Tạp chí khoa học số 2004-11-05

In the Presidency Conclusions of the ECM meeting held in March 2004, the council “sees merit inenhanced support for basic research of the highest quality and the case for specific fundin

E DITORIAL

A crucial day for the future of European research will be 26 November 2004 The European

Council of Ministers (ECM), representing the 25 member states of the European Union(EU), will discuss whether and how the EU will support basic research in the years tocome Ultimately, the question of whether to establish a European Research Council (ERC)will have to be answered, and it is the hope of ERC supporters that the ECM will endorsetheir plan this year

In the Presidency Conclusions of the ECM meeting held in March 2004, the council “sees merit inenhanced support for basic research of the highest quality and the case for specific funding will be examined.”

Despite this positive antecedent and support from 52 major European research

organizations (see the open letter in Science, 6 August 2004, p 776) and

politicians, the result of the impending meeting is far from predictable,and some EU member states as well as parts of the European Commissioncontinue to be reluctant

There are reasons for this reluctance The ERC is meant to beindependent from institutions of the EU and national governments

To have a noticeable impact, its budget must be on the order of atleast 1 to 2 billion euros annually This sum means that almost one-fifth of the future EU research budget will no longer be under thedirect control of the Commission and its dealings with nationalgovernments There is also fear among national ministries that ifcompetition among the very best European research groups is takenseriously, “their” national groups may not succeed Any legal form forthe ERC must ensure that no criteria other than scientific quality, based onpeer-reviewed competition, have an impact on its decision-making Thisincludes the total exclusion of any “juste retour” considerations, which are notonly purely unscientific but also inhibit fair competition and international competitiveness in basic research

Such guarantees for the decision-making process are an essential prerequisite for the ERC’s long-term success

To resolve these issues, the ECM will have to decide on enhanced financial perspectives for Europeanresearch funding and establish a constituting committee to draw up a charter for an ERC that proposes agovernance structure that fits the objectives and prepares the necessary steps for its implementation Thecommittee should consist of a small number of eminent researchers and research managers (perhaps no morethan 10 to 12 people) The members should have outstanding international reputations and several years ofexperience in research policy-making or research management, and should include Nobel laureates Theyshould be appointed in their personal capacity and refrain from representing their respective discipline orinstitution It is essential that the committee be in a position to act as a guarantor for the operational autonomy

of an ERC This can only be achieved if its members pursue no individual interests and truly embody theEuropean basic research community at its highest level, and thus wholeheartedly enjoy its trust and support Theirmidwife role implies that the constituting committee will be dissolved as soon as the ERC is in full operation

The window of opportunity for creating an ERC and moving toward a truly European research base hasnever been as wide open as it is now And yet it is crucial that the necessary steps be taken right now tomake sure that a newly established ERC can start its operations on time, at the beginning of the 7th FrameworkProgramme It is clear that if a negative decision on 26 November should prevail, this window might beclosed for a long time

Should the ECM decide to postpone its decision until 2005 or even to ignore the needs and demands ofEuropean researchers and research organizations, the consequences would be enormous The EuropeanResearch Area would not enter its next and crucial phase, in which the very best of basic research willprovide the foundation for the innovations of tomorrow A further loss of some of the best researchers could

be inevitable Europe cannot afford such negative consequences The ECM has to act now

Wilhelm Krull and Helga Nowotny

Wilhelm Krull is the secretary general of the Volkswagenstiftung Helga Nowotny is chair of the European Research AdvisoryBoard and directs the Society in Science program at ETH Zürich Both were members of the Expert Group examining the needfor establishing a European Research Council (ERCEG)

Decisive Day for European Research

5 NOVEMBER 2004 VOL 306 SCIENCE www.sciencemag.org952

Seeking harmony on nanotechnology

Th i s We e k

Planetary scientists knew that Saturn’s huge,

haze-enshrouded moon Titan—by scientists’

reckoning the solar system’s “largest expanse

of unexplored terrain”—would prove to have

a most unearthly makeup From theory and

observation, they had recognized or inferred

a cold-hardened bedrock of ice,

liquefied-methane clouds, a gasoline-like drizzle, and a

coating of pervasive organic goo Still, as the

Cassini spacecraft now orbiting Saturn

homed in last week for its first close pass by

Titan, scientists expected to glimpse

reason-ably familiar features, including continents,

seas, clouds, and rain Guess again

After staring at the images returned by

Cassini for a few bleary-eyed hours, “we’re

not quite sure what we’re looking at,” said

camera team leader Carolyn Porco of the

Space Science Institute in Boulder,

Col-orado “There isn’t much we’re absolutely

confident about right now.” Two days’ more

reflection didn’t help much Team members

did become confident that Titan has a

youth-ful surface reshaped by surprisingly active

geological processes, a couple of which look

similar to other icy moons of the outer solar

system “Things are starting to make a

minute amount of sense,” said

planetary geologist Laurence

Soderblom of the Cassini team and

the U.S Geological Survey in Flagstaff, zona Yet Cassini scientists have not foundthe most anticipated feature of Titan: seas, or

Ari-at least lakes, of liquid hydrocarbons

From the two flybys of the Voyagerspacecraft in 1981 and later Earth-based ob-servations, Titan looked as if it could be

a fairly simpleworld The organic chemical factory in itsdense atmosphere, where solar radiation in-teracts with methane, would account for theenveloping yellowish haze and scatteredclouds, presumably composed of methaneand ethane At Titan’s surface temperature of

94 degrees above absolute zero,the haze particles of organiccrud would settle onto the water-ice surface of brighter

“continents” or into the darkerhydrocarbon “seas.” Thislight-dark patterning could

be seen from Earth in penetrating infrared telescopicimages Astronomers could only speculate,given chemists’ inference of methane-ethanerain, that Titan’s dark regions were seas, al-though radar astronomers reported that theirradar beams often glinted off the surface likesunlight on a sea Geologists weren’t count-ing on finding anything more on the conti-nents than the inevitable cratering by largeimpacts; Titan has no strong source of inter-nal heat that could still be reshaping the sur-face, say by repaving it with icy “lava,” ashappens on some other icy moons

haze-On first examining the Cassini data, teammembers found anything but the expected

“When you increase [imaging] resolution by

a factor of 10,” said Porco, “you always findsomething new It’s just that we can’t figureout what that is.” The camera found only therare, fleeting cloud outside the south polar re-gion, and the Visible Infrared Mapping Spec-trometer (VIMS) could detect no methane inthe polar clouds Surprisingly, neither Cassi-ni’s imaging radar nor its camera could findany obvious impact craters “Maybe there’ssomething burying them,” said interdiscipli-nary science team member Jonathan Lunine

of the University of Arizona (UA), Tucson

That something could have been an eon’sworth of haze particles settling to the sur-

face Radar team leader Charles Elachi, rector of the Jet Propulsion Laboratory inPasadena, California, reported that along thestrip of surface probed by the radar’s altime-ter, the surface is so flat that its height varied

di-by only 100 meters over many hundreds ofkilometers And the radar, operating in itsradiometer mode, found a heat-emission sig-nature consistent with a coating of organicmaterial rather than bare ice One area, atleast, “is really covered in organics,” saidradar team member Ralph Lorenz of UA

Hundreds of meters of organic crud mightexplain the flatness of the region the altime-ter surveyed, but where the camera and radarreturned images of areas other than thosesurveyed by the altimeter, something moregeologically active is suggested “We haveimages of very complex geological process-es,” says Lunine “There’s enormous diver-sity [of features] in just one area of the radardata.” There are what may be ridgessqueezed up from a plain, says Soderblom,and bright “pieces of string” squigglingacross the landscape that are familiar fromNeptune’s big moon Triton The strings

Titan Remains Mysterious

With a Hint of the Familiar

P L A N E TA R Y S C I E N C E

Haze cutter Infrared imaging pierces Titan’s

Ah, youth Radar imaging of Titan has revealed

abundant geologic features but no obvious impact craters, implying geologic rejuvenation

may be steep scarps eaten into the edges ofthin layers of icy “volcanic” flows, he says.

In both radar and VIMS images, there arestrong hints of so-called grooved terrainthat the Voyagers imaged over large parts

of Jupiter’s Ganymedeand other outer-solar-system moons AndElachi reported thatsome areas of Titanthat are dark in radarimages, at least, give a

“kind of impression of lakes of liquid.”

Planetary geologist James Head of BrownUniversity in Providence, Rhode Island,agrees that Titan’s apparently grooved terrainreminds one of Ganymede’s, but he notes that

there is still no ment about how thosestructures form onGanymede And theCassini camera hasyet to see sunlightglinting off the sur-face of Titan, as itmust if much of themoon is covered byliquid Indeed, most

agree-of the flyby analysis

so far may be flat wrong, says Soderblom,who has long experience with Voyager flybys

of icy moons “This is certainly among thestrangest worlds of the solar system,” he says.Cassini team members agree that suchstrangeness will no doubt demand patient in-tegration of observations from the space-craft’s dozen instruments during the 44 Titanflybys planned for the next 4 years And thosedata will be anchored by the observations to

be made by the Huygens probe as it drifts tothe Titan surface on 14 January after releasefrom Cassini in December There it may add

to the glamour of the mysterious Titan: Whatcould be stranger than an alien craft tossed on

a wine-dark sea of liquefied natural gas?

Two-way invasions

How inflammation promotes cancer

F o c u s

So long Cassini will send

the Huygens probe (goldendisk) to land on the moonTitan in January

The H5N1 avian influenza virus that has astated poultry flocks in Asia and raised thespecter of a global human pandemic seems to

dev-be dev-becoming more dangerous New lab periments suggest that the virus can replicatecopiously in the guts of domestic ducks with-out making them sick What’s more, thisyear’s strain survives longer in the environ-ment than last year’s, and ducks shed largerquantities of it Together, the findings indicatethat controlling the virus could be even morechallenging than previously thought

ex-“This requires a rethinking of the entirestrategy to control H5N1 in poultry,” saysKlaus Stöhr, coordinator of the World HealthOrganization’s (WHO’s) Global InfluenzaProgram “We have to ask if it will be possi-ble to eliminate the virus among chickens ifthere is a reservoir nearby which doesn’tshow it is harboring the virus.”

The experiments, carried out at the St

Jude Children’s Research Hospital in phis, Tennessee, a WHO collaborating centerfor animal influenza viruses, have not yetbeen published, but WHO went public withthe findings because of their potential im-pact on animal and public health efforts

Mem-Although disconcerting, the findings arenot completely unexpected K S Li ofShantou University Medical College in Chi-

na and colleagues at the University of Hong

Kong and other institutions reported in ture on 8 July that asymptomatic domestic

Na-ducks sampled at live animal markets insouthern China were shedding H5N1, sug-

gesting that ducks could be a key factor inthe transmission of the virus

In the new lab study, researchers infectedducks with several strains of the H5N1 viruscollected from chickens and humans earlierthis year and compared the results to those of

a separate study that used strains collected in

2003 Ducks infected with the

2004 strains shed more virus forlonger time periods than didducks infected with earlier strains

And most of the ducks showed nosigns of illness—a surprise for avirus that is so pathogenic inchickens it typically kills themwithin 24 hours of infection

Yi Guan, a microbiologist atthe University of Hong Kong and

a member of the group that

pub-lished the July Nature paper, notes

that past strains of H5N1 did killducks “It is natural that the viru-lence of the virus is gradually de-creasing in aquatic birds,” he says,because viruses that cause rapiddeath have less chance to reproduce But hesays it is unusual and worrying that the virus

is compatible in one bird species yet lethal inanother when the two are often raised togeth-

er The 2004 virus also seems to have come more stable, surviving in the environ-ment for 6 days at a temperature of 37°C,compared to 2 days for older strains

be-Stöhr says additional studies are needed

to determine how widespread the virus is in

domestic duck flocks in Asia, if the virus ismoving between ducks and migratory birds,and what risks infected ducks pose to hu-mans Although there have been no provencases of viral transmission from ducks to hu-mans, WHO now says it must be consideredwhen tracking the route of human infections

The United Nations Food and ture Organization already suggests thatchickens and other poultry species be raisedseparately But enforcing this recommenda-tion is difficult In many Asian countries,free-ranging ducks and chickens mingle andfrequently share the same water supplies Sofar, says Stöhr, only Thailand is trying to getfarmers to keep ducks and chickens apart

Agricul-–DENNISNORMILE

Ducks May Magnify Threat of Avian Flu Virus

I N F E C T I O U S D I S E A S E S

Reservoir New studies suggest that domestic ducks may be

key in avian flu transmission

www.sciencemag.org SCIENCE VOL 306 5 NOVEMBER 2004

Harvard Engineering Growth

Engineering is one discipline likely to get

a boost under a planned Harvard sity hiring boom The Cambridge, Massa-chusetts, campus intends to expand itsfaculty dramatically during the nextdecade, says William Kirby, dean of thefaculty of arts and sciences The pushstems from a report earlier this year urg-ing the university to improve its faculty-to-student ratio (Science, 7 May, p 810).Engineering and applied sciencescould grow from around 60 to 100 pro-fessors; life, physical, and social sciencesalso could win increases in a hiring planthat administrators hope to have in place

Univer-by early next year –ANDREWLAWLER

Parliament’s Blast on Beagle 2

A committee of U.K parliamentarians hasblasted the British government and the European Space Agency (ESA) for fundingdecisions that it says contributed to lastyear’s loss of the Beagle 2 Mars lander (Science, 27 August, p 1227) Beagle 2’sbudget plan was an amateurish “gentle-man’s agreement” that “may have increasedthe risks of an already risky project,” thelawmakers say

Six years ago, ESA selected Beagle 2 to befunded from national coffers.After a consor-tium failed to raise the needed $52 million,ESA and the U.K government bailed it out.But the U.K.’s failure “to provide an adequateguarantee of support early” and ESA’s lack of

“sufficiently close monitoring” probablydoomed the lander, the lawmakers conclud-

ed However, David Southwood, ESA’s head

of science, insists it was not up to ESA tofund the project:“If you want to get themaximum benefit for your scientists and engineers, you have to make sure they haveadequate funding.” –DANIELCLERY

Arctic Warming Accelerating

A new review concludes that the Arctic iswarming nearly twice as fast as the rest ofthe world.The mercury could rise 4° to 7°C

by 2100, says the Arctic Climate Impact Assessment, due to be released at a 9 No-vember science symposium in Iceland

The polar region is warming “at rates

we had not anticipated even 5 years ago,”says Robert Corell, a senior fellow of theAmerican Meteorological Society, who ledthe 4-year exercise involving eight na-tions The consequences are apparent inmelting glaciers and sea ice and thawingpermafrost, the 139-page report notes.Arctic governments are expected to offerrecommendations on 24 November, and adetailed analysis is due from the samegroup in January –JOCELYNKAISER

ScienceScope

Stung by memories of the bruising battles

over genetically modif ied organisms

(GMOs), leaders from industry, academia,

and environmental organizations met in

Houston, Texas, last week to launch a new

fo-rum for hashing out concerns over

nanotech-nology, the nascent field of building materials

up from the atomic scale The rollout of the

International Council on Nanotechnology

(ICON) had its share of hiccups Three

envi-ronmental organizations balked at becoming

founding members, charging that ICON is

too industry-focused But two of those groups

still took part in the meeting as “guests” and

say they will consider joining down the road,

a development that most see as an

improve-ment over the GMO fiasco

“All of the parties have a signif icant

amount of goodwill and want to talk That’s

not a bad starting point,” says Pat Mooney,

who heads the ETC Group, an environmental

organization based in Ottawa, Canada, that

spearheaded the attack against GMOs and has

advocated a cautious approach to

nanotech-nology “There is a lot of trust that needs to be

built from all sides,” says Stephen Harper,

who heads environmental, health, and safety

policy for the computer chip giant Intel in

Washington, D.C

Researchers with Rice University’s

Cen-ter for Biological and Environmental

Nano-technology dreamed up ICON to begin

building that trust, write reports to help

reg-ulatory agencies, and possibly even fund

original research But Mooney and other

en-vironmentalists say they were initially wary

because ICON’s funding comes from

indus-trial members, and some of the proposed tial research projects seemed aimed at con-vincing the public that nanotechnology issafe rather than addressing basic concernsabout the revolutionary technology

ini-At the Houston meeting, participantsavoided potential flash points—such asbacking safety studies or researchingwhether nanotechnology will benefit devel-oping countries—to focus initially on a sub-ject virtually everyone could agree on: deter-mining how to describe various nano-sized clumps of matter Nomencla-ture is particularly tricky for nanoma-terials because different sized nano-particles of a material often havedrastically different properties

Straw-shaped carbon nanotubes, forexample, conduct electricity either assemiconductors or metals depending

on the pitch at which the atoms windaround the straws Such complexitybedevils agencies responsible for reg-ulating the handling, manufacture,and release of nanomaterials

Many nanotechnology expertslaud ICON’s early focus on lan-guage “It will be the key to gettingthe regulations right It’s exactlywhat needs to be done,” says DavidRejeski, director of the foresight andgovernance project at the WoodrowWilson International Center forScholars in Washington, D.C., another or-ganization working to build bridges betweendisparate nanotechnology stakeholders Butwith hundreds of products containing nano-materials already on the market and the fieldexpected to balloon to a $1 trillion industry

in less than a decade, Rejeski says, muchmore needs to be done Mooney and otherenvironmentalists urge governments to actquickly to ensure that products don’t posehealth and environmental threats and to in-clude countries around the globe in discus-sions about how this emerging industry willaffect their economies and societies

Other nanotechnology forums are ing up to tackle those issues In February, forexample, the Rockefeller Foundation is co-sponsoring a meeting in Alexandria, Egypt,called the Global Dialogue on Nanotechnol-ogy and the Poor Leaders from governmentsand other organizations will grapple withproblems such as ensuring access to revolu-tionary technologies and promoting researchthat benefits poor people as well as rich—andhope the cautionary tale of GMOs gives thisstory a happier ending –ROBERTF SERVICE

spring-Nanotech Forum Aims to Head Off

Replay of Past Blunders

S C I E N C E P O L I C Y

Outsiders Nanoparticles’ ability to penetrate living

cells raises environmental concerns

5 NOVEMBER 2004 VOL 306 SCIENCE www.sciencemag.org956

P ARIS —Never before have French scientists

been as single-minded as they were at a

2-day gathering last week in the Alpine town

of Grenoble, when more than 900 of them

voted for a raft of proposals on everything

from cash to careers in the name of

over-hauling fundamental research With

remark-able unanimity, they approved a wish list for

François Fillon, minister for education and

research, who has promised to take a reform

bill to Parliament next spring This effort

builds on last year’s showdown with the

gov-ernment in which more than 2000 lab

direc-tors and research team leaders quit their

ad-ministrative duties to protest funding and

staffing cutbacks (Science, 16 April, p 368).

After the government backed down,

hun-dreds of scientists across the country

organ-ized into working groups to prepare a string

of reports, which were then honed into last

week’s proposals They call for more

coher-ent governmcoher-ent oversight and stronger

sup-port of scientific careers The proposals

in-clude creating a single research and highereducation ministry, an independent higherscience council to advise the government

on strategy, a new body to evaluate all researchers, a long-term jobs plan for researchers, and more crossover betweenagency and academic science They ask that

lecturers’ teaching loads be halved and thatuniversities be reformed in depth Doctoralstudents, who now have no health or socialsecurity coverage, should be given properpay and working conditions, and postdocs

should be given associate searcher contracts of up to 3 years.But how much of the wish listwill find its way into the parlia-mentary bill is another matter

re-“We don’t know what the ment will do with it, but we shallremain vigilant,” says AlainTrautmann, co-director of the cellbiology department at the CochinInstitute and spokesperson for theprotest movement Although Fil-lon told the Grenoble conference

govern-he would also take account ofsome 20 other reform proposals,Trautmann says, “ours outstrip theothers by a wide margin” in repre-senting the community

All the major political parties are payingattention: Their leaders were on hand lastweek, echoing a commitment to raise re-search spending to the European Union goal

of 3% of gross domestic product (GDP) by

2010 And researchers were warmed by lon’s clear admission that the governmentmade mistakes last year when his predeces-sor Luc Ferry and former research chiefClaudie Haigneré were in office

Fil-Although the show of hands in favor ofthe reform proposals was almost unanimous,the preceding 2 days of debate were far fromdocile They were peppered with applause,boos, and a rowdy protest from an anti-science lobby that was silenced when thedemonstrators were hustled from the confer-ence room The high spot, some delegatessaid, was the summing-up by EdouardBrézin, French Academy of Sciences vicepresident and co-president of the Initiativeand Proposal Council (CIP), which wasformed earlier this year to produce a consen-sus for change among scientists “It was amoment of great emotion,” says Trautmann

“We had to produce [a document] thatcontained neither overspecific recommenda-tions nor a compromise that was so general

we would have looked ridiculous,” saidBrézin, who will take over as academy presi-dent from endocrinologist Etienne-EmileBaulieu next January “We have succeeded

in drawing up precise and realistic proposalsand have not simply issued a union-type de-mand for more money, more posts.”

The unions are happy with the result Theproposals “constitute a working frameworkfor the scientific community,” says the lead-ing research union, SNCS But universitypresidents are far from happy They regret

“the absence of important elements” theyhad suggested, such as merging agency

Researchers Back a 70-Page Agenda to

Reform Agencies, Boost Careers

F R E N C H P R O T E S T S

Denouement Earlier protest marches have paid off this

fall by drawing political attention to researchers’ demands

Hot on the Trail of Cosmic Rays

C AMBRIDGE , U.K.—Researchers are closing

in on finding the source of galactic cosmic

rays, charged particles that whiz around

the galaxy in huge numbers and constantly

bombard Earth’s upper atmosphere

Im-ages published this week in Nature show

the production of high-energy gamma rays

around the remnant of a supernova,

known as RX J1713.7-3946 The pictures,

taken with a new gamma ray telescope

called the High Energy Spectroscopic

System (HESS) in Namibia (Science, 3

Sep-tember, p 1393), mark the first time

re-searchers have produced a resolved image

of a supernova remnant at such high

ener-gies Astrophysicists believe that thesegamma rays, with energies of about 1012

electron volts (TeV), are produced at thesame time as cosmic rays and so mark thelocation of their source, but they haven’tgot proof yet “This strong signal is abreakthrough,” says Karl Mannheim of theUniversity of Würzburg in Germany “Butthere are many open questions.”

Cosmic rays travel at speeds produced

in the most powerful particle accelerators

Theorists believe that when particlesstreaming out of a supernova remnant hitinterstellar gases, protons and other lightnuclei get boosted by the shock wave andproduce a few TeV gamma rays as abyproduct The problem is, electronsstreaming from a supernova remnant canalso generate TeV gamma rays, withoutcosmic rays being involved The HESS teamshould be able to figure out whether pro-tons or electrons are the culprits by study-ing the supernova at other wavelengths,such as radio waves, to figure out the den-sity of matter around it “We’re just gear-ing up that,” says HESS spokespersonWerner Hofmann of the Max Planck Insti-tute for Nuclear Physics in Heidelberg

–DANIELCLERY

H I G H - E N E R G Y A S T R O P H Y S I C S

NE W S O F T H E WE E K

and university researcher status to replace

the separate agency and university academic

researcher categories, and turning the

lead-ing agencies into mere suppliers of cash

CIP is putting finishing touches to the

70-page document before presenting it to

Fillon and Research Minister François

d’Aubert on 9 November Most changes are

modest, but the wording may prompt further

discussion For example, the report will

ex-plicitly reject the idea of making the

Nation-al Research Agency, which will be created

next year, the lead provider of research

funds Instead, delegates agreed, ment subsidies to agencies should continue

govern-to provide the bulk of the cash Even all thatwould not be enough to put French researchback on track, according to Baulieu He told

a parliamentary conference on Tuesday thatthe government must inject an extra €1 bil-lion into research each year for 5 years, or atotal of €15 billion, if it is to achieve the 3%

of GDP target The French reform game isclearly far from over

–BARBARACASASSUS

Barbara Casassus is a writer in Paris

The European Parliament (EP) is asking its

25 member nations to place a ium on the use of high-intensity navalsonars implicated in the mass deaths ofwhales Last week’s nonbinding resolutioncites “increasing scientific and public con-cerns” about ocean noise impacts oncetaceans and calls for an in-depth study

morator-of the issue

Over the last decade, researchers havelinked several strandings to sonar, butthey are still uncertain of exactly how thesound pulses harm cetaceans (Science,

24 October 2003, p 547) The

moratori-um and study are a needed step towardclarifying the science and “searching forclean technologies,” says marine mam-malogist Antonio Fernández of the Uni-versity of Las Palmas de Gran Canaria inthe Canary Islands But the resolutionmay have little impact unless it is adopt-

ed by the European Commission and theCouncil of Ministers, which will considerthe measure next year

or burns

Ethics bodies in France and the UnitedKingdom have recently rejected proposalsfor similar transplants due to concernsabout tissue rejection, with some expertspredicting a 50% failure rate But MariaSiemionow, the clinic’s director of plasticsurgery research, believes the transplantscould improve on current skin-graftingtechniques, which can leave significantscarring and allow little or no facial expression, she told the Cleveland PlainDealer The procedure would likely involve removing skin and fat depositsfrom cadaver donors, but the patient isn’tlikely to look like the donor after surgerybecause the underlying muscles andbones are what shape the face

The first change of face may not comeanytime soon: Clinic doctors caution that

it may take months to find appropriatedonors and patients

–SEANBRUICH

M OSCOW —President Vladimir Putin has

be-gun a campaign to trim Russia’s vast

net-work of state-funded research institutions

Although there is general agreement that the

current system is broken, no consensus

ex-ists on what is needed Many scientex-ists

worry that Putin’s proposed fix could do

se-rious damage to the Russian Academy

of Sciences (RAS),whereas others arguethat additional struc-tural changes are alsonecessary

Speaking here lastweek to the UnitedPresident’s Council

on Science, ogy, and Education,Putin squelched ru-mors that the govern-ment wants to elimi-

Technol-nate RAS (Science,

29 October, p 795) “Nobody wants to ruin

RAS,” he said “We want to preserve the

academy The question is how to adapt it to

present realities.” That could be hard work,

he acknowledged “We expect a thoughtful

and effective modernization We expect

pro-posals on a drastic reorganization of work.”

Putin’s comments appear to give a green

light to a plan, formally known as the

Con-cept of the Management of State Research

Organisations, that would privatize, merge,

sell off, or close most of the academy’s

rough-ly 5000 research institutions while increasing

support for the 100 to 200 institutes that

re-main (Science, 24 September, p 1889)

Sci-ence and education minister Andrey Fursenko

has promised a $26,000 boost by 2008 in per

capita spending—including salaries,

equip-ment, and other costs of doing science—for

the nearly 150,000 scientists expected to

re-main at RAS At the same time, however, the

proposal would put RAS on a shorter leash

The Russian cabinet would be given the

au-thority to approve RAS’s operating charter,for example, and to ratify the choice of anRAS president by academicians

Such drastic moves have drawn criticismfrom RAS trade unions, which have de-manded Fursenko’s resignation They alsoprompted an angry letter to Putin from anumber of Russian Nobel Prize winners,who fear that RAS will be swallowed up bythe Russian bureaucracy “We indeed areworried about this concept, which has scaredeveryone so much,” says 2003 physics No-belist Vitaly Ginzburg “The academy must

in no circumstances become a state ment; this is the main point.”

depart-Shortly after the meeting, RAS reportedchanges to its leadership that are believed

to reflect Putin’s new priorities Valery Kozlov, former RAS vice president ofyouth policy, was appointed vice president

in charge of the RAS restructuring Vicepresident Gennady Mesyats, a fierce oppo-nent of the reform plan, saw his authorityover financial issues handed to AlexandrNekipelov, who was appointed vice presi-dent Nekipelov refused comment on thechanges, and Mesyats was unavailable

In the next few months the Russian Duma (parliament) is expected to take upchanges in the tax laws and other provisionsthat would be needed to implement the Con-cept But Boris Saltykov, who pushed unsuc-cessfully for reform as Russia’s first scienceminister in the mid-1990s, believes that evenstronger medicine is needed to correct whatails RAS “If you give money to an old Sovi-

et quasi-structure like RAS, it will be used

as ineffectively as in the past and will nothave any substantial effect,” he says “While

it is necessary to retain some form of RAS,its very structure must be reformed But noone even speaks about structural reform.”

–ANDREYALLAKHVERDOV AND

Change agent Putin

wants RAS to “adapt

to present realities.”

www.sciencemag.org SCIENCE VOL 306 5 NOVEMBER 2004 959

B ERLIN —Janez Potoc∨nik

should be safely installed

as the European Union’s

(E.U.’s) new research

commissioner by now,

but his future remains in

question after the entire

slate of new

commis-sioners, the E.U.’s

policy-making body, was pulled

last week just hours

be-fore an approval vote by

the European Parliament

The unprecedented move

was forced on

President-designate Jose Manuel

Barroso following

con-troversial comments by

the nominee for justice The would-be

com-missioners now face a shakeup; the final

out-come is unlikely to be settled before early

De-cember

The Parliament has only a yes or no vote

on the entire 25-member commission, so the

furor over Italian nominee Rocco Buttiglioni’s

comments on homosexuality and the role of

women forced Barroso to withdraw his wholeslate before an expected no vote on 27 Octo-ber The comments were enough to convince amajority of the Parliament that he was not fit

to oversee the E.U.’s antidiscrimination laws

Buttiglioni resigned on 30 October, butthe aftershocks of his replacement couldtrigger further changes Several other pro-

posed commissioners got low marks fromParliament, and Barroso must now decidewhether to move some commissioners tonew positions in an effort to find a slate like-

ly to win the Parliament’s approval There is

a chance that Potoc∨nik could be shifted inthe resulting reorganization

Many European scientists are hoping thatdoesn’t happen In contrast to Buttiglioni,Slovakia’s Potoc∨nik got high marks from theCommittee on Industry, Research, and Ener-

gy that questioned him During his 3-hourhearing on 1 October, he voiced support forthe creation of the European Research Coun-

cil (ERC) that would fund basic research ence, 29 October, p 796) and a doubling of

(Sci-the budget for (Sci-the next Framework program,

which funds E.U science (Science, 25 June,

p 1885)

Potoc∨nik has also gotten positive earlymarks from observers outside the E.U bu-reaucracy “I was very favorably impressed,”says Jose Mariano Gago, former Portuguesescience minister and head of the Initiative forScience in Europe, a group lobbying for thedevelopment of the ERC, who met with Potoc∨nik last month “I think he’s a man ofgreat capacity He is totally committed to advancing scientif ic development inEurope.” Europe’s scientists will have towait to see if he gets a chance to do so

–GRETCHEN VOGEL

New Research Commissioner

Caught in Controversy’s Wake

E U R O P E A N U N I O N

T OKYO —It may be a truism that scientific

progress has widened the gap between the

industrial and the developing world But

next week in Kyoto, a group of Nobelists,

government officials, and business leaders

will discuss how scientific advances might

also be used to bridge that gap

The occasion is the first Science and

Technology in Society forum The driving

force behind the project is Koji Omi, a

veter-an Japveter-anese legislator veter-and former science

minister who envisions the forum as an

an-nual event Rather than just assembling

sci-entists and engineers, including nine Nobel

laureates, Omi decided to reach out to media

representatives, politicians, and captains of

industry, too In three preparatory meetings,

a group of 40 founding

members—includ-ing Sydney Brenner of the Salk Institute in

La Jolla, California; Lu Yongxiang,

presi-dent of the Chinese Academy of Sciences;

David King, scientific adviser to the U.K

government; and Henry McKinnell, CEO of

Pfizer Inc.—agreed to concentrate initially

on five areas: energy and the environment,

bioethics, information technologies and their

social impact, capacity building, and human

health and bioterrorismissues “Our hope is tobuild a network of peo-ple who will thinkdeeply about these prob-lems,” says Omi

Bruce Alberts, dent of the U.S NationalAcademy of Sciencesand also a foundingmember, says that bridg-ing the divide betweenadvanced and developingcountries is an importantsubtheme of the forum,which will feature twin-ned sessions to exploreparticular problemsthrough the perspective

presi-of both advanced tries and the developingworld For example, speakers from Japan, theUnited Kingdom, and Germany will lead adiscussion on “The Promise of Clean Ener-gy,” while officials from India, China, andSudan will offer their views on “The Chal-lenge of Meeting Energy Needs in Develop-

coun-ing Countries.” Albertsviews the forum as “anexperiment,” with the testbeing whether partici-pants use it as a spring-board for tackling in-tractable global problemssuch as the digital divide.That kind of exchange

is also the fond wish of

1999 chemistry NobelistAhmed Zewail, a native

of Egypt and a professor

at the California Institute

of Technology in dena Zewail, anotherfounding member, hopesthis “dialogue of cul-tures” will lead to more

Pasa-“support for science cation and the technolo-gies that can help developing countries.”

edu-The 11 to 14 November meeting is by vitation only The Japanese government put

in-up about a quarter of the $4 million neededfor the meeting, with Japanese businessescontributing the rest –DENNISNORMILE

Japan to Host Global Forum on Societal Impacts

S C I E N C E P O L I C Y

Going deep Koji Omi wants “deep

thinkers” to explore how science can leviate societal woes

al-N E W S O F T H E WE E K

5 NOVEMBER 2004 VOL 306 SCIENCE www.sciencemag.org960

May 2003 was an up-and-down

month for Daniel Klessig On

7 May, he resigned as

pres-ident of the Boyce

Thomp-son Institute (BTI) for

Plant Research in Ithaca,

New York, for health

reasons But 9 days

later, his lab

pub-lished a paper in

Cell that promised

to be a major

ad-vance His team

had identif ied a

long-sought

zyme that

en-ables plants to

make nitric

ox-ide, especially in response to

infec-tions Last Friday, however, Klessig and

some of his co-authors retracted the Cell

pa-per and a subsequent one published in the

Proceedings of the National Academy of

Sci-ences (PNAS) that was dependent on the

ear-lier research “This is the worst experience

I’ve had in my life, and that includes

open-heart surgery,” says Klessig

According to Klessig, an “absolutely

criti-cal” experiment described in the Cell paper

cannot be reproduced by members of his lab,

and he now considers the experiment’s data to

be “shaky.” Meena Chandok, the first author

on both papers, did not approve the

retrac-tions BTI is investigating the matter Klessig

acknowledges that Chandok left his

laborato-ry in April: “I can’t say a lot more than what’s

in the retraction.” (Science was unable to

reach Chandok.)

The initial Cell paper was considered a

breakthrough because plant biologists have

recently come to recognize that nitric oxide

may have as many roles as a signaling

mole-cule in plants as it does in animals, where it

regulates immunity, nerve cell

communica-tion, and much more Yet investigators had

been unable to identify a plant version of a

nitric oxide synthase (NOS), the enzyme that

creates the gas in animals But in the Cell

re-port, Chandok, Klessig, and their colleagues

described a tobacco plant protein that

pro-duced nitric oxide in response to infection

“The paper was very important, as it

was the first discovery of a nitric oxide

synthase enzyme … in plants; it provided a

mechanism for making NO in response to

pathogen infection, and it indicated that

plants have a novel mechanism for

produc-ing NO compared with animals,” says plant

biologist Nigel Crawford of the University

of California, San Diego “This is a major

setback for the field.”

The tobacco tein identif ied

pro-by the Ithacateam bore littleresemblance toanimal NOS en-zymes, but a fol-low-up paper bythe group in the 25

sup-In the Cell retraction, Klessig’s team says

that the data showing that the recombinantform of the tobacco protein has NO-making

properties cannot be reproduced, and they

call the original data in the Cell paper

“unre-liable.” Although Klessig’s lab continues tostudy the protein and hasn’t ruled out that itmay be an NOS, he advises others not topursue the matter for now “It’s importantthat the rest of the scientific community notbase their research on this unreliable datathat we are no longer confident in,” he says

In any case, the hunt for a responsive NOS in plants may be over Last

pathogen-October in Science, Crawford’s team fied a gene in the mustard plant Arabidopsis thaliana that encodes a protein, dubbed At-

identi-NOS1, with a strong resemblance to a NOSfound in snails, and they showed that it pro-duces the gas in response to hormonal sig-nals Now, in a paper in the 2 November

PNAS, a group headed by Jörg Durner of the

GSF-National Research Center for ment and Health in Neuherberg, Germany,reports that mustard plants exposed to thebacterial cell surface components activate At-NOS1 as a way to make NO and induce de-fensive genes Mutating the gene left theplants more vulnerable to bacterial infection

Environ-“It’s a quite convincing story,” says Klessig

–JOHNTRAVIS

NO-Making Enzyme No More:

P L A N T B I O L O G Y

Ethicists to Guide Rationing of Flu Vaccine

Despite emergency efforts by the Centers forDisease Control and Prevention (CDC),medical experts have not been able to solve

a pressing numbers problem: how to ute 61 million or so existing doses of fluvaccine among the 95 million Americans inhigh-risk categories

distrib-So, with only half the expected supply ofvaccine available and flu season imminent,CDC has turned to another class of experts—

ethicists The agency, facing calls for ance from state and local health officials asthey struggle to ration the vaccine, hasformed its first-ever ethics panel to help

guid-guide flu vaccine distribution The member group held its first two meetings lastweek, and CDC expects to issue recommen-dations later this month Although this panel

five-is a stopgap measure, CDC officials are ning to establish a more permanent ethicscommittee, which will consider a range of issues, including vaccine distribution

plan-The flu panel can’t act soon enough Over

40 million doses of flu vaccine have alreadybeen shipped by their maker, Aventis; CDCdoesn’t know how many have been adminis-tered Last week, the Department of Healthand Human Services announced that it had

located 5 million more doses andwas hoping to purchase them, ifthey meet safety criteria Mean-while, state health officials arestruggling with difficult ques-tions Should an elderly individ-ual, who is in danger of dyingfrom the flu but has already livedmuch of his or her life, receivevaccine before a 2-year-old, who

is likelier than most people to getsick but will probably survive?(Last year, 36,000 people died offlu in the United States; 150 wereyoung children.) If a nursinghome has only enough vaccine foreither its staff or its patients, whoshould receive it?

I N F E C T I O U S D I S E A S E S

Picking and choosing Ranking the flu vaccine needs of babies

versus those of the elderly is one of many thorny questions

NE W S O F T H E WE E K

The crisis erupted on 5 October after

British authorities banned distribution of

vaccine made at Chiron’s U.K plant because

of contamination Within 4 hours of hearing

the news, frantic CDC officials convened a

meeting of their Advisory Committee on

Im-munization Practices (ACIP) to revise

rec-ommendations on who should get flu shots

Before the shortage, CDC put the number of

recommended recipients at 185 million

Based on various studies assessing a

per-son’s risk of contracting and dying from flu,

ACIP members managed to shrink it to 95

million For example, the committee agreed

that only very young children needed to be

vaccinated and that healthy 50- to

64-year-olds didn’t Available data, however, didn’t

permit ACIP to reduce the number further

“ACIP and CDC will go as far as the

science will allow us to go,” says Lance

Rodewald, director of the immunization

services division in CDC’s national

immu-nization program “There was a sense that

we needed some help.”

Five ethicists have agreed to provide it:

Daniel Callahan of the Hastings Center in

Garrison, New York; John Arras of the

Uni-versity of Virginia in Charlottesville;

Kath-leen Kinlaw of Emory University in Atlanta,

Georgia; Thomas Beauchamp of

George-town University in Washington, D.C.; and

Robert J Levine of Yale University in New

Haven, Connecticut

The panel is first focusing on the

rela-tive responsibilities of CDC and local

health departments in making

recommen-dations about rationing The group may

consider questions such as the life

expectancy of different populations with

and without the vaccine, says Levine It

may also examine how to prevent shortages

of other vaccines in the future Eight of 11

childhood vaccines, for example, have been

in short supply in recent years

“What the CDC is doing now is

some-thing that we need,” says Paul Offit, chief of

infectious diseases at Children’s Hospital of

Philadelphia Although vaccinating health

care workers—as most states are doing—

can help contain influenza’s spread, “what

are the ethics of immunizing a healthy adult

who works in the hospital and is unlikely to

die?” Offit asks

The issues are tough enough when they

concern routine flu immunization The

CDC panel will also likely consider how

vaccine should be rationed if a deadly new

flu pandemic emerges, possibly from the

avian flu strains now circulating in Asia

There, CDC won’t be alone: On 11

Novem-ber, the World Health Organization will

hold a summit with country officials and

16 vaccine companies to determine how to

boost vaccine and antiviral supplies now

Data from the Children’s Environmental Exposure Research Study (CHEERS) willhelp the U.S Environmental ProtectionAgency (EPA) determine safe levels for in-fants and toddlers But the study is under at-tack because it is partially funded by a chemi-cal trade association, and because parents willnot be informed of the potential danger totheir children from continued use of pesticides

in the home “The nuances of this study raise

a series of questions,” says environmentalhealth researcher Bruce Lanphear of the Uni-versity of Cincinnati, Ohio

Although many studies have looked at cupational exposure to pesticides and otherchemicals, little is known about how pesti-cides get into children’s bodies CHEERShopes to fill that gap by following 60 childrenfor 2 years Parents will track what their chil-dren eat and how active they are, while alsokeeping track of the pesticides used in thehouse Field teams will visit every 6 months tosample floors and other surfaces and to collecturine to analyze for metabolites of the pesti-cides The study is being conducted in andaround Jacksonville in northeastern Florida

oc-Each participating family will be paid $970and will get to keep the camcorder they used

to record their child’s movements

EPA originally planned to study only ticides, but last month it received a $2 million

pes-contribution from the American ChemistryCouncil to add several other kinds of com-mon household chemicals, including flameretardants, perfluorinated chemicals, and in-gredients in plastic products The council,which represents 136 chemical companies,says it wants any government regulation to bebased on facts about exposure

That arrangement “erodes EPA’s ence of research to regulate chemicals,” saysRichard Wiles of the Environmental WorkingGroup (EWG), a nonprofit advocacy organi-zation based in Washington, D.C EWG alsoaccused the agency of, in effect, paying par-ents to expose their children to pesticides EPA denies those charges “People wouldnot do this just for the money,” says LindaSheldon, acting director of EPA’s Human Ex-posure and Atmospheric Sciences Division.Although household use of pesticides is onecriterion for enrollment, parents are not re-quired to continue applying pesticides duringthe duration of the study, she adds

independ-The study has been approved by EPA’sInstitutional Review Board (IRB), whichexamined subject protection and other is-sues Green lights were also given by IRBsfor EPA’s partners—the Centers for DiseaseControl and Prevention and the DuvalCounty Health Department—and one of itscontractors

Sheldon says EPA approached the can Chemistry Council (ACC) for the addi-tional funding because it has a program of re-search grants ACC did not pick the additionalchemicals, she says, and it will have no role inthe study nor any say over publication ACC’sCarol Henry seconds that: “We have no con-trol over the study.” Although EWG noted that

Ameri-an advisory committee will contain industryscientists, EPA picked them and will rely onthem solely for technical advice, mainly aboutanalytical techniques, says EPA’s Roy Fort-mann, a project leader

Some researchers say the study has cal shortcomings, especially because EPAdoesn’t plan to warn parents explicitly aboutthe potential risks of pesticides to children

ethi-“It’s unconscionable,” says Leo Trasande ofMount Sinai School of Medicine in NewYork, who works on biomonitoring and chil-dren’s health “EPA’s policy is that if you ap-ply pesticides according to the labels, thenthey should be safe for indoor use,” Sheldonsays Of course, it’s the CHEERS study thatwill provide EPA with the data needed to de-termine ultimately what’s safe for children

–ERIKSTOKSTAD

With reporting by Jocelyn Kaiser

EPA Criticized for Study of Child Pesticide Exposure

T O X I C O L O G Y

Source? An EPA study will determine how young

children are exposed to household pesticides andother chemicals

N E W S O F T H E WE E K

With high-domed skulls built like battering

rams, dinosaurs called pachycephalosaurs

look for all the world as if they must have

butted heads Paleontologists imagined the

males sparring for mates as bighorn sheep

do, and the idea was bolstered by radiating

bony structures that apparently strengthened

the head against impacts But did they

actu-ally knock noggins?

To find out, Mark Goodwin of the

Uni-versity of California, Berkeley, and John

Horner of Montana State University in

Boze-man did something that would

give most museum curators the

heebie-jeebies: They sawed open

the skulls to examine the

fos-silized bone tissue The answer

was trapped within the domes,

Goodwin says, and histology—

the study of tissues—was the only

way to get it

Preserved in the bone, as in

many fossils, was a beautiful

record of the original tissue,

down to the level of individual

cells That’s beyond the

resolu-tion of computed tomography

scanners By studying

pachy-cephalosaurs of various ages,

Horner and Goodwin determined

that the radial structures were

ephemeral features associated

with fast-growing bones of juveniles There

was no sign of stress to the skull bones, they

reported in the spring issue of Paleobiology.

“I didn’t see any evidence that they

head-butted,” Goodwin says However, he and

Horner did f ind bundles of so-called

Sharpey’s fibers, which anchor ligaments

and also thick pads of keratin to bone

Horner and Goodwin speculate that this may

have secured a crest to the top of the head,

perhaps for display

More and more paleontologists are putting

their fossils under the knife—the rock saw,

actually—to gain new insights into their

biol-ogy “The microstructure includes a

tremen-dous amount of information,” says Armand

de Ricqlès of the Université Paris VII After

removing a slice of bone, they glue it to aglass slide and then grind it until it is trans-parent Studying this “thin section” of bonetissue with microscopes can explain the ori-gin of strange structures, such as the thickheads of pachycephalosaurs and the plates ofstegosaurs, and help test hypotheses abouttheir function “I get quite excited about thepotential of using bone microstructure toflesh ancient animals out and make themmore real,” says Anusuya Chinsamy-Turan ofthe University of Cape Town, South Africa

Paleohistology is already shedding light

on the question of how sauropods and nosaurs attained their gigantic sizes and oth-

tyran-er evolutionary patttyran-erns It can tell adult mals from juveniles, and it provides the onlyway to determine how old extinct animalswere when they died and how quickly theygrew—key questions for studying their pop-ulation biology and ecology “We’re on thecusp of being able to learn a lot about the bi-ology of these animals—things we thoughtwe’d never be able to tease out of the bones,”

ani-says Lawrence Witmer of Ohio UniversityCollege of Osteopathic Medicine in Athens

This week at the Society of Vertebrate ontology (SVP) annual meeting in Denver,Colorado, paleontologists unveiled a bumper

Pale-crop of histological studies, from a possible

determination of the sex of a Tyrannosaurus rex to identification of an island of dwarf

sauropods “This field is about to explode,”says paleobiologist Gregory Erickson ofFlorida State University, Tallahassee

Diverse tissues

Paleohistology has a long history For 150years, paleontologists have used the tech-nique to classify ancient fishes But not un-til the early 20th century did they begin to

compare the microstructure ofvarious fossil land animals.Studies by Rodolfo Amprino ofthe University of Turin, Italy, led

to an observation in 1947 that isnow called “Amprino’s rule”:The rate of an animal’s growthstrongly influences the type oftissue deposited in its bones Arapidly growing bone has manyblood vessels Its characteristic

“f ibrolamellar” texture ismarked by quickly depositedfibers and holes that are thenfilled by bony structures calledprimary osteons In contrast, abone growing more slowly has atexture called lamellar-zonalwith fewer blood vessels and afinely layered appearance

Dinosaurs typically have bone tissuethat more closely resembles the quicklygrowing bones of large birds and mammalsthan the slow, lamellar-zonal tissue of rep-tiles In 1969 and in later papers, de Ric-qlès suggested that the dinosaur bone tissuemight indicate fast, continuous growth and

an active metabolism like that of mammalsand birds This idea played an importantrole in the “renaissance” that changed theperception of dinosaurs from sluggish rep-tiles to active, possibly warm-blooded ani-mals Tissue type turned out not to be asimple indicator of metabolism, but it doesindicate the general pace of growth

Bone tissue offers another way to stand the growth of ancient animals Bones C

With a wealth of good specimens now at their disposal, paleontologists are probing elegantly

preserved fossil bone tissue for clues to how long-extinct animals grew and lived

Dinosaurs Under the Knife

N e w s Fo c u s

Supersize me Bone-based growth curves show how T rex dwarfed its kin.

sometimes lay down dark lines, called lines

of arrested growth (LAGs), which represent

periods when growth slowed or stopped for a

while LAGs are common in amphibians and

reptiles, but modern birds typically lack them

because they complete their growth in less

than a year In 1981, Robin Reid of Queen’s

University of Belfast reported that dinosaurs

showed the lines, too By counting them like

tree rings, paleontologists can infer how

many years a bone has grown, and by

exten-sion how long the dinosaur lived This

tech-nique of skeletochronology is now widely

used by biologists studying modern reptiles

and amphibians thanks to Jacques Castanet

and others in de Ricql`es’s laboratory in Paris,

which has trained many paleohistologists

Interpreting fossils can be tricky For one

thing, an animal’s body continually dissolves

primary bone—to extract calcium or to repair

microfractures—and then deposits secondary

bone, erasing the bone’s early history To

ac-count for missing LAGs, researchers must

make assumptions about their spacing and

about bone deposition rates—no simple task,

because in living animals, deposition rates

vary widely from species to species and even

between bones in the same individual

Tem-perature and diet affect bone growth, too

One solution is to look at many

speci-mens of various ages, so that juvenile bone

fills in the missing picture for adults “As

long as I have enough individuals and a

di-versity of bones, I can reconstruct what

was going on,” says Kristi Curry Rogers of

the Science Museum of Minnesota in St

Paul By counting LAGs, researchers can

assemble a series with individuals of

vari-ous ages Then, using techniques for

esti-mating an animal’s mass from the size of

its bones, they plot how various types of

dinosaurs typically grew over time Most

growth series are partial, but the hadrosaur

Maiasaura is known from embryo to adult.

When growth curves were published in

the 1990s, they revealed startling facts

about dinosaurs In a 1999 Journal of

Vertebrate Paleontology paper, for

exam-ple, Curry Rogers showed that the giant

sauropod Apatosaurus reached full size—

25 meters long—in just 8 to 11 years, not

the decades that had long been assumed

“People would have laughed!” says Kevin

Padian of the University of California,

Berkeley The quick growth rate

compli-cates a long-standing puzzle: How did

sauropods, with their relatively small

mouths and simple teeth, manage to get so

big, particularly during the Jurassic, when

only cycads and other plants of meager

nu-trition were growing?

Researchers have also established a

general pattern for dinosaurs, compared to

other groups In a pair of 2001 Nature

pa-pers, two groups—Padian’s team and ickson and colleagues—used differenttechniques to plot growth curves for sever-

Er-al dinosaur species Both concluded thatdinosaurs grew faster than reptiles Thelarger dinosaurs packed on weight at apace comparable with that of mammals,but none grew as blindingly fast as modernbirds do Growth curves also show that,

like birds and mammals, dinosaurs grewfast when young, then slowed down andstopped growing as adults By contrast,nondinosaurian reptiles such as crocodilesgrow more slowly

Growth curves have been used to gate how dinosaurs evolved various patterns

investi-of growth In August, Erickson and several

co-authors reported in Nature how T rex

evolved to its formidable size, relative to

other tyrannosaurids (Science, 13 August,

p 930) Rather than extend its growth phase,

T rex accelerated its adolescent growth

spurt—packing on up to 2 kilograms a day

Sauropods show similar changes, according

to a paper by Martin Sander of the sity of Bonn, Germany, and colleagues, in

Univer-press at Organisms, Diversity & Evolution.

“That’s not information you can get fromgross anatomy,” notes Allison Tumarkin-Deratzian of Vassar College in Poughkeep-sie, New York “The only way you can pindown accelerated rates of growth versus ex-tended period of growth in the fossil record

is by looking at histology.”

Giants and dwarfs

As a rule, dinosaurs and other vertebratesevolved to get bigger through the ages Butthe bones show that some bucked the trend

At this week’s SVP meeting, Sander and Octavio Mateus of the Museum of Lourinha

in Portugal and colleagues announced thatsmall sauropods discovered in Germany arenot juveniles but “the first unequivocal case

of dwarfing for any dinosaur.” The 10 viduals range in size from 1.8 to 6.2 meterslong—much smaller than the 23-meter-longbrachiosaurids to which they are closely

indi-related—but the tightly spaced growth lines

in their bone tissue clearly show that theywere full grown The growth curve, based

on seven leg bones, suggests that the dwarfsmay have been sexually mature at as young

“island dwarf ” sauropods, titanosaursfrom Argentina and Romania

Birds are another group that reducedtheir body size relative to their dinosauri-

an ancestors, and histology is helping searchers figure out how that happened

re-By comparing their tissue with those ofthe most birdlike dinosaurs, Padian andothers have argued that they shrank byshortening the amount of time they spendgrowing most rapidly (Most birds reachfull size within a few weeks.) “It’s a verysmart idea,” says Luis Chiappe of the Nat-ural History Museum of Los AngelesCounty So even though they grow at afaster rate than dinosaurs, they end upsmaller—which would have been a key

www.sciencemag.org SCIENCE VOL 306 5 NOVEMBER 2004 965

step toward evolving the ability to fly

Following up on research begun by

de Ricqlès in the 1970s, Chinsamy-Turan is

also looking at the beginnings of another

fast-growing group: the mammals Research on

their therapsid ancestors shows that some of

these so-called mammallike reptiles were still

growing like reptiles, while others show some

distinct evidence of more mammalian growth

patterns Now she’s looking at Mesozoic

mammals from the Gobi desert “I have

begged and really pleaded” to get access to

these rare specimens, she says, to compare

them to modern mammals

Paleo-exotica

Sometimes zeroing in on ancient bones

turns up exotic results At the SVP

meet-ing, Horner, and Mary Schweitzer and

Jen-nifer Wittmeyer of North Carolina State

University in Raleigh, described tissue,

never reported from a dinosaur before,

from the femur of a T rex The tissue has a

random structure and is much richer in

blood vessels than surrounding tissue The

researchers propose that it functioned like

tissue that female birds use to store

calci-um for making eggshell If so, it would be

the first time paleontologists have

deter-mined gender—and reproductive status—

from a dinosaur bone Some skeptics,

however, think the tissue structure might

be the result of injury or disease

Other novel tissues have been reported

from flying reptiles Pterosaur bones have

plywoodlike tissue made of layers stacked

so that bone fibers run at right angles in

alternate layers Such crisscrossing

struc-tures are common in f ish scales, but

Horner, Padian, and de Ricqlès were the

first to describe them in a four-limbed

ver-tebrate In 2000 in the Zoological Journal

of the Linnean Society, they speculated

that the tissue is an adaptation for the mechanical demands of flight

bio-Another unusual tissue has been found insquat, armored dinosaurs called anky-losaurs Examining the bony plates calledscutes, Sander and Torsten Scheyer, also ofthe University of Bonn, found bundles ofstructural f ibers

arranged parallel,perpendicularly, andobliquely to the scute

surface—a light, strong design that wouldhave resisted impacts from all directions,they speculate in next month’s issue of the

Journal of Vertebrate Paleontology “It’s a

highly developed, composite material, like abulletproof vest, that would prevent penetra-tion of sharp objects,” Sander says

Histology can also be used to test theses about the function of bizarre struc-tures that no longer exist in the world

hypo-Stegosaurus plates have long attracted

attention, and a prevalent idea is that theywere used to regulate body temperature

Horner, Padian, de Ricqlès, and RussellMain, now a graduate student at HarvardUniversity, decided to test that idea Aftermaking thin sections of stegosaur platesand the smaller scutes of related dinosaurs,the team discovered that Stegosaur plateshad evolved simply by expanding the keel

of scutes “We saw nothing special aboutthe stegosaur plates” that would be anadaptation for thermoregulation, Mainsays Moreover, structures originally de-scribed as blood vessels probably weren’t

The plates were probably used instead forspecies recognition, the group proposes in

a paper in press at Paleobiology.

Indirectly, bone histology can even shedlight on long-vanished animals’ behavior Cu-rious about whether baby hadrosaurs wouldhave stayed in the nest or struck out on theirown after hatching, Horner’s team looked at

the bone tissue of Maiasaura embryos, as well

as embryos of alligators and ratite birds “We

didn’t have much evidence til we looked at the histology

un-of the bones,” Horner says.Unlike the ossified bones ofalligators and ostriches, the tis-sue at the end of the hadrosaurlimb bones consisted of calci-fied cartilage, suggesting thathatchlings couldn’t walk im-mediately They reported these

findings in the Journal of tebrate Paleontology in 2000.

Ver-Walking, running, ing, flying: Bones actively respond to thestresses of these and other physical activi-ties On the one hand, this can complicatethe interpretation of bone tissue when re-searchers are trying to establish growthrates But because physical activity affectsbone, it may also be possible to extract thathistory from bone tissue, for example bystudying the orientation of the strutlike tra-becular tissue inside bones, which is oftenoriented perpendicularly to the major axis

jump-of strain “It’s tricky and requires a certainamount of interpretation,” cautions JohnHutchinson of the Royal Veterinary Col-lege in London “There’s still a lot of workthat needs to be done in modern animals tosee how strain impacts bone remodeling.”

A good amount of that work is going on.For example, Main is studying goats to de-termine how biomechanics affects theirbone histology He hopes to find signals thatcould enable fossils to reveal posture,among other details Other researchers areseeking similar clues in alligators, croco-diles, and birds “Modern animals are some

of the great unsung heroes of dinosaur ontology,” says Curry Rogers

pale-Better known heroes are playing a keyrole too, especially when they are abundant.Horner, for example, continues to mine arich deposit of hadrosaurs, with individuals

of all ages and sizes “We’re cutting dreds and hundreds of slides,” says Horner,who has a technician working on histologyfull-time Once his group and others naildown what’s normal for bone tissues, theymay be able to probe the many influencesthat affect bone, extracting informationabout sexual dimorphism, climate, gait, andmuch else “We’ve just begun to scratch thesurface,” says Chinsamy-Turan

hun-–ERIKSTOKSTAD

Dino tissue Vessel-rich Troodon leg bone (above) grew faster than a riblike Deinonychus bone (inset).

Hepatitis B virus infects hundreds of millions

of people worldwide, causing jaundice,

fa-tigue, liver damage, and joint pain More

ominously, investigators have indicted it in

another role: as co-conspirator in a

far-ranging case they’ve been building for years

linking chronic inflammation and cancer

Re-searchers have long known that patients with

persistent hepatitis B infections experience

inflammation and scarring of liver tissue and

an increased risk of liver cancer Other

sources of chronic inflammation, including

the ulcer-causing bacterium Helicobacter

py-lori and an immune disorder known as

ulcer-ative colitis, predispose patients to cancers of

the stomach and colon

Based on their experience with these

diseases, researchers estimate that

inflamma-tion contributes to the development of at least

15% of all cancers Much less clear, however,

is exactly how it does its dirty work The

inflammation-cancer connection is especially

puzzling in light of other work suggesting

that in some circumstances the immune

sys-tem, which sustains inflammation, has the

op-posite effect: inhibiting tumor development

A flurry of results published over the past few

months may now be resolving the mystery

The new work has implicated an

inflam-mation-induced protein called NF-κB as a

key player It is an intermediary in promoting

the cellular changes leading to the

uncon-trolled growth of cancer cells and also to later

changes that help metastatic cells escape

from the original tumor and spread to new

sites in the body

Other studies have pointed to a source of

trouble in the inflammatory cells that creep

into a growing tumor, suggesting that they

produce numerous substances that can

con-tribute to tumor growth and survival,

includ-ing some that might trigger increased NF-κB

activity Research into the

inflammation-cancer link “is a very exciting area that is

de-veloping very rapidly,” says Raymond

DuBois of Vanderbilt University Medical

Center in Nashville, Tennessee

The excitement stems in part from the

ex-pectation that this emerging understanding

could lead to improved cancer prevention and

treatment Epidemiological studies have

al-ready shown that people who regularly take

NSAIDs—nonsteroidal anti-inflammatory

drugs—have a lower risk of developing

can-cer than people who don’t take the drugs Butthe efficacy of NSAIDs is not ideal, and thefirst generation of these drugs, such as as-pirin, can cause life-threatening stomach ul-cers Now, even the newer NSAIDs—the so-called COX-2 inhibitors, which were de-signed to avoid that side effect—may haveproblems: On 30 September, the pharmaceu-tical company Merck removed its blockbusterCOX-2 inhibitor Vioxx from the market be-cause it increased patients’ risks of havingheart attacks and strokes

Researchers hope that if they learn howchronic inflammation leads to cancer, theywill be able to design new drugs that counterits effects The information may also aid thedevelopment of vaccines and other strategies

to enhance immune attacks on tumors But

researchers will have to toe a fine line: Thenew work suggests that some approachesmight enhance the growth of tumors ratherthan kill them “It’s a complicated field, butit’s extremely important,” says Albert Bald-win of the University of North CarolinaSchool of Medicine in Chapel Hill (UNC)

Jekyll-and-Hyde macrophages

One line of evidence linking inflammation tocancer comes from the somewhat surprisingfinding that immune cells can foster tumordevelopment Lisa Coussens, Douglas Hana-han, and their colleagues at the University ofCalifornia School of Medicine in San Fran-cisco have found that this can happen in amodel of skin cancer they devised in micethat ordinarily develop invasive skin carcino-mas by age 1

The tumors also show infiltration by mune cells, an inflammatory response the re-searchers initially viewed in a positive light asthe body’s attempt to destroy the tumor cells.But when they turned down the immune re-sponse by crossing their mice with anotherstrain lacking a particular type of T immunecell, the resulting animals showed reduced tu-mor formation rather than the expected in-crease “If we abate inflammation, we abatecancer development,” Coussens says

im-Similarly, Jeffrey Pollard and his leagues at Albert Einstein College of Med-icine in New York City have evidence thatinfiltration with inflammatory cells calledmacrophages promotes both the develop-ment of breast cancers and their eventualspread to other sites in the body The teamshowed this by knocking out the gene forcolony-stimulating factor 1 (CSF-1), a so-called cytokine that attracts macrophages,

col-in a mouse stracol-in genetically engcol-ineered todevelop mammary cancers

Compared to controls, Pollard says,

“the incidence and rate of early [tumor]growth in the resulting animals was notchanged, … but progression to malignancywas slowed, and there was almost nometastasis.” Consistent with this, Pollard’steam and others have also shown that CSF-1 levels are elevated in human breast,ovarian, and uterine cancers and that thiselevation correlates with a poor prognosis.Why the immune system can promotecancer development at times and in other cas-

es help keep it in check remains unclear Butactivating macrophages appears to be espe-cially dangerous; many of the weapons theseimmune system scavengers release can pro-mote cancer The noxious substances includehighly reactive forms of oxygen that cancause carcinogenic mutations Cells are espe-cially vulnerable to these assaults when theyare dividing rapidly, as may be the case when

Inflammation and Cancer:

The Link Grows Stronger

Research into a long-suspected association between chronic inflammation and cancer

reveals how the immune system may be abetting tumors

C a n c e r Re s e a r c h

Macrophage

Reactive oxygen Cytokines (TNF-α) Growth factors Angiogenic factors Proteases

NF-κB

NF-κB IκB IκB

ATP

CSF-1

Degraded

COX-2 and Bcl

IKK

P

Vicious cycle Macrophages produce several

sub-stances that can enhance tumor growth, including

target tissue cells and in macrophages themselves

Tumor cells produce substances such as CSF-1and COX-2 that give a further boost to inflam-matory processes, as well as proteins such asBcl that inhibit apoptosis

966

NE W S FO C U S

tissues try to repair the damage caused by

vi-ral or bacterial infections Macrophages also

produce growth factors, enzymes that can

help cancer cells escape from tumors and

mi-grate through the body, and still other

pro-teins that stimulate the formation of blood

vessels needed for tumor growth

Inflammation hallmark implicated

Although the overall picture is far from

com-plete, recent evidence has now firmly tied

one protein, NF-κB, into the

cancer-promot-ing action of inflammatory cells Suspicion

fell on the protein several years ago It is

highly active in both inflammatory cells, such

as macrophages, and in other cells of

in-flamed tissues Indeed, NF-κB activity is

“al-most a hallmark of inflammation,” says a

re-searcher involved in this work, Yinon

Ben-Neriah of Hebrew University-Hadassah

Med-ical School in Jerusalem, Israel

Researchers have also found that the

pro-tein is abnormally active in some cancers and

that that portends a bad prognosis This may

be due to the fact that NF-κB activity leads to

inhibition of the programmed cell death

(apoptosis) that can eliminate defective cells,

thus contributing both to cancer development

and resistance to drug and radiation therapies

But the case against NF-κB was largely

circumstantial—until this summer, when

Michael Karin and his colleagues at the

University of California, San Diego, helped

uncover a smoking gun They found that

the protein contributes to cancer

develop-ment in two distinct ways (The results

ap-peared in the 6 August issue of Cell.)

For these experiments, the researchers

turned to a mouse model of colitis-associated

cancer developed about 8 years ago by a

team led by Isao Okayasu of Kitasato

Uni-versity School of Medicine in Japan This

involves giving young mice a single

injec-tion of a cancer-triggering chemical and

then dosing them repeatedly with a salt that

irritates the intestinal lining, causing

chron-ic inflammation of the colon This treatment

produces numerous colon tumors

To test the involvement of NF-κB, the

Karin team used two different strains of

ge-netically altered mice; in one, the gene for an

essential NF-κB activator enzyme called

IKKβ had been knocked out only in the cells

that give rise to macrophages, and in the

sec-ond strain the same gene was knocked out in

the epithelial cells lining their intestines In

the absence of the IKKβ gene, neither cell

type could activate its NF-κB

Loss of NF-κB function in macrophages

reduced tumor incidence in the animals by

about 50%, and the tumors that did form

were smaller than those in controls Further

work traced this reduced tumor growth to loss

of growth factors produced by inflammatory

cells in response to NF-κB activation

The picture was different in mice whoseintestinal epithelial cells couldn’t activate NF-κB These animals “had an 80% reduc-tion in tumor incidence,” Karin says Thatdifference was not due to reduced intestinalinflammation because it was more severe

in the knockouts than in the controls stead, with NF-κB eliminated, apoptosiswas no longer inhibited in intestinal cells,which presumably helped the knockout ani-mals eliminate cells with cancer-promotingmutations “Apoptosis is sort of like a safe-

In-ty mechanism,” Karin explains “It makessure tumor formation is low.”

Something similar appears to be going on

in inflammation-associated liver cancer, asdescribed by Ben-Neriah and his colleagues

in Nature (published online on 25 August).

These researchers started with a geneticallyaltered strain of mice that develop severe liver inflammation—and cancer Then theymade a second genetic alteration, adding agene for a natural NF-κB inhibitor called

IκB that carried a regulatory sequence allowing the gene to be turned off at will bygiving the animals an antibiotic

When the IκB gene was on—and NF-κB was inactive—the livers of themice showed early precancerous changessuch as increased cell division and the for-mation of small growths called adenomasfor 7 months But compared to controls,ver y few of those adenomas—about10%—progressed to full-fledged cancers

When the researchers then turned the IκBgene off, Ben-Neriah says, “we started to seetumors as if there had been no [earlier]

NF-κB switch-off.” This suggests that theprotein is not required for the early changesthat put cells on the path to cancerous trans-

formation but is required for later progression

to malignancy Again, it apparently promotesliver cancer development by inhibiting apop-tosis

What’s more, NF-κB may be needed forone of the most dangerous features of cancercells: their ability to spread throughout thebody and seed new tumors Some of this

work, described in the August Journal of Clinical Investigation, comes from a team led

by Thomas Wirth of the University of Ulm,Germany Working with cultured mammaryepithelial cells that had been transformedwith the cancer-causing Ras oncogene, the re-searchers showed that the NF-κB protein isneeded for something called the epithelial-mesenchymal transition, in which normally

sedentary epithelial cells undergo changesthat allow them to migrate

Consistent with those f indings, theGerman team also showed that cells inwhich NF-κB had been inhibited by addi-tion of an active IκB gene form far fewerlung metastases when injected into mice

“NF-κB may be involved both early andlate” in cancer development, Wirth says If

so, it would be a good target for potentialtherapeutic drugs

Karin’s team provides further tion that inflammation-induced NF-κB activity is needed for growth of tumormetastases To mimic metastatic growth,they injected colon cancer cells into miceand observed them forming metastases inthe animals’ lungs When the researchersalso injected a bacterial lipopolysaccharide(LPS) to stimulate an inflammatory re-sponse in the mice, the metastases “prettymuch doubled in size,” Karin says That re-sponse also required a protein called tumor

confirma-Metastasis link Breast cancer cells producing CSF-1 (top) contain copious numbers of

macrophages (brown stain) and produce numerous lung metastases Cells that don’t make theprotein (bottom) contain few macrophages and show little tendency to metastasize

NE W S FO C U S

968

necrosis factor–α (TNF-α) that’s made by

macrophages But when the researchers

performed the same experiment with

tu-mor cells bearing an NF-κB inhibitor, the

tumors actually shrank following LPS

in-jection due to increased apoptosis

Although much evidence supports the

idea that NF-κB promotes cancer by

in-hibiting apoptosis, it may tribute in numerous other ways

con-as well Some may be direct growths of the protein’s role ininflammation The genes activat-

out-ed by NF-κB include the onethat makes COX-2, an enzymeneeded for the synthesis of apro-inflammatory compoundcalled PGE-2 This can bring inmore immune cells to maintainthe inflammation and furtherprod tumor growth COX-2 alsopromotes blood vessel growth

Actions such as these may plain why COX-2 inhibitors havecancer-fighting effects

ex-Some NF-κB actions may beindependent of inflammation, however Al-though NF-κB activity is not necessary forcancerous transformation by Ras, Baldwinsays, it does foster cancer growth, similar

to what Karin’s team found in their ments But when Ras activates NF-κB, theconsequences may be different than when

experi-TNF-α or other inflammatory factors do Inwork published in the 15 October issue of

Cancer Cell, Baldwin and his colleagues

an-alyzed the genes turned on by NF-κB lowing its activation by Ras The genes af-fected included several that make growth-promoting proteins, but for reasons not yetunderstood, none of the inflammation-promoting genes typically activated in re-sponse to TNF-α Although this fosterscancer growth, NF-κB activity in response

fol-to Ras apparently activates a different set

of genes than when TNF-α is the activator However NF-κB works, it’s looking moreand more like a good target for anticancerdrugs “There are definitely ways to take ad-vantage of this,” Karin predicts The pharma-ceutical industry is currently working to de-velop NF-κB inhibitors, and even some low-tech compounds such as the active ingredi-ents in green tea and red wine, which arethought to have anticancer properties, areturning out to act on the protein “Almostevery cancer preventive is an NF-κB in-hibitor,” Baldwin says –JEANMARX

To the unknowing eye, the reeds growing

along the Yangtze River near Shanghai—a

burst of green in the summer, with tips that

turn golden brown in the fall—belong on

pic-ture postcards But ecologists know them as a

biological experiment run amok The salt

marsh grass Spartina alterniflora, a native of

eastern North America introduced in 1979 to

check erosion, has now spread across

south-eastern China, choking estuaries, crowding

out native grasses, and reducing feed and

habitat for fish and migratory birds

Across the Pacific, the tiny holes in the

bark of maples, willows, and elms in New

York, New Jersey, and Illinois come from the

Asian longhorned beetle Anoplophora

glabripennis is an unwelcome hitchhiker

from China that most likely arrived in the

United States a decade ago aboard wooden

shipping crates Unless checked, the beetle

threatens to bore its way through billions of

dollars’ worth of valuable timber, shade, and

maple syrup trees

Different continents, different species,

dif-ferent routes of introduction—but a common

problem Since China opened its doors to the

West in the early 1980s, its burgeoning tradewith the United States (see graph) has meantmore marine organisms in ballast water andinsects in packing crates accidentally trans-ported across the Pacific in both directions

The influx has been supplemented by tional introductions by commercial U.S nurs-eries and horticultural collectors looking forexotic Chinese specimens Many of thesespecies flourish thanks to similar habitats andclimates in the two countries The result, says

inten-Li Bo, a plant ecologist at Shanghai’s FudanUniversity, is that “American species can easi-

ly get established in China, and Chinesespecies can easily get established in America.”

Many introduced species never expandbeyond a beachhead, and most horticulturalspecies are well behaved But for reasons thatare still being debated, some introductionslead to ecological disaster This summer Chi-nese and U.S scientists held two meetings*in

hopes of sharing information and ing strategies that might help avert dire con-sequences from this two-way traffic “There

develop-is a chance we could stop thdevelop-is wave of sive species exchange between China andthe United States” through stricter inspec-tions and stiffer regulations, believes PeterAlpert, a plant ecologist at the University ofMassachusetts, Amherst But he says “there

inva-Expanding Trade With China

Creates Ecological Backlash

Scientists in the United States and China are scrambling to cope with an unintended

consequence of increasing economic ties—a two-way flow of unwelcome plants and animals

I n va s i v e S p e c i e s

Partners in crime Inflammatory cells such as the macrophages

shown here are turning out to boost tumor growth and spread

Invaders Arkansas rivers are filling with carp

from China originally imported for aquafarming

*Beijing International Symposium on Biological

Inva-sions, 8–15 June 2004 Biological Invasions: SpeciesExchanges between Eastern Asia and North America,Portland, Oregon, 2 August 2004, held in conjunctionwith the Ecological Society of America meeting

are as yet no effective policies to control

species exchanges.”

Proceed with caution

Controlling invasives is already a major

headache for both countries David

Pi-mentel, an entomologist and systems

ecolo-gist at Cornell University in Ithaca, New

York, concluded in 2000 that invasive

species cost the United States more than

$137 billion per year A similar study

pre-sented this summer at the Beijing

sympo-sium by Xu Haigen of the Nanjing Institute

of Environmental Sciences, working with

colleagues there and at the Nanjing Forestry

University, concluded that invasives had

caused $2.4 billion in damages to eight

ma-jor Chinese industries alone Many

re-searchers on both sides say these numbers

are probably low and certain to rise

The first bilateral efforts to control

inva-sive species in Asia and North America

occurred in the 1980s, when U.S

scien-tists visited China in search of natural

predators for species imported as

horti-cultural ornamentals in the late 19th and

early 20th centuries Many of these, such

as Chinese privet, are now widespread in

North America Those initial surveys led to

the creation of a Sino-American Biological

Control Laboratory in Beijing in 1989 to

identify and exchange insects that eat

inva-sive plant species

One apparent success story is a leaf beetle

(Diorhabda elongata) native to Xianjiang

Province It attacks tamarisks (Tamarix

ramo-sissima), an invasive Eurasian species that

has displaced native plants along rivers and in

arid regions of the American west The beetle

was released 7 years ago, and it “looks like it

will be a terrific success” in reducing the

spread of tamarisk, says Robert Pemberton,

an entomologist at the U.S Department of

Agriculture’s (USDA’s) Agricultural Research

Service in Fort Lauderdale, Florida So too

does the 1980 introduction into China of two

beetles, originally from South America, to

control the spread of water hyacinth

Introducing insects to eat invasive plants

is a laborious and risky process The insects

must be carefully screened to ensure that they

don’t attack native plant species, as happened

with the weevil Rhinocyllus conicus, brought

into Canada and the United States from

Eu-rope in the late 1960s to battle exotic thistles

It turned out to have a taste for native thistles

as well Ding Jianqing, an entomology

post-doc at Cornell University who previously

headed a biocontrol lab at the Chinese

Acad-emy of Agricultural Sciences, says that

con-cerns about unintended consequences have

limited the number of introductions in the

past 15 years to just four predatory insects

from the United States to China, and only

three in the opposite direction

In situations in which eradication is cult or impossible, scientists would settle forpreventing further spread of the invasiveplant But they are hoping that early interven-tion may actually stem the infestation of theAsian longhorned beetle USDA researchershave been cooperating with their Chinesecounterparts to develop better methods to de-tect tree infestation early as well as to identify

diffi-possible lures to attract and kill the beetles

But this is still a work in progress

Similarly, Duane Chapman, a fisheriesbiologist at the U.S Geological Survey’s Co-lumbia Environmental Research Center inMissouri, is struggling to reduce damagefrom bighead and silver carp imported fromTaiwan in the 1970s for aquaculture farms

in Arkansas The carp escaped and are nowcrowding out native species Chapman andChinese counterparts are summarizing andtranslating Chinese studies of the carp intheir native habitat, information which maylead to more effective control strategies

A closer look

Despite good working relationships betweenindividual scientists, U.S and Chinese re-searchers say that they won’t make a seriousdent in controlling invasive species untiltheir governments make it a higher priority

The United States spent $1 billion this year

on efforts such as inspections of cargo andbaggage for accidental introductions andvetting deliberately imported species A

2002 National Academy of Sciences (NAS)study says that USDA’s Animal and PlantHealth Inspection Service (APHIS) inter-cepts more than 53,000 arthropods,pathogens, and noxious plants each year

But that’s only a tiny fraction of what needs

to be done, say scientists Only 2% of cargoshipments are checked, they note And al-though packing crates infested with theAsian longhorned beetle have been stopped

at ports in 17 states, according to USDA,enough slipped through to start an invasion.There’s also growing concern about de-liberately introduced horticultural speci-

mens “We’re going to have to take a muchmore aggressive position on evaluatingwhether a species can become a threat toany aspect of the U.S biota prior to its re-lease into the marketplace,” says RichardMack, an ecologist at Washington State Uni-versity in Pullman Mack chaired the panel

that produced the 2002 NAS report ing Invasions of Nonindigenous Plants and Plant Pests Its recommendations are being

Predict-discussed by scientists and federal officials,

he says, “but I don’t see any strong ment yet” to implement them

move-Fudan’s Li says China has only recentlybegun to conduct inspections and implementquarantine requirements Last year the gov-ernment published a “black list” of speciesnot to be brought into the country, but itsimply covered known problems, such as

Spartina alterniflora This month, however,

the Chinese Academy of Agricultural ences, with the support of the ministries ofScience and Technology and of Agriculture,

Sci-is holding a workshop to develop a nationalinvasive species strategy that would coverprevention, early detection, and on-the-ground management That strategy shouldgive scientists on both sides of the Pacificmore ammunition to battle an increasing on-slaught of invasive species

Along for the ride These salt marsh grass and longhorned beetle species have expanded their

habitats thanks to increased trade between the United States and China

Remediating Rocky Flats

I N THEIR P OLICY F ORUM “A VOIDING DESTRUC

-tive remediation at DOE sites” (12 Mar., p

1615), F W Whicker and colleagues applaud

the “risk-based” cleanup of U.S Department

of Energy (DOE) sites and point to Rocky

Flats near Denver as a success story

Unfortunately, various assumptions about risk

as well as certain features of the Rocky Flats

cleanup show the risk-based approach to be

seriously flawed

Those who assess risk and set radiation

exposure standards have systematically

excluded affected populations from every step

of the process Risk calculation, particularly as

encoded in U.S standards, is weighted against

the most vulnerable populations

Rocky Flats will become a wildlife refuge

after cleanup Thus, the

plutonium-contami-nated site is being cleaned on the surface to

protect wildlife refuge workers The DOE will

rely on not-yet-specified institutional controls

to contain larger quantities left below a depth

of 3 feet The National Academy of Sciences

says such controls will not last (1) Given the

24,400-year half-life of plutonium, the Rocky

Flats cleanup is a short-term response to a

long-term problem Whicker et al say “natural

attenuation” will take care of smaller

concen-trations left behind But plutonium left in the

environment constitutes an essentially

perma-nent danger in particles too small to see but

not too small to inhale, ingest, or otherwise

take into the body Although Whicker et al are

sanguine about wildlife, genetic effects on

such populations are poorly understood

Whicker and his colleagues say

risk-based cleanup will save money A

closed-door decision that imposed fiscal limits on

activities at Rocky Flats made cost, not

risk, the real driver for cleanup No one can

predict what costs, monetary and

other-wise, future generations may face Contrary

to Whicker and colleagues’ comments,what is happening at Rocky Flats sets apoor precedent for other sites

L E R OY M OORE

Rocky Mountain Peace and Justice Center, PostOffice Box 1156, Boulder, CO 80306, USA E-mail:leroymoore@earthlink.net

Reference

1 T Leschine et al., Long-Term Institutional Management

of U.S Department of Energy Legacy Waste Sites

(National Academies Press, Washington, DC, 2000).

Response

I N OUR P OLICY F ORUM , WE POINTED TO THE

Rocky Flats Environmental Technology Site

as a case where a portion of a contaminated but ecologically valuable grass-land ecosystem was spared from costly anddamaging remediation because of congres-sional legislation converting much of the site

plutonium-to a wildlife refuge This meant that cleanupstandards were based on a less stringentscenario, namely, a wildlife refuge worker,rather than a site resident

Moore states that “[t]hose who assess riskand set radiation exposure standards havesystematically excluded affected populationsfrom every step of the process.” In the case ofRocky Flats, a Citizens Advisory Board(www.rfcab.org/PI.html) and several otherorganizations have sought to involve andinform the local public and to provide oppor-tunities for public input to cleanup criteria andenvironmental decisions at Rocky Flats Wealso cannot agree with Moore’s blanket state-ment that “[r]isk calculation, particularly asencoded in standards, is freighted against themost vulnerable populations.” In our experi-ence, when precise knowledge is lacking,worst-case assumptions erring toward the side

of conservatism tend to be the rule, rather thanthe exception, in setting governmental radia-tion protection standards and in risk assess-ments related to development of cleanupcriteria

Indeed, we endorse most of the broad ciples about the need for effective publicinvolvement as outlined in a recent article co-

prin-authored by Moore (1) Our main argument is

that the cleanup process itself can also createhuman health risks by mobilizing contami-nants to air and water and causing construc-tion accidents, as well as producing environ-mental degradation and instability at both thecleanup and disposal sites

Moore also states that “plutonium left in theenvironment constitutes an essentially perma-nent danger…” The danger is related to theconcentrations in the environment and theamounts that get into and decay in the body, notjust the fact that it is there Plutonium from

Image not

available for

online use.

The Rocky Flats nuclear weapons plant in 1979.

The Rocky Flats site is undergoing cleanup by

the Department of Energy.

LE T T E R S

historic nuclear weapons testing fallout, and

naturally occurring radioactivity exist in soil

virtually everywhere on Earth’s surface

Potentially, there is a small risk from this

natural radioactivity and from the very small

amounts of plutonium that may be left in soil

after cleanup We argue that it may not be

physically or economically possible to remove

the plutonium or natural radioactivity from soil

everywhere, but even if it were, such removal

is not risk free The question thus becomes one

of where to draw the line between engineered

cleanup that can spread risks to a larger

popu-lation and leaving the material in place where

it can be effectively managed

We believe that risk-based cleanup will

save money and that, underpinned by the

honest and credible application of sound

scientific knowledge, it can prevent

unjusti-fied soil excavation, preserve valuable

ecosys-tems, and provide ample protection of human

health and the environment Also, federal

commitments are in place that ensure periodic

assessment of site conditions and the goal of

protecting future generations (2).

F W W HICKER , 1 T G H INTON , 2

M M M AC D ONELL , 3 J E P INDER III, 1

L J H ABEGGER 3

Health Sciences, Colorado State University, Fort

Savannah River Ecology Laboratory, Aiken, SC 29802,

National Laboratory, Argonne, IL 60439, USA

References

1 L Ledwidge et al., Health Phys 87, 293 (2004).

2 U.S Environmental Protection Agency,

“Compre-hensive five-year review guidance” (EPA 540R01007,

EPA, Washington, DC, June 2001).

Net Environmental

Benefit Analysis

I N THEIR P OLICY F ORUM “A VOIDING

destructive remediation at DOE sites” (12

Mar., p 1615), F W Whicker et al state,

“[W]e are aware of no specific protocol or

set of criteria to identify and promote the

preservation of ecologically valuable, but

slightly contaminated, sites” and “[i]f the

criteria of size, ecological value, current

risk associated with land use, and projected

risk reduction through natural attenuation

were established, [the policy of risk-based

end states] could be more uniformly

administered.” A decision framework for

comparing net environmental benefits of

multiple risk management alternatives for

chemical contamination, including

remedi-ation, ecological restorremedi-ation, and natural

attenuation, is available, although it has not

been endorsed by DOE (1, 2)

Principles of net environmental benefit

analysis (NEBA) have been used in the

context of oil spill remediation since the

Exxon-Valdez spill (3), and comparisons of

ecological states (i.e., injured and restored) areregularly undertaken in determinations ofcompensatory restoration in Natural Resource

Damage Assessments (4) NEBA is consistent

with guidance from the U.S EnvironmentalProtection Agency to “weigh… ecologicaleffects of active remediation alternatives andpassive alternatives when selecting a final

response” [(5), (pp 6–7)] and with DOE’s

emphasis on risk-based end-states Thecomparison of alternative ecological statesshould be supported by additional R&D ofnonmonetary valuation methods such as

Habitat Equivalency Analysis (4), vative (6) ecological exposure-response

nonconser-models for chemical contamination and ical disturbance, and dynamics of ecologicalrecovery Moreover, even if future land useplans permit residential developments on landthat is slightly contaminated, it may be appro-priate to balance the reduction of ecologicalversus human risk, including risk from

phys-proposed remediation (7)

R EBECCA A E FROYMSON

Environmental Sciences Division, Oak RidgeNational Laboratory, MS 6036, Oak Ridge, TN

37831, USA E-mail: efroymsonra@ornl.gov

References and Notes

1 R.A Efroymson et al.,“A framework for net environmental

benefit analysis for remediation or restoration of leum-contaminated sites” (ORNL/TM-2003/17, Oak Ridge National Laboratory, Oak Ridge, TN, January 2003) (available at www.esd.ornl.gov/programs/ecorisk/ documents/NEBA-petrol-s-report-RE.pdf).

petro-2 R A Efroymson et al., Environ Manage 34, 315 (2004).

3 NOAA Hazardous Materials Response Branch,

“Excavation and rock washing treatment technology: net environmental benefit analysis” (National Oceanic and Atmospheric Administration, Seattle, WA, 1990).

4 NOAA Damage Assessment and Restoration Program,

“Habitat equivalency analysis: an overview” (National Oceanic and Atmospheric Administration, Seattle, WA, 2000).

5 S D Luftig, “Issuance of final guidance: ecological risk assessment and risk management principles for Superfund sites” (OSWER Directive 9285.7-28, Office of Emergency and Remedial Response, U.S Environmental Protection Agency, Washington, DC, 1990).

6 Nonconservative so that unbiased comparisons can

be made.

7 G W Suter II et al., Risk Anal 15, 221 (1995).

Response

W E APPRECIATE E FROYMSON ’ S COMMENTS

concerning the application of “net

environ-Letters to the Editor

Letters (~300 words) discuss material published

in Science in the previous 6 months or issues of

general interest They can be submittedthrough the Web (www.submit2science.org) or

by regular mail (1200 New York Ave., NW,Washington, DC 20005, USA) Letters are notacknowledged upon receipt, nor are authorsgenerally consulted before publication.Whether published in full or in part, letters aresubject to editing for clarity and space

LE T T E R S

mental benefit analysis” for making decisions

on the management of contaminated sites

within the DOE nuclear weapons complex

We are not presently in a position to offer

critique or endorsement of the specific

frame-works and approaches referenced by

Efroymson However, we encourage DOE to

investigate the merits of such concepts with

the goal of achieving cleanup decisions that

provide optimal protection of both human

health and environmental quality

F.W.W HICKER , 1 T G H INTON , 2 M M M AC D ONELL , 3

J E P INDER III, 1 L J H ABEGGER 3

Health Sciences, Colorado State University, Fort

Savannah River Ecology Laboratory, Aiken, SC 29802,

National Laboratory, Argonne, IL 60439, USA

Support for Steiger’s

Policies

J OCELYN K AISER ’ S ARTICLE ON W ILLIAM

Steiger, the point person on international

health for U.S Health and Human Services

(HHS) Secretary Tommy Thompson (“The

man behind the memos,” News Focus, 10

Sept., p 1552), is decidedly one-sided She

interviewed me at length, and although Isupport his policies and admire his profes-sionalism, none of my views, or those ofother supporters, are mentioned

Scientists cry “academic freedom”

when their travel is cut, as happened withthe AIDS conference in Bangkok Steiger,

as well as journalists and health experts,knew this conference had become unruly

and unproductive (1)

Kaiser writes that Steiger’s support of theBush administration’s “controversial” position

on abstinence in HIV prevention programs

“ruffled feathers” among researchers Sheneglects to mention how promotion of ab-stinence is dramatically reducing HIV/AIDS

infections in Uganda (2)

Furthermore, Steiger is doing his job whenrefusing to fund any conference that under-mines the administration’s approach toprocuring safe and effective AIDS drugs

Kaiser fails to mention how the tion’s position has been supported by theWorld Health Organization’s (WHO) removal

administra-of five AIDS drugs from its list because administra-ofunproven quality South Africa has banned atleast one of the drugs that the Administrationrefused to buy until tested

The fact that Steiger approves HHS staffinvolvement in WHO activities may upset

some researchers Yet all government agenciesroutinely approve which researchers attendmeetings as U.S representatives It is Steiger’sjob to coordinate the U.S position with inputsfrom many agencies with expertise mostappropriate for any particular conference Hedoes this job in the best interests of thecountry, and does it well

C AROL C A DELMAN

Hudson Institute, Center for Science in PublicPolicy, 1015 18th Street, NW, Suite 300,Washington, DC 20036, USA

References

1 For example, see these op-ed pieces: S Mallaby, “AIDS

activists misfiring,” Washington Post, 18 July 2004, p A17; L Garrett, “Bragging in Bangkok,” N.Y Times, 16

D AN F ERBER ’ S N EWS F OCUS “D EAD ZONE FIX

not a dead issue” (10 Sept., p 1557) gives toomuch credence to assertions that reducingnitrogen pollution would not shrink the extent

of hypoxia in the Gulf of Mexico The EPAreport focuses solely on ratios of concentra-tions of dissolved inorganic nitrogen (DIN)

5 NOVEMBER 2004 VOL 306 SCIENCE www.sciencemag.org978

and dissolved inorganic phosphorus (DIP),

mainly in the lower Mississippi River, to

suggest that inputs of phosphorus could

control phytoplankton production on the

continental shelf and, thus, the scale of

summer hypoxia

Decomposition of plankton biomass along

the inner shelf west of the river mouth

depletes dissolved oxygen in the denser

bottom waters In the spring and early

summer, plankton biomass accumulates in

surface waters enriched with DIN but often

with extremely low DIP concentrations If

there were severe phosphorus limitation, howcan this biomass be grown? Almost certainly,

it is because there are other phosphorussources, including recycling from organicmaterial, large reservoirs in bottom sediments,and the deeper Gulf of Mexico Surface-waterorganisms rapidly take up any DIP supplied,keeping DIP concentrations very low, indi-cating that DIN:DIP ratios are notoriously

unreliable indicators of nutrient limitation (1).

These other phosphorus sources may beunconnected or indirectly connected toseasonal river inputs and, thus, may prove

difficult if not impossible to control.More comprehensive assessment stronglyindicates that nutrient pollution, particularly

in the form of nitrogen from MississippiBasin agriculture, is the principal cause ofhypoxia and that improved agricultural prac-tices coupled with restoration of wetlands in

the river basin are the only solutions (2).

With better understanding, it might proveeffective to reduce both nitrogen and phos-phorus inputs as is being pursued elsewhere

(3), but a solid body of science indicates that

substantial reductions in nitrogen loads are

required to reduce the extent of hypoxia (4)

and, further, that curtailing phosphoruswithout reducing nitrogen inputs might actu-ally extend the effects of overenrichment to a

1 W K Dodds, J N Am Benthol Soc 22,171 (2003).

2 N N Rabalais, R E Turner, D Scavia, Bioscience 52,

129 (2002).

3 D F Boesch, Estuaries 25, 886 (2002).

4 D Scavia, D Justi´c,V J Bierman, Estuaries 27, 419 (2004).

5 H W Paerl, L M Valdes, A R Joyner, M F Piehler,

Environ Sci Technol 38, 3068 (2004).

6 J D Hagy, W R Boynton, C W Keefe, K V Wood,

Estuaries 27, 634 (2004).

TECHNICAL COMMENT ABSTRACTS

COMMENT ON“The Origins of Genome Complexity”

Vincent Daubin, Nancy A Moran

Lynch and Conery (Reports, 21 November 2003, p 1401) claim a universal relation between genetic population

size and genomic size and complexity, but their treatment of bacteria is invalid.Their estimates of polymorphism

for bacteria largely reflect evolutionary divergence of independent clonal lineages rather than selection efficiency

within cohesive species An alternative measure of genetic drift shows no relation to genome size

Full text at www.sciencemag.org/cgi/content/full/306/5698/978a

RESPONSE TOCOMMENT ON“The Origins of Genome Complexity”

Michael Lynch, John S Conery

Daubin and Moran claim that prokaryotes do not have larger effective population sizes than eukaryotes, and also

argue that genetic drift is a minor force in prokaryotic genome evolution These arguments are mutually

incon-sistent and are contrary to a substantial body of empirical and theoretical work

Full text at www.sciencemag.org/cgi/content/full/306/5698/978b

Comment on ‘‘The Origins of

Genome Complexity’’

Lynch and Conery (1) argued that genome

complexity reflects a history of genetic drift

caused by small genetic population size (Ne),

which in turn enables the spread of mildly

deleterious selfish elements and duplications

Under this argument, large organisms, which

tend to have small Ne, also exhibit larger

(complex) genomes, and the small size of

bacterial genomes reflects their position at

the extreme large end of the range for Ne

Genome size can vary among bacteria by at

least an order of magnitude, but the fraction

of noncoding DNA is consistently low

(È15%) Consequently, larger bacterial

ge-nomes are expected to result from more

selection, because they maintain more

func-tional DNA Indeed, large genome size in

bacteria commonly has been considered an

adaptation to changing environments (2–6)

Furthermore, small Nein symbiotic bacteria

appears to result in reduced genomes through

gene loss (7, 8) Thus, the relation between

Ne and genome size in bacteria is, if

anything, expected to be the opposite of that

proposed by Lynch and Conery (1)

The claimed relationship between Ne and

genome size is based on inferring Neu (where

u is the mutation rate per nucleotide) from

estimates of polymorphism using sequence

divergences among isolates (1) This

ap-proach, however, has several drawbacks,

which are especially severe in bacteria One

concern is the assumption of similar mutation

rates across bacteria, which is required if

estimates of Neu are interpreted as directly

reflecting variation in Ne Lynch and Conery

(1) justify this assumption by citing a survey

(9) that included only a single bacterial strain;

other evidence indicates that bacterial lineages

vary substantially in mutation rates (10)—as

would be expected, because their genomes

differ markedly in content of genes known to

impact mutation Thus, polymorphism levels

reflect mutational input as well as Ne

Even more problematic is the assumption

that taxonomic names correspond to

cohe-sive species, which is implicit in the use by

Lynch and Conery (1) of polymorphism data

to estimate Ne Although this assumption can

be considered approximately true for most

sexual eukaryotic species, it is often far from

valid for bacterial strains grouped under a

particular species name These usually

con-stitute phylogenetically and ecologically

distinct clusters separated by niche and

genetic boundaries; thus, genetic drift does

not act within the namedBspecies[ as a wholebut within smaller, divergent subunits Theposition of bacteria at the extreme high end ofthe range of Neu values cited by Lynch andConery (1) is based in large part on lumpingindependently evolving strains for calcula-tions of polymorphism For example, thehighest Neu value listed by Lynch andConery, for the free-living marine cyanobac-terial genus Prochlorococcus, is derived fromdivergent species with different ecologicalniches (11), dramatically different gene in-ventories reflecting habitat differences, andextreme differences in genomic base compo-sition (12) This use of divergence to calcu-late Neis comparable to using the divergence

of cat, dog, human, and mouse to calculate Nefor the species Bmammal,[ an exercise thatwould give an enormous population size

Similarly, Lynch and Conery (1) treatSalmonella enterica as a cohesive species,yet it consists of distinct, highly clonallineages living in different vertebrate hosts(13, 14) Lateral gene transfer, which isrelatively common in bacteriaEand reportedfor the S enterica data set (13)^ will alsoradically affect the calculated degree ofpolymorphism if undetected (or ignored)

The complexities of relating polymorphism

to Ne in bacteria are widely recognized,having been noted in early enzyme electro-phoresis studies, which revealed that poly-morphism levels were unexpectedly low inview of the census population sizes ofbacterial species (15)

Even if issues of population subdivision,recombination, and variation in mutationalinput could be resolved, estimates of Nebased

on polymorphism would reflect the veryrecent evolutionary past rather than the longerperiod over which genome features evolved.Major reductions in polymorphism are ob-served for recently derived pathogens relative

to ecologically generalized parental speciesEsee, for example, (16)^, but genomic traits,including genome size, are little altered

A more reliable index of genetic driftover evolutionary time is the ratio of Ka(nonsynonymous substitutions per site) to Ks(synonymous substitutions per site) for alarge set of genes, based on comparisons ofrelated species The Ka/Ksratio, which is al-most always less than one, is widely used as

an indicator of the extent of purifying lection acting to conserve coding sequencesEsee, for example, (17)^ Although Ka/Kscan

se-be elevated in particular genes as a result ofpositive selection for amino acid changes orselection conserving codon choice, a pattern

of genome-wide differences in Ka/Ks cates persistent levels of differences in Ne.Therefore, the proposal of Lynch and Conery(1) predicts a positive correlation between

indi-Ka/Ksand genome size and, thus, higher Ka/

Ks ratios in eukaryotes Another advantage

of this approach is that it is not invalidated

by variation in mutation rate among lineages

No clear relation is observed betweenaverage Ka/Ks and genome size acrossbacteria, whether all genes or a common set

of genes are used (Fig 1) Because Ksvaluescan be depressed by selection favoringcertain codons, we used different approaches

to define a genome-wide Ka/Ksbut found norelation with genome size (Fig 1 and TableS1) Furthermore, genomes known to have

(26) versus genomesize for 23 pairs ofrelated bacteria (filledcircles) and severalpairs of eukaryotes(open circles) Pairsspan the range of bac-terial genome size andinclude all available

g e n o m e p a i r s f o r

d e p e n d s o n g e n efunction, we selectedthe single-copy orthologs (defined as reciprocal unique BLAST hits) present in at least 15 of the 23genome pairs, to ensure that comparisons among genomes encompass similar gene sets Similarlack of relation was observed using the complete set of orthologs for each genome pair and using

four pairs of eukaryotic genomes (in order of increasing genome size: yeasts, nematodes, flies, andmammals) are presented Data and taxon names for bacteria are presented in the SupportingOnline Material

strong codon biases, such as E coli-S.

enterica or Bacillus species, show low Ka/

Ks, and cases known to have low codon bias,

such as Buchnera (18), have high Ka/Ks,

which indicates that purifying selection on

codon use has less effect on Ka/Ksthan does

selection on amino acid residue

As is evident in Fig 1 Eand as noted

previously based on more limited data (19)^,

average Ka/Ks can be very high in certain

bacteria, including not only symbiotic

line-ages but also others, such as several groups

of soil-dwelling bacteria At least for

sym-bionts, high Ka/Kscorresponds to very low

levels of nucleotide polymorphism (20, 21),

implicating small Neas the likely cause The

conventional view—that large Ne is typical

of bacteria—was initially established on the

basis of the limited early sequence data

available for E coli and S enterica, which,

as it turns out, show exceptionally low Ka/Ks

(Supporting Online Material) Genomic data

and polymorphism studies now suggest that

this early view was simplistic and that Ne

varies widely across bacterial groups

For comparison, we have included Ka/Ks

values calculated for eukaryotes Ederived

from genome-wide comparisons of orthologs

for fungi, mammals, nematodes, and flies

(Fig 1)^ These fall within the range of

values found in bacteria, which suggests that

there is no consistent difference in the

ef-ficacy of purifying selection between

pro-karyotes and eupro-karyotes, despite the gap

between their genome sizes

As Lynch and Conery (1) argue, Ne

probably does affect the ability to eliminate

useless or deleterious DNA However, their

model supposes that lack of effective

selec-tion will lead to genome growth because of

the dynamics of selfish elements That

supposition may be true for eukaryotes butdoes not hold for bacteria, which display amutational bias in nature favoring deletionand preventing major accumulation of so-called junk DNA (22–24) This deletionalbias may itself reflect vulnerability to selfishDNA (23), resulting from distinctive biolog-ical features of bacteria, including lack of anuclear envelope or meiotic sex Althoughbacteria undergoing severe genetic drift mayexperience outbreaks of selfish elements, theprincipal outcome of reduced population size

is the random inactivation and deletion ofgenes participating in basic cellular processes(7), ultimately resulting in an overall genomereductionEsee, for example, (25)^ Hence, themodel of Lynch and Conery does not explainthe differences in genome sizes amongbacteria Their conclusions that Neis consist-ently enormous in bacteria and that theresulting low levels of genetic drift underliethe lack of genomic and phenotypic com-plexity in these organisms are not warranted

Vincent DaubinDepartment of Biochemistry and

Molecular BiophysicsUniversity of Arizonadaubin@email.arizona.edu

Nancy A MoranDepartment of Ecology andEvolutionary BiologyUniversity of ArizonaTucson, AZ 85721, USAnmoran@email.arizona.edu

References and Notes

1 M Lynch, J S Conery, Science 302, 1401 (2003).

2 K T Konstantinidis, J M Tiedje, Proc Natl Acad.

Sci U.S.A 101, 3160 (2004).

3 K E Nelson et al., Environ Microbiol 4, 799 (2002).

4 U Dobrindt, J Hacker, Curr Opin Microbiol 4, 550 (2001).

5 H Ikeda et al., Nature Biotechnol 21, 526 (2003).

6 S D Bentley et al., Nature 417, 141 (2002).

10 H Ochman, Mol Biol Evol 20, 2091 (2003).

11 E Urbach, D J Scanlan, D L Distel, J B Waterbury,

S W Chisholm, J Mol Evol 46, 188 (1998).

12 G Rocap et al., Nature 424, 1042 (2003).

13 F S Wang, T S Whittam, R K Selander, J Bacteriol.

17 M Lynch, J S Conery, Science 290, 1151 (2000).

18 C Rispe, F Delmotte, R C van Ham, A Moya, Genome Res 14, 44 (2004).

19 H Ochman, S Elwyn, N A Moran, Proc Natl Acad Sci U.S.A 96: 12638 (1999).

20 P Abbot, N A Moran, Mol Ecol 11, 2649 (2002).

21 D J Funk, J J Wernegreen, N A Moran, Genetics

25 J Parkhill et al., Nature Genet 35, 32 (2003).

26 J M Comeron, J Mol Evol 41, 1152 (1995).

27 M Nei, T Gojobori, Mol Biol Evol 3, 418 (1986).

28 Z Yang, Comput Applic Biosci 13, 555 (1997).

29 We acknowledge A Rokas, C Seoighe, and A Cutter for sharing their sets of eukaryotic orthologous genes

in closely related eukaryote species, and H Ochman for comment on this manuscript Funding was from Department of Energy award DEFG0301ER63147 to

H Ochman and NSF award EF-0313737 to N.A.M Supporting Online Material

www.sciencemag.org/cgi/content/full/306/5698/978a/ DC1

SOM Text Table S1

29 March 2004; accepted 5 August 2004

T E C H N I C A L C O M M E N T

978a

Response to Comment on ‘‘The

Origins of Genome Complexity’’

Our study (1) argued that the long-term

genetic effective size of a population (Ne)

plays a central role in dictating the types of

genomic evolution that can occur and that

many aspects of eukaryotic genome

com-plexity may have arisen owing to a reduction

in Ne that began near the time of origin of

eukaryotes and became much more

pro-nounced in lineages of multicellular species

(1) If these general principles naturally

extend to prokaryotes, then the diminutive

genomes of prokaryotes may stem largely

from the enhanced efficiency of selection

against very mildly deleterious insertions,

and one need not resort to arguments based

on perceived constraints of cell morphology

Daubin and Moran (2), by contrast, imply

that the power of genetic drift is as great in

prokaryotes as in eukaryotes and that unique

selective forces must account for streamlined

prokaryotic genomes Their arguments,

how-ever, are inconsistent with both empirical

observation and population-genetic theory

First, contrary to the suggestion in (2),

the strong correlation between genome size

and gene number in prokaryotes is consistent

with theoretical expectations The key issue

here is why, despite the range in variation

among taxa, the average number of genes

harbored within prokaryotic genomes is so

much smaller than that in eukaryotes

Numerous results (1, 3–5) show that microbes

have gene duplication rates comparable to or

greater than those in multicellular

eukary-otes, so the small size of prokaryotic

ge-nomes can be explained only by a reduced

rate of duplicate-gene retention Theoretical

work demonstrates that a common

mecha-nism of duplicate-gene preservation in

ani-mals and plants, subfunctionalization of

preexisting functions by degenerative

muta-tions (6, 7), is nearly inoperable in

popula-tions with large effective sizes (8, 9) In

contrast, the preservation of duplicate genes

by rare beneficial mutations to new functions

is much more likely in populations of

enor-mous absolute size (prokaryotes)

Thus, if the small number of duplicate

genes retained in prokaryotes owe their

maintenance to neofunctionalization (or,

per-haps, dosage requirements), as Daubin and

Moran imply, the tight correlation between

gene number and genome size (10, 11) does

not contradict our model Rather, the

ob-served pattern is entirely consistent with our

argument that selection efficiency in

prokary-otes is typically too great for substantialmobile-element and intron proliferation—

which, in turn, means that most such genomesconsist largely of functional genes This said,although the mobile-element contribution tothe genome content of prokaryotes is almostalwaysG10%, a positive scaling with genomesize is entirely continuous with that foreukaryotic genomes (12), consistent with ourtheory The reduction in gene number inendosymbiotic prokaryotes is also consistentwith our hypothesis that random genetic drift

is a major force in genomic evolution If asubstantial fraction of genes are only weaklyfavorable to organismal fitness, then theadoption of an endosymbiotic life style wouldmagnify the likelihood of loss of such genes,because functions supplied by the host wouldreduce the selective advantage of generetention and because the reduction in Newould reduce the efficiency of selection

Second, Daubin and Moran object to theuse of standing variation at silent sites,(, as

an indicator of Nefor prokaryotes (13) Theyargue that ( reflects variation in u, themutation rate per base pair per generation,

as well as variation in Ne This point, ever, only strengthens the conclusion that Ne

how-in prokaryotes substantially exceeds that how-inmost eukaryotes (1) Many factors can in-fluence the mutation rate within and amonggenomes, but all evidence suggests that u ismuch lower in prokaryotes than in eukary-otes, increasing with generation time andnumber of germ-line cell divisions in thelatter (14, 15) A composite direct estimate

of u for the eubacterium Escherichia coli,0.36 10–9

(16), is similar to that for thearchaebacterium Sulfolobus acidocaldarius,0.26 10–9 (17), and not greatly differentfrom that for the single-celled yeast Saccha-romyces cerevisiae, 0.19  10–9

(14) Incontrast, a variety of methods suggest that u

; 22.0  10–9

in humans (18, 19) and on theorder of 2.0 to 9.0  10–9

in invertebrates(Drosophila and Caenorhabditis) and plants(14, 20, 21) This È100-fold increase in ufrom the smallest prokaryotes to the largesteukaryotes implies that the decline in Newith increasing organismal size is muchmore pronounced than the already strongdecline of ( we depicted Efigure 1 in (1)^

This is a conservative conclusion because itdoes not factor in the enhanced efficiency ofselection on silent sites in microbes relative

to multicellular species

Third, the contention by Daubin andMoran that population subdivision invali-dates application of the Neconcept is incor-rect Genome evolution is a long-term process,defined by the effective number of individuals

in the entire species, not by the size of localdemes Not only is the concept of Neentirelyvalid (provided some cohesion exists betweenvarious population segments), but it wasspecifically developed to deal with populationstructures that deviate from an ideal random-mating situation (22)

Daubin and Moran do raise an importantpoint—the inherent difficulty in definingspecies boundaries in prokaryotes (23).However, although molecular surveys ofprokaryotes reveal clades within lineagesclassically defined as species, all genealog-ical data exhibit such structure The key issue

is whether the observed patterns in otes represent inappropriate mixes of two ormore permanently separated lineages orsimple stochastic variation in lineage struc-ture, including variation among demes exhib-iting local adaptation Even in an idealpopulation, the expected degree of separation

prokary-of the two most deeply branching lineages in

a neutral gene genealogy is half the depth ofthe entire tree (24), and population subdivi-sion will induce deeper furrows in a tree

To avoid arbitrary decisions in estimating(, we took species designations as imposed

by investigators to be the natural boundaries

of analysis That could have resulted in someupwardly biased estimates of prokaryotic(,but that is also true for eukaryotic species(particularly unicellular species) To exam-ine this problem further, we have supple-

prokaryotes with recently published studies(Supporting Online Material) and with genesequence surveys that had not been previ-ously converted to(, conservatively restrict-ing analyses to individual clades when noted

by the authors The average estimate of (from 39 taxa, 0.0412 (0.0123), remains quitehigh relative to estimates in eukaryotes Inmany of these studies, the authors indicateevidence for within-species recombination

In the extreme case of Prochlorococcus, evenwhen different light-adapted clades are treat-

ed separately, average( is still 0.4748 cause the gradient in ( is already greatlybiased downwardly relative to Ne, it remainsextremely unlikely that the average Ne ofprokaryotes is anywhere near as low as that

Be-in eukaryotes

The implication by Daubin and Moranthat microbial lineages with silent-site diver-gence similar to that of phenotypically di-verse mammals must also have equivalentlevels of ecological or functional genetic

divergence fails to consider the basic

princi-ples that lead to the maintenance of

near-neutral variation Provided Neis large enough,

there is nothing to prevent the development of

very high ( while maintaining stability of

protein functions To begin to address these

issues, we have sequenced a number of

nuclear genes in global isolates within

Para-mecium species (25), whose species

desig-nations have been confirmed by laboratory

crosses The average( for silent-site variation

is 0.04, higher than that for all but a single

eukaryote in figure 1 of (1) and slightly less

than the divergence between cat and dog (24)

Unless lineages within currently described

microbial species can be shown to be

completely impervious to gene flow, there is

no reason to abandon the use of ( as a

surrogate measure of recent Ne

Although a meaningful measure of Ne

that extends beyond the applicable time scale

for( (the past 2Negenerations for haploids)

is highly desirable, we do not share the

enthusiasm of Daubin and Moran for the

ratio of replacement-site (Ka) to silent-site

(Ks) substitutions among species as an

indicator of long-term Ne Although it is

common to use Ka/Ksas a relative measure

of the width of the selective sieve (that is, as

1.0 minus the fraction of mutations that are

eliminated by selection) within species, there

are substantial problems in employing this

logic among species with dramatically

dif-ferent Ne Under this ideal interpretation, Ks

is an unbiased estimate of the mutation rate,

but selection on silent sites associated with

codon usage can cause Ks to decline with

increasing Ne, inducing a tendency for Ka/Ks

to increase

Although an increasing fraction of mildly

deleterious mutations at replacement sites is

also expected to be purged with larger Ne,

whether the reduction in Kawill exceed the

reduction in Ks, yielding the qualitative

be-havior that Daubin and Moran expect,

de-pends on (i) the distribution of mutational

effects, and (ii) the effect of selection on

linked loci As to point (i), a plausible case

can be made that most beneficial mutations

have very small effects (26), in which case

Ka could actually increase with Ne in some

range of Ne (where the increased fixation of

beneficial mutations exceeds the reduction in

purging of deleterious mutations) Recent

work supports this view in showing that a

substantial fraction of Kaseparating

metazo-an species consists of adaptive mutations

(27, 28) Because selection should be more

efficient in species with higher Ne, this effect

is likely to be even greater in prokaryotes

As to (ii), the effect of selection on linked

loci is especially pronounced in species with

major clonal phases of reproduction and is

probably the main reason that Ne scales

substantially less than linearly with absolute

population size (29) The case has been madethat with background selection on beneficialmutations, Ks could first decrease and thenincrease with increasing Ne as very mildlydeleterious mutations at silent sites are fixed

by hitchhiking with beneficial mutations atreplacement sites (30) If that is true, thenour conclusions based on silent-site diversityare even more conservative than suggestedabove At any rate, although the overall level

of adaptation is expected to generally crease with Ne, this need not be reflected in amonotonic negative scaling between Ne andbetween-species Ka/Ks

in-The arguments of Daubin and Moranabout the behavior of Ka/Ksin species withhigh versus low levels of codon bias provide

no insight into these issues, because theyneglect the fact that codon bias is a function

of both mutation and selection An absence

of codon bias cannot be taken at face value

as evidence for the absence of selection, norcan the presence of codon bias be taken asevidence of selection In addition, theseinterpretive issues aside, the data reported

in (2) have fundamental quantitative tainties Silent sites are saturated withchanges in many of their comparisons (Ks91.0 in 12 cases), and many of their Ka/Ksestimates are ratios of average Kato average

uncer-Ks Because the expected value of a ratio isunequal to the ratio of expectations, suchratios are biased with respect to what Daubinand Moran claim to be measuring Moreover,this treatment explicitly contradicts thecontention by Daubin and Moran that theuse of Ka/Kscontrols for potential variation

in mutation rates among genes This diversecombination of conceptual and analyticalproblems raises enough questions aboutstatistical validity and biological meaning

to render the indices of Daubin and Moranuninterpretable

Finally, Daubin and Moran imply thatprokaryotes harbor an endogenous mutationaldeletion bias that keeps genomes streamlinedwithout the need for direct selection Thisassumption is crucial to their line of thinking;

Mira et al (10) have argued that there is noobservable selective cost of genome size inprokaryotes However, the references cited

by Daubin and Moran do not actuallyconsider the distribution of mutationaleffects, but rather simply report on thesubstitution process Thus, as in the case ofcodon bias, they have failed to separatemutational from selective effects Althoughsome of the cited studies involve divergence

of pseudogenes, it is risky to assume thatmutational changes in such sequences areentirely neutral In species with large enough

Ne to promote codon bias, one cannot ruleout the ability of natural selection to promotedeletions over insertions in otherwise selec-tively neutral sequence It is, moreover,

becoming increasingly clear that numerouspseudogene sequences in a variety of specieshave major functional consequences (31–33).The one study that has provided an unbiasedestimate of the mutational size spectrumsuggests a substantial excess of insertionrelative to deletion mutations (34)

Although many questions regarding crobial genomic evolution remain unanswered,the idea that eukaryotes have experiencedreductions in Ne no longer seems to be indoubt Individual estimates of Ne may beinaccurate, but the general patterns are clearand, because of the intrinsic biases discussedabove, are likely to be even more pronouncedthan what we have suggested Contrary tothe assertion of Daubin and Moran, we donot suggest that random genetic drift isabsent in prokaryotes; indeed, given thelow estimates of prokaryotic Ne relative totheir enormous absolute population sizes,

mi-we are suggesting quite the opposite.Although the data in our study and asubstantial body of theoretical work supportthe idea that a reduction in the relative power

of random genetic drift from microbes tomulticellular species induces radical differ-ences in the types of genomic evolution thatcan proceed, we see no compelling reason toabandon the idea that the same principles ofpopulation genetics guide genomic evolution

in all prokaryotes and eukaryotes

Michael LynchDepartment of BiologyIndiana UniversityBloomington, IN 47405, USAmlynch@bio.indiana.edu

John S ConeryDepartment of Computer and

Information ScienceUniversity of OregonEugene, OR 97403, USA

References

1 M Lynch, J S Conery, Science 302, 1401 (2003).

2 V Daubin, N A Moran, Science 306, 978 (2004); www.sciencemag.org/cgi/content/full/306/5698/978a.

3 M Lynch, J S Conery, Science 290, 1151 (2000).

4 M Lynch, J S Conery, J Struct Funct Genom 3, 35 (2003).

5 S D Hooper, O G Berg, Mol Biol Evol 20, 945 (2003).

6 A Force et al., Genetics 151, 1531 (1999).

7 V E Prince, F B Pickett, Nature Rev Genet 3, 827 (2002).

8 M Lynch, A Force, Genetics 154, 459 (2000).

9 M Lynch, M O’Hely, B Walsh, A Force, Genetics

12 M Lynch, J S Conery, in preparation.

13 We refer to silent-site heterozygosity as (, with the observed variation being divided by two in the case of prokaryotes and four in the case of diploid eukaryotes

to estimate N e u See Supporting Online Material in (1).

T E C H N I C A L C O M M E N T

978b

14 J W Drake, B Charlesworth, D Charlesworth, J F.

Crow, Genetics 148, 1667 (1998).

15 M Lynch et al., Evolution 53, 645 (1999).

16 H Ochman, Mol Biol Evol 20, 2091 (2003).

17 D W Grogan, G T Carver, J W Drake, Proc Natl.

Acad Sci U.S.A 98, 7928 (2001).

18 F Giannelli, T Anagnostopoulos, P M Green, Am J.

21 M Lynch, Mol Biol Evol 14, 914 (1997).

22 S Wright, Evolution and the Genetics of Population.

vol 2, The Theory of Gene Frequencies (University of Chicago Press, Chicago, 1969).

23 J Maynard Smith, E J Feil, N H Smith, Bioessays 22,

30 Y Kim, Mol Biol Evol 21, 286 (2004).

31 S Hirotsune et al., Nature 423, 91 (2003).

32 K Yamada et al., Science 302, 842 (2003).

33 E S Balakirev, F J Ayala, Annu Rev Genet 37, 123 (2003).

34 D R Denver, K Morris, M Lynch, W K Thomas, Nature 430, 679 (2004)R

Supporting Online Material www.sciencemag.org/cgi/content/full/306/5698/978b/ DC1

Whether we scientists are inspired,

bored, or infuriated by philosophy,

all our theorizing and

experimenta-tion depends on particular philosophical

background assumptions This hidden

influ-ence is an acute embarrassment

to many researchers, and it is

therefore not often

acknowl-edged Such fundamental

no-tions as reality, space, time, and

causality—notions found at the

core of the scientific

enter-prise—all rely on particular

metaphysical assumptions about

the world The situation for

brain scientists is no different

The thinking of psychologists,

neuroscien-tists, clinicians, and others who care about

the subjective mind and how it relates to the

objective brain is constrained, by and large,

by terms René Descartes introduced in the

mid-17th century Everything under the sun

consists of one of two substances (hence this

sort of philosophy is termed dualism): it is

ei-ther physical stuff that has extension (res

ex-tensa) or mental stuff, thinking substance (res

cogitans; by this he meant consciousness).

Humans are unique in that they are made up

of both substances (for Descartes took the

position that animals are not conscious) Of

course, classical dualism is vehemently

re-jected by the large majority of today’s

scien-tists Nonetheless, these Cartesian terms

re-main immensely influential; they continue

to frame the modern mind-body debate,

making a resolution unlikely So argues John

Searle—the University of California,

Berkeley philosopher who put forth the

Chinese room argument against the notion

that computers can understand anything—in

his latest book, Mind: A Brief Introduction

In this short book that can easily be read

on a transatlantic flight, Searle succinctly

outlines the confusing variety of -isms that

seek to explain the relationship between

mind and brain These include property and

substance dualism, idealism, and

material-ism and its offspring (logical and

method-ological behaviorism, physicalism,

epiphe-nomenalism, functionalism, and eliminative

reductionism) Over the past 100 years, there

have been three successive dominant tific approaches to problems of the mind: be-haviorism (which denied the existence ofconsciousness), computational functional-ism (which asserted that consciousness is to

scien-the brain what a program is to acomputer), and the contempo-rary focus on brain regions andneurons involved in the genesis

of specific conscious percepts

Philosophers have no such fying framework Some respond

uni-to the existence of consciouspercepts, feelings, and thoughts(sometimes referred to asqualia) that constitute experi-enced life by denying their existence Othersaccept them as given but declare that theyare forever beyond the

pale of a reductionistic,scientific explanation

Still others try to stand them as part of thenatural order of things

under-Searle falls within the ter camp To him, “Con-sciousness is a system-level,biological feature in muchthe same way that digestion,

lat-or growth, lat-or the secretion ofbile are system-level, biologicalfeatures.” The crucial problem inunderstanding the mind is to ex-plain how mental phenomena re-late to the material substrate of thebrain In particular, what is the causal rela-tionship between the mental and the physi-cal? Refuting the entire Cartesian frame-work, Searle argues that (i) consciousness iscausally reducible to the relevant micro vari-ables (that is, to its neurobiological sub-strate), yet (ii) consciousness is not ontolog-ically reducible to these brain processes Theimplications of this syncretistic dualistic-materialistic account are profound

The first statement avers that the causalpowers of consciousness are exactly thesame as the causal powers of the neuronalcorrelates of consciousness (the NCC);

neither more nor less Consciousness ispart of the ordinary physical world; it is not

something over and above it (res cogitans

or the mind cannot act on its own, withoutthe brain) The second one implies that thefirst-person perspective of the experienc-ing subject is real and is not identical to the

objective description of the underlyingneuronal processes that are sufficient forthe conscious experiences (the third-person perspective of an external observ-er) Even though feelings are caused bybrain processes, they have their own exis-tence (As anybody who has ever sufferedfrom a tooth pain knows only too well; thesodium, potassium, calcium, and other ionssloshing around the brain that are suffi-cient for the pain are not the same as theawful feeling itself.) Searle argues—in linewith what Francis Crick and I have beenadvocating—that the inchoate science ofconsciousness needs to move from a re-search program that characterizes theNCC, through tests of their causal powers,

to an ultimate theory of consciousness It isdifficult to convey in a few sentences theforcefulness and common sense of Searle’sposition, which he labels biological natu-ralism A believer in the scientific method,

he accepts no mysticism or denial of theobvious Whether Searle has truly untan-gled the Gordian knot of the mind-bodyproblem remains to be seen But hisviews are compatible with every-thing we know about the worldand consciousness

The book also ers related topics, inparticular intentional-ity (another age-old philosophical conun-drum concerned with meaning), causation,the unconscious, and the self As a proper un-derstanding of mental causation may radical-

cov-ly contradict the traditional image that peoplehave formed of themselves throughout theages and across cultures, Searle spends twofascinating chapters on this problem It iseasy enough to state The universe is causal-

ly closed: that is, anything that happened pened because it was caused by somethingelse Given the causes, the effect had to oc-cur There is no choice When I climb and myfeet slip, I fall if the force of gravity on mybody exceeds the force that my hands exert

hap-on the rock It is as simple as that On the

oth-er hand, thoth-ere is the profound expoth-erience offreedom of will Except when drunk, hypno-tized, under some powerful emotion (such asrage), or in a similar condition, I am free atthe psychological level Nobody compelled

me to go climbing, to pick this particularroute, or to make this sequence of moves

0-19-of Philosophy

The reviewer is in the Computation and Neural

Systems Program, Division of Biology, California

Institute of Technology, Pasadena, CA 91125, USA.

Quale of red Searle asks how

the quale (or qualitative pect) of seeing red arises inthe brain

Now these are obviously absolutely

irrecon-cilable convictions Can both be true? Many

have argued that there is no freedom To this

Searle retorts, “If…I am in a restaurant and I

am confronted with a menu and the waiter

asks me what I would like, I cannot say ‘I’m

a determinist, I’ll just wait and see what

hap-pens,’ because even that utterance is only

in-telligible to me as an exercise of my free

will.” Yet where does freedom come in?

Refreshingly, Searle admits to puzzlement

(In my experience, it is rare for a philosopher

to admit not understanding something.) Why

should primates have evolved a large

deci-sion-making apparatus (most of the

pre-frontal cortex is concerned with making

deci-sions of various kinds) and the perception of

freedom of action if it is all a big illusion?

Searle does propose that quantum

indetermi-nacy may be important here However, at this

point it is mysterious how an injection of

ran-domness at the subneuronal level should lead

to indeterminacy of a useful kind at the

be-havioral level We will have to see how this

plays out over the years

Mind finishes with a chapter whose title

says it all, “Philosophy and the Scientific

World-View.” That masterful, three-page

essay should top the required-reading list

in every high school and college around the

world Although I have a few

disagree-ments with Searle, I believe that every

thinking person concerned about the mind

and its place in the world should own a

copy Easy to read, the book keeps

philo-sophical jargon to a minimum Pound per

pound, you don’t get much better value

H I S T O R Y O F S C I E N C E

Reading Between

the Lines

Charlotte Sleigh

What if Alfred Russel Wallace, not

Charles Darwin, had gone down

in history as the primary force

be-hind modern evolutionary theory? Besides

having a more ungainly name (Wallaceism,

anyone?), the science would have been

sub-tly yet powerfully altered Wallace’s greater

emphasis on what is now called

biogeogra-phy, and on interspecific (rather than

in-traspecific) relationships, might have

pro-duced an evolutionary science much more

like ecology, in both its specialist and

pop-ular senses In this form, where the

in-traspecific competitive analogies to Adam

Smith were less evident, the science might

have prospered in the Soviet Union We

might also have found that the ongoing bates about teaching evolution in schoolswere obviated, for Wallace increasinglycame to explain evolution as a

de-theistic process, operating der the guiding hand of God

un-In the event, Wallace—

whose 1858 essay on tation by selection was present-

transmu-ed to the Linnean Society at thesame time as Darwin’s ab-stract—declined a place inWestminster Cathedral along-side the “Devil’s Chaplain.” He even went so

far as to write a book entitled Darwinism,

thus effectively erasing his own claim to ashare of the priority

Because of Wallace’s religious stance,and because of his diffidence (he alsolacked a Huxley to act as spin doctor), hehas been written off by many as an also-ran

of science: a potential great who was flected from the true faith of positivism bythe unscientific Victorian temptations of

de-spiritualism and socialism An Elusive Victorian, like some other recent books,

seeks to challenge such simplistic and venient judgments Exploring the develop-ment of Wallace’s politics and spirituality

con-in the context of his developcon-ing science,Martin Fichman (a historian of science atYork University) convincingly argues that

“there is an identifiable coherence…inWallace’s thought and activities.”

Fichman takes a resolutely intellectualapproach, trawling Wallace’s correspon-dence and extensive published oeuvre Hehas also read through Wallace’s book collec-tion, noting the annotations and underliningsand taking them as clues to the Victorian’sdeveloping thoughts In so doing, Fichmancompellingly elucidates connections be-tween Wallace’s philosophy and a compre-hensive range of his predecessors’ and con-temporaries’ beliefs Some of these namesare expected (for example, Edward Bellamy,

the American socialist and author of Looking Backward), though Fichman’s discussions

are no less illuminating for that cluding the Swedish mystic EmanuelSwedenborg and the American scholarsWilliam James and Charles SandersPeirce—are a surprise James and Peircehelp demonstrate links between Wallace andrespectable, mainstream thought of the late19th and early 20th centuries

Others—in-Although Wallace was more willingthan the ever-cautious Darwin to explorethe significance of evolution to human be-ings, he initially thought that the undevel-oped potential of the intellect of “savages”

indicated that the brain’s evolution couldnot be accounted for by natural selection

Natural selection (or survival of the fittest,

as Wallace preferred to call it) provided

just-good-enough solutions to selectivepressures; it could never explain the devel-opment of potential that was not exercised

Hence, Wallace concluded,humans must be exempt fromthe evolutionary explanation.Later in life, Wallace modifiedhis views Sexual selection be-came for him a means to inte-grate the evolution of humansand socialism Only whenwomen were economically in-dependent (when their maritalchoices were not constrained by the finan-cial inequalities of capitalist society) couldthey make rational choices regarding a de-sirable father for their children Fichmanmakes a convincing argument in this re-spect, countering the claims of historianRobert M Young that the two possibili-ties—socialism and human evolution—were an either-or choice for Wallace.All this solid scholarship does not makefor a light read; Fichman offers relatively fewflashes of biographical or psychological col-

or It is not enlightening to be told that “Thecuriosity to explore new ideas has long beenrecognized by sociologists of science as anindispensable component of the psychologi-cal make-up of those individuals who choosescience as a career Wallace was no excep-tion.” Nor do the pictures provide any addi-tional stimulation—the standard portraitswill all be familiar to those who have doneany reading on 19th-century science

On occasion, this lack of color is mental to the book Spiritualism and so-cialism were more than ideas; they werelived responses to the experience of 19th-century society Though Fichman acknowl-edges this in word, readers would havebeen helped if he had presented it more indeed What, exactly, did spiritualist prac-tice involve? What was it like to be present

detri-at a séance? In whdetri-at other religious tices did spiritualists engage? Whatprotests and events book-ended Wallace’ssocialist publications? How much was hisrent? Did his family and friends thrive orsuffer under late-Victorian political econo-my? Fichman need not fear that such de-tails would detract from his scholarship;they would in fact enrich it

prac-Answers to such questions would vide historians and readers the selective en-vironment in which Wallace, as the book’ssubtitle suggests, evolved His optimisticbelief in the perfectibility of humankindthrough rational social organization demon-strates the Enlightenment inheritance of19th-century science, while his theism andspiritualism reflect an adaptation of that sci-ence to a society seeking faith and meaning

pro-in the Golden Age and beyond, pro-into theapocalypse of the Great War

The reviewer is at the Centre for History and Cultural

Studies of Science, University of Kent, Canterbury,

CT2 7NX, UK E-mail: c.l.sleigh@kent.ac.uk

Chemistry was first introduced as an

academic discipline in medical

facul-ties, academies, botanical gardens,

and museums in the late

17th century Over the next

100 years, it became an

established part of the

European intellectual world

Eighteenth-century chemists

were teachers and

profes-sors, authors of learned

books and experimental

es-says, members of academies

and scholarly societies, and

frequent visitors to coffee

shops and salons Yet, they

differed markedly from other

savants of the time They

were passionate experimenters who spent

many hours in their laboratories

Further-more, they were learned practitioners:

apothecaries, metallurgical officials,

con-sultants, inspectors of manufactures,

entre-preneurs, and members of state committees

and technological boards (1, 2).

The connection between chemistry and

pharmacy went back to medieval times The

distillation vessels used in 18th-century

apothecaries’ laboratories originated in the

late medieval alchemical tradition Chemical

operations such as distillations and

extrac-tions with solvents were also not invented in

apothecary guilds but learned from

15th-century alchemists Pharmacopoeias and

other apothecary books of the 18th century

included recipes for hundreds of chemical

medicines originally introduced by the

Paracelsian medical-chemical movement

Inversely, almost all 18th-century chemical

textbooks presented numerous recipes for

the fabrication of medicines and described

their properties and medical virtues

The chemist-apothecary was a widely

respected persona throughout 18th-century

Europe Andreas Sigismund Marggraf

(1709–1782), for example, had completed

an apothecary apprenticeship and, between

1735 and 1753, administered his father’s

apothecary shop in Berlin He later becamefamous for his chemical experiments andpublications, as well as for his directorship

of the Berlin Academy’sphysics class Marggraf hadalso learned assaying withJohann Friedrich Henckel(1678–1744) in Freiberg,done the first experimentsfor the extraction of sugarfrom beets, and was an am-bitious naturalist whoowned a large mineral col-lection Many 18th-centurychemist–physicians alsoproduced and sold their own

chemical remedies (3)

A number of

18th-centu-ry chemists also were involved in miningand metallurgy, particularly in Germany andSweden Travels to mining districts and vis-its to mines, salt-works, and metallurgicalfactories were an important part of the tech-nical education of 18th-century chemists Inthis way, they gathered information aboutthe processes of mining, smelting, and as-saying; the extraction of salts; and the prop-erties and uses of machines and materials

They brought back from their travels ples of minerals, as well as improved naturalhistorical knowledge about minerals, moun-tains, and strata of rocks

sam-But not only occasional travel createdbonds between academic chemistry and theworld of mines and smelting factories

From 1683, the Swedish Board of Minesmaintained a chemical laboratory wherechemist–mining officials analyzed miner-als and mapped the Swedish mineral re-

sources (4) In the mid-18th century, this

laboratory became a pioneering place forthe use of the blowpipe in mineral analysis

Many German chemists also held positions

as mining and metallurgical councilors inmining towns such as Freiberg, Brunswick,and Schemnitz, where they were chargedwith the control and improvement of thetechnology, economy, and organization oflabor in mines and smelting works andwith the analysis of minerals Most of themwere also teachers, authors of chemical andmetallurgical treatises, and members of

academies and scientific societies Thisdual carrier as chemist and salaried miningand metallurgical councilor continuedthroughout the 18th century

Eighteenth-century chemists were alsoactively involved in other arts and crafts.For example, in the 1740s, the Prussianking Frederick II commissioned threechemists to study the manufacture ofporcelain One of them, Johann HeinrichPott (1692–1777), established a porcelainworks in Freienwalde, funded by thePrussian king In France, Joseph PierreMacquer (1718–1784) made similar inves-

tigations, which culminated in the tion of the first French porcelain at Sèvres.Chemists were also involved in thesearch for surrogates for precious importedcommodities such as sugar, tobacco, cof-fee, brandy, and liqueurs The extraction ofsugar from beets, for example, which hadbeen initiated by Marggraf in the 1840s,was pursued by German chemists Mostsuccessful was Franz Carl Achard(1753–1821), who in the 1790s received asalary and an estate from the Prussian kingFriedrich Wilhelm II to establish a sugarmanufacture In the royal manufactures ofFrance, many chemists held leading posi-

produc-tions as inspectors (5).

B E Y O N D T H E I V O R Y T O W E R

Not a Pure Science: Chemistry in

the 18th and 19th Centuries

Ursula Klein

The author is with the Max Planck Institute for the

History of Science, Wilhemstrasse 44, 10117 Berlin,

Germany E-mail: klein@mpiwg-berlin.mpg.de

it was their experimental expertise that was sought after by governments and broader society Chemical theory began to play a key role in 19th-century experimentation, including the emergence of the synthetic chemical industry.

This year's essay series highlightsthe benefits that scientists, science,and technology have brought tosociety throughout history

Inventory of a 17th- to 19th-century apothecary’s laboratory.

It was less their theoretical knowledge

than their experimental and natural

histori-cal knowledge about a broad range of

mate-rials, their ability to identify matemate-rials,

ex-perimental skills, and their familiarity with

chemical analysis that equipped chemists

for their various technological occupations

as hybrid technologist-savants Until the

mid-18th century, the equipment in

aca-demic chemical laboratories did not differ

substantially from that of apothecaries’

shops, assaying and smelting workshops,

and distilleries We know from drawings of

chemical laboratories and instruments, and

from their verbal descriptions, that

18th-century chemists relied heavily on the

in-struments and materials provided by

ordi-nary craftsmen and merchants Their

smelt-ing and testsmelt-ing furnaces, bellows, crucibles,

calcination dishes, and balances were

large-ly the same as those used in the workshops

of assayers and smelters Evaporation

ves-sels, crystallizing dishes, phials, retorts,

alembics, pelicans, receivers, and

transmis-sion vessels were common instruments both

in the chemical and pharmaceutical

labora-tory (see the figures), while simple retorts

and receivers were shared with distillers for

fabricating mineral acids, alcoholic spirits,

and fragrant oils

As these technological occupations

il-lustrate, chemistry never was a pure

sci-ence But in the early 18th century,

chemists began to develop areas of inquiry

that were largely unfamiliar outside of

aca-demia They refined techniques of

chemi-cal analysis, restructured the relations

be-tween chemical analysis and theories of

chemical composition, explored cycles of

analysis and resynthesis of substances in

the laboratory, and analyzed experimental

results to establish laws of chemical

affini-ty between substances

These scientific activities culminated in

the so-called chemical revolution in the last

third of the 18th century, but they were

largely confined to explorations of

inorgan-ic substances (mineral acids, alkalis, metals,

and salts) that could be purified to a

com-paratively high degree with the available

laboratory techniques By contrast,

sub-stances extracted from plants and animals

were largely excluded from these

develop-ments Until the early 19th century, plant

and animal chemistry—or “organic

chem-istry” as it was then often called—included

inquiries into the natural history and

physi-ology of plants and animals and maintained

most of its traditional bonds with pharmacy

and other arts Between the late 1820s and

the 1840s, this changed fundamentally

when, spurred in particular by the work of

French and German chemists, a new form

of organic or “carbon chemistry” emerged

In the new carbon chemistry, the

investi-gation of the constitution (later “structure”)

of organic compounds and their substitutionreactions became a prominent objective

This objective was largely disconnectedfrom pharmacy and other commercial ap-plications Substitution reactions procuredchlorinated hydrocarbons and other organ-

ic molecules that were unknown both in ture and the extant commercial world Mostchemists involved in the emerging subdis-cipline of carbon chemistry—such asJustus Liebig (1803–1873) and FriedrichWöhler (1800–1882) in Germany and JeanDumas (1800–1884) and Auguste Laurent(1808–1853) in France—maintained theirconnections with pharmacy and industry

na-in areas other than carbon chemistry

However, the applicability of substitutionproducts and the usefulness of investiga-tions into the structure of organic com-pounds were not on their agenda

Yet, from the mid-1850s, it was exactlythis new type of chemical expertise that be-came most useful in the emerging synthetic-dye industry Why was this the case? Theconnections between chemical science andtechnology in the new synthetic-dye indus-try that began to develop after WilliamHenry Perkin’s synthesis of mauve in 1856are complex But one contribution of the sci-ence of carbon chemistry to the synthetic-dye industry was clearly crucial: chemicaltheory embodied in chemical formulae

Linear chemical formulae, like H2O for ter, had been introduced by the Swedishchemist Jacob Berzelius (1779–1848) in

wa-1813 They presented the composition ofchemical compounds according to a theory

of definite quantitative units or portions ofsubstances With atomism, this new quanti-tative theory shared the assumption of dis-continuous composition of substances Butthe algebraic form of Berzelian formulaeavoided narrow definitions in terms of

“atoms,” which many chemists rejected asmetaphysical entities Letters, numbers, andadditivity were sufficient to represent quan-titative units of elements and discontinuous

composition of compounds Differentarrangements of letters visually showed howunits of elements were combined with eachother The structural formulae of the 1860sdisplayed chemical and spatial arrange-ments in an even more pictorial form Beginning in the late 1820s, chemistsused chemical formulae as tools on paper

to model the constitution of organic pounds Using chemical formulae as papertools, chemists reduced the complexity inthe “jungle of organic chemistry” (F.Wöhler) Chemical formulae enabled them,for example, to order organic chemical re-actions by formula equations that distin-guished between a main reaction, side reac-tions, and successive reactions

com-In the 1860s, chemical formulae had come an emblem not only of academicchemistry but also of the synthetic-dye in-dustry Quantitative chemical theory wasimplemented in the new alliance betweencarbon chemistry and the synthetic-dye in-dustry in the form of paper tools that weresubordinated to chemists’ experimental and

be-technological goals (6) Compared with the

connections between academic chemistryand the arts and crafts in the 18th century,this role played by chemical theory and for-mulae was a novelty

References and Notes

1 K Hufbauer, The Formation of the German Chemical Community (1720–1795) (Univ of California Press, Berkeley, 1982).

2 C C Gillispie, Science and Polity in France at the End

of the Old Regime (Princeton Univ Press, Princeton,

NJ, 1980).

3 W Schneider, Geschichte der pharmazeutischen Chemie (Verlag Chemie, Weinheim, 1972).

4 T M Porter,Ann Sci 38, 543 (1981).

5 A Nieto-Galan, Colouring Textiles: A History of Natural Dyestuffs in Industrial Europe, vol 217 of Boston Studies in the Philosophy of Science (Kluwer, Boston, 2001).

6 U Klein, Experiments, Models, Paper Tools: Cultures

of Organic Chemistry in the Nineteenth Century T Lenoir, H U Gumbrecht, Eds (Stanford Univ Press, Stanford, CA, 2003).

7 D J Diderot, J LeRond d’Alembert, Encyclopédie ou Dictionnaire Raisonné des Sciences, des Arts, et des Métiers, Facsimile reprint of the 1st ed of 1751–1780 (Frommann, Stuttgart, 1966), 35 vols C

An 18th-century chemical laboratory.

Giving up smoking is easy I’ve done

it hundreds of times.” This quote by

Mark Twain sums up the addictive

nature of tobacco consumption Much has

been written about the sociopsychological

factors involved in cigarette smoking

However, the warding effects oftobacco consump-tion, and the cravingassociated with with-drawal, reflect a chemical dependence The

re-alkaloid nicotine found in tobacco leaves is

the addictive compound that drives

smok-ing behavior, and its effects suggest a

di-rect action on the central nervous system

(see the figure) In addition to its many

other physiological effects, nicotine also

increases mental alertness and improves

memory But how do nanomolar

concen-trations of nicotine exert such effects on

the brain? On page 1029 of this issue,

Tapper and colleagues (1) take an

impor-tant step forward in answering this

ques-tion They show that a mutation in the α4

subunit of nicotinic acetylcholine

recep-tors (nAChRs) expressed by neurons

low-ers the threshold for the induction of

nico-tine dependence These results not only

identify one of the nAChR subunits

in-volved in nicotine dependence, but also

re-veal the mechanism by which nicotine

pro-duces its addictive effects

The neuronal nAChRs are a family of

ligand-gated ion channels widely

ex-pressed in the central and peripheral

nerv-ous systems They are activated by the

en-dogenous neurotransmitter acetylcholine

The nAChRs are closely related to the

re-ceptors at the neuromuscular junction,

which mediate the transmission of nerve

impulses to our muscles The nAChRs are

cation-permeable channels formed from

the coassembly of five subunits, and their

activation causes excitation of the neuron

Eleven neuronal nAChR subunits have

been identified in humans, and these

com-bine to give different receptor subtypes

with distinct functional properties and

pharmacological characteristics (2) At the

molecular level, the nAChR is possibly thebest characterized of all the ligand-gatedreceptors Nevertheless, given the exis-tence of numerous receptor subtypes andthe lack of accurate in-

formation about whichsubunits make upfunctional receptors invivo, the mechanismwhereby binding ofnicotine to this recep-tor leads to depend-ence is still poorlyunderstood

The distribution ofnAChRs in the nervoussystem has been re-vealed by radiolabelednicotine or related li-gands These analysesshow that brains fromsmokers exhibit an in-creased number (that

is, “up-regulation”) ofhigh-affinity nicotine-binding sites, suppos-edly containing the α4and β2 nAChR sub-units Paradoxically, invitro experiments dem-onstrate that chronicexposure to nicotinecan “desensitize” thesereceptors, making themless sensitive to the nat-ural neurotransmitter acetylcholine It isthought that the up-regulation of bindingsites might reflect an increased number ofreceptors that may compensate for nicotine-induced desensitization

Previous attempts to assess the bution of nAChRs to the physiology andbehavior of nicotine addiction used knock-out mice lacking specific nAChR subunits

contri-These studies showed that mice lacking theβ2 subunit exhibit reduced sensitivity to the

addictive effects of nicotine (3) Blockade

of the high-affinity α4- and β2-containingreceptors by a competitive antagonist alsoattenuates the rewarding effect of nicotine

(4) The elegant experiments of Tapper et

al further support the crucial role of the α4β2 nAChRs in nicotine addiction (1).

These authors postulated that if the geneencoding the α4 subunit of the nAChR isengineered to confer a higher affinity tonicotine, animals expressing such a muta-tion should be more prone to its addictiveeffects Accordingly, the authors foundthat mice genetically modified to expressthis mutant α4 subunit display a highersensitivity to the multiple effects of nico-tine, with lower doses required to in-crease locomotor activity, sensitization,and reward

In addition to demonstrating an increase

in nicotinic sensitivity distinguishable both

in cellular and behavioral tests, the Tapper et

al work provides important information for

understanding addiction The cognitive inforcing effects of nicotine have beendemonstrated to be beneficial in neurode-generative disorders such as Alzheimer’s andParkinson’s diseases, suggesting that com-pounds targeted to nAChRs may be useful inrestoring a cholinergic deficit Genetic stud-ies have revealed the existence of polymor-phisms in genes encoding nAChR subunits.The relevance of these mutations is high-lighted by a rare form of nocturnal epilepsy

re-N E U R O S C I E re-N C E

What Genes Tell Us About Nicotine Addiction

R C Hogg and D Bertrand

The authors are in the Department of Neuroscience,

Centre Medical Universitaire, Geneva, Switzerland.

Unraveling the addictive power of nicotine Nicotine contained in

tobacco leaves diffuses in the smoker’s brain within a few seconds

nAChRs mediates both the negative (addiction) and positive (increasedmental alertness) effects of nicotine

that is genetically transmissible (5, 6).

Interestingly, these mutations have been

found in the α4 and β2 genes Functional

analyses of these mutated nAChRs show that

they all display increased sensitivity to the

natural neurotransmitter acetylcholine (7).

The findings of Tapper et al not only

vide direct evidence of how nicotine

pro-motes dependence, but also raise

fundamen-tal questions about the genetics of addiction

Is it possible that polymorphisms present inthe human population could determine oursusceptibility to addiction? If this is the case,

we can predict that polymorphisms in ronal nicotinic receptor genes could be asso-ciated with tobacco dependence A coordi-nated study of polymorphisms in nAChRgenes, smoking behavior, and functionalcharacterization of mutated receptors would

neu-be required to answer this question

References

1 A R Tapper et al., Science 306, 1029 (2004).

2 N Le Novere, P J Corringer, J P Changeux, J.

5 O K Steinlein,Prog Brain Res 145, 275 (2004).

6 I E Scheffer, S F Berkovic,Trends Pharmacol Sci 24,

428 (2003).

7 D Bertrandet al., Epilepsia 43 (suppl 5), 112 (2002).

Today’s information economy is driven

by the electronics and photonics

tech-nologies, which use electrons and

pho-tons, respectively, to carry, store, and process

information An emerging branch of

photon-ics, called plasmonphoton-ics, aims to use

nano-structured materials that support “surface

plasmons” (see the figure, top left) for these

purposes Plasmonics can potentially achieve

highly complex miniaturized devices by

con-trolling and manipulating light on the

nanometer scale (1–3) Several

plasmon-ic devplasmon-ices—including filters (1), wave

guides (1, 3), polarizers (4), and

nanoscopic light source (5)—have been

demonstrated However, for plasmonics

to reach its full potential, active

plasmon-ic devplasmon-ices (6) such as switches and

mod-ulators will be required

On page 1002 of this issue, Andrew

and Barnes (7) take an important step

toward the realization of an active

plas-monic device by combining thin

poly-mer films containing molecular

chro-mophores with thin silver films (see the

figure) The chromophores are the

mo-lecular equivalent of a Wi-Fi

transmit-ter and receiver, transferring energy and

hence information across the silver film

with the help of surface plasmons

When light interacts with a

free-elec-tron metal, such as a thin (10 to 200 nm)

silver film, its surface electrons oscillate

collectively and absorb, scatter, or

re-radi-ate the incident photons The resulting

sur-face electromagnetic field propagates in

the plane of the film (the x and y

direc-tions) with ranges of around 10 to 100 µm,

but decays exponentially in the z direction

(see the figure, top left) with a range of

200 to 300 nm The field intensity in the z

direction is amplified by a factor of 10 to

100 relative to the incident intensity These

propagating electromagnetic modes areproperly termed surface plasmon polari-tons, but are often referred to simply assurface plasmons

The molecular plasmonic device structed by Andrew and Barnes consists oftwo polymer layers, one containing donor(D) chromophore molecules and the othercontaining acceptor (A) fluorophore mole-cules (see the figure, bottom left) Theselayers are deposited on either side of a thin(30 to 120 nm) silver film to form a sand-wich structure that is supported on a quartzsubstrate (see the figure, right panel) Thedonors absorb incident light and transferthis excitation energy by dipole-dipole in-teractions to the acceptors The latter thenemit their characteristic fluorescence

con-The distance dependence of this dipole interaction, which is commonly re-ferred to as fluorescence resonance energytransfer or FRET, usually places an upperlimit of about 10 nm on the distance be-

dipole-tween D and A (8) This length scale is

comparable to the mensions of biologicalmacromolecules FRET

di-is therefore widelyused to measure thedistances betweensites in biomoleculesthat are labeled withdonors and acceptors.However, a number ofbiological and physicalproblems require dis-tance measurementswith a range longerthan 10 nm At presentthis cannot be accom-plished by FRET

Andrew and Barnesshow that by couplingsurface plasmons andFRET, energy can betransferred with highefficiencies of up to

~70% over distancesthat are 15 to 20 timeslonger than the 10 nmobserved in free solu-tion or in biomole-cules The authors sup-port the observation of the acceptor emis-sion spectrum with two additional lines ofevidence First, time-resolved measure-ments of the donor and acceptor emissionspectra provide an independent demon-stration of surface plasmon–mediatedFRET Second, angle-resolved lumines-cence studies on samples with a nanofab-ricated sinusoidal grating etched into theSiO2substrate conclusively show the par-ticipation of surface plasmon polaritonmodes

This groundbreaking research raises anumber of important questions Can otherplasmonic nanostructures, such as nano-particle or nanohole arrays, be fabricatedwith integrated molecular components toact as switches? What molecular proper-

D Ag A Quartz

30–200 nm

+++ – – – +++ – – – x z

Metal Dielectric medium

Toward active plasmonic devices (Top left) Surface plasmons are surface

dielectric medium (Bottom left) Donor and acceptor molecules used by Andrew and Barnes (Right) Schematic representation of the plasmonic de-

vice created by Andrew and Barnes Light arrives from the left (blue arrow), isabsorbed by the donor molecules in the blue layer, and is transferred acrossthe silver film with the help of surface plasmons The acceptor molecules inthe green layer absorb the light and emit their characteristic fluorescence

The author is in the Chemistry Department,

Northwestern University, Evanston, IL 60208, USA.

ties, such as thermal or light-driven color

change, can be used to make active

de-vices? Can other light-driven surface

processes be mediated in an analogous

manner? Is it possible to use molecular

plasmonic nanostructures to measure large

distances between sites in biological

struc-tures such as viruses and organelles? These

and other questions will engage both

theo-rists and experimentalists

Andrew and Barnes suggest the use of

surface plasmon–mediated energy transfer

to improve the light output of organic

light-emitting diodes, which have great promise

as low-cost, flexible, portable displays (9,

10) Surface plasmons may also facilitate

charge separation in synthetic harvesting nanostructures Other potentialapplications for molecular plasmonics in-

light-clude nanoscale optical spectroscopy (11), surface-enhanced spectroscopy (12), sur- face plasmon resonance sensing (13, 14), and nanolithography (15).

References and Notes

1 W L Barnes, A Dereux, T W Ebbesen,Nature 424,

824 (2003).

2 C L Haynes et al., J Phys Chem B 107, 7337 (2003).

3 S A Maier et al., Nature Mater 2, 229 (2003).

4 C L Haynes, R P.Van Duyne,Nano Lett 3, 939 (2003).

5 H J Lezec et al., Science 297, 820 (2002).

6 A V Krasavin, N I Zheludev,Appl Phys Lett 84, 1416

(2004).

7 P Andrew, W L Barnes,Science 306, 1002 (2004).

8 F Kulzer, M Orrit,Annu Rev Phys Chem 55, 585 (2004).

9 L H Smith, J A Wasey, W L Barnes, Appl Phys Lett.

84, 2986 (2004).

10 S Wedge et al., Appl Phys Lett 85, 182 (2004).

11 B Pettinger et al., Phys Rev Lett 92, 096101

(2004).

12 G C Schatz, R P Van Duyne, in Handbook of Vibrational Spectroscopy; J M Chalmers, P R Griffiths, Eds (Wiley, New York, 2002), vol 1, pp 759–774.

13 A J Haes, R P Van Duyne,Anal Bioanal Chem 379,

920 (2004).

14 J M Brockman, B P Nelson, R M Corn, Annu Rev.

Phys Chem 51, 41 (2000).

15 W Srituravanich et al., Nano Lett 4, 1085 (2004).

16 Supported by the Air Force Office of Scientific Research Multidisciplinary University Research Initiative program (F49620-02-1-0381) and the National Science Foundation (EEC-0118025, DMR-

0076097, CHE-0414554).

Many of today’s tiniest

sensors are built

with microsystems

technology, a set of design

and production tools that

us-es techniquus-es similar to

inte-grated circuit production to

make micrometer-scale

me-chanical structures Since

the 1970s, researchers have aimed to create

smarter sensors and transducers—in which

data are gathered and interpreted within the

sensor or transducer—using this

technolo-gy From accelerometers and gyroscopes to

image and pressure sensors, these smart

sensors have higher functionality and

per-formance, are more reliable and often

dras-tically smaller, and can be cheaper to

man-ufacture than conventional sensors

In many smart sensors, the sensor and

its computational chip may be next to each

other (see the figure, left panel) This

arrangement is not, however, ideal for all

applications For example, micromirror

ar-rays for projection display applications (1)

necessitate huge parallelism for controlling

the individual elements They therefore

re-quire “monolithic integration” (see the

fig-ure, right panel), in which each mirror

ele-ment is directly interfaced with its

process-ing electronics

Monolithic integration saves volume,

electric power, and possibly cost, but

achieving it is not straightforward, because

very different materials and processingtechniques must be combined on the samesubstrate There are currently three ap-proaches to realize monolithic integration

of sensors and electronics: processing themicrosystems first and the integrated cir-

cuits last, typically next to the sensors (2);

mixing the fabrication of both (3, 4); and

processing the integrated circuits first and

the microsystems last (5, 6), typically on

top of the circuitry

This third method could be the HolyGrail of smart microsystem processing, be-cause the ability to process and interconnectdevices in a modular manner above any un-derlying signal-processing circuitry is a sub-stantial simplification Unfortunately, at-tempts to realize this method by academiaand corporate R&D have not had the ex-pected widespread success Some industryleaders have questioned whether this ap-proach is worth pursuing, because the effortand time needed to develop a generic mi-crosystems module may not be adequatelyoffset by cost savings

The problem is indeed complex Oneneeds a good sensor material and a viableprocess technology for making the sensor,without at the same time affecting or de-

stroying the underlying circuitry (for ample, by excessive temperature or bychemical reactions) The major hurdleturns out to be temperature: The finishedcircuitry cannot withstand temperaturesabove 450°C—yet the most widely usedmicrosystems material, polysilicon (poly-Si), must be deposited and treated above

ex-800°C to ensure good chanical and electrical prop-erties The processes clearlyare incompatible

me-Many materials, such asorganic resists or metals,can be deposited and fabri-cated into microsystems atlow temperatures, but thesesuffer from degraded de-vice properties, high mechanical stress orstress gradient, and limited reliability.However, one material may combine allthe required properties Poly-SiGe is high-

ly suitable for reliable micromachining cause of its high melting point (well above900°C), high elastic constants (around 150GPa), and low losses (allowing it to vi-

be-brate or resonate if needed) (6) By

com-bining chemical vapor deposition andplasma-enhanced chemical vapor deposi-tion, poly-SiGe with mechanical qualitiessimilar to those of the abundant poly-Sican be manufactured at a deposition tem-perature of 450°C and at adequate deposi-

tion rates (100 nm/min) (7) Practical

films (for example, with a thickness of 10µm) have been generated with low tensile

stress and a very low strain gradient (7) In

situ doping has led to a poly-SiGe

materi-al with low resistivity (8) Moreover, the

electrical connection to the circuitryshows a reasonable contact resistivity Thegood mechanical and electrical propertiesmentioned above are needed for gyro-scopes and other kinematic sensors (such

Chris Van Hoof, Kris Baert, Ann Witvrouw

The authors are in the Microsystems Components

and Packaging Division, IMEC, Kapeldreef 75, 3001

Leuven, Belgium E-mail: chris.vanhoof@imec.be

Integrated circuits Microsensor Integrated circuits

without placing any restrictions on the

stan-dard integrated circuitry

Future smart sensors for personal health,

comfort, and safety monitoring (such as shock

sensors for fall detection and glucose sensors

for diabetics) will also require power autonomy

Apart from the necessary extreme

miniaturiza-tion for ubiquitous operaminiaturiza-tion, such sensors may

even have to be disposable This requirement

can create a whole new paradigm: In contrast

to the present applications, where reliability

and durability of microsystems sensors is of

utmost importance (for example, in airbag

de-ployment and electronic stability control

sys-tems in automobiles), disposable sensors may

not need to be as durable as current sensors

A number of materials are available for

dis-posable sensors Silicon will probably remain

the material of choice for signal processing in

future smart sensors because of its huge cessing power per unit area (and the smaller thearea, the lower the cost) With silicon circuitry

pro-in place, tpro-iny autonomous sensors will requirematerials compatible with this circuitry tomake the microsystems Besides poly-SiGe, anumber of reasonably durable metal alloys(such as Mg-alloyed aluminum) are pursuedbecause they can be produced at lower cost

But smart sensors may take a differentformat altogether—for example, as ultraflatelements integrated into textile materials or

even into paper (9) In this case,

semicon-ducting polymers may be coated on the tile or paper to form transistors and performmany of the sensing functions of today’s sil-icon-based microsystem devices This, inturn, may point to metals or polymers ascheap sensor materials Whatever microsys-

tex-tem application is envisioned, the next yearsmay see the market introduction of the newsensor materials discussed above

References and Notes

1 L J Hornbeck,Proc SPIE 3013, 27 (1997).

2 J Smith, S Montague, J Sniegowski, J Murray, P McWhorter, in Proc IEDM ’95 (IEEE, Washington, DC, 1995), pp 609–612.

3 W Kuehnel, S Sherman,Sens Actuat A 45, 7 (1994).

4 Analog Devices Inc (www.analog.com).

5 S Sedky, A Witvrouw, K Baert,Sens Actuat A 97-98,

Con-8 B C.-Y Lin, T.-J King, R T Howe,Proc MRS 782, 43

(2004).

9 F Ender et al., Appl Phys Lett 84, 2673 (2004).

10 The authors’ work on SiGe was partially funded by the IST project SiGeM (IST-2001-37681).

Bacteria are endowed with a

consider-able degree of internal organization

Thanks to fluorescence microscopy,

it is now clear that many bacterial

compo-nents—DNA as well as proteins—are

found in specific subcellular locations

Indeed, the discovery of prokaryotic

ho-mologs of both tubulin and actin, which are

key components of eukaryotic cellular

or-ganization, has overturned the textbook

credo that cytoskeletons are exclusive to

eukaryotes (1) On page 1021 of this issue,

Garner et al (2) provide the first

quantita-tive biochemical study of a bacterial

actin-like protein called ParM They show that

ParM is a dynamic polymerization engine

that drives the segregation of DNA

plas-mids during bacterial cell division

Accumulating evidence suggests that

bacterial proteins related to tubulin and

actin may be involved in bacterial cell

divi-sion and DNA segregation For example,

FtsZ, a bacterial homolog of tubulin, is the

principal component of the cytokinetic

Z-ring that constricts the middle of the

divid-ing bacterium (see the figure, A) (3).

MinCD is an inhibitor of Z-ring assembly

that oscillates rapidly from pole to pole,

ensuring that the Z-ring forms only at the

bacterial midcell position (4) The MinCD

proteins are organized into extended brane-associated coiled structures thatwind throughout the bacterium between

mem-the two poles (5) Rod-shaped bacteria

contain actin homologs such as MreB,which forms helical filaments just beneaththe bacterial cell surface that help to deter-mine cell shape and may contribute to

DNA segregation (see the figure, B) (6).

Both the FtsZ ring and the MreB helicesare highly dynamic structures that undergo

continuous cycles of remodeling (6, 7)

Bacterial plasmids are autonomous netic elements composed of DNA that en-code genes conferring resistance to, for ex-ample, antibiotics or heavy metals Suchplasmids must be segregated evenly be-tween the daughter cells during bacterial

ge-cell division The par (partitioning) locus

of the R1 drug-resistance plasmid encodesthree components: a centromere-like site in

the DNA (parC), a DNA-binding protein

(ParR), and a protein with adenosinetriphosphatase (ATPase) activity (ParM)

These components form a minimalist totic spindle, which positions pairs of plas-mids at opposite ends of the rod-shapedbacterium, ensuring equal distribution ofthe plasmids between the daughter cells

mi-An important clue to how the R1 par

operon works came from the discoverythat ParM forms dynamic filamentousstructures that extend along the longitudi-

nal axis of the bacterium Escherichia coli

(see the figure, C) (8) These ParM

fila-ments are visible in only about 40% of abacterial population at any given time, indi-cating that they are transient R1 plasmidsare located at the ParM filament tips Thisimplies that polymerization of ParM mole-cules into filaments could be the drivingforce that pushes the plasmids to opposite

poles of dividing rod-shaped bacteria (9).

Although ParM (like MreB) shows lessthan 15% sequence similarity to actin, itsthree-dimensional structure is very similar.Moreover, ParM assembles into two-stranded helical filaments that are almostidentical to the filamentous (F) actin of eu-

karyotic cells (10) Assembly of F-actin

from monomeric subunits is kinetically larized such that assembly at the end of onefilament (the barbed end) is favored overthe other end (the pointed end) At equilib-rium, the result is that ATP-actin subunitsassociate with the barbed end and ADP-actin subunits dissociate from the pointedend, leading to treadmilling of actinmonomers from the barbed end to the

po-pointed end (11)

By dual fluorescent labeling of ParM

filaments assembled in vitro, Garner et al.

show that, unlike either F-actin or tubules, ParM filaments polymerize bi-directionally with equal rates of addition atboth ends This is a surprising observationgiven the structural polarity of the ParM

micro-polymer (10) The authors examined

indi-vidual ParM filaments using total internalreflection fluorescence microscopy anddiscovered that ParM polymerization isvery rapid In fact, at physiological con-centrations of about 10 µM, ParM fila-ments extend from pole to pole of a bac-terium in just 10 seconds Even more sur-prising, after symmetrical bidirectionalgrowth, the ParM polymers suddenlyswitch to rapid unidirectional disassembly

M I C R O B I O L O G Y

Dynamic Instability of

a Bacterial Engine

Jakob Møller-Jensen and Kenn Gerdes

J Møller-Jensen is in the MRC Laboratory of

Molecular Biology, Cambridge CB2 2QH, UK.

K Gerdes is in the Department of Biochemistry and

Molecular Biology, University of Southern Denmark,

DK-5230 Odense, Denmark E-mail: kgerdes@

bmb.sdu.dk

This behavior differs from F-actin

tread-milling and instead resembles the

cata-strophic event of microtubule disassembly

when microtubule plus-ends switch from

growth to shrinkage (defined as dynamic

polymer instability)

The switch from ParM polymer

elonga-tion to shortening is regulated by

hydroly-sis of ATP to ADP, because ParM

fila-ments treated with ATP analogs that

can-not be hydrolyzed do can-not show this switch

and remain stable (2, 8) Garner and

col-leagues confirm the previous finding that

the ATPase activity of ParM is stimulated

by polymerization They then show that

ADP-ParM dissociates very rapidly from

the polymer The measured rate of ParM

filament disassembly far exceeds that of

ATP hydrolysis, suggesting that the

fila-ments quickly depolymerize when

nu-cleotide hydrolysis catches up with

polym-erization at one of the filament ends

Consistently, addition of a small amount of

a mutant form of ParM (that cannot

hy-drolyze ATP) to the polymerization

reac-tion led to stabilizareac-tion of the filaments

(2) Because fluorescent labeling does not

discriminate between the two ends of the

ParM filament, it remains to be

deter-mined whether disassembly can be

initiat-ed at either end in a random fashion or

whether it is initiated at one end only

When studied in isolation, F-actin

treadmilling is far too slow to account for

the dynamics of the actin-based motility

observed in living cells Instead, the

turnover of F-actin is regulated by a

multi-tude of actin-binding proteins that act at

different stages during subunit recycling

and hence trim the dynamic potential of

F-actin to suit a variety of cellular functions

(11) By contrast, ParM is highly dynamic

even in the absence of additional proteins

and so probably has no need for auxiliary

factors (2) Given its plasmid origin, ParM

is probably designed to operate only during

the partitioning of plasmids prior to

bacte-rial cell division Hence, ParM does not

need to be tightly integrated into other

bac-terial protein interaction networks

How does the dynamic behavior of

ParM translate into rapid plasmid

segrega-tion? A previous model proposed that

plas-mid copies located in the plas-middle of the

bacterium are paired via their centromeric

parC sites and are bound together by ParR

(12) The paired plasmids then serve as a

nucleation point for ParM polymerization,

which in turn pushes the plasmids apart to

the opposite poles of the bacterium (8, 9).

Consistent with the ParM localization

pat-tern in vivo, this model postulated that

ParM filaments in opposite orientations

constitute the DNA segregation apparatus

(9) Garner et al (2) suggest an alternative

model based on their observation thatParM has a high propensity for self-assem-bly and therefore is likely to polymerizewithout the need for nucleation These in-vestigators propose that transient ParMprotofilaments formed throughout the bac-terium “search” the cytoplasmic lumen forDNA just as microtubules do in eukaryot-

ic cells In this model, the ParR-parC

com-plex, rather than acting as a nucleation tor for polymerization, instead serves to

fac-stabilize ParM filament ends while ing the insertion of new subunits at thepolymer tips If true, this is a surprisinglysimple and elegant solution to the puzzle

allow-of how a mitotic apparatus accomplishesDNA segregation F-actin barbed ends aresimilarly stabilized by proteins called

formins (13) Furthermore, Garner and

colleagues propose that bidirectionallyelongating ParM filaments are capped atboth ends by the segregating plasmids.Given that ParM polymers are structurallypolarized as a result of the head-to-tail

arrangement of their subunits (10), this

mechanism would require that the

ParR-parC complex is capable of interacting

with ParM filaments in opposite tions Further analysis is needed to seewhether this is the case

orienta-A second type of three-component titioning system is known to geneticallystabilize low–copy number plasmids such

par-as P1 and F factor In this cpar-ase, a like partitioning ATPase (ParA) that con-tains a Walker-box motif oscillates alonghelical tracks over the bacterial nucleoid

MinD-(see the figure, D) (14, 15) How the

dy-namic behavior of these proteins relates tothe symmetrical bidirectional partitioning

of plasmids during bacterial cell divisionpresents a fascinating conundrum Severallines of evidence implicate the actin ho-molog MreB in the active segregation of

bacterial chromosomal DNA (16–18) It

will be interesting to learn whether MreBperforms a function similar to that of ParMand whether it constitutes an equally potentpolymerization engine

3 E F Bi, J Lutkenhaus,Nature 354, 161 (1991).

4 D M Raskin, P A de Boer, Proc Natl Acad Sci U.S.A.

7 J Stricker, P Maddox, E D Salmon, H P Erickson,Proc.

Natl Acad Sci U.S.A 99, 3171 (2002).

8 J Møller-Jensen, R B Jensen, J Lowe, K Gerdes, EMBO

J 21, 3119 (2002).

9 J Møller-Jensenet al., Mol Cell 12, 1477 (2003).

10 F van den Ent, J Møller-Jensen, L A Amos, K Gerdes,

J Lowe,EMBO J 21, 6935 (2002).

11 T D Pollard, G G Borisy,Cell 112, 453 (2003).

12 R B Jensen, R Lurz, K Gerdes, Proc Natl Acad Sci.

cy-MinCD (green coil) oscillates from pole to pole

of the bacterium and thereby prevents tion of the Z-ring (yellow) at the poles Instead,the Z-ring, which is composed of the tubulin ho-

forma-molog FtsZ, forms at the midcell position (B)

MreB, a bacterial homolog of actin, forms helicalfilaments (red) just beneath the cell surface thathelp to determine cell shape and may con-tribute to the segregation of chromosomal

DNA (C) ParM polymerizes in an

ATP-depend-ent manner into double-stranded helical ments (blue) similar to actin filaments TheParM filaments have plasmids (pink) at theirends The polymerization of ParM filaments pro-vides the driving force to segregate plasmids toopposite poles of the bacterium during cell divi-

fila-sion (D) The ParA filament (yellow) oscillates

over the nucleoid (blue) and thereby positionsthe plasmids (pink) at the midcell position

Autophagy in Health and Disease:

A Double-Edged Sword

Takahiro Shintani and Daniel J Klionsky*

Autophagy, the process by which cells recycle cytoplasm and dispose of excess or

defective organelles, has entered the research spotlight largely owing to the discovery of

the protein components that drive this process Identifying the autophagy genes in yeast

and finding orthologs in other organisms reveals the conservation of the mechanism of

autophagy in eukaryotes and allows the use of molecular genetics and biology in different

model systems to study this process By mostly morphological studies, autophagy has

been linked to disease processes Whether autophagy protects from or causes disease is

unclear Here, we summarize current knowledge about the role of autophagy in disease

and health

C ellular homeostasis requires a

con-stant balance between biosynthetic

and catabolic processes Eukaryotic

cells primarily use two distinct mechanisms

for large-scale degradation, the proteasome

and autophagy; but only autophagy has

the capacity to degrade entire organelles

The three types of autophagy are

macro-autophagy, micromacro-autophagy, and

chaperone-mediated autophagy (1) Here, we will focus

on macroautophagy, hereafter called

autoph-agy, which plays an important physiological

role in human health In autophagy, a

double-or multi-membrane–bound structure, called

the autophagosome or autophagic vacuole, is

formed de novo to sequester cytoplasm Then,

the vacuole membrane fuses with the some to deliver the contents into the organellelumen, where they are degraded and the re-sulting macromolecules recycled (Fig 1)

lyso-Autophagy occurs at basal levels in mosttissues and contributes to the routine turnover

of cytoplasmic components However, agy can be induced by a change of environ-mental conditions such as nutrient depletion

autoph-In addition to turnover of cellular components,autophagy is involved in development, differ-entiation, and tissue remodeling in various or-ganisms (2) Autophagy is also implicated incertain human diseases Paradoxically, autoph-agy can serve to protect cells but may also con-tribute to cell damage (Table 1) Here, we willsummarize the current connections betweenautophagy and human disease and aging

Programmed Cell DeathAutophagy is involved in programmed celldeath (PCD) Type I PCD, apoptosis, is char-

acterized by condensation of cytoplasm andchromatin, DNA fragmentation, and cell frag-mentation into apoptotic bodies, followed byremoval and degradation of the dying cells

by phagocytosis Type II PCD (autophagic)

is characterized by the accumulation of tophagic vesicles (autophagosomes and auto-phagolysosomes) and is often observed whenmassive cell elimination is demanded or whenphagocytes do not have easy access to thedying cells One feature that distinguishesapoptosis from autophagic cell death is thesource of the lysosomal enzymes used formost of the dying cells’ degradation Apopto-tic cells use phagocytic cell lysosomes forthis process, whereas cells with autophagicmorphology use the dying cells’ endogenouslysosomal machinery It has been unclearwhether autophagy directly executes celldeath or is the secondary effect of apopto-sis A recent study, however, suggests thatautophagy might cause cell death (3) Caspaseinhibitor–induced autophagic cell death is se-verely affected by RNA interference (RNAi)with ATG7 and beclin 1 expression, two geneswhose products are essential for autophagy(3)

au-Two key molecules that control PCD aremembers of the death-associated protein ki-nase (DAPk) family Both DAPk and DAPk-related protein kinase-1 (DRP-1) promotedeath in a way that depends on their ki-nase activities DAPk predominantly acti-vates apoptosis through a caspase-dependent

University of Michigan, Life Sciences Institute, and

Departments of Molecular, Cellular, and

Develop-mental Biology and of Biological Chemistry, Ann

Arbor, MI 48109–2216, USA.

*To whom correspondence should be addressed.

E-mail: klionsky@umich.edu

Table 1 Possible roles of autophagy in health and disease.

Disease state Beneficial effects of autophagy Negative effects of autophagy

Cancer Acts as a tumor suppressor; may be involved in type II

PCD in cancer cells, could limit cell size or may remove damaged organelles that could generate free radicals and increase mutations.

May allow survival of cancer cells within the nutrient-poor environment of a tumor, could prevent cell death, and may protect against some cancer treatments.

Liver disease Allows removal of nonfunctional endoplasmic

retic-ulum resulting from accumulation of aggregated

Neurodegeneration Allows the removal of protein aggregates before they

990