. (C) Detailed interactions of BAH with H3. A magnified view of the top part of (A). Magenta dashes connect residues forming potential hydrogen bonds. Five LRS (Q76, D77, F78, K79, and T80) residues in helix a1 and loop L1 of H3 that contact the BAH domain in the structure. BAH W86 is within 4 Å of the H3Q76 carbonyl, T80 side chain, and K79 Ca. There is a potential hydrogen bond between the H3D77 side-chain carbonyl and the BAH N205 side-chain amide. K79 could potentially form three hydrogen bonds with the BAH domain, one to the side chain of E84 and two to the E140 side chain. K79 conformation is further stabilized by van der Waals interactions with BAH W86 and H4E74. The T80 side chain interacts with the main chain of L138 and S139. BAH R75 forms polar interactions, one www.sciencemag.org
SCIENCE
centration of each half of this BAH homodimer interface. The complex visualized here is anticipated to be one of the central components for establishment of the silent state of chromatin in yeast. The BAH domain of Sir3 binds to an extensive histone surface within the nucleosome, causing structural transitions in both BAHSir3 and the H4 tail of the nucleosome. The correlation between mutations that affect silencing by Sir3 and amino acids that form physical contacts between BAHSir3 and the nucleosome show that this structure is important in the generation of silencing. The importance of a broad contiguous face in the interaction is underscored by our finding that mutations initially isolated as suppressors of H4 tail mutations, such as D205N, enhance interactions in the body of the nucleosome that are physically distant from the tail interactions. Numerous previous
Downloaded from www.sciencemag.org on November 17, 2011
How might the interface between the nucleosome and BAHSir3 be integrated in a larger structure containing full-length Sir3 to compact long regions of chromatin? The Sir3 protein has features not studied here that contribute to silencing, including acetylation of the N terminus and dimerization determined by C- terminal regions (22, 48–51). In addition, interactions involving other proteins, especially Sir4, might be important, although overexpression of Sir3 alone can increase the size of the silent domain, implicating Sir3 as a fundamental architectural protein in establishing these extended domains (52, 53). To understand Sir3 oligomerization, we will need to determine structures of full-length Sir3 with nucleosome arrays. Even in light of these caveats, there are features of the crystal packing of the BAHSir3-NCP structure that suggest a possible contribution of the BAH domain to nucleosome compaction.
of which is a potential hydrogen bond with the main chain of H3 residues D77 and F78. Additionally, a hydrogen bond might also be formed between the BAH E140 side-chain carbonyl and the main-chain amide of H3 T80. (D) Detailed interactions of BAH with H4 and H2B. A magnified view of the bottom part of (A). Magenta dashes connect residues forming potential hydrogen bonds. Two residues at the tip of loop 3 (L79 and N80) interact with histones H4 and histone H2B. They make van der Waals contacts with histone H4 residues E74, H75, and K77. Additionally, the BAH N80 side chain could form hydrogen bonds with main-chain carbonyl of H4E74 and side chain of H2B R89. L79 interacts with three H2B residues in helix a3 (R89, T93, and Q92). BAH N77 and T78 main-chain carbonyls make charged interactions and a potential hydrogen bond with side chains of H2B residues R96 and Q92, respectively. The side chain of BAH N77 can additionally interact with four residues of H2B located in helix aC. (E) Interaction of the BAH domain with the acidic patch, same view as in Fig. 1 (front). A positively charged BAH patch (residues 28 to 34) is in close proximity to acidic residues E61, E64, D90, and E92 of H2A as well as residue E110 of H2B. (F) Crystal packing interaction. VOL 334
18 NOVEMBER 2011
981
studies have implicated nucleosomes as being important for regulation through either their physical location on the genome relative to regulatory sites or their covalent modification to specify docking of regulatory complexes. We extend these examples by describing an extensive interface between a regulatory factor and the core histones of the nucleosome, thereby showing how the nucleosome can be a direct component of regulation. It is instructive to note how covalent modification of histones affects formation of this complex. Both acetylation of H4K16 and methylation of H3K79 are expected to disrupt several interactions that contribute to the BAHSir3-NCP interface. Acetylation of K16 is the more important of these modifications in vivo and would disrupt a larger number of molecular interactions based on the structure. Thus, with this complex, covalent modification of histones does not create a docking interface but rather has the potential to disrupt contacts and thereby cause a substantial change in the energetics of interaction. References and Notes 1. L. N. Rusche, A. L. Kirchmaier, J. Rine, Annu. Rev. Biochem. 72, 481 (2003). 2. S. Loo, J. Rine, Annu. Rev. Cell Dev. Biol. 11, 519 (1995). 3. S. J. McBryant, C. Krause, C. L. Woodcock, J. C. Hansen, Mol. Cell. Biol. 28, 3563 (2008). 4. F. Martino et al., Mol. Cell 33, 323 (2009). 5. A. Johnson et al., Mol. Cell 35, 769 (2009). 6. P. S. Kayne et al., Cell 55, 27 (1988). 7. P. C. Megee, B. A. Morgan, B. A. Mittman, M. M. Smith, Science 247, 841 (1990). 8. E. C. Park, J. W. Szostak, Mol. Cell. Biol. 10, 4932 (1990). 9. L. M. Johnson, P. S. Kayne, E. S. Kahn, M. Grunstein, Proc. Natl. Acad. Sci. U.S.A. 87, 6286 (1990). 10. L. M. Johnson, G. Fisher-Adams, M. Grunstein, EMBO J. 11, 2201 (1992). 11. J. H. Park, M. S. Cosgrove, E. Youngman, C. Wolberger, J. D. Boeke, Nat. Genet. 32, 273 (2002).
12. J. S. Thompson, M. L. Snow, S. Giles, L. E. McPherson, M. Grunstein, Genetics 163, 447 (2003). 13. M. N. Kyriss, Y. Jin, I. J. Gallegos, J. A. Sanford, J. J. Wyrick, Mol. Cell. Biol. 30, 3503 (2010). 14. J. Dai, E. M. Hyland, A. Norris, J. D. Boeke, Genetics 186, 813 (2010). 15. M. Braunstein, R. E. Sobel, C. D. Allis, B. M. Turner, J. R. Broach, Mol. Cell. Biol. 16, 4349 (1996). 16. D. J. Mahoney, J. R. Broach, Mol. Cell. Biol. 9, 4621 (1989). 17. S. Loo, J. Rine, Science 264, 1768 (1994). 18. R. Schnell, J. Rine, Mol. Cell. Biol. 6, 494 (1986). 19. L. Sussel, D. Shore, Proc. Natl. Acad. Sci. U.S.A. 88, 7749 (1991). 20. K. A. Nasmyth, Cell 30, 567 (1982). 21. D. E. Gottschling, O. M. Aparicio, B. L. Billington, V. A. Zakian, Cell 63, 751 (1990). 22. S. Ehrentraut et al., Genes Dev. 25, 1835 (2011). 23. V. Sampath et al., Mol. Cell. Biol. 29, 2532 (2009). 24. E. M. Stone, C. Reifsnyder, M. McVey, B. Gazo, L. Pillus, Genetics 155, 509 (2000). 25. A. Norris, M. A. Bianchet, J. D. Boeke, PLoS Genet. 4, e1000301 (2008). 26. J. R. Buchberger et al., Mol. Cell. Biol. 28, 6903 (2008). 27. J. J. Connelly et al., Mol. Cell. Biol. 26, 3256 (2006). 28. Materials and methods are available as supporting material on Science Online. 29. R. D. Makde, J. R. England, H. P. Yennawar, S. Tan, Nature 467, 562 (2010). 30. F. van Leeuwen, P. R. Gafken, D. E. Gottschling, Cell 109, 745 (2002). 31. Z. Hou, J. R. Danzer, C. A. Fox, J. L. Keck, Protein Sci. 15, 1182 (2006). 32. K. Luger, A. W. Mäder, R. K. Richmond, D. F. Sargent, T. J. Richmond, Nature 389, 251 (1997). 33. Q. Yu, L. Olsen, X. Zhang, J. D. Boeke, X. Bi, Genetics 188, 291 (2011). 34. C. J. Fry, A. Norris, M. Cosgrove, J. D. Boeke, C. L. Peterson, Mol. Cell. Biol. 26, 9045 (2006). 35. O. M. Aparicio, B. L. Billington, D. E. Gottschling, Cell 66, 1279 (1991). 36. M. Oppikofer et al., EMBO J. 30, 2610 (2011). 37. G. J. Hoppe et al., Mol. Cell. Biol. 22, 4167 (2002). 38. K. Luo, M. A. Vega-Palas, M. Grunstein, Genes Dev. 16, 1528 (2002). 39. L. N. Rusché, A. L. Kirchmaier, J. Rine, Mol. Biol. Cell 13, 2207 (2002). 40. J. C. Tanny, D. S. Kirkpatrick, S. A. Gerber, S. P. Gygi, D. Moazed, Mol. Cell. Biol. 24, 6931 (2004).
Active Starvation Responses Mediate Antibiotic Tolerance in Biofilms and Nutrient-Limited Bacteria Dao Nguyen,1† Amruta Joshi-Datar,2 Francois Lepine,3 Elizabeth Bauerle,2 Oyebode Olakanmi,4 Karlyn Beer,2 Geoffrey McKay,1 Richard Siehnel,2 James Schafhauser,1 Yun Wang,5 Bradley E. Britigan,4,6* Pradeep K. Singh2 Bacteria become highly tolerant to antibiotics when nutrients are limited. The inactivity of antibiotic targets caused by starvation-induced growth arrest is thought to be a key mechanism producing tolerance. Here we show that the antibiotic tolerance of nutrient-limited and biofilm Pseudomonas aeruginosa is mediated by active responses to starvation, rather than by the passive effects of growth arrest. The protective mechanism is controlled by the starvation-signaling stringent response (SR), and our experiments link SR-mediated tolerance to reduced levels of oxidant stress in bacterial cells. Furthermore, inactivating this protective mechanism sensitized biofilms by several orders of magnitude to four different classes of antibiotics and markedly enhanced the efficacy of antibiotic treatment in experimental infections.
I 982
n the laboratory, marked antibiotic tolerance can be produced by starving bacteria for nutrients (1). Starvation also contributes to tol-
erance during infection, as nutrients become limited when they are sequestered by host defenses and consumed by proliferating bacteria (2, 3).
18 NOVEMBER 2011
VOL 334
SCIENCE
41. M. Onishi, G. G. Liou, J. R. Buchberger, T. Walz, D. Moazed, Mol. Cell 28, 1015 (2007). 42. A. Hecht, T. Laroche, S. Strahl-Bolsinger, S. M. Gasser, M. Grunstein, Cell 80, 583 (1995). 43. Y. Park, J. Hanish, A. J. Lustig, Genetics 150, 977 (1998). 44. C. Liu, A. J. Lustig, Genetics 143, 81 (1996). 45. M. S. Singer et al., Genetics 150, 613 (1998). 46. H. H. Ng, D. N. Ciccone, K. B. Morshead, M. A. Oettinger, K. Struhl, Proc. Natl. Acad. Sci. U.S.A. 100, 1820 (2003). 47. A. J. Barbera et al., Science 311, 856 (2006). 48. H. Liaw, A. J. Lustig, Mol. Cell. Biol. 26, 7616 (2006). 49. G. G. Liou, J. C. Tanny, R. G. Kruger, T. Walz, D. Moazed, Cell 121, 515 (2005). 50. D. A. King et al., J. Biol. Chem. 281, 20107 (2006). 51. S. J. McBryant, C. Krause, J. C. Hansen, Biochemistry 45, 15941 (2006). 52. H. Renauld et al., Genes Dev. 7, (7A), 1133 (1993). 53. A. Hecht, S. Strahl-Bolsinger, M. Grunstein, Nature 383, 92 (1996). Acknowledgments: This work was supported by grant GM043901 from NIH (to R.E.K.). K-J.A. was supported in part by a fellowship from the Human Frontier Science Program. We thank the staff at Beamlines 24-IDC/E at Argonne National Laboratory, especially K. Rajashankar and F. Murphy, for excellent assistance with data collection. We thank R. Sternglanz for BAH domain constructs. We thank T. Schwartz for use of the high-throughput crystallization facility, as well as helpful discussions and critical reading of the manuscript. We thank F. Winston for critical reading of the manuscript. We thank S. Jenni and D. Kostrewa for helpful discussions; S. Tan and K. Luger for help with technical aspects of forming nucleosomes; J. Cochrane, S. Bowman, S. Miller, M. Simon, and K. Bouazoune for critical reading of the manuscript, and members of the Kingston laboratory for helpful discussions. Coordinates and structure factors have been deposited in the Protein Data Bank with accession code 3TU4.
Supporting Online Material www.sciencemag.org/cgi/content/full/334/6058/977/DC1 Materials and Methods SOM Text Figs. S1 to S5 Table S1 References (54–68) 8 July 2011; accepted 29 September 2011 10.1126/science.1210915
One of the most important causes of starvationinduced tolerance in vivo is biofilm growth, which occurs in many chronic infections (4–6). Starvation in biofilms is due to nutrient consumption by cells located on the periphery of biofilm clusters and by reduced diffusion of substrates through the biofilm (7). Biofilm bacteria show extreme tolerance to almost all antibiotic classes, and supplying limiting substrates can restore sensitivity (8). 1 Departments of Medicine, Microbiology and Immunology, McGill University, 1650 Cedar Avenue, L11.513, Montreal, Quebec H3G 1A4, Canada. 2Departments of Medicine and Microbiology, University of Washington School of Medicine, 1959 Northeast Pacific Street, Seattle, WA 98195–7242, USA. 3 Department of Microbiology, INRS Armand Frappier, 531 Boulevarde des Prairies, Laval, Quebec H7V 1B7, Canada. 4Department of Internal Medicine, University of Cincinnati, Medical Sciences Building 6065, Post Office Box 670557, Cincinnati, OH 45267, USA. 5Department of Civil and Environmental Engineering, Northwestern University, A222 Technological Institute, 2145 Sheridan Road, Evanston, IL 60208, USA. 6Veterans Administration Medical Center–Cincinnati, 3200 Vine Street, Cincinnati, OH 45220, USA.
*Present address: College of Medicine, University of Nebraska, 42nd and Emile, Omaha, NE 68198, USA. †To whom correspondence should be addressed. E-mail:
[email protected]
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
REPORTS
REPORTS To investigate the relative contributions of growth arrest and starvation physiology to tolerance, we sought experimental conditions in which nutrient-limited cells could be studied in the presence and absence of starvation responses. Many bacterial species sense and respond to nutrient limitation using a regulatory mechanism known as the stringent response (SR). Carbon, amino acid, and iron starvation activate the SR by inducing the relA and spoT gene products to synthesize the alarmone (p)ppGpp. This signal regulates the expression of many genes and is also involved in virulence (10–12). We inactivated the SR by disrupting relA and spoT in Pseudomonas aeruginosa, which causes lethal acute and chronic infections and is a model organism for studying biofilms. SR inactivation eliminated (p)ppGpp production stimulated by the starvation-inducing serine analog, serine hydroxamate (SHX) (Fig. 1A) (13). Note that SHXinduced starvation produced a nearly identical pattern of growth arrest in the wild type and DrelA spoT mutant (Fig. 1B). This allowed us to compare antibiotic tolerance in starvationarrested cells with and without SR-activated responses. In wild-type bacteria, serine starvation reduced the number of bacteria killed by
ofloxacin; the difference was ~2300-fold (Fig. 1C). In contrast, serine starvation reduced killing by only ~34-fold in the DrelA spoT mutant (Fig. 1C), despite the fact that growth was arrested in both strains (Fig. 1B). SHX treatment may not replicate typical starvation physiology, thus we studied stationary-phase cultures and biofilms where nutrient limitation occurs spontaneously (7). Whereas stationaryphase growth of wild-type P. aeruginosa produced ~106 ofloxacin-tolerant bacteria, the DrelA spoT mutant produced <104 (Fig. 1D). In biofilms, inactivation of the SR reduced the number of ofloxacin-tolerant cells by a factor of 103 (Fig. 1E). The susceptibility of the mutant in stationary phase and biofilms was restored by complementation with wild-type copies of relA and spoT (Fig. 1, D and E). A possible explanation for the marked tolerance of wild-type biofilms was that the SR restrained growth and the activity of antibiotic targets under the conditions we tested. However, growth curves of stationary-phase cultures and biofilms revealed that both the wild-type and DrelA spoT mutant strains had ceased growing before antibiotics were added (fig. S1). We also directly measured the activity of functions targeted
Downloaded from www.sciencemag.org on November 17, 2011
How does starvation produce such pronounced antibiotic tolerance? A leading hypothesis implicates the inactivity of antibiotic targets in growtharrested cells as a central mechanism (9). Target inactivity could block antibiotic action because bactericidal agents subvert their targets to produce toxic products. Thus, if targets are inactive, quinolones will likely generate fewer DNA breaks, aminoglycosides will produce less protein mistranslation, and b-lactams will cause lower levels of peptidoglycan accumulation that trigger cell lysis. However, growth arrest during starvation occurs in the context of pervasive physiological changes induced by starvation responses. This fact raises the possibilities that tolerance depends on these adaptive responses and that growth arrest and target inactivity per se are not sufficient. Identifying tolerance mechanisms is important to devising new therapeutic strategies. For example, if tolerance is inseparably linked to target inactivity, sensitizing cells could require stimulating bacterial growth, a worrisome approach during infection. Alternatively, if physiological adaptations are critical, disrupting starvation response mechanisms could enhance bacterial killing.
*** Fig. 1. SR inactivation impairs starvation-induced, stationary-phase, and biofilm antibiotic tolerance. (A) Detection of (p)ppGpp by thin-layer chromatography. The E. coli relA+ strain expresses an inducible relA. (B) Growth curves of wildtype and DrelA spoT strains, with and without SHX treatment. OD, optical density (absorbance) at 600 nm. (C) Ofloxacin tolerance of log-phase bacteria after SHX-induced starvation. CFU, colony-forming units. Error bars, SD. **P ≤ 0.001 versus wild type. (D) Ofloxacin tolerance of stationary-phase www.sciencemag.org
SCIENCE
wild-type, DrelA spoT, and DrelA spoT +SR strains. Error bars, SD. *P ≤ 0.05 or **P ≤ 0.001 versus wild type. (E) Antibiotic killing of biofilms treated with ofloxacin (30 mg/ml), meropenem (300 mg/ml), colistin (300 mg/ml), and gentamicin (50 mg/ml). Error bars, SD. *P ≤ 0.05 or ***P ≤ 0.0005 versus wild type. (F) Rates of protein and DNA synthesis in biofilms measured by [35S]methionine and [3H]adenine incorporation. Error bars, SD. *P ≤ 0.05 versus wild type. VOL 334
18 NOVEMBER 2011
983
REPORTS growth arrest per se are not responsible for the tolerance of stationary-phase and biofilm bacteria, and that active SR-mediated responses are required. We decided to focus subsequent work on biofilms because their extreme antibiotic tolerance contributes to the persistence of chronic
Fig. 2. HAQs mediate antibiotic susceptibility in the DrelA spoT mutant. (A) Endogenous levels of hydroxyl radicals (OH•) inbiofilms. OH• was measured using the probe HPF (3′-p-hydroxyphenyl fluorescein). Error bars, SD. **P ≤ 0.005 versus wild type. (B) Autolysis occurs in the DrelA spoT mutant after prolonged growth on agar (arrow). Scale bar, 2.5 mm. (C) Spontaneous cell death in DrelA spoT biofilms detected by viability staining (live cells are green and dead cells red). Images were acquired with the same microscope settings. (D) HAQ measurements by LC-MS. Error bars, SD. *P ≤ 0.01 versus wild type. (E) Antibiotic killing of biofilms treated with ofloxacin (30 mg/ml), meropenem
infections (4). The sensitizing effect of SR inactivation was seen with extended treatment times (fig. S3) and in biofilms grown for longer periods (fig. S4). Although SR inactivation sensitized biofilms grown in microtiter wells (fig. S5) and on filters on agar plates (Fig. 1E), we did not see an effect in a reactor system in which medium
(300 mg/ml), colistin (300 mg/ml), and gentamicin (50 mg/ml). CFU, colonyforming units. Error bars, SD. *P ≤ 0.05 or ***P ≤ 0.0005 versus wild type. (F) Relation between HAQ levels, ofloxacin tolerance, and [OH•] in wild-type and DrelA spoT biofilms. Strains producing graded HAQ expression in the wild type include (a) DpqsA control, (b) wild-type control, and (c) DpqsA pqsA-E+. Strains producing graded HAQ expression in DrelA spoT include (d) DrelA spoT pqsA control, (e) DrelA spoT pqsA pqsA-C+, (f) DrelA spoT control, and (g) DrelA spoT pqsA pqsA-E+. Error bars, SD. Biofilm killing **P ≤ 0.001 versus DpqsA control; OH• levels *P ≤ 0.05 versus DpqsA control.
Fig. 3. SR inactivation impairs oxidative defenses. (A, B, and C) SOD and catalase activity in biofilms as measured by native protein activity gel staining (A) and biochemical assays (B and C). Error bars, SD. An image of intact gels from (A) is shown in fig. S11. *P ≤ 0.05 or **P ≤ 0.001 versus wild type. (D) Biofilms lacking HAQs show similar ofloxacin tolerance with or without an intact SR. Error bars, SD. (E) Antibiotic tolerance in E. coli biofilms treated with ofloxacin (30 mg/ml) and tobramycin (50 mg/ml). Error bars, SD. **P ≤ 0.005 versus wild type. CFU, colony-forming units.
984
18 NOVEMBER 2011
VOL 334
SCIENCE
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
by several antibiotics at the time of drug treatment. Despite being more sensitive to killing, biofilms formed by the DrelA spoT mutant showed similar rates of protein and RNA synthesis (Fig. 1F and fig. S2) and lower rates of DNA synthesis compared with the wild-type strain (Fig. 1F). These data indicate that reduced drug target activity or
REPORTS tolysis phenotype to the overproduction of 4hydroxy-2-alkylquinoline molecules (HAQs) by P. aeruginosa (17). HAQs function in intercellular signaling and iron chelation (18–20). HAQs also have prooxidant effects, and overexpressing HAQs in wild-type P. aeruginosa modestly increased susceptibility to antibiotics [~25% more killing by ciprofloxacin (21)]. Liquid chromatography–mass spectrometry (LC-MS) analysis confirmed that the DrelA spoT mutant produced higher levels of HAQs than the wild-type strain (Fig. 2D). Of note, the DrelA spoT mutant was deficient in production of prooxidant phenazines (22) (fig. S9), which made it unlikely that these molecules caused oxidative stress in the DrelA spoT mutant. To investigate whether HAQ overproduction mediated the antibiotic sensitivity of DrelA spoT mutant biofilms, we inactivated pqsA and thus eliminated HAQ biosynthesis in this strain (Fig. 2D). Notably, wild-type levels of tolerance to ofloxacin, colistin, gentamicin, and meropenem were restored (Fig. 2E). Disrupting pqsA in the DrelA spoT mutant also abolished autolysis of colonies (Fig. 2B) and restored wild-type OH• levels in biofilms (Fig. 2A). We used gene expression constructs that generated varying amounts of
Fig. 4. SR inactivation improves antibiotic efficacy in murine infections and blocks the emergence of resistant mutants. (A) Ofloxacin treatment is more effective against lethal infections produced by the DrelA spoT strain than in infections caused by wild-type or DrelA spoT pqsA P. aeruginosa. Graphs represent pooled data from three independent experiments, with at least 15 mice per group. **P ≤ 0.005 versus treated wild-type infections. (B) Ofloxacin treatment is more effective in subcutaneous biofilm infections if the SR is inactivated. CFU, colony-forming units. Graphs represent pooled data from two independent experiments, with at least six mice per group. Error bars, SEM. **P ≤ 0.001 versus treated wild-type infections. (C) Resistant mutants emerge after prolonged exposure to ofloxacin in the wild type but not the DrelA spoT strain. **P ≤ 0.005 versus wild type. www.sciencemag.org
SCIENCE
VOL 334
HAQs (23) to determine whether a dose-response relation existed between HAQs and antibiotic susceptibility. As shown in Fig. 2F, modest increases in HAQ levels substantially enhanced antibiotic sensitivity in DrelA spoT biofilms. HAQ expression also increased OH• levels in DrelA spoT biofilms (Fig. 2F). The SR has pleiotropic effects on bacterial physiology. Thus, we considered the possibility that antibiotic sensitivity depends on other defects produced by SR inactivation, in addition to elevating HAQs. To test this, we expressed the HAQ gene constructs described above in wildtype P. aeruginosa. In contrast to the sensitivity produced in the DrelA spoT mutant, progressive increases in HAQ levels had minimal effects on antibiotic susceptibility in bacteria with an intact SR, even though higher HAQ levels were achieved (Fig. 2F and fig. S10). Expressing HAQs in wild-type biofilms also failed to increase OH• levels (Fig. 2F) The different responses of wild-type and DrelA spoT biofilms to high HAQ levels led us to hypothesize that the mutant had impaired antioxidant defenses, as this defect could sensitize cells to the prooxidant effect of HAQs. We measured catalase and superoxide dismutase (SOD) activity in biofilms and found that SR inactivation significantly decreased both (Fig. 3, A to C, and fig. S11). SOD and catalase levels were also low in the DrelA spoT pqsA triple mutant (Fig. 3, B and C), thus impaired oxidant defenses were independent of HAQ overproduction. These findings suggest that both impaired antioxidant defenses and HAQ overproduction are required for antibiotic sensitivity. To test this idea further, we compared the antibiotic susceptibility of DpqsA and DrelA spoT pqsA mutant biofilms and found no difference (Fig. 3D). This comparison was informative as neither strain expressed HAQs, but DpqsA biofilms produce SOD and catalase at near wild-type levels (fig. S12), whereas SOD and catalase are low in DrelA spoT pqsA biofilms (Fig. 3, B and C). These data show that isolated increases in HAQ levels or decreases in SOD and/or catalase activity fail to change antibiotic susceptibility in the biofilm conditions we tested. Taken together, the data are consistent with a model in which the SR mediates the antibiotic tolerance of P. aeruginosa biofilms by both curtailing HAQ production and inducing antioxidant defenses (fig. S13). Although the SR is conserved in almost all Gram-positive and Gram-negative bacteria, HAQ biosynthetic genes are not. This led us to investigate whether the SR mediated tolerance in species that do not produce HAQs. Inactivation of relA spoT in Escherichia coli decreased the number of antibiotic-tolerant bacteria by over 65-fold (Fig. 3E). The E. coli DrelA spoT mutant biofilms also had reduced catalase and elevated OH• levels (fig. S14). These results show that the SR mediates biofilm tolerance in another Gram-negative pathogen, in addition to P. aeruginosa, and raises
18 NOVEMBER 2011
Downloaded from www.sciencemag.org on November 17, 2011
flowed continuously (fig. S6). We also measured sensitivity to antibiotics with four different mechanisms of action and found that SR inactivation increased the number of bacteria killed by a factor of 102 to 105 in both the laboratory strain and clinical isolates (Fig. 1E and fig. S7). Our finding that the SR mediated resistance to drugs that interact with different cellular targets suggested that it disrupts a killing mechanism common to diverse agents. Recent work indicates that, regardless of their primary targets, bactericidal antibiotics induce hydroxyl radical (OH•) production and kill cells by oxidative damage (14–16). This finding led us to hypothesize that SR inactivation might sensitize biofilms by increasing endogenous oxidative stress. We found that SR inactivation raised OH• levels in biofilms (Fig. 2A) and increased biofilm killing by the oxidants paraquat and phenazine methosulfate (fig. S8), which is also consistent with increased endogenous oxidant production. What could account for the increased endogenous oxidative stress in the DrelA spoT mutant? A clue about the mechanism emerged when we noted spontaneous cell death in the central areas of DrelA spoT colonies (Fig. 2B) and biofilm clusters (Fig. 2C). Previous work linked this au-
985
the possibility that the control of oxidant stress may be a common mechanism. To investigate the effect of targeting the SR to increase antibiotic activity in lethal infections, we infected mice with stationary-phase P. aeruginosa. Whereas ofloxacin failed to increase the survival of mice infected with wild-type bacteria, it was highly effective against the DrelA spoT strain (Fig. 4A). Furthermore, eliminating HAQ biosynthesis abolished the susceptibility of the mutant in vivo (Fig. 4A), as was seen in vitro (Fig. 2E). Inactivation of the SR also increased antibiotic activity in a murine bioflm model (Fig. 4B). Finally, because tolerance allows bacteria to survive sustained drug exposure, tolerant subpopulations are thought to be an important source of genetic antibiotic-resistant mutants (9, 24). As shown in Fig. 4C, SR inactivation eliminated the emergence of ofloxacin-resistant clones in conditions promoting adaptive resistance. Whether cells recognize it or not, starvation will eventually stop growth and the activity of antibiotic targets. However, the capacity to sense and respond to starvation allows bacteria to arrest growth in a regulated manner that maximizes chances for long-term survival. Our data show that interfering with this orderly process sensitizes experimentally starved, stationary-phase, and biofilm bacteria to antibiotics, without stimulating their growth. Furthermore, our experiments suggest that starvation responses protect
by curtailing the production of prooxidant metabolites and increasing antioxidant defenses. Thus, antibiotic-tolerant states may depend on physiological adaptations without direct connections to antibiotic target activity or to drug uptake, efflux, or inactivation. Identifying these adaptations, and targeting them to enhance the activity of existing drugs, is a promising approach to mitigate the public health crisis caused by the scarcity of new antibiotics. References and Notes 1. R. H. Eng, F. T. Padberg, S. M. Smith, E. N. Tan, C. E. Cherubin, Antimicrob. Agents Chemother. 35, 1824 (1991). 2. J. C. Batten, P. A. Dineen, R. M. McCune Jr., Ann. N. Y. Acad. Sci. 65, 91 (1956). 3. W. McDermott, Yale J. Biol. Med. 30, 257 (1958). 4. M. R. Parsek, P. K. Singh, Annu. Rev. Microbiol. 57, 677 (2003). 5. C. A. Fux, J. W. Costerton, P. S. Stewart, P. Stoodley, Trends Microbiol. 13, 34 (2005). 6. K. Lewis, Nat. Rev. Microbiol. 5, 48 (2007). 7. P. S. Stewart, M. J. Franklin, Nat. Rev. Microbiol. 6, 199 (2008). 8. G. Borriello, L. Richards, G. D. Ehrlich, P. S. Stewart, Antimicrob. Agents Chemother. 50, 382 (2006). 9. B. R. Levin, D. E. Rozen, Nat. Rev. Microbiol. 4, 556 (2006). 10. M. Cashel, D. R. Gentry, V. J. Hernandez, D. Vinella, in Escherichia coli and Salmonella, F. C. Neidhardt, Ed. (ASM Press, Washington, D.C., 1996), vol. 1, pp. 1458-1496. 11. M. F. Traxler et al., Mol. Microbiol. 68, 1128 (2008). 12. S. L. Vogt et al., Infect. Immun. 79, 4094 (2011). 13. Materials and methods are available as supporting material on Science Online.
H2S: A Universal Defense Against Antibiotics in Bacteria Konstantin Shatalin,1 Elena Shatalina,1 Alexander Mironov,2 Evgeny Nudler1* Many prokaryotic species generate hydrogen sulfide (H2S) in their natural environments. However, the biochemistry and physiological role of this gas in nonsulfur bacteria remain largely unknown. Here we demonstrate that inactivation of putative cystathionine b-synthase, cystathionine g-lyase, or 3-mercaptopyruvate sulfurtransferase in Bacillus anthracis, Pseudomonas aeruginosa, Staphylococcus aureus, and Escherichia coli suppresses H2S production, rendering these pathogens highly sensitive to a multitude of antibiotics. Exogenous H2S suppresses this effect. Moreover, in bacteria that normally produce H2S and nitric oxide, these two gases act synergistically to sustain growth. The mechanism of gas-mediated antibiotic resistance relies on mitigation of oxidative stress imposed by antibiotics. ntil recently H2S has been known merely as a toxic gas. It is now associated with beneficial functions in mammals from vasorelaxation, cardioprotection, and neurotransmission to anti-inflammatory action in the gastrointestinal tract (1–3). The ability of H2S to function as a signaling molecule parallels the action of another established gasotransmitter, nitric oxide
U
1 Department of Biochemistry, New York University School of Medicine, New York, NY 10016, USA. 2State Research Institute of Genetics and Selection of Industrial Microorganisms, Moscow 117545, Russia.
*To whom correspondence should be addressed E-mail:
[email protected]
986
(NO) (3–5). Like NO, H2S is produced enzymatically in various tissues (1–3). Three H2Sgenerating enzymes have been characterized in mammals: cystathionine b-synthase (CBS), cystathionine g-lyase (CSE), and 3-mercaptopyruvate sulfurtransferase (3MST). CBS and CSE produce H2S predominantly from L-cyst(e)ine (Cys). 3MST does so via the intermediate synthesis of 3-mercaptopyruvate produced by cysteine aminotranferase (CAT), which is inhibited by aspartate (Asp) competition for Cys on CAT (1) (fig. S1). In contrast to mammal-derived H2S, bacteriaderived H2S has been known for centuries but
18 NOVEMBER 2011
VOL 334
SCIENCE
14. M. A. Kohanski, D. J. Dwyer, B. Hayete, C. A. Lawrence, J. J. Collins, Cell 130, 797 (2007). 15. D. J. Dwyer, M. A. Kohanski, J. J. Collins, Curr. Opin. Microbiol. 12, 482 (2009). 16. X. Wang, X. Zhao, Antimicrob. Agents Chemother. 53, 1395 (2009). 17. D. A. D’Argenio, M. W. Calfee, P. B. Rainey, E. C. Pesci, J. Bacteriol. 184, 6481 (2002). 18. E. Déziel et al., Proc. Natl. Acad. Sci. U.S.A. 101, 1339 (2004). 19. F. Bredenbruch, R. Geffers, M. Nimtz, J. Buer, S. Häussler, Environ. Microbiol. 8, 1318 (2006). 20. S. P. Diggle et al., Chem. Biol. 14, 87 (2007). 21. S. Häussler, T. Becker, PLoS Pathog. 4, e1000166 (2008). 22. Y. Wang, D. K. Newman, Environ. Sci. Technol. 42, 2380 (2008). 23. L. A. Gallagher, S. L. McKnight, M. S. Kuznetsova, E. C. Pesci, C. Manoil, J. Bacteriol. 184, 6472 (2002). 24. D. F. Warner, V. Mizrahi, Clin. Microbiol. Rev. 19, 558 (2006). Acknowledgments: We thank M. Cashel, C. Manoil, and J. Burns for providing strains; A. Hunziker and J. Harrisson for technical assistance; and J. Penterman, J. Mougous, M. Parsek, and C. Manoil for helpful discussions. Funding was provided by the Burroughs Wellcome Fund (D.N. and P.K.S), Cystic Fibrosis Foundation (P.K.S), NIH (P.K.S), and Canadian Institutes for Health Research (D.N.).
Supporting Online Material www.sciencemag.org/cgi/content/full/334/6058/982/DC1 Materials and Methods Figs. S1 to S14 Table S1 References (25–46) 12 July 2011; accepted 27 September 2011 10.1126/science.1211037
was considered to be only a byproduct of sulfur metabolism, with no particular physiological function in nonsulfur microorganisms. Likewise, little is known about the metabolic pathways involving H2S in mesophilic bacteria. However, analysis of bacterial genomes has revealed that most, if not all, have orthologs of mammalian CBS, CSE, or 3MST (figs. S1 and S2), which suggested an important cellular function(s) that preserved these genes throughout bacterial evolution. We became interested in the role of these enzymes after establishing that endogenous NO protects certain Gram-positive bacteria against antibiotics and oxidative stress (6–8). Considering some functional similarities between mammalian gasotransmitters (1–3), we hypothesized that bacterial H2S may, similarly, be cytoprotective. To determine whether CBS, CSE, or 3MST produces H2S in bacteria, we inactivated each enzyme genetically or chemically in four clinically relevant and evolutionarily distant pathogenic species: Bacillus anthracis (Sterne), Pseudomonas aeruginosa (PA14), Staphylococcus aureus (MSSA RN4220 and MRSA MW2), and Escherichia coli (MG1655). The first three species have the CBS/CSE operon, but not 3MST, whereas E. coli carries 3MST, but not CBS/CSE. The chromosomal organization of H2S genes (fig. S3) and the strategy we used for their replacement prevented any polar effects. We monitored H2S production in wild-type (wt) and mutant cells using lead acetate [Pb(Ac)2], which reacts
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
REPORTS
REPORTS and growth curves obtained for wt and 3MST and CBS/CSE mutant E. coli, P. aerugenosa, S. aureus, and B. anthracis in the presence of several representative antibiotics confirmed the results of the screen and generalized them to both Grampositive and -negative species (Fig. 1C and figs. S6 and S7). 3MST overexpression resulted in increased resistance to spectinomycin (fig. S6A), whereas chemical inhibition of CBS, CSE, or 3MST rendered bacteria more sensitive to different antibiotics (Fig. 1C and fig. S6B). An H2S donor, NaHS, suppressed the antibiotic sensitivity of CBS-CSE– and 3MST-deficient cells (Fig. 1C and figs. S6C and S7). Taken together, these results establish that endogenously produced H2S confers multidrug resistance. H2S-mediated cytoprotection resembles that of NO, which defends certain Gram-positive bacteria against some of the same antibiotics as does H2S (8). NO-mediated protection relies, in part, on its ability to defend bacteria against oxidative stress imposed by antibiotics (6–8). To examine whether H2S acts by a similar mechanism, we performed detailed analyses of its effect on bacterial killing by the representative antibiotics, gentamicin (Gm), ampicillin (Ap), and nalidixic acid (NA) (Fig. 2). All three have been shown
Fig. 1. Endogenous H2S protects bacteria against antibiotic toxicity. (A) H2S production by B. anthracis, S. aureus, P. aeruginosa, and E. coli depends on CBS/CSE and 3MST, respectively. Lead acetate–soaked paper strips show a PbS brown or black stain as a result of reaction with H2S. Strips were affixed to the inner wall of a culture tube, above the level of the liquid culture of wt or mutant bacteria, for 18 hours. CBS/CSE and 3MST inhibitors PAG/AOAA (inh) and aspartate (Asp, 3.2 mM), respectively, were added as indicated. Numbers (%) show the relative decrease in H2S production due to chemical or genetic inhibition of CBS/CSE and 3MST. pMST indicates the E. coli strain that expresses an extra copy of the 3MST gene under a strong pLtetO-1 promoter. (B) www.sciencemag.org
SCIENCE
to exert their bactericidal effect via oxidative stress (8, 10). Indeed, pretreatment of cells with 2,2′-dipyridyl, an iron chelator that suppresses the damaging Fenton reaction (11), or the hydroxyl radical scavenger thiourea, substantially decreased the toxicity of Gm (Fig. 2A). Note that wt and H2S-deficient cells became equally resistant to Gm in the presence of dipyridyl or thiourea (Fig. 2A). Moreover, the H2S donor added together with Gm was as effective as dipyridyl or thiourea in protecting against antibiotics but failed to further protect cells that had already been pretreated with antioxidants (Fig. 2A). Thus, H2S, like NO, acts by suppressing the oxidative component of antibiotic toxicity. Consistently, H2S-generating enzymes provided protection against antibiotics only under aerobic conditions. Anaerobically grown CBS/CSE-deficient B. anthracis cells were as resistant to NA or pyocyanin as wt bacteria (Fig. 2B and fig. S8). The above results suggested that H2S bolsters the antioxidant capacity of bacterial cells. Indeed, H2S-deficient B. anthracis, E. coli, S. aureus, and P. aeruginosa displayed higher susceptibility to peroxide than their wt counterparts, whereas NaHS rendered them more resistant to the peroxide (Fig. 2C and figs. S9 to S11).
Downloaded from www.sciencemag.org on November 17, 2011
specifically with H2S to form a brown lead sulfide stain. The rate of change of staining on a Pb(Ac)2-soaked paper strip is directly proportional to the concentration of H2S (9). Deletion of CBS/CSE in B. anthracis and P. aerugenosa or 3MST in E. coli greatly decreased or eliminated PbS staining (Fig. 1A). Similar results were obtained when DL-propargylglycine (PAG), aminooxyacetate (AOAA), or Asp were used, respectively, as specific inhibitors of CSE, CBS, or 3MST (Fig. 1A). Addition of Cys markedly increased PbS staining for all wt, but not CSE-CBS– or 3MST-deficient bacteria (Fig. 1B). In addition, overexpression of the chromosomal 3MST gene from a strong pLtetO-1 promoter in E. coli resulted in increased production of H2S (Fig. 1A). We conclude that all three enzymes produce H2S endogenously from Cys during exponential growth of bacteria in rich media. To elucidate the physiological role of H2S, we first compared wt and 3MST-deficient E. coli in a phenotype microarray (PMA) (fig. S4 and table S1). Whereas these strains showed little or no growth defects (fig. S5), a large number of antibiotics, highly diverse in structure and function, preferentially suppressed growth of 3MSTdeficient cells (Table 1 and table S1). The killing
Cysteine (Cys) is a substrate for bacterial CBS/CSE and 3MST. Addition of Cys (25 mM for E. coli; 200 mM for other species) greatly stimulated H2S synthesis in wt, but not in CBS/CSE- or 3MST-deficient strains. (C) H2S suppresses antibiotic-mediated bacterial killing. Representative survival curves show the effect of CBS/CSE (B. anthracis) and 3MST (E. coli) deletions or CBS/CSE inhibition (S. aureus and P. aeruginosa) by PAG/AOAA (inh) on Gm-mediated (50 mg/ml) killing. Where indicated, NaHS (0.2 mM) was added before the antibiotic challenge (see Materials and Methods). The percentage of surviving cells was determined by counting colony-forming units (CFU) and is shown as the mean T SD from three experiments. VOL 334
18 NOVEMBER 2011
987
Formation of double-strand DNA breaks (DSBs) is the primary cause of bacterial death from peroxide (12, 13). These DSBs result from the Fenton reaction (14), which can also be triggered by antibiotics (8, 10, 15, 16). To examine whether H2S protects bacteria from the damaging Fenton reaction, we monitored chromosomal DNA integrity by pulsed-field gel electrophoresis (PFGE) (Fig. 2D). The intact E. coli chromosome does not migrate into the agarose gel but remains at the origin (17), whereas linear chromosomes containing a single DSB migrate as a 4.6-Mb species (Fig. 2D, lane 1). Absent antibiotic or H2O2, DNA isolated from wt or H2S-deficient cells was retained almost entirely at the origin (lanes 2 and 3). However, treatment of cells with a sublethal dose of H2O2 or ampicillin resulted in a greater linearization (DSBs) of the chromosome in 3MSTdeficient cells (lanes 4 and 6). Overexpression of 3MST suppressed this linearization (lane 8), as did treatment with NaHS (lanes 9 and 10). These results were corroborated by polymerase chain reaction (PCR) analysis of B. anthracis, E. coli, and P. aeruginosa genomic DSBs as a function of H2S production (fig. S12) and further supported by the ability of H2S to suppress the Gminduced SOS response (fig. S13). Taken together, these results directly implicate endogenous H2S in the mitigation of chromosomal damage inflicted by antibiotics. The antioxidant effect of endogenous H2S can also be explained, in part, by its ability to augment the activities of catalase and superoxide dismutase (SOD) (Fig. 2E). The rate of H2O2 degradation in crude extracts of wt E. coli cells was >1.5 times that of 3MST-deficient cells and was increased further in cells that overexpressed 3MST (Fig. 2E). SOD activity was also proportional to the level of 3MST expression (Fig. 2E). Thus, H2S increases bacteria resistance to oxidative stress and antibiotics by a dual mechanism (fig. S14) of suppressing the DNA-damaging Fenton reaction via Fe2+ sequestration (Fig. 2A and figs. S10 and S15) and stimulating the major antioxidant enzymes catalase and SOD (Fig. 2E). The latter is essential for long-term protection but is less important during the first moments of oxidative stress. Indeed, katE and sodA E. coli mutants are well protected by NaHS during the first minutes of H2O2 exposure but then quickly loose viability (Fig. 2F). This cytoprotective mechanism of H2S parallels that of NO (8), which suggests that bacteria that produce both gases may benefit from their synergistic action. To test this hypothesis, we examined the effect of simultaneously inhibiting H2S and NO on B. anthracis growth. We were unable to generate a strain of B. anthracis in which both bacterial nitric oxide synthase (bNOS) and CBS/CSE were genetically inactivated, which suggested that the absence of both gases is incompatible with B. anthracis survival. Indeed, B. anthracis Dnos cells containing an isopropylb-D-thiogalactopyranoside (IPTG)–inducible CBS/CSE conditional knockout could grow only
988
in the presence of IPTG (fig. S16). Notably, the amount of NO produced in H2S-deficient cells or the amount of H2S produced in NO-deficient cells was greater than that produced in wt cells (Fig. 3A), which indicated that one gas compensates for the lack of the other. Also, the activity of both CBS/CSE and bNOS was stimulated in response to antibiotics (Fig. 3A). Moreover, H2O2 and antibiotics (e.g., erythromycin) substantially induced CBS/CSE gene expression (Fig. 3B) and H2S production (Fig. 3B and fig. S17). Furthermore, chemical inhibition of CBS/CSE in bNOS-deficient cells or inhibition of bNOS in CBS/CSE-deficient cells sensitized B. anthracis to antibiotics to a much greater extent than did each mutation alone (Fig. 3C). These results in concert with our previous study (8) demonstrate the synergistic and specific protective effects of H2S and NO against antibiotics. Notably,
in contrast to bNOS, which is present in only a small number of Gram-positive species (18), H2S enzymes are essentially universal (fig. S1). Because H2S equilibrates rapidly across cell membranes, a fraction of cells that generate this gas in culture or in biofilms could, in principle, defend the entire population. Indeed, wt E. coli cells effectively protect 3MST-deficient cells from Gm toxicity in exponentially growing coculture (fig. S18). Because endogenous H2S diminishes the effectiveness of many clinically used antibiotics, the inhibition of this “gaskeeper” should be considered as an augmentation therapy against a broad range of pathogens. Bacterial CBS, CSE, and 3MST have diverged substantially from their mammalian counterparts (fig. S2), which suggest that it is possible to design specific inhibitors targeting these enzymes.
Table 1. 3MST protects E. coli against different classes of antibiotics. A representative list of chemicals from the Phenotype MicroArray that preferentially suppressed the growth of 3MST-deficient cells (DsseA). Major classes of antibiotics are indicated by type (column 4). Negative numbers indicate the relative growth inhibition of the 3MST-deficient strain compared with that of the wt strain (as provided by Biolog Inc.) (fig. S4). The minimum inhibitory concentration drop for DsseA, as determined by Biolog, for two representative antibiotics, norfloxacin and troleandomycin, is 12- and 7-fold, respectively. DsseA inhibition –265 –212 –183 –117 –295 –142 –93 –372 –262 –251 –421 –179 –520 –498 –468 –294 –235 –203 –136 –271 –352 –212 –429 –231 –102 –198 –201 –160 –159 –346 –227 –216 –186 –182 –171 –164
18 NOVEMBER 2011
Chemical
Biological effect
Type
Novobiocin Norfloxacin Nalidixic acid Oxolinic acid Acriflavine 9-Aminoacridine Proflavine Trifluoperazine Promethazine Chlorpromazine Streptomycin Apramycin Tylosin Oleandomycin Erythromycin Josamycin Spiramycin Troleandomycin Chloramphenicol Fusidic acid Rifamycin SV Trimethoprim 2,4-Diamino-6,7-diisopropylpteridine Polymyxin B Colistin Vancomycin Cefsulodin Cephalothin Cefoperazone Oxacillin Nafcillin Phenethicillin Penicillin G Cloxacillin Moxalactam Carbenicillin
DNA intercalation DNA intercalation DNA intercalation DNA intercalation DNA intercalation DNA intercalation DNA intercalation DNA intercalation DNA intercalation DNA intercalation Protein synthesis Protein synthesis Protein synthesis Protein synthesis Protein synthesis Protein synthesis Protein synthesis Protein synthesis Protein synthesis Protein synthesis RNA synthesis DNA-RNA synthesis DNA-RNA synthesis Membrane Membrane Cell wall Cell wall Cell wall Cell wall Cell wall Cell wall Cell wall Cell wall Cell wall Cell wall Cell wall
Quinolone Quinolone Quinolone Quinolone Acridine Acridine Acridine Phenothiazine Phenothiazine Phenothiazine Aminoglycoside Aminoglycoside Macrolide Macrolide Macrolide Macrolide Macrolide Macrolide Amphenicol Steroid Ansamycin Antifolate Antifolate Polymyxin Polymyxin Glycopeptide Cephalosporin Cephalosporin Cephalosporin Lactam Lactam Lactam Lactam Lactam Lactam Lactam
VOL 334
SCIENCE
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
REPORTS
Fig. 2. H2S protects against antibiotic-inflicted oxidative damage (A) H2S acts by diminishing reactive oxygen species (ROS)– mediated antibiotic toxicity. E. coli cells were pretreated with the iron chelator, 2,2′-dipyridyl (0.05 mM) or the ROS scavenger thiourea (15 mM) for 3 min, followed by treatment with Gm. Cells were grown in triplicate at 37°C with aeration using a Bioscreen C automated growth analysis system. The curves represent averaged values from three parallel experiments with a margin of error of less than 5%. (B) Endogenous H2S renders cells more resistant to NA in aerobic conditions, but fails to do so in anaerobic conditions. A paper disk saturated with 20 mg/ml NA was placed on wt or CBS/CSE-deficient B. anthracis lawns that were grown aerobically or anaerobically for the next 18 hours. Zone borders are marked with dashed lines. (C) Endogenous H2S renders bacteria resistant to hydrogen peroxide. Agar plates seeded with the indicated bacteria were incubated overnight with a filter paper disk saturated with 0.125 or 0.45 M H2O2 placed atop the bacterial lawn. CBS/CSE- or 3MST-deficient cells formed a clear 5- to 10-mm zone around the disk, whereas wt cells grew a complete lawn and so demonstrated strong H2S-dependent resistance to hydrogen peroxide. (D) Pulsed-field gel analysis of chromosomal DSBs. Lane 1: 4.6 Mb linearized E. coli chromosomes (I-SceI); lanes 2 and 3: DNA from wt and DMST cells; lanes 6 to 8: DNA from wt, 3MST-deficient, and 3MST-overproducing cells after treatment with 10 mg/ml Amp; lanes 9 and 10: DNA from NaHS-treated cells after Amp treatment; and lane 11: concatemers from 0.05 to 1.0 Mb. “% linear” indicates the relative increase in linearized chromosomal DNA. The values are the average of three independent experiments (P < 0.1). (E) Stimulating effect of H2S on H2O2 degrading activity and SOD activity in crude extracts of wt and 3MSTdeficient E. coli cells. Total H2O2 degrading activity was measured as described in (7). Catalase activity at 100% is 30 mM H2O2 min–1•mg–1. Values shown are the means T SEM from three experiments. SOD activity was measured using a tetrazolium-based assay kit. (F) Dual protective effect of H2S against oxidative stress: Catalase and SOD are required for prolonged defense against
H2O2 toxicity mediated by NaHS but not for immediate protection. Wt, katE, and sodA E. coli cells were grown in Luria-Bertani broth (LB) to absorbance (optical density) OD600 of ~1.0, treated with NaHS (200 mM) for the indicated time intervals (min), followed by the addition of H2O2 (2 mM) for 10 min. Cell survival was determined by counting CFU and is shown as the mean TSD from three independent experiments.
Fig. 3. Synergistic action of H2S and NO in B. anthracis. (A) Compensatory induction of endogenous H2S and NO. In vivo production of NO in response to deletion of CBS/CSE or cefuroxime (Cef) (20 mg/ml) challenge was detected using the Cu(II)-based NO fluorescent sensor (CuFL) (19) (left bars). Cells were grown in LB to OD600 of ~0.5 followed by addition of freshly prepared CuFL (20 mM) and Cef. Fluorescence was measured in the total culture after 18 hours of incubation using a real-time fluorometer (PerkinElmer LS-55). H2S was measured using Pb(Ac)2 as in Fig. 1A (right bars). (B) H2S induction in response to antibiotic (erythromycin) or H2O2 challenge. The plot shows b-galactosidase activity (Miller units) expressed by B. anthracis cells harboring a chromosomal transcriptional fusion of the cbs/cse promoter and leader region to a promoterless lacZ gene. Bacteria were grown in LB medium until OD600 of ~0.6 followed by the addition of 0.5 mg/ml erythromycin or 2 mM H2O2. The bottom panel shows PbS brown or black stain, which is proportional to the amount of H2S produced. B. anthracis cells were grown in
96-well plates in LB + Cys (200 mM) covered with lead acetate–soaked paper by using a Bioscreen C automated growth analysis system. (C) Representative OD growth curves of wt (black curves), CBS/CSE-deficient (red) or bNOSdeficient (green) B. anthracis (Sterne) cells. Acriflavine, PAG/AOAA, NaSH, and the NOS inhibitor N-nitro-L-arginine methyl ester (L-NAME) were added as indicated. Cells were grown in triplicate at 37°C with aeration using a Bioscreen C automated growth analysis system. The curves represent the averaged values (P < 0.05).
www.sciencemag.org
SCIENCE
VOL 334
18 NOVEMBER 2011
Downloaded from www.sciencemag.org on November 17, 2011
REPORTS
989
REPORTS 10. M. A. Kohanski, D. J. Dwyer, B. Hayete, C. A. Lawrence, J. J. Collins, Cell 130, 797 (2007). 11. N. R. Asad, A. C. Leitão, J. Bacteriol. 173, 2562 (1991). 12. O. I. Aruoma, B. Halliwell, E. Gajewski, M. Dizdaroglu, J. Biol. Chem. 264, 20509 (1989). 13. S. I. Liochev, I. Fridovich, IUBMB Life 48, 157 (1999). 14. J. A. Imlay, Annu. Rev. Microbiol. 57, 395 (2003). 15. M. A. Kohanski, D. J. Dwyer, J. J. Collins, Nat. Rev. Microbiol. 8, 423 (2010). 16. B. W. Davies et al., Mol. Cell 36, 845 (2009). 17. B. Birren, E. Lai, Pulsed-Field Gel Electrophoresis: A Practical Guide (Academic Press, New York, 1993). 18. I. Gusarov et al., J. Biol. Chem. 283, 13140 (2008). 19. M. H. Lim, D. Xu, S. J. Lippard, Nat. Chem. Biol. 2, 375 (2006). Acknowledgments: We thank E. Avetissova for technical assistance, S. Mashko for materials, and members of
Wolbachia Enhance Drosophila Stem Cell Proliferation and Target the Germline Stem Cell Niche Eva M. Fast,1 Michelle E. Toomey,1,2 Kanchana Panaram,1 Danielle Desjardins,1* Eric D. Kolaczyk,3 Horacio M. Frydman1,2† Wolbachia are widespread maternally transmitted intracellular bacteria that infect most insect species and are able to alter the reproduction of innumerous hosts. The cellular bases of these alterations remain largely unknown. Here, we report that Drosophila mauritiana infected with a native Wolbachia wMau strain produces about four times more eggs than the noninfected counterpart. Wolbachia infection leads to an increase in the mitotic activity of germline stem cells (GSCs), as well as a decrease in programmed cell death in the germarium. Our results suggest that up-regulation of GSC division is mediated by a tropism of Wolbachia for the GSC niche, the cellular microenvironment that supports GSCs.
W
olbachia are maternally transmitted intracellular bacteria infecting a large number of invertebrates such as insects and parasitic worms (1). Many invertebrates that harbor these bacteria are either the vectors (for instance, mosquitoes) or the causative agent (for example, filarial nematodes) of devastating human infectious diseases. By understanding the biology at the interface between Wolbachia and their hosts, advances in the treatment of filarial diseases and the control of disease vectors are made possible (2–7). Furthermore, Wolbachia can dramatically alter host reproduction, affecting the evolutionary history of numerous invertebrates (1). Therefore, understanding how Wolbachia affect their hosts is an important ecological, evolutionary, and human health question. To investigate the influence of Wolbachia on their hosts at the cellular level, we used the 1 Department of Biology, Boston University, Boston, MA 02215, USA. 2National Emerging Infectious Disease Laboratory, Boston University, Boston, MA 02118, USA. 3Department of Mathematics and Statistics at Boston University, Boston, MA 02215, USA.
*Present address: Medical University of South Carolina, Charleston, SC 29412, USA. †To whom correspondence should be addressed. E-mail:
[email protected]
990
Drosophila gonad, a powerful experimental system. We have previously shown that in Drosophila melanogaster, Wolbachia target the somatic stem cell niche (SSCN) (Fig. 1A), the microenvironment that supports the somatic stem cells, in the female ovary (8). Further work shows that Wolbachia also target the somatic stem cell niche in the ovary of other insects (9, 10). Here, we report two additional stem cell niches preferentially colonized (i.e., cell tropism) by Wolbachia: the female germline stem cell niche (GSCN) (Fig. 1A) and the hub, at the apical tip of the testis (discussed below). In a D. mauritiana stock infected with Wolbachia wMau, we consistently noticed an intense accumulation of bacteria in the GSCN, the structure harboring the GSCs (infection frequency = 91 T 5.7%, N = 958 germaria) (see Wolbachia, labeled green in Fig. 2, A and B, Fig. 3A, and fig. S1A). This GSCN accumulation was absent in D. melanogaster (GSCN infection frequency = 0%, N = 180 germaria) (see fig. S1, B compared to A). Electron microscopy (EM) and three-dimensional reconstruction of confocal images show that the vast majority of the cytoplasmic volume of the GSCN is occupied by Wolbachia wMau [see Fig. 1B, the Wolbachia cells (a red asterisk indicates a single bacterial cell) occupy most of the GSCN (shown in green)
18 NOVEMBER 2011
VOL 334
SCIENCE
the Nudler laboratory for valuable comments and discussion. This research was supported by the NIH Director’s Pioneer Award, Biogerontology Research Foundation, and Dynasty Foundation (E.N.). Provisional patent application has been filed: U.S. Patent No. 61/438,524 “Methods for treating infections by targeting bacterial H2S-producing enzymes.” by K.S. and E.N.
Supporting Online Material www.sciencemag.org/cgi/content/full/334/6058/986/DC1 Materials and Methods SOM Text Figs. S1 to S20 Table S1 References (20–34) 15 June 2011; accepted 14 September 2011 10.1126/science.1209855
compared with the noninfected control in fig. S1C; see also movie S1]. Because GSCN function is essential for stem cell maintenance and activity (11), we hypothesized that the high levels of infection in the niche would impair its associated stem cells to a certain degree. An easy readout of GSC activity is egg production, because every egg produced originates from the division of a stem cell associated with the GSCN (Fig. 1A′). The total number of eggs laid per Wolbachia-infected female was 3.5 times higher than that observed in noninfected flies (herein referred to as “W–”; the genetic background of the W– flies was homogenized by successive backcrossing to infected males, as shown in fig. S2). This experiment was repeated under different temperature, humidity, and age conditions [see supporting online material (SOM) methods and table S1] (12). Under these different conditions, infected flies (W+) still produced approximately fourfold more eggs than the noninfected females (Fig. 1C and table S1). Given these levels of egg production, we reasoned that W+ ovaries contain GSCs that are more active. To test this possibility, we measured the frequency of GSC division in W+ and W– flies using three different markers for three distinct phases of the cell cycle. We performed the initial assessment with the use of an antibody to phospho-histone H3, which labels cells in mitosis (Fig. 2, A and C, and fig. S3G) (12). The labeling of GSCs in W+ flies was, on average, 2.7 (T 0.23)– fold higher than in W– flies (Fig. 2E and table S2). This increase could indicate either a higher GSC division in infected germaria or an arrest during the mitotic phase of the cell cycle. We further investigated GSC proliferation using two additional markers: incorporation of the thymidine analog BrdU, an indicator of DNA synthesis during S phase (fig. S3, A, D, and G), and a particular fusome morphology characteristic of GSCs in G2 (fig. S3, B, E, and H). The fusome is a germline-specific organelle that assumes the shape of an exclamation mark (!) during G2 (13, 14). Both markers corroborated a higher GSC proliferation rate in W+ (Fig. 2E). In nine independent experiments using three different methods, stem cell division in W+ flies was, on average, doubled (2.12 T 0.66) (table S2). For
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
References and Notes 1. H. Kimura, Antioxid. Redox Signal. 12, 1111 (2010). 2. M. M. Gadalla, S. H. Snyder, J. Neurochem. 113, 14 (2010). 3. R. Wang, Antioxid. Redox Signal. 12, 1061 (2010). 4. A. K. Mustafa, M. M. Gadalla, S. H. Snyder, Sci. Signal. 2, re2 (2009). 5. B. Lima, M. T. Forrester, D. T. Hess, J. S. Stamler, Circ. Res. 106, 633 (2010). 6. I. Gusarov, E. Nudler, Proc. Natl. Acad. Sci. U.S.A. 102, 13855 (2005). 7. K. Shatalin et al., Proc. Natl. Acad. Sci. U.S.A. 105, 1009 (2008). 8. I. Gusarov, K. Shatalin, M. Starodubtseva, E. Nudler, Science 325, 1380 (2009). 9. B. A. Forbes, Bailey and Scott's Diagnostic Microbiology (Mosby, St. Louis, MO, ed. 10, 1998).
REPORTS events that modulate egg production in Drosophila, the first in the germarium (Fig. 1A, left red arrow) and the second during the onset of vitellogenesis (Fig. 1A, right red arrow) (15, 18). In the parasitic wasp Asobara tabida, removal of Wolbachia causes sterility through massive cell death in mid-oogenesis, at the previtellogenic stages (16). Therefore, we initially measured PCD at these stages. We found that the differences in PCD between W+ and W– previtellogenic egg chambers were highly variable and not significant regarding Wolbachia’s effects at this developmental point (fig. S4 and table S3) (12). Accordingly, we measured the levels of PCD in the germarium. Using two different assays— DNA fragmentation in fixed tissue [terminal de-
Fig. 1. Wolbachia target the GSCN, and infection increases egg production. (A) Drosophila ovariole with the germline shown in light blue and the somatic follicle cells in white. Egg chambers are formed in the germarium (left) and mature into the egg. The upward-pointing green arrow indicates GSC (dark blue) division, which positively affects egg production [see inset (A′): GSC divides asymmetrically, and one daughter cell exits the GSCN (green) and forms the egg’s germline (light blue)]. The downward-pointing red arrows indicate developmental points where the onset of PCD reduces egg production, either in the germarium or in previtellogenic egg chambers. The black asterisk indicates the onset of vitellogenesis. (Lower left) A magnified view of the germarium shows both the SSCN (green arrowhead) and the GSCN (yellow bracket), formed by the terminal filament (light green) and the cap cells (dark green), which contact the GSCs (blue arrowhead). (B) Electron micrographs of a GSCN (green) and the GSC (blue) in infected D. mauritiana. Most of the cytoplasm of the cap cells (GSCN) is occupied by Wolbachia wMau (red asterisk) (see also movie S1). Scale bar, 1 mm. (Inset) Magnified view of the GSCN. (C) Fold change of total amount of eggs laid per infected female (W+, green) under different conditions normalized to noninfected females (W–, yellow). Relative egg production was measured in triplicate for each condition: room temperature [(RT), 20 and 46 days (d), light green] or at 25°C (20 days, dark green). Wolbachia significantly induced fecundity gains at all conditions (Student’s t test, PRT 20 days = 6.5 × 10−4, PRT 46 days = 3.9 × 10−4 and P25°C 20 days = 1.7 × 10−2) (table S1) (12). www.sciencemag.org
SCIENCE
VOL 334
oxynucleotidyl transferase–mediated deoxyuridine triphosphate nick end labeling (TUNEL)] (Fig. 2, B and D) and visualization of dead cells via live imaging (Acridine orange) (fig. S3, C and F)— Wolbachia infection consistently decreased PCD in the germarium by approximately one-half as compared with noninfected flies (Fig. 2F and table S4) (12). Wolbachia-driven reduction of PCD in the germarium was statistically significant (Fig. 2 and table S4). Together, these results indicate that the increase in egg production in W+ D. mauritiana is due to both increased GSC mitosis and decreased PCD in the germarium. Next, we examined the mechanistic foundation for Wolbachia’s manipulation of GSC mitotic activity. Considering that GSCN regulates stem cell physiology (19), we designed an experiment to test whether levels of Wolbachia in the GSCN correlate with mitotic activity of the
Downloaded from www.sciencemag.org on November 17, 2011
all three methods, the increase in probability of GSC division in W+ was statistically significant (Fig. 2E, SOM methods, and table S2) (12). Although substantial, a twofold increment in GSC mitotic activity by itself does not suffice to explain the fourfold increase in egg production in infected flies. An additional cellular event that could alter egg production in a Wolbachia-dependent manner could be cell death in the ovary. Programmed cell death (PCD) is a known key regulator of egg production in D. melanogaster (15). Furthermore, previous studies in wasps and human neutrophils have shown that the presence of Wolbachia or Wolbachia-derived proteins, respectively, inhibits host apoptosis (16, 17). We quantified the influence of Wolbachia infection on two developmentally regulated PCD
Fig. 2. Wolbachia infection increases GSC mitotic activity and suppresses PCD in the germarium. Representative confocal images of D. mauritiana germaria infected [W+, Wolbachia shown in green (A and B)] and noninfected [W–, (C and D)] are shown. Arrowheads indicate the presence (red arrowhead) or absence (blue arrowhead) of GSC division [pH3 (phospho-histone H3), red in (A) and (C)] or PCD [TUNEL, red in (B) and (D)]. Germline is labeled with anti-Vasa (blue). Scale bars, 10 mm. (E and F). Average fold difference for each marker indicated below the graphs, normalized to W– (mean of triplicates, 15 independent experiments total). Infection significantly affects GSC mitosis (E) and PCD (F) for all markers (logistic regression, PpH3 = 5.4 × 10−3, N = 621 germaria; PBrdU = 2.0 × 10−2, N = 1061; PFusome = 4.3 × 10−3, N = 695; PTUNEL = 8.0 × 10−3, N = 802; PAcridine orange = 1.2 × 10−7, N = 754) (see also tables S2 and S4) (12). AO, Acridine orange; error bars indicate SD.
18 NOVEMBER 2011
991
Fig. 3. High levels of Wolbachia at the GSCN up-regulate GSC mitosis. (A and B) Niches (yellow brackets) from infected flies are classified as highly infected [HN, (A)] or with low infection [LN, (B)]. Fusome staining (red) shows GSCs in the HN dividing [“!” morphology in (A)]. Scale bar, 5 mm. (C) Frequency of HN (solid green bars) and LN (hatched green bars) in four independent experiments. The numbers in each category and the total number of germaria analyzed are indicated for each experiment. (D) For each germarium counted, the frequency of GSC division was determined by either fusome morphology (Exp. 1 and 2) or BrdU incorporation (Exp. 3 and 4). HN significantly increases GSC mitosis (logistic regression, P = 2.4 × 10−2). (E and F) In infected testes of D. mauritiana, Wolbachia also target the stem cell niche (hubs, yellow arrowheads) at high [HN, (E)] and low levels [LN, (F)]. (F) pH3 staining (white) labels a dividing testis stem cell adjacent to an HN niche. Scale bars, 5 mm. GSC (fig. S5). During this assay, we used only Wolbachia-infected flies. Even though in these W+ flies most of the GSCNs were highly infected (91 T 6.5%, N = 788) (Fig. 3, A and C), there is a small population of niches that have either very low levels of or no Wolbachia present. These distinct types of niches were termed “LN” (low infection in the niche) (Fig. 3B), and their infected counterparts are referred to as “HN” (high infection in the niche) (Fig. 3A; see also fig. S6, A compared to B, and fig. S7). Because these distinct populations of GSCs are present inside the same infected flies, all of the environmental and systemic factors are exactly the same. In four independent experiments, the mitotic activity of GSCs residing in LN niches was substantially lower or absent in comparison with HN niches (Fig. 3, C and D). There is a statistically significant association of GSC mitosis with the high density of Wolbachia in the niche (P = 2.4 × 10−2) (table S5) (12). This observation favors a mechanism in which Wolbachia’s infection in the niche modulates stem cell activity, although it does not rule out a contribution from systemic or stem cell–intrinsic signals (see SOM text S1 and figs. S8 and S9). We found that Wolbachia wMau also target the hub, a group of somatic cells that form the niche supporting germline and somatic stem cells of the testis (20). In males, both the targeting of the hub (64%, N = 77) (Fig. 3, E and F) as well as the up-regulation of GSC division did not occur to the same degree as in females (fig. S10 and table S6). It is possible that phenotypic consequences of niche tropism are diverse in males. Wolbachia and other maternally inherited endosymbionts can evolve drastically different germline-manipulation phenotypes between sexes (21).
992
The vast majority of insects have symbiotic associations with bacteria that are vertically transmitted through the egg cytoplasm (22). Because of maternal transmission, these host-bacteria partnerships evolve to favor the reproductive success of infected mothers (1, 23–25). In the Drosophila genus, there are several reports of Wolbachiainduced changes in fecundity, including cases of rapid evolution of both partners, changing from a parasitic to mutualistic association in 20 years (24, 26–29). There is little understanding of these dramatic and widespread interferences with host reproduction at the cellular and molecular level (30). Here, we have identified two cellular events that are manipulated by Wolbachia. The combination of Wolbachia-induced alterations of both PCD in the germarium and GSC mitosis results in higher egg production, which further promotes Wolbachia spreading through maternal transmission. These findings provide the cellular mechanisms for Wolbachia’s effects on host fecundity observed in this infected D. mauritiana strain over its noninfected counterpart (see SOM text S2). Advancing our understanding of how endosymbionts subvert the cellular processes of insects will also be relevant to the growing efforts toward controlling human infectious diseases through symbiotic bacteria (3–7, 31).
7. T. Walker et al., Nature 476, 450 (2011). 8. H. M. Frydman, J. M. Li, D. N. Robson, E. Wieschaus, Nature 441, 509 (2006). 9. T. Hosokawa, R. Koga, Y. Kikuchi, X. Y. Meng, T. Fukatsu, Proc. Natl. Acad. Sci. U.S.A. 107, 769 (2010). 10. L. Sacchi et al., Tissue Cell 42, 328 (2010). 11. T. Xie, A. C. Spradling, Science 290, 328 (2000). 12. Materials and methods are available as supporting material on Science Online. 13. M. de Cuevas, A. C. Spradling, Development 125, 2781 (1998). 14. H. Lin, L. Yue, A. C. Spradling, Development 120, 947 (1994). 15. D. Drummond-Barbosa, A. C. Spradling, Dev. Biol. 231, 265 (2001). 16. B. A. Pannebakker, B. Loppin, C. P. Elemans, L. Humblot, F. Vavre, Proc. Natl. Acad. Sci. U.S.A. 104, 213 (2007). 17. C. Bazzocchi et al., Parasite Immunol. 29, 73 (2007). 18. T. L. Pritchett, E. A. Tanner, K. McCall, Apoptosis 14, 969 (2009). 19. S. J. Morrison, A. C. Spradling, Cell 132, 598 (2008). 20. M. de Cuevas, E. L. Matunis, Development 138, 2861 (2011). 21. J. H. Werren, Proc. Natl. Acad. Sci. U.S.A. 108, 10863 (2011). 22. K. Hilgenboecker, P. Hammerstein, P. Schlattmann, A. Telschow, J. H. Werren, FEMS Microbiol. Lett. 281, 215 (2008). 23. M. Turelli, A. A. Hoffmann, Nature 353, 440 (1991). 24. A. R. Weeks, M. Turelli, W. R. Harcombe, K. T. Reynolds, A. A. Hoffmann, PLoS Biol. 5, e114 (2007). 25. A. G. Himler et al., Science 332, 254 (2011). 26. A. A. Hoffmann, M. Turelli, L. G. Harshman, Genetics 126, 933 (1990). 27. A. J. Fry, M. R. Palmer, D. M. Rand, Heredity 93, 379 (2004). 28. K. T. Reynolds, L. J. Thomson, A. A. Hoffmann, Genetics 164, 1027 (2003). 29. D. Poinsot, H. Mercot, Evolution 51, 180 (1997). 30. L. R. Serbus, C. Casper-Lindley, F. Landmann, W. Sullivan, Annu. Rev. Genet. 42, 683 (2008). 31. B. Weiss, S. Aksoy, Trends Parasitol. (2011). Acknowledgments: We thank K. McCall, G. Cooper, D. Waxman, C. Bradham, and A. Boxer for valuable suggestions in the manuscript; E. Wieschaus, the McCall Lab, T. Blute, D. Gantz, and M. Bisher for help and support with PCD and EM experiments; E. Wieschaus, T. Schüpbach, R. Lehmann, P. Lasko, D. Stern, V. Orgogozo, and M. Ramos for fly stocks and reagents; members of the Frydman Lab for assistance and suggestions during the realization of this work; J. Li and D. Robson for help with MatLab software; and A. Mahowald for sharing his unpublished results and encouraging us to analyze Wolbachia in the testis. Finally, we would like to thank the anonymous reviewers for their helpful comments. This work was supported by funds from Boston Univ. and National Institute of Allergy and Infectious Diseases (grant 1K22AI74909-01A1 to H.M.F.). The data described in this paper are available in the SOM.
Supporting Online Material References and Notes 1. J. H. Werren, L. Baldo, M. E. Clark, Nat. Rev. Microbiol. 6, 741 (2008). 2. K. M. Pfarr, A. M. Hoerauf, Mini Rev. Med. Chem. 6, 203 (2006). 3. K. Bourtzis, Adv. Exp. Med. Biol. 627, 104 (2008). 4. C. J. McMeniman et al., Science 323, 141 (2009). 5. Z. Kambris, P. E. Cook, H. K. Phuc, S. P. Sinkins, Science 326, 134 (2009). 6. G. L. Hughes, R. Koga, P. Xue, T. Fukatsu, J. L. Rasgon, PLoS Pathog. 7, e1002043 (2011).
18 NOVEMBER 2011
VOL 334
SCIENCE
www.sciencemag.org/cgi/content/full/science.1209609/DC1 Materials and Methods SOM Text Figs. S1 to S10 Tables S1 to S6 References (32–58) Movie S1 10 June 2011; accepted 7 October 2011 Published online 20 October 2011; 10.1126/science.1209609
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
REPORTS
REPORTS
Jian Xu,1,2 Cong Peng,1* Vijay G. Sankaran,1,5* Zhen Shao,1 Erica B. Esrick,1,3 Bryan G. Chong,1 Gregory C. Ippolito,4 Yuko Fujiwara,1,2 Benjamin L. Ebert,3 Philip W. Tucker,4 Stuart H. Orkin1,2† Persistence of human fetal hemoglobin (HbF, a2g2) in adults lessens the severity of sickle cell disease (SCD) and the b-thalassemias. Here, we show that the repressor BCL11A is required in vivo for silencing of g-globin expression in adult animals, yet dispensable for red cell production. BCL11A serves as a barrier to HbF reactivation by known HbF inducing agents. In a proof-of-principle test of BCL11A as a potential therapeutic target, we demonstrate that inactivation of BCL11A in SCD transgenic mice corrects the hematologic and pathologic defects associated with SCD through high-level pancellular HbF induction. Thus, interference with HbF silencing by manipulation of a single target protein is sufficient to reverse SCD.
re+
re-
EpoR-Cre-
BCL11A GAPDH
B
EpoR-Cre-
EpoR-Cre+
D
16 14
100 -/-
0.1
8 6
E18.5
E16.5
E14.5
4
0.01
1
100
2
8 10 15 20 30
6
2
Age (Weeks)
0 0
Peripheral Blood
Fetal Liver
E
10
Bcl11a
1
E12.5
Human γ-globin expression (%)
12
10
2
Mx1-CreMx1-Cre+
∗
1 Division of Hematology/Oncology, Children’s Hospital Boston and Department of Pediatric Oncology, Dana-Farber Cancer Institute, Harvard Stem Cell Institute, Harvard Medical School, Boston, MA 02115, USA. 2Howard Hughes Medical Institute, Boston, MA 02115, USA. 3Dana-Farber Cancer Institute, Brigham and Women’s Hospital, Harvard Stem Cell Institute, Harvard Medical School, Boston, MA 02115, USA. 4Institute for Cellular and Molecular Biology, The University of Texas at Austin, Austin, TX 78712, USA. 5Broad Institute and Whitehead Institute for Biomedical Research, Cambridge, MA 02142, USA.
*These authors contributed equally to this work. †To whom correspondence should be addressed. E-mail:
[email protected]
Fig. 1. BCL11A loss in adult mice reverses gsilencing. (A) Expression of BCL11A protein in CD71+Ter119+ fetal liver (FL) and bone marrow (BM) cells of control (EpoR-Cre-) and BCL11A knockout (EpoR-Cre+) b-YAC mice. Glyceraldehyde phosphate dehydrogenase (GAPDH) was analyzed as a loading control. (B) Expression of human fetal (g) globin genes was monitored by quantitative reverse transcription polymerase chain reaction (qRT-PCR) in FL cells (E12.5 to E18.5) or peripheral blood of postnatal animals (1 to 30 weeks old). Data are plotted as percentage of g-globin Hba-a1/a2 Hba-x over total b-like human globin gene levels in conHbb-y Hbb-bh1 trol (EpoR-Cre-) and BCL11A knockout (EpoR-Cre+) b-YAC mice (N ≥ 4 per genotype at each time Hbb-b1/b2 point). Results are means T SD. All g-globin levels for the different genotypes are significantly different (P < 1 × 10−5, two-tailed t test). (C) Immunohistochemistry for HbF was performed on E16.5 FLs from EpoR-Cre- and EpoR-Cre+ animals. (D) Transcriptional profiling of control (Bcl11a+/+) and Bcl11a–/– (EpoR-Cre+) CD71+Ter119+ erythroid cells (N = 3 per genotype). Probes corresponding to mouse a- and b-globin genes are indicated by ar6 8 10 12 14 16 +/+ rows. Hba-x, z-globin; Hba-a1/a2, a-globin; Hbb-y, Bcl11a ey-globin; Hbb-bh1, bh1-globin; Hbb-b1/b2, b-globin. (E) Expression of human g-globin genes was monitored by qRT-PCR in control (Mx1-Cre-) and BCL11A knockout (Mx1-Cre+) mice before and after polyinosinepolycytidine (pIpC) (N ≥ 4 per genotype at each time point; *P < 1 × 10−5).
EpoR-Cre+
Ep
oR Ep
C
oR
-C
-C
reEp
Ep
oR
oR
-C
BM
re+
FL
A
critical to the pathogenesis of sickle cell disease (SCD) and the b-thalassemias. As increased HbF lessens the severity of these conditions (1, 2), elu-
-C
T
he switch from fetal (HbF, a2g2) to adult hemoglobin (HbA, a2b2), a paradigm for transcriptional control in development, is
cidation of mechanisms to relieve HbF silencing in adult erythroid cells has been a long-sought goal. Here, we demonstrate that inactivation of one component involved in HbF regulation, the transcription factor BCL11A, provides phenotypic correction of mice that model SCD. Our findings provide a crucial proof of principle for targeted reactivation of HbF. SCD, the first “molecular disease,” is caused by substitution of valine for b-6 glutamic acid in the b-globin chain of adult hemoglobin (3). The
Downloaded from www.sciencemag.org on November 17, 2011
Correction of Sickle Cell Disease in Adult Mice by Interference with Fetal Hemoglobin Silencing
∗
∗
∗
∗
3
6
9
12
4
Human γ-globin expression (%)
10
1
0.1
0.01 Before
1
Weeks after inactivation of BCL11A www.sciencemag.org
SCIENCE
VOL 334
18 NOVEMBER 2011
993
REPORTS mutated, assembled hemoglobin, HbS (a2bS2), undergoes polymerization upon deoxygenation, resulting in erythrocyte deformation, hemolysis, and morbid complications secondary to microvascular occlusion. HbS polymerization is highly sensitive to inhibition by HbF (4). The level of HbF varies among adult individuals and is inherited as a quantitative trait. Genome-wide association studies (GWAS) recently provided critical insight into loci controlling HbF. Three loci harboring genetic variants
A
LCR
with large effects were identified, including the b-globin cluster itself, a HBS1L-MYB gene interval, and the gene encoding BCL11A (5–7). Functional studies demonstrate that BCL11A serves as transcriptional repressor of HbF expression (8–10). BCL11A controls the developmental switch from embryonic to adult b-globin in the mouse and the silencing of HbF expressed from human b-globin locus transgenes in mouse fetal liver. Moreover, BCL11A contributes to HbF silencing in cultured, primary human erythroid cells
δ
Aγ
Gγ
ε
(8, 11). While providing compelling support for BCL11A as a regulator of globin switching and HbF silencing in development, these findings do not address its in vivo role at the adult stage and potential as a therapeutic target for reactivation of HbF. Besides the GWAS-identified regions, numerous nuclear factors have been implicated in globin switching and/or HbF silencing (12). Apart from their proposed roles in globin regulation, many of these are essential for proper maturation of erythroid cells, thereby complicat-
β
B P < 0.0001
-53-50
+6 +17 +50
+210
Unmethylated CpG Methylated CpG
HBG2
+1 +/+
βYAC
E14.5 FL
3-wk BM
30-wk BM
E14.5 FL
3-wk BM
30-wk BM
-2 56 -1 62 -5 3 -5 0 +6 +1 7 +5 0
-2 5 -1 6 6 -5 2 3 -5 0 +6 +1 7 +5 0
-2 5 -1 6 6 -5 2 3 -5 0 +6 +1 7 +5 0
-2 56 -1 62 -5 3 -5 0 +6 +1 7 +5 0
-2 5 -1 6 6 -5 2 3 -5 0 +6 +1 7 +5 0
βYAC
-2 5 -1 6 6 -5 2 3 -5 0 +6 +1 7 +5 0
-/-
-2 5 -1 6 6 -5 2 3 -5 0 +6 +1 7 +5 0
Bcl11a
Human BM Pro-E
-2 56 -1 62 -5 3 -5 0 +6 +1 7 +5 0
Bcl11a
Human FL Pro-E
40 30 20 10
P = 0.0002
0.03 0.02 0.01 0
5-azaD +/+
Bcl11a
D
ChIP Fold Enrichment
6 0 30 0 12
12 0 40
96%
81%
82%
89%
40%
59%
66%
BCL11A H3K9ac HDAC1
Human γ-globin expression (%)
20%
C
30
H3K27me3 Pol II
0
5
4
3
2
ε
1
Gγ
20
10 P = 0.011
0.03 0.02 0.01 0 +/+
Bcl11a
LCR
994
P < 0.0001
SAHA
β
δ
Aγ
-/-
Bcl11a
-/-
Bcl11a
Fig. 2. DNA demethylation and HDAC inhibition enhance residual g-globin expression. (A) CpG methylation within the Gg promoter was measured by bisulfite sequencing analysis in CD71+Ter119+ erythroid progenitors in control (Bcl11a+/+) and Bcl11a–/– (EpoR-Cre+) b-YAC mice at various development stages. Human primary FL and BM proerythroblasts (Pro-E) were analyzed as controls. The percentage of methylated CpG dinucleotides is shown for each sample. A diagram of the human b-globin cluster is shown on the top. (B) Expression of human g-globin genes was monitored by qRT-PCR in control
(Bcl11a+/+) and Bcl11a–/– (EpoR-Cre+) b-YAC mice before and after 5-azaD treatment (N = 6 per genotype per treatment). (C) In vivo chromatin occupancy of BCL11A, acetyl-H3 lysine 9 (H3K9ac), HDAC1, trimethyl-H3 lysine 27 (H3K27me3), and RNA Pol II was determined by ChIP-chip in primary human erythroid progenitors. A genome browser representation of binding patterns at the human b-globin cluster is shown. (D) Expression of human g-globin genes was monitored by qRT-PCR in control (Bcl11a+/+, N = 8) and Bcl11a–/– (EpoR-Cre+, N = 12) b-YAC mice before and after SAHA treatment.
Table 1. Correction of hematologic parameters and urine concentration defect in SCD mice by inactivation of BCL11A. All animals were analyzed at 8 to 10 weeks after birth. Data are means T SEM. P values were calculated by two-tailed t test between the SCD and SCD/Bcl11a–/– animals; *P < 0.002, **P < 1 × 10−5.
N = 17, 14, and 5 for the control, SCD, and SCD/Bcl11a–/– animals, respectively. RBC, red blood cell; Hb, hemoglobin; Hct, hematocrit; MCV, mean corpuscular volume; MCH, mean corpuscular hemoglobin content; MCHC, mean corpuscular hemoglobin concentration; Retic, reticulocytes; RDW, red cell distribution width.
Mice
RBC (×106/ml)
Hb (g/dl)
Hct (%)
MCV (fl)
MCH (pg)
Control SCD SCD/Bcl11a–/–
10.1 T 0.2 6.4 T 0.5 9.8 T 0.4*
13.1 T 0.3 7.8 T 0.6 13.6 T 0.7**
44.2 T 1.0 28.3 T 1.9 46.2 T 1.4**
44.1 T 1.4 44.8 T 1.6 47.2 T 0.9
13.0 T 0.4 12.3 T 0.6 13.8 T 0.2
29.7 T 0.6 3.1 T 0.6 27.5 T 0.6 38.2 T 3.9 29.3 T 0.8 7.0 T 0.3*
SCIENCE
www.sciencemag.org
18 NOVEMBER 2011
VOL 334
MCHC (g/dl)
Retic (%)
RDW (%)
Urine concentration (mOsm)
19.0 T 0.7 26.9 T 0.5 23.4 T 0.4*
2440 T 213 1037 T 82 2133 T 333*
Downloaded from www.sciencemag.org on November 17, 2011
-162
-256
Human γ-globin expression (%)
50
REPORTS We further examined several experimental features (YAC copy number and time of Cre-mediated gene inactivation) that might contribute to the observed silencing and showed that they were noncontributory (see SOM text and fig. S5). Transcriptional profiling of BCL11A-null erythroid cells purified from adult bone marrow was used to assess the quality of erythroid maturation. BCL11A-null and wild-type erythroid cells exhibited highly similar patterns of gene expression, characterized by a Pearson correlation coefficient (r2) of 0.9736 for the log2 normalized intensities (Fig. 1D). The expression of known erythroid transcriptional regulators, including GATA1, FOG1, NF-E2, KLF1, SOX6, and MYB, was comparable between the groups. The most differentially expressed genes were mouse embryonic b-like and a-like globin genes (Fig. 1D, fig. S6, and tables S2 and S3). Thus, BCL11A is highly selective in controlling targets in erythroid cells, and only expression of the globin genes is substantially affected in its absence. These findings establish roles for BCL11A in HbF silencing but fail to demonstrate whether g-globin genes that are fully silenced during normal development can be reactivated upon loss of BCL11A. Thus, we introduced the interferoninducible Mx1-Cre allele into BCL11A floxed b-YAC mice (fig. S7). Efficient excision of floxed
Fig. 3. Inactivation of BCL11A rescues sickle cell defects in humanized SCD mice. (A) Representative blood smears of control, SCD, and SCD/Bcl11a–/– mice are shown at 1000x magnification. (B) RBC life span is significantly extended in SCD/Bcl11a–/– mice at every time point compared with SCD mice (N ≥ 4; P < 0.01). Results are means T SEM. (C) Correction of splenomegaly in SCD/Bcl11a–/– mice (N ≥ 3 per genotype). Results are means T SEM. (D) Expression of fetal (g) and sickle adult (bs) globin genes was monitored by qRT-PCR in the peripheral blood of control, SCD, and SCD/Bcl11a–/– animals (8 to 10 weeks old; N = 5, 6, and 4, respectively). Results are means T SEM. (E) Distribution of HbF in red cells. Representative graphs for control, SCD, and SCD/Bcl11a–/– animals are shown. The same scale is used in all three graphs, and the mean percentage of F cells (HbF/HbA double-positive) is shown (N = 5, 6, and 4, respectively).
A
SCD
Control
Biotinylated RBC (%)
100
SCD/ Bcl11a
C
Control SCD -/SCD/Bcl11a
80
1.4
60 40 20
P < 0.001
-/-
P < 0.001
1.2
Spleen weight (g)
B
BCL11A alleles in adult mice was not associated with significant changes in blood counts (fig. S8) except for a decline in total B cells, consistent with a role in lymphopoiesis (14). Developmentally silenced g-globins were reexpressed to 13.8% of total b-like human globins 1 week after inactivation of BCL11A and sustained thereafter (Fig. 1E). As this level of g-derepression closely approximates that in EpoR-Cre BCL11A conditional mice, the substantial component of HbF silencing dependent on BCL11A is reversible. Similarly, the mouse embryonic ey- and bh1globin genes were derepressed on BCL11A loss (fig. S9). Partial silencing of g-globin expression in BCL11A-null erythroid cells points to additional silencing pathways that act independently of BCL11A. Two epigenetic pathways, DNA methylation and histone deacetylation, have been implicated in HbF control (15–17). DNA methylation of the g-globin promoters progressively increased in BCL11A-null erythroid cells in correlation with the gradual, but partial, silencing of g-globin expression (Fig. 2A). Administration of the DNA methylation inhibitor 5-aza-2′-deoxycytidine (5-azaD) to normal b-YAC mice led to a very small increment in g-globin mRNA (Fig. 2B). In contrast, 5-azaD treatment was synergistic to BCL11A loss, leading to 37.9% g-globin mRNA.
Downloaded from www.sciencemag.org on November 17, 2011
ing their consideration as targets for therapeutic manipulation. To address in vivo roles of BCL11A, we employed stringent genetic tests in mice carrying the human b-globin gene cluster as a yeast artificial chromosome transgene (b-YAC mice). Knockout of BCL11A interrupts silencing of endogenous b-like embryonic genes and human g-globin genes in mouse fetal liver (9). Because BCL11A-null mice are postnatally lethal, we examined the contribution of BCL11A to g-silencing in adults through conditional inactivation of BCL11Awith erythroid-selective EpoR-GFP Cre alleles (13) (Fig. 1A). Mice harboring erythroid-specific inactivation of BCL11A developed normally. Erythropoiesis in fetal liver and adult bone marrow appeared normal in the absence of BCL11A (fig. S1). As in the conventional knockout, hemoglobin switching failed to occur in fetal liver, such that g constituted >80% of the b-like human globins (Fig. 1B and fig. S2). HbF was robustly expressed in definitive erythroid cells of E16.5 (embryonic day 16.5) fetal liver (Fig. 1C and fig. S3). After birth, the level of g-globin declined progressively to a residual level of ~11% in 30-week and older adults (Fig. 1B and fig. S2). Mouse embryonic b-like globin genes (ey and bh1) were also up-regulated in BCL11A-null erythroid cells throughout development (fig. S4).
1.0 0.8 0.6 0.4 0.2
0 0
3
6
9
12
15
0
Post-biotin interval (d)
E
40
Control
10 4
P = 0.00017
20 10
11 a -/cl
SC D D /B
7.2%
3
10
3
10
3
10
2
10
2
10
2
10
1
10
1
10
1
10 0
10 0 0
10
1
10
2
10
3
10
4
-/-
85.1%
10
10
SCD/ Bcl11a
10 4
10 0 10 0
10 1
10 2
10 3
10 4
10 0
10 1
10 2
10 3
10 4
HbF
SC
C on
tro
l
0
SCD SCD/Bcl11a
SCD
10 4
3.6%
30
HbA or HbS
s
γ/(γ+β ) (%)
D
-/-
Control
www.sciencemag.org
SCIENCE
VOL 334
18 NOVEMBER 2011
995
REPORTS
996
mice was also reversed in SCD/Bcl11a–/– mice (fig. S12). To relate these phenotypic findings to HbF reactivation, we determined the steady-state level of g-globin expression and cellular distribution of HbF in SCD/Bcl11a–/– mice. Expression of g-globin genes was greatly elevated (P = 0.00017) in adult SCD/Bcl11a–/– mice compared with SCD mice (28.3% versus 1.3% of total b-like human globins) (Fig. 3D). The cellular distribution of HbF in red cells was assessed by staining for F cells using antibodies against HbF and HbA (which also recognizes HbS). Peripheral blood of control and SCD mice contained few F cells (3.6% and 7.2%, respectively) (Fig. 3E). In contrast, peripheral blood of SCD/Bcl11a–/– mice exhibited strong pancellular staining of HbF, and F cells accounted for 85.1% of total RBCs (P = 3.9 × 10−5) (Fig. 3E). The level of HbF expression achieved in the absence of BCL11A exceeds the estimated ~15 to 20% HbF thought to be necessary to virtually eliminate SCD phenotypes in patients (4). In addition, we introduced BCL11Anull alleles into an independent SCD mouse strain (22). Similar to the Berkeley SCD mice, inactivation of BCL11A corrected the sickle cell phenotypes due to HbF reactivation (fig. S13). Thus, our findings are not limited to a single SCD mouse model. Described more than 100 years ago (23), SCD is the first genetic disorder for which a causative mutation was identified (3). Despite extensive study of sickle hemoglobin containing erythrocytes and globin expression, effective therapy has been elusive and empirical. The correction of the phenotypes of SCD mice with removal of BCL11A provides proof of principle and critical in vivo evidence on behalf of therapeutic targeting of BCL11A in patients with SCD. We anticipate that our findings should apply similarly to the b-thalassemias. Several formidable barriers need to be overcome before these findings are translated into new therapies. As a transcription factor, BCL11A is a challenging therapeutic target. Interfering with BCL11A expression and/or function can be pursued by several strategies. Depletion of BCL11A mRNA by systemic administration of RNA inhibitory molecules (24, 25) or by somatic gene transfer of short hairpin RNA (26) merits exploration. Substantial advances in the transplantation of gene-modified autologous hematopoietic stem cells (27) make the latter approach attractive, particularly in that reactivation of g-globin expression is accompanied by reciprocal down-regulation of mutant b-globin on the same chromosome. Screening for small molecules (or peptides) that inhibit BCL11A function, either directly or through disruption of proteinprotein interactions, represents another approach (28). Recent advances in chemical biology enhance the prospects for discovery of drugs to accomplish this goal (29). The correction of SCD in mice by genetic manipulation of a single component involved in globin gene regulation constitutes a requisite step in translating new insights
18 NOVEMBER 2011
VOL 334
SCIENCE
in HbF silencing into mechanism-based, improved therapy for the major hemoglobin disorders. References and Notes 1. A. P. Kraus, B. Koch, L. Burckett, BMJ 1, 1434 (1961). 2. C. L. Conley, D. J. Weatherall, S. N. Richardson, M. K. Shepard, S. Charache, Blood 21, 261 (1963). 3. V. M. Ingram, Nature 178, 792 (1956). 4. O. S. Platt et al., N. Engl. J. Med. 330, 1639 (1994). 5. S. Menzel et al., Nat. Genet. 39, 1197 (2007). 6. M. Uda et al., Proc. Natl. Acad. Sci. U.S.A. 105, 1620 (2008). 7. G. Lettre et al., Proc. Natl. Acad. Sci. U.S.A. 105, 11869 (2008). 8. V. G. Sankaran et al., Science 322, 1839 (2008). 9. V. G. Sankaran et al., Nature 460, 1093 (2009). 10. J. Xu et al., Genes Dev. 24, 783 (2010). 11. A. Wilber et al., Blood 117, 2817 (2011). 12. A. Wilber, A. W. Nienhuis, D. A. Persons, Blood 117, 3945 (2011). 13. A. C. Heinrich, R. Pelanda, U. Klingmüller, Blood 104, 659 (2004). 14. P. Liu et al., Nat. Immunol. 4, 525 (2003). 15. J. DeSimone, P. Heller, L. Hall, D. Zwiers, Proc. Natl. Acad. Sci. U.S.A. 79, 4428 (1982). 16. T. J. Ley et al., N. Engl. J. Med. 307, 1469 (1982). 17. J. E. Bradner et al., Proc. Natl. Acad. Sci. U.S.A. 107, 12617 (2010). 18. T. M. Ryan, D. J. Ciavatta, T. M. Townes, Science 278, 873 (1997). 19. C. Pászty et al., Science 278, 876 (1997). 20. R. Pawliuk et al., Science 294, 2368 (2001). 21. M. E. Fabry et al., Blood 97, 410 (2001). 22. J. Hanna et al., Science 318, 1920 (2007). 23. J. B. Herrick, Arch. Intern. Med. 6, 517 (1910). 24. D. M. Dykxhoorn, J. Lieberman, Cell 126, 231 (2006). 25. R. E. Lanford et al., Science 327, 198 (2010). 26. C. Li, P. Xiao, S. J. Gray, M. S. Weinberg, R. J. Samulski, Proc. Natl. Acad. Sci. U.S.A. 108, 14258 (2011). 27. K. L. Shaw, D. B. Kohn, Sci. Transl. Med. 3, 97ps36 (2011). 28. R. E. Moellering et al., Nature 462, 182 (2009). 29. A. N. Koehler, Curr. Opin. Chem. Biol. 14, 331 (2010). Acknowledgments: We thank K. Peterson and H. Fedosyuk for b-locus mice, K. Gaensler and M. Groudine for the A20 strain of b-locus mice, T. M. Townes for the SCD mice, and M. Nguyen, R. Okabe, and J. Alves for technical assistance. We thank D. Higgs, D. A. Williams, L. I. Zon, D. E. Bauer, H. F. Xie, M. Kerenyi, T. Menne, J. Harriss, and J. Decker for advice and discussions. This work was supported by funding from NIH (S.H.O. and B.L.E.); the National Cancer Institute (NCI) and the Marie Betzner Morrow Endowment (P.W.T.); and a Center Award from the National Institute of Diabetes and Digestive and Kidney Diseases (S.H.O.). G.C.I. was supported by a National Research Service Award/NCI postdoctoral fellowship, and J.X. is a Howard Hughes Medical Institute–Helen Hay Whitney Foundation fellow. The authors declare no conflict of interest.
Supporting Online Material www.sciencemag.org/cgi/content/full/science.1211053/DC1 Materials and Methods SOM Text Figs. S1 to S13 Tables S1 to S3 References (30–43) 12 July 2011; accepted 4 October 2011 Published online 13 October 2011; 10.1126/science.1211053
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
By chromatin immunoprecipitation coupled to microarrays (ChIP-chip) analysis, we observe that histone deacetylase 1 (HDAC1) occupies the g-globin promoters in primary adult human erythroid precursors (Fig. 2C). Notably, the binding peaks of HDAC1 overlap peaks of trimethyl-H3 Lys27, consistent with the very-low-level expression of g-globin genes in adult erythroid cells. Administration of an HDAC inhibitor, suberoylanilide hydroxamic acid (SAHA), also synergized with BCL11A in derepression of g-globin mRNA (Fig. 2D). The level of histone H3 acetylation was elevated in bone marrow cells and within g-promoter regions after treatment (fig. S10), suggesting that the enhanced transcription of g-globin genes is related to the increase in histone acetylation. These results show that loss of BCL11A markedly enhances the effects of DNA demethylation and HDAC inhibitors in reactivating HbF expression and suggest that BCL11A down-regulation might be combined with known HbF inducers for efficient HbF augmentation. The above findings illustrate favorable features of BCL11A as a target for reactivation of HbF in the b-hemoglobinopathies. HbF silencing in the adult is strongly, although not exclusively, dependent on BCL11A. Red cell production is largely unaffected by its absence. Moreover, loss of BCL11A enhances effects of known HbF inducers. We next explored whether impairment of BCL11A is sufficient to ameliorate disease symptoms in a mouse model of SCD. Transgenic knockout mice expressing exclusively human sickle hemoglobin faithfully recapitulate the principal hematologic and histopathologic features of the human disease and are widely used to evaluate pharmacologic and genetic therapies (18–22). “Berkeley” SCD mice (19), which contain targeted deletions of murine a and b globins plus human fetal (Gg, Ag) and adult (d, bs) globin transgenes in a normal genomic configuration and express low levels of HbF in adults, have been employed in our primary analysis. We introduced BCL11A-null alleles into SCD mice (SCD Bcl11a fl/fl EpoR-Cre+, hereafter called SCD/Bcl11a–/– mice). As in typical human patients, SCD mice exhibit severe hemolytic anemia, reticulocytosis, and shortened RBC survival (Fig. 3, A and B, and Table 1). Hematologic parameters, including RBC counts and hemoglobin (Hb) content, were corrected in SCD/Bcl11a–/– mice (Table 1). Reticulocyte counts, red cell distribution width, and numbers of white blood cells (WBCs) and platelets were near control values (Table 1 and fig. S11). In stark contrast to SCD mice, sickled cells were absent in SCD/Bcl11a–/– mice (Fig. 3A), and red cell survival was markedly improved (Fig. 3B). Accordingly, spleen size, a measure of compensatory erythropoiesis, was dramatically reduced (Fig. 3C). In humans and SCD mice, RBC sickling leads to reduced medullary blood flow and impairs urine-concentrating ability. Urine osmolality of SCD/Bcl11a–/– mice was normal, heralding improved renal function (Table 1). Organ histopathology present in SCD
REPORTS
Nadia Dominici,1,2 Yuri P. Ivanenko,1 Germana Cappellini,1 Andrea d’Avella,1 Vito Mondì,3 Marika Cicchese,3 Adele Fabiano,3 Tiziana Silei,3 Ambrogio Di Paolo,3 Carlo Giannini,4 Richard E. Poppele,5 Francesco Lacquaniti1,2,6* How rudimentary movements evolve into sophisticated ones during development remains unclear. It is often assumed that the primitive patterns of neural control are suppressed during development, replaced by entirely new patterns. Here we identified the basic patterns of lumbosacral motoneuron activity from multimuscle recordings in stepping neonates, toddlers, preschoolers, and adults. Surprisingly, we found that the two basic patterns of stepping neonates are retained through development, augmented by two new patterns first revealed in toddlers. Markedly similar patterns were observed also in the rat, cat, macaque, and guineafowl, consistent with the hypothesis that, despite substantial phylogenetic distances and morphological differences, locomotion in several animal species is built starting from common primitives, perhaps related to a common ancestral neural network. ich repertoires of complex behaviors are created from the flexible combination of a small set of modules (1–9). A locomotor module is a functional unit—implemented in a neuronal network of the spinal cord—that generates a specific motor output by imposing a spatiotemporal structure to muscle activations (1, 3, 6). Each module involves a basic activation pattern (temporal structure) with variable weights of distribution (spatial structure) to different muscles (Fig. 1A). Thus, the electromyographic (EMG) activity of trunk and leg muscles during human adult locomotion is explained by few basic patterns independent of locomotion mode, direction, speed, and body support (3, 4, 8). These patterns may be regarded as locomotor primitives in a computational sense, because they are the building blocks from which locomotor activities are constructed. But are they primitives also in a developmental sense—that is, are they related to precursors present at or before birth? There are alternative hypotheses about the development of motor patterns (5, 9–12): (i) Adult-like stereotyped patterns are selected from an initially much larger number of motor patterns. (ii) Primitive patterns are discarded, replaced by entirely new patterns. (iii) Primitive patterns are retained and tuned, while new patterns are added during development. Clearly, these hypotheses imply different constraints on the development of the underlying neural networks.
R
1 Laboratory of Neuromotor Physiology, IRCCS Santa Lucia Foundation, 00179 Rome, Italy. 2Centre of Space Bio-medicine, University of Rome Tor Vergata, 00173 Rome, Italy. 3Department of Pediatrics, University of Rome Tor Vergata, Sant’Eugenio Hospital, 00144 Rome, Italy. 4Neonatology Unit, Sant’Eugenio Hospital, 00144 Rome, Italy. 5Department of Neuroscience, University of Minnesota, Minneapolis, MN 55455, USA. 6Department of Neuroscience, University of Rome Tor Vergata, 00133 Rome, Italy.
*To whom correspondence should be addressed. E-mail:
[email protected]
To discriminate among these possibilities, we compared the locomotor output of neonates with those of toddlers, preschoolers, and adults (13). Stepping can be evoked in newborns, but it is very irregular and typically disappears ~4 to 6 weeks postnatally (unless trained or supported by water buoyancy) to reappear at ~6 to 8 months, evolving into intentional walking (10–12). Meanwhile, size, mass, and proportions of the body segments have changed dramatically, as have the biomechanical requirements for locomotion.
A
Basic patterns
We elicited stepping in neonates supported on a table (Fig. 1B and movie S1) and recorded kinematics, contact forces, and EMG activity from up to 24 muscles simultaneously. EMG was modulated sinusoidally with the step cycle, with many extensor muscles coactivated over most of stance, and flexor muscles coactivated mainly during swing (Fig. 2A). Toddlers exhibited more complex, individuated EMG profiles, which were similar to those of preschoolers and adults, but sinusoidal modulation persisted in some muscles. Motoneuron outputs gradually became pulsatile, as in adults. To quantify the basic patterns underlying bilateral muscle activations, we applied a nonnegativematrix-factorization (13–14) to the EMGs pooled across steps. Because the EMG profiles were similar relative to a cycle regardless of its period (fig. S1), all data were normalized to the cycle. In each group of subjects, we could reproduce the EMG profiles of all recorded muscles by combining two to four basic patterns (Fig. 2B) with appropriate weights (Fig. 2C). Patterns were ordered on the basis of the relative timing of the peak. Two sinusoidal-like patterns accounted for 89 T 6% (mean T SD) of variance across neonates (fig. S2A). One pattern peaked at ~30% of the cycle, and the other one at ~75% (fig. S2C). For both, the duration at half-maximum was ~35% of the cycle (fig. S2B). In toddlers, we found two kinds of basic patterns, and a total of four patterns accounting for 90 T 4% of variance. The peak durations of patterns 1 and 3 were ~20% of the
Weights
Downloaded from www.sciencemag.org on November 17, 2011
Locomotor Primitives in Newborn Babies and Their Development
Muscle activations
w1
m1
p1
m2
x
=
w2
m3
p2 m1 m2 m3 m4 m5
w3 p3
0
% cycle
100
m4 m5
0
% cycle
100
B
Touch down
Mid-stance
Toe off
Mid-swing
Touch down
Fig. 1. (A) Schematic of motor modules. Simulated example of muscle activity profiles as weighted sum of basic patterns: mi(t)=∑ j pj(t)wij. The outputs of the first (green), second (blue), and third (magenta) modules are summed together to generate overall muscle activation (black envelope). (B) Illustration of a step cycle in a neonate.
www.sciencemag.org
SCIENCE
VOL 334
18 NOVEMBER 2011
997
cycle. The shape of patterns 2 and 4 was similar to that of the neonate (neonate-toddler correlation r = 0.88 and 0.98, respectively), as were peak duration (~35%) and timing (25 and 75%, respectively). Therefore, we infer that patterns 2 and 4 are retained from the stage of newborn stepping, whereas patterns 1 and 3 develop after that stage. In preschoolers, all four patterns (accounting for 91 T 2% of variance) showed transitional shapes, the average peak timing being intermediate between that of the toddler and the adult. Patterns 2 and 4 were quite variable across preschoolers, with a time shift of the peak relative to the cycle that correlated with age (r = 0.86 and 0.79 for patterns 2 and 4): the older the child, the closer the pattern to the adult. These transitional patterns strongly suggest a continuous development of the corresponding motor modules. In adults, we also found four patterns explaining 89 T 4% of variance, with peak duration 15 to 20% of the cycle. The timing of patterns 1 and 3 was similar to that in toddlers and preschoolers. Patterns 2 and 4 were timed on each foot contact, instead of midstance or midswing as in neonates and toddlers. The results depended little on the specific EMG decomposition technique (fig. S3). The neural underpinnings of these developmental changes remain speculative (Fig. 3A). Neonate stepping mainly reflects spinal and brainstem control, as shown by stepping anencephalic infants (10). Subsequent development stems from a growing integration of supraspinal, intraspinal, and sensory control (10–12). The lack of a specific activation pattern timed on foot contact in neonates could depend on immature sensory and/or descending modulation of stepping. Indeed, in the absence of sensory and descending modulation (e.g., during fictive locomotion), the spinal circuitry of animals may produce sinusoidal-like patterns (15), similar to those of human neonates. The addition of basic patterns in the first months of life may imply a functional reorganization of interneuronal connectivity, additional functional layers in the spinal central-pattern-generators (CPGs), and/or more powerful descending and sensory influences on CPGs (9, 15, 16). We found a good correlation between the developmental changes of the patterns and parallel changes in locomotion biomechanics (Fig. 3, B to D). In neonates, two activation patterns were sufficient for planar covariation (4) of segment motion (Fig. 3B), but posture was flexed, feet were lifted high during swing (Fig. 3C), and stereotyped heel-to-toe shifts of pressure during stance were absent (Fig. 3D). Pattern 2 provided (partial) body support during stance while pattern 4 drove the limb during swing, but there was no specific activation at either touch-down or lift-off. In toddlers, the new patterns (1 and 3) were timed at touch-down and lift-off, providing shear forces to decelerate and accelerate the body. The other two patterns (2 and 4) were similar to those of the neonate, as were the corresponding kinematic and kinetic events. In adults, the four patterns were accurately timed around the four
998
Fig. 2. Recorded EMG profiles and derived basic patterns. (A) Ensemble-averaged (across all subjects of each group) EMG profiles during the step cycle, aligned with stance onset in the right leg. Shaded areas are the experimental data, and black traces the profiles reconstructed as weighted sum of the patterns extracted from the ensemble. ES, erector spinae; GM, gluteus maximus; TFL, tensor fasciae latae; Add, adductor longus; HS, hamstrings; VM, vastus medialis; VL, vastus lateralis; RF, rectus femoris; MG, gastrocnemius medialis; LG, gastrocnemius lateralis; Sol, soleus; TA, tibialis anterior. (B) Basic patterns from averaged (across steps) EMG profiles in 10 subjects of each group (black). Patterns from ensemble EMG averages (colored). (C) Normalized weights of the ensemble patterns in color scale. critical events of the gait cycle: heel strike, weight acceptance, forward propulsion, and lift-off (3, 4, 8). Accordingly, the legs were kept much straighter than in children, and the planar covariation of segment motion was adjusted to fully exploit the inverted pendulum mechanism, with minimum muscle activity to support and propel the body. Center of pressure shifted smoothly heel-to-toe. Habitual erect, bipedal mode of locomotion sets humans apart in the animal kingdom and may have been a crucial initiating event in human evolution. Given the unique biomechanical features of human locomotion, it is not surprising that its muscle activity profiles differ markedly from those of other animals. Also, the developmental time course can differ (17): Small-brained
18 NOVEMBER 2011
VOL 334
SCIENCE
animals tend to walk independently shortly after birth, whereas independence is achieved by human infants only after ~1 year. These observations beg the question: Are the basic motor patterns unique to humans, or are they shared by other vertebrates with legged terrestrial locomotion? We applied the same analysis used in humans to the published recordings available from a few other mammals (rat, cat, rhesus monkey) and a bird (guineafowl) (13). Guineafowls are bipedal, whereas the other three species are quadrupedal. In neonate stepping rats, we found two patterns (accounting for 81% of variance) nearly identical to those of human neonates (human-rat correlation r = 0.94 and 0.98 for patterns 2 and 4, Fig. 4A). In all examined adult animals, we found four patterns with two types of modulation
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
REPORTS
REPORTS pools
logical differences of vertebrates (18). Thus, we identified similar activation patterns in animal species, some of which probably diverged more than 100 million years ago (17). Locomotion in humans and other species may be built starting from common, largely inborn primitives. CPGs coordinating muscle activity in legged vertebrates may have emerged during evolution from a common ancestral network (19). Conserved locomotor primitives may be a systems-level analog of the core modules of evolutionary biology (16, 20). Genes, proteins, cell types, and networks are conserved across several species with different lineages, constructing diverse natural forms and functions. Old elements are not discarded in favor of a totally new design, but they become adapted to solve novel problems efficiently.
C Leg posture and D Center of
coordination
foot contact
pressure
Toddler
Neonate
& pattern network
B Intersegmental
Downloaded from www.sciencemag.org on November 17, 2011
A Rhythm Premotor Motor
References and Notes
Foot
Adult
80 40 0 0
Shank -40
40 0
0
Thigh
100 nce % Sta
Descending
Fig. 3. Relationship between neural control modules and key biomechanical features of locomotion. (A) Motor rhythms and patterns generated by CPGs under descending and sensory influence (conceptual scheme). Activation patterns are distributed to different motoneuronal pools via a premotor network, dynamically reconfigurable through flexible weights. Intersegmental coordination (B), stick diagrams (C), and shifts of the center of pressure (D) are color-matched to the corresponding activation patterns. In toddlers, the first pattern (red) is timed at foot strike, the second (violet) at weight acceptance, the third (cyan) at forward propulsion, the fourth (green) at lift-off. In newborns, there are only two patterns, corresponding to the second and fourth of toddlers. Planar covariation of thigh elevation angle versus shank and foot angles identifies counterclockwise loops, with foot strike and lift-off at the top and bottom (B). Fig. 4. Comparison of activation patterns for locomotion in humans and other vertebrates. (A) Average patterns of human newborns are superimposed on those of neonatal rats; (B) patterns of human toddlers are superimposed on those of adult rats, cats, monkeys, and guineafowls; and (C) patterns of human adults stand alone.
A
B
C p1
1
p2
0 0
% cycle
100
p3
p4
Human neonate Neonatal rat
(accounting for 90 to 93% of variance, Fig. 4B): The two shorter pulses overlapped patterns 1 and 3 of human toddlers, whereas the two longer pulses overlapped patterns 2 and 4 of toddlers (toddler-animal correlation r = 0.94 T 0.04). The human developmental path appears to diverge from that of other animals after the stage of independent locomotion in toddlers, perhaps to accommodate discrete arm movements (such
Monkey Toddler Cat Rat Guineafowl
Human adult
as reaching and grasping an object) within rhythmic locomotion (4). Upper limb and trunk muscles can be engaged independently of locomotor activations (involving some of the same muscles) only if the latter occur as the brief events of adults (Fig. 4C). Our observations are consistent with the idea that motor patterns are highly conserved across substantial phylogenetic distances and morpho-
www.sciencemag.org
SCIENCE
VOL 334
1. S. Grillner, Science 228, 143 (1985). 2. F. A. Mussa-Ivaldi, S. F. Giszter, E. Bizzi, Proc. Natl. Acad. Sci. U.S.A. 91, 7534 (1994). 3. Y. P. Ivanenko, R. E. Poppele, F. Lacquaniti, J. Physiol. 556, 267 (2004). 4. Y. P. Ivanenko, G. Cappellini, N. Dominici, R. E. Poppele, F. Lacquaniti, J. Neurosci. 25, 7238 (2005). 5. D. Aronov, A. S. Andalman, M. S. Fee, Science 320, 630 (2008). 6. E. Bizzi, V. C. Cheung, A. d’Avella, P. Saltiel, M. Tresch, Brain Res. Rev. 57, 125 (2008). 7. D. Sheynikhovich, R. Chavarriaga, T. Strösslin, A. Arleo, W. Gerstner, Psychol. Rev. 116, 540 (2009). 8. D. J. Clark, L. H. Ting, F. E. Zajac, R. R. Neptune, S. A. Kautz, J. Neurophysiol. 103, 844 (2010). 9. C. B. Hart, S. F. Giszter, J. Neurosci. 30, 1322 (2010). 10. H. Forssberg, Exp. Brain Res. 57, 480 (1985). 11. E. Thelen, D. W. Cooke, Dev. Med. Child Neurol. 29, 380 (1987). 12. J. F. Yang, M. J. Stephens, R. Vishram, J. Physiol. 507, 927 (1998). 13. Materials and methods are available as supporting material on Science Online. 14. D. D. Lee, H. S. Seung, Nature 401, 788 (1999). 15. S. Grillner, Neuron 52, 751 (2006). 16. O. Kiehn, Curr. Opin. Neurobiol. 21, 100 (2011). 17. M. Garwicz, M. Christensson, E. Psouni, Proc. Natl. Acad. Sci. U.S.A. 106, 21889 (2009). 18. P. C. Wainwright, Curr. Opin. Neurobiol. 12, 691 (2002). 19. S. Grillner, T. M. Jessell, Curr. Opin. Neurobiol. 19, 572 (2009). 20. J. Gerhart, M. Kirschner, Proc. Natl. Acad. Sci. U.S.A. 104 (suppl. 1), 8582 (2007). Acknowledgments: We thank W. Gerstner, M. Molinari, P. Viviani, and M. Zago for helpful comments on earlier versions of the manuscript. This work was supported by the Italian Ministry of Health, Italian Ministry of University and Research (PRIN grant), Italian Space Agency (DCMC and CRUSOE grants), and European Union FP7-ICT programs (MINDWALKER grant 247959 and AMARSi grant 248311).
Supporting Online Material www.sciencemag.org.org/cgi/content/full/334/6058/997/DC1 Materials and Methods SOM Text Figs. S1 to S4 Table S1 Movie S1 References (21–27) 4 July 2011; accepted 28 September 2011 10.1126/science.1210617
18 NOVEMBER 2011
999
Rational Choice, Context Dependence, and the Value of Information in European Starlings (Sturnus vulgaris) Esteban Freidin* and Alex Kacelnik† Both human and nonhuman decision-makers can deviate from optimal choice by making context-dependent choices. Because ignoring context information can be beneficial, this is called a “less-is-more effect.” The fact that organisms are so sensitive to the context is thus paradoxical and calls for the inclusion of an ecological perspective. In an experiment with starlings, adding cues that identified the context impaired performance in simultaneous prey choices but improved it in sequential prey encounters, in which subjects could reject opportunities in order to search instead in the background. Because sequential prey encounters are likely to be more frequent in nature, storing and using contextual information appears to be ecologically rational on balance by conditioning acceptance of each opportunity to the relative richness of the background, even if this causes context-dependent suboptimal preferences in (less-frequent) simultaneous choices. In ecologically relevant scenarios, more information seems to be more. cquiring information involves time and energy (1–3). It follows that informationacquisition mechanisms should evolve if these costs are, on average, offset by the benefits of using knowledge to modify and improve decisions in the animal’s natural environment. However, in some situations decision rules that
A
Department of Zoology, University of Oxford, South Parks Road, Oxford OX1 3PS, UK. *Present address: Centro de Recursos Naturales Renovables de la Zona Semi-árida (CERZOS), Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Camino La Carrindanga Kilómetro 7, 8000 Bahía Blanca, Argentina. †To whom correspondence should be addressed. E-mail:
[email protected]
disregard available knowledge outperform alternatives that use it—an informational “less-ismore effect” (4). The frequency and importance of such paradoxes in natural scenarios is not known, but they do occur in both humans and nonhumans performing experimental tasks (5–15) and can be related to well-known breaches of economic rationality. Some examples include sunk costs, state-dependent learning, and contextdependent utility. The “sunk cost fallacy” is committed when knowledge of irrecoverable, retrospective costs increases preference for alternatives known to have had greater cost, distorting cost-free choices (5–8). “State-dependent valuation learning” occurs because items obtained
Fig. 1. Schematic of the sequence of trials in a session in which the context was (A) signaled or (B) unsignaled by the color of the trial-initiating light (x). On the right of each of the main columns, there are amplified representations of exemplar trials. Trials started with the x-light flashing. A peck to x led to either a no-choice trial or a choice (either simultaneous or sequential) after a 2-s random interval (b in Eqs. 1 and 2). Pecking options A3, B8, C13, and D35 yielded two food pellets after the delays indicated in the suffixes (in seconds). Pecking RAB or RCD caused the next trial to start immediately, hence choosing the R-option served to reject the food option available on that trial. Sessions consisted of two consecutive contexts (AB and CD), each comprising blocks of no-choice, simultaneous-choice, and sequentialchoice trials. A simultaneous choice trial in context AB consisted of B8 paired against A3 (8 out of 10 choices) or against C13 (2 out of 10 choices), and in context CD of C13 paired against D35 (8 out of 10 choices) or B8 (2 out of 10
1000
18 NOVEMBER 2011
VOL 334
when in greater need yield larger benefits, so that memory for accrued benefits hinders simultaneous choices when the subject’s state is identical for all options (9–11). Last, “context-dependent utility” results from the fact that perceived utility depends on background opportunities; thus, memory for context has the same hindering potential as state-dependent valuation learning, by enhancing preference for options associated with poorer contexts (12–15). We examined the impact of contextual information on choice using a laboratory representation of foraging decision-making by European starlings (Sturnus vulgaris). Starlings’ main foraging technique is to walk in short grass areas, briefly stopping to probe the ground searching for grubs, and then restarting their walking to cover new ground. Geographically or temporally identifiable foraging sites can be thought of as contexts that may contain a different assortment of prey of diverse quality. In a given site, upon detection of clues indicating the possible presence of a prey starlings either dig deeper to pursue the opportunity or walk on to continue searching. This involves a sequential decision in which the relative advantage of rejecting an opportunity to keep walking depends on information of both the attributes associated with each opportunity (prey species, capture time, capture probability) and the properties of the context (prey type distribution, intercapture intervals). Occasionally, a bird may simultaneously observe signs for two potential prey types and thus face a simultaneous choice. Generally, in these cases the optimal choice depends on the attributes of the prey types, regardless of the context. Thus, contextual information is irrelevant and sometimes can lead to losses. For example, if each of the items in a choice set is associated to
choices). A sequential choice trial in context AB consisted of either A3 or B8 paired against RAB, and in context CD of either C13 or D35 paired against RCD. As soon as subjects pecked at an option in a choice, the other one was turned off and disabled for that trial. IBI, inter-block interval (~10 min long); ICI, inter-context interval (45 min long) (20). SCIENCE
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
REPORTS
REPORTS vided the same amount of food, profitability comparisons are based on immediacy, the reciprocal of the delay between responding and food delivery. The two pairs appeared in two temporal contexts, hereafter referred to as “contexts” AB and CD (Fig. 1). For options B8 and C13, this arrangement causes the ranking of their objective profitabilities (B8 > C13) to be the opposite of the memory for their respective within-context ranking (B8 < A3, C13 > D35). Thus, any bias toward favoring C13 over B8 reflects the influence of contextual information. We manipulated the amount of contextual information in two conditions: In condition “context signaled,” the color of the trial-initiating light signaled whether the present time block (context) was AB or CD. In condition “context unsignaled,” this light was uncorrelated with context (Fig. 1). We examined sequential and simultaneous choices in each context. See further details in the caption of Fig. 1 (20). Figure 2A shows that varying the amount of contextual information affected subjects’ preference in simultaneous B8-versus-C13 choices. Starlings preferred B8 over C13 (the rate-maximizing choice) when context was unsignaled (t test against indifference: t7 = 3.84, P < 0.004) but not when context was signaled (t7 = 0.82, P = 0.22). A Wilcoxon matched pairs test showed a significant effect of context signaling on the proportion of choices for B8 over C13 (n = 8 starlings, z = 2.24, P = 0.025), but neither the context in which preference was measured (whether AB or CD) nor its interaction with condition were significant [analysis of variance (ANOVA), condition: F1,7 = 5.67, P = 0.049; context: F1,7 = 2.36, P = 0.17; and condition x context interaction: F1,7 < 1, P = 0.87), proving that the effects were mediated by contextual memory and not by context at the time of choice [supporting online material (SOM) text]. We discuss sequential decision-making following Charnov’s diet choice model (18). We assume that a forager encounters food options on average every b seconds of searching and can meet either of two option types, X and Y,
with probabilities pX and 1 − pX, respectively. If attacked, option X delivers aX food units after dX seconds, whereas option Y yields aY after dY seconds. Arbitrarily, we set option X as having greater profitability, defined as the ratio a/d. Equations 1 and 2 present the rate of intake of an ideal forager that is either a generalist (accepts all options; Eq. 1) or a specialist [only takes the option with the highest payoff in each context, rejecting the lower-profitability option upon encounter (we assume that identifying, handling, and if appropriate, rejecting items takes no time at all); Eq. 2]. To identify which of these is the ratemaximizing strategy, we plugged the experimental parameters into Eqs. 1 and 2 and compared the resulting rates (SOM text). Generalist rate ¼
pX aX þ ð1 − pX ÞaY (1) pX dX þ ð1 − pX ÞdY þ b
Specialist rate ¼
pX aX pX dX þ b
ð2Þ
The comparison shows that a specialist that took only the better option in each context (accepted A3 and C13 but rejected B8 and D35) would obtain a mean rate of intake 17% higher than would a generalist that consumed all options (averaging across contexts). Intuitively, this happens because the specialist uses the time that the generalist dedicates to exploit poorer options to search for the most profitable alternatives in the context. In terms of our experimental situation, optimal (ratemaximizing) sequential choices consist of always accepting option A3 and rejecting B8 in context AB (choose RAB over B8) and always accepting option C13 and rejecting D35 (choose RCD over D35) in context CD (Fig. 1). In contrast with the results in simultaneous choices, observed preferences in sequential choices came closer to the predictions of the ratemaximizing model just described when more contextual information was available: Signaling the context increased the proportion of choices of
Downloaded from www.sciencemag.org on November 17, 2011
the memory for a different background, then using background-relative information in a simultaneous choice may lead to preference for the worse item. Experimental animals are known to fall for this (12, 14, 15). Sequential encounters are different. According to rate-maximizing models in optimal foraging theory (OFT) [and profit maximization in microeconomics (16)], optimal sequential decisions depend on the attributes of the present opportunity and its background. To maximize foraging rate of gain, an individual should pursue each opportunity if doing so confers a higher expected outcome than foraging in the background. For this reason, far from being a potentially confusing factor, in sequential encounters contextual information is at the core of optimal decision-making. The lost-opportunity rationale is the basis of the two main paradigms in OFT, the marginal value theorem (17) and the diet choice model (18, 19). The fundamental difference in the role of contextual information between sequential encounters (in which it is crucial) and simultaneous encounters (in which it is irrelevant or harmful) gives us an opportunity to investigate the apparent paradox of context-dependent decision-making, in which agents gather contextual information but can experience losses by using it. We show that providing more contextual information leads to stronger context-dependent valuation of prey types, helping decision-makers (DMs) in sequential foraging but hindering them in simultaneous prey choices. Eight European starlings made both sequential and simultaneous decisions while we manipulated the availability of contextual information. The basic procedure was as follows: Four arbitrary stimuli were consistently paired with specific time intervals between the animal’s choice and food delivery (an analog of the pursuit time for each potential prey item). They were arranged in two pairs. The objective profitabilities of each of them ranked as follows: A3 > B8 > C13 > D35, with the suffixes indicating the delays (seconds) in each option (Fig. 1). Because all options pro-
Fig. 2. Results of (A) simultaneous choices for B8 over C13 and (B) sequential choices for B8 (over RAB) and for (C) C13 (over RCD). Solid diamonds indicate the mean (T1 SEM). Empty diamonds indicate individual subjects. Asterisks indicate significant statistical differences between conditions. www.sciencemag.org
SCIENCE
VOL 334
18 NOVEMBER 2011
1001
REPORTS Fig. 3. Subjects reward expectation as a function of delay, expressed as mean pecks per second in no-choice trials (open symbols, context signaled; solid symbols, context unsignaled; n = 8 starlings) for options A3 (◆,◇), B8 (■,□), C13 (▲,△), and D35 (●,◯).
1002
P = 0.90, respectively) (Fig. 3), implying that the effect of contextual information was not mediated by distortions of memory for physical attributes of the alternatives. In the wild, memory for contextual information may be highly adaptive because it enhances sensitivity to background opportunities. For starlings in their typical foraging settings, simultaneous choices are rare, and the occasional loss caused by context influence in such cases is likely to be overridden by the benefit they confer in sequential decisions. From a reverse engineering perspective, the widespread finding of context dependence across many species supports the inference that sequential decision-making has probably been a strong influence in the evolution of valuation and choice mechanisms across a majority of taxa. Thus, although the costs that these mechanisms cause in controlled experiments seem highly relevant to modern-day shopping decisions (13, 23), they are likely to have been less important in nature (24, 25). So far, however, it has not been possible to quantify the relative importance of different kinds of decisions in the ecological circumstances of any species. In conclusion, the advantageous influence of context dependence in sequential choices may be relevant for a variety of decision issues in which the relative value of incentives has been highlighted, from the study of heuristics and biases in animal (7–12, 14, 15) and human (4, 13, 23) decision-making to research on incentive relativity in behavior (26) and brain functioning (27).
18 NOVEMBER 2011
VOL 334
SCIENCE
T. D. Johnston, Adv. Stud. Behav. 12, 65 (1982). R. Dukas, J. Theor. Biol. 197, 41 (1999). F. Mery, T. J. Kawecki, Science 308, 1148 (2005). D. G. Golstein, G. Gigerenzer, in Handbook of Experimental Economics Results, C. R. Plott, V. L. Smith, Eds. (North-Holland, Amsterdam, 2008), pp. 987–992. 5. E. Aronson, J. Abnorm. Soc. Psychol. 63, 375 (1961). 6. H. R. Arkes, C. Blumer, Organ. Behav. Hum. Dec. 35, 124 (1985). 7. T. S. Clement, J. R. Feltus, D. H. Kaiser, T. R. Zentall, Psychon. Bull. Rev. 7, 100 (2000). 8. A. Kacelnik, B. Marsh, Anim. Behav. 63, 245 (2002). 9. L. Pompilio, A. Kacelnik, Anim. Behav. 70, 571 (2005). 10. L. Pompilio, A. Kacelnik, S. T. Behmer, Science 311, 1613 (2006). 11. J. M. Aw, R. I. Holbrook, T. Burt de Perera, A. Kacelnik, Behav. Processes 81, 333 (2009). 12. T. W. Belke, Anim. Learn. Behav. 20, 401 (1992). 13. I. Simonson, A. Tversky, J. Mark. Res. 29, 281 (1992). 14. T. A. Waite, Behav. Ecol. 12, 318 (2001). 15. L. Pompilio, A. Kacelnik, Proc. Natl. Acad. Sci. U.S.A. 107, 508 (2010). 16. P. A. Samuelson, W. D. Nordhaus, Economics (McGraw-Hill, New York, 2010). 17. E. L. Charnov, Theor. Popul. Biol. 9, 129 (1976). 18. E. L. Charnov, Am. Nat. 110, 141 (1976). 19. D. W. Stephens, J. R. Krebs, Foraging Theory (Princeton Univ. Press, Princeton, 1986). 20. Materials and methods are available as supporting material on Science Online. 21. A. C. Catania, in Theory of Reinforcement Schedules, W. N. Schoenfeld, Ed. (Appleton-Century-Crofts, New York, 1970), pp. 1–42. 22. S. Roberts, J. Exp. Psychol. Anim. B. 7, 242 (1981). 23. J. Huber, J. W. Payne, C. Puto, J. Consum. Res. 9, 90 (1982). 24. D. W. Stephens, D. Anderson, Behav. Ecol. 12, 330 (2001). 25. M. S. Shapiro, S. Siller, A. Kacelnik, J. Exp. Psychol. Anim. B. 34, 75 (2008). 26. C. Flaherty, Incentive Relativity (Cambridge Univ. Press, Cambridge, 1996). 27. W. Schultz, Curr. Opin. Neurobiol. 14, 139 (2004). Acknowledgments: This research was supported by the Programme Alban (the European Union program of High Level Scholarships for Latin America) scholarship E04D031814AR and Overseas Research Student Scheme Award UK to E.F., and UK Biotechnology and Biological Sciences Research Council grant BB/G007144/1 to A.K. The authors declare no competing financial interests or conflict of interest. E.F. performed the experiment and collected, processed, and analyzed the data. Both authors shared the experimental design and writing of the paper.
Supporting Online Material www.sciencemag.org/cgi/content/full/334/6058/1000/DC1 Materials and Methods SOM Text References (28, 29) 10 June 2011; accepted 7 October 2011 10.1126/science.1209626
www.sciencemag.org
Downloaded from www.sciencemag.org on November 17, 2011
RAB over B8 and of C13 over RCD. (Wilcoxon matched pairs tests of the proportion of choice with condition as a factor; n = 8 starlings; RAB over B8, z = 1.96, P = 0.049; C13 over RCD, z = 2.38, P = 0.017) (Fig. 2, B and C, and SOM text). Context effects could be mediated either by context dependence for the memory of the options’ utility or for the memory of the options’ physical attributes (15). These possibilities can be differentiated when, as here, options only differ in delay to outcome because there is an independent measure of attribute knowledge. Key pecking during options’ delay to food indicates the subjects’ expected time of reward (21, 22). Options B8 and C13 were the alternatives with the most similar profitabilities, and choices involving these options were the most affected by manipulation of contextual information. We were particularly interested in testing first, whether subjects discriminated between them, and second, whether contextual information influenced memory of reward immediacy. The answers were yes and no, respectively (Fig. 3). Pecking patterns during the delay to food prove that they discriminated between B8 and C13 and knew that B8 involved a shorter delay. This can be inferred from the higher pecking rate toward B8 than C13 during the initial 8 s of responding to each stimulus (ANOVA, F1,7 = 15.12, P = 0.006); after that, B8 delivered its food reward and was turned off (Fig. 3). Also, timing discrimination was not affected by the differential amount of contextual information (ANOVA, condition and option x condition interaction, both Fs < 1, P = 0.98, and
References and Notes 1. 2. 3. 4.
LIFE SCIENCE TECHNOLOGIES
Produced by the Science/AAAS Publishing Office AAAS/ScienceCustom Business Office Feature
NEW PRODUCTS: PRODUCTS AUTOMATION IHC/ISH Platform The Leica BOND RX is a truly open, fully automated immunohistochemistry (IHC) and in situ hybridization (ISH) platform, which delivers researchers the freedom they need to make breakthrough discoveries for new cancer treatments. The efficiency of full automation is matched to unlimited reagent selection and freely customizable protocols. Where previous research platforms were restrictive, the Leica BOND RX offers unprecedented choice with researchers able to select the ideal reagents, dispense sequences, and incubation conditions for any study. Drug discovery trials and cancer research are further enhanced by exceptional staining quality and total tissue care. With precise automation, Leica BOND RX delivers the consistency needed for detailed analysis while the unique Covertile system provides maximum protection for every sample so vital information is not lost. Early and late-stage researchers will also appreciate the speed and efficiency of Leica BOND RX. For info: 800-248-0123
www.leica-microsystems.com
Robotic Autosampler Compatible with Thermo Scientific GC and GC/MS systems, the TriPlus RSH Autosampler sets new standards in automation and provides advanced liquid handling cycles that enable automated functionality beyond traditional liquid, headspace, and solid-phase microextraction (SPME) injections. Designed to provide exceptional repeatability and reliability, the TriPlus RSH Autosampler delivers high-quality analytical results and features precise built-in automation to improve productivity and prevent human error. The groundbreaking Automatic Tool Change (ATC) capability enables seamless operation by automating the exchange of syringes for different tasks in a single, unattended sequence prior to sample injection. Another significant feature provided by this new sampling system is the ability to expand unattended operations and productivity by offering unprecedented sample capacity. A sample vial capacity of 648 2 mL vials, combined with multiple 100 mL wash/waste bottles, allows the TriPlus RSH to achieve weekendlong unattended operations. Thermo Fisher Scientific
for quickly developing rugged, reliable solid-phase extraction (SPE) methods. The powerful, high throughput system is dedicated specifically for SPE, eliminates bottlenecks and allows users the ability to realize the full benefits of today’s powerful analytical instruments. RapidTrace+ is fully compatible with 1, 3, and 6 mL SPE columns. Productivity is increased with a five-position rack for 40 mL scintillation vials, allowing for larger sample volumes to be processed. The modular design of the RapidTrace+ allows for up to 10 workstations to be linked and controlled by one computer. In its full modular configuration, these units can process up to 100 samples per hour. Easy-to-use and intuitive software dramatically increases the speed and reduces the effort involved in method development. Methods can be developed in minutes and recalled in seconds; with the ability to insert, delete, or modify steps as well as easily adjust parameters. Biotage For info: 800-446-4752
www.biotage.com
Automated Cell Culture System
SPE Workstation
The CompacT SC is an automated T-flask cell culture system designed to fully control expansion, differentiation, and maintenance of stem cells. The new CompacT SC, which can be used to aseptically culture multiple different stem cell lines in parallel using T75 and T175 flasks, has a number of features which makes controlling the growth and differentiation of stem cells an easy, convenient task. The system includes automated cell imaging and analysis, allowing scientists to program the CompacT SC to monitor cell morphology and growth curves. It also permits researchers to set the system to respond to any preprogrammed changes by automatically passaging cells and adding new media, reagents, or vectors at appropriate times. This means scientists can consistently maintain cell lines in the state they need them, either as pluripotent cells or during cell differentiation and transfection. Processes can be performed at any time, without researchers having to be in the laboratory. TAP Biosystems
The RapidTrace+ SPE Workstation is a robust, automated system
For info: +44-(0)-1763-227200
For info: 800-532-4752
www.thermoscientific.com
Automated DNA/NA Isolation Kits The Certal Extraction Kits—the QIAsymphony Certal Residual DNA Kit and Vaccine NA Kit—are designed for the automated DNA isolation of host cell residues and viral nucleic acids in biopharmaceuticals. With the new kits, residual DNA is automatically isolated with the QIAsymphony system from a variety of sample matrices, including bioprocess purification buffer, cell culture supernatant, or vaccine preparations. The kits provide an automated benchtop system that minimizes manual hands-on time with optimized protocols that ensure consistent, reproducible results that meet regulatory requirements. Qiagen For info: 800-426-8157
www.qiagen.com
www.tapbiosystems.com
Electronically submit your new product description or product literature information! Go to www.sciencemag.org/products/newproducts.dtl for more information. Newly offered instrumentation, apparatus, and laboratory materials of interest to researchers in all disciplines in academic, industrial, and governmental organizations are featured in this space. Emphasis is given to purpose, chief characteristics, and availability of products and materials. Endorsement by Science or AAAS of any products or materials mentioned is not implied. Additional information may be obtained from the manufacturer or supplier.
1004
18 NOVEMBER 2011
VOL 334
SCIENCE
www.sciencemag.org/products
Downloaded from www.sciencemag.org on November 17, 2011
Leica Microsystems
Methods for Discc overy a nd Detecc tio on
WEDNESDAY, NOVEMBER 30, 2011 12 noon U.S. Eastern 9 a.m. Pacific 5 p.m. UK
Both large and small changes to DNAcollectively described as genetic variantsare known contributors to human disease. The detection and characterization of single nucleotide polymorphisms (SNPs) and copy-number variations (CNVs) are an important research focus for scientists. This webinar will introduce two leading technologies useful for discovering such genetic variants: targeted DNA resequencing and comparative genomic hybridization which, used together, are a powerful combination for both SNP and CNV detection.
PARTICIPANTS: Hakon Hakonarson, M.D., Ph.D. Childrens Hospital of Philadelphia Philadelphia, PA
Ray Hershberger, M.D. University of Miami Miami, FL
Andrew J. Sharp, Ph.D. Mount Sinai School of Medicine New York, NY
During the webinar, our expert panel of researchers will:
REGISTER NOW!
Talk about best practices for applying these technologies to detect genetic variations
Sign Up At:
Share their experiences using these technologies in human disease research
www.sciencemag.org/webinar
Answer your questions live and in real time!
Webinar sponsored by
Brought to you by the Science/AAAS Custom Publishing OfÃce
SPECIAL OFFER
$99
ANTIBODY SPECIAL
Find out more @ www.epitomics.com/products/promo
MONO
RabMAbs = BETTER STEM CELL ANTIBODIES ®
• Antibodies to pluripotent transcription factors: LIN28, Nanog, OCT4, SOX2 and more
• Validated in WB, IHC, IF, IP and FACS • Over 6800 antibodies and counting
Why RabMAbs make better stem cell antibodies: The rabbit immune system responds better to difficult antigens than the mouse immune system. This is critical for the development of high quality antibodies. Epitomics Rabbit Monoclonal Antibodies (RabMAbs) provide the combined benefits of superior antigen recognition of the rabbit immune system with the specificity and consistency of a monoclonal antibody, bringing yo the highest quality antibody possible. Epitomics has developed various high quality stem cell marker antibodies, including a panel of RabMAbs to core pluripotent cell markers (e.g. LIN28, Nanog, OCT4, SOX2).
te
r Antibo
d
s
Be
ie
t
Find out more @ www.epitomics.com/stemcell
B
e
e
Antibody Satisfaction Guarantee tt
er Scie
n
c
SOX2
Nanog
Immunohistochemical analysis of paraffin-embedded embryonal carcinoma using anti-SOX2 RabMAb (cat. #2683-1)
Immunohistochemical analysis of paraffin-embedded human dysgerminoma tissue using anti-Nanog RabMAb (cat. #3369-1)
www.epitomics.com
US & Canada | 1-877-772-2622 Outside NA | 650-583-6688
GE Healthcare Life Sciences
Is it science or art? Check out our gallery… Scientists from around the world have been submitting inspirational images from their cellular research into GE Healthcare’s 2011 IN Cell Analyzer Image Competition. Now it’s your chance to vote for your favorite image and receive a limited edition, 2012 Image Competition calendar* featuring twelve of the most popular images as voted by you. The winning images will be displayed on NBC Universal’s high-definition screen in Times Square, New York, so you decide – which image best captures the art in science? Cast your vote today at www.gelifesciences.com/incellcompetition
imagination at work
* Offer available to the first 2 000 eligible registrants unless prohibited by applicable local law, regulations or General Electric’s internal policies or rules relating to governments and healthcare professionals. This advert features winning images from previous IN Cell Analyzer Image Competitions. GE, imagination at work, and GE monogram are trademarks of General Electric Company. © 2011 General Electric Company—All rights reserved. First published Nov. 2011. GE Healthcare Bio-Sciences AB, Björkgatan 30, SE-751 84 Uppsala, Sweden. 11/2011
Science Mobile App Now Available for Android Phones
They say you never know when inspiration i ti will ill strike. t ik Download the Science mobile app for Android devices and be ready the next time you’re inspired to read the latest news, research, and career advice from Science on your mobile phone.
FFeatures include: • Summaries and abstracts from Science, Science Translational Medicine, and Science Signaling. • Ability to e-mail full-text links. • The latest news from ScienceNOW.
To download the Science mobile app for Android visit content.aaas.org/mobile, visit the Android Market on your phone, or just scan this barcode.
• Career advice articles from Science Careers. • Access to the Science weekly podcast and other multimedia. • Content caching for reading without wi-fi access.
Free up your time
Automated sample and assay technologies by QIAGEN Automated solutions from sample to result: ■ The widest choice of sample processing protocols ■ Low-, medium-, and high-throughput automation ■ Leading solutions for molecular testing ■ Plug-and-play automated sample preparation
AUTOPortf1111S1WW
■ Quantitative, real-time PCR detection ■ Automated analysis of DNA fragments and RNA ■ High-resolution sequence-based DNA detection and quantification Making improvements in life possible — www.qiagen.com
Sample & Assay Technologies
Built for performance. DNA Ligases from New England Biolabs. Why compromise any aspect of performance in your cloning experiments? DNA ligases from NEB set the industry standard for function and purity. Continuous product development and extensive quality controls ensure success in your ligation reactions. DNA ligases from NEB – building a strong foundation for your cloning experiments.
Experience extreme purity with NEB’s T4 DNA Ligase M
NEB Lot 101
NEB Lot 102
A
Supplier B C
D
SPECIAL OFFER Purchase NEB 5-alpha Competent Cells with NEB’s Quick Ligation Kit and receive the item of lesser value for FREE
Equivalent amounts of protein were loaded and analyzed by Coomassie gel and silver stained using SilverXpress™. Marker M is NEB’s Broad Range Protein Marker (NEB #P7702).
For more information, visit
www.neb.com/DNAligases
SilverXpress™ is a trademark of Invitrogen. Monarch caterpillar © Maryellen Baker/Botanica
Call for Proposals The Progressive Multifocal Leukoencephalopathy (PML) Consortium, a not-for-profit pharmaceutical collaboration with the mission of reducing the occurrence, morbidity and mortality of PML, is seeking proposals for research into the pathogenesis of, diagnosis of, and risk factors associated with PML.
The complete Science archive 1880–1996
Funding is available through the Consortium for research grants. The recommended range of funding per grant is up to $300,000 over two years. Areas of interest include: >>
Fully integrated with Science Online (1997–today)
PML Pathogenesis (JCV virology and host resistance mechanisms)
>>
PML Diagnosis
>>
PML Risk Factors
>>
Treatment of PML (therapeutic and preventative)
Available to institutional customers through a site license. Contact
[email protected] for a quote.
Applicants with related proposals in other disease areas included in the scope of the RFP that meet the program requirements are encouraged to apply.
Information: www.sciencemag.org/classic
Please visit www.pmlconsortium.org for more information.
© 2007 JupiterImages Corporation
The deadline for initial proposals is January 31, 2012.
A Journal with Impact from AAAS, the publisher of Science
Science Translational Medicine Integrating Medicine and Science A recent journal article features the sequencing of fetal DNA from plasma of a pregnant woman to permit prenatal, noninvasive genome-wide screening to diagnose fetal genetic disorders. Sci Transl Med 8 December 2010: Vol. 2, Issue 61, p. 61ra91 DOI: 10.1126/scitranslmed.3001720
Recommend an institutional subscription to your library today! ScienceOnline.org/recommend
ScienceTranslationalMedicine.org
Indexed in MEDLINE/PubMed
QuantStudio™ 12K Flex Real-Time PCR System Do what you never thought possible. With 5 interchangeable blocks, you can analyze from 1 to over 12,000 data points per run in any assay format that is right for your project. The QuantStudio™ 12K Flex system transforms the leading real-time PCR platforms into a single platform combining maximum throughput, scalability, and flexibility. Now, targeted discovery, confirmation, high-throughput screening, and everything in between, can be done on a single system.
See what the system can do at www.appliedbiosystems.com/quantstudio For Research Use Only. Not intended for human or animal therapeutic or diagnostic use. © 2011 Life Technologies Corporation. All rights reserved. The trademarks mentioned herein are the property of Life Technologies Corporation or their respective owners. CO124128 1011
i get data faster with 20x more assay choices
i am TaqMan
®
Speed through samples with the largest collection of preoptimized assays With 20x more predesigned and preoptimized assays in multiple sizes and formats, TaqMan® Assays help you detect your desired target virtually 100% of the time. Browse the world’s largest selection of real-time PCR assays for gene expression, SNP genotyping, microRNA and noncoding RNA, pathogen research, copy number variation, and protein analysis today. And as always, all TaqMan® Assays carry our “QPCR” Guarantee.* Shop now at www.appliedbiosystems.com/taqman *Certain terms and conditions apply. To obtain a copy of the full terms and conditions of the guarantee, visit www.lifetechnologies.com/taqmanguarantee. For Research Use Only. Not intended for any animal or human therapeutic or diagnostic use. © 2011 Life Technologies Corporation. All rights reserved. The trademarks mentioned herein are the property of Life Technologies Corporation or their respective owners. TaqMan® is a registered trademark of Roche Molecular Systems, Inc., used under permission and license. CO24202 1011
the difference between
breakthrough and try again
Gibco® cell culture products are the most trusted for quality and consistency The leaders in cell culture, Gibco® products are as uncompromising in their performance as you are in your dedication to scientific research. That’s why they are more cited than any other brand of cell culture products. From classical Gibco® media to our premium GlutaMAX™ media and FBS, you’ll see the Gibco® difference. Shop now at www.lifetechnologies.com/discovergibco Gibco. Every little thing matters. ©2011 Life Technologies Corporation. All rights reserved. The trademarks mentioned herein are the property of Life Technologies Corporation or their respective owners. For research use only. Not intended for any animal or human therapeutic or diagnostic use. CO24200 1011
AAAS Annual Meeting
Be Part of a Winning Team
Vancouver is a dynamic, multicultural city set in a spectacular natural environment. It was the proud host of the 2010 Olympic and Paralympic Winter Games. In 2012 it hosts the world’s annual Olympiad of Science, 16-20 February. AAAS organizes an international conference annually— 4 days of symposia, lectures, seminars, workshops, and poster sessions that cover every area of science, technology, and education. We also organize a unique community science showcase that offers an array of hands-on demonstrations and other family friendly activities.
They represent nearly all U.S. states and territories as well as up to 55 countries. Technologically savvy, our attendees are always looking for new and improved tools and resources. Take this opportunity to select or design your sponsorship and secure your exhibit space. For more information about sponsorship opportunities and the exhibition, contact Isabel Patterson, AIM Meetings, (703) 549-9500;
[email protected]. For the program, go to www.aaas.org/meetings.
Attendees constitute a highly educated, global market that promotes the effective use of research and practice.
Presenting sponsor
AAAS thanks for its generous support of the Science Journalism Awards.
Learn how current events are impacting your work. ScienceInsider, the new policy blog from the journal Science, is your source for breaking news and instant analysis from the nexus of politics and science. Produced by an international team of science journalists, ScienceInsider offers hard-hitting coverage on a range of issues including climate change, bioterrorism, research funding, and more. Before research happens at the bench, science policy is formulated in the halls of government. Make sure you understand how current events are impacting your work. Read ScienceInsider today.
www.ScienceInsider.org
Breaking news and analysis from the world of science policy
LIFE SCIENCE TECHNOLOGIES
Produced by the Science/AAAS Publishing Office AAAS/ScienceCustom Business Office Feature
NEW PRODUCTS PRODUCTS: AUTOMATION IHC/ISH PLATFORM The Leica BOND RX is a truly open, fully automated immunohistochemistry (IHC) and in situ hybridization (ISH) platform, which delivers researchers the freedom they need to make breakthrough discoveries for new cancer treatments. The efficiency of full automation is matched to unlimited reagent selection and freely customizable protocols. Where previous research platforms were restrictive, the Leica BOND RX offers unprecedented choice with researchers able to select the ideal reagents, dispense sequences, and incubation conditions for any study. Drug discovery trials and cancer research are further enhanced by exceptional staining quality and total tissue care. With precise automation, Leica BOND RX delivers the consistency needed for detailed analysis while the unique Covertile system provides maximum protection for every sample so vital information is not lost. Early and late-stage researchers will also appreciate the speed and efficiency of Leica BOND RX. Leica Microsystems For info: 800-248-0123
www.leica-microsystems.com
ROBOTIC AUTOSAMPLER Compatible with Thermo Scientific GC and GC/MS systems, the TriPlus RSH Autosampler sets new standards in automation and provides advanced liquid handling cycles that enable automated functionality beyond traditional liquid, headspace, and solid-phase microextraction (SPME) injections. Designed to provide exceptional repeatability and reliability, the TriPlus RSH Autosampler delivers high-quality analytical results and features precise built-in automation to improve productivity and prevent human error. The groundbreaking Automatic Tool Change (ATC) capability enables seamless operation by automating the exchange of syringes for different tasks in a single, unattended sequence prior to sample injection. Another significant feature provided by this new sampling system is the ability to expand unattended operations and productivity by offering unprecedented sample capacity. A sample vial capacity of 648 2 mL vials, combined with multiple 100 mL wash/waste bottles, allows the TriPlus RSH to achieve weekendlong unattended operations. Thermo Fisher Scientific
for quickly developing rugged, reliable solid-phase extraction (SPE) methods. The powerful, high throughput system is dedicated specifically for SPE, eliminates bottlenecks and allows users the ability to realize the full benefits of today’s powerful analytical instruments. RapidTrace+ is fully compatible with 1, 3, and 6 mL SPE columns. Productivity is increased with a five-position rack for 40 mL scintillation vials, allowing for larger sample volumes to be processed. The modular design of the RapidTrace+ allows for up to 10 workstations to be linked and controlled by one computer. In its full modular configuration, these units can process up to 100 samples per hour. Easy-to-use and intuitive software dramatically increases the speed and reduces the effort involved in method development. Methods can be developed in minutes and recalled in seconds; with the ability to insert, delete, or modify steps as well as easily adjust parameters. Biotage For info: 800-446-4752
www.biotage.com
AUTOMATED CELL CULTURE SYSTEM
SPE WORKSTATION
The CompacT SC is an automated T-flask cell culture system designed to fully control expansion, differentiation, and maintenance of stem cells. The new CompacT SC, which can be used to aseptically culture multiple different stem cell lines in parallel using T75 and T175 flasks, has a number of features which makes controlling the growth and differentiation of stem cells an easy, convenient task. The system includes automated cell imaging and analysis, allowing scientists to program the CompacT SC to monitor cell morphology and growth curves. It also permits researchers to set the system to respond to any preprogrammed changes by automatically passaging cells and adding new media, reagents, or vectors at appropriate times. This means scientists can consistently maintain cell lines in the state they need them, either as pluripotent cells or during cell differentiation and transfection. Processes can be performed at any time, without researchers having to be in the laboratory. TAP Biosystems
The RapidTrace+ SPE Workstation is a robust, automated system
For info: +44-(0)-1763-227200
For info: 800-532-4752
www.thermoscientific.com
AUTOMATED DNA/NA ISOLATION KITS The Certal Extraction Kits—the QIAsymphony Certal Residual DNA Kit and Vaccine NA Kit—are designed for the automated DNA isolation of host cell residues and viral nucleic acids in biopharmaceuticals. With the new kits, residual DNA is automatically isolated with the QIAsymphony system from a variety of sample matrices, including bioprocess purification buffer, cell culture supernatant, or vaccine preparations. The kits provide an automated benchtop system that minimizes manual hands-on time with optimized protocols that ensure consistent, reproducible results that meet regulatory requirements. Qiagen For info: 800-426-8157
www.qiagen.com
www.tapbiosystems.com
Electronically submit your new product description or product literature information! Go to www.sciencemag.org/products/newproducts.dtl for more information. Newly offered instrumentation, apparatus, and laboratory materials of interest to researchers in all disciplines in academic, industrial, and governmental organizations are featured in this space. Emphasis is given to purpose, chief characteristics, and availability of products and materials. Endorsement by Science or AAAS of any products or materials mentioned is not implied. Additional information may be obtained from the manufacturer or supplier.
1004
18 NOVEMBER 2011
VOL 334
SCIENCE
www.sciencemag.org/products
WEBINaAbleR Now avail on demand.
FACTS&FICTION Careers in Industry and Academia Trying to figure out the next step in your career? Join us for a roundtable discussion that will look at facts and fiction surrounding academic and industry career options for PhD-level scientists. Get some nuts and bolts advice on how to research career options, what questions to ask, and how to best prepare for various careers. • Do industry and academic careers require different skill sets? • Do industry jobs have better compensation? Less autonomy? • Do academic scientists have less work/life balance?
For answers view our roundtable discussion for free at:
ScienceCareers.org/webinar
Produced by the Science/AAAS Business Office.
Bioidentify. Easily identify the gene targets of your microRNA. Introducing the MISSION® Target ID Library.
The MISSION Target ID Library enables transcriptome-wide identification of gene targets regulated by microRNA and other non-coding RNA. Simplify target identification with these easy steps: transfect library, select, transfect miRNA and select target clones.
sigma.com/targetid
biosilencing ©2011 Sigma-Aldrich Co. LLC. All rights reserved. SIGMA and SIGMA-ALDRICH are trademarks of Sigma-Aldrich Co. LLC, registered in the US and other countries. MISSION is a registered trademark of Sigma-Aldrich Co. LLC. Where bio begins is a trademark of Sigma-Aldrich Co. LLC.
For full advertising details, go to ScienceCareers.org and click For Employers, or call one of our representatives.
Tracy Holmes Worldwide Associate Director Science Careers Phone: +44 (0) 1223 326525
UNITED STATES & CANADA E-mail:
[email protected] Fax: 202-289-6742 Tina Burks Midwest/West Coast/ South Central/Canada Phone: 202-326-6577
Candidates must have defended their PhD thesis after January 15, 2004. Successful candidates will be appointed as head of a group of up to 6 people for a period of 5 years. The budget (up to 1,500,000€ over 5 years) includes the salary for the group leader (if necessary), a three-year postdoctoral position, a technician, part-time secretarial assistance, basic laboratory equipment, a substantial contribution to running costs and essential large equipment, and access to on-campus facilities including state-of-the-art technology platforms. Candidates should send their formal applications by E-mail to the Director of Scientific Evaluation, Prof. Alain Israël, at the Institut Pasteur (
[email protected]). The Application shall comprise the following (in order) in a single pdf file: 1. A brief introductory letter (candidates are encouraged to contact the coordinators of the LabEx, Pascale Cossart (
[email protected]) or Philippe Sansonetti (
[email protected]) 2. A Curriculum Vitae and a full publication list. 3. A description of past and present research activities (up to 5 pages with 1.5 spacing). 4. The proposed research project (up to 10 pages with 1.5 spacing) and how it would fit in the defined topic. 5. The names of 3 scientists from whom letters of recommendation can be sought, together with the names of scientists with a potential conflict of interest from whom evaluations should not be requested.
Elizabeth Early East Coast & Industry Phone: 202-326-6578 Marci Gallun Sales Administrator Phone: 202-326-6582
Faculty Position
Online Job Posting Questions Phone: 202-312-6375
The Koch Institute for Integrative Cancer Research at the Massachusetts Institute of Technology (http://ki.mit.edu/) invites applications for a junior or senior faculty appointment, Assistant Professor or above. The Koch Institute is an NCI-designated Cancer Center which features research across a wide range of areas in cancer biology and cancer-oriented engineering.
EUROPE & REST OF WORLD E-mail:
[email protected] Fax: +44 (0) 1223 326532 Simone Jux Phone: +44 (0)1223 326529 Alex Palmer Phone: +44 (0) 1223 326527 Customer Service Phone: +44 (0) 1223 326528
JAPAN Makiko Hara Phone: +81 (0) 90-9853-9982 E-mail:
[email protected] Fax: +81 (0) 3-6369-4491
CHINA & TAIWAN Ruolei Wu Phone: +86-1367-1015-294 E-mail:
[email protected] All ads submitted for publication must comply with applicable U.S. and non-U.S. laws. Science reserves the right to refuse any advertisement at its sole discretion for any reason, including without limitation for offensive language or inappropriate content, and all advertising is subject to publisher approval. Science encourages our readers to alert us to any ads that they feel may be discriminatory or offensive.
at free copy today Download your ooklets
.org/b ScienceCareers
This is an open search with regard to field of study and specific research focus, but clear cancer relevance of the proposed research program is essential. Areas of interest include, but are not limited to: imaging, proteomics, single-cell analysis, systems biology, metastasis, stem cell biology, and novel approaches to detecting, monitoring and treating cancer. The candidate(s) will be expected to develop and lead an internationally competitive research program as well as participate in undergraduate and graduate teaching. The successful candidate(s) will have laboratory space in the recently opened Koch Institute building and a faculty appointment in an appropriate department at MIT. The deadline for the complete application with letters is December 1, 2011. Please apply at https://academicjobsonline.org/ajo/jobs/926 The David H. Koch Institute for Integrative Cancer Research at MIT is an Equal Employment Opportunity/Affirmative Action Employer.
Brought to you by the AAAS/Science Business Office
online @sciencecareers.org
s r e e r a C e c n Sciedvertising A
The Institut Pasteur announces an international call for candidates wishing to create independent young researcher groups on its Paris campus in France. The call responds -in the framework of the French Government’s program entitled “Investments for the Future”, - to the recent award to Institut Pasteur of a Laboratory of Excellence (LabEx) named Integrative Biology of Emerging Infectious Diseases (IBEID). Four groups will be created with one group in each of the four following topics: - Antibiotic resistance - Novel regulatory mechanisms in bacteria - Signaling in microbial communities - Arboviruses A complete description of these topics can be found at: http://www.pasteur.fr/ip/resource/filecenter/ document/01s-00004f-0v3/ibeid-g5-1.pdf The deadline for applications is January 15, 2012. Short-listed candidates will be called for interview in February-March 2012 and decisions will be announced by June 30.
online @sciencecareers.org
Faculty Position in Metabolism Research Faculty Position in Cancer Research The Children’s Research Institute (CRI) at the University of Texas Southwestern Medical Center seeks applications for a tenure-track faculty position in Dallas, TX at the Instructor, Assistant Professor, or Professor level in the area of cancer biology. Outstanding investigators at any academic rank will be considered. Candidates must have a Ph.D., M.D. or equivalent degrees, a track record of outstanding research, and the ability to direct an independently-funded research program. The UT Southwestern Medical Center has a distinguished history of excellence in disease-related basic science research. The CRI is a new institute dedicated to recruiting outstanding scientists dedicated to solving fundamental problems in human disease and to providing a dynamic, stimulating and highly collaborative scientific environment. Major areas of focus within the CRI will include stem cell biology and cancer in addition to metabolism. Please submit a CV, a 2-page summary of past accomplishments and research plans, and ask three references to submit letters to Beth Morris at
[email protected]. CRI is a collaborative venture with Children’s Medical Center of Dallas
The Children’s Research Institute (CRI) at the University of Texas Southwestern Medical Center seeks applications for a tenure-track faculty position in Dallas, TX at the Instructor, Assistant Professor, or Professor level in the area of metabolism and disease. Outstanding investigators at any academic rank will be considered. Candidates must have a Ph.D., M.D. or equivalent degrees, a track record of outstanding research, and the ability to direct an independently-funded research program. Areas of specific interest for this position include metabolomics, metabolic flux analysis, mitochondrial biology and other areas of metabolism relevant to human health and development. In addition to analytical equipment dedicated to the investigator’s studies, CRI members will also have access to a metabolomics core housing instrumentation for ultra-high pressure liquid chromatography, triple-quadrupole mass spectrometry and gas chromatography/mass spectrometry. NMR spectroscopy, 13C dynamic nuclear polarization, a human 7-Tesla MRI and a state-of-the-art mouse metabolic phenotyping facility are also available on campus to provide an unparalleled breadth of metabolic analysis. The UT-Southwestern Medical Center has a distinguished history of excellence in disease-related basic science research. The CRI is a new institute dedicated to recruiting outstanding scientists dedicated to solving fundamental problems in human disease and to providing a dynamic, stimulating and highly collaborative scientific environment. Major areas of focus within the CRI will include stem cell biology and cancer in addition to metabolism. Please submit a CV, a 2-page summary of past accomplishments and research plans, and ask three references to submit letters to Beth Morris at
[email protected]. CRI is a collaborative venture with Children’s Medical Center of Dallas
UT Southwestern is an Equal Opportunity/Affirmative Action Employer.
UT Southwestern is an Equal Opportunity/Affirmative Action Employer.
Excellence in Action
Kent State University
KENT Bioengineering Initiative
Multiple Faculty Positions
STATE
Bioengineering Initiative
Multiple Faculty Positions
Kent State University is undertaking an initiative to significantly enhance its capacity in cutting-edge bioengineering research and to create bioengineering programs. This initiative builds upon: our recent substantial investments in faculty and infrastructure in the life and health sciences; our founding of the new Kent State University College of Public Health; world-class psychology and biology research programs; our many collaborative research relationships with major medical institutions in Northeast Ohio; and the growing emphasis on medical devices and related advanced materials in Kent State’s nationally recognized Liquid Crystal Institute™. As an initial step in this initiative, Kent State is seeking to recruit several tenure-track faculty members, one at a senior level and up to three at a junior level. The person filling the senior position will be expected to take a leadership role in building the research and degree programs in this initiative and have considerable input into the planned recruitment of additional faculty. Bioengineering faculty will serve as a catalyst for increasing and enhancing biomedical research across the university and will play a central role in the translation of Kent State research from the laboratory to the clinic.
Current research strengths across the many disciplines at Kent State are particularly compatible with the bioengineering fields of tissue regeneration, biomaterials, biocompatibility, sensors and implanted devices, and for this reason, faculty with these research specialties are particularly encouraged to apply. However, those candidates with other bioengineering research interests will also be considered.
Qualifications: To be considered for the senior position, candidates are expected to have a Ph.D. in bioengineering or a closely related field, an internationally recognized research reputation, a record of successfully competing for a high level of extramural funding, and must have current extramural support. The university offers competitive salary and benefits, substantial start-up resources and the opportunity to drive the growth of this new program. To be considered for the junior positions, candidates are expected to have a Ph.D. in bioengineering or a closely related discipline and postdoctoral experience in one of the areas listed above. Candidates with current extramural funding or the clearly demonstrated ability to compete successfully for funding will be given the highest consideration. Review of applications will begin immediately and continue until the positions have been filled. However for full consideration, applications should be received by Jan. 15, 2012. Kent State University, Kent State and KSU are registered trademarks and may not be used without permission. Kent State University, an equal opportunity, affirmative action employer, is committed to attaining excellence through the recruitment and retention of a diverse workforce. 11-2615
For more information and to apply for these positions, please visit:
www.kent.edu/bioengineering
online @sciencecareers.org
TEMASEK RESEARCH FELLOWSHIP (TRF) The Nanyang Technological University (NTU) and the National University of Singapore (NUS) invite outstanding young researchers with a PhD Degree in science or technology to apply for the prestigious TRF awards.
LABORATORY SCIENTIST
The TRF scheme provides selected young researchers an opportunity to conduct and lead research that is relevant to defence. It offers:
– LONG ISLAND, NY – The Winthrop-University Hospital Research Institute is actively recruiting a PhD or MD laboratory scientist to join the current group of clinical and basic researchers studying bone and mineral metabolism.The scientist will direct the laboratory effort of the bone and mineral research program.The successful candidate will be capable of developing innovative programs of research focused on the causes and complications of osteoporosis, and other disorders of bone and mineral metabolism. It is expected that senior candidates will have an established track record and existing funding, and junior candidates will have a high degree of potential for becoming successful independent investigators. Appointments from Assistant to Full Professor will be made commensurate with the level of academic achievements.As the Nassau County clinical campus of Stony Brook School of Medicine, Winthrop provides a vibrant, interactive environment offering numerous opportunities for multidisciplinary collaborative investigations. Construction of a new research facility and a generous recruitment package including support for a team of researchers in clinical,basic,translational and outcomes research,create an exceptionally exciting and unique opportunity. Winthrop is located in a highly desirable geographic location within easy access to New York City and the beaches of Long Island. Please send letter of interest & curriculum vitae,to: John F.Aloia, MD, Chief Academic Officer. Email:
[email protected], or call for more information: 516-663-2442.
3-year research grant that commensurate with the scope of work, with an option to extend up to a further 3 years. Possible tenure-track academic appointment with the university at the end of the TRF. Attractive and competitive remuneration.
Fellows may lead and conduct research, and publish in these areas: 1. 2. 3. 4. 5.
Advanced Protective Materials Bio-mimetic Aerodynamics Cyber Security Sensemaking Technology Sensor Systems and Signal Processing
Other fundamental areas of science or technology, where a breakthrough would be of interest to defence and security, will also be considered. Singapore is a globally connected cosmopolitan city-state with a supportive environment and vibrant research culture. For more information and application procedure, please visit
Bone and Mineral Metabolism
NTU – http://www3.ntu.edu.sg/trf/index_trf.html NUS – http://www.nus.edu.sg/dpr/funding/trf.htm
WUH is committed to affirmative action, equal opportunity, and the diversity of its workforce. EOE-AA-m/f/d/v
Closing date: 10 February 2012 (Friday) Shortlisted candidates will be invited to Singapore to present their research plans, meet local researchers and identify potential collaborators in May 2012.
Institute for Cancer Care
V
I
S
I
Institute for Digestive Care
T
A
T
Institute for Family Care
:
W
Institute for Heart Care
W
W
.
W
Institute for Neurosciences
I
N
T
H
Institute for Lung Care
R
O
P
Institute for Specialty Care
.
O
R
G
Faculty Positions in Nanyang Technological University of Singapore (School Of Chemical and Biomedical Engineering) The School, comprising the Division of Chemical and Biomolecular Engineering and Division of Bioengineering invites applications for tenuretrack faculty appointments at the Assistant, Associate or Full Professor level in:
FACULTY POSITION BIOLOGY DEPARTMENT BOSTON COLLEGE The Boston College Biology Department seeks outstanding candidates for a tenure-track faculty position at the level of ASSISTANT PROFESSOR. Boston College provides competitive start-up funds and research space, with the expectation that the successful candidate will establish, or bring to the university, a vigorous, externally-funded research program. The successful applicant will have access to well-equipped animal facilities, core laboratories with state of the art instrumentation for flow cytometry and fluorescence microscopy and computational resources. We seek colleagues whose research focuses on microbe-host interactions and/or mechanisms of microbial pathogenesis to expand our existing program in microbiology. Special consideration will be given to candidates whose experimental or computational research program complements current faculty interests in cell biology, genetics, genomics, and bioinformatics (see http://www.bc.edu/biology for profiles of current faculty research programs). In addition, the successful candidate will be expected to train graduate students and participate in the undergraduate teaching mission of the Biology Department. This appointment will begin on or after July 1, 2012. Applicants should prepare a curriculum vitae, a statement of present and future research plans, and arrange for three letters of reference. All documents should be submitted as .pdf files to:
[email protected]. Applications received by January 17, 2012 are assured of full consideration. Review of applications will continue until the position is filled. Boston College is an Affirmative Action, Equal Opportunity Employer committed to improving diversity.
Chemical and Biomolecular Engineering Candidates must have at least a B.Sc. or B.Eng. in Chemical Engineering, a Ph.D. in Chemical Engineering or other closely related areas from reputable universities. Research expertise is required in one of the following areas: Electrochemical Engineering, Energy and Carbon Capture, Food Engineering, Control, Pharmaceutical Engineering, Printed Electronics, Synthetic Biology, and Microbial Engineering. Bioengineering Candidates must have at least a B.Sc. or B.Eng. in Life Sciences, Mechanical Engineering, Electrical Engineering, Computer Science/ Engineering or Bioengineering, a Ph.D. in Bioengineering or other closely related areas from reputable universities. Research expertise is required in one of the following areas: Microbiology, antibacterial drug discovery, pathogen resistance studies, antimicrobial agents and mechanism, and system biology. Creative and energetic individuals who show extraordinary promise or accomplishment in other related areas will also be considered. In addition to the qualifications and research expertise, candidates should have interests in undergraduate and graduate teaching, a track record of excellent publications and preferably successful grant applications. Details of the School can be found on http://www.scbe.ntu.edu.sg/Pages/ default.aspx APPLICATION PROCEDURE Qualified candidates are invited to submit an application. Guidelines for Submitting an Application for Faculty Appointment are available at http:// www.ntu.edu.sg/ohr/Career/SubmitApplications/Pages/Faculty.aspx . The post applied for should be clearly stated. Electronic submission of application is encouraged and can be forwarded to: Chairman, Search Committee School of Chemical and Biomedical Engineering NANYANG TECHNOLOGICAL UNIVERSITY E-mail:
[email protected] Website: www.scbe.ntu.edu.sg
online @sciencecareers.org
online @sciencecareers.org
SYRACUSE UNIVERSITY Assistant/Associate Professor- Biology (028464)
ASSOCIATE DEAN FOR GRADUATE STUDIES AND RESEARCH The State University of New York College of Optometry invites applications and nominations for the position of Associate Dean for Graduate Studies and Research. The Associate Dean directs the Graduate Center for Vision Research (GCVR), which is the administrative organization supporting the college’s graduate and research programs. We seek an active and innovative researcher and scholar to lead and work with faculty to provide vision and expertise around research and graduate programs aligned with the institution’s educational, research and patient care mission and goals. The college seeks to be a leader in basic, translational, and clinical research on the eye, visual neurophysiology, visual perception, and clinical optometric care. The college’s research faculty make up an active and growing group with research funding from NIH, NSF, DoD, foundations, and industry. The Associate Dean reports to the Vice President and Dean for Academic Affairs and is a member of the Dean’s Council. As director of the GCVR, the Associate Dean is responsible for the administration, evaluation, and development of the college’s graduate degree programs (M.S. and Ph.D. in Vision Research). The Associate Dean serves as the institution’s liaison for all matters related to research, is a member of the SUNY Eye Institute Steering Committee, and is responsible for all graduate faculty and research development. The Associate Dean will work closely and collaboratively with other department chairs around programs, research, and faculty development. The Associate Dean oversees the research support from the Optometric Center of New York for the Schnurmacher Institute of Vision Research, and all sources of intramural research funding. Additional responsibilities include administrative oversight of the Clinical Research Center, the Biological Research Facilities, and SUNY Research Foundation operations at the college. The successful candidate will have an O.D., Ph.D., or both with a strong background in vision research and experience in graduate education in optometry, ophthalmology, or vision science. The candidate will be expected to build the graduate programs and both basic and clinical research programs at the college. The successful candidate will also receive a faculty appointment; tenure and rank will be determined by experience. Candidates with an active research program are desirable. Applicants should send a letter of interest, CV, the names and contact information of three references, and other supporting material to: Associate Dean for Graduate Studies and Research Search Committee, C/O Jean Pak, Office of Academic Affairs, SUNY College of Optometry, 33 West 42nd St, New York, NY 10036 or by email to Jean Pak, Academic Programs Coordinator,
[email protected]. Review of applications will begin immediately and will continue until the position is filled. The State University of New York College of Optometry is an Affirmative Action, Equal Opportunity Employer.
The Department of Biology at Syracuse University (SU) invites applications for a tenure-track Assistant/Associate Professor position to support ongoing interests in epigenetics, chromatin, and small RNA biology at SU. We seek applicants who utilize biochemical, genetic, and/or genomic approaches to address chromatin- and/or small RNA-based mechanisms of epigenetic regulation within a developmental context such as, but not limited to, stem cell biology, cell-signaling, or cellular differentiation. The successful candidate will occupy space in the Life Sciences Complex, a new research facility designed to support collaborative research. This position is part of the epigenetics research focus at SU. The successful candidate is expected to develop an independent, extramurally funded research program and will be expected to interact effectively with colleagues in Biology as well as with colleagues from other departments at SU, SUNY-Upstate Medical University, and SUNY-College of Environmental Science and Forestry. The successful candidate will also be expected to teach undergraduate and graduate courses and develop new courses as appropriate to his/her expertise and the needs of the Department of Biology. Competitive salary, start-up funds, and laboratory space will be provided. Candidates must have a PhD in any area of biology relevant to this search and productive postdoctoral research experience. For full consideration applicants must complete an online application at www.sujobopps.com, (#028464) and attach the following documents. Please attach documents as follows: FILE 1 - a cover letter outlining the candidate's qualifications, a 2-3 page statement of research experience, interests and philosophy, a curriculum vitae, and contact information for three professional references to provide letters of recommendation. FILE 2 - recent publication #1 FILE 3 - recent publication #2 Review of applications will begin December 5, 2011. For questions, please e-mail Eleanor Maine, Chair of the Search Committee,
[email protected]. Syracuse University is an Affirmative Action/Equal Opportunity Employer; qualified women and minority candidates are especially encouraged to apply.
UIC
FACULTY POSITIONS DEPARTMENT OF MICROBIOLOGY AND IMMUNOLOGY The Department of Microbiology and Immunology in the College of Medicine at the University of Illinois at Chicago (UIC) is seeking to fill two tenured/tenure track faculty positions at the level of Assistant, Associate, or Full Professor. Candidates with research interests encompassing areas relevant to obesity and/or diabetes (e.g. microbiomes, virus induced fatty liver and inflammation), to infection and immunology of the lung, or to the immunology of transplantation and autoimmunity are especially encouraged to apply. Each successful faculty candidate is expected to maintain a vigorous independent research program and participate actively in the teaching, research, and graduate training programs in the department. Generous laboratory space and start-up funds are available. Applicants are required to have a Ph.D., M.D. or equivalent doctorate level degree, and a proven track record in research, as evidenced by consistent scholarly publications and extramural peer reviewed grant funding. UIC is the largest institution of higher learning in the Chicago area and is a major center for research and education. UIC’s College of Medicine is part of the Illinois Medical District, the largest complex of medical centers in the United States. The Department of Microbiology and Immunology occupies over 30,000 square feet of renovated or new space in the recently completed college of medicine research building. The faculty members in the department have active and interdisciplinary research programs in cellular and molecular immunology, microbial pathogenesis, host-pathogen interactions, virology, and structural biology. The department also maintains active collaborations with colleagues within the UIC Colleges of Dentistry, Pharmacy, and Liberal Arts and Sciences. Applicants must apply online at https://jobs.uic.edu/. Required application materials are: current curriculum vitae, bibliography, statement of current research interests, and the names and contact information for three references. Review of applicants will begin December 5, 2011 and continue until the positions are filled. For more information about the Department of Microbiology and Immunology, please visit our Web Site: http://www.uic.edu/depts/ mcmi/index.htm. The University of Illinois at Chicago is an Affirmative Action/Equal Opportunity Employer. Women and minorities are strongly encouraged to apply.
MICROBIOLOGY/IMMUNOLOGY FACULTY POSITIONS The Department of Biological Sciences, College of Sciences, Old Dominion University, invites applications for several tenure track/tenured Faculty positions (beginning July 2012) at the Assistant Professor, Associate Professor, or Professor level as part of a major recruitment initiative. We are particularly interested in: (1) investigators in host-pathogen interactions and molecular pathogenesis; (2) molecular or cellular immunologists employing contemporary molecular biology approaches. A successful candidate must establish and maintain a vigorous research program that will attract peer-reviewed funding, and provide excellent education to our graduate and undergraduate students. All applicants must have a Ph.D., M.D. or an equivalent degree in an appropriate field. Applicants at the Associate Professor or Professor level must demonstrate substantial research accomplishments, a consistent record of independent peer-reviewed funding, and have active competitive grants. Interactions are encouraged with other departments/units of the College and the University, as well as the Eastern Virginia Medical School. A cross-appointment with the Center for Molecular Medicine (Chris D. Platsoucas, Ph.D., Center Director) is available. State salary support and competitive start-up packages are available. The Department of Biological Sciences receives substantial support from state funds, as well as from research grants from federal and other granting agencies. The department has strong Ph.D. and M.S. graduate programs that currently enroll over 125 students. The College of Sciences is undergoing a major research expansion. Over the last three years research grant awards to the College have increased by 78% to $20.6 million in FY2010. Old Dominion University (www.odu.edu) is a state supported, Carnegie doctoral research extensive institution enrolling more than 24,000 students including 6,000 graduate students. Interested individuals should submit curriculum vitae, a statement of research achievements and research plans, and the names, addresses, e-mail addresses and phone numbers of three references to Wayne Hynes, Ph.D., Professor and Chairman, Department of Biological Sciences, Old Dominion University at
[email protected]. Review of applicants will begin immediately and continue until the positions are filled. Old Dominion University is an affirmative action, equal opportunity institution and requires compliance with the Immigration Reform and Control Act of 1986.
The University invites applications for four tenure track Assistant Professor positions in RNA science and technology. CELL/DEVELOPMENTAL BIOLOGIST, Department of Biological Sciences: conducting research on the role of RNA, including but not exclusive to non-coding or microRNA molecules, in post-transcriptional gene regulation or other cellular and/or developmental processes. CHEMIST/BIOCHEMIST, Department of Chemistry: conducting research in RNA science and its applications in areas such as, but not limited to, modified nucleosides, synthesis, imaging, and analytical chemistries as it pertains to RNA structure/function relationships, including interactions with proteins and other RNAs. RNA VIROLOGIST, Department of Biological Sciences: conducting research with mammalian RNA viruses in any one or more areas including but not exclusively: genome structure, mechanisms of genome replication and packaging, gene expression and regulation, host-range and cell specificity, evolution of emerging diseases. BIOINFORMATICIST, Department of Chemistry: conducting research on the structure/function of RNA molecules through both computational and wet-lab research. or MACROMOLECULAR X-RAY CRYSTALLOGRAPHER, Department of Chemistry: conducting research on solving nucleic acid/protein structures particularly that of RNA. The successful candidate will be expected to collaborate with RNA Institute researchers studying RNA structure and RNA-protein interactions. The Department plans to add researchers in both areas over the next two years and will fill the position this year with the best candidate. All positions will be affiliated with the RNA Institute (http://www.albany.edu/rna) with state-of-the-art laboratories housed in the Life Sciences Research Building (http://www.albany.edu/lifesciences). The Institute brings together more than 35 investigators from the College of Arts & Sciences, the College of Nanoscale Science and Engineering, the School of Public Health, and regional institutions including the Wadsworth Center, Rensselaer Polytechnic Institute, and Albany Medical College. This creates an outstanding environment for research collaborations. Instructional responsibilities will be consistent with the position and those of the faculty in the home department, and the interests of the candidate. Submit applications for CELL/DEVELOPMENTAL BIOLOGY at: http://albany.interviewexchange.com/jobofferdetails.jsp?JOBID=27938 Submit applications for CHEMIST/BIOCHEMIST at: http://albany.interviewexchange.com/jobofferdetails.jsp?JOBID=27907 Submit applications for RNA VIROLOGIST at: https://albany.interviewexchange.com/jobofferdetails.jsp?JOBID=28449 Submit applications for BIOINFORMATICIST or MACROMOLECULAR X-RAY CRYSTALLOGRAPHER at: https://albany.interviewexchange.com/jobofferdetails.jsp?JOBID=28447 Applications must include a CV with publications cited in detail and any present or past grant funding, statement of research interests, statement of teaching interests, and a minimum of three letters of reference as directed by the above websites. The successful candidates for both positions will be offered a competitive salary, start-up package, and research space in the Life Sciences Research Building. Qualifications for all Candidates: Ph.D. from a college or university accredited by the U.S. Department of Education or an internationally recognized accrediting organization and a strong publication record reflecting significant scientific accomplishments. Applicants must address in their applications their ability to work with and instruct a culturally diverse population. Preferred qualifications include productive post-doctoral training and the potential or demonstrated ability, to obtain independent extramural funding. Review of applications will begin on December 15, 2011 and continue until the positions are filled. The University at Albany is an EEO/AA/IRCA/ADA employer.
Faculty and Post Doctoral Positions Staff Scientist Faculty Position The Barrow Neurological Institute seeks new faculty members for a newly-initiated Brain Tumor Research Center to develop high-impact and sustainable research programs that advance our understanding of pediatric and adult brain tumors. Appointments may be at the Assistant, Associate or Full Professor level. Membership to the appropriate basic science or clinical department, appointment to graduate programs, excellent space, ongoing partial salary support, and startup funding will be provided. An advanced degree and substantial research background are required. The Barrow Neurological Institute, located in Phoenix, Arizona, is one of the busiest clinical neuroscience centers in the world and is ranked by US News & World Report as one of the top 10 Neurology & Neurosurgery programs in the country. Applicants may apply in any area of brain tumor research, although strategic interests include molecular neuro-oncology, epigenetics, neuro-epidemiology, stem cell biology, neurofibromatosis, immunotherapy, and developmental neurobiology. Please send CV, summary of current research (one page), outline of future research (two pages), and 3 letters of recommendations to Debbie Nagelhout (debbie.nagelhout@bnaneuro .net), academic assistant to Nader Sanai, M.D., Director, Barrow Brain Tumor Research Center, Barrow Neurological Institute. Applications will be accepted until February 1, 2011 and apply online to the Staff Scientist position at http://www.chwcareers.org/Facilities/StJosephs-Hospital-and-Medical-Center-AZ/Careers/index.htm. Postdoctoral Fellow Positions The Barrow Neurological Institute has a POSTDOCTORAL FELLOW positions available in laboratory of Nader Sanai, M.D. at the Barrow Neurological Institute (Phoenix, AZ). Our laboratory investigates the neurobiological basis of neural and glial CNS progenitors (see Nature; 2004: Feb 19 and Nature; 2011: Oct 20), particularly in the context of brain tumors. We use molecular biology, histopathology, biochemistry, tissue culture assays and mouse models to characterize stem and progenitor cells. Our work also extends into molecular imaging of neural precursors, applying high field-strength magnetic resonance spectroscopy through the Arizona State University Magnetic Resonance Research Center. The Barrow Neurological Institute is a translationally-oriented basic science research facility located in Phoenix, Arizona and located adjacent to the state’s largest tertiary care hospital. It is in an urban center of academic, biotechnology, and pharmaceutical research, neighboring Arizona State University and the University of Arizona School Of Medicine. The Sanai Laboratory is a member of the Barrow Brain Tumor Research Center. Initial appointment is for two years with the possibility of renewal. Interested candidates should e-mail a statement of research interests, CV and the names and contact information for three references to the e-mail address to
[email protected] and apply online at http://www.chwcareers.org. The Barrow Neurological Institute is seeking a POSTDOCTORAL FELLOW to join the bioengineering laboratory of Rachael W. Sirianni, Ph.D. Our primary goal is to engineer multifunctional nanocarriers for targeted drug delivery to brain tumors. Qualified applicants will hold a Ph.D. in a relevant biomedical discipline with expertise in one or more of the following areas: targeted drug delivery, biomaterials, primary culture of human neuronal cells, blood-brain barrier physiology, and molecular imaging. Experience with molecular biology or in vivo models of CNS disease is preferred but not necessary. Interested candidates should send a cover letter, a summary of research interests, curriculum vitae, and the names and contact information for three references to rachael.siri
[email protected]. Applications will also be accepted online, at http: //www.chwcareers.org. Affirmative Action/Equal Opportunity Employer.
online @sciencecareers.org
Four Faculty Positions in RNA Research
Concurso abierto
Tenure-Track Assistant Professor in Biology with Emphasis on Biological Science Education
The Department of Biology invites applications for a tenure-track assistant professor in Biology Science Education to join the faculty in August 2012. Requirements for the position include a Ph.D. in a biological discipline, relevant post-doctoral experience in the teaching and learning of science and a commitment to excellence in STEM education research and teaching. The ideal candidate will have a primary research focus in science education in the biological sciences. The candidate will be expected to: (1) deliver and refine the biology curriculum in the university’s new Common Academic Program (CAP); (2) participate in the collaborative efforts of CAP science education with colleagues in biology and other STEM disciplines and professional organizations; (3) engage in research, writing and other scholarly activities that contribute to discipline-based education research, and (4) develop an extramurally funded research program. Teaching expectations include courses at the undergraduate and graduate level appropriate to the individual’s expertise. Preference will be given to those candidates with classroom teaching experience at the postsecondary level and a record of publication in the teaching and learning of science. We also prefer a qualified candidate who has experience in teaching and advising students from diverse backgrounds. In addition, the faculty hire will be expected to participate in the outside speaker program (BIO 501) and in the graduate special topics journal club course (BIO 601), and supervise an occasional undergraduate seminar course (BIO 299 or BIO 420). This individual will also serve as an academic advisor to undergraduate students. The University of Dayton, founded in 1850 by the Society of Mary, is a top ten Catholic research university. The University seeks outstanding, diverse faculty and staff who value its mission and share its commitment to academic excellence in teaching, research and artistic creativity, the development of the whole person, and leadership and service in the local and global community.
Application deadline is December 31, 2011. To attain its Catholic and Marianist mission, the University is committed to the principles of diversity, inclusion and affirmative action and to equal opportunity policies and practices. We act affirmatively to recruit and hire women, traditionally under-represented minority groups, people with disabilities and veterans.
The Department of Physiology and Biophysics, Faculty of Medicine, invites applications for a position at the ASSISTANT PROFESSOR level available for the 2012 academic year. Qualified candidates will have demonstrated research expertise in CARDIOVASCULAR physiology, with an emphasis on molecular or cellular research in cardiovascular physiology/pathophysiology. The candidate is expected to develop an extramurally funded research program and to participate in the teaching mission of the Department. The successful candidate will have ample opportunity to develop active and synergistic research collaborations with scientists in the Cardiovascular Research Group consisting of basic and clinician scientists from a number of Departments in the Faculty of Medicine at Dalhousie University. Current research strengths in this group exist in the areas of ion channel physiology, regulation of the cardiovascular system and molecular mechanisms of cardiovascular disease. Applicants must have a Ph.D. and/or M.D. degree, several years of postdoctoral training, excellent communication skills, and a strong record of peer-reviewed publications. This is a salaried, probationary tenure-track appointment; however, the candidate will be expected to apply for external salary support from appropriate granting agencies. Interested applicants should submit a curriculum vitae along with a brief description of research experience and interests as well as teaching experience and interests by email to
[email protected]. Applicants should also arrange to have three letters of reference sent directly to: Dr. Paul R. Murphy, Head, Department of Physiology and Biophysics, Faculty of Medicine, Dalhousie University, Sir Charles Tupper Medical Building, 5850 College Street, Halifax, Nova Scotia B3H 4R2, Canada. Review of applications will begin in January, 2012. To guarantee consideration, your application should be submitted by January 31, 2012. All qualified candidates are encouraged to apply; however, Canadians and permanent residents will be given priority. Dalhousie University is an Employment Equity/Affirmative Action employer. The University encourages applications from qualified Aboriginal people, persons with a disability, racially visible persons and women.
SE BUSCA Personas con título de grado y posgrado en áreas disciplinares afines, menores a sesenta (60) años, residentes en el país y/o en el exterior, con antecedentes de idoneidad y capacidad disciplinar suficiente y con experiencia en dirección y gestión universitaria de unidades pertenecientes a Instituciones de Educación Superior. El Instituto de Investigación e Ingeniería Ambiental es un ámbito de investigación académica y desarrollo tecnológico que se propone brindar respuestas adecuadas a las problemáticas ambientales de la sociedad. Para obtener más información sobre condiciones ofrecidas y plazos de presentación escribir a:
[email protected]
To apply for this position please visit this site: http://jobs.udayton.edu/applicants/Central?quickFind=52616
Dalhousie University Halifax, Nova Scotia Assistant Professor Position Cardiovascular Physiology
Decano del Instituto de Investigación e Ingeniería Ambiental (3iA)
DIRECCIÓN DE DISEÑO
online @sciencecareers.org
UNIVERSITY of
3iA
Instituto de Investigación e Ingeniería Ambiental L A U N I V E R S I D A D P Ú B L I C A M E T R O P O L I TA N A
FACULTY POSITION Department of Anatomy and Cell Biology The Department of Anatomy and Cell Biology (ACB) at the University of Pennsylvania School of Dental Medicine invites applications for a full-time tenure-track or tenured faculty position at the Assistant Professor, Associate Professor or Professor level. We are seeking faculty with outstanding academic accomplishments in cellular and/or molecular biology complementing our departmental research strengths, current NIH grant funding and ability to teach gross anatomy and/ or histology/embryology to first year dental students. The ACB Department faculty currently investigate basic and translational approaches to skeletal and smooth muscle physiology, mechanobiology, extracellular matrix synthesis in fibrosis, mechanisms of lysosomal signaling and physiology, and craniofacial development. Applicants with a Ph.D. or dual degree (DMD-Ph.D., MD-Ph.D., DVMPh.D.) are invited to submit a statement of research and teaching interests, curriculum vitae, and names with contact information for three references. Review of applications will begin January 1, 2012, and continue until the position is filled. Applicants can apply directly at the University of Pennsylvania website: https://facultysearches.provost.upenn.edu/ applicants/jsp/shared/frameset/Frameset.jsp?time=1320782626017 or by submitting materials to the Chair of the Search Committee: Carolyn W. Gibson, Ph.D. Dept. of Anatomy and Cell Biology School of Dental Medicine University of Pennsylvania 240 S. 40th Street Philadelphia, PA 19104 E-mail:
[email protected] The University of Pennsylvania is an Equal Opportunity Affirmative Action Employer; women and minority candidates are strongly encouraged to apply.
Biotechnology and Biomedicine Research Centre of the Academy of Sciences and the Charles University at Vestec (BIOCEV) www.biocev.eu | R&D focus: Biotechnology; Biomedicine | position sought: •BIOCEV Director Central European Institute of Technology (CEITEC) www.ceitec.eu | R&D focus: Life sciences; Advanced materials and technologies |positions sought: • Scientific Director • Executive Director ELI: Extreme Light Infrastructure – Beamlines facility www.eli-beams.eu | R&D focus: Ultra high intensity laser | positions sought: • ELI Beamlines Project Director • Technical Director of the ELI Beamlines facility ELI: Extreme Light Infrastructure – Delivery Consortium (ELI DC): Consortium of three ELI laser facilities: ELI Beamlines (in the Czech Republic), ELI Attosecond (in Hungary) and ELI Nuclear Physics (in Romania) www.extreme-light-infrastructure.eu | R&D focus: Ultra high intensity laser | positions sought: • ELI DC Director General • ELI DC Scientific and Technical Director St. Anne’s University Hospital Brno - International Clinical Research Center (FNUSA-ICRC) www.fnusa-icrc.org/en | R&D focus: Clinical research - cardiovascular and neurological diseases | positions sought: • Chair of the International Clinical Research Center • Director for Strategic Interna¬tional partnerships IT4Innovations (IT4I) www.it4i.eu |R&D focus: Information technology; Supercomputing | positions sought: • Managing Director of IT4I • Head of the Supercomputing Centre
For further details regarding any particular R&D center and the respective positions, please visit the following webpage: → www.msmt.cz/strukturalni-fondy/search-committees ← Interested candidates are welcome to send a letter of interest, including a brief description of research and leadership experience, CV and bibliography to:
[email protected] until November 28th 2011. For informal enquiries about any aspect of the post, please contact relevant contact person stated at the end of each individual advertisement that can be found on the webpage above. All applications will be assessed by a special selection board (so called Search Committee), consisting of both international and Czech R&D experts and representatives of the particular center, that shall meet under the coordination and supervision of the Managing Authority of the OP RDI in December 2011. Top selected candidates will be invited for interviews taking place in January and February 2012 in the Czech Republic (concrete dates to be communicated at a later stage).
BME/DGHI Joint Faculty Position
The Department of Biomedical Engineering in the Pratt School of Engineering at Duke University invites applications for a tenure track faculty position in medical imaging. The assistant professor level is preferred but all levels will be considered. Candidates should hold a Ph.D. or equivalent in biomedical engineering or a related field. Applications are sought in the field of medical ultrasound including signal processing, molecular imaging, transducers and ultrasound therapy. Applications are also sought in the fields of photo-acoustics and foundational medical image processing. Candidates should have the necessary experience and skills to teach undergraduate courses in biomedical electronics and imaging. Applicants should electronically submit: a curriculum vitae, names and contact information for three references, a teaching and research statement, and expected date of availability to the following website: https://academicjobsonline.org/ ajo/jobs/1173 Applications received by December 31, 2011 will be given consideration. Duke University is an Affirmative Action/ Equal Opportunity Employer.
The Department of Biomedical Engineering (BME) at Duke University and the Duke Global Health Institute (DGHI) seek a faculty candidate who is committed and will lead in development of bioengineering tools for health applications in low and middle income countries. We are particularly interested in candidates who are developing engineering approaches to address global health disparities, for example as related to the diagnosis and/or treatment of infectious or non-communicable diseases in resourcelimited settings. Specific areas of engineering science and technology may include, for example, nanotechnology and biomaterials, gene and drug delivery, imaging, genomic technologies, bioanalytical systems, and biomolecular modeling, all with an emphasis on field applications. Rank is open and will depend upon the qualifications of the candidate. We seek applicants with an affinity and aptitude for collaboration, and a research and teaching focus that is cross-disciplinary. Candidates must have a doctorate in biomedical engineering, or a related field of science or engineering, and an outstanding record of accomplishment. The successful candidate will be expected to develop an active externally funded research program, to build strong ties between DGHI and BME, initiate collaborative research with other faculty at Duke University and its Medical Center, and have a strong commitment to teaching at the undergraduate and graduate levels. Evidence of prior work in global health applications of biomedical engineering will be a substantial advantage for the successful candidate. The application materials must include a curriculum vitae (with address, phone number, and e-mail address); statements of research and teaching interests; and names, addresses, phone numbers, and e-mail addresses of three referring individuals, who will provide letters of reference at the time of the submitted application. In a separate document, please provide a summary of the following: (1) how your research interests interface with the development of bioengineering tools for addressing critical health challenges in the developing world; (2) your strategy for developing strong research ties between the Duke GHI and BME Department; and (3) your plans for developing a course that applies engineering tools to both specific and broad aspects of global health. Please note that preference will be given to candidates who clearly indicate the impact that their research and teaching has had, or will have, on health problems facing populations in low and middle income countries. For full consideration, please submit your application by December 31, 2011. To apply please submit your electronic application to: https://academicjobsonline.org/ajo/jobs/1215. Duke University is an Affirmative Action/Equal Opportunity Employer that is committed to increasing the cultural and intellectual diversity of its faculty. Applications from women and underrepresented minority groups are strongly encouraged.
online @sciencecareers.org
The Operational Programme Research and Development for Innovation (OP RDI) - major R&D programme in the Czech Republic co-funded by the European Structural Funds and the Ministry of Education, Youth and Sports of the Czech Republic - is inviting applications for positions in top scientific and executive management of five largest and most prestigious newly emerging R&D European Centers of Excellence:
online @sciencecareers.org
Picture yourself as a AAAS Science & Technology Policy Fellow Make a Difference. Help give science a greater voice in Washington, DC! Since 1973, AAAS Fellows have applied their skills to federal decision-making processes that affect people in the U.S. and around the world, while learning first-hand about the government and policymaking. Join the Network. Year-long fellowships are available in the U.S. Congress and federal agencies. Applicants must hold a PhD or equivalent doctoral-level degree in any behavioral/social, biological, computational/mathematical, earth, medical/health, or physical science, or any engineering discipline. Individuals with a master’s degree in engineering and three years of post-degree professional experience also may apply. Federal employees are not eligible and U.S. citizenship is required. Apply. The application deadline for the 2012-2013 AAAS Fellowships is 5 December. Fellowships are awarded in the spring and begin in September. Stipends range from $74,000 to $97,000. Note: Additional fellowships are available through approximately 30 scientific society partners. Individuals are encouraged to apply with AAAS as well as with any scientific societies for which they qualify.
Full details at: fellowships.aaas.org
Enhancing Public Policy, Advancing Science Careers Kathy Kahn, PhD Interdisciplinary Biological Sciences, University of Missouri 2004-06 AAAS Fellow at the U.S. Department of Agriculture, Biotechnology Group in the Foreign Agricultural Service Now a program officer at the Bill and Melinda Gates Foundation, Agriculture Development Group
THE DEPARTMENT OF MEDICINE AT YALE UNIVERSITY SCHOOL OF MEDICINE is seeking faculty at the Assistant or Associate Professor level with independent research programs in applied translational research. Innovative multi-disciplinary research involving the development of biomarkers for personalized medicine, therapeutic clinical trials or human-based studies of disease pathogenesis is encouraged. The sectional appointment within the Department will depend upon the scientific and clinical expertise of the faculty member, who will be expected to participate in teaching/mentoring of students and/or clinical care. Office space and a generous start-up package will be provided by the Department’s newly developed Program of Translational Research and laboratory space (if needed) will be provided by the section. Applicants should have an M.D., Ph.D., or M.D./ Ph.D., and demonstrated success in competing for extramural funding. Candidates will be evaluated based on their publications in translational research, potential to maintain an externally funded research program, and the degree to which their research intersects with other members of the program. Please reply with CV, cover letter including a description of the research program, and the names of three references. Review of applications will commence on December 15, 2011 and will continue until a successful candidate is identified. For additional information and inquiries, and to submit an application, please contact: Director, Program of Applied Translational Research Yale University School of Medicine P.O. Box 208029 333 Cedar Street New Haven, CT 06520-8029 E-mail c/o
[email protected] Yale University is an Affirmative Action/Equal Opportunity Employer.
Founded in 1897, Zhejiang University is committed to the highest standards of excellence in education and research, and has been at the forefront of academic leadership in China. The College of Life Sciences with rich history (see http://www.cls.zju.edu.cn) aspires to become a top academy for life science education and research. Multiple positions for Associate Professor and Professor are open in following areas: Biochemistry and Molecular Biology Stem cell research and cell Biology Developmental Biology Genetics Microbiology,Pathobiology and Immunology Plant Sciences Ecology and Evolution Biology Marine Biology Bioinformatics Applicants should have strong research profile and potential capacity to conduct innovative research in these areas, and are expected to teach both undergraduate and graduate courses. The University and the College provide state-of-the-art research facilities and strong supporting staffs. Competitive start-up support, salary and benefits will be offered according to individual qualification and experience. Postdoctoral positions are also available with very competitive fellowship. Please submit your full CV, the cover letter, a list of three references and your future plan to Dr. Qin Lu at
[email protected] or call 86-(571)88206483. The positions will be open until they are filled by appropriate candidates.
DIRECTOR WWAMI School of Medical Education University of Alaska Anchorage University of Washington School of Medicine The University of Alaska Anchorage (UAA) is seeking a Director for the WWAMI (Washington, Wyoming, Alaska, Montana, Idaho) School of Medical Education. The position includes appointment as Assistant Dean at the University of Washington School of Medicine (UWSOM). The largest of three universities in the University of Alaska system and its designated health campus, UAA is located in the University/Medical District of Anchorage. The School of Medical Education is an academic unit within the newly formed College of Health, housed in the Health Sciences Building that opened in summer 2011. As the comprehensive urban university in Alaska, UAA serves more than 20,000 students on the main campus and four community campuses. Nestled between the ocean and snow-capped peaks, Anchorage’s 290,000 residents enjoy a metropolitan lifestyle balanced with year-round outdoor recreational opportunities in a pristine environment, with easy access to some of America’s largest national forests and parks. The Director is primarily responsible for assuring the excellence of the medical education program, promoting research and other scholarly activities of the faculty, and managing the School’s budget including general fund and sponsored programs. Other responsibilities include: assisting the UWSOM and the other WWAMI sites to develop and deliver quality medical education throughout the region; leading the expansion of medical education in Alaska; promoting the goals and activities of the School of Medical Education to municipal and state agencies, and to the general public; leading development of strategic alliances in the University/Medical district; facilitating growth of biomedical research in the School and of biomedical health sciences at UAA and in the University of Alaska system; and giving leadership to the growth of educational pathway programs to guide Alaska’s K-16 students toward careers in medicine and biomedical research. Candidates must hold an M.D., Ph.D., or equivalent in a medically relevant discipline; have experience in medical education, or administrative duties in a medical or academic setting; and have a record of research, teaching, or other professional activity sufficient for appointment as a senior member of the faculty. The Director will have the opportunity to continue an active research and graduate education program. Preference will be given to candidates with: academic administrative experience, including program planning, curriculum development, budget management, and supervision of faculty, staff and students; teaching experience of medical students in the basic science areas; biomedical research experience in basic, translational or clinical areas, including external funding; and experience working with diverse groups including state and federal agencies, healthcare providers, and/or community organizations. Salary for this 12-month appointment will be commensurate with experience, and compensation includes a competitive benefits package. The candidate should qualify for a senior faculty rank within his or her discipline at the time of appointment to be eligible for tenure at UAA. The Director reports to the Dean of the UAA College of Health; as an Assistant Dean in the UWSOM, the Director reports to the Vice Dean for Academic Affairs and works closely with the Vice Dean for Regional Affairs. The Director will have an affiliate academic appointment in an appropriate department of the UWSOM but will not accrue tenure rights in that school. To apply, see UAA Quick Link: www.uakjobs.com/applicants/ Central?quickFind=75731. PCN: 301225. Review of applications begins November 28, 2011; search will remain open until the position is filled. Start date is negotiable. For further information contact Dr. Robert Furilla, Interim Director, (907) 786-4789;
[email protected]. UAA is an AA/EO Employer and Educational Institution. Applicants must be eligible for employment under the Immigration Reform and Control Act of 1986 and subsequent amendments. Your application for employment with UAA is subject to public disclosure.
online @sciencecareers.org
Applied Translational Research
online @sciencecareers.org
POSITIONS OPEN
The Department of Biological Sciences (website: http://biology.uark.edu) at the University of Arkansas solicits applications for a tenure-track ASSISTANT PROFESSOR working in any area of microbiology (Position #Y12926). The successful candidate will have a Ph.D., postdoctoral experience, and will be expected to establish an extramurally supported research program, supervise graduate and undergraduate research, and teach at the graduate and undergraduate levels, including a core course in ecology, evolution, genetics, or cell biology. Review of completed applications will begin December 10, 2011, and will continue until the position is filled. Applications should include curriculum vitae, a statement of current and future research plans, teaching interests, and three letters of recommendation. Application materials should be sent or electronically sent to: Dr. Mack Ivey (e-mail:
[email protected]), Department of Biological Sciences, 601 SCEN, 1 University of Arkansas, Fayetteville, AR 72701. The University of Arkansas is an Equal Opportunity/Affirmative Action Employer. Applicants must have proof of legal authority to work in the United States at the time of hire. All applicants are subject to public disclosure under the Arkansas Freedom of Information Act.
NORTHEAST TERRITORY MANAGER Due to our continuing growth, the DiaPharma Group, an Ohio-based provider of hemostasis research and diagnostic products, requires a sales representative to cover the Northeast region of the United States. Our products are utilized by industry, reference, research, and specialty hospital laboratories in the blood coagulation field. Requirements include: B.S./M.S. degree in Biochemistry with strong communication and organizational skills. Successful laboratory sales track record with recent hemostasis/coagulation instruments/reagents a must! Competitive benefits package includes medical, dental, and vision insurance coverage and a 401(k) plan. Extensive travel required within the seven state region. If you meet the above requirements, please electronically send your resume to e-mail: employment@ diapharma.com.
ELINGS PRIZE FELLOWSHIPS in Experimental Science The California Nanosystems Institute (CNSI) at the University of California–Santa Barbara is pleased to announce the 2011–12 competition for the Elings Prize Fellowships in Experimental Science. The Elings Prize POSTDOCTORAL FELLOWSHIPS provide a salary of $60,000/year for two years, renewable for a third, with benefits and research funds. Applicants must indicate with which CNSI-associated experimental group(s) they will pursue research. See website: http://www.cnsi.ucsb.edu/fellowships. Applicants holding a Ph.D. in science or engineering should submit a cover letter, curriculum vitae, a onepage research proposal, and arrange for three supporting letters, all submitted via the website: https://fellow. cnsi.ucsb.edu by January 31, 2012. The University of California is an Equal Opportunity/Affirmative Action Employer.
1016
POSITIONS OPEN
POSITIONS OPEN
MICROBIOLOGY FACULTY POSITION Req. #01077 The Department of Microbiology at the University of Chicago invites applications for a tenure-track faculty position at the rank of ASSISTANT PROFESSOR, although, depending on qualifications, candidates may be proposed for a senior appointment. Applicants must have a Ph.D. or M.D.-Ph.D. and relevant postdoctoral training. The successful candidate is expected to develop an extramurally supported research program focusing on host-pathogen interactions of viral, parasitic, fungal, or bacterial infectious agents. Candidates are also expected to contribute to departmental teaching. The University of Chicago maintains extensive core facilities in support of microbiological research including facilities for experiments with gnotobiotic animals and risk group 1–3 infectious agents. Competitive salaries and startup packages will be provided. Review of applications will begin on January 1, 2012 and continue until the position is filled. Interested applicants should submit a cover letter, curriculum vitae, the names and contact information for at least three references, and a statement of research interests emphasizing career goals at the following website: http:// academiccareers.uchicago.edu/applicants/ Central?quickFind=51966. The University of Chicago is an Affirmative Action/ Equal Opportunity Employer. Website: http:// tinyurl.com/6qp5llo.
FACULTY POSITION in Microbial Genomics at The University of Texas at San Antonio The Department of Biology and the South Texas Center for Emerging Infectious Diseases (STCEID) at the University of Texas at San Antonio (UTSA) are seeking outstanding candidates for a tenure-track faculty position in microbial pathogenesis with an emphasis on genomics. Rank is at the ASSISTANT PROFESSOR level. This position will join the microbiology/immunology research group focused on aspects of bacterial, fungal, and viral infections and host response. For further details and information on how to apply, please see website: http://scjobs.sciencemag. org/JobSeekerX/ViewJob.asp?cjid=77051& accountno=1316.
ASSISTANT PROFESSOR The College of Arts and Sciences at the University of Miami seeks applicants and nominations for four tenure-track appointments in Complexity Science, including complex networks and complex systems, from disciplines across the physical, biological, computational, mathematical, statistical, medical, economic, and social sciences. Successful candidates will join a vigorous research program already under way across departments and schools within the university. Appointments may be made at any level and joint appointments are possible between different departments and schools. Candidates should have a Ph.D. in a suitable quantitative discipline, and have experience in analyzing real-world data as well as mathematical and computational modeling. To be considered at ASSOCIATE or FULL PROFESSOR level the candidate must have independent teaching and research experience. The successful candidate will be expected to teach at both undergraduate and graduate levels and to develop and maintain internationally recognized research. Applicants should forward curriculum vitae, research plan, and a statement of teaching philosophy as well as arrange to have three letters of recommendation sent to the Chair of the Search Committee. All materials must be sent electronically to e-mail: complexitystudies@ miami.edu. Review of applications will begin November 15, 2011 and continue until the positions are filled. Information about the College can be found at website: http://www.as.miami. edu/. The University of Miami is an Affirmative Action/ Equal Opportunity University that values diversity and has progressive work-life policies. Women, persons with disabilities, and members of other underrepresented groups are encouraged to apply. Position #043514.
18 NOVEMBER 2011
VOL 334
SCIENCE
ASSISTANT PROFESSOR Biological Sciences For position details and application process, visit website: http://jobs.plattsburgh.edu and select BView Current Openings.[ SUNY College at Plattsburgh is an Equal Opportunity Employer committed to excellence through diversity.
TENURE-TRACK FACULTY POSITION in Molecular Immunology The Department of Molecular Biology at the University of Wyoming seeks an outstanding scientist to fill a tenure-track faculty position in Immunology at the ASSISTANT PROFESSOR level. The successful candidate will establish an extramurally funded research program, participate in teaching undergraduates and/or WWAMI medical students, and will contribute to Departmental and interdepartmental graduate programs (see links at website: http://uwacadweb.uwyo. edu/UWmolecbio/). Salary and startup package will be competitive. Candidates must have a Ph.D. degree or equivalent, postdoctoral research experience, and clear evidence of research productivity. Applications must be sent electronically to e-mail:
[email protected] as a single PDF file labeled with your last name and first initial that includes a cover letter, curriculum vitae, description of research interests, and description of teaching interests and philosophy. In addition, please have three letters of recommendation sent electronically to e-mail:
[email protected]. The Department of Molecular Biology includes 16 faculty with diverse research interests and significant extramural support. The University of Wyoming enrolls È13,500 students, including È3,000 graduate students. Laramie is located in the Rocky Mountain region of southeastern Wyoming, about 120 miles from Denver. In addition to opportunities for academic excellence, the University of Wyoming and Laramie offer a small college-town environment with extraordinary cultural and outdoor recreational opportunities. Screening of applications will begin on December 15, 2011 and continue until a suitable candidate is identified. The University of Wyoming is committed to diversity and endorses principles of affirmative action. We acknowledge that diversity enriches and sustains our scholarship and promotes equal access to our educational mission. We seek and welcome applications from individuals of all backgrounds, experiences, and perspectives. The University of Wyoming is dedicated to ensuring a safe and secure environment for our faculty, staff, students, and visitors. To achieve that goal, we will conduct a background investigation on the successful candidate.
www.sciencecareers.org
TOPICS INCLUDE Application of single cell genomics in microbial ecology and bioprospecting
Systems biology/transcriptional networks
Cloud computing as a platform for large scale sequence analysis
An international gathering offering invited presentations, tours, workshops and tutorials on sequence-based bioinformatics, data management systems and new sequencing technologies, and poster sessions. Short talks will be chosen from submitted abstracts.
For more information and to register, scan code or visit: http://1.usa.gov/JGI-UM7
POSITIONS OPEN
Omics in the Arctic: Genome-enabled contributions to carbon cycle and biogeochemical research in high-latitude ecosystems
online @sciencecareers.org
MEETINGS
online @sciencecareers.org
POSITIONS OPEN
POSITIONS OPEN
ASSISTANT PROFESSOR Medicinal Chemistry University of Connecticut, Storrs, CT The Department of Pharmaceutical Sciences at the University of Connecticut invites applications for a tenure-track faculty position at the Assistant level, starting August 23, 2012. The Department is seeking a scientist with a strong background in original research and teaching and a planned research focus in the broadly defined area of medicinal chemistry, including, drug discovery, mechanism of drug action, and characterization of therapeutically relevant macromolecular targets. Areas of expertise may include but are not limited to structural biology, synthetic organic chemistry, chemical biology, and biochemistry. This will complement the interests of the faculty in the Department of Pharmaceutical Sciences, which includes integrative approaches in drug discovery and design. The Department is housed in a recently completed 200,000 square foot state-of-the-art building in the science quad of the University of Connecticut and encourages faculty interdisciplinary interactions with other programs of the University such as Chemistry, Physiology/Neurobiology, and Molecular and Cell Biology. A competitive salary and startup funds will be provided. The successful candidate is expected to develop a strong extramurally funded research program and effectively participate in teaching at the graduate and professional levels. Applicants must possess a Ph.D. degree, good oral and written communication skills, and a strong background in research and teaching. A preferred applicant would also contribute through research, teaching, and/or public engagement to the diversity and excellence of the learning experience. Applicants should submit a cover letter, curriculum vitae, brief statement of research and teaching interests, names, telephone numbers and full addresses of three references via Husky Hire website: http://www.jobs.uconn.edu. Review of applications will start on December 15, 2011 and continue until the position is filled. The University of Connecticut is an Equal Employment Opportunity/Affirmative Action.
ASSISTANT PROFESSOR University of California, Irvine Department of Pharmaceutical Sciences The Department of Pharmaceutical Sciences, at the University of California, Irvine invites applications for a tenure-track faculty position in computational pharmaceutical sciences. The appointment will be made at the Assistant Professor level. The successful candidate will be expected to establish an active research program employing computational methods to address fundamental problems in biomolecular structure and function, pharmacogenomics, pharmacometrics, pharmacokinetics, or any other area of computational biology or chemistry relevant to the pharmaceutical sciences. Interested applicants should submit curriculum vitae, at least three letters of reference, and a brief outline of future research plans. Electronic submissions are preferred. Electronic application instructions can be found at website: https://recruit.ap.uci.edu. Review of applications will begin December 1, 2011. To ensure full consideration, applications and all supporting materials should be received by this time. The position will remain open until filled. The University of California, Irvine is an Equal Opportunity Employer committed to excellence through diversity and strongly encourages applicants from all qualified applicants, including women and minorities.
1018
DIRECTOR Systems Biology Institute Yale University Yale University seeks to appoint the inaugural Director of the recently established Systems Biology Institute. The Institute is one of several science and engineering initiatives located at Yale_s West Campus, a 137-acre campus that has provided the University with unparalleled opportunities to stimulate cuttingedge, interdisciplinary research. The Director will lead the development of innovative research programs at the Institute and oversee the recruitment of Institute faculty, who will have primary appointments in relevant departments within the Faculty of Arts and Sciences, the School of Medicine and the School of Engineering and Applied Science. Candidates must have a Ph.D. in a relevant discipline, an outstanding record of research that demonstrates originality in addressing significant questions in the study of dynamical systems biology, and a proven record as a leader. The Search Committee is particularly interested in individuals with broad and innovative visions for the development of systems biology as a quantitative discipline and the ability to see commonalities and potential synergies across a number of fields. Applicants should create a profile at website: https://academicjobsonline.org/ajo/ jobs/1275 and upload a statement of research plan, curriculum vitae, and up to five reprints of published work(s). Applicants should also arrange for three references to upload their letters of recommendation. For further information, contact: Kelly Locke at e-mail:
[email protected] or P.O. Box 27390, West Haven, CT 06516-7390. The review of applications will begin on 15 December 2011. Yale University is an Affirmative Action/Equal Opportunity Employer. Yale values diversity among its faculty, students, and staff and strongly encourages applications from women and underrepresented minorities.
TENURE-TRACK FACULTY POSITIONS in Immunology/Inflammation The Department of Pathology, Microbiology, and Immunology, School of Medicine, University of South Carolina (USC) invites applications for two tenure-track ASSISTANT PROFESSOR positions in Immunology/Inflammation. Outstanding applicants working in an area complementing our existing faculty research interests (website: http://pmi.med.sc. edu/) will be considered. The candidates must have a Ph.D. or equivalent, and at least three years of postdoctoral research experience. Preference will be given to candidates who have shown evidence of independence and currently active grant funding. Successful candidates are expected to develop a strong extramurally funded research program. They must participate in the teaching mission of the department. Competitive salary and startup funds are available. Please submit curriculum vitae and statement of research plans to: Dr. Mitzi Nagarkatti, Chair, Department of Pathology, Microbiology, and Immunology, University of South Carolina School of Medicine, Columbia, SC 29208 or e-mail:
[email protected]. Kindly arrange to submit three letters of recommendation. The search will start immediately and continue until the positions are filled. USC Columbia is an Equal Opportunity/Affirmative Action Employer, and encourages applications from women and minorities and is responsive to the needs of dual career couples.
Nontraditional Careers:
Opportunities Away From the Bench
Webinar Want to learn more about exciting and rewarding careers outside of academic/industrial research? View a roundtable discussion that looks at the various career options open to scientists and strategies you can use to pursue a nonresearch career.
Now Available On Demand
www.sciencecareers.org/ webinar
Produced by the Science/AAAS Business Office.
Find your future here.
↓
www.ScienceCareers.org
18 NOVEMBER 2011
VOL 334
SCIENCE
www.sciencecareers.org
