Science - October 2, 2020 USA

FDA documents reveal slipshod Freshwater conservation requires Thirst research wins
oversight of clinical trials p. 24 specifc planning pp. 38 & 117 neurobiology prize p. 45

$15
2 OCTOBER 2020

SPECIAL ISSUE

NEURODEGENERATION

Understanding the brain’s decline



Sethuraman Ruha Benjamin
Panchanathan
Princeton University
National Science
Foundation

Anthony S. Fauci

National Institute of
Allergy and Infectious

Diseases

Mary L. Gray Claire Fraser

Microsoft Research AAAS President
University of Maryland

School of Medicine

The 2021 Annual Meeting will convene entirely online, February 8-11
with related pre-released materials available starting in late January.
Meeting registration will be available in November.

For more information, please visit:

aaas.org/meetings | #AAASmtg

ADVERTISEMENT

Eppendorf: Working for
a better world for 75 years

Eppendorf is supporting essential can be regarded as the archetypes of that to go hand in hand with significant
businesses and institutions in the modern medical technologies now reductions in the workload of labora-
fight against the coronavirus. In the used routinely all over the world. Eva tory staf.” This is a Herculean task, as
year of the company’s 75th birthday, van Pelt points to the important prod- Fruhstorfer points out, and one that
we are thus fulfilling the mission ucts developed by Eppendorf since its requires digitalization, the best minds
of our founders – to help improve at Eppendorf and the special spirit that
people’s living conditions – in a “The mission of our exists within the company.
very special way. company, as defined by
its founders, is as current Strong teamwork for quality
“The mission of our company, as de- and relevant today as and innovation
fined by its founders, is as current and it was then.” “The pronounced sense of community
relevant today as it was then,” says within the company is one of the keys
Eva van Pelt, Co-CEO of Eppendorf Eva van Pelt, to Eppendorf’s long-standing success,”
AG. This commitment to improve Co-CEO Dr. Fruhstorfer explains. The well-be-
people’s living conditions drove a ing of every employee was always im-
small group of technical experts led by foundation and leaves no doubt portant to the founders of Eppendorf,
the company’s founders, Dr. Heinrich that the company’s innovative spirit and this attitude remains an integral
Netheler and Dr. Hans Hinz, toward and determination to advance techni- part of the company’s distinctive
the healthcare sector in the first weeks cally will continue to be reflected in culture to this day. Mutual respect is
after World War II. In August 1945, its products in the future. Nowadays, just as important as the open-minded
they began repairing urgently needed however, Eppendorf no longer operates attitude toward others that can be
but defective medical equipment in the field of medical technology, but encountered at Eppendorf throughout
and instruments belonging to Univer- as a globally successful manufacturer the world. In the eight decades of
sity Medical Center Eppendorf in of laboratory equipment with around Eppendorf’s presence on the market,
Hamburg. The group was so success- 4,000 employees all over the world. these two qualities have been the
ful in that task that it soon received basis for dialogue, the transfer of
its first development orders for new “In our work, we cooperate closely knowledge, exchanges of experience
types of equipment from the medical and trustfully with our partners and and global networking. All of these
center’s departments. customers to address the issues of factors have contributed to the high
the future,” adds Co-CEO Dr. Peter reliability, good service and special
Eppendorf – an innovative pioneer Fruhstorfer. He notes that Eppendorf quality of the company’s products
in medical technology is driven by the idea of enabling labo- and thus to the eWcient support
The innovations developed by ratories to work eWciently, sustainably provided by equipment, consumables
Eppendorf – such as the Stimulator, and with high quality. “Specifically, and service that has long made
a device to stimulate muscles and our aim is to speed up laboratory Eppendorf indispensable for many
nerves, or the Ophtalmochirurg, a kind processes while at the same time laboratories.
of forerunner to eye lasers for the reducing potential errors. And we want
treatment of retinal detachment –

1945 1949 1961 1963 1963

Dr. Heinrich Netheler and Eppendorf develops a photometer Eppendorf launches the first piston- A microliter system consisting The “Eppi®“ reaction vessel
Dr. Hans Hinz develop medical diag- for clinical applications. Over time, stroke microliter pipette, heralding of reaction vessels, mixers, centrifuges is launched on the market. It quickly
nostic products; in 1946 their company the age of precise and fast pipetting.
was renamed “Elektromedizinische it became a world standard for and pipettes is launched. became indispensable in medical
chemical and biochemical analyses. It facilitates and improves laboratory and scientific
Werkstätten GmbH.“
work enormously. laboratories worldwide.

ADVERTISEMENT

Pioneer in accuracy, products are always up to date. Ideally,
reliability and innovation they are even a step ahead of the
times, like the piston-stroke pipette
> In Liquid Handling, the product was when it was launched. Other ex-
range extends from manual amples are the Eppendorf centrifuges
pipettes to electronic pipettes with their remarkable durability and
and comprehensive pipetting Eppendorf‘s high-tech plastic consum-
systems. ables for use in laboratories.

> Whether tubes, plates, pipette State-of-the-art technology and
tips, stem cell cultureware or the sophisticated ergonomics of the
cuvettes: The unique product laboratory equipment ensure easy and
features of Eppendorf Consum- reliable handling, allowing users to
ables accelerate and simplify concentrate fully on their research.
routine tasks and ensure greater
reliability in every laboratory. “The pronounced sense
of community within
> Eppendorf is one of the world`s the company is one of the
market leaders and full- keys to Eppendorf’s
spectrum supplier in the field long-standing success.”
of centrifugation: the product
range extends from benchtop Dr. Peter Fruhstorfer,
centrifuges, premium floor- Co-CEO
standing centrifuges and ultra-
centrifuges to clinical and
automated centrifuges.

Above High-tech developments In this way, Eppendorf actively
View into the clean together with users contributes to research activities
room production: In its Business Areas, Eppendorf around the world every day – and at
high-quality labo- focuses all its ideas and developments the moment, to the fight against the
ratory consumables on the needs of the user. Customer Sars-CoV-2 virus – in keeping with the
made of high-tech requirements determine what inno- mission that the founders impressed
plastics vations are developed, and these are on the company 75 years ago: tttttttttt
constantly discussed and tested with to improve people’s living conditions.
Bottom left users through agile working meth-
Production halls and ods – a process that, by the way, was More information at:
high-bay warehouse used in Eppendorf’s early days, laying eppendorf.com
at the Eppendorf the foundation for the great market
location Oldenburg success of its equipment.
in Holstein
The result is state-of-the-art equip-
Bottom right ment technology with sophisticated
Eppendorf centri- ergonomics, ensuring that Eppendorf’s
fuges of the latest
generation
at the Leipzig
location

1978 1996 2003 2018 Today

The first Multipette® with Eppendorf develops the air The Eppendorf line of epMotion The 4 liter versatile Multipurpose Discover more product
Combitips® is launched and oil micro-injectors CellTram – automated liquid handling systems Centrifuge 5910 R with universal innovations online:
on the market and for pressure control during manual rotor and adapter concept and the
becomes a bestseller. is designed to help smallest footprint in the market eppendor f.com / 75-years
micro-injection and automate routine pipetting tasks
dosing of liquids. expands the centrifuge range.
to free up time.

ADVANCING BIOPHARMACEUTICALS

With over 80 years’ experience in healthcare, Fujifilm is
committed to enhancing patient care and addressing
unmet medical needs. As a world-leading contract
biopharmaceutical manufacturer (CDMO), we’re helping
accelerate the development of new drugs with a complete
range of services. Our state-of-the-art facilities deliver
precise control of process conditions and unmatched
productivity. Utilizing our cell biology and bioprocessing
expertise, we provide high-quality cell culture media to
our biopharma partners for the production of many of
the world’s leading drugs. At Fujifilm, we’ll NEVER STOP
pushing scientific boundaries to help develop novel
therapies and life-changing treatments that create a
healthier world for all.

Follow Fujifilm Life Sciences on

Image is an actual manufacturing facility of Fujifilm in Denmark.
FUJIFILM and Fujifilm Value from Innovation are trademarks of FUJIFILM Corporation.
©2020 FUJIFILM Corporation. All rights reserved.

CONTENTS 19 Stem cell studies probe origins
of the placenta
2 OCTOBER 2020 • VOLUME 370 • ISSUE 6512 Lab models of organ will track how
it emerges—and what can go wrong
SPECIAL SECTION
By K. Servick
NEURODEGENERATION
20 The ‘bat man’ tackles COVID-19
ILLUSTRATION: SIMON PRADES INTRODUCTION ON THE COVER After a career investigating why so many
48 A cruel end to too many lives Neurodegeneration in later life robs viruses come from bats, Linfa Wang
us of our abilities and our memories, eyes a new challenge By K. Kupferschmidt
REVIEWS slowly and inexorably, like leaves
50 Glymphatic failure as a final common falling from a tree. A greater under- 22 A call for diagnostic tests to
pathway to dementia standing of the pathophysiological report viral load
M. Nedergaard and S. A. Goldman underpinnings within the brain Measure could help officials know
who is most contagious By R. F. Service
56 Beyond aggregation: Pathological may provide
phase transitions in neurodegenerative crucial hints that FEATURES
disease C. Mathieu et al. will help us delay
and perhaps 24 Official inaction
61 Translating genetic risk of even reverse A Science investigation shows that FDA
Alzheimer’s disease into mechanistic symptoms. oversight of clinical trials is lax, slow moving,
insight and drug targets A. Sierksma et al. See page 48. and secretive—and that enforcement is
Illustration: declining By C. Piller
66 Microglia modulate Simon Prades
neurodegeneration in Alzheimer’s and 27 Disgraced researchers can
Parkinson’s diseases T. Bartels et al. SEE ALSO PERSPECTIVE p. 32 still reap drug industry payouts
By C. Piller
QQ群: 1074370165 17 China’s bold climate pledge earns
praise—but is it feasible? PODCAST
NEWS Plan to achieve carbon neutrality by 2060
would make China a global leader, but INSIGHTS
IN BRIEF abandoning coal will be hard
LETTERS
14 News at a glance By D. Normile
30 NextGen Voices:
IN DEPTH 18 Record U.S. and Australian fires Funding fix: Spend time
raise fears for many species
16 Europe builds ‘digital twin’ of Earth Scientists say fires likely wiped out PERSPECTIVES
to hone climate forecasts some rare Australian organisms,
Ingesting more data than ever before, and worry U.S. blazes now threaten more 32 Shifts and drifts in prion science
exascale model will simulate the impact of Important questions remain unanswered
climate change on humans By P. Voosen By J. Pickrell and E. Pennisi since prions were discovered four
decades ago By A. Aguzzi and E. De Cecco
SCIENCE sciencemag.org EDITORIAL p. 13
NEURODEGENERATION SECTION p. 48

34 Mutational selection in
normal urothelium
Mutations in normal tissue point to causes
of DNA damage and set the stage for cancer

By S. G. Rozen

RESEARCH ARTICLES pp. 75 & 82

36 Coaxing stem cells to repair
the spinal cord
Spinal cells in mice can be induced to
generate protective oligodendrocytes
after injury By C. G. Becker and T. Becker

RESEARCH ARTICLE p. 73

37 A rival to superalloys at high
temperatures
Slip-pathway activation provides plasticity
in a multiprincipal element alloy with
high-temperature strength By J. Cairney

RESEARCH ARTICLE p. 95

2 OCTOBER 2020 • VOL 370 ISSUE 6512 7



CONTENTS

38 & 117 89 Coronavirus
Selective and cross-reactive SARS-CoV-2
PHOTO: CLAUS MEYER/MINDEN PICTURES 38 A boost for freshwater conservation RESEARCH T cell epitopes in unexposed humans
Integrating freshwater and terrestrial J. Mateus et al.
conservation planning has high returns IN BRIEF
95 Metallurgy
By R. Abell and I. J. Harrison 70 From Science and other journals Multiplicity of dislocation pathways
in a refractory multiprincipal
REPORT p. 117 RESEARCH ARTICLES element alloy F. Wang et al.

POLICY FORUM 73 Neurodevelopment PERSPECTIVE p. 37
A latent lineage potential in resident neural
40 How to fix the GDPR’s frustration stem cells enables spinal cord repair 101 Organic chemistry
of global biomedical research E. Llorens-Bobadilla et al. A universal system for digitization
Sharing of data for research beyond the EU and automatic execution
must improve By J. Bovenberg et al. RESEARCH ARTICLE SUMMARY; FOR FULL TEXT: of the chemical synthesis literature
DX.DOI.ORG/10.1126/SCIENCE.ABB8795 S. H. M. Mehr et al.
BOOKS ET AL. PERSPECTIVE p. 36
REPORTS
43 Flawed research and 74 Solar cells
its enduring repercussions Vapor-assisted deposition of highly efficient, 108 Solar cells
A journalist recounts how he exposed stable black-phase FAPbI3 perovskite solar Impact of strain relaxation on performance
problems with a study linking vaccines and cells H. Lu et al. of a-formamidinium lead iodide
autism By P. A. Offit perovskite solar cells G. Kim et al.
RESEARCH ARTICLE SUMMARY; FOR FULL TEXT:
44 Training tomorrow’s scientists DX.DOI.ORG/10.1126/SCIENCE.ABB8985 113 Developmental biology
Prioritizing STEM diversity, equity, and An adhesion code ensures robust pattern
inclusion requires rethinking graduate Mutation formation during tissue morphogenesis
education By A. Huderson 75 Extensive heterogeneity in somatic T. Y.-C. Tsai et al.

PRIZE ESSAY mutation and selection in the human 117 Conservation planning
bladder A. R. J. Lawson et al. Integrated terrestrial-freshwater planning
45 The origins of thirst 82 Macroscopic somatic clonal expansion doubles conservation of tropical aquatic
Sensory signals arise throughout the body in morphologically normal human species C. G. Leal et al.
and converge in the brain to regulate drinking urothelium R. Li et al.
PERSPECTIVE p. 38
By C. A. Zimmerman PERSPECTIVE p. 34
121 Sex determination
The mouse Sry locus harbors a cryptic
exon that is essential for male sex
determination S. Miyawaki et al.

Electrocalorics
125 Giant temperature span in electrocaloric

regenerator A. Torelló et al.

129 A high-performance solid-state
electrocaloric cooling system
Y. Wang et al.

D E PA R T M E N T S

13 Editorial
Fire in our future

By William Wallace Covington and Stephen Pyne

NEWS STORY p. 18

138 Working Life
A leap of faith By Rachel Mason

Science Staff ................................................ 10
New Products............................................. 134
Science Careers ......................................... 135

SCIENCE (ISSN 0036-8075) is published weekly on Friday, except last week in December, by the American Association for the Advancement of Science, 1200 New York Avenue, NW, Washington, DC 20005. Periodicals mail
postage (publication No. 484460) paid at Washington, DC, and additional mailing offices. Copyright © 2020 by the American Association for the Advancement of Science. The title SCIENCE is a registered trademark of the AAAS. Domestic
individual membership, including subscription (12 months): $165 ($74 allocated to subscription). Domestic institutional subscription (51 issues): $2148; Foreign postage extra: Air assist delivery: $98. First class, airmail, student, and
emeritus rates on request. Canadian rates with GST available upon request, GST #125488122. Publications Mail Agreement Number 1069624. Printed in the U.S.A.
Change of address: Allow 4 weeks, giving old and new addresses and 8-digit account number. Postmaster: Send change of address to AAAS, P.O. Box 96178, Washington, DC 20090–6178. Single-copy sales: $15 each plus shipping and
handling available from backissues.sciencemag.org; bulk rate on request. Authorization to reproduce material for internal or personal use under circumstances not falling within the fair use provisions of the Copyright Act can be obtained
through the Copyright Clearance Center (CCC), www.copyright.com. The identification code for Science is 0036-8075. Science is indexed in the Reader’s Guide to Periodical Literature and in several specialized indexes.

SCIENCE sciencemag.org 2 OCTOBER 2020 • VOL 370 ISSUE 6512 9

BOARD OF REVIEWING EDITORS (Statistics board members indicated with S)

Editor-in-Chief Holden Thorp, [email protected] Adriano Aguzzi, U. Hospital Zürich Robert Kingston, Harvard Med. School
Takuzo Aida, U. of Tokyo Nancy Knowlton, Smithsonian Institution
Executive Editor Monica M. Bradford Leslie Aiello, Wenner-Gren Foundation Etienne Koechlin, École Normale Supérieure
Editors, Research Valda Vinson, Jake S. Yeston Editor, Insights Lisa D. Chong Deji Akinwande, UT Austin Alex L. Kolodkin, Johns Hopkins U.
Judith Allen, U. of Manchester Julija Krupic, U. of Cambridge
DEPUTY EDITORS Julia Fahrenkamp-Uppenbrink (UK), Stella M. Hurtley (UK), Phillip D. Szuromi, Sacha Vignieri SR. EDITORIAL FELLOW Marcella Alsan, Harvard U. Thomas Langer, Max Planck Inst. Cologne
Andrew M. Sugden (UK) SR. EDITORS Gemma Alderton (UK), Caroline Ash (UK), Brent Grocholski, Pamela J. Hines, Di Jiang, Sebastian Amigorena, Institut Curie Mitchell A. Lazar, U. of Penn.
Marc S. Lavine (Canada), Ian S. Osborne (UK), Beverly A. Purnell, L. Bryan Ray, H. Jesse Smith, Keith T. Smith (UK), Jelena Stajic, James Analytis, UC Berkeley Wendell Lim, UC San Francisco
Peter Stern (UK), Valerie B. Thompson, Brad Wible, Laura M. Zahn ASSOCIATE EDITORS Michael A. Funk, Priscilla N. Kelly, Tage S. Rai, Trevor Archer, NIEHS, NIH Jianguo Liu, Michigan State U.
Seth Thomas Scanlon (UK), Yury V. Suleymanov LETTERS EDITOR Jennifer Sills LEAD CONTENT PRODUCTION EDITORS Harry Jach, Paola Arlotta, Harvard U. Luis Liz-Marzán, CIC biomaGUNE
Lauren Kmec CONTENT PRODUCTION EDITORS Amelia Beyna, Jeffrey E. Cook, Chris Filiatreau, Julia Katris, Nida Masiulis, Suzanne M. Johan Auwerx, EPFL Omar Lizardo, UCLA
White SR. EDITORIAL COORDINATORS Carolyn Kyle, Beverly Shields EDITORIAL COORDINATORS Aneera Dobbins, Joi S. Granger, Jeffrey David Awschalom, U. of Chicago Jonathan Losos, Washington U. in St. Louis
Hearn, Lisa Johnson, Maryrose Madrid, Ope Martins, Shannon McMahon, Jerry Richardson, Hilary Stewart (UK), Alana Warnke, Clare Baker, U. of Cambridge Ke Lu, Chinese Acad. of Sciences
Alice Whaley (UK), Anita Wynn PUBLICATIONS ASSISTANTS Jeremy Dow, Alexander Kief, Ronmel Navas, Brian White EXECUTIVE Nenad Ban, ETH ZÜrich Christian Lüscher, U. of Geneva
ASSISTANT Jessica Slater ASI DIRECTOR, OPERATIONS Janet Clements (UK) ASI SR. OFFICE ADMINISTRATOR Jessica Waldock (UK) Franz Bauer, Pontificia U. Católica de Chile Jean Lynch-Stieglitz, Georgia Inst. of Tech.
Ray H. Baughman, U. of Texas at Dallas Fabienne Mackay, QIMR Berghofer
News Editor Tim Appenzeller Carlo Beenakker, Leiden U. Anne Magurran, U. of St.Andrews
Yasmine Belkaid, NIAID, NIH Asifa Majid, U. of York
NEWS MANAGING EDITOR John Travis INTERNATIONAL EDITOR Martin Enserink DEPUTY NEWS EDITORS Elizabeth Culotta, Lila Guterman, Philip Benfey, Duke U. Oscar Marín, King’s College London
David Grimm, Eric Hand (Europe), David Malakoff SR. CORRESPONDENTS Daniel Clery (UK), Jon Cohen, Jeffrey Mervis, Elizabeth Gabriele Bergers, VIB Charles Marshall, UC Berkeley
Pennisi ASSOCIATE EDITORS Jeffrey Brainard, Catherine Matacic NEWS REPORTERS Adrian Cho, Jennifer Couzin-Frankel, Jocelyn Kiros T. Berhane, Columbia U. Christopher Marx, U. of Idaho
Kaiser, Kelly Servick, Robert F. Service, Erik Stokstad, Paul Voosen, Meredith Wadman INTERNS Lucy Hicks, Cathleen O'Grady Bradley Bernstein, Mass. General Hospital David Masopust, U. of Minnesota
CONTRIBUTING CORRESPONDENTS Warren Cornwall, Ann Gibbons, Mara Hvistendahl, Sam Kean, Eli Kintisch, Kai Kupferschmidt Joseph J. Berry, NREL Geraldine Masson, CNRS
(Berlin), Andrew Lawler, Mitch Leslie, Eliot Marshall, Virginia Morell, Dennis Normile (Shanghai), Elisabeth Pain (Careers), Alessandra Biffi, Harvard Med. School C. Robertson McClung, Dartmouth College
Charles Piller, Michael Price, Tania Rabesandratana (Barcelona), Joshua Sokol, Emily Underwood, Gretchen Vogel (Berlin), Peer Bork, EMBL Rodrigo Medellín, U. Nacional Autónoma de México
Lizzie Wade (Mexico City) CAREERS Donisha Adams, Rachel Bernstein (Editor), Katie Langin (Associate Editor) COPY EDITORS Julia Chris Bowler, École Normale Supérieure Graham Medley, LSHTM
Cole (Senior Copy Editor), Cyra Master (Copy Chief) ADMINISTRATIVE SUPPORT Meagan Weiland Ian Boyd, U. of St. Andrews Jane Memmott, U. of Bristol
Emily Brodsky, UC Santa Cruz C. Jessica Metcalf, Princeton U.
Creative Director Beth Rakouskas Ron Brookmeyer, UCLA (S) Baoxia Mi, UC Berkeley
Christian Büchel, UKE Hamburg Edward Miguel, UC Berkeley
DESIGN MANAGING EDITOR Marcy Atarod GRAPHICS MANAGING EDITOR Alberto Cuadra PHOTOGRAPHY MANAGING EDITOR William Douthitt Dennis Burton, Scripps Res. Tom Misteli, NCI, NIH
WEB CONTENT STRATEGY MANAGER Kara Estelle-Powers DESIGN EDITOR Chrystal Smith DESIGNER Christina Aycock GRAPHICS Carter Tribley Butts, UC, Irvine Yasushi Miyashita, U. of Tokyo
EDITOR Nirja Desai INTERACTIVE GRAPHICS EDITOR Xing Liu SENIOR SCIENTIFIC ILLUSTRATORS Valerie Altounian, Chris Bickel György Buzsáki, New York U. School of Med. Alison Motsinger-Reif, NIEHS, NIH (S)
SCIENTIFIC ILLUSTRATOR Alice Kitterman SENIOR GRAPHICS SPECIALISTS Holly Bishop, Nathalie Cary SENIOR PHOTO EDITOR Emily Petersen Blanche Capel, Duke U. Danielle Navarro, U. of New South Wales
PHOTO EDITOR Kaitlyn Dolan WEB DESIGNER Jennie Pajerowski Annmarie Carlton, U. of California, Irvine Daniel Nettle, Newcastle U.
Nick Chater, U. of Warwick Daniel Neumark, UC Berkeley
Chief Executive Officer and Executive Publisher Sudip Parikh M. Keith Chen, UCLA Beatriz Noheda, U. of Groningen
Zhijian Chen, UT Southwestern Med. Ctr. Helga Nowotny, Vienna Sci., Res. & Tech. Fund
Publisher, Science Family of Journals Bill Moran Ib Chorkendorff, Denmark TU Rachel O’Reilly, U. of Birmingham
James J. Collins, MIT Harry Orr, U. of Minnesota
DIRECTOR, BUSINESS SYSTEMS AND FINANCIAL ANALYSIS Randy Yi DIRECTOR, BUSINESS OPERATIONS & ANALYSIS Eric Knott DIRECTOR OF ANALYTICS Robert Cook-Deegan, Arizona State U. Pilar Ossorio, U. of Wisconsin
Enrique Gonzales MANAGER, BUSINESS OPERATIONS Jessica Tierney SENIOR BUSINESS ANALYST Cory Lipman, Meron Kebede FINANCIAL Alan Cowman, Walter & Eliza Hall Inst. Andrew Oswald, U. of Warwick
ANALYST Alexander Lee ADVERTISING SYSTEM ADMINISTRATOR Tina Burks SENIOR SALES COORDINATOR Shirley Young DIGITAL/PRINT STRATEGY Carolyn Coyne, U. of Pittsburgh Isabella Pagano, Istituto Nazionale di Astrofisica
MANAGER Jason Hillman QUALITY TECHNICAL MANAGER Marcus Spiegler ASSISTANT MANAGER DIGITAL/PRINT Rebecca Doshi SENIOR CONTENT Roberta Croce, VU Amsterdam Margaret Palmer, U. of Maryland
SPECIALISTS Steve Forrester, Jacob Hedrick, Antoinette Hodal, Lori Murphy PRODUCTION SPECIALIST Kristin Wowk DIGITAL PRODUCTION Ismaila Dabo, Penn State U. Elizabeth Levy Paluck, Princeton U.
MANAGER Lisa Stanford CONTENT SPECIALIST Kimberley Oster ADVERTISING PRODUCTION OPERATIONS MANAGER Deborah Tompkins Jeff L. Dangl, U. of North Carolina Jane Parker, Max Planck Inst. Cologne
DESIGNER, CUSTOM PUBLISHING Jeremy Huntsinger SR. TRAFFIC ASSOCIATE Christine Hall SPECIAL PROJECTS ASSOCIATE Sarah Dhere Chiara Daraio, Caltech Giovanni Parmigiani, Dana-Farber Cancer Inst. (S)
Nicolas Dauphas, U. of Chicago Daniel Pauly, U. of British Columbia
ASSOCIATE DIRECTOR, BUSINESS DEVELOPMENT Justin Sawyers GLOBAL MARKETING MANAGER Allison Pritchard DIGITAL MARKETING MANAGER Christian Davenport, U. of Michigan Samuel Pfaff, Salk Inst. for Biological Studies
Aimee Aponte JOURNALS MARKETING MANAGER Shawana Arnold MARKETING ASSOCIATES Tori Velasquez, Mike Romano, Ashley Frans de Waal, Emory U. Julie Pfeiffer, UT Southwestern Med. Ctr.
Hylton DIGITAL MARKETING SPECIALIST Asleigh Rojanavongse SENIOR DESIGNER Kim Huynh Claude Desplan, New York U. Philip Phillips, UIUC
Sandra DÍaz, U. Nacional de CÓrdoba Matthieu Piel, Institut Curie
DIRECTOR AND SENIOR EDITOR, CUSTOM PUBLISHING Sean Sanders ASSISTANT EDITOR, CUSTOM PUBLISHING Jackie Oberst Ulrike Diebold,TU Wien Kathrin Plath, UCLA
Hong Ding, Inst. of Physics, CAS Martin Plenio, Ulm U.
DIRECTOR, PRODUCT & PUBLISHING DEVELOPMENT Chris Reid DIRECTOR, BUSINESS STRATEGY AND PORTFOLIO MANAGEMENT Sarah Whalen Dennis Discher, U. of Penn. Katherine Pollard, UC San Francisco
ASSOCIATE DIRECTOR, PRODUCT MANAGMENT Kris Bishop PRODUCT DEVELOPMENT MANAGER Scott Chernoff PUBLISHING TECHNOLOGY Jennifer A. Doudna, UC Berkeley Elvira Poloczanska, Alfred-Wegener-Inst.
MANAGER Michael Di Natale SR. PRODUCT ASSOCIATE Robert Koepke SPJ ASSOCIATE Samantha Bruno Fuller Raissa M. D'Souza, UC Davis Julia Pongratz, Ludwig Maximilians U.
Bruce Dunn, UCLA Philippe Poulin, CNRS
DIRECTOR, INSTITUTIONAL LICENSING Iquo Edim ASSOCIATE DIRECTOR, RESEARCH & DEVELOPMENT Elisabeth Leonard MARKETING MANAGER William Dunphy, Caltech Jonathan Pritchard, Stanford U.
Kess Knight SENIOR INSTITUTIONAL LICENSING MANAGER Ryan Rexroth INSTITUTIONAL LICENSING MANAGER Marco Castellan MANAGER, Christopher Dye, U. of Oxford Félix A. Rey, Institut Pasteur
AGENT RELATIONS & CUSTOMER SUCCESS Judy Lillibridge SENIOR OPERATIONS ANALYST Lana Guz FULFILLMENT COORDINATOR Melody Scott Edwards, Harvard U. Trevor Robbins, U. of Cambridge
Stringer SALES COORDINATOR Josh Haverlock Todd Ehlers, U. of TÜbingen Joeri Rogelj, Imperial College London
Jennifer Elisseeff, Johns Hopkins U. Amy Rosenzweig, Northwestern U.
DIRECTOR, GLOBAL SALES Tracy Holmes US EAST COAST AND MID WEST SALES Stephanie O'Connor US WEST COAST SALES Lynne Stickrod Tim Elston, U. of North Carolina Mike Ryan, UT Austin
US SALES MANAGER, SCIENCE CAREERS Claudia Paulsen-Young US SALES REP, SCIENCE CAREERS Tracy Anderson ASSOCIATE DIRECTOR, Andrea Encalada, U. San Francisco de Quito Shimon Sakaguchi, Osaka U.
ROW Roger Goncalves SALES REP, ROW Sarah Lelarge SALES ADMIN ASSISTANT, ROW Bryony Cousins DIRECTOR OF GLOBAL Nader Engheta, U. of Penn. Miquel Salmeron, Lawrence Berkeley Nat. Lab
COLLABORATION AND ACADEMIC PUBLISHING RELATIONS, ASIA Xiaoying Chu ASSOCIATE DIRECTOR, INTERNATIONAL COLLABORATION Grace Karen Ersche, U. of Cambridge Nitin Samarth, Penn State U.
Yao SALES MANAGER Danny Zhao MARKETING MANAGER Kilo Lan ASCA CORPORATION, JAPAN Kaoru Sasaki (Tokyo), Miyuki Tani Barry Everitt, U. of Cambridge Jürgen Sandkühler, Med. U. of Vienna
(Osaka) COLLABORATION/CUSTOM PUBLICATIONS/JAPAN Adarsh Sandhu Vanessa Ezenwa, U. of Georgia Erica Ollmann Saphire, La Jolla Inst.
Michael Feuer, The George Washington U. Alexander Schier, Harvard U.
DIRECTOR, COPYRIGHT, LICENSING AND SPECIAL PROJECTS Emilie David RIGHTS AND LICENSING COORDINATOR Jessica Adams RIGHTS AND Toren Finkel, U. of Pittsburgh Med. Ctr. Wolfram Schlenker, Columbia U.
PERMISSIONS ASSOCIATE Elizabeth Sandler CONTRACTS AND LICENSING ASSOCIATE Lili Catlett Gwenn Flowers, Simon Fraser U. Susannah Scott, UC Santa Barbara
Peter Fratzl, Max Planck Inst. Potsdam Rebecca Sear, LSHTM
MAIN HEADQUARTERS EDITORIAL PRODUCT ADVERTISING AAAS BOARD OF DIRECTORS Elaine Fuchs, Rockefeller U. Anuj Shah, U. of Chicago
& CUSTOM PUBLISHING Eileen Furlong, EMBL Vladimir Shalaev, Purdue U.
Science/AAAS [email protected] CHAIR Steven Chu Jay Gallagher, U. of Wisconsin Jie Shan, Cornell U.
1200 New York Ave. NW advertising.sciencemag.org/ PRESIDENT Claire M. Fraser Daniel Geschwind, UCLA Beth Shapiro, UC Santa Cruz
Washington, DC 20005 NEWS products-services PRESIDENT-ELECT Susan G.Amara Karl-Heinz Glassmeier, TU Braunschweig Jay Shendure, U. of Washington
TREASURER Carolyn N. Ainslie Ramon Gonzalez, U. of South Florida Steve Sherwood, U. of New South Wales
SCIENCE INTERNATIONAL [email protected] [email protected] Sandra González-Bailón, U. of Penn Brian Shoichet, UC San Francisco
CHIEF EXECUTIVE OFFICER Elizabeth Grove, U. of Chicago Robert Siliciano, Johns Hopkins U. School of Med.
Clarendon House INFORMATION FOR AUTHORS CLASSIFIED ADVERTISING Nicolas Gruber, ETH ZÜrich Lucia Sivilotti, U. College London
Clarendon Road Sudip Parikh Hua Guo, U. of New Mexico Alison Smith, John Innes Centre
Cambridge, CB2 8FH, UK sciencemag.org/authors/ advertising.sciencemag.org/ BOARD Cynthia M. Beall Kip Guy, U. of Kentucky College of Pharmacy Richard Smith, U. of North Carolina (S)
science-information-authors science-careers Rosina M. Bierbaum Taekjip Ha, Johns Hopkins U. Mark Smyth, QIMR Berghofer
SCIENCE CHINA Ann Bostrom Christian Haass, Ludwig Maximilians U. Pam Soltis, U. of Florida
REPRINTS AND PERMISSIONS [email protected] Stephen P.A. Fodor Sharon Hammes-Schiffer, Yale U. John Speakman, U. of Aberdeen
Room 1004, Culture Square S. James Gates, Jr. Wolf-Dietrich Hardt, ETH ZÜrich Tara Spires-Jones, U. of Edinburgh
No. 59 Zhongguancun St. sciencemag.org/help/ JOB POSTING CUSTOMER SERVICE Laura H. Greene Louise Harra, U. College London Allan C. Spradling, Carnegie Institution for Sci.
Haidian District, Beijing, 100872 reprints-and-permissions Kaye Husbands Fealing Jian He, Clemson U. V. S. Subrahmanian, Dartmouth College
employers.sciencecareers.org Maria M. Klawe Carl-Philipp Heisenberg, IST Austria Ira Tabas, Columbia U.
SCIENCE JAPAN MEDIA CONTACTS Robert B. Millard Ykä Helariutta, U. of Cambridge Sarah Teichmann, Wellcome Sanger Inst.
[email protected] Alondra Nelson Janet G. Hering, Eawag Rocio Titiunik, Princeton U.
ASCA Corporation [email protected] William D. Provine Hans Hilgenkamp, U. of Twente Shubha Tole, Tata Inst. of Fundamental Res.
Sibaura TY Bldg. 4F, 1-14-5 MEMBERSHIP AND INDIVIDUAL Kai-Uwe Hinrichs, U. of Bremen Wim van der Putten, Netherlands Inst. of Ecology
Shibaura Minato-ku MULTIMEDIA CONTACTS SUBSCRIPTIONS Deirdre Hollingsworth, U. of Oxford Reinhilde Veugelers, KU Leuven
Tokyo, 108-0073 Japan Lora Hooper, UT Southwestern Med. Ctr. Bert Vogelstein, Johns Hopkins U.
[email protected] sciencemag.org/subscriptions Fred Hughson, Princeton U. Kathleen Vohs, U. of Minnesota
[email protected] Randall Hulet, Rice U. David Wallach, Weizmann Inst. of Sci.
MEMBER BENEFITS Auke Ijspeert, EPFL Jane-Ling Wang, UC Davis (S)
INSTITUTIONAL SALES Akiko Iwasaki, Yale U. Jessica Ware, Amer. Mus. of Natural Hist.
AND SITE LICENSES aaas.org/membercentral Stephen Jackson, USGS and U. of Arizona David Waxman, Fudan U.
Kai Johnsson, EPFL Jonathan Weissman, UC San Francisco
sciencemag.org/librarian Peter Jonas, ISTAustria Chris Wikle, U. of Missouri (S)
Matt Kaeberlein, U. of Washington Terrie Williams, UC Santa Cruz
Science serves as a forum for discussion of important issues related to the advancement of science by publishing material on William Kaelin Jr., Dana-Farber Cancer Inst. Ian A. Wilson, Scripps Res. (S)
which a consensus has been reached as well as including the presentation of minority or conflicting points of view. Accordingly, Daniel Kammen, UC Berkeley Yu Xie, Princeton U.
all articles published in Science—including editorials, news and comment, and book reviews—are signed and reflect the individual V. Narry Kim, Seoul Nat. U. Jan Zaanen, Leiden U.
views of the authors and not official points of view adopted by AAAS or the institutions with which the authors are affiliated.

10 2 OCTOBER 2020 • VOL 370 ISSUE 6512 sciencemag.org SCIENCE

International
Research Leader
Grants

Stage 1 application period for both calls
September 9, 2020 – October 21, 2020

Exceptional Grants for Exceptional Scientists
Two calls for applications within the areas of biomedicine &
biotechnology. International Research Leader Grants from
the Novo Nordisk Foundation are for outstanding scientists
to establish and run their laboratories in Denmark.

Novo Nordisk Foundation Novo Nordisk Foundation

Laureate Research Grants Young Investigator Awards

» Individual grants up to » Individual grants up to

DKK 50 million over 7 years DKK 25 million over 7 years
(EUR ~6.7 million, USD ~7.4 million) (EUR ~3.4 million, USD ~3.7 million)

» For principal investigators who have » For principal investigators who have
directed an independent research group for directed an independent research group for
7 or more years. less than 7 years.

» Grant holders can apply for continued Lau- » Award holders can apply for further funding
reate Research Grant funding, up to DKK 35 from other Novo Nordisk Foundation grant
million over 7 additional years. programs.

To learn more, please visit: novonordiskfoundation.com

At Vertex, science leads the way,
but our people lead the science.

We believe that by bridging science, technology and business,
we can transform our growing global biotechnology company
in ways that will further our mission to create transformative

medicines for people with serious diseases. We’re hiring in
Cell and Gene Therapy, Manufacturing, Clinical Development

and Medical Affairs, Research, Quality and more.

Find open positions at

careers.vrtx.com

©2020 Vertex Pharmaceuticals Incorporated
Vertex and the Vertex triangle logo are registered trademarks of Vertex Pharmaceuticals Incorporated.

EDITORIAL

Fire in our future

I t can seem like Earth itself is on fire. In places such Within limits, we can dampen fire intensities by modi- William Wallace
as Australia and California for which fire is a natu- fying the landscapes that fire feeds upon, and we can Covington
ral feature, landscapes are burning at historic if harden communities to keep embers blown from the is Emeritus Founding
not epic scales. In the Arctic and Greenland, where countryside from metastasizing into urban confla- Executive Director
fire is rare, tundra is smoldering and melting per- grations. The strategies are the same as those used to of The Ecological
mafrost. In Amazonia, Indonesia, and Mediterra- contain urban fire. Concepts like the home ignition Restoration Institute
nean Europe, fires are interacting with the land zone—the house and its immediate surroundings— and Emeritus
Professor of Forestry
clearing of rainforest, the draining of peatlands, and identify points of vulnerability. Long-extant programs at Northern Arizona
University, Flagstaff
the abandonment of rural lands to create damaging, like Firewise, which also add concepts like defensible AZ, USA. w.wallace.
[email protected]
even lethal, conditions. space, promote suites of tried-and-tested techniques to
Stephen Pyne
There is no single driver except humanity behind this communities in nearly all kinds of environments. is an emeritus
professor at Arizona
outbreak. But increasingly, anthropogenic climate change In montane forests like the ponderosa pine of the State University,
Tempe, AZ, USA,
is recognized as an enabler, performance enhancer, and Southwest, research shows that thinning and burning are and a fire historian.
stephen.pyne@
globalizer. Fire seasons are lengthening, fire severity is effective methods to reduce fuel loads and allow surface asu.edu

escalating, and collateral damages are compounding. fires to return. But many techniques are available, includ-

Is this a “wicked” problem so entangled with scien- ing prescribed grazing, the use of managed wildfire, and

tific and social complexities that solu- varieties of mechanical treatments like

tions are impossible? We think not. chipping and masticating. Most places

We need to unbundle “fire” in all its “There is will need a cocktail of treatments, ap-
shape-shifting avatars into manageable no single propriate to their local conditions.
pieces. Some issues will have technical driver except
solutions—fires sparked by powerlines Smart treatments, done well, will en-
can be prevented. Some involve knotty hance ecological integrity at the same
ecological processes: Lands that have time that they reduce hazardous fu-
had fires removed can suffer an eco- els. Thinning, for example, resembles
woody weeding and unlike logging

logical fire deficit for which reinstating humanity…” removes the small stuff that powers
flame can be as complicated as restor- fire. Moreover, fire is a biochemical

ing a vanished species. Most of the process, not just a flaming woodchip-

problems involve clashes of cultural per. Fire as fire matters biologically.

values over how we get energy, organize our economy, Good fire can provide herd immunity against bad fire.

and choose to live on the land. These will demand a Yet all these interventions will be overpowered un-

political resolution. less climate change is brought to heel. Paradoxically, as

Scales matter. Some reforms can be applied imme- we ratchet down our binge-burning of fossil fuels, we’ll

diately and locally, as with protecting towns. Others have to ratchet up our burning of living landscapes to

will require decades of work across countries and grant them the robustness they will need to survive the

regions. Restoring a suitable regimen of fire to tens stresses to come.

of millions of hectares will be an arduous exercise in Science can’t do all the intellectual lifting. Fire is sys-

adaptive management. Confronting the effects of cli- temic: We need a systemic cultural response. We need

mate change will likely prove a century-long quest, but art, new narratives and a poetry of flame, a revamp-

unless we reverse trends, they will overwhelm what- ing of liability laws to make controlled burning a de-

ever type of management is implemented. We need to fault choice, a restoration of traditional knowledge to

pursue all levels simultaneously. broaden techniques and purposes, a politics that can

Begin with ignition. Research shows that nationally, see the flames behind the smoke and engage with those

97% of the fires that have threatened houses are started who must live with its choices. In the end, science can

by people. There will always be accidental ignitions, advise; it can’t decide.

and in the West and Florida, lightning kindles many But we need a solid empirical basis for the tough de-

fires. But prevention programs can reduce the risk to cisions heading our way. We need what science can do

manageable levels. best, and the best of what science can do.

Still, fires will escape. The power of fire, however,

resides in its capacity to spread and inflict damages. –William Wallace Covington and Stephen Pyne

SCIENCE sciencemag.org 10.1126/science.abe9780
2 OCTOBER 2020 • VOL 370 ISSUE 6512 13

NEWS

The global toll of COVID-19 this week passed 1 million deaths. A mourner cries near where a relative was laid to rest at the Cementerio General in La Paz, Bolivia.
This summer, the cemetery saw as many as 90 burials per day and built three new pavilions and two mass graves.

IN BRIEF
Edited by Jeffrey Brainard

U.S. is far from herd immunity Trump wants shorter student visas Even if finalized, however, the changes are PHOTO: GASTON BRITO/GETTY IMAGES
unlikely to take effect if President Donald
COV I D -19 | Fewer than one in I M M I G RAT I O N | The Trump administra- Trump loses the November election.
10 Americans carried antibodies to the tion last week proposed tightening visa
pandemic coronavirus in late July, policies for international students, a Genome gaps nearly filled
according to a nationwide study of step it says is necessary to monitor their
dialysis patients. One of the largest of its academic progress and protect national GENETICS | Despite the fanfare that
kind, the study indicates that the United security. But many university officials greeted the first human genome sequence
States is a long way from reaching “herd believe the new rules, if adopted, would 17 years ago, researchers knew it was in-
immunity” to COVID-19, when a large reinforce the message that foreign stu- complete. For the sake of speed and because
portion of the population would become dents aren’t welcome. One change would of limitations in technology, teams had
resistant to the virus that causes the impose a 4-year limit on their stay; a visa skipped over regions of nearly identical
disease, slowing the rate of new infec- is now good for the duration of a stu- DNA, which are difficult to map correctly.
tions. Researchers reported last week in dent’s academic program. And the limit Last week, using newer technologies, a
The Lancet that they found the virus in would only be 2 years for students from team of 70 researchers released a human
8% of samples from leftover plasma of 59 countries where more than 10% of visa cell line’s genome sequence with almost all
28,503 people. Residents of neighbor- holders violate the terms of their visit. those difficult parts deciphered. It’s a first
hoods that are predominantly Black or The 4-year limit would apply to students attempt to sequence each chromosome from
Hispanic or are densely populated were from China and India, which combined telomere to telomere, meaning from one
significantly more likely to have antibod- are home to more than half the total end to the other. The work fills in about
ies to the virus. The researchers also number of international students in the 135 million missing base pairs, bringing the
found regional differences: Thirty-three United States. Any student wanting to total to 3.057 billion bases, the Telomere-to-
percent of New York state samples had remain to complete a degree would need Telomere consortium announced on Twitter
antibodies, but only 3.8% of California to have “a compelling academic reason” or and in the Genome Informatics Section of
samples did. face “circumstances beyond their control.” GitHub. Still missing are arrays of genes

14 2 OCTOBER 2020 • VOL 370 ISSUE 6512 sciencemag.org SCIENCE

coding for ribosomal RNA, an essential part A partisan divide was smaller in other countries. Still,
of the cell’s protein-building machinery. respondents in all countries had high
The consortium hopes to finish mapping Share of survey respondents, by political regard for scientists, with 36% express-
remaining stretches within 1 year. orientation, who trust scientists “a lot” ing “a lot of trust” in them, matching the
military for most trusted and besting
France to pair science, media Left Right Diference perceptions of government and business
leaders and the news media.
POLICY | France plans to bring scientists United States 62 20 +42
and journalists closer together in an effort U.S. census dispute is extended
to boost public access to reliable scientific Canada 74 35 +39
information and combat misinformation. DEMOGRAPHY | The political battle over
Parliament is expected to approve the Australia 68 39 +29 when to stop tracking down Americans
project in coming weeks as part of a 10-year who haven’t completed the 2020 U.S.
science plan. The science ministry said United Kingdom 62 35 +27 census continued this week. After a
the plan is needed “at a time when French court ruling last week overturned the
society is crossed by currents of irrationality Germany 55 38 +17 Trump administration’s deadline of
and doubts about progress and knowledge.” 30 September, the Census Bureau said
An early draft of the bill called for a sci- Italy 47 31 +16 on 28 September it would finish the
ence media center similar to those in the fieldwork by 5 October. Critics have
United Kingdom and Germany. But after France 34 35 –1 worried ending the work too soon could
critics said funders of a French center could cause a large undercount. In April,
distort its agenda and threaten journalistic Brazil 22 22 0 the Census Bureau said the COVID-19
independence, the bill was reworded. It now pandemic had forced delays in deploying
calls for a “series of actions or network.” m some 500,000 door-to-door enumerators.
But in early August, the White House
HHMI mandates open access U.S. leads in science trust gulf said the fieldwork must wrap up by the
end of September. City and state officials
PU B L I S H I N G | The Howard Hughes PUBLIC OPINION | When it comes to trust- and civil rights groups sued to block that
Medical Institute (HHMI), one of the largest ing scientists, Americans are the most plan, leading to last week’s ruling.
research philanthropies, said on 1 October split ideologically, a survey of 20 coun-
it will begin to require its staff scientists to tries has found. Some 62% of Americans
make research papers in which they played who identified their politics as left wing
a leading role immediately free to read. said they held “a lot” of trust in scien-
The new policy will replace the current tists “to do what is right” for the public,
requirement that HHMI papers become the Pew Research Center reported on
open access within 12 months of publica- 29 September. But only 20% of those who
tion. HHMI-supported scientists can comply identified with the right wing expressed
with the rule, which takes effect in January that level of confidence. The polarization
2022, by publishing in journals with open-
access options or depositing a near-final, Trade in Sri Lanka’s
peer-reviewed version in a free archive. critically endangered
They can also use HHMI funds to publish in Knuckles pygmy
hybrid journals, which allow authors to pay lizard was banned
for open-access publication but also charge last year.
subscriptions, but only if the journal is tran-
CREDITS: (GRAPHIC) C. SMITH/SCIENCE; (DATA) PEW RESEARCH CENTER; (PHOTO) FRANK CANON sitioning to immediate open access for all C O N S E R VAT I O N
content. (For nonprofit publishers’ journals,
this provision takes effect in 2023.) HHMI Unregulated reptile trade threatens species
also said this week it will join Coalition S,
which includes funders mostly in Europe, in M ore than one-third of reptile species, or nearly 4000, have been offered for sale
its campaign to flip journals to immediate online in recent years, researchers at the Xishuangbanna Tropical Botanical
open access. The new policy, which applies Garden in China found when they searched 151 websites for ads for pet reptiles
to papers on which an HHMI scientist is placed between 2000 and 2019. Some species for sale are endangered, reports
the first, last, or corresponding author, may the study, published on 29 September in Nature Communications. More than
divert papers from top-tier journals that half of reptiles imported to the United States are captured from the wild, and the trade
lack open-access options, acknowledges bio- of most reptiles is not regulated, the study notes, even though researchers don’t know
chemist Erin O’Shea, HHMI’s president. In enough about some 1500 species to tell whether they are endangered. The Convention
2019, 13% of all papers in Cell, 5% in Nature, on International Trade in Endangered Species of Wild Fauna and Flora (CITES) moni-
and 7% in Science would have been subject tors the trade of just 9% of reptile species, typically those sold in larger volumes, such
to the new requirement, the institute said. as crocodiles and pythons, whose skin is used in shoes, wallets, and other items.
O’Shea says the open-access model helps In all, CITES regulates trade of 856 reptile species and has banned sales for 98. The
accelerate science, so she hopes those sub- International Union for Conservation of Nature lists 1406 as species of concern.
scription journals and others will find ways
to help HHMI authors comply.

SCIENCE sciencemag.org 2 OCTOBER 2020 • VOL 370 ISSUE 6512 15

At 1-kilometer resolution, a European climate model
(left) is nearly indistinguishable from reality (right).

IN DEPTH deputy director of research, who coordinated
Extreme Earth and has been advising the Eu-
CLIMATE CHANGE ropean Union on the new program.

Europe builds ‘digital twin’ of Typical climate models run at resolu-
Earth to hone climate forecasts tions of 50 or 100 kilometers; even top
ones like ECMWF’s “European” model
Ingesting more data than ever before, exascale model run at 9 kilometers. The new model’s
will simulate the impact of climate change on humans 1-kilometer resolution will enable it to di-
rectly render convection, the vertical trans-
By Paul Voosen tailed real-time data than ever before. The port of heat critical to the formation of IMAGES: (LEFT TO RIGHT) ECMWF; © EUMETSAT
project, which will be described in detail in clouds and storms, rather than relying on an
T he European Union is finalizing two workshops later this month, will start algorithmic approximation. “I call it the third
plans for an ambitious “digital twin” next year and run on one of the three super- dimension of climate modeling,” says Bjorn
of planet Earth that would simulate computers that Europe will deploy in Fin- Stevens, a climate scientist at the Max Planck
the atmosphere, ocean, ice, and land land, Italy, and Spain. Institute for Meteorology. The model will also
with unrivaled precision, provid- simulate the ocean in fine enough detail to
ing forecasts of floods, droughts, and Destination Earth rose out of the ashes capture the behavior of swirling eddies that
fires from days to years in advance. Destina- of Extreme Earth, a proposal led by the Eu- are important movers of heat and carbon.
tion Earth, as the effort is called, won’t stop ropean Centre for Medium-Range Weather
there: It will also attempt to capture human Forecasts (ECMWF) for a billion-euro flag- In Japan, pioneering runs of a 1-kilometer
behavior, enabling leaders to see the impacts ship research program. The European Union global climate model have shown that di-
of weather events and climate change on ultimately canceled the flagship program, rectly simulating storms and eddies leads to
society and gauge the effects of different cli- but retained interest in the idea. Fears that better short-term rainfall predictions. But it
mate policies. Europe was falling behind China, Japan, and should also improve climate forecasts over
the United States in supercomputing led to periods of months and years. Recent work
“It’s a really bold mission, I like it a lot,” the European High-Performance Computing has shown climate models are not captur-
says Ruby Leung, a climate scientist at the Joint Undertaking, an €8 billion investment ing predictable changes in wind patterns
U.S. Department of Energy’s (DOE’s) Pacific to lay the groundwork for eventual “exascale” that drive swings in regional temperature
Northwest National Laboratory. By render- machines capable of 1 billion billion calcu- and rainfall—probably because the models
ing the planet’s atmosphere in boxes only lations per second. The dormant Extreme fail to reproduce storms and eddies (Science,
1 kilometer across, a scale many times finer Earth proposal offered a perfect use for such 31 July, p. 490).
than existing climate models, Destination capacity. “This blows a soul into your digital
Earth can base its forecasts on far more de- infrastructure,” says Peter Bauer, ECMWF’s The high resolution will also enable Des-
tination Earth to base its forecasts on more
detailed data. Weather models suck in obser-
vations of temperature and pressure from
satellites, weather stations, aircraft, and
buoys to guide their simulations. But coarse
grids mean the models can’t assimilate mea-
surements that don’t average well or cover
broad areas, such as fractures opening up in
sea ice. Destination Earth will close this gap,
says Sandrine Bony, a cloud scientist at the
Pierre Simon Laplace Institute. “The scales
that are resolved are closer to the scales that
are measured.”

The model will also incorporate real-time
data charting atmospheric pollution, crop
growth, forest fires, and other phenomena
known to affect weather and climate, says
Francisco Doblas-Reyes, an earth system
scientist at the Barcelona Supercomputing
Center. “If a volcano goes off tomorrow, that’s
important for the risk of tropical precipita-
tion failure in a few months.” And it will
fold in data about society, such as energy
use, traffic patterns, and human movements
(traced by mobile phones).

The goal is to allow policymakers to di-
rectly gauge how climate change will impact
society—and how society could alter the tra-
jectory of climate change. For example, the
model could predict how climate change

16 2 OCTOBER 2020 • VOL 370 ISSUE 6512 sciencemag.org SCIENCE

NEWS

will affect agriculture and migration pat- CLIMATE CHANGE
terns in Brazil—and also how cuts in ethanol
subsidies might limit deforestation in the China’s bold climate pledge
Amazon. Currently, climate scientists extract earns praise—but is it feasible?
regional results from global climate models
and pass them to experts in agriculture or Plan to achieve carbon neutrality by 2060 would make
economics to understand effects on human
behavior. Now, says Erin Coughlan de Perez, China a global leader, but abandoning coal will be hard
a climate hazard scientist at the Red Cross
Red Crescent Climate Centre, modelers are QQ群: 1074370165
“moving from just forecasting what weather
will be, to what the weather will do.” By Dennis Normile opia,” Vance Wagner of Energy Foundation
China wrote in a piece published online by
Getting there won’t be easy. Exascale C hina’s surprise pledge last week to the nonprofit China Dialogue.
supercomputers rely on both traditional cut its net carbon emissions to zero
computer chips as well as graphical pro- within 40 years has reignited hopes China had previously said its CO emis-
cessing units (GPUs), which are efficient of limiting global climate change to 2
at handling intensive calculations. GPUs tolerable levels. The country is the
are good for running model components in world’s largest producer of carbon sions would peak “around” 2030, a target
parallel and training artificial intelligence dioxide (CO ), accounting for 28% of global most analysts considered within reach. But
algorithms—two techniques Destination achieving carbon neutrality before 2060
Earth will lean on to enhance performance. 2 will require drastically reducing the use of
But old climate modeling code will have fossil fuels in transportation and electric-
to be reworked. ECMWF has a head start: emissions, and its move may inspire other ity generation and offsetting any remaining
It is adapting its forecast model to a GPU- countries to follow suit. But observers warn emissions through carbon capture and stor-
based environment, and last year tested it at that China faces daunting challenges in age or planting forests.
1-kilometer resolution for four simulated reaching its goals. Kicking its coal habit will
months on Summit, the U.S. supercomputer be particularly hard. China has not yet revealed details of
that was the world’s fastest until a Japanese how it will do this. But a research group at
machine recently eclipsed it. “We aim to have CO emissions peak be- Tsinghua University presented a $15 tril-
2 lion, 30-year road map on 27 September
The massive amount of data generated that calls for ending the use of coal for
by the model will be a problem of its own. fore 2030 and achieve carbon neutrality electricity generation around 2050, dra-
When the Japanese team ran its 1-kilometer- before 2060,” Chinese President Xi Jinping matically increasing nuclear and renew-
scale experiment, it took half a year to extract told the United Nations General Assembly able power generation, and relying on
something useful from a couple days of data, via a video link on 22 September. That’s “a electricity for 80% of China’s energy con-
Doblas-Reyes says. “There’s a bottleneck very significant and encouraging announce- sumption by 2060.
when we try to access the data and do some- ment,” says Josep Canadell, an earth system
thing clever with it.” A big part of Destination scientist at Australia’s Commonwealth Sci- Coal is both the biggest challenge and
Earth will be solving this problem, designing entific and Industrial Research Organisa- an opportunity. Last year, the carbon-heavy
ways to analyze model results in real time. tion. He says the new targets “won’t likely fuel accounted for about 58% of China’s
let us to stop at 1.5° Celsius [of global warm- total energy consumption and 66% of its
As an operational system, Destination ing],” the preferred target set in the 2015 electricity generation. In coal-producing
Earth will likely run at several time scales, Paris agreement. “But below 2° might still regions, coal is also used to heat buildings.
Bauer says. One will be near daily, perhaps be consistent with [Xi’s] announcement.” Recent advances in renewable energy have
targeting individual extreme weather events China’s commitment also “ratchets up pres- made replacing coal easier than cutting oil
weeks or months in the future. Runs in the sure on other major emitters” to set more use in transportation and emissions from
other mode—long-term predictions—would ambitious targets “while further isolating farm fields and livestock. “The power sec-
be less frequent: perhaps a decadelong pre- the Trump administration in its climate my-
diction of the climate made every half-year
PHOTO: XU CONGJUN/AP PHOTO or so. “If this works, it could be a template A coal-fired power plant in Jiangsu province. Coal accounted for 58% of China’s energy consumption last year.
for other countries to follow,” Bauer says.

The Europeans aren’t alone in planning
for exascale climate models. “We’re head-
ing in that direction as well, but we’ve yet to
reach that level of effort,” says Leung, who
serves as chief scientist for DOE’s earth sys-
tem model.

Stevens says it’s thrilling to be involved in
a truly planetary-scale information system
that can reveal not just the proverbial but-
terfly effect in weather and climate, but also
how local human actions manifest globally.
“That’s the story of globalization. That’s the
story of the Anthropocene. And this is the
scientific platform that will allow you to ex-
plore those.” j

SCIENCE sciencemag.org 2 OCTOBER 2020 • VOL 370 ISSUE 6512 17

NEWS | IN DEPTH

tor is the part of the energy system where Many koalas were killed by Australia’s record wildfires, jeopardizing the survival of some populations.
zero emission technologies are the most
mature and economically competitive,” says CONSERVATION BIOLOGY
Lauri Myllyvirta, an air pollution analyst
at the Centre for Research on Energy and Record U.S. and Australian fires
Clean Air in Helsinki. Zero-carbon electric- raise fears for many species
ity could make charging electric vehicles
cleaner and supplant coal for heating. Scientists say fires likely wiped out some rare Australian
organisms, and worry U.S. blazes now threaten more
But it will require a U-turn. A recent
study by Myllyvirta and colleagues found By John Pickrell and Elizabeth Pennisi mate. “We are in uncharted territory here,” PHOTO: PETER PARKS/AFP/GETTY IMAGES
that China’s coal-fired generating capacity says ecologist S. Mažeika Patricio Sullivan
grew by about 40 gigawatts (GW) in 2019, F or the past 3 months, arachnologist of Ohio State University, Columbus. “We
to about 1050 GW. Another 100 GW is under Jess Marsh has been searching for just don’t know how resilient species and
construction and coal interests are lobbying the Kangaroo Island assassin spider. ecosystems will be to wildfires of the mag-
for even more plants. “This is all despite Early this year, during the worst fire nitude, frequency, and intensity that we are
significant overcapacity in the sector,” with season ever recorded in Australia, currently experiencing in the U.S. West.”
plants running at less than 50% of capac- a wildfire charred the spider’s only
ity and many coal-power companies losing known home on an island off the nation’s Australia’s postfire experience offers
money, the study said. Canadell says the south coast. Now, Marsh fears the tiny, rusty cause for anxiety, researchers say. From
building boom is the result of misplaced brown arachnid is another of the many Aus- September 2019 to March, more than
incentives to build coal plants and create tralian species that the blazes have put on 11 million hectares burned, mostly in the
construction jobs. He predicts many of the a path to extinction: Countless hours of continent’s southeastern forests, killing at
new plants will barely be used or become scouting haven’t revealed a single survivor. least 34 people. More than 20% of the na-
stranded assets that have to be written off. “Its habitat is completely incinerated,” says tion’s total forest cover was lost, researchers
Marsh, who is affiliated with Charles Dar- at Western Sydney University reported in
A related challenge will be reforming win University. February. Even normally fire-proof rainfor-
the electricity market. Renewable energy ests and wetlands were scorched (Science,
is increasingly cost competitive with coal, She isn’t the only field biologist worried 20 December 2019, p. 1427). By one esti-
says Li Shuo, a climate policy adviser to that the record wildfires around the globe mate, released early this year by the Aus-
Greenpeace China. But regulators allocate are inflicting lasting damage on species and tralian government, 114 threatened plant
operational time among electricity plants ecosystems. Even as Australia tallies the and animal species lost 50% to 80% of their
to match generation to demand, with little damage from its blazes, the worst fires in habitats; 327 species saw more than 10% of
consideration of economic or environmen- more than 70 years are burning in Califor- their ranges burn.
tal implications, Li says. The system over- nia, Oregon, and Washington; so far, they
whelmingly favors coal-fired generation, have consumed some 2 million hectares, Those estimates, however, were based on
partly because it doesn’t suffer from the killing at least 35 people. As in Australia, satellite data, says John Woinarski, also at
variability of wind and solar power. The scientists fear the loss of habitat has threat- Charles Darwin University. To get better as-
uncertain market access has already slowed ened species with small populations or re- sessments, researchers have been trying to
investment in renewables, Li says. Given stricted ranges, and could potentially lead visit burned sites, an effort complicated by
the power of coal and construction inter- to permanent ecological changes if burned the COVID-19 pandemic.
ests, the needed reforms will take consider- landscapes fail to rebound in a warming cli-
able political will. In some cases, they’ve reported good
news. There was grave concern for the en-
Expanding nuclear power presents chal-
lenges as well. The 2011 Fukushima Daiichi
nuclear disaster in Japan sent ripples of
concern through China, which mandated
additional safety measures that made new
plants more expensive. Public opposition is
also growing. China has 48 nuclear power
reactors in operation and 12 under con-
struction, according to the World Nuclear
Association. The government had aimed
for 58 GW of nuclear capacity by this year
but did not get beyond 52 GW.

China’s Five-Year Plan for 2021–25, now
being drafted, may contain concrete mea-
sures to help realize Xi’s ambitious target.
“China’s interest in climate change has
waned in recent years, due to the slowing
down of economic growth and the U.S.
withdrawal from the Paris agreement,” says
Zhang Junjie, an environmental economist
at Duke Kunshan University. “The commit-
ment on carbon neutrality reignited hopes
for China’s climate action.” j

18 2 OCTOBER 2020 • VOL 370 ISSUE 6512 sciencemag.org SCIENCE

IMAGE: ANDY WALKER/MIDLAND FERTILITY SERVICES/SCIENCE SOURCE dangered Kangaroo Island glossy black Research Station and M. Zachariah Peery DEVELOPMENTAL BIOLOGY
cockatoo after 75% of its habitat burned. from the University of Wisconsin (UW),
But observers have seen large flocks move Madison, found. In 2015, some 22% of nest- Stem cell
to unburned areas of the island, says ing sites used by the birds in 2014 were not studies probe
Karleah Berris of Natural Resources Kanga- reoccupied and still are empty, Jones says, origins of
roo Island. And many of the birds appear and this year’s fires could add to the losses. the placenta
to have bred and fledged young. “It seems Western fires also threaten the white-
they are coping with the reduction in food headed woodpecker, found only in pine Lab models of organ will
by [moving] to where the food is,” she says. forests in the Pacific Northwest and Califor- track how it emerges—and
Researchers were also worried about the nia, and the Grace’s warbler, limited to pine what can go wrong
Kangaroo Island dunnart, a shrew-size car- and oak forests in the southwestern United
nivorous marsupial. Even before the fires, States and northern Mexico, says wild- By Kelly Servick
just 500 or so remained, and they lost 95% life biologist Vicki Saab, also at the Rocky
of their habitat to the flames. But auto- Mountain Research Station. T he placenta—a Frisbee-size hunk of
mated cameras have revealed that at least tissue that chaperones a fetus in the
some dunnarts survived, and managers Plants that have small ranges and are uterus only to be tossed aside in the
moved quickly to build fences to protect the found in burned areas, such as the Coulter delivery room—has mysterious begin-
remaining animals from feral cats. pine in California, might also face trouble, nings. The organ emerges from cells
says Camille Stevens-Rumann, a fire eco- that develop alongside the embryo,
Other findings are more ominous. In New logist at Colorado State University, Fort and that have been difficult to grow in the lab.
South Wales, fires killed about one-third Collins. “California especially has a lot of Now, researchers have devised a way to de-
of the state’s koalas, a government inquiry endemic plant species that could be very rive and observe early precursors of placental
found in July. It warned that the marsupial much impacted,” she says. cells in a dish. They have found a method of
would be extinct in the state by 2050 if dra- “reprogramming” adult cells, reverting them
matic measures are not taken to conserve it. The longer term consequences for eco-
And in the state’s Nightcap National Park, systems are harder to predict, research- The outer layer of cells (thin ring above) in a several-
a survey found that fires destroyed 10% or ers say. In both Australia and the Western days-old human embryo goes on to form the placenta.
more of the remaining stands of several criti- United States, many ecosystems are adapted
cally endangered rainforest trees. Some spe- to fire and even require it to thrive. “Many to a primitive state, that can prompt them to
cies were down to fewer than 200 trees before of the old-growth forests we know and love become trophoblast stem cells (TSCs), which
the fires, says botanist Robert Kooyman in the Pacific Northwest were born of large give rise to placental cells.
of Macquarie University; they are now “cer- and severe fires centuries ago,” says Brian
tainly a few steps closer to extinction.” Harvey, a wildfire ecologist at the Univer- The method promises a window on how
sity of Washington, Seattle. Fires can also defects in placental development may lead
Such concerns have prompted scientists help create a mosaic of habitats that sup- to infertility, miscarriage, and preeclampsia,
to ask Australia’s government to expand its port a wealth of species, he and others note. a dangerous complication of pregnancy. “It’s
endangered species list. At least 41 verte- like gaining a toehold on Mars,” says repro-
brates that were not endangered before the But climate change adds to the ductive biologist Susan Fisher at the Univer-
fires now face existential threats, Woinarski uncertainty about how forests will respond sity of California, San Francisco. “We know
and others reported in July in Nature this time. “The postfire climate is likely to be almost nothing about the early steps.”
Ecology & Evolution. An additional 21, al- warmer and drier than when the parent trees
ready tagged as threatened, might need established long ago,” Harvey says, making it Those steps begin just days after a sperm
greater protection. Marsh has recom- harder for ecosystems to recover, and bod- and egg join. “The first decision in human
mended adding 16 invertebrate species ing more fire in the future. “Just a little more life is to set aside the placental, supportive
found on Kangaroo Island to the list, in- drought can lead to much bigger fires,” says
cluding the assassin spider. “That species is Monica Turner, a fire ecologist at UW who
really hanging in the balance,” she says. calls climate change “a threat multiplier.”

In the United States, researchers say it’s Already, some ecosystems in North Amer-
too soon to know how many species the ica that have had frequent or intense burns
fires have put in jeopardy. But there are are not regenerating. In some places, such
already worrying reports. In Washington, as the sagebrush ecosystem of the Great
biologists estimate the fires have killed 50% Basin west of the Sierra Nevada mountain
of the state’s endangered pygmy rabbits, range and forests in the Klamath Mountains
which inhabit sagebrush flats that burned along the California-Oregon border, invasive
this year. They believe only about 50 of shrubs or grasses appear to have taken over.
North America’s smallest rabbit remain. Of- Because the invaders burn frequently, they
ficials estimate the flames have also killed appear to be preventing seedlings from ma-
30% to 70% of the state’s sage grouse and turing. In Australia, researchers have similar
sharp-tailed grouse, birds that also depend concerns. In the state of Victoria, forests of
on sagebrush. alpine ash, a towering eucalyptus tree found
in moist regions, historically experienced
In California, the impact of fires in 2014 fires less than once a century or so. Now,
may offer a preview. After flames swept some forests have been hit by five fires in the
through habitat of the endangered spot- past 20 years, and scientists fear some of the
ted owl, many of the birds abandoned stately groves will disappear for good. j
nesting sites, biologists Gavin Jones of
the U.S. Forest Service’s Rocky Mountain John Pickrell is a journalist in Sydney.

SCIENCE sciencemag.org 2 OCTOBER 2020 • VOL 370 ISSUE 6512 19

NEWS | IN DEPTH

cells,” says Kathy Niakan, a developmental are key to maintaining a pregnancy. David, VOICES OF THE PANDEMIC

biologist at the Francis Crick Institute, whose a co-author on that paper, separately used The ‘bat man’

team reported key molecular signals for that gene expression data from human embryos tackles

initial step in Nature last week. These cells to estimate that his own group’s lab-derived COVID-19

go on to form the trophoblast, a multilayered TSCs are equivalent to those seen 8 to After a career investigating
why so many viruses
ring that surrounds the embryo and helps it 10 days after fertilization, the team reported come from bats, Linfa Wang
eyes a new challenge
implant into the wall of the uterus. Some of on 15 September in a preprint on bioRxiv.
By Kai Kupferschmidt
these cells, TSCs, then give rise to cell types It will be important to thoroughly compare
B y pure chance, Linfa Wang, one of the
that will make up the bulk of the placenta, these induced TSCs to placenta-derived and world’s foremost experts on emerg-
ing viruses, was in the Chinese city
which enables mother and fetus to exchange ES cell–derived TSCs, says Washington Uni- of Wuhan in January. The biologist
was visiting collaborators at the Wu-
nutrients and gases and helps protect the fe- versity in St. Louis stem cell biologist Thorold han Institute of Virology (WIV) just
as SARS-CoV-2 was starting to spread from
tus from the mother’s immune system. Theunissen, whose team recently derived the city to the rest of the world. Even among
those experts there was little fear then. “I was
Scientists have derived TSC-like cells TSCs from ES cells. That analysis should in- mixing with all the lab people,” Wang says.
“We would go to a restaurant every night.”
from unused embryos created for in vitro clude comparing the chemical tags on DNA
Only when he left on 18 January did he
fertilization (IVF) or from the placentas of that influence cell function and sizing up realize how serious the situation was. At the
airport, staff checked his temperature three
terminated pregnancies, but both are lim- how efficiently the cells differentiate into dif- times before he could board his flight home
to Singapore. Five days later, Wuhan, a city
ited resources. And in a dish, these cells ferent types of specialized trophoblast cells. of 11 million people, was shut down. Wang
later learned that a woman on his plane had
have tended to mature and stop dividing. Induced TSCs could now be used to study carried the virus; luckily, he was not infected.

The same has been true of TSC-like cells genetic defects that can end a pregnancy, says Wang, who heads the Emerging Infec-
tious Diseases Program at Duke-NUS Medi-
created from cultured embryonic stem (ES) Soumen Paul, a stem cell biologist at the Uni- cal School in Singapore, immediately got to
work developing a new assay that can detect
cells and from induced plurip- versity of Kansas Medical Cen- antibodies against SARS-CoV-2 in blood
samples—an indication of prior infection.
otent stem (iPS) cells—mature “It’s like ter. By making TSCs from cells The tool could help untangle how the pan-
cells reprogrammed to an ES- gaining a toehold from women with infertility demic began. So far, the evidence is that the
like state. and watching them develop in virus originated in bats, animals Wang has
long argued are uniquely suited to harboring
But in 2017, Tohoku Univer- on Mars.” the lab, researchers could pin- viruses that pose a danger to humans. Now,
sity stem cell biologist Takahiro point how abnormal tropho- he hopes his assay can help trace the path of
the virus to humans and pinpoint when and
Arima and colleagues described Susan Fisher, blast cells prevent the embryo where it first spilled over.
a broth of nutrients and other University of California, from implanting in the uterus
compounds that could make or from developing normally The work is a natural next chapter for
TSCs from IVF embryos or San Francisco once implanted. Wang, who has been tracking viruses from
bats to humans for more than 2 decades.
first trimester placentas thrive in a dish. Or TSCs could help root out causes of Marion Koopmans, a virologist at Erasmus
Medical Center, credits him for essentially
“An enormous amount of work that was preeclampsia, in which a pregnant woman launching the field of bat immunology and
developing the tools to pursue it. “He has
never possible before became possible,” says suddenly develops high blood pressure that made a heroic effort to establish a very chal-

William Pastor, a stem cell biologist at McGill sometimes can be relieved only by inducing Science’s COVID-19 reporting is supported by the
Pulitzer Center and the Heising-Simons Foundation.
University. This year, Pastor’s group and two an early delivery. Preeclampsia is thought to

others showed this culture medium could stem from a defect of the placenta, perhaps

also coax certain types of ES cells to become in the way it invades the uterine wall and

self-renewing TSCs. interacts with the mother’s blood vessels,

To make TSCs that genetically match a pa- Pastor says. Researchers should now be able

tient, however, researchers want to be able to to make TSCs from umbilical cord blood or

start from mature skin or blood cells. In the from a baby’s blood or skin cells to observe

two new studies, teams led by stem cell bio- how placental precursor cells emerge and in-

logists Jose Polo at Monash University and teract with uterine cells.

Laurent David at the University of Nantes The new TSCs could also add realism

found ways to convert adult skin cells into to synthetic embryo models—stem cell–

“induced” TSCs. Both teams had been study- derived structures that mimic early hu-

ing how gene expression changes as mature man development in a lab dish. So far, they

cells are reprogrammed into iPS cells. They haven’t included trophoblast or other such

noticed that along the way, some expressed “extra-embryonic” cells, says Jianping Fu, a

genetic signatures of so-called trophecto- bioengineer developing such models at the

derm cells, which give rise to the trophoblast. University of Michigan, Ann Arbor. But sig-

“That was very weird,” Polo says, because a nals from these cells are critical to normal

cell’s decision to become trophectoderm hap- embryo growth, he says. Adding them would

pens so early in development—not anywhere take the models “to the next level.”

along the expected path backward from skin Better approximations of real embryos

cell to iPS cell. But by culturing the cells in will raise ethical concerns. The U.S. Na-

the newly available medium, the researchers tional Institutes of Health has not released

managed to push them to become TSCs. formal guidelines, but Fu says the agency

In a 16 September Nature paper, Polo’s discouraged him from including tropho-

team reported that these induced TSCs blast tissue in a recent grant application.

could develop into two major types of tro- But he thinks such experiments should pro-

phoblast cells and, like the cells surround- ceed. “When you mix the cells together, al-

ing an embryo, secrete human chorionic lowing them to self-organize … they will do

gonadotropin, a hormone whose signals amazing things.” j

20 2 OCTOBER 2020 • VOL 370 ISSUE 6512 sciencemag.org SCIENCE

ILLUSTRATION: KATTY HUERTAS lenging research line, which needed to start lent singer and to see him and Shi Zhengli SARS-CoV-2 attaches to when it invades cells.
from scratch,” she says. do a duet is very special.” Researchers then add a solution containing
the fragment of the viral spike protein that
As a child growing up in Shanghai dur- Now, Wang hopes to home in on the can bind to ACE2. If the binding takes place,
ing the Cultural Revolution, Wang would origin of SARS-CoV-2—an effort that will an enzyme turns the solution blue and then
listen to Mao Zedong’s speeches through a likely require screening thousands of ani- yellow. But when a sample contains anti-
loudspeaker in kindergarten. “I was think- mals and humans for signs of a prior infec- bodies against SARS-CoV-2, they prevent
ing: ‘My God … how does his voice trans- tion. The gold standard for doing that is the binding, blocking the colorful reaction.
fer from Beijing to Shanghai?’” Electrical called a virus neutralization assay, which Wang’s assay works on a variety of species
engineering became his passion. But after combines human cells and live virus with almost as well as the gold standard, says
getting into the prestigious East China Nor- a blood sample to see whether the sample Peiris, who has been using it for several
mal University, Wang was dismayed when contains antibodies that keep the virus weeks in infected cats, dogs, and hamsters.
the faculty assigned him to study biology. “I from binding to the cells. But using live vi-
thought, ‘I don’t like plants, I don’t like ani- rus means working in a high-level biosafety “This is an extremely interesting ap-
mals,’” he says. Going to a renowned univer- lab—expensive and very slow work. An al- proach,” says Isabella Eckerle, a virologist
sity felt like going to heaven, he says, “but ternative called an enzyme-linked immuno- at the Centre for Emerging Viral Diseases
the wrong door of heaven, basically, because sorbent assay (ELISA) is much easier to han- at the University of Geneva. Eckerle and
I went to a biology department.” dle, but a distinct version must be developed colleagues validated the test for WHO and
for every animal species. “You need to have a published the result as a preprint in late
Secretly listening to Voice of America,
Wang eventually became so profi- “I am nowfascinatedwith bats September. “Especially for screen-
cient at English that he was chosen [but] I am still not an animal fan.” ing potential plasma donors or
for a scholarship to study abroad. He when looking for the animal reser-
did a Ph.D. in molecular biology at Linfa Wang, Duke-NUS Medical School voir it should be really useful.”
the University of California, Davis,
and later moved to Australia, where whole panel of ELISAs that are optimized for Wang hopes to use the test to
he studied infectious diseases in ani- different bat species, and raccoon dogs, and screen animals and people in South-
mals. His career took a turn when civet cats, and pangolins, and God knows east Asia to identify “intermediate
a new virus emerged in the leafy what,” says Malik Peiris of the University of hosts”—species that may have picked
Brisbane suburb of Hendra in 1994, Hong Kong. “It’s a never-ending business.” up the virus from bats and transmit-
killing 14 horses and a trainer. Wang ted it to people—and learn whether
managed to sequence the virus, later Wang’s new assay, published in July in it crossed over into humans before
named Hendra virus, and helped de- Nature Biotechnology and now produced by the fateful outbreak in Wuhan.
velop a vaccine for horses. The virus Genscript Biotech, replaces the human cells
turned out to be transmitted by bats. and live SARS-CoV-2 virus of the gold stan- The bigger question that drives his
A few years later Wang worked on dard assay with human and viral proteins, work is: why bats? Over the past de-
another novel virus, Nipah virus, also eliminating the need for a high-security cade he has started to piece together
from bats. Intrigued, Wang scoured lab. The sample is tested on a plate impreg- an evolutionary story as convoluted
the literature and found numerous nated with angiotensin-converting enzyme as his own path to bats. As the only
other viruses linked to bats. 2 (ACE2), the human receptor protein that flying mammals, bats expand huge
amounts of energy. This eventually
Then came severe acute respira- damages their DNA, and Wang con-
tory syndrome (SARS). After the tends that they have adapted, in part,
World Health Organization (WHO) by dampening immune responses to
declared the epidemic over in July DNA damage. RNA viruses like SARS-
2003, it put together a mission of CoV-2 can cause similar damage, so
eight scientists, including Wang, to the upshot is that bats tolerate low
investigate the origins of the virus levels of viruses in a kind of peace-
in China. Wang had a hunch bats ful coexistence. “That’s why they are
could be the source, but the rest of such a good reservoir,” Wang says.
the team was skeptical. At a meet-
ing in Beijing, Wang met the head of Koopmans is not yet convinced by
WIV, who suggested he collaborate Wang’s immune system argument—
with a scientist at her institute: Shi Zhengli, bat ecology may play a greater role,
who was then studying viruses in fish and she says. For instance, bats often
shrimp. “She was the only virologist who be- range over wide territories, potentially
lieved me and was willing to collaborate with picking up a greater variety of viruses than
me,” Wang says. other animals, and in many bat species mil-
lions of animals roost together, making it
The two have since co-authored dozens easier for viruses to spread. But she says
of papers, including one in Science in 2005 that thanks to Wang’s work, there’s no doubt
that pinpointed horseshoe bats as a reser- that bats are key viral reservoirs.
voir of SARS-like coronaviruses. They also It’s an ironic legacy for a student who
like to team up in karaoke bars to sing clas- studied biology despite disliking animals. “I
sic Chinese ballads, says Peter Daszak, a re- am now fascinated with bats,” he concedes.
searcher at the EcoHealth Alliance, a New But, perhaps appropriately given what he has
York City nonprofit, and a longtime collabo- learned about emerging infections, he says:
rator with Wang and Shi. “Linfa is an excel- “I am still not an animal fan in the sense of
keeping animals near me.” j

SCIENCE sciencemag.org 2 OCTOBER 2020 • VOL 370 ISSUE 6512 21

NEWS | IN DEPTH

COVID-19

A call for diagnostic tests to report viral load

Measure could help officials know who is most contagious

By Robert F. Service Clinical Infectious Diseases, members of the have high CT values, suggesting the PCR

College of American Pathologists urged cau- has identified genetic material from non-

E ver since the coronavirus pandemic tion in interpreting CT values. infectious viral debris. Current guidelines
began, battles have raged over testing: Nevertheless, Mina, Gaglia, and others from the Centers for Disease Control and
Which tests should be given, to whom, Prevention and World Health Organization,
and how often? Now, epidemiologists argue that knowing whether CT values are which call for patients to isolate themselves
and public health experts are opening high or low can be highly informative. “Even for 10 days after onset of symptoms, recog-
a new debate. They say testing centers with all the imperfections, knowing the viral nize they are not likely to be infectious after
load can be extremely powerful,” Mina says.

should report not just whether a person is Early studies showed that patients in the that period. But Mina and others say the

positive, but also a number known as the first days of infection have CT values below recent findings also suggest that a patient

cycle threshold (CT) value, which indicates 30, and often below 20, indicating a high who has undergone multiple tests with high

how much virus an infected person harbors. level of virus; as the body clears the corona- CT values is likely at the tail end of their

Advocates point to new research indicat- virus, CT values rise gradually. More recent infection and need not isolate themselves.

ing that CT values could help doctors flag studies have shown that a higher viral load He adds that contact tracers should triage

patients at high risk for serious disease. can profoundly impact a person’s conta- their efforts based on CT values. “If 100 files

Recent findings also suggest the numbers giousness and reflect the severity of disease. land on my desk [as a contact tracer], I will

could help officials determine who prioritize the highest viral loads

is infectious and should therefore first, because they are the most in-

be isolated and have their contacts fectious,” Mina says.

tracked down. CT value is an im- Broad access to CT values could

perfect measure, advocates con- also help epidemiologists track out-

cede. But whether to add it to test breaks, Mina says. If researchers

results “is one of the most pressing see many low CT values, they could

questions out there,” says Michael conclude an outbreak is expanding.

Mina, a physician and epidemio- But if nearly all CT values are high,

logist at Harvard University’s T.H. an outbreak is likely waning. “We

Chan School of Public Health. have to stop thinking of people as

Standard tests identify SARS- positive or negative, and ask how

CoV-2 infections by isolating and positive?” Mina says.

amplifying viral RNA using a pro- CT values could also help clini-

cedure known as the polymerase cians flag patients most at risk for

chain reaction (PCR), which relies severe disease and death. A re-

on multiple cycles of amplification port in June from researchers at

to produce a detectable amount of Weill Cornell Medicine found that

RNA. The CT value is the number among 678 hospitalized patients,

of cycles necessary to spot the vi- 35% of those with a CT value of

rus; PCR machines stop running at Positive coronavirus tests could reveal a person’s infectiousness, too. 25 or less died, compared with

that point. If a positive signal isn’t 17.6% with a CT value of 25 to

seen after 37 to 40 cycles, the test is negative. In a study published this week in Clini- 30 and 6.2% with a CT value above 30. In

But samples that turn out positive can start cal Infectious Diseases, researchers led by August, researchers in Brazil found that

out with vastly different amounts of virus, Bernard La Scola, an infectious diseases ex- among 875 patients, those with a CT value

for which the CT value provides an inverse pert at IHU-Méditerranée Infection, exam- of 25 or below were more likely to have se-

measure. A test that registers a positive result ined 3790 positive samples with known CT vere disease or die.

after 12 rounds, for a CT value of 12, starts values to see whether they harbored viable Gandhi agrees that having access to CT

out with more than 10 million times as much virus, indicating the patients were likely in- values could help clinicians identify people

viral genetic material as a sample with a CT fectious. La Scola and his colleagues found at high risk for developing symptoms. Nev- PHOTO: MARTIN BERNETTI/AFP/GETTY IMAGES

value of 35. that 70% of samples with CT values of 25 or ertheless, she and others note that a high

But the same sample can give different CT below could be cultured, compared with less viral load doesn’t necessarily lead to disease;

values on different testing machines, and dif- than 3% of the cases with CT values above 35. some 40% of people who contract SARS-

ferent swabs from the same person can give “It’s fair to say that having a higher viral load CoV-2 stay healthy even though they have a

different results. “The CT value isn’t an abso- is associated with being more infectious,” similar amount of virus to patients who fall

lute scale,” says Marta Gaglia, a virologist at says Monica Gandhi, an infectious diseases ill. “As a physician, having the CT value is not

Tufts University. That makes many clinicians specialist at the University of California, the only thing I will use” to diagnose and

wary, Mina says. “Clinicians are cautious by San Francisco. track patients, says Chanu Rhee, a hospital

nature,” Mina says. “They say, ‘If we can’t rely Conversely, people often test positive for epidemiologist at Brigham and Women’s

on it, it’s not reliable.’” In an August letter in weeks or even months after they recover but Hospital. “But I do still find it helpful.” j

22 2 OCTOBER 2020 • VOL 370 ISSUE 6512 sciencemag.org SCIENCE

Congratulations AAAS® and Science® are registered trademarks of the American Association for the Advancement of Science, USA. Eppendorf® and the Eppendorf Brand Design are registered trademarks of Eppendorf AG Germany.
to Christopher Zimmerman, Ph.D. All rights reserved, including graphics and images. Copyright © 2020 by Eppendorf AG. Photography: Sameer A. Khan.
Postdoctoral Fellow
Princeton Neuroscience Institute

Meet the Winner

Eppendorf & Science Prize for Neurobiology The annual US$25,000 Eppendorf & Science Prize for
Congratulations to Christopher Zimmerman on winning Neurobiology honors scientists, like Dr. Zimmerman,
the 2020 Eppendorf & Science Prize for his work on the for their outstanding contributions to neurobiological
neural circuits that govern thirst and drinking behavior. research. Christopher Zimmerman is the 19th
Dr. Zimmerman discovered that sensory signals originating recipient of this international award. Due to coronavirus
throughout the body come together within individual (COVID-19), the 2020 winner and finalists will be
neurons in the brain to produce the sense of thirst. honored at the Prize Ceremony in November 2021 in
He demonstrated that this new class of body-to-brain Chicago.
signals predicts changes in hydration before they occur
and, as a result, adjusts our level of thirst preemptively. You could be next to win this prize.
Dr. Zimmerman’s research has revealed fundamental If you are 35 years of age or younger and currently
principles of ingestive behavior and provided neural performing neurobiological research, you could be next
mechanisms to explain aspects of everyday human to win the 2021 Prize.
experience.
Deadline for entries is June 15, 2021.

Learn more at: www.eppendorf.com/prize

NEWS

FEATURES

OFFICIAL INACTION ILLUSTRATION: STEPHAN SCHMITZ

A Science investigation shows that FDA oversight of clinical trials is lax,
slow moving, and secretive—and that enforcement is declining

By Charles Piller

24 2 OCTOBER 2020 • VOL 370 ISSUE 6512 sciencemag.org SCIENCE

NEWS

F or nearly a decade, the Food and and Harris failed to properly report abnor- documents obtained by Science through the
Drug Administration (FDA) cited mal lab test results. He also did not disclose Freedom of Information Act (FOIA). Inspec-
osteopath Michael Harris for egre- that trial participants were taking opioid, tors told Harris he could be subject to fines,
gious errors in clinical trials he was antidepressant, or antipsychotic drugs— permanent disqualification from clinical re-
overseeing. Agency inspectors found which could have skewed results or posed search in the United States, and legal pros-
a litany of problems at Harris’s pri- safety concerns. The agency said Aspen’s re- ecution. Repeat problems and a raft of new
vate firm, Aspen Clinical Research cords were disorganized, contradictory, and ones emerged during inspections in 2014,
in Orem, Utah, which had contracts sometimes backdated in a way that “begs the 2015, and 2019. Each time, in responses to
to evaluate many drugs, including question of the authenticity and veracity of FDA, Harris admitted some transgressions,
ones aiming to treat postoperative pain, pe- data collected.” strenuously disputed others, and promised
diatric schizophrenia, and migraines. FDA to improve.
found there were serious lapses in obtain- Those “serious, ongoing deviations”
ing informed consent from trial volunteers, might constitute “fraud, scientific mis- Through all that, FDA never formally
unqualified staff made medical assessments, conduct,” and “significant human subject sanctioned Harris or pursued other penal-
protection violations,” according to FDA ties. The agency never made public the al-
leged offenses or told trial participants they
might have been put at risk. Nor did it tell
companies sponsoring some of the trials
that their data might have been compro-
mised. (The documents Science obtained
were heavily redacted, making it impossible
to know which trials were in doubt and,
thus, which volunteers might have been
harmed or endangered.) Meanwhile, phar-
maceutical and medical device companies
continued to contract with Aspen. Since
2011, they have paid the firm millions of dol-
lars for work on at least 65 trials, and Aspen
is now recruiting people for nine new trials
on Alzheimer’s disease, autism, depression,
and other serious disorders.

Harris declined to comment on the FDA
reports and the agency refused to discuss
its dealings with him. Wayne Croft, Harris’s
business partner at Aspen, also would not
reply to questions, but he describes FDA’s
treatment of Harris as bureaucratic harass-
ment that drove away business. “Anytime he
hears about the FDA,” Croft says, Harris ex-
periences “post-traumatic stress.”

Scrutiny of the agency’s clinical trial
oversight for the past 11 years suggests its
interactions with Harris and Aspen are
commonplace. FDA’s enforcement of clinical
research regulations is often light-handed,
slow-moving, and secretive, Science has
found in an investigation that included a
review of some 1600 agency inspection and
enforcement documents for other trials that
FDA said violated rules and law. Almost all
were acquired via FOIA requests, including
many made by FDAzilla, a commercial ser-
vice that tracks the agency.

The period examined covered former Pres-
ident Barack Obama’s administration and
the first 3 years of President Donald Trump’s
term. Clear corrections of inspector-reported
dangerous or unlawful clinical trial prac-
tices were the exception, even amid signs
that trial participants were harmed and that
data underpinning evidence-based medicine
were corrupted. On the rare occasions when
FDA formally warned researchers of findings
that they had broken the law, the agency of-
ten neglected to ensure that fixes occurred,

SCIENCE sciencemag.org 2 OCTOBER 2020 • VOL 370 ISSUE 6512 25

NEWS | FEATURES

Science found. Moreover, the agency fre- oversee trials locally. Over the period Science Even voluntary action designations declined

quently closed cases on the basis of unveri- examined, FDA conducted roughly 6700 in- sharply under Trump, whereas “no action”

fied claims by those accused. spections of clinical researchers or IRBs. inspections spiked. The agency disqualified

The apparent neglect appears to be wors- FDA inspections can result in a range of an average of three investigators per year un-

ening. The agency issued 99 “warning let- responses, from “no action” declarations; der Obama but just two total during Trump’s

ters” for serious clinical trial transgressions to “official action indicated” (OAI) reports, first 3 years.

during Obama’s first 3 years in office, 36 in which require a violator to clean up serious FDA did not dispute Science’s figures on

his last 3 years, and just 12 during the first transgressions; to warning letters, which OAIs and warning letters—but the agency

3 years under Trump. “It certainly looks like threaten further regulatory action unless said Trump-era policies were not a factor.

FDA is enforcing clinical trial requirements corrections are made promptly. In extreme “The number of warning letters can ebb and

much less frequently, which is troubling for cases, the agency can even disqualify a scien- flow,” the agency wrote in an email.

protecting subjects’ welfare and ensuring tist from clinical research. For extreme problems in a clinical trial,

the validity of data for our medical prod- But Science found that FDA rarely levels when patients or data reliability are at imme-

ucts,” says Patricia Zettler, a lawyer who sanctions. The agency almost always rules diate and serious risk, FDA can use warning

reviewed Science’s findings. She worked for that no action is warranted or requests vol- letters to immediately halt the trial or restrict

FDA from 2009 to 2012, rising to associate untary corrections, according to documen- an IRB from approving new trials. The agency

chief counsel. tation on all inspections that drew an OAI did so seven times during Obama’s first term

Jill Fisher, a social scientist at the Uni- designation or more severe enforcement ac- and not once since Trump took office.

versity of North Carolina, Chapel Hill, who tion. Since 2009, FDA has used the OAI des-

studies clinical trials, is also troubled that ignation to direct only 291 of the inspected LITTLE FOLLOW-UP

when an FDA inspection exposes apparent researchers or institutions—about 4%—to Among the 291 OAI cases Science identi-

mistakes, recklessness, or fraud in a trial, correct serious, illegal, and potentially dan- fied, only 71 resulted in a clear regulatory

endpoint—such as disqualification,

a “closeout letter” certifying that

A watchdog loses its bite corrections were completed, or an-
other formal statement. For the
Despite a steady or rising enforcement budget, the Food and Drug Administration has wielded its power to enforce laws meant remaining 220, no clear outcomes
to safeguard volunteers for clinical trials—and ensure the integrity of testing data—less and less frequently since the beginning could be found in public documents
of the Obama administration. Under President Donald Trump, the number of enforcement actions has plummeted.

Investigator Warning Cases classifed Clinical trial enforcement or data banks, leaving trial partici-
disqualifcations letters “ofcial action indicated” budget ($ millions)* pants and others in the dark. FDA
can also deem violations “not cor-
Obama rectable,” ending its enforcement
First 3 years with neither compliance nor any

Obama form of disclosure. From outside the
Last 3 years agency, it’s impossible to know how

Trump often that occurs.
First 3 years In addition to searching public

0 10 20 0 40 80 0 40 80 120 160 0 50 100 150 documents and FDA databases and

*Combined *scal year budgets for the periods shown, constant 2019 dollars filing FOIA requests for key docu-

ments in nearly all 291 cases, Science

neither the agency nor the scientists run- gerous clinical trial problems. Furthermore, filed additional FOIA requests in 28 of those

ning it are obliged to notify participants. FDA officials sometimes downgraded a cases for every essential document, including

She calls that failure to inform “a travesty.” problem originally classified by inspectors any that described a resolution. After about

At the least, Fisher says, volunteers in a as OAI—in which compliance must be con- 10 months, FDA released documents describ-

trial should be told when the integrity of firmed by a follow-up inspection or other ing how 16 cases were resolved. Most were

a research center running the study or its actions—to “voluntary action.” based on written promises from researchers.

investigator is in question. Holly Fernandez Lynch, a lawyer and bio- In one, the researcher pledged to never again CREDITS: (GRAPHIC) N. DESAI/SCIENCE; (DATA) FOOD AND DRUG ADMINISTRATION

ethicist at the University of Pennsylvania conduct clinical studies and then violated

LIGHT ENFORCEMENT (UPenn), says the agency prefers voluntary that pledge, according to records posted by

Although best known for overseeing food action. “When people fail to comply, it’s hard drug companies that hired him for clinical

safety and evaluating experimental drugs, to know what to do about it, short of com- trials—with no apparent FDA response.

vaccines, and medical devices for marketing pletely ending a trial or all research at that In an email, FDA defended its approach,

approval, FDA performs many other tasks institution,” which might deprive patients of noting its “strategy to focus inspectional re-

with its $5.7 billion annual budget (for fis- a chance at an effective treatment in trials, sources on higher-risk facilities.” FDA tries

cal year 2019). Among them is overseeing Lynch notes. to address clinical trial problems early in

most clinical research in the United States Under Trump, the number of FDA inspec- the process, it also noted, so “an issue that

and some conducted elsewhere for the U.S. tions per year has increased significantly, might warrant a warning letter could be

market—including many high-stakes vac- yet enforcement actions have nosedived resolved before the problem rises to that

cine and drug trials for COVID-19. The (see graphic, above). About 6% of FDA in- level.” (Agency officials would not agree to

agency now deploys 102 inspectors to re- spections were classified as OAI during an interview. Spokespeople instead emailed

spond to whistleblower complaints of dan- the Obama administration, although the selective responses to written questions.)

gerous operations, conduct routine visits to proportion began to decline in his second Yet FDA sometimes issues closeout letters

trial sites, and review records of those sites term. During Trump’s first 3 years, however, even when nothing material has changed. In

or the institutional review boards (IRBs) that OAI reports fell to less than 1% of the total. 2016, the agency sent a warning letter to the

26 2 OCTOBER 2020 • VOL 370 ISSUE 6512 sciencemag.org SCIENCE

Disgraced researchers can still reap drug industry payouts

ILLUSTRATION: STEPHAN SCHMITZ W hen the Food and Drug Adminis- he pleaded guilty to two federal crimes: investigators. The 22 that responded all
tration (FDA) wields its ultimate illegal possession of anabolic steroids said the two remain qualified physicians
penalty, disqualification, against and (in his role as sole owner of a private who were not hired for clinical trials. None
clinical researchers who it de- practice) “knowingly and willfully” making would say whether they knew about the
termines have violated the law, a “false, fictitious, or fraudulent statement FDA sanctions, or Farber’s criminal convic-
falsified data, or committed grave errors or or representation” to the sponsor of a tion, when they made the payments.
misconduct, they can no longer run human clinical trial. Federal prosecutors found that
trials in the United States. But that doesn’t during the clinical trial, for an experimental Among the 42 other physicians FDA
always sever their financially rewarding cream to treat Sun-damaged skin, Farber disqualified since 2005, data from 2013
relationships with big pharma. delegated patient exams that he claimed to to 2019 show five others took in about
have conducted himself to an assistant. He $10,000 to $36,000 in drug company
In 2008, FDA filed a public notice that was sentenced to probation and medical payments after their bans. One was Miami
it had disqualified Texas urologist James supervision of his practice and paid about internist Farid Marquez, barred from re-
Vestal after its inspectors discovered $220,000 in fines and restitution, mostly search in 2015 after FDA found what it said
egregious problems in clinical trials he had to the sponsor of the skin cream trial. And were phony documents in his work for drug
led of an experimental hormonal treatment he was barred from future earnings for giants Boehringer Ingelheim and Eli Lilly.
for advanced prostate cancer. The agency conducting trials. But those companies seem to have no hard
said Vestal admitted to fabricating medical feelings. Although Marquez no longer gets
exams, faking signatures, enrolling ineligible But Farber has since recouped those their research money, Boehringer Ingel-
patients, and other actions that “exposed losses in other ways, thanks to the pharma- heim has treated him to at least
[his] subjects to unnecessary risks.” ceutical industry. From 2013 to 2019 alone, 168 meals since his disqualification; and
Open Payments data show, he took in Lilly, 17. Marquez declined to comment.
Yet a Science examination of corporate about $665,000 from 45 medical research
compensation disclosures from the federal companies—including Actavis, Allergan, “The most favorable sort of interpreta-
Open Payments database showed that from Bayer, and Genentech—for consulting tion is that the companies’ internal vetting
2013 to 2019, 27 drugmakers—including and teaching. Pfizer paid him the most, process is deeply broken,” says Vinay
heavyweights Bayer, the Johnson & Johnson $90,000. He received $169,000 in other Prasad, a hematologist-oncologist at the
subsidiary Janssen, and Sanofi—paid Vestal fees, food, and travel perks. Asked whether University of California, San Francisco,
about $422,000, including $340,000 for he informed the companies about his FDA who studies how drug industry funding
consulting and teaching. (The system only disqualification and legal history, Farber influences research and medical practice.
began to record pharmaceutical compensa- declined to comment. “The more pessimistic interpretation is
tion to physicians in 2013.) Vestal did not that they turned a blind eye because the
respond to requests for comment. Science asked 33 drug companies why investigators were not disqualified for
they paid Vestal or Farber to teach or something that hurts the companies’ bot-
In 2005, FDA disqualified Philadelphia consult after FDA expelled them as clinical tom lines.” —C.P.
dermatologist Harold Farber just before

SCIENCE sciencemag.org 2 OCTOBER 2020 • VOL 370 ISSUE 6512 27

NEWS | FEATURES

medical director of a Texas research firm. who studies how drug funding affects re- deaths to Hennepin’s IRB—a required prac-
He had enrolled trial participants who suf-
fered from kidney impairments, an agency search. Steven Joffe, a medical ethicist at tice to verify that the trial itself does not pose
inspection reported, and then overdosed
them with an experimental drug that could UPenn, agrees, citing an obligation to be safety hazards. In a response to FDA, Driver
have further damaged their kidneys. The
physician responded with a plan to prevent transparent to the people who have volun- acknowledged his reporting error in three
future overdoses, but FDA replied that it was
inadequate. Yet on the same day as that re- teered their bodies for clinical research: “We of the deaths but did not address the other
ply, the agency sent a letter closing the case.
owe that back to them.” 10 deaths, and FDA documented no further
“The agency should follow through and
complete actions—so they don’t initiate a information on them. In an email to Science,
process and never resolve it,” says Joshua
Sharfstein, vice dean at the Johns Hopkins “SYSTEMIC PROBLEM” IGNORED Driver wrote that FDA had never discussed
Bloomberg School of Public Health and a
top FDA official under Obama. “And it’s im- A clinical trial scandal that erupted in Min- those additional deaths with him, but he ac-
portant for the agency to share information
about its processes, so people can under- nesota in 2018 illustrates the potential public knowledged failing to report the deaths of the
stand what happened.”
health hazards posed by FDA’s lax enforce- 13 FDA initially flagged, as well as 10 others
When FDA does follow up on the prob-
lems it identifies, the process can be slow. ment. That year, the consumer advocacy in the trial that it did not document. Driver
Agency documents and databases covering
the 291 OAI cases show dozens of cases that group Public Citizen and more than 60 cli- attributed all 23 deaths to acute illnesses or
took 10 months to 14 years to be escalated
to a warning letter. Some delays can be nicians and medical ethicists alleged that injuries unrelated to the study drugs, includ-
“egregious,” Lynch says.
a large Minneapolis community hospital ing sepsis, stroke, and heart attack.
In 2015, the agency issued an OAI after it
inspected a Pennsylvania hospital IRB that operated by Hennepin Healthcare had vio- Driver became a member of the hospital’s
oversees more than 100 trials annually. The
agency’s findings read like a functional col- lated informed consent requirements in tri- IRB in 2016 and was on it when the group
lapse: records glitches, including failure to
keep a list of IRB members and their quali- als comparing ketamine, a powerful sedative approved or reviewed the later ketamine tri-
fications or record attendance at meetings;
approval of trials without a quorum; sloppy and anesthetic, with other drugs for people als flagged by Public Citizen. Science showed
reviews of trials; a lack of required written
policies on how to determine the risk of de- who came to the emergency room with agi- the FDA documentation of his earlier ket-
vices being tested on patients; neglecting to
promptly inform FDA or the hospital’s own tation. The critics said the emergency de- amine trial to Michael Carome, who directs
leaders about trial problems; and poor or
no training for IRB members. According to partment doctors did not obtain informed health research at Public Citizen. “We have
FDA documents, the agency waited 4 years
to follow up. In a second inspection, in 2019, consent from trial partici- a group of physicians, all in
many of the same problems were found, re-
sulting in a warning letter. Only then did the “It certainly lookspants, even though ketamine the emergency department,
IRB finally resolve the deficiencies. over a period of several years
was much more likely to
A St. Louis IRB received citations about
similar problems in a 2012 warning letter, cause serious breathing prob- like FDA is enforcing … not complying with FDA
which noted that pervasive lapses at the IRB lems and movement disor- clinical trial regulations regarding human
were also found in 2003 and in 1998. That ders. Many patients who subjects protection,” says Car-
IRB finally addressed the issues after the received ketamine developed requirements much ome, a physician himself. The
warning, according to an FDA closeout letter. such symptoms. clinical trial missteps exposed

That sparse, slow enforcement puts par- After Public Citizen’s alle- less frequently, “a pattern of sloppiness and
ticipants and the research process itself at gations, FDA inspectors con- carelessness” by Hennepin
risk, critics say. Given the extremity of vio-
lations that lead to little or no punishment, which is troubling.”firmed the lapses in informed researchers and their IRB, he
Fisher asks: “What would it take for the FDA adds, raising broad doubts
to take serious action to ensure that clinical consent. They concluded that
research is being conducted with the most
integrity possible?” Hennepin’s IRB had waived Patricia Zettler, about its oversight of human

The secrecy of the process also means trial the requirement to obtain former FDA lawyer subjects research.
participants are kept unaware of possible
harm, violating “a basic human right,” says consent and did not institute The IRB chair, a hospital
Vinay Prasad, a hematologist-oncologist at
the University of California, San Francisco, additional safeguards “to protect the rights spokesperson, Miner, and Driver all declined

and welfare of those subjects.” And inspec- to comment about the later clinical trial

tors said researchers running the trials ne- problems FDA identified. In a public state-

glected to properly report serious adverse ment, Hennepin defended the trials while

events. Nevertheless, the agency did not clas- apologizing for “not involving the commu-

sify the inspection as OAI or send a warning nity in a proactive way” before the studies.

letter. Hennepin rejected most of the FDA Hennepin said it would increase training for

findings, although the health system agreed clinical researchers and improve study mon-

to some procedural changes and suspended itoring and public disclosure. But the health

one trial for review. system refused to answer questions such as

FDA already knew of alleged problems in why—after years of FDA inspections cited

Hennepin’s ketamine research at the same serious problems with ketamine trials—the

hospital before the Public Citizen allega- IRB and emergency department launched

tions, according to new documents obtained new trials with no special precautions to

by Science. In 2014, the agency inspected a protect patients.

Hennepin trial led by Brian Driver, a staff Hennepin’s ketamine controversy also

physician, and James Miner, then and now raises questions about FDA oversight. De-

chief of emergency medicine at the hospital spite the problems it identified as early

where the later trials took place. The trial as 2014, FDA never issued a warning let-

compared ketamine with another drug used ter or otherwise demanded that Hennepin

to sedate patients who required breathing researchers rectify trial procedures. The

tubes. As in the later trials, the agency de- agency imposed no restrictions on the hos-

tailed failures to properly obtain informed pital’s clinical research or IRB operations.

consent from patients. FDA never mandated that trial participants

A 2016 follow-up review by FDA of the who suffered adverse effects—or their loved

ketamine trial detailed failures by Driver ones—be advised about inspection findings.

and Miner to properly report 13 patient FDA declined to comment on Hennepin’s

28 2 OCTOBER 2020 • VOL 370 ISSUE 6512 sciencemag.org SCIENCE

clinical trials—or on any specific cases de- 2018 to disqualify him. disqualification stemmed from FDA’s review

scribed in the Science investigation. Baker’s abuses were not mere technicali- of just a few of those trials. Not long before

“FDA is just dropping the ball here,” says ties. FDA said they “placed unnecessary risks his disqualification, the agency asked Baker

Carome, a former research regulation execu- to human subjects and jeopardized the in- whether any data or patient records had also

tive at the Department of Health and Hu- tegrity of data.” Inspectors found informed been doctored in his other completed trials.

man Services. “I think the institution should consent breakdowns involving children and Baker said he lacked the capacity to review

have gotten a very strong warning letter for adults—some of whom were injected with those studies. As far as Science could ascer-

the [latter] ketamine trials, and that letter experimental allergy or immunology drugs tain, FDA did not investigate the other trials.

perhaps should have noted the prior history or biologics despite taking other drugs, such Baker declined to say whether he had in-

with Dr. Driver, to signal a systemic problem.” as cough medicine and antibiotics, that formed participants in his many trials that he

By not coming down hard on IRBs and could interact with them. He failed to prop- had been disqualified. Science also contacted

the scientists running clinical trials, FDA erly document or follow up on a child’s emer- multiple companies that sponsored Baker’s

lets those parties avoid possible negligence gency department visit for severe abdominal trials during the period for which FDA found

lawsuits or other legal liability if patients are pain. People were enrolled against eligibility falsified information. Asked whether they

harmed, Prasad says. He adds that the lack rules—with some participating in more than told trial volunteers about the government’s

of enforcement also lets the drug companies one trial, for example, or having conditions findings, none responded.

sponsoring the trials off the hook. “FDA is that should have disqualified them. FDA said Baker, who declined to answer most ques-

a regulatory agency charged with protecting Baker kept substandard case histories, over- tions submitted by Science, contended in an

the public’s best interests. But at times email that his patients suffered no ad-

it behaves like an attorney working on verse events. But he also added: “I agree

behalf of the companies.” with [the] way the FDA controls the re-

search world and regret running afoul

SLOW-MOTION DISQUALIFICATIONS with those studies.”

For the most extreme violations of rules Had FDA acted earlier, Baker’s clini-

governing human subjects research, cal trial venture—the Baker Allergy,

FDA can brandish the threat of a pub- Asthma and Dermatology Research

licly announced disqualification—effec- Center—might have lost a lot of busi-

tively the death penalty for a clinical ness. The French drug giant Sanofi

research career. But how the agency and other companies paid him at least

has used that sanction captures many $1.2 million for trials conducted be-

flaws of FDA’s clinical trial oversight. tween the time FDA detected his fal-

Since the beginning of Obama’s pres- sified exams in 2016 through 2019,

idency, the agency has disqualified just according to the federal Open Pay-

24 clinical trial investigators, mostly for ments system, which tracks pharma-

deliberate and repeated falsification of ceutical industry payments to doctors.

data or other information such as pa- Yet the work kept flowing from

tient referrals for a trial. And although Sanofi to his research business even

PHOTO: CLOUDSHINE MEDIA, “BAKER ALLERGY, ASTHMA & DERMATOLOGY (COVID-19 SAFETY),” 2 APRIL 2020/YOUTUBE.COM disqualification bars a scientist from re- after he was disqualified. Sanofi as-

ceiving any experimental drug or device signed responsibility for Baker’s trials

or conducting any trial involving FDA- James Baker is one of just two scientists disqualified from U.S. to another doctor at his center, which

regulated drugs or other substances, clinical research by FDA during President Donald Trump’s first 3 years. Baker operated with his dermato-

it can sting less than it sounds. Drug logist wife in the same building as

companies often continue to lavish money dosed patients on experimental medicines, their medical practice.

and prestige on disgraced investigators (see and repeatedly conducted trials without IRB Likewise, the disqualification leaves Baker

sidebar, p. 27). approval. Baker conceded to agency inspec- free to operate his other company, Xtract

Investigators under a cloud also can con- tors that one of his employees, who had no Solutions, a suburban Portland, Oregon, pur-

tinue to run lucrative trials while FDA con- relevant medical credentials, prescribed veyor of allergy and immunology technology

siders whether to disqualify them, a process drugs to trial volunteers. She worked on to hospitals and physicians. In 2014, FDA is-

that takes, on average, more than 3 years. 30 of his trials. sued a warning letter to Baker and Xtract for

Scientists or IRBs facing FDA accusations, FDA finally disqualified Baker after he allegedly selling “adulterated devices” and

including those that ultimately lead to dis- was caught backdating many clinical trial failing to establish and maintain quality-

qualification, have collectively conducted or reports beginning in 2013. The records indi- control procedures for its syringe-filling

overseen thousands of trials during the past cated he performed patient exams in Oregon systems—meant to speed up vaccinations

decade, Science found in a review of public when he was actually in Las Vegas, Hawaii, during pandemics.

records. Take Oregon allergist-immunologist and elsewhere attending medical meetings Xtract eventually dropped its injection

James Baker, one of only two disqualifica- and, in one instance, a wedding. (FDA in- device. But the company still sells software

tions by Trump’s FDA. spectors had earlier caught him similarly to guide doctors and hospitals on a subject

Between 2009 and 2017, FDA inspected falsifying records in 2015, when he admitted Baker has ample experience, if not always

Baker-led trials on four occasions and each he had “gone fishing” on days he purportedly success, with: FDA regulatory compliance. j

time found violations. Time after time, saw trial patients.)

Baker promised to do better, according to FDA documents show Baker conducted With reporting by Meagan Weiland. This story was

agency documents. Again and again, he at least 115 clinical trials from 2005 until supported by the Science Fund for Investigative

broke his promises, FDA documents con- 2017, often testing drugs for which compa- Journalism. For more on how the story was reported,

clude. The agency, however, waited until nies later sought agency approval. Yet his see http://scim.ag/FDA_Methods.

SCIENCE sciencemag.org 2 OCTOBER 2020 • VOL 370 ISSUE 6512 29

INSIGHTS Hone your skills ILLUSTRATION: ROBERT NEUBECKER

LETTERS How can you turn the time created by
this pandemic into a blessing? Faced with
NEXTGEN VOICES: ASK A PEER MENTOR similar challenges, I have tried to upgrade
my skills to work remotely, manage effec-
Funding fix: Spend time tively, network extensively, and expand
my knowledge through webinars, online
As coronavirus disease 2019 (COVID-19) restrictions isolate conferences, and seminars. I suggest using
scientists in their homes, close labs, and cancel conferences, this time to increase your knowledge by
some researchers are overwhelmed by professional and domestic reading extensively in your research area.
responsibilities, whereas others, like the advice-seeker below, have Write reviews, book chapters, or a short
a surfeit of time but a dearth of funding. Here, peer mentors ofer piece analyzing old data. Find collabora-
advice to this researcher. Do you need advice for the COVID-19 tors who can give you access to a lab that is
era? See the box on page 31 to submit your own questions. operational. Apply for funding that is open
Follow NextGen Voices on Twitter with hashtag #NextGenSci. to young researchers or that requires col-
Read previous NextGen Voices survey results at https://science. laboration with national or international
sciencemag.org/collection/nextgen-voices. —Jennifer Sills partners. Attend interesting webinars, and
never miss an opportunity to present your
Dear NextGen VOICES peer mentors, work to the research community.
I am a young researcher with research proposals but no funding. My grant
applications are still pending. My contract requires me to publish, but I can’t Charu Lata
conduct experiments or produce reliable results without funding. With restric- Council of Scientific and Industrial Research–
tions in place for COVID-19, I have limited access to my lab. How can I use National Institute of Science Communication and
this time most effectively to ensure that I can stay in academia long-term? Information Resources, New Delhi, Delhi, 110067,
Sincerely, Funding Fix India. Email: [email protected]

30 2 OCTOBER 2020 • VOL 370 ISSUE 6512 Have you considered using publicly
available data as a source for research?
I suggest that you stay in your area of
research and ask some new questions.
In this era of genomics, the wealth of
publicly available databases allows you to
conduct research at home. In the medical
field, medical records and big data discov-
ery are considered an essential part of the
health infrastructure and represent a valu-
able resource for translational research.

Ruty Mehrian-Shai
Pediatric Hemato-oncology, Sheba Medical
Center, Ramat Gan, 52621, Israel.
Email: [email protected]

Are you sure that spending all your time
doing experiments in the lab is the best
way to obtain academic achievements? In
my experience, analyzing experimental
results, organizing figures, and rethinking
scientific ideas at regular intervals are more
important to improving quality of research.
Consider using this time to better under-
stand your current results. Prepare research
figures that help you tell the scientific story
of your findings. Review the literature
with an eye toward adjusting the aims and
methods of your ongoing research. Practice
communicating your work through oral and
poster presentations in virtual meetings.
These activities will prepare you to use your
time more productively once you return to
the lab.

Bo Cao
Core Research Laboratory, The Second Affiliated
Hospital, School of Medicine, Xi’an Jiaotong
University, Xi’an, Shaanxi, 710004, China.
Email: [email protected]

sciencemag.org SCIENCE

Connect and collaborate avoid doing tangential experiments that Would your papers, science, and current
are not necessary to support your pri- datasets benefit from a shift in paradigm?
Have you reached out to your depart- mary hypothesis. This introspection will Early in my career, I found that working
mental peers and administrators for make your time in the lab more efficient with oral historians exposed me to a new
direction? As a faculty member in a non– when you return. way of seeing the archaeological landscape.
tenure-track position, I have found that It made my science more accessible to
it’s helpful to form a peer group to brain- Felix Man-Him Cheung a general audience, brought community
storm ideas. It might also be useful to ask School of Biomedical Sciences, University of support for my work, and even helped
your school’s administration about their Hong Kong, Hong Kong, China. pass legislation protecting archaeological
expectations for the current pandemic Email: [email protected] remains in the countries where I recorded
situation. Perhaps they will be flexible stories. I encourage you to keep an open
about your contractual obligations. How can you use this time to protect your mind and be flexible. Taking a step away
Naga Rama Kothapalli mental health? We always complain about from the lab could create an opportunity
not having enough time to read and learn to examine your experimental paradigm,
Department of Chemistry and Biochemistry, everything we need. Now that many of us expand your literature searches to include
University of Oklahoma, Norman, OK 73019, USA. have plenty of time, it is important not to fields outside your own, and seek differ-
Email: [email protected] waste this gift. Remember to save some time ent perspectives by opening dialogue with
for leisure. Read a romance novel, play a members of another field.
Have you thought about how much time game, or listen to music. When you go back
you have been devoting to conducting to the lab, you need not only the knowledge Felicia Beardsley
lab work, writing manuscripts, securing to complete your experiments but also the Department of Anthropology, University
grants, managing students, and reading? mental health to continue your work. of La Verne, La Verne, CA 91750, USA.
Many young scientists are pressured to Email: [email protected]
focus on publications and funding to the Wagner Eduardo Richter
exclusion of other important aspects of a Department of Chemical Engineering, Have you thought about taking a break
career in academia. This time presents an Technological Federal University–Paraná, from research? Moving away from aca-
opportunity to find a balance. Consider Paraná, Brazil.Email: [email protected] demia for a short period helped me acquire
helping the graduate students and post- new skills and gain momentum later in
doctoral students in your lab. Spending Explore a new field my career. I took a break during the third
some time bonding with these researchers, year of my Ph.D. program to do an intern-
professionally and personally, will help Have you considered reading current ship in industry, where I started to look at
you support them. Help them contribute literature outside of your primary area of problems with a new perspective. Later, I
by asking them to think in innovative expertise? Waiting for funding outcomes incorporated some of that experience into
ways. Fostering new networks will help necessary to commence data collection is my thesis, my postdoc work, and even my
you excel in the post-COVID world. stressful, but it does provide opportuni- tenure track position. I suggest that you
Garima Singh ties to spend time in other ways. Reading take a break and look for “virtual” intern-
widely enabled me to bring novel per- ship opportunities. You may acquire new
Fleming Fellowships (Antimicrobial Resistance), spectives to my research areas, which led skills that you can apply to your research
South Asia, Delhi, India. to my highest-funded and most reward- later. You might even realize that there are
Email: [email protected] ing projects. Although reading broadly many other options for you to apply your
will not result in immediate publications, knowledge and skills, which will allow you
Step back and rethink it can help differentiate you from your to make better career decisions.
peers, which will ultimately help you stay
How can you use this time in non-tradi- in academia long-term. Xiao-Yu Wu
tional ways? Science teaches us to adapt Department of Mechanical and Mechatronics
when new information arrives or situations Samuel Nathan Kirshner Engineering, University of Waterloo, Waterloo, ON
change. I have found that using my creativ- School of Information Systems and Technology N2L 3G1, Canada.
ity has helped inspire my best science and Management, University of New South Wales, Email: [email protected]
service. I suggest that you find ways to be Sydney, NSW 2052, Australia.
innovative by making art, learning about a Email: [email protected] 10.1126/science.abe6015
new topic or perspective, volunteering for
peer review, doing outreach, educating oth- NEXTGEN VOICES: CALL FOR QUESTIONS
ers, or taking part in citizen science.
Daniel Ari Friedman Need advice? Ask your peers!

Department of Entomology, University of Do your COVID-19 experiences differ from this young scientist, who faces too much
California, Davis, CA 95616, USA and Remotor time and not enough funding? Are you affected by increased family duties, cancelled
Consulting Group, Davis, CA 95616, USA. travel, health uncertainties, or limited career options? Are you facing unique chal-
Keybase: @docxology lenges that others have overlooked? Science would like to support you by asking
readers to provide peer mentoring advice. Do you have a question that you would like
What is the one experiment that you are your peers to address? Send it to Science at the link below!
most excited to complete? I have noticed
from my work-from-home experience To submit, go to www.sciencemag.org/nextgen-voices-covid-19-questions
that it is good to stop and rethink the Please submit by 30 October. If your question is selected, Science will post it anonymously
research from time to time. By prioritiz- and ask young scientists to respond with advice to be published in a later issue.
ing different experiments, goals become
clearer. You might realize that you can 2 OCTOBER 2020 • VOL 370 ISSUE 6512 31

SCIENCE sciencemag.org

PERSPECTIVES

1 Scrapie is a
highly infectious
disease of sheep

that is caused
by neurotoxic prion
protein aggregation.

LANDMARK: NEURODEGENERATION ist without its respective gene, discovered
in hamsters the gene encoding the cellu-
Shifts and drifts in prion science lar prion protein (PrPC), whose misfold-
ing yields tightly packed aggregates called
Important questions remain unanswered since prions PrPSc. It is generally believed that prion
were discovered four decades ago replication occurs when coalesced PrPSc is
broken down into smaller species. Those
By Adriano Aguzzi and Elena De Cecco bly limited. Prusiner postulated that prions species then accrue further PrPSc, in a pro- PHOTO: WAYNE HUTCHINSON/MINDEN PICTURES
carry on their replicative cycle without the cess akin to the growth of crystals, and
P aradigm shifts are drivers of scientific participation of nucleic acids. This hypoth- eventually break again, perpetuating their
progress, yet the shifters of the para- esis, reminiscent of John Griffith’s 1967 sug- replicative cycle. Infectious prion seeds
digms often experience scorn rather gestion of the existence of self-replicating then move to neighboring cells and wreak
than immediate applause. That was proteins (2), had the potential to explain the havoc in the central nervous system by in-
the fate of Stanley Prusiner’s 1982 prodigious resistance of the scrapie agent to ducing vacuolation (“spongiosis”) within
paper claiming—to the initial amuse- DNA-damaging radiation. neurons (see the figure).
ment of his colleagues—that scrapie, a de-
generative disease that affects the central Daniel Carleton Gajdusek, who won a Does this mean that PrPSc is the prion?
nervous system of sheep, is caused by “pro- Nobel Prize for showing that Kuru was a hu- Weissmann’s discovery in 1993 that prion
teinaceous infectious particles,” which he man disease transmitted by cannibalism in protein (Prnp)–ablated mice are resistant
called prions (1). Prusiner’s intuition, which Papua New Guinea, proposed in 1959 that to scrapie (4) was designed to disprove
earned him the 1997 Nobel Prize, is influ- the neurodegenerative disorders Kuru, scra- the protein-only hypothesis but failed to
encing our approach to an ever-expanding pie, and Creutzfeldt-Jakob disease (CJD) do so. However, Prnp deletion in mice also
variety of seemingly unrelated diseases and are caused by “slow viruses.” Indeed, prions fell short of proving the prion hypothesis.
physiological processes, and its implications behave similarly to neurotropic viruses in If PrPC were the receptor of an imaginary
reverberate to the present day. many surprising ways, including the colo- “scrapie virus,” its ablation may also render
nization of extraneural organs followed mice resistant to scrapie. More direct evi-
The two decades preceding Prusiner’s pa- by neuroinvasion of the brain through pe- dence for Prusiner’s ideas emerged in 2001
per had witnessed the immense success of ripheral nerves (3). Yet, Prusiner purified from Claudio Soto’s landmark experiment:
molecular biology, including the cracking the agent and found it to be smaller than a Repeated cycles of PrPSc fragmentation,
of the genetic code; the elucidation of DNA virus: No informational nucleic acid would when followed by addition of PrPC and ag-
replication, transcription, and translation; fit into it. Over time, the group of prion dis- gregate regeneration, can multiply prions
and the cloning of genes. These discover- eases grew to include other human (fatal fa- ad libitum (5). These findings strengthen
ies prompted Francis Crick to conceptual- milial insomnia) and animal (bovine spon- the hypothesis that the transfer of struc-
ize the “central dogma”: Information flows giform encephalopathy, also called mad tural information can occur horizontally
unidirectionally from DNA to proteins. But cow disease, and chronic wasting disease) between proteins.
although religious dogmas may be eternal, disorders, but no causative virus has been
the shelf life of scientific dogmas is inevita- identified and their prion etiology is now More recently, the prion concept has been
well accepted. applied, sometimes overenthusiastically, to
Institute of Neuropathology, University of Zürich, virtually all diseases characterized by pro-
Rämistrasse 100, CH-8091 Zürich, Switzerland. Email: But prions did not contradict Crick’s cen- gressive deposition of aggregated proteins
[email protected] tral dogma after all. Charles Weissmann, in the central nervous system, whether in-
refusing to believe that a protein could ex- fectious or not—and even to physiological
processes such as memory formation (6).
a-Synuclein aggregates can self-propagate

32 2 OCTOBER 2020 • VOL 370 ISSUE 6512 sciencemag.org SCIENCE

INSIGHTS

Three centuries of prion science

The timeline shows key prion-related discoveries. In 1982, Prusiner suggested that the prion protein (PrP) is the infectious cause of spongiform encephalopathies,
including Kuru, scrapie, and Creutzfeldt-Jakob disease (CJD). These insights have had implications for many neurodegenerative diseases involving prionoids, but
many questions still remain unanswered.

2 CJD brain with vacuolation (spongiosis) 3 An antiprion molecule targets prion aggregates 4 Structure of PrP aggregate (PrPSc)

1700 1800 1900 1935 1950 1975 2000
1 2
3 4

1732 1898 1936 1959 1982 1993 1996 2001 2007 2016 2019

Scrapie Neuronal Scrapie Similarities Prusiner Mice without Cellular Prp Protein Spectral PrPC controls A plausible
reported vacuolation transmissibility between isolates the the prion protein (PrPC) is misfolding discrimination myelin model of
in sheep recognized as recognized scrapie, CJD, scrapie agent gene (Prnp) essential cyclic of prion homeostasis prion
a feature of and Kuru and names it are resistant for prion amplifcation strains structure
scrapie reported “prion” to prions neurotoxicity

GRAPHIC: MELISSA THOMAS BAUM/SCIENCE; (IMAGES, LEFT TO RIGHT) K. FRONTZEK; in the brains of Parkinson’s disease patients so, most material aggregated in vitro may be of toxicity—and therefore also a target for
U. S. HERMANN ET AL., SCI. TRANSL. MED. 7, 299RA123 (2015); G. SPAGNOLLI, ADAPTED FROM (12) (7), in cultured cells, and in mice (8). This noninfectious and may not be informative of therapeutic interventions.
implies that a-synuclein is a de facto prion the structure of the prion or of its replicative
and that its handling demands high bio- mechanism. High-resolution three-dimensional struc-
safety standards. Similar arguments were tures of prions are also required to solve
made for tau and amyloid-b (Ab) aggre- Of all the models that have been proposed the long-standing question of prion strains,
gates, the major hallmarks of Alzheimer’s so far, the most plausible suggests that the which share the same PrP sequence and yet
disease (9). However, prions caused many prion consists of fibrils arranged as four- cause distinct diseases (e.g., “hyper” and
epidemics, whereas infectiousness has not rung b-solenoids (12) stacked either head-to- “drowsy” phenotypes in minks), the traits of
been conclusively demonstrated for other tail or head-to-head. Cryo–electron micros- which are maintained over successive rounds
protein aggregates—and specifically not copy of purified glycosylphosphatidylinositol of infection. Viral strains are defined by spe-
through oral transmission. Protein aggre- (GPI)–anchorless prion fibrils (13) supports cific polymorphisms in their respective ge-
gates that were not shown to be serially this model, thus providing the first high- nomes, and the existence of strains in prion
transmissible across multiple generations magnification images of infectious prions, al- diseases was long thought to be incontro-
of hosts are better regarded as “prionoids,” beit the resolution does not suffice to deter- vertible evidence for the involvement of
even if they share molecular mechanisms of mine the precise arrangement of the mono- nucleic acids. However, after four decades
amplification with bona fide prions in vitro. mers within the fibrils. These structures are of failed attempts to isolate any scrapie-
quite different from those of tau, a-synu- specific genomes, strains are now thought to
As predicted by Prusiner in the closing clein, and Ab and also differ from recombi- be caused by different PrPSc conformations
lines of his paper, the “prion revolution” nant PrP fibrils—all of which are arranged in that can be distinguished with conformer-
boosted research in the field of neurodegen- long fibers with no cavity. Hence, PrPSc has sensitive fluorescent polythiophenes.
eration by providing an intellectual frame- distinctive structural characteristics, but it is
work that might explain many aspects of unknown whether and how these peculiari- Embarrassingly for the prion field, no
Alzheimer’s disease, Parkinson’s disease, ties relate to their frightening infectivity. definitive structural evidence for these
and many other neurodegenerative diseases presumptions has come forward, and the
featuring protein aggregates. Although cel- The link between the generation of PrP “strainness” of bona fide infectious prions
lular PrPC is now known to be crucial for aggregates and their neurotoxicity is also un- is still diagnosed using imperfect surrogate
the maintenance of peripheral myelin (10), clear. A large body of evidence (14) indicates biomarkers such as differential resistance
our understanding of prions has essentially that PrPC is necessary for toxicity, perhaps to disaggregation and proteolysis. By con-
stagnated for more than a decade and may because extracellular PrPSc oligomers dock trast, conformational heterogeneity was
now be lagging behind that of prionoids. to PrPC on the surface of diverse cell types. reported to correlate with distinct clinical
What is really known about prions, after Another aspect specific to prion infections phenotypes in some prionoid pathologies,
almost 40 years since Prusiner’s discovery? pertains to the peculiar morphology of the although the stability of different conforma-
damage that it wreaks on the brain. Of all tions in serial transmission experiments is
One crucial obstacle to advancing prion aggregation-prone proteins, prions are the not yet fully established.
research is the lack of high-resolution struc- only ones that cause extensive intraneuro-
tures of PrPSc owing to its insolubility, its nal vacuolation (spongiosis), the severity of But how stable are prion strains across
noncrystalline aggregational state, and the which increases during disease progression. generations? RNA viruses achieve maximal
persistent difficulties in preparing high- This phenomenon is as much intriguing as fitness by creating quasispecies, clouds of
purity infectious material de novo from it is mysterious. To date, almost nothing is variants in precarious equilibrium between
recombinant protein (11). This raises the known about the cellular and molecular adaptive mutagenesis and error catastro-
possibility that infectious aggregates may pathologies underlying vacuole formation; phe. Notably, prions can also engender qua-
constitute a sparsely populated conforma- yet its ubiquity in all known prion diseases sispecies whose monoclonal constituents
tional variant within such preparations. If suggests that vacuolation is a prime driver can be isolated from cultured cells by apply-
ing various kinds of selective pressure (15).

SCIENCE sciencemag.org 2 OCTOBER 2020 • VOL 370 ISSUE 6512 33

INSIGHTS | PERSPECTIVES

The structural mechanisms underlying this CANCER
phenomenon are unknown and may in-
volve conformational selection of distinct Mutational selection in
PrPSc species. The conformational selection normal urothelium
model predicates the coexistence of multi-
ple conformers within a single infected or- Mutations in normal tissue point to causes of
ganism, some of which may replicate more DNA damage and set the stage for cancer
efficiently in their host under certain envi-
ronmental circumstances. The incubation By Steven G. Rozen selective advantage (indicated by high preva-
time of prion infections can vary immensely lence of a clone) usually stemmed from muta-
between different strains, and the delay in C ells have elaborate machinery to pre- tions in genes encoding proteins involved in
the onset of the pathology might reflect the serve the integrity of their genomes, histone modification and chromatin remod-
time needed for such selection to occur. which nevertheless relentlessly gather eling. Within these genes, truncating, and
new mutations over time. Recent presumably inactivating, mutations were
PrPSc conformer heterogeneity may also technical advances have enabled often strongly selected. The affected genes
underlie the barriers that control interspecies high-resolution delineation of these prominently included KMT2D (histone-lysine
prion transmission, the strength of which is accumulated mutations and their spatial or- N-methyltransferase 2D) and KDM6A (lysine-
variable and depends both on host factors ganization in tissues (1–8). It is now possible specific demethylase 6A). These chromatin
and on prion strains. Although prion propa- to deduce which mutations in which genes remodeling genes are also mutated in many
gation from cows to humans results in vari- allowed cells to outcompete their neighbors cancers of the urothelium, suggesting that
ant CJD, sheep prions appear to be largely and colonize nearby regions of normal tissue the nonmalignant expanded clones driven
innocuous to humans. This species barrier (clonal expansion). One can also sometimes by these genes regularly, but not frequently,
relies both on the structural diversity of the ascertain which endogenous mutational pro- become malignant. This contrasts with the
PrPSc contained in the inoculum and the PrPC cesses or external mutagens caused these so- esophagus, in which some genes that often
of the host, which cannot always interact matic mutations. These recent findings have drive clonal expansion in the normal tissue
with the misfolded conformer efficiently. profound implications for understanding only rarely act as drivers in cancers (3, 6).
aging and the early stages of cancer initia-
The ideas promulgated by Prusiner tion. On pages 75 and 82 of this issue, Lawson By examining many samples per indi-
have undergone a marked metamorphosis. et al. (9) and Li et al. (10), respectively, delve vidual, Lawson et al. found that, in some
Templated nucleation of protein aggregates into the somatic mutations lurking in the individuals, the urothelium showed strong
is now known to underlie not only diseases normal urothelium—the lining of the blad- selective preference for mutations in partic-
but also many physiological processes, some der and ureter—and relate them to cancers ular genes. For example, one person had 35
of which bear little resemblance to the origi- in these tissues. distinct KDM6A mutations distributed over
nal set of diseases that attracted Prusiner’s multiple clones and 2 different ARID1A (AT-
attention. Notably, the structural predictions Although the findings of the two studies rich interactive domain-containing protein
of the prion model were verified for several are broadly consistent, substantial method- 1A, also involved in chromatin remodeling)
prionoids but not for prions. As such, many ological differences account for some diver- mutations, whereas another person had 4
of the questions raised by Prusiner in 1982— gence. Lawson et al. studied the urothelium different KDM6A mutations and 20 different
prion structure, mechanism of replication, of the bladder, primarily in organ donors ARID1A mutations distributed over multiple
and drivers of toxicity—are still open. Based from Europe, whereas Li et al. studied the clones. It was not possible to make such ob-
on historical evidence, addressing these nonmalignant urothelium of both the blad- servations with the study design of Li et al.,
questions in the prion arena may, once again, der and the ureter in cancer patients from but with a larger sample of individuals, Li et
provide answers that will also apply to more China. Additionally, Lawson et al. studied al. may have been better able to assess the
prevalent neurodegenerative diseases. j an average of ~100 tiny (~0.01 mm2) urothe- prevalence of mutations in different genes.
lial samples each from 15 organ donors and
REFERENCES AND NOTES 5 cancer patients; by contrast, Li et al. stud- Both studies found that, with the excep-
ied an average of 1.3 large (~2 mm2) samples tion of the chromatin remodeling genes,
1. S. B. Prusiner, Science 216, 136 (1982). of normal tissue from 120 cancer patients. most other genes that are commonly mu-
2. J. S. Griffith, Nature 215, 1043 (1967). The studies also differed in their sequencing tated in urothelial cancers were rarely mu-
3. M. Prinz et al., Nature 425, 957 (2003). approaches. Lawson et al. sequenced a mix tated in normal urothelium. These include
4. H. Büeler et al., Cell 73, 1339 (1993). of whole genomes, whole exomes, and a tar- the well-established cancer driver genes
5. G. P. Saborio, B. Permanne, C. Soto, Nature 411, 810 geted panel of 321 cancer-associated genes, PIK3CA (phosphatidylinositol 4,5-bisphos-
whereas Li et al. sequenced whole exomes. phate 3-kinase catalytic subunit a), FGFR3
(2001). (fibroblast growth factor receptor 3), and
6. R. Hervas et al., Science 367, 1230 (2020). Despite the methodological differences, RB1 (RB transcriptional corepressor 1). Thus,
7. J.-Y. Li et al., Nat. Med. 14, 501 (2008). both studies analyzed somatic mutations in mutations in these driver genes might be
8. K. C. Luk et al., Proc. Natl.Acad. Sci. U.S.A. 106, 20051 normal urothelium to show that clonal ex- later events that finally trigger malignant
pansion occurred. Both studies found that a transformation of cells harboring mutations
(2009). that drive clonal expansion (see the figure).
9. F. Clavaguera et al., Nat. Cell Biol. 11, 909 (2009). Centre for Computational Biology and Programme in Both studies also found few large-scale copy
10. A. Küffer et al., Nature 536, 464 (2016). Cancer and Stem Cell Biology, Duke–National University of number alterations in the normal urothe-
11. G. Legname et al., Science 305, 673 (2004). Singapore (NUS) Medical School, 169857 Singapore. Email: lium, in contrast to the abundant large-scale
12. G. Spagnolli et al., PLOS Pathog. 15, e1007864 (2019). [email protected]
13. E.Vázquez-Fernández et al., PLOS Pathog. 12, e1005835

(2016).
14. S. Brandner et al., Nature 379, 339 (1996).
15. J. Li, S. Browning, S. P. Mahal,A. M. Oelschlegel, C.

Weissmann, Science 327, 869 (2010).

ACKNOWLEDGMENTS

The authors thank K. Frontzek for the micrograph of a CJD
brain and G. Spagnolli for the PrPSc molecular model.The
Aguzzi lab is supported by the Nomis Foundation and the
Swiss National Science Foundation.

10.1126/science.abb8577

34 2 OCTOBER 2020 • VOL 370 ISSUE 6512 sciencemag.org SCIENCE

Mutations in the bladder and ureter linings by Li et al.), but very limited in Europe (the
location of individuals studied by Lawson et
Cells in the normal urothelium that lines the ureter and bladder accumulate mutations caused by endogenous al.) (11–13). AA mutagenesis was previously
mutagenic processes (e.g., APOBEC cytidine deaminases) or by exogenous mutagens (e.g., AA or tobacco detected in a few normal urothelial samples
smoke). Some mutations (e.g., in the chromatin modifying genes KMT2D and KDM6A) confer competitive (13, 14), and Li et al. now show that AA muta-
advantages that drive cells to colonize larger regions of the urothelium. Additional mutations (e.g., in the genes genesis is widespread in normal urothelium.
TP53, PIK3CA, FGFR3, or RB1) and perhaps other changes are needed to trigger malignant transformation. This suggests that it may be possible to re-
liably assess the AA mutational signature in
Kidney cells or DNA shed from normal urothelium
and, using this signature, noninvasively as-
Ureter sess previous AA exposure (13). A noninvasive
test for AA exposure would offer substantial
Bladder Section of urothelium from benefits to research into the epidemiology of
the bladder or ureter AA-associated disease and to secondary pre-
vention of cancer and kidney failure in AA-
Sequence and detect somatic mutations exposed individuals.

Mutational signatures suggest causes of mutations Detect genes that drive clonal expansion An unexpected finding was the strong dif-
ference in driver mutation preferences be-
Endogenous Exogenous tween individuals, with, for example, one per-
APOBECs son having multiple independent mutations
Clonal expansion in KDM6A and few in ARID1A, and another
Smoking in normal urothelium: person with the opposite pattern. It will be
KMT2D and KDM6A interesting to see how general this phenom-
AA-containing herbs mutations enon is and whether driver preferences can
predict cancer risk or, very speculatively,
AA, aristolochic acid; FGFR3, fbroblast growth factor receptor 3; Cancer Malignant transformation suggest prophylactic therapies. Each study
KDM6A, lysine-specifc demethylase 6A; triggered by mutations in other explored mutations in a different dimen-
KMT2D, histone-lysine N-methyltransferase 2D; genes (e.g., TP53, PIK3CA, FGFR3, sion. One studied one or two samples from
PIK3CA, phosphatidylinositol 4,5-bisphosphate 3-kinase catalytic subunit a; RB1) or other factors a broad swath of individuals, and the other
RB1, RB transcriptional corepressor 1; TP53, tumor protein 53. was a deep study of many samples from a few
individuals. To understand the implications
GRAPHIC: MELISSA THOMAS BAUM/SCIENCE copy number alterations found in urothelial activation, and whether it is a cause or an ef- of differences in preferences of driver gene
cancers. Lawson et al. detected copy number fect of transformation, are unknown. mutations between individuals, a study that
alterations in only 28% of exomes, whereas is simultaneously broad and deep is required.
the average bladder cancer has 200 per Lawson et al. sequenced the entire ge- Nevertheless, the shared message of Li et al.
exome. Thus, genome instability occurs late nomes of a subset of the urothelial samples, and Lawson et al. is that histologically nor-
in the malignant transformation of the uro- which enabled them to uncover previously mal urothelium contains many clones that
thelium, a pattern also seen in skin, esopha- unknown mutational signatures. One of these are only a few steps away from turning malig-
geal, and colon transformation (1, 3, 5). signatures correlated with tobacco smoking. nant but that rarely do. Studies using mouse
This may solve the puzzle that no specific models of the development of urothelial
Slowing the accumulation of mutations in smoking-associated mutational signature cancers, similar to a recent study of the de-
normal tissue might slow aging and reduce has been identified in bladder cancer, even velopment of esophageal cancer (15), would
the risk of cancer, and therefore it is important though tobacco smoking is a well-established further delineate the mechanisms by which
to understand the causes of these mutations. risk factor for bladder cancer and causes a mutations in the normal human urothelium
The two studies analyzed their respective highly recognizable signature in lung cancer. sometimes give rise to cancer. j
sequencing data for “mutational signatures” The smoking signature in lung cancer bears
that might point to these causes. Mutational no resemblance to the smoking-associated REFERENCES AND NOTES
signatures are the patterns of single base mu- signature identified by Lawson et al., and the
tations within distinct sequences of preced- mechanisms by which tobacco might cause it 1. I. Martincorena et al., Science 348, 880 (2015).
ing and following bases that can distinguish are unknown. Li et al. did not detect this sig- 2. M.A. Lodato et al., Science 359, 555 (2018).
various mutational processes. Both studies nature, possibly because they analyzed whole 3. I. Martincorena et al., Science 362, 911 (2018).
identified the mutational signature caused by exomes rather than whole genomes. 4. S. F. Brunner et al., Nature 574, 538 (2019).
APOBEC cytidine deaminase activity in more 5. H. Lee-Six et al., Nature 574, 532 (2019).
than half of the individuals studied, although A prominent difference between the mu- 6. A.Yokoyama et al., Nature 565, 312 (2019).
it was not found in every sample from those tational signatures detected in the two stud- 7. L. Moore et al., Nature 580, 640 (2020).
individuals. This signature is almost always ies is the presence of the signature caused by 8. K.Yoshida et al., Nature 578, 266 (2020).
found in bladder cancer and often in cancer aristolochic acid (AA) in more than half of 9. A. R.J. Lawson et al., Science 370, 75 (2020).
of the ureter; it is also found in many other the ureter samples and about a third of the 10. R. Li et al., Science 370, 82 (2020).
cancer types (11). The two studies show that bladder samples examined by Li et al. AA is 11. L. B.Alexandrov et al., Nature 578, 94 (2020).
increased APOBEC activity seems to accom- a kidney toxin and carcinogen that occurs 12. A.W.T. Ng et al., Sci.Transl. Med. 9, eaan6446 (2017).
pany malignant transformation in urothelial naturally in some herbs used as medicine. AA 13. H. Lu et al., Theranostics 10, 4323 (2020).
cancer. The mechanisms responsible for this exposure is widespread in East Asia, includ- 14. Y. Du et al., Eur. Urol. 71, 841 (2017).
ing China (the location of patients studied 15. B. Colom et al., Nat. Genet. 52, 604 (2020).

ACKNOWLEDGMENTS

This work was supported by Singapore National
Medical Research Council grant MOH-000032/MOH-
CIRG18may-0004 and by the Singapore Ministry of Health
via the Duke-NUS Signature Research Programmes. I thank
A. Boot for comments.

10.1126/science.abe0955

SCIENCE sciencemag.org 2 OCTOBER 2020 • VOL 370 ISSUE 6512 35

INSIGHTS | PERSPECTIVES

R E G E N E R AT I O N

Coaxing stem cells to repair the spinal cord

Spinal cells in mice can be induced to generate protective oligodendrocytes after injury

By Catherina G. Becker and Thomas Becker dendrocyte production. Injury induces ep- more than one-third of ependymal progeny,
endymal cells to proliferate, which changes the astrocytic scar, which also consists of
S pinal cord injuries destroy neurons, ax- gene accessibility. Moreover, spinal injury cells derived from ependymal cells, was not
onal processes, and oligodendrocytes induces inflammation (3) and attendant sig- depleted. The astrocytic scar has protective
that provide insulation and protection naling molecules that may influence gene ex- functions in wound healing (4), so it is im-
of axons by means of membrane wrap- pression in ependymal cells, but these factors portant that any therapeutic approach does
pings, called myelin sheaths. None of are largely unknown. not compromise scar tissue. Together, these
these cellular structures are efficiently findings indicate that ependymal cells in the
replaced after injury. This can lead to lifelong The observed increase in oligodendrocyte spinal cord can be reprogrammed for oligo-
disabilities, including paralysis. Endogenous numbers was substantial. High numbers dendrocyte production after injury in mice.
stem cells exist in the spinal cord, but after of cells may have been reached because the
injury they produce mainly astrocytic scar progeny of ependymal cells comprises prolif- Full differentiation of oligodendrocytes is
tissue, no neurons, and very few oligoden- erative oligodendrocyte progenitors that can a prerequisite for efficient impulse propaga-
drocytes (1). On page 73 of this issue, Llorens- be considered transit-amplifying cells for oli- tion in axons. One problem with naturally
Bobadilla et al. (2) show that by overexpress- godendrocytes. Even though oligodendrocyte occurring remyelination is that new my-
ing a single factor in spinal stem cells, they production was boosted from less than 1% to elin sheaths are usually thinner than those
can boost post-injury production of oligoden- that were originally present, compromising
drocytes in mice. This leads to better remy- Induced protection conduction velocity. It will be interesting to
elination of remaining axons that lost their find out whether new myelin displays “full
myelin and to improved axonal impulse con- Chromatin in adult ependymal cells in the mouse spinal thickness” and is comparable to myelin
duction in vivo. The study raises hope that cord is accessible to the transcription factor OLIG2 that has arisen during development. It will
endogenous stem cells in the injured spinal (oligodendrocyte transcription factor 2). After spinal also be interesting to determine whether
cord can be recruited to generate neural cell cord injury, myelin is lost near the injury site and can new myelin persists for longer than the 3
types in a more balanced way after injury to be lost from spared axons. Overexpression of OLIG2 in months reported in this study to indicate
promote recovery of function. ependymal cells in mice reprograms these stem cells permanent repair.
to produce more oligodendrocytes, which form new
One of the problems with spinal cord in- myelin near the injury and improve axon conduction. Llorens-Bobadilla et al. found that induced
jury is secondary cell death around an injury myelin improved conductance velocity in
site that leads to the loss of not only neurons, Spinal cord a mouse model of spinal cord injury. The
but also oligodendrocytes and the myelin model is a contusion injury, which resembles
sheaths they produce. This in turn causes de- Ependymal the physical impact associated with injuries
nuding of spared axons, which, bereft of their cells in humans. Conductance velocity in spared
insulation and trophic support, function axons above the injury site was improved
inefficiently and may ultimately degener- OLIG2 in OLIG2-overexpressing animals compared
ate. Llorens-Bobadilla et al. analyzed mouse to injured controls, ostensibly because of
ependymal cells (stem cells in the lining of Corticospinal improved myelination. Below the injury, no
the spinal cord central canal) to find that neurons such effects were observed, likely because of
chromatin regions with binding motifs for the scarcity of spared axons. Consequently,
the oligodendrocyte-determining transcrip- Preexisting New recovery of function was not better in OLIG2- GRAPHIC: KELLIE HOLOSKI/SCIENCE
tion factors OLIG2 (oligodendrocyte tran- oligodendrocyte oligodendrocyte overexpressing animals.
scription factor 2) and SOX10 were accessible
even though the transcription factors were Destroyed Compression To bring about functional recovery, com-
not expressed. This suggested a latent capac- myelin injury binations with other treatments would be
ity of ependymal cells to generate oligoden- needed (5) because although oligodendro-
drocytes. Indeed, inducing overexpression cytes can support axons, remyelination alone
of OLIG2 in ependymal cells in vivo strongly is insufficient to boost recovery (6). For exam-
increased the accessibility of OLIG2 binding ple, axon growth can be enhanced by modu-
sites and the production of oligodendrocytes lating intrinsic axon growth propensity (7)
from these cells after spinal injury. or the inhibitory environment (8). Moreover,
transplanted neural stem cells can form new
Neither promoter accessibility nor oligo- neurons that integrate and improve neuronal
dendrocyte production increased without an impulse conduction over an injury site (9).
injury, indicating that factors in addition to Electrical stimulation of spared fibers may
OLIG2 expression are necessary to realize the also lead to more efficient myelination (10).
latent potential of ependymal cells for oligo- In all approaches, newly grown axons could
benefit from additional myelination capacity
University of Edinburgh, Edinburgh, Scotland, UK. Email: from induced oligodendrocytes.
[email protected]; [email protected]
To realize the potential of ependymal cells
in human spinal cord injuries, it will be nec-

36 2 OCTOBER 2020 • VOL 370 ISSUE 6512 sciencemag.org SCIENCE

essary to determine whether similar stem M E TA L LU R GY
cells exist near the human spinal cord central
canal in sufficient numbers (11). The lumen A rival to superalloys at high
of the central canal progressively disappears temperatures
during childhood, but cells with stem cell po-
tential can be isolated from the human spinal Slip-pathway activation provides plasticity in a multiprincipal
cord (12). It will be necessary to determine element alloy with high-temperature strength
how similar these cells are to mouse ependy-
mal cells and whether their chromatin is sim- By Julie Cairney iron enhances its strength, creating steel,
ilarly poised to generate oligodendrocytes.
and adding yet another element, nickel,
From a therapeutic viewpoint, efficient
means of inducing gene expression are A lthough conventional alloys are improves its corrosion response, creating
needed. Viral delivery systems are under based mainly on one element, recent stainless steel. Sophisticated superalloys
development that evade immune detec- design efforts have focused on mul- have complex compositions that provide
tion and can be switched off—an important tiprincipal element alloys (MPEAs) high performance near their melting point,
safety feature to avoid unwanted prolifera- that contain substantial quantities but they are still based on a primary ele-
tion of cells (13). Such strategies could be of several elements. Success with this ment, usually nickel, cobalt, or iron.
used to treat any demyelination after spinal
injury or in demyelinating diseases, such as approach requires a robust understanding To expand the alloy design space, more
multiple sclerosis.
of the mechanistic origin of MPEA proper- recent efforts have shifted toward the de-
Ependymal cells in the mouse spinal cord
have a stem cell potential that sets them ties. On page 95 of this issue, Wang et al. velopment of alloys that contain substantial
apart from other spinal cell types. For exam-
ple, Llorens-Bobadilla et al. found that astro- (1) report the deformation behavior in a quantities of three or more elements (see
cytes did not produce oligodendrocytes after
forced OLIG2 expression. What else could ep- promising body-centered cubic (bcc) MPEA the figure). These materials are variously re-
endymal cells do? In rats, stem cells derived
from whole spinal cord, including ependymal with good room-temperature plasticity and ferred to as MPEAs, complex concentrated
cells, formed neurons when transplanted
into the hippocampus, a neurogenic region high strength at the temperatures at which alloys, or high-entropy alloys (a subset of
of the brain (14). In anamniotes (salaman-
ders and fishes), spinal ependymal progeni- conventional alloys would soften. They MPEAs). Some of these new alloys display
tor cells generate neurons after injury in situ
(12). This indicates a potential for neurogen- observed multiplanar, multicharacter dis- unprecedented combinations of strength,
esis in spinal stem cells across vertebrates. It
might be informative to determine the nature location slip that was not expected in bcc ductility, high-temperature performance, or
of the gene regulatory programs activated by
anamniotes to produce neurons from epen- functional properties (2, 3). Each
dymal cells in the spinal cord and whether
these could be activated by gene therapy in new composition can result in the
nonregenerating mammals to contribute to
repair after injury. j A more equal union formation of different phases and
microstructures within the alloy,
REFERENCES AND NOTES The schematics represent the atomic structure of conventional which can in turn be altered by
and multifunctional alloys. Conventional alloys are based mechanical deformation and heat
1. K. Meletis et al., PLOS Biol. 6, e182 (2008). mainly on a single element, compared with multicomponent
2. E. Llorens-Bobadilla et al., Science 370, eabb8795
alloys that contain several elements in similar proportions. treatment. With such an enormous
(2020).
3. A. D. Greenhalgh, S. David, F. C. Bennett, Nat. Rev. Elements Conventional Multiprincipal range of possibilities, traditional
A dilute alloy component alloy trial-and-error approaches are inef-
Neurosci. 21, 139 (2020). B fective, and researchers are turning
4. M.A.Anderson et al., Nature 532, 195 (2016). C to computational and combinato-
5. G. Courtine, M.V. Sofroniew, Nat. Med. 25, 898 (2019). D rial approaches to predict the el-
6. G.J. Duncan et al., Nat. Commun. 9, 3066 (2018). emental combinations that could
7. J. Ruschel et al., Science 348, 347 (2015). lead to alloys with desirable prop-
8. B.T. Lang et al., Nature 518, 404 (2015).
9. S. Ceto, K.J. Sekiguchi,Y.Takashima,A. Nimmerjahn, M. erties (4, 5). However, for these ap-

H.Tuszynski, Cell Stem Cell 27, 430 (2020). proaches to be successful, it is criti-
10. S. Mensch et al., Nat. Neurosci. 18, 628 (2015).
11. D. Garcia-Ovejero et al., Brain 138, 1583 (2015). cal that the alloy design process is
12. C. G. Becker,T. Becker,J.-P. Hugnot, Prog. Neurobiol. 170,
systems. This property is attributed to vari- guided by an understanding of the origins
67 (2018).
13. E. R. Burnside et al., Brain 141, 2362 (2018). ations in the glide resistance along the core of the specific properties that are desired.
14. L. S. Shihabuddin, P.J. Horner,J. Ray, F. H. Gage, J.
of dislocations, created by the atomic-scale Refractory MPEA alloys, with their ex-
Neurosci. 20, 8727 (2000).
fluctuations in composition that are charac- cellent high-temperature strength, show
ACKNOWLEDGMENTS
teristic of MPEAs. This mechanism explains great promise. They are composed of
Supported by the EU Cofund ERANET NEURON consor-
tium NEURONICHE with contributions from MRC (MR/ the plasticity and could be used to guide the combinations of three or more of the ele-
R001049/1), Spinal Research, and Wings for Life (C.G.B.) and
by BBSRC project grant BB/R003742/1 (T.B.). design of new MPEA candidates for high- ments chromium (Cr), molybdenum (Mo),

10.1126/science.abe1661 temperature applications in aerospace and niobium (Nb), vanadium (V), tantalum

power generation. (Ta), tungsten (W), hafnium (Hf ), titanium

GRAPHIC: KELLIE HOLOSKI/SCIENCE For centuries, alloy design has involved (Ti), or zirconium (Zr) at nearly equal con-

taking a base metal and adding small centrations. The alloys usually have a bcc

amounts of other elements to improve the crystal structure and, for many of the most

properties. For example, adding carbon to desirable combinations, are single-phase

solid solutions. However, unlike dilute

School of Aerospace, Mechanical and Mechatronic bcc alloys, many of these alloys show ex-
Engineering, University of Sydney, Sydney, Australia. Email: cellent retention of strength up to 1900 K
[email protected] (6). Aerospace, petrochemical, and power-

SCIENCE sciencemag.org 2 OCTOBER 2020 • VOL 370 ISSUE 6512 37

INSIGHTS | PERSPECTIVES

generation industries all require tough be operative in addition to those expected C O N S E R VAT I O N
components that are exposed to high tem- for a bcc alloy at room temperature. Wang
peratures. Superalloys, the best available et al. attribute this multiplanar, multichar- A boost for
option, have an operational limit of around acter dislocation slip to variations in the
1100 K (2). This material constraint affects glide resistance for dislocations caused by freshwater
the efficiency of potential new technologies the atomic-scale chemical fluctuations in
for power generation by limiting operating composition along the core of the disloca- conservation
temperatures (7). It is also a serious issue in tion. Atomistic simulations show that the
aerospace applications. Presently, aircraft plane that has the lowest stress required Integrating freshwater and
parts that are exposed to hot environments for the movement of dislocations can vary terrestrial conservation
in engines, or components that will be ex- in this system depending on the local planning has high returns
posed to the high temperatures caused by atomic configuration.
hypersonic travel, require complex and ex- By Robin Abell and Ian J. Harrison
pensive ceramic thermal barrier coatings to The implication of this finding is that
withstand the service environment (8). there are additional pathways for disloca- S ystematic conservation planning—
tion slip, which is desirable for plasticity a data-driven process for prioritiz-
MoNbTaW, MoNbTaVW (9), HfNbTaTiZr and toughness. This observation explains ing biodiversity conservation re-
(10), and HfMoNbTiZr (11) are all examples the plasticity and supports an explana- sources—has been strongly biased
of MPEAs that display excellent high-tem- tion for high-temperature strength based over the past two decades toward
perature strength (1). However, many of around the slip mechanism instead of solid- terrestrial and marine species and
these systems have limited room-tempera- solution strengthening. Deformation was systems (1). Freshwater ecosystems, which
ture ductility, which is characteristic of bcc studied at room temperature, and future are among the most threatened on Earth,
alloys. The lack of ductility in conventional work at high temperature may reveal more have received less attention. Wetland ex-
bcc systems is related to the mobility of de- details about the active slip systems. tent is estimated to have declined globally
fects (dislocations), but dislocations could by nearly 70% since 1900 and, on average,
behave differently in bcc MPEAs because Activation of additional slip pathways as freshwater vertebrate populations declined
of local variations in composition along a design goal will require renewed effort to by 84% between 1970 and 2016 (2). There
their core (2, 6). Local compositional fluc- understand how atoms are arranged along is an urgent need for prioritizing resources
tuations are intrinsic to MPEAs, in which the core of the dislocation. Despite sugges- toward freshwater conservation. On page
the elements that surround each individual tions of local chemical order in MPEAs, ex- 117 of this issue, Leal et al. (3) show that
atom vary (see the figure, right). The high- perimental verification has been ambigu- such prioritizations need not be a zero-sum
temperature strength in MPEA alloys has ous (3). In Wang et al.’s work, atom probe game: Integrated cross-realm conservation
been attributed to solid-solution strength- tomography suggested that the atoms were planning can, for a negligible reduction
ening by regions of concentrated solute, the randomly distributed. A smaller number in terrestrial benefits, increase freshwater
mobility of certain dislocation structures, or of species in this alloy, as compared with benefits up to 600%.
both effects (12). Understanding the details many other MPEAs, has reduced the com-
of the dislocation structure and motion is plexity of analysis. Further atom probe These results are important because, at this
crucial for a mechanism-guided search for work could be carried out on carefully cho- moment, the global conservation community
the best refractory bcc alloys across the im- sen model MPEAs to minimize overlapping is setting targets for the next 30 years. The
mense range of possible compositions. peaks in data. Last, if bcc MPEA alloys are Convention on Biological Diversity (CBD)
truly to rival superalloys for high-temper- is defining its Post-2020 Global Biodiversity
Wang et al. present a new MPEA MoNbTi ature use, consideration must be given to Framework, to replace the expiring Aichi
alloy with good room-temperature strength factors beyond the strength, ductility, and Biodiversity Targets. There is an opportunity
but considerably lower density than many toughness. This includes oxidation resis- with the renewed CBD framework to create
of the other refractory MPEA options. tance, creep strength, fatigue strength, and a policy environment and commitments de-
Density is important for transport appli- routes for manufacture, offering directions signed to “bend the curve” of freshwater bio-
cations, especially in rotating parts, where for future research. j diversity loss (4).
lower density increases the allowable ser-
vice temperature by decreasing the stress REFERENCES AND NOTES Leal et al.’s findings, from their studies in
caused by self-loading. The alloy displays Pará, Brazil, are of particular relevance to
homogeneous plasticity in microscale ten- 1. F.Wang et al., Science 370, 95 (2020). two targets in the draft framework, which
sion tests performed at room temperature 2. D. B. Miracle, O. N. Senkov, Acta Mater. 122, 448 (2017). will be negotiated during the first quarter
in a scanning electron microscope. To un- 3. E. P. George, D. Raabe, R. O. Ritchie, Nat. Rev. Mater. 4, of 2021 and finalized at the 15th meeting of
derstand the deformation process, Wang et the Conference of the Parties to the CBD,
al. used a sophisticated experimental setup 515 (2019). scheduled for May 2021. One target aims
that combines microscale mechanical test- 4. A.Abu-Odeh et al., Acta Mater. 152, 41 (2018). for a proportion of global land and sea ar-
ing and advanced microscopy. They used 5. T. Borkar et al., Acta Mater. 116, 63 (2016). eas to be under spatial planning, as a pre-
a focused ion beam (FIB) to prepare cross 6. O. N. Senkov, S. Gorsse, D. B. Miracle, Acta Mater. 175, cursor to protecting and restoring natural
sections underneath nano-indents and ecosystems. The second aims for a propor-
performed in situ deformation on a single- 394 (2019).
crystal specimen, again prepared with a 7. Y.Ahn et al., Nucl. Eng.Technol. 47, 647 (2015). Conservation International, Arlington, VA, USA.
FIB. The tensile axis was aligned with the 8. B. Liu et al., J. Mater. Sci.Technol. 35, 833 (2019). Email: [email protected]
[001] direction so that the four 1/2k111l–type 9. O. N. Senkov, G. B.Wilks,J. M. Scott, D. B. Miracle,
Burgers vectors are equally stressed. They
observed unexpected nonscrew dislocation Intermetallics 19, 698 (2011).
structures. Multiple slip systems appear to 10. O. N. Senkov et al., J. Mater. Sci. 47, 4062 (2012).
11. N. N. Guo et al., Mater. Des. 81, 87 (2015).
12. F. Maresca,W.A. Curtin, Acta Mater. 182, 235 (2020).

ACKNOWLEDGMENTS

The author thanks M. Griffith for feedback on the manuscript
and acknowledges funding from the Australian Research
Council Future Fellowship (FT180100232).

10.1126/science.abd6587

38 2 OCTOBER 2020 • VOL 370 ISSUE 6512 sciencemag.org SCIENCE

The Amazonas Lowlands ecoregion contains a rich and diverse fauna. The main types of vegetation include seasonally flooded forests. There are strong ecological links
between the aquatic and terrestrial habitats. Rivers and forests supply multiple ecosystem services to the communities living in the region.

PHOTO: CLAUS MEYER/MINDEN PICTURES tion of the planet to be conserved using a into larger landscape strategies, in particu- policy must be matched with the resources
combination of conventional protected ar- lar through entire watershed management to enact it. At present, freshwater systems
eas and other effective area-based conserva- plans (9, 10). The International Union for receive a tiny percentage of overall envi-
tion measures (5). The formulation of these Conservation of Nature’s approach for ronmental funding; for example, only 3.2%
targets may yet change, but the underlying identifying freshwater Key Biodiversity of the environmental funding provided by
intent will likely remain: We need to pro- Areas (KBAs), which then serve as spatial European foundations, and about 8% pro-
tect more of our planet than we currently conservation planning inputs to critical vided by North American foundations, goes
do, and we need to use spatial planning to site networks, is one applied example [e.g., to freshwater conservation in those regions
guide that protection. Lake Victoria, East Africa (11)]. These ho- or elsewhere (14). These funding challenges
listic approaches consider not only aquatic are likely to get worse, in the face of new
Historically, protected areas have been habitats of importance but also the land- economic crises and humanitarian needs
designed and managed first for terrestrial scapes in which these habitats are embed- in a world shaped by coronavirus disease
ecosystems and their species, with fresh- ded, which strongly regulate flow, sedi- 2019 and other potential pandemics. There
water considered as an afterthought, if at ment, and nutrient regimes in rivers, lakes, are always trade-offs associated with any
all. For example, at least 1249 large dams and wetlands. Freshwater KBAs and simi- prioritization—in spatial planning, policy,
are located in protected areas, with the pro- lar efforts acknowledge the impossibility and resource allocation—but the study by
cesses of “downgrading, downsizing, and of securing freshwater species and systems Leal et al., combined with complementary
degazettement” having legalized dam con- without addressing the lands draining to analyses of the multiple benefits of healthy
struction within many of these areas (6). them. Leal et al.’s study shows not only fresh waters and their watersheds, suggests
A study of continental Africa (7) showed high reward for achieving freshwater con- that those trade-offs may be more accept-
that the proportion of freshwater species servation targets from such an integrated able than we think. j
whose ranges are substantially covered by approach but also low risk to achieving
protected areas and Ramsar sites (wetlands paired terrestrial targets. REFERENCES AND NOTES
designated to be of international impor-
tance under the Ramsar Convention) is The good news from Leal et al.’s study— 1. S. Linke, E.Turak, M. G.Asmyhr, G. Hose, Aquat. Conserv.
much smaller than for birds or mammals. namely that when we conserve freshwater 29, 1063 (2019).
Globally, about 70% of river reaches (by species and habitats, we receive multiple
length) have no protected areas in their up- biodiversity cobenefits—also applies to the 2. C. Baker et al.,“A deep dive into freshwater: Living Planet
stream catchments, rendering them partly delivery of ecosystem services. Two-thirds Report 2020”(World Wildlife Fund, 2020).
conserved at best (8). These and other of today’s human population lives down-
shortfalls can be traced back in part to a stream from protected areas, and more 3. C. G. Leal et al., Science 370, 117 (2020).
lack of freshwater-focused conservation than one-quarter of water provisions sup- 4. D.Tickner et al., Bioscience 70, 330 (2020).
planning and investment (1). plied by the world’s protected areas are 5. CBD, Updated Zero Draft of the Post-2020 Global
exposed to low levels of threat (12). There
At the same time, as Leal et al. rightly is a win-win from protecting important Biodiversity Framework; www.cbd.int/article/
observe, conventional protected areas are source water areas for human communi- zero-draft-update-august-2020.
no silver bullet for conserving freshwater ties; an assessment of the likely source 6. M. L.Thieme et al., Conserv. Lett. 13, e12719 (2020).
biodiversity. Instead, they are part of a catchments of 4000 cities supplying water 7. W. R.T. Darwall et al., Conserv. Lett. 4, 474 (2011).
toolbox of options that should also em- to up to 1.7 billion people showed that 85% 8. R.Abell, B. Lehner, M.Thieme, S. Linke, Conserv. Lett. 10,
brace a growing set of strategic adaptive of the area of these catchments overlaps 384 (2017).
management approaches for freshwater with high biodiversity value freshwater 9. R.Abell,J. D.Allan, B. Lehner, Biol. Conserv. 134, 48
protection. Among these options is area- ecoregions (13). (2007).
based protection of systems beyond pro- 10. R. Flitcroft et al., in Freshwater Ecosystems in Protected
tected areas, including lands with collec- Achieving this wealth of benefits will Areas: Conservation and Management, C. M. Finlayson,
tive tenure rights for indigenous peoples require strong policy that recognizes the A. H.Arthington,J. Pittock, Eds. (Routledge, 2018),
and local communities. connections between terrestrial and fresh- pp. 190–203.
water systems and that treats those systems 11. C.A. Sayer,J.A. Carr,W. R.T. Darwall, Fish. Manag. Ecol.
Recommendations already exist for inte- as equal in importance. And this strong 26, 435 (2019).
grating freshwater conservation priorities 12. I.J. Harrison et al., Aquat. Conserv. 26, 103 (2016).
13. R.Abell et al., Aquat. Conserv. 29, 1022 (2019).
14. Synchronicity Earth Insight,“Freshwater”;
www.synchronicityearth.org/wp-content/uploads/
2018/02/Synchronicity-Earth-Freshwater-Insight.pdf.

10.1126/science.abe3887

SCIENCE sciencemag.org 2 OCTOBER 2020 • VOL 370 ISSUE 6512 39

INSIGHTS

POLICY FORUM can no longer be used, as it has been invali-
dated by the recent decision of the CJEU (2).
DATA SHARING
Standard contractual clauses, which bind
How to fix the GDPR’s frustration data transferees to comply with certain data
of global biomedical research protection standards when they receive and
process personal data, are commonly used
Sharing of data for research beyond the EU must improve for cross-border transfer in the commercial
context, but they pose particular difficulties
By Jasper Bovenberg1, David Peloquin2, propriate rights with respect to the use of for transfers to certain types of data recipi-
Barbara Bierer3, Mark Barnes2,4, their data, and that data users be required ents, including governmental agencies such
Bartha Maria Knoppers5 to follow certain standards in processing as the U.S. National Institutes of Health or
those data. But balancing these concerns universities outside the EU. Such entities
S ince the advent of the European Union against the concerns over research should are often barred by their own national laws
(EU) General Data Protection Regula- be informed by the generally scientific re- from agreeing to certain terms required to
tion (GDPR) in 2018, the biomedical search–friendly approach of the GDPR. be included in the standard contractual
research community has struggled to Current interpretations of the GDPR fail to clauses, including those specifying auditing
share data with colleagues and con- recognize how research uses of personal data of data systems by a foreign entity and sub-
sortia outside the EU, as the GDPR differ from other uses, particularly because mission to the jurisdiction of foreign courts
limits international transfers of personal data used for research purposes are often (3). Many research entities that are arms of
data. A July 2020 ruling of the Court of pseudonymized, used to derive generalizable sovereign governments either lack autho-
Justice of the European Union (CJEU) rein- knowledge that can benefit society, and can rization to waive their sovereign immunity
forced obstacles to sharing, and even data be used in this way without identification of, or have a long-standing policy not to waive
transfer to enable essential research into or perceptible harm to, data subjects. Thus, such immunity. Moreover, because the EU
coronavirus disease 2019 (COVID-19) has the balance between privacy of the individ- data transferors are often private universi-
been restricted in a recent Guidance of the ual and the benefit to society in the research ties or research institutes and transferees
European Data Protection Board (EDPB). context is different than in other contexts, are governmental or parastatal entities, the
We acknowledge the valid concerns that such as many commercial contexts in which individually negotiated interstate transfer
gave rise to the GDPR, but we are concerned data are used to construct a profile of an in- agreements contemplated by the GDPR for
that the GDPR’s limitations on data transfers dividual to permit targeted advertising with transfers between two public bodies are not
will hamper science globally in general and demonstrable impact on the individual. routinely available as an alternative to the
biomedical science in particular (see the text standard contractual clauses (4).
box) (1)—even though one stated objective of GLOBAL SHARING OF RESEARCH DATA
the GDPR is that processing of personal data The rationale behind the GDPR’s limitations Although the CJEU has upheld the validity
should serve humankind, and even though on transfers of data outside the EU is sim- of at least one set of the standard contractual
the GDPR explicitly acknowledges that the ple: When personal data are transferred to clauses to permit cross-border data transfer,
right to the protection of personal data is not non-EU countries, the level of protection en- it has also ruled that a data exporter and the
absolute and must be considered in relation sured in the EU should not be undermined. recipient of personal data using the clauses
to its function in society and be balanced The limitations aim to ensure that the are required to verify, prior to any transfer,
against other fundamental rights. We exam- “GDPR travels with the data.” Several routes whether the level of protection required by
ine whether there is room under the GDPR for valid transfer of research data have been EU law is respected in the importing country
for EU biomedical researchers to share data proposed, which we discuss below. (2). It also made clear that recipients out-
from the EU with the rest of the world to side the EU must return any received data
facilitate biomedical research. We then pro- Data may be transferred on the basis of or destroy them “in their entirety” when
pose solutions for consideration by either the “an adequacy decision.” This means that the their domestic laws no longer allow them to
EU legislature, the EU Commission, or the European Commission has decided that the comply with the EU clauses (2). The verifica-
EDPB in its planned Guidance on the pro- third country or international organization tion must consider, as regards any access by
cessing of health data for scientific research. in question ensures an “adequate level of public authorities of the importing country
Finally, we urge the EDPB to revisit its recent protection.” Such a transnational data trans- to the personal data transferred, the rele-
Guidance on COVID-19 research. fer does not require any specific authoriza- vant aspects of the legal system of the im-
tion. However, to date, adequacy decisions porting country (2). Such an assessment on
Concerns that gave rise to the GDPR in- are in place for only a limited number of a case-by-case basis (and its monitoring on
clude that data subjects be informed of use countries worldwide: Andorra, Argentina, an ongoing basis) will probably be beyond
of their personal data and be afforded ap- Canada (commercial organizations), Israel, the capabilities of most, if not all, EU re-
Japan, New Zealand, Switzerland, Uruguay, searchers and their institutions. In essence,
1Legal Pathways Life Sciences Law, Haarlem, Netherlands. and the self-governing dependencies of the this requires resource-limited private par-
2Ropes & Gray LLP, Boston, MA, USA. 3Multi-Regional Isle of Man, Guernsey, Jersey, and the Faroe ties to undertake the adequacy assessment
Clinical Trials Center of Harvard University and Brigham Islands. The adequacy decision that was process that would typically be done by the
and Women’s Hospital, Cambridge, MA, USA. 4Yale Law in place for the United States, the EU-U.S. European Commission.
School, New Haven, CT, USA. 5Centre of Genomics and “Privacy Shield” framework, was available
Policy, Faculty of Medicine, McGill University, Montreal, only to for-profit organizations and today Even if researchers would somehow be
Quebec, Canada. Email: [email protected] able to complete such an assessment (and
to monitor it on a going-forward basis), the
standard contractual clauses present com-
plications for multi-party research collabora-
tions, when the recipient organization needs
to share the data with other organizations

40 2 OCTOBER 2020 • VOL 370 ISSUE 6512 sciencemag.org SCIENCE

ILLUSTRATION: MIXMAGIC/ISTOCK.COM in their own country or in another “third disfavors the use of consent as a legal basis SHARING DATA FOR COVID-19 RESEARCH
country” in order to complete the research. and condition for the use of the data for re- Sadly, the EDPB Guidelines published on 21
Unfortunately, the standard contractual search processing, as it does not believe that April 2020 regarding data for COVID-19 re-
clauses are not clear regarding such “onward such consent can be freely given, and this search (11) lack both any sense of urgency
transfers” and the mechanisms they offer; logic might also be extended to consent and any consideration of the public good, and
unambiguous consent by the data subject asked for transfer of data out of the EU. fail to take into account other fundamental
to the onward transfer, or adherence to the rights, societal interests, and scientific con-
clauses by the onward transferee(s), are of- Recently, in at least one long-standing re- siderations. They stress that consent must be
ten not viable options. search collaboration involving the NIH, an “specific,” the “derogations and limitations in
EU research institute agreed to permit the relation to the protection of data used in re-
Although the GDPR provides that enti- transfer of genetic data from Finland to the search must apply only in so far as is strictly
ties may enter into bespoke clauses that United States on the basis that the transfer necessary,” and that “the current COVID-19
are tailored to the circumstances, such is necessary “for important reasons of public outbreak does not suspend or restrict the
bespoke clauses must be approved by the interest” (8). This recognized GDPR deroga- possibility of data subjects to exercise their
national competent supervisory authority tion to the prohibition on cross-border trans- rights.” The Guidelines go so far as to state
(5). Yet in many EU jurisdictions, the lack fer of personal data requires that the “public that “storage periods” (timelines) must be
of guidance from the EDPB on the require- interest ... shall be recognized in Union law set for COVID data. This is an inexplicable
ments for bespoke clauses means that the or in the law of the Member State to which limitation, and arguably even violates the
competent authorities have not yet estab- the controller is subject” (9). Some examples GDPR’s exemption to the storage limitation
lished a process for the review of bespoke of when this public interest provision may
clauses [e.g., (6)]. for data processed for scientific pur-
be relied upon include international data ex- poses. Pandemic researchers need
For prospective research, such change between competition authorities, tax access to past, present, and future
as interventional clinical trials, in or customs administrations, financial super- collections of human biospecimens
which data subjects provide in- visory authorities, services competent for so- and associated personal data, to an-
formed consent at the time they cial security matters, or for public health, such ticipate new waves of infections and
enroll in the study, researchers have as tracing for contagious diseases and/or any new mutations. Guidance on the
often relied on the explicit consent of elimination of doping in sport (10). The “compatibility presumption” for re-
the data subject as the means to legit- EDPB, however, fails to provide clear guide- search, announced in January 2019,
imize data transfer. However, under lines and only complicates matters by stat- is simply being postponed, leaving
the GDPR, this “transfer consent” is ing that “the derogation only applies when the research community in the dark.
subject to a number of requirements it can also be deduced from EU law or the
and limitations. The researcher must law of the Member State to which the con- Luckily, the EDPB does consider
inform data subjects about the pos- troller is subject, that such data transfers COVID research as qualifying as an
sible risk that their personal data are possible for important public interest ‘‘important public interest” to al-
will be transferred to a country for purposes including in the spirit of reciproc- low international transfer of data.
which there is no adequacy decision ity for international cooperation” (7). The However, the EDPB also notes that
or appropriate safeguards. Pursuant EDPB claims that this transfer mechanism, this derogation may only be justified
to Guidance from the EDPB, invoking data as an exception to the requirement of an ad- for “initial transfers,” as a “temporary mea-
subjects’ consent as a basis for transfer is equacy decision or appropriate safeguards, sure,” and that repetitive transfers as part of a
limited to occasional and “not repetitive” although not expressly limited to “occa- long-lasting research project cannot proceed
transfers. Consent therefore is not a viable sional” or “not repetitive” transfers, must be under this derogation. Thus, even in the im-
option for research consortia, data reposi- interpreted restrictively. portant context of COVID-19 research when
tories, and legacy collections that store data the ability to transfer personal data across
for the global research community. Also, the In sum, the inability to find a suitable international borders for research purposes
general GDPR requirements for a valid con- mechanism amid the above legal bases for is urgently needed, the EDPB minimizes the
sent continue to apply, including that it must transfer has stymied research collaborations ability of the research community to rely on
be free, informed, specific and explicit, and between the EU and the rest of the world, re- derogations that are included in the very text
subject to immediate withdrawal by the data sulting in the cessation or harmful delays of of the GDPR.
subject at any time (withdrawal of consent critical data flows (see the text box) (1).
should be as easy as giving consent). Upon BEYOND EUROPE IN THE PUBLIC INTEREST
withdrawal of consent, processing must be The above limitations for transfer of rese-
stopped, unless there is another legal basis arch data outside the EU appear to be at
to continue. odds with the generally research-friendly
intent of the GDPR. During its drafting and
In addition, since the advent of the GDPR, negotiation, the European Parliament and
ethics committees have asked researchers to the European Council paid extensive atten-
provide a detailed list of all countries that will tion to research issues, which has resulted
receive data collected as part of the study (7). in several provisions that may facilitate the
Yet at the outset of a data-intensive research processing of personal data for scientific re-
study, it is usually not possible to know all of search. Processing data for research is deemed
the countries to which data may be sent, given compatible with the initial purpose of data
the large number of collaborating parties and collection, and data may be stored longer if
service providers involved in multinational for research purposes. The GDPR explicitly
collaborative studies. Moreover, with respect allows data subjects to give general consent
to data gathered in an interventional clinical rather than specific consent for processing
trial of a medicinal product, EDPB Guidance

SCIENCE sciencemag.org 2 OCTOBER 2020 • VOL 370 ISSUE 6512 41

INSIGHTS | POLICY FORUM

Examples of biomedical research frustrated by the GDPR These clauses should reflect the specific con-

Inefficient distributed analysis of international data text, purposes, and practices of such trans-
The International Genomics of Alzheimer’s Consortium and the U.S.-based Alzheimer’s
Disease Sequencing Project based at the University of Pennsylvania have been unable to pool fers—for example, review of sharing or access
personal data on a single server because EU investigators believe that the GDPR prevents
them from sharing the European personal data with U.S.-based researchers. This creates a requests by independent Data Access Com-
scientifically compromised, inefficient, and more expensive distributed analysis of interna-
tional Alzheimer’s disease data because investigators must run identical analyses on segre- mittees. Fourth, the EDPB should (i) issue
gated pools of data in different locations. This distributed analysis model both slows research
and limits the scope of research projects in which they can engage. guidance for the approval by the competent

Protections in place, but struggling to identify a transfer mechanism supervisory authorities of bespoke clauses
European research centers used to send de-identified human genetic data to the Imputation
Server hosted by the University of Michigan. The server has been certified by an outside audi- for specific research studies and (ii) issue
tor for conformance with recognized information technology security and privacy standards
[National Institute of Standards and Technology (NIST)]. Measures are in place to secure guidance identifying when data processing
physical security of the location, space, and equipment and for identification and authentica-
tion (logging in). Users upload their private data, which is not accessed by server administra- for scientific research, if carried out outside
tors. Once imputation is complete, the results are encrypted and uploaded files are deleted.
Server administrators do not have access to users’ private encryption passwords. Measures of the EU by a non-EU entity, would fall un-
are also in place for encryption of data during storage and transmission. Server administrators
cannot access completed imputation data. Despite the measures and protections in place, EU der GDPR standards. Finally, with respect to
centers are now unable to send their data for imputation to the Michigan Imputation Server, as
they struggle to identify a viable transfer mechanism under the GDPR. COVID-19 research, we recommend that the

for research purposes. The GDPR also allows friendly provisions for the sharing of data for EDPB revisit its Guidance on processing of
for an exception to the notice requirement science beyond the borders of Europe should
when providing notice proves impossible or be part and parcel of the GDPR. The GDPR health data for scientific research, to reaffirm
would involve a disproportionate effort—in legislature has failed to take this crucial as-
particular, processing for scientific research pect of sharing data with scientific collabo- the validity of broad consent and to clarify
purposes. The GDPR further exempts from rators around the globe into account when
the right of erasure personal data processed drafting the research provisions. that the exemption for transfers of research
for scientific research purposes if erasure is
likely to render impossible or seriously im- REFORM AND GUIDANCE data for important reasons of public interest
pair the achievement of the objectives of the We suggest a number of solutions, in the
processing. Furthermore, the GDPR explic- form of GDPR reform per se, dialogue be- is not restricted to time-limited, occasional,
itly provides for an exemption to the right tween the Commission and the EDPB and the
to object when personal data are processed Commission’s global counterparts, or as part and nonrepetitive transfers with respect to
for scientific research purposes, and permits of the Guidance planned by the EDPB on the
member states to enact derogations from processing of health data for the purpose of COVID-19 research.
various data subject rights in the research scientific research. First, we recommend that
context. Notably, all exemptions are subject the GDPR transfer mechanisms be expanded We believe that our recommendations can
to appropriate safeguards for the rights and by adding processing necessary for scientific
freedoms of the data subject, such as techni- research as an express public interest, subject help to redress the unfortunate consequences
cal and organizational measures, including to appropriate safeguards, such as pseudony-
pseudonymization. mization (coding), data protection by design created by the existing GDPR approach to
and default, and the requirements of notice
The common rationale behind these ex- and choice [e.g., (12)]. This basis for global international transfers of research data and
ceptions and exemptions is the notion that sharing of research data should also extend to
scientific research is a “public interest” and onward transfers. Second, it should be clari- will enable the biomedical research commu-
the notion that the GDPR should facilitate fied that pseudonymized data should not be
processing of personal data in the public in- considered personal data in the hands of an nity to share data beyond the EU for scien-
terest. Likewise, the conduct of science has entity that does not possess the key needed to
been acknowledged by EDPB as a legitimate re-identify such data, as was understood by tific research, while ensuring a high level of
interest. Missing, however, from the GDPR many researchers and institutions under the
list of research-friendly provisions is an ap- law preceding the GDPR (1, 13, 14). Third, as protection for data subjects. j
preciation of the international dimensions of part of its ongoing modernization of the stan-
research and, consequently, a corresponding dard contractual clauses, the EU Commission REFERENCES AND NOTES
appropriate provision to enable scientific re- should adopt specific standard contractual
search data transfers across the globe. As the clauses for scientific biomedical research. 1. R. Eiss, Nature 584, 498 (2020).
conduct of science is a global affair, research- 2. CJEU Case C-311/18, 16 July 2020 (“Schrems II”),

specifically paragraphs 104–105, 135–143, and 203.
3. European Commission, Standard Contractual Clauses;

https://ec.europa.eu/info/law/law-topic/data-
protection/international-dimension-data-protection/
standard-contractual-clauses-scc_en.
4. GDPR,Articles 46(2)(a), 46(3)(b); see also EDPB,
Guidelines 2/2020 on Articles 46(2)(a) and 46(3)(b)
of Regulation 2016/679 for Transfers of Personal Data
Between EEA and non-EEA Public Authorities and Bodies
version 1 (18 January 2020).
5. GDPR,Article 46(3).
6. United Kingdom, Information Commissioner’s Office,
Guide to the GDPR: International Transfers; https://ico.
org.uk/for-organisations/guide-to-data-protection/
guide-to-the-general-data-protection-regulation-gdpr/
international-transfers/.
7. GDPR, Recital 42, EDPB 2/2018 Guidelines on
Derogations of Article 49 of Regulation 2016/679
(adopted 25 May 2018).
8. GDPR,Article 49(1)(d).
9. GDPR,Article 49(4).
10. GDPR, Recital 112.
11. EDPB, Guidelines 03/2020 on the Processing of Data
Concerning Health for the Purpose of Scientific Research
in the Context of the COVID-19 Outbreak (21 April 2020).
12. PHG Foundation of the University of Cambridge, The
GDPR and Genomic Data: The Impact of the GDPR and
DPA 2018 on Genomic Healthcare and Research (May
2020); www.phgfoundation.org/documents/gdpr-and-
genomic-data-report.pdf.
13. United Kingdom, Information Commissioner’s Office,
Anonymisation: Managing Data Protection Risk: Code
of Practice (November 2012); https://ico.org.uk/
media/1061/anonymisation-code.pdf.
14. CJEU Case C-582/14 of 19 October 2016 (“Breyer”).

ACKNOWLEDGMENTS

J.B., D.P., and M.B. provide legal counsel to the biomedical
research community, inter alia on issues of data protection
and data transfers. B.M.K. received funding from Genome
Canada/Genome Quebec and under EU-CIHR grant agree-
ments No. 825903 euCanSHare and No. 160202 EUCANCan.

10.1126/science.abd2499

42 2 OCTOBER 2020 • VOL 370 ISSUE 6512 sciencemag.org SCIENCE

Andrew Wakefield addresses the media outside
the General Medical Council on 28 January 2010.

procedures often not indicated by the chil-

dren’s symptoms. “The Royal Free would

become the Mecca, or Lourdes,” he writes,

“for the desperately questing families of de-

velopmentally challenged children.”

Wakefield, we learn, had received £435,643

(the equivalent of $846,000 today) to con-

duct studies that would help build a legal

case against MMR vaccine producers 2 years

prior to the Lancet publication. And although

he reported that the children in his study

were referred to his hospital through routine

channels, many came from an antivaccine

group called JABS and the lawyer preparing

to sue vaccine makers. In June 1997, further

undermining the sentiment the physician

would convey at the 1998 press conference

(“It’s a moral issue for me”), Wakefield sub-

BOOKS et al. mitted a patent for a product that claimed to
treat so-called “autistic enterocolitis,” rid the

body of harmful toxins, and immunize safely

against measles.

PUBLIC HEALTH Deer reveals that Wakefield also misrep-

Flawed research and resented clinical, biological, and molecular
data. Although the 12 children in his paper
were described as having suffered develop-

its enduring repercussions mental problems within 2 weeks of vaccina-
tion, for example, Deer discovered that some
had begun displaying symptoms before re-

A journalist recounts how he exposed problems ceiving the MMR vaccine whereas others did
not begin exhibiting symptoms until months

with a study linking vaccines and autism afterward. Moreover, there were instances in
which normal intestinal biopsy specimens

were mischaracterized as colitis, and Wake-

By Paul A. Offit Wakefield became an international hero. field’s claim that the measles vaccine virus

A biopic starring British actor Hugh Bonn- genome was present in intestinal epithelial

O n 26 February 1998, the Royal Free eville portrayed him as a courageous man cells of children with autistic enterocolitis
Hospital in London held a press willing to speak truth to power. In the was inconsistent and irreproducible.
conference to announce that a study United States, Wakefield testified before the
conducted by one of the hospital’s Congressional Committee on Government As a consequence of these and other rev-
clinicians would be published in Reform and appeared on 60 elations, The Lancet retracted the paper,
The Lancet, one of the world’s oldest Minutes with Ed Bradley. Then
and Wakefield lost his license to
practice medicine. Subsequent

and most prestigious medical journals. Sit- along came Brian Deer, an in- studies have shown that children

ting at the front of the room was the senior vestigative reporter working for who receive the MMR vaccine

author, Andrew Wakefield, who explained The Sunday Times. Deer would are at no greater risk of develop-

that the combination measles-mumps- become the first to expose the mental delays than those who do

rubella (MMR) vaccine could cause develop- clinician’s undisclosed financial not receive it. Nonetheless, the

mental delays, including autism. Wakefield associations and unearth trou- damage was done. The Wake-

argued that the MMR vaccine suppressed bling problems with the Lancet field study helped to accelerate

PHOTO: SHAUN CURRY/AFP/GETTY IMAGES the immune system in some children, free- paper. In The Doctor Who Fooled The Doctor Who the antivaccination movement
ing the measles vaccine virus to damage the the World, Deer recounts in vivid Fooled the World that has imperiled children and
intestine, which allowed encephalopathic detail how he came to learn that led to the resurgence of once-
proteins to enter the circulation, cross the Wakefield and his study were Brian Deer controlled diseases.
Johns Hopkins University

Press, 2020. 408 pp.

blood-brain barrier, and destroy brain cells. not what they appeared to be. Although many people think

He called for MMR vaccinations to cease Deer reveals that children admitted to the they know this now-infamous story, it is

until more research could be conducted. Royal Free Hospital with developmental de- likely they are unaware of all its dramatic de-

lays in the 1990s were subjected to a regime tails. Curious lay readers and vaccine experts

The reviewer is director of the Vaccine Education Center known as the Wakefield protocol, which alike are sure to learn something worthwhile
and an attending physician in the Division of Infectious entailed magnetic resonance imaging, elec- from Deer’s well-chronicled account. j
Diseases at Children’s Hospital of Philadelphia, Philadelphia, troencephalography, spinal taps, abdominal
PA 19104, USA. Email: [email protected] x-rays, blood tests, and intestinal biopsies— 10.1126/science.abb3587

SCIENCE sciencemag.org 2 OCTOBER 2020 • VOL 370 ISSUE 6512 43

INSIGHTS | BOOKS

E D U C AT I O N

Training tomorrow’s scientists

Prioritizing STEM diversity, equity, and inclusion requires rethinking graduate education

By Ashley Huderson how their teaching and mentoring ap- rethink the practices and the qualities we

H ow do you engage with an institu- proaches affect these issues. In Equity in associate with good scientists. The only
tion that was not designed with
women or people of color in mind? Science, Julie Posselt demonstrates how we way to ensure that science is equitable, she
How do you dismantle years of dis-
crimination and unequal treatment? can take steps to mitigate systemic discrim- argues, is to restructure scientific culture
When addressing such questions, I
am reminded of a saying my mother liked ination in STEM education. through a lens that respects and shows
to use. When asked “How do you eat an
elephant?” she would reply, “One bite at Posselt examines existing equity, diver- cultural differences and encourages those
a time.”
sity, and inclusion efforts across a number in nonmarginalized positions to acknowl-
This approach works in many contexts
but is especially applicable to the challenges of graduate STEM fields, including as- edge their power and privilege and the ben-
faced by academic institutions as science,
technology, engineering, and mathemat- tronomy, physics, chemistry, geology, and efits that are conferred to certain groups.
ics (STEM) fields aim to increase diversity
and become more inclusive and equitable. psychology. Her case studies— We must also center the voices,
Although the importance of these elements
to the scientific enterprise is implied, many which include an ethnographic needs, and stories of people from
STEM educators hesitate to participate in
discussions surrounding diversity, equity, assessment of a geology field marginalized groups.
and inclusion because they think they lack
the knowledge and/or training needed to course; a comparative analy- Posselt argues that we must
engage intelligently or because they fear
that their efforts will not result in change. sis of the trajectories of high- design STEM graduate programs
However, as institutions begin to prioritize
diversifying STEM, faculty must consider diversity STEM graduates; and with a diversity, equity, and in-

The reviewer is at the American Society of Mechanical an overview of the barriers, clusion lens that dismantles tra-
Engineers, Washington, DC 20036, USA; the Department
of Biology, University of the District of Columbia, lessons learned, and design of ditional community norms and
Washington, DC 20008, USA; and STEM Innovation
Consulting, Washington, DC 20018, USA. several STEM Ph.D. programs— values, including assumptions
Email: [email protected]
reveal key ways that privilege Equity in Science about scholastic ability, admis-
and power operate in scientific Julie R. Posselt sion requirements, and cur-
organizations and have cre- riculum structure. She lays out
ated a culture of exclusion and Stanford University Press, recommendations for the reten-
2020. 240 pp.

sameness. The book closes with targeted tion and recruitment of traditionally mar-

recommendations for how individuals, de- ginalized groups in graduate programs.

partments, and scholarly societies can cre- These recommendations include down-

ate systemic and sustainable change. playing or eliminating Graduate Record

So how does Posselt suggest we disman- Examination scores as admission criteria;

tle our “elephant”? What does that first bite creating “bridge” programs that create a

look like? She argues that we must revisit clear and intentional pathway through

and reorganize the practices and priori- the STEM pipeline at various critical

ties that have been socialized throughout junctions; providing faculty mentors who

STEM culture, and she encourages readers share students’ identities; and tracking

to reimagine the community’s boundaries program-level data, disaggregated by race

of what constitutes good science and to or ethnicity and gender. She even provides

recommendations for improving equity

and inclusion within existing scientific

collaborations, including ways to manage

impervious and wayward colleagues. Here,

she advises how to assess a collaborator’s

willingness to change and discusses how to

overcome different types of resistance.

Posselt argues that advancing the move-

ment for diversity, equity, and inclusion in

science requires more effective collabora-

tion across boundaries that typically sepa-

rate scholars. She highlights how these

collaborations tend to lie at the intersec-

tion of diverse identities, including gen-

der, race, economic status, and discipline.

Although its conclusions and recommen- PHOTO: ISTOCK.COM/POBA

dations are not exactly novel, the book

succeeds in illustrating the depth to which

diversity, equity, and inclusion are lacking

at every level of STEM culture. j

To increase STEM workforce diversity, we must dismantle norms that discriminate against marginalized groups. 10.1126/science.abd9804

44 2 OCTOBER 2020 • VOL 370 ISSUE 6512 sciencemag.org SCIENCE

PRIZE ESSAY

Scanning confocal fluorescence
micrograph of a coronal section of the

subfornical organ in the mouse
brain, showing thirst neurons (yellow)

and cell nuclei (magenta).

NEUROBIOLOGY absorbed for many minutes (5, 6), and that
eating stimulates prandial drinking long
The origins of thirst before the ingested food enters the blood-
stream (7, 8). How does the brain bridge
Sensory signals arise throughout the body and converge these disparate time scales to dynamically
in the brain to regulate drinking adjust our sense of thirst?

PHOTO: CHRISTOPHER ZIMMERMAN By Christopher A. Zimmerman sor (2, 3). He ultimately discovered a small I reasoned that we might gain new insight
area within the hypothalamus where even into this longstanding question by record-
W e experience thirst every day, minute amounts of salt triggered immedi- ing the activity of thirst-promoting neurons
but where does this sensation ate, voracious drinking. Subsequent studies in living animals. My colleagues and I thus
come from? In the 1950s, Bengt established that Andersson’s osmosensor began by genetically labeling the SFO neu-
Andersson proposed a tantalizing encompasses the subfornical organ (SFO), a rons that comprise Andersson’s osmosensor
answer: Our brains might contain brain region that is distinctively suited to and confirming that these cells are essential
an “osmosensor” (1) that governs detecting blood osmolarity because it lies for dehydration-induced drinking (9). We
thirst, which consists of a group of cells that outside the blood-brain barrier (4). then set out to observe the neural dynamics
sense when we are dehydrated by directly underlying thirst in behaving mice (10, 11).
monitoring the osmolarity of the blood. In The osmosensor model is powerful be-
a series of pioneering experiments, cause it explains how dehydration gener- THIRST NEURONS ARE MORE THAN SIMPLE
Andersson systematically infused DEHYDRATION SENSORS
salt into the brains of goats in an ates thirst, but it has a crucial If SFO neurons are genuine osmosen-
attempt to locate this osmosen- shortcoming: Drinking behavior sors, then we would expect them to sim-
is regulated on a fast, moment-by- ply encode an animal’s dehydration level.
Princeton Neuroscience Institute, moment basis that cannot be ex- Consistent with this idea, our initial fiber
Princeton University, Princeton, NJ, USA. plained by slow changes in blood photometry recordings demonstrated that
Email: [email protected] osmolarity. Consider that drinking these neurons are dose-dependently acti-
immediately satiates thirst, even vated by increases in blood osmolarity (10).
though the water imbibed is not
It was therefore surprising to discover
that SFO neurons are also rapidly regulated

SCIENCE sciencemag.org 2 OCTOBER 2020 • VOL 370 ISSUE 6512 45

INSIGHTS | PRIZE ESSAY

during eating and drinking, well in advance GRAND PRIZE interact? To answer this question, we used PHOTOS: (TOP TO BOTTOM) SAMEER A. KHAN; MELISSA PENLEY CORMIER/RESEARCHGRAPHICS/UMBC; VICKY PAPAVASILEIOU
of any impact food and drink might have WINNER microendoscopic imaging to track the activ-
on the blood (10). For example, their activ- ity of single neurons during dehydration,
ity decreases every time a mouse licks from Christopher drinking, and intragastric infusion (11).
a water bottle and increases with every Zimmerman This revealed a simple processing logic: The
bite of food. This counterintuitive finding signals arising from the mouth, gut, and
indicated that SFO neurons—long viewed Christopher Zimmer- blood converge onto the same individual
as merely passive sensors of dehydration— man received his under- thirst neurons, thereby enabling every cell
must receive a second class of signals that graduate degrees from to continuously integrate information about
operate on the fast time scale of behavior. the University of Pittsburgh and a Ph.D. current hydration status with the predicted
from the University of California, San consequences of ongoing ingestion.
LAYERS OF SIGNALS ARISE FROM THE Francisco. His thesis research focused
DIGESTIVE TRACT DURING INGESTION on the neural mechanisms that govern In a parallel series of experiments, we
To pinpoint the origin of these signals, we thirst and drinking behavior. Zimmerman showed that downstream brain regions use
traced the flow of water through the diges- is currently a postdoctoral fellow at the this integrated representation to coordinate
tive tract of the mouse. We found that fluid Princeton Neuroscience Institute, where the various components of the body’s re-
detection in the mouth triggers a near-in- he continues to study the neural pro- sponse to dehydration, including not only
stantaneous inhibitory signal that closely cesses underlying motivated behaviors. drinking but also cardiovascular adjust-
tracks the volume ingested (10). Tempera- ments, hormone secretion, and changes to
ture sensing contributes to this process— FINALIST emotional valence (11, 14).
SFO neurons are most efficiently inhibited
by drinking cold water, a phenomenon Tara LeGates CONCLUSIONS
that could be reproduced through isolated Thirst is governed by a sensory system,
oral cooling. This may explain why we ex- Tara LeGates received analogous to vision or hearing. Unlike these
perience cold drinks as especially thirst- her B.S. in Biopsycholo- exterosensory systems, however, the neural
quenching and pleasurable (12, 13). gy from Rider University dynamics underlying thirst were previously
and a Ph.D. from Johns unknown. Our recordings revealed that
Using an intragastric infusion paradigm, Hopkins University. She thirst is regulated by layers of signals that
we next discovered that the osmolarity of completed a postdoctoral fellowship arise throughout the body and converge
ingested fluids is precisely measured in the at the University of Maryland School onto individual neurons in the forebrain.
gastrointestinal tract and then rapidly trans- of Medicine, where she established the This convergence occurs at the first node
mitted to the brain by the vagus nerve (11). importance of the strength and plastic- in the thirst system—the SFO—and gener-
This gut-to-brain osmolarity signal sustains ity of hippocampus-nucleus accumbens ates a real-time estimate of the body’s need
the inhibition of SFO neurons produced by synapses and reward behavior. LeGates for water that downstream nodes use to
oral volume signals and satiates thirst if is now an assistant professor at the Uni- dynamically adjust drinking, valence, and
pure water is drunk. By contrast, detection versity of Maryland, Baltimore County cardiovascular physiology (10, 11, 14). Our
of hypertonic fluids in the gut causes SFO (UMBC). Her lab studies how neuronal findings reveal fundamental principles
activity to rebound to the “thirsty” state. circuits integrate information to regulate that govern ingestive behavior (15, 16) and
Thus, drinking generates layers of signals behavior and their alterations in psychiat- provide neural mechanisms that can po-
that enable thirst neurons to predict how ric disorders. www.sciencemag.org/ tentially explain long-enigmatic elements
ingested fluids will affect hydration in the content/370/6512/46.1 of everyday human experience, including
future and then adjust drinking preemp- the speed of thirst satiation, the prevalence
tively. This simple model explains how FINALIST of drinking during meals, and the thirst-
drinking can rapidly quench thirst yet also quenching power of oral cooling. j
be properly calibrated to match an animal’s Riccardo Beltramo
level of dehydration (5, 6). REFERENCES AND NOTES
Riccardo Beltramo re-
Does the body notify the thirst system ceived his undergradu- 1. E. B.Verney, Proc. R. Soc. London Ser. B 135, 25 (1947).
about other behaviors that affect hydration? ate degree from the 2. B.Andersson, Acta Physiol. Scand. 28, 188 (1953).
We found that eating triggers additional sig- University of Turin and 3. B.Andersson, S. M. McCann, Acta Physiol. Scand. 33,
nals that activate SFO neurons in anticipa- a Ph.D. from the Italian
tion of food absorption (10). This activation Institute of Technology. After his doctoral 333 (1955).
drives prandial drinking or, if water is un- training, Beltramo joined the Howard 4. M.J. McKinley et al., The Sensory Circumventricular
available, suppresses further feeding. This Hughes Medical Institute at the Uni-
suggests a neural basis for the widespread versity of California, San Diego and the Organs of the Mammalian Brain (Springer, 2003).
coordination of eating and drinking (7, 8). University of California, San Francisco, 5. R.T. Bellows, Am.J. Physiol. 125, 87 (1938).
where he is completing his postdoctoral 6. B.J. Rolls et al., Am.J. Physiol. Regul. Integr. Comp.
To test the causal role of the body-to- work. He studies sensory perception in
brain signals identified by our record- the mouse visual system, focusing on Physiol. 239, R476 (1980).
ing experiments, we used optogenetics to understanding how cortical and subcor- 7. T.J. Fitzsimons,J. Le Magnen, J. Comp. Physiol. Psychol.
precisely manipulate each of them during tical neural circuits process visual infor-
behavior. This allowed us to confirm that mation to drive behavior. www.sciencemag. 67, 273 (1969).
these signals are necessary for thirst sa- org/content/370/6512/46.2 8. F. Bellisle,J. Le Magnen, Physiol. Behav. 27, 649 (1981).
tiation, prandial thirst, and dehydration- 9. C.A.Zimmerman, D. E. Leib,Z.A. Knight, Nat. Rev.
induced anorexia (10, 11), and thus account SIGNALS CONVERGE ONTO INDIVIDUAL
for most normal drinking behavior. NEURONS TO DYNAMICALLY ADJUST THIRST Neurosci. 18, 459 (2017).
The discovery of diverse inputs to SFO neu- 10. C.A.Zimmerman et al., Nature 537, 680 (2016).
rons raises the fundamental question of 11. C.A.Zimmerman et al., Nature 568, 98 (2019).
how signals are processed by the individual 12. G. Kapatos, R. M. Gold, Science 176, 685 (1972).
cells that comprise the thirst system. Do 13. D. Boulze, P. Montastruc, M. Cabanac, Physiol. Behav.
they flow in segregated “streams” or do they
30, 97 (1983).
14. D. E. Leib et al., Neuron 96, 1272 (2017).
15. B. B. Lowell, N. Engl.J. Med. 380, 459 (2019).
16. C.A.Zimmerman,Z.A. Knight, Curr. Opin. Neurobiol. 64,

79 (2020).

10.1126/science.abe1479

46 2 OCTOBER 2020 • VOL 370 ISSUE 6512 sciencemag.org SCIENCE

Where The Most YWZZQQSF
Science ADVANCED
Patch Clamp CRISPR-enabled
Gets drugs for one?
Social. System
Available Whatever projects you’re working on,
from exploring mechanisms of disease
Today to creating personalized therapies,
you can count on our innovative
Sigma-Aldrich® portfolio to drive
your work forward. With cutting-
edge products and deep technical
expertise in genome engineering,
we’re here to help you make the
next great breakthrough.

Unleash the impossible

To find out more, visit:
SigmaAldrich.com/UnleashTheImpossible

AAAS.ORG/ Visit
COMMUNITY www.dpatch.com

to learn more

AAAS’ Member MilliporeSigma, the vibrant M and Sigma-Aldrich are trademarks
Community is a one-stop of Merck KGaA, Darmstadt, Germany or its affiliates. All other
destination for scientists trademarks are the property of their respective owners.
Detailed information on trademarks is available via publicly
and STEM enthusiasts accessible resources.
alike. It’s “Where
© 2020 Merck KGaA, Darmstadt, Germany and/or its affiliates.
Science Gets Social”: a All Rights Reserved.
community where facts
matter, ideas are big and The Life Science Business of Merck KGaA, Darmstadt, Germany
there’s always a reason operates as MilliporeSigma in the US and Canada.
to come hang out, share,

discuss and explore.

+1.415.883.0128
[email protected]
www.sutter.com

SPECIAL SECTION

ji-xian-sheng

NEURODEGENERATION

A cruel end to too many lives

By Stella Hurtley and Gemma Alderton

N eurodegeneration is an all-too-common disease, but there is still a lot to learn about how to
set of symptoms seen in people as they translate this knowledge into therapies.

reach life’s later stages. Most famously, Disrupted sleep often occurs before or in the

neurodegeneration can lead to memory early stages of neurogenerative disease. An impor-

loss, robbing people of their tant function of sleep is to allow waste

quality of life and their essen- INSIDE products—including misfolded proteins

tial personality. REVIEWS associated with neurodegeneration—to

Much fundamental research Glymphatic failure as a be cleared from the brain via the glym-

has led to a greater understand- fnal common pathway phatic system.
ing of the pathophysiological processes to dementia p. 50 Microglia, the major tissue-resident
involved in neurodegeneration and has
Beyond aggregation: immune cells of the brain, directly modu-
Pathological phase

revealed that accumulation of misfolded, transitions in late neuronal function in development
and in neurodegenerative diseases such as
neurodegenerative

toxic proteins is frequently a cause of disease p. 56

neurodegeneration. This understanding Translating genetic Alzheimer’s and Parkinson’s diseases.

is being refined toward the view that a risk of Alzheimer’s Elucidating the biological foundations
disease into of neurodegenerative disease will be key
common underlying principle of neuro- mechanistic insight

degenerative misfolded proteins involves and drug targets p. 61 to the development of targeted therapies

pathological phase transitions. Microglia modulate to slow or prevent the onset of demen-
Many diferent disorders can cause neurodegeneration tia. In such late-onset diseases, delaying
in Alzheimer’s and symptoms by even a few years can make
neurodegeneration, but by far the most Parkinson’s diseases an enormous diference to the quality of
prevalent is Alzheimer’s disease. Much is
p. 66

already known about the genetic under- RELATED ITEM life of patients, as well as their families
pinnings of susceptibility to Alzheimer’s PERSPECTIVE p. 32 and caregivers.

PHOTO: CREDIT GOES HERE AS SHOWN; CREDIT GOES HERE AS SHOWN

48

ILLUSTRATION: SIMON PRADES Neurodegeneration can stop people from
being able to perform the most basic

tasks in life. However, possibly the most
difcult thing is the gradual loss of

their most precious memories of loved ones.

49

NEURODEGENERATION

REVIEW circadian clock, controls synaptic protein pro-
duction. Under continued wakefulness, proteins
Glymphatic failure as a final common pathway involved in synaptic signaling are continuously
to dementia produced, whereas proteins needed for restor-
ative metabolic processes are not translated.
Maiken Nedergaard1,2* and Steven A. Goldman1,2* Thus, extended wakefulness is associated with
a dysregulation of translation that enables the
Sleep is evolutionarily conserved across all species, and impaired sleep is a common trait of the diseased sustained potentiation of excitatory transmis-
brain. Sleep quality decreases as we age, and disruption of the regular sleep architecture is a frequent sion; this supports a critical homeostatic role
antecedent to the onset of dementia in neurodegenerative diseases. The glymphatic system, which of sleep that cannot occur in the awake state.
clears the brain of protein waste products, is mostly active during sleep. Yet the glymphatic system It is intriguing to speculate that the depth of
degrades with age, suggesting a causal relationship between sleep disturbance and symptomatic recovery sleep, detected as slow-wave activity,
progression in the neurodegenerative dementias. The ties that bind sleep, aging, glymphatic clearance, controls the translation of proteins needed to
and protein aggregation have shed new light on the pathogenesis of a broad range of neurodegenerative restore metabolic homeostasis.
diseases, for which glymphatic failure may constitute a therapeutically targetable final common pathway.
The glymphatic and lymphatic systems
L ittle can replace the rejuvenating feeling fails to fulfill the homeostatic function of sleep.
of a good night’s sleep. Our mood and This prediction has been supported by numer- A fundamental tenet of brain homeostasis is
affect, as well as our ability to attend, ous studies of night-shift workers, who as a that protein clearance must approximate pro-
focus, and problem-solve, are all directly group are predisposed to stress, obesity, cog- tein synthesis. Is removal of protein waste also
linked to how well we sleep. The benefits nitive deficits, and an elevated risk of neuro- controlled by the sleep-wake cycle? Until 2012
of sleep are cumulative; they are not restricted degenerative diseases (10–13). One of the most it was believed that the brain, singular among
to the morning hours or even to a given day. prominent current models of sleep posits that organs, was recycling all of its own protein
Good sleepers live longer, weigh less, have a the purpose of sleep is to restore synaptic ho- waste (21). Only a small number of proteins
reduced incidence of psychiatric disorders, meostasis (14). The synaptic homeostasis hy- were known to be transported across the blood-
and remain cognitively intact longer (1–4). pothesis of sleep is based on the observations brain barrier, and these did not include most of
that wakefulness is associated with the sus- the primary proteins made or shed by brain
Why do we sleep? tained potentiation of excitatory transmission, cells (22). In the absence of lymphatic vessels or
as well as with the structural expansion of post- any overt pathways for fluid export, it was un-
The idea that our brains rest during sleep to synaptic dendritic spines (15, 16). The larger size clear how protein waste might exit the mature
preserve energy was both posited and rejected of spines during wakefulness increases their brain parenchyma. The default conclusion was
in the 1950s, when electroencephalographic postsynaptic currents and thereby strengthens that the classical cellular protein degradation
(EEG) recordings of brain activity made it clear excitatory transmission. This model is sup- pathways—autophagy and ubiquitination—
that rapid eye movement (REM) sleep, which ported by the observation that sleep depriva- must be responsible for all central nervous sys-
comprises ~20% of normal sleep, is linked to tion is linked to an increased risk of seizures in tem (CNS) protein recycling (23).
cortex-wide neuronal activation (5, 6). Indeed, predisposed individuals (17). It is only during
energy consumption declines by only 15% in subsequent recovery sleep that excitatory trans- This supposition, that the brain must re-
the remaining non-REM (NREM) periods of mission tone and spine volume fall, each re- cycle its own waste, was questioned after the
sleep. Borbély proposed 40 years ago that the turning to its sleep-associated baseline (18). discovery of the glymphatic system (24). The
sleep-wake cycle is determined by the interac- glymphatic system is a highly organized cere-
tion of two processes: a circadian oscillator, Recent studies in mice have offered molec- brospinal fluid (CSF) transport system that
which cycles with the solar day, and a homeo- ular insights into the synaptic homeostasis shares several key functions, including the
static drive for sleep (7). A key element in that hypothesis by mapping the impact of the sleep- export of excess interstitial fluid and proteins,
model is that a sleep deficit (i.e., sleep depri- wake cycle on synaptic gene expression (19, 20). with the lymphatic vessels of peripheral tis-
vation) causes a quantifiable “pressure to go to These studies showed that genes involved in sues (Fig. 1A). Indeed, both the brain’s CSF and
sleep.” Subsequent NREM sleep is both longer synaptic signaling were predominantly tran- peripheral lymph are drained together into
and deeper than normal, and the antecedent scribed before the mice woke up, whereas tran- the venous system, from which protein waste
sleep loss can be identified post hoc by an scripts of genes involved in metabolism rose a is removed and recycled by the liver (25). Yet
increase in EEG slow-wave activity during few hours before their expected bedtime. Thus, brain tissue itself lacks histologically distinct
recovery sleep (8). Slow-wave activity is char- the circadian clock dictates the transcription lymphatic vessels. Rather, fluid clearance from
acterized by a wave of synchronous local neu- of genes in anticipation of the tasks appropri- the brain proceeds via the glymphatic pathway,
ral firing that typically begins in the frontal ate for the time of day. Similarly, translation of a structurally distinct system of fluid transport
cortex and propagates posteriorly, occurring mRNAs into proteins largely followed tran- that uses the perivascular spaces created by the
roughly every second during NREM sleep (9). scription, so that proteins involved in synaptic vascular endfeet of astrocytes (26). The endfeet
One of the predictions of the Borbély model is signaling were produced during wakefulness, surround arteries, capillaries, and veins, serving
that daytime sleep is lighter, because it is not whereas those with a role in metabolism were as a second wall that covers the entire cerebral
aligned with the circadian clock, and hence translated during sleep. Surprisingly, when vascular bed. The perivascular spaces are open,
the mice were kept awake longer than normal, fluid-filled tunnels that offer little resistance
1Center for Translational Neuromedicine, Faculty of Health the translation of proteins involved in synaptic to flow. This is in sharp contrast to the dis-
and Medical Sciences, University of Copenhagen, 2200 signaling continued during sleep deprivation, orientingly crowded and compact architecture
Copenhagen, Denmark. 2Center for Translational concurrently with suppressed production of of adult brain tissue, the neuropil, through
Neuromedicine, University of Rochester Medical Center, proteins associated with metabolism (19, 20). which interstitial fluid flow is necessarily
Rochester, NY 14642, USA. Thus, the behavioral state, rather than the slow and restricted—akin to a marsh, flow-
*Corresponding author. Email: [email protected] (M.N.); ing to the glymphatic system’s creeks and
[email protected] (S.A.G.) then rivers (27). The glymphatic system’s peri-
vascular tunnels are directly connected to the

50 2 OCTOBER 2020 • VOL 370 ISSUE 6512 sciencemag.org SCIENCE