Molecular Biology
Conservation of Regulation of G1-to-S transcription in eukaryotes.
Eukaryotes from yeast to humans coordinately express a family
of genes during the G1 phase in preparation for traversing the
cell cycle. Although these transcriptional regulators in yeast
and humans have no primary sequence homology, their function is
conserved. The transcriptional repressor Whi5 and the corepressor
Nrm1 in yeast perform functions analogous to the mammalian
pocket proteins Rb and p107, respectively. Nrm1, which
represses transcription as cells enter S phase, is a target for
the DNA replication checkpoint, and p107 may be similarly
regulated to ensure genome integrity. Work done by Rob de
Bruin, Ph.D., research associate, in the laboratory of Curt
Wittenberg, Ph.D., professor.
Paul R. Schimmel, Ph.D.,
Ernest and Jean Hahn Professor
of Molecular Biology and
Chemistry, and Wei Zhang,
Ph.D., Research Asssociate
MOLECULAR BIOLOGY
DEPAR TMENT OF
MOLECULAR BIOLOGY
2007
THE SCRIPPS RESEARCH INSTITUTE
Lluis Ribas De Pouplana,
Ph.D.
Adjunct Assistant Professor
Barcelona Science Park
Barcelona, Spain
Ehud Keinan, Ph.D.
Adjunct Professor
Technion-Israel Institute of
Technology
Haifa, Israel
Ashok Deniz, Ph.D.
Associate Professor
Richard A. Lerner, M.D.,
Ph.D.**
President, Scripps Research
Lita Annenberg Hazen
Professor of
Immunochemistry
Cecil H. and Ida M. Green
Chair in Chemistry
S TA F F
Peter E. Wright, Ph.D.*
Professor and Chairman
Cecil H. and Ida M. Green
Investigator in Medical
Research
Ruben Abagyan, Ph.D.
Professor
Carlos F. Barbas III, Ph.D.**
Professor
Janet and W. Keith Kellogg II
Chair, Molecular Biology
Rajesh Belani, Ph.D.
Adjunct Assistant Professor
Scripps Mercy Hospital
San Diego, California
Ola Blixt, Ph.D.
Assistant Professor of
Molecular Biology
Michael N. Boddy, Ph.D.
Assistant Professor
Charles L. Brooks III, Ph.D.
Professor
David A. Case, Ph.D.
Professor
Geoffrey Chang, Ph.D.*
Associate Professor
Eli Chapman, Ph.D.
Assistant Professor of
Molecular Biology
Jerold Chun, M.D., Ph.D.
Professor
Aymeric Pierre De Parseval,
Ph.D.***
Assistant Professor of
Molecular Biology
Public Health Research
Institute
Newark, New Jersey
H. Jane Dyson, Ph.D.
Professor
John H. Elder, Ph.D.****
Professor
Martha J. Fedor, Ph.D.*
Associate Professor
James A. Fee, Ph.D.
Professor of Research
Scott Lesley, Ph.D.
Assistant Professor of
Biochemistry
Tianwei Lin, Ph.D.
Assistant Professor
Elizabeth D. Getzoff,
Ph.D.*****
Professor
Ian MacRae, Ph.D.
Assistant Professor
David B. Goodin, Ph.D.
Associate Professor
Clare McGowan, Ph.D. †
Associate Professor
David S. Goodsell Jr., Ph.D.
Associate Professor of
Molecular Biology
Duncan E. McRee, Ph.D.
Adjunct Associate Professor
ActiveSight
San Diego, California
Joel M. Gottesfeld, Ph.D.
Professor
Jennifer Harris, Ph.D.
Assistant Professor of
Biochemistry
Christian A. Hassig, Ph.D.
Adjunct Assistant Professor
Kalypsis, Inc.
San Diego, California
Peter B. Hedlund, M.D.,
Ph.D.
Assistant Professor of
Molecular Biology
John E. Johnson, Ph.D.
Professor
Gerald F. Joyce, M.D.,
Ph.D.**
Professor
Dean, Faculty
David P. Millar, Ph.D.
Associate Professor
Louis Noodleman, Ph.D.
Associate Professor
183
Harold Scheraga, Ph.D.
Adjunct Professor
George W. and Grace L. Todd
Professor of Chemistry,
Emeritus
Cornell University
Ithaca, New York
Paul R. Schimmel, Ph.D.**
Ernest and Jean Hahn
Professor of Molecular
Biology and Chemistry
Anette Schneemann, Ph.D.
Associate Professor
Subhash C. Sinha, Ph.D.*
Associate Professor of
Medicinal Chemistry
Gary Siuzdak, Ph.D.
Adjunct Associate Professor
Director, Center for Mass
Spectrometry
Vaughn V. Smider, Ph.D.
Assistant Professor of
Molecular Biology
Robyn L. Stanfield, Ph.D.
Assistant Professor
James Stevens, Ph.D.***
Assistant Professor
Centers for Disease Control
and Prevention
Atlanta, Georgia
Arthur J. Olson, Ph.D.
Professor
Raymond C. Stevens,
Ph.D. †††
Professor
James C. Paulson, Ph.D. ††
Professor
Charles D. Stout, Ph.D.
Associate Professor
Vijay Reddy, Ph.D.
Associate Professor
Peiqing Sun, Ph.D.
Associate Professor
Steven I. Reed, Ph.D. †
Professor
J. Gregor Sutcliffe, Ph.D.
Professor
Paul Russell, Ph.D. †
Professor
John A. Tainer, Ph.D.*
Professor
Michel Sanner, Ph.D.
Associate Professor
Fujie Tanaka, Ph.D.
Associate Professor
184 MOLECULAR BIOLOGY
Elizabeth A. Thomas, Ph.D.
Assistant Professor
James R. Williamson,
Ph.D.*****
Professor
Associate Dean, Kellogg
School of Science and
Technology
2007
Peter Sobieszcsuk, Ph.D.††††
Core Manager, Consortium
for Functional Glycomics
THE SCRIPPS RESEARCH INSTITUTE
Kelly Lee, Ph.D.
Joseph W. Arndt, Ph.D.***
Biogen Idec
Cambridge, Massachusetts
Richard R. Rivera, Ph.D.
Karunesh Arora, Ph.D.
Kenji Sugase, Ph.D.***
Suntory Institute for
Bioorganic Research
Osaka, Japan
Mabelle Ashe, Ph.D. ††††
Philip Arno Venter, Ph.D.
Jamie Mitchell Bacher,
Ph.D.***
Rincon Pharmaceuticals
La Jolla, California
Julio Kovacs, Ph.D.
S E N I O R S TA F F
SCIENTIST
Quansheng Zhou, Ph.D.
Ian A. Wilson, D.Phil.*
Professor
S TA F F S C I E N T I S T S
Curt Wittenberg, Ph.D. †
Professor
Kurt Wüthrich, Ph.D.
Cecil H. and Ida M. Green
Professor of Structural
Biology
Svitlana Berezhna, Ph.D.
Reto Horst, Ph.D.
Sung-Hun Bae, Ph.D.
Ying Chuan Lin, Ph.D.
Maria Martinez-Yamout, Ph.D.
Xiang-Lei Yang, Ph.D.
Assistant Professor of
Molecular Biology
Xiaoqin Ye, M.D., Ph.D.***
University of Georgia
Athens, Georgia
Wojciech Augustyniak,
Ph.D.
Garrett M. Morris, Ph.D.
Chiaki Nishimura, Ph.D.
Todd O. Yeates, Ph.D.
Adjunct Professor
University of California
Los Angeles, California
Jeffrey Speir, Ph.D.
Qinghai Zhang, Ph.D.
Assistant Professor
Xueyong Zhu, Ph.D.
Mutsuo Yamaguchi, Ph.D.
R E S E A R C H A S S O C I AT E S
Melanie Ann Adams, Ph.D.
Fabio Agnelli, Ph.D.
Hanna-Stina Martinsson
Ahlzén, Ph.D.
Reetesh Raj Akhouri, Ph.D.
Alexander Ivanov Alexandrov,
Ph.D.
Stephen G. Aller, Ph.D.
SENIOR RESEARCH
Manidipa Banerjee, Ph.D.
Konstantinos Beis, Ph.D.***
Imperial College London
London, England
Christine Beuck, Ph.D.
David Boehr, Ph.D.
Yannick Bomble, Ph.D.
David Bostick, Ph.D.
Giovanni Bottegoni, Ph.D.
SERVICE FACILITIES
Beatriz Gonzalez Alonso,
Ph.D.
A S S O C I AT E S
John Chung, Ph.D.***
Manager, Nuclear Magnetic
Resonance Facilities
Juniper Networks, Inc.
Sunnyvale, California
Gerard Kroon
Manager, Nuclear Magnetic
Resonance Facilities
David Barondeau, Ph.D.***
Texas A&M University
College Station, Texas
Juliano Alves, Ph.D.***
Genomics Institute of the
Novartis Research
Foundation
San Diego, California
Tommy Bui, Ph.D.
Yu An, Ph.D.**
Millipore
Temecula, California
Justin E. Carlson, Ph.D. ††††
Brigitte Anliker, Ph.D.***
University of Düsseldorf
Düsseldorf, Germany
Qing Chai, M.D., Ph.D.***
Applied Molecular Evolution
San Diego, California
Andrew James Annalora,
Ph.D.
Ramya Thirumalai
Chakravarthy, Ph.D.
Maria Alejandra GamezAbascal, Ph.D.
Munehito Arai, Ph.D.
Amarnath Chatterjee, Ph.D.
Elsa Garcin, Ph.D.
Roger Armen, Ph.D.
Susana Chaves, Ph.D.
Kirk Beebe, Ph.D.***
Metabolon, Inc.
Durham, North Carolina
Ryan Burnett, Ph.D.
Bo Ma, Ph.D.
Core Manager, Consortium
for Functional Glycomics
Michael E. Pique
Director, Computer Graphics
Development
Nahid Razi, Ph.D.
Assistant Core Manager,
Consortium for Functional
Glycomics
Lintao Bu, Ph.D.
Zhongguo Chen,
Paul Card, Ph.D. ††††
Rosa Maria Cardoso, Ph.D.
Andrew Barry Carmel, Ph.D.
Ph.D. ††††
Adrienne Elizabeth Dubin,
Ph.D.
Li Fan, Ph.D.
MOLECULAR BIOLOGY
Jianhan Chen, Ph.D.***
Kansas State University
Manhattan, Kansas
Weihsu Claire Chen, Ph.D.
Yen-Ju Chen, Ph.D.
Ying Chen, M.D.
Zhiyong Chen, Ph.D.***
Trius Therapeutics, Inc.
San Deigo, California
Srinivas Chittaboina, Ph.D.
2007
Claire Louise Dovey, Ph.D.
Zhanna Druzina, Ph.D.***
SGX Pharmaceuticals, Inc.
San Diego, California
Li-Lin Du, Ph.D.***
National Institute of
Biological Sciences
Beijing, China
Michelle Duquette-Huber,
Ph.D.
THE SCRIPPS RESEARCH INSTITUTE
Shoufa Han, Ph.D.***
Xiamen University
Xiamen, China
Rodney Harris, Ph.D.
Ilja V. Khavrutskii, Ph.D.***
University of California
San Diego, California
Joreg Hinnerwisch, Ph.D.***
ALTANA Chemie AG
Wesel, Germany
Reza Khayat, Ph.D.
Dae Hee Kim, Ph.D.
Stephen Edgcomb, Ph.D.
Wen-Xu Hong, Ph.D.
Chung Jen Chou, Ph.D.
Susanna V. Ekholm-Reed,
Ph.D.
Li-Chiou Chuang, Ph.D.
Daniel Felitsky, Ph.D.
Yunfeng Hu, Ph.D.***
Tanabe Research Laboratories
U.S.A., Inc.
San Diego, California
Linda Maria Columbus,
Ph.D.***
University of Virginia
Charlottesville, Virginia
Allan Chris Merrera Ferreon,
Ph.D.
Gladys Completo, Ph.D.
Yann Gambin, Ph.D.
Stephen Connelly, Ph.D.
Shannon E. Gardell,
Ph.D. ††††
Josephine Chu Ferreon, Ph.D.
Adam Corper, Ph.D.
Charles Gersbach, Ph.D.
Carla P. Da Costa, Ph.D. ††††
Surya Kanta De, Ph.D.***
Burnham Institute for Medical
Research
La Jolla, California
Jennifer Knight, Ph.D.
Minsun Hong, Ph.D.
Eda Koculi, Ph.D.
Zheng-Zheng Huang, Ph.D.
Kwan Hoon Hyun, Ph.D.***
CrystalGenomics
Seoul, Korea
Masanori Imai, Ph.D.***
Hokkaido Pharmaceutical
University
Hokkaido, Japan
Irina Kufareva, Ph.D.
Shantanu Kumar, Ph.D.***
University of California
San Diego, California
Sharon Kwan, Ph.D.
Bianca Lam, Ph.D.
Emma Langley, Ph.D.
Joshua Gill, Ph.D.
Thamara Janaratne, Ph.D.
Chang-Wook Lee, Ph.D.
Edith Caroline Glazer, Ph.D.
Kai Jenssen, Ph.D.***
Merz Pharma
Frankfurt, Germany
Chul Won Lee, Ph.D.
Rajib Kumar Goswami, Ph.D.
Bettina Groschel, Ph.D.
Margaret Alice Johnson,
Ph.D.
Hai-Ming Guo, Ph.D.
June Hyung Lee, Ph.D.***
New York University
New York, New York
Sophie Lefebvre, Ph.D.
Steven Johnson, Ph.D.
Edward Lemke, Ph.D.
Susanna Juraja, Ph.D.
Robert de Bruin, Ph.D.
Mahender Gurram, Ph.D.
Gerald Edward Dodson,
Ph.D.
Milka Kostic, Ph.D.***
Cell Press
Cambridge, Massachussetts
Jason Lanman, Ph.D.
Min Guo, Ph.D.
Paula Desplats, Ph.D.***
University of California
San Diego, California
Bethany Koehntop, Ph.D.
Veli-Pekka Jaakola, Ph.D.
Sandro Cosconati, Ph.D.
Qizhi Cui, Ph.D.
National Research Council
Ottawa, Ontario, Canada
Yang Khandogin, Ph.D.***
University of Oklahoma
Norman, Oklahoma
Deron Herr, Ph.D.
Scott Eberhardy, Ph.D.***
Schering-Plough Corporation
Union, New Jersey
Ji Woong Choi, Ph.D.
Andrey Aleksandrovich
Karyakin, Ph.D.
David M. Herman, Ph.D.
Kenichi Hitomi, Ph.D.
Jungwoo Choe, Ph.D.***
University of Seoul
Seoul, Korea
185
Liao Liang, Ph.D.
Bong Kwan Han, Ph.D.***
Cornell University
Ithaca, New York
Christian Kannemeier,
Ph.D.***
PrimeCell Therapeutics L.L.C.
Irvine, California
Byung Woo Han, Ph.D.
Mili Kapoor, Ph.D.
Vasco Liberal, Ph.D.
William M. Lindstrom, Ph.D.
Bin Liu, Ph.D.
186 MOLECULAR BIOLOGY
Kunheng Luo, Ph.D.
Ann MacLaren, Ph.D.***
Biogen Idec
San Diego, California
2007
THE SCRIPPS RESEARCH INSTITUTE
Wataru Nomura, Ph.D.***
Tokyo Medical and Dental
University
Tokyo, Japan
Sanjay Adrian Saldanha,
Ph.D.***
Translational Research
Institute, Scripps Florida
Rebecca E. Taurog, Ph.D.
Ewan Richardson Taylor,
Ph.D.
Amy Odegard, Ph.D.
Mariana Santa-Marta, Ph.D.
Leonardo Teixeira, Ph.D.
Darly Joseph Manayani,
Ph.D.
Lisa Renee Olano, Ph.D.
Andre Schiefner, Ph.D.
Hua Tian, Ph.D.
Jeff Mandell, Ph.D.***
Ionian Technologies, Inc.
San Diego, California
Brian L. Olson, Ph.D.***
St. Cloud University
St. Cloud, Minnesota
Lauren J. Schwimmer, Ph.D.
Mauricio Carrillo Tripp, Ph.D.
Jennifer S. Scorah, Ph.D.
Oleg Trott, Ph.D
Santiago Cavero Martinez,
Ph.D.
Akira Onoda, Ph.D.
Alim Seit-Nebi, Ph.D.
Mary O’Reilly, Ph.D.
Pedro Serrano-Navarro, Ph.D.
Ulrich Tschulena, Ph.D.***
German Cancer Research
Center
Heidelberg, Germany
Brian Paegel, Ph.D.
Craig McLean Shepherd,
Ph.D.***
Burnham Institute for Medical
Research
La Jolla, California
Tsutomu Matsui, Ph.D.
Derrick Meinhold, Ph.D.
Eva Mejia Ramirez De
Arellano, Ph.D.
Elena Menichelli, Ph.D.
Jonathan Mikolosko, Ph.D.
Mauro Mileni, Ph.D.
Maki Minakawa, Ph.D.
Marissa Mock, Ph.D.***
Biocatalytics
Pasadena, California
Biswaranjan Mohanty, Ph.D.
Samrat Mukhopadhyay, Ph.D.
Tetsuji Mutoh, Ph.D.
Sujatha Narayan, Ph.D. ††††
Mir Hussain Nawaz, Ph.D.
Hung Nguyen, Ph.D.
Tuan Nguyen, Ph.D.
Sung-Jean Park, Ph.D.
Sandeep Patel, Ph.D.***
University of Delaware
Newark, Delaware
Stephanie Pebernard, Ph.D.
Bill Francesco Pedrini, Ph.D.
Robert Pejchal, Ph.D.
Vladimir Pelmenschikov,
Ph.D.
Jefferson Perry, Ph.D.
Suzanne Peterson, Ph.D.
Jessica Petrillo, Ph.D.
David S. Shin, Ph.D.
Develeena Shivakumar,
Ph.D.***
University of Chicago
Chicago, Illinois
David A. Shore, Ph.D.
Daniela Andrea Slavin, Ph.D.
Elisabetta Soragni, Ph.D.
Kyoko Noguchi, M.D., Ph.D.
Sorin Tunaru, Ph.D.
Naoto Utsumi, Ph.D.
Frank van Drogen, Ph.D.***
Institute für Biochemie
Zürich, Switzerland
Ajay Vashisht, Ph.D.
Sangita Venkataraman, Ph.D.
Petra Verdino, Ph.D.
My Vo, Ph.D.
Pawel Stanczak, Ph.D.
William Frederick Waas,
Ph.D.***
Defined Health
Florham Park, New Jersey
Thomas Steinbrecher, Ph.D.
Shun-ichi Wada, Ph.D. ††††
Shih-Che Su, Ph.D.
Jun Wang, Ph.D.
Sebastian Sudek, Ph.D.
Jessica Williams, Ph.D.
Magnus Sundstrom, Ph.D.
Robert Scott Williams, Ph.D.
Michael T. Sykes, Ph.D.
Eric L. Wise, Ph.D.***
Wayne State University
Detroit, Michigan
Surya Venkata Sripada, Ph.D.
Goran Pljevaljc̆ić, Ph.D.
Stephanie Pond, Ph.D.
John Prudden, Ph.D.
Christopher L. Reyes,
Ph.D.***
Biogen Idec
San Diego, California
George Nicola, Ph.D.
Tadateru Nishikawa,
Ph.D.***
Bruker BioSpin
Tsukuba, Japan
Julie L. Tubbs, Ph.D.
Kimberly A. Reynolds, Ph.D.
Blair R. Szymczyna, Ph.D.
Jin-Kyu Rhee, Ph.D.
Hiroaki Tateno, Ph.D.***
National Institute of
Advanced Industrial
Sceince and Technology
Tsukuba, Japan
Gabriela Ring, Ph.D.
Riturparna Sinha Roy, Ph.D.
Jonathan Wojciak, Ph.D.***
Lpath, Inc.
San Diego, California
Vance Wong, Ph.D.
MOLECULAR BIOLOGY
Timothy I. Wood, Ph.D.***
Walter Reed Army Medical
Center
Washington, D.C.
2007
Ognian V. Bohorov, Ph.D.
Dennis Carlton, B.S.
Vadim Cherezov, Ph.D.
Wei Xie, Ph.D.
Ellen Yu-Lin Tsai Chien, Ph.D.
Lan Xu, Ph.D.***
BioVerdant
San Diego, California
Xiaoping Dai, Ph.D.
Marc Deller, D.Phil
Yoshiki Yamada, Ph.D.
THE SCRIPPS RESEARCH INSTITUTE
Joseph David Ng, Ph.D.
University of Alabama
Huntsville, Alabama
Victoria A. Roberts, Ph.D.
University of California
San Diego, California
Robert D. Rosenstein, Ph.D.
Lawrence Berkeley National
Laboratory
Berkeley, California
Gye Won Han, Ph.D.
Tohru Yamagaki, Ph.D.
Wenge Han, Ph.D.
Atsushi Yamagata, Ph.D.***
Tokyo University
Tokyo, Japan
Michael Allen Hanson, Ph.D.
Debbie Tahmassebi, Ph.D.
University of San Diego
San Diego, California
Hope Johnson, Ph.D.
Yongjun Ye, Ph.D.***
VIRxSYS Corporation
Gaithersburg, Maryland
Marcy A. Kingsbury, Ph.D.***
Indiana University
Bloomington, Indiana
Kye Sook Yi, Ph.D.
Padmaja Natarajan, Ph.D.††††
Yong Yin,
Ph.D. ††††
Lin Wang, Ph.D.
Sung-Il Yoon, Ph.D.
Kenji Yoshimoto, Ph.D.
Naoto Yoshizuka, M.D., Ph.D.
Yuan Yuan, Ph.D.***
Millipore
Temecula, California
Markus Zeeb, Ph.D.***
Boehringer Ingelheim GmbH
Biberach an der Riss,
Germany
VISITING
I N V E S T I G AT O R S
Stephen J. Benkovic, Ph.D.
Pennsylvania State University
University Park, Pennsylvania
Wayne A. Fenton, Ph.D.
Yale University
New Haven, Connecticut
Astrid Graslund, Ph.D.
Stockholm University
Stockholm, Sweden
Arne Holmgren, M.D., Ph.D.
Karolinska Institutet
Stockholm, Sweden
Qing Zhang, Ph.D.
Wei Zhang, Ph.D.
Barry Honig, Ph.D.
Columbia University
New York, New York
Yong Zhao, Ph.D.
S C I E N T I F I C A S S O C I AT E S
Enrique Abola, Ph.D.
Andrew S. Arvai, M.S.
** Joint appointments in the
Department of Chemistry and
The Skaggs Institute for
Chemical Biology
*** Appointment completed; new
location shown
**** Joint appointment in the
Department of Molecular and
Integrative Neurosciences
* **** Joint appointments in the
Department of Immunology and
The Skaggs Institute for
Chemical Biology
†
††
Ying Zeng, Ph.D.
Haile Zhang, Ph.D.
* Joint appointment in The
Skaggs Institute for Chemical
Biology
Arthur Horwich, M.D.
Yale University
New Haven, Connecticut
Tai-Huang Huang, Ph.D.
Academica Sinica
Taipei, Taiwan
Joint appointment in the
Department of Cell Biology
Joint appointment in the
Department of Molecular and
Experimental Medicine
†††
Joint appointment in the
Department of Chemistry
††††
Appointment completed
187
188 MOLECULAR BIOLOGY
2007
Peter E. Wright, Ph.D.
Chairman’s Overview
esearch in the Department of Molecular Biology
encompasses a broad range of disciplines, extending from structural and computational biology at
one extreme to molecular genetics at the other. During
the past year, our scientists have continued to make rapid
progress toward understanding the fundamental molecular events that underlie the processes of life. Major
advances have been made in elucidating the structural
biology of signal transduction, receptor recognition, and
viral assembly; understanding mechanisms of viral infectivity; determining the structures of membrane proteins
and multidrug transporters; understanding the molecular basis of nucleic acid recognition and DNA repair;
and determining the mechanisms of protein folding and
ribosome assembly.
Progress has been made in elucidating the molecular events involved in regulation of the cell cycle, tumor
development, induction of sleep, the molecular origins
of neuronal development and CNS disorders, the regulation of transcription, and decoding of genetic information in translation. Finally, new advances have been made
in the design of novel low molecular weight compounds
that can specifically regulate genes and in biomolecular
engineering, building novel functions into viruses, antibodies, zinc finger proteins, RNA, and DNA. Progress in
these and other areas is described in detail on the fol-
R
THE SCRIPPS RESEARCH INSTITUTE
lowing pages, and only a few highlights are mentioned
here. The Department of Molecular Biology is also home
to 2 major National Institutes of Health initiatives: the
Joint Center for Structural Genomics and the Consortium
for Functional Glycomics.
Research in the laboratory of Raymond Stevens
has revealed the structural basis for the high-affinity
interaction of botulinum toxin with nerve cell receptors. Botulinum toxin is a potent bacterial toxin that
causes paralysis by binding at neuromuscular junctions at femtomolar concentrations to block neurotransmitter release. Dr. Stevens and coworkers determined
the x-ray structure of type B botulinum toxin bound to
the recognition motif of synaptotagmin II, an integral
membrane protein found in synaptic vesicles. In addition to binding synaptotagmin, the botulinum toxin binds
simultaneously to ganglioside molecules in the presynaptic membrane to form a very high-affinity multivalent
complex in which the toxin is oriented optimally to
facilitate the formation of pores. The molecular insights
gained from this structural analysis may form the basis
for designing new small-molecule therapeutic agents
to treat botulism or for generating novel cross-reactive
antibodies that can bind with high affinity to several
botulinum toxin subtypes. Dr. Stevens and his colleagues
have generated such cross-reactive antibodies and have
determined structures of the antibodies complexed with
toxin to develop new strategies for evolving antibodies
with broadened specificity.
New advances were made in understanding the
behavior of natively disordered proteins. It is now recognized that many eukaryotic proteins contain long unstructured regions or are entirely unfolded in their native state.
Ashok Deniz and coworkers applied single-molecule
fluorescence and fluorescence correlation methods to
characterize the prion protein Sup35, which functions
as a translation termination factor in yeast. Sup35 can
switch to a self-replicating amyloid state under physiologic conditions, a process analogous to formation of
amyloid fibrils by misfolded human proteins in many
neurodegenerative diseases. Dr. Deniz and coworkers
showed that the monomeric collapsed state of Sup35
consists of an ensemble of rapidly fluctuating conformations. Assembly of the monomers into oligomeric intermediates is an essential first step for nucleating the
formation of amyloid fibers. The novel fluorescence
techniques developed by scientists in the Deniz laboratory hold exceptional promise for characterizing the
structure and dynamics of the toxic intermediates formed
MOLECULAR BIOLOGY
2007
during the aggregation processes that lead to Alzheimer’s
and other neurodegenerative diseases.
In my laboratory, Jane Dyson and I provided the first
insights into the mechanism by which a natively disordered transcription activation domain recognizes its
target and folds into a compact structure upon binding.
Recent work by Arthur Horwich and colleagues has
provided important insights into the topology of substrate proteins folding within the cavity of the chaperonin GroEL-GroES.
Recent research by Peiqing Sun and coworkers has
provided novel insights into a signaling pathway that
mediates cellular senescence and, when activated,
inhibits tumor development. These studies have indicated that the p38-regulated/activated protein kinase
(PRAK) is an essential component of the senescence
pathway. Inactivation of PRAK in normal human cells
prevents senescence and makes them more susceptible
to oncogenic transformation. Mice lacking the PRAK
gene have increased susceptibility to skin cancer and
have accelerated development of lymphoma. Mechanistically, PRAK appears to function by phosphorylating
and activating the tumor suppressor p53. These studies
suggest a potential new approach to cancer therapy:
using molecules that can activate the PRAK pathway.
Carlos Barbas, Subhash Sinha, Richard Lerner, and
colleagues have reported a major advance in the synthesis of therapeutic antibodies. These scientists have
developed catalytic antibodies that effectively target
integrin αvβ3 expressed by human breast cancer cells
and catalyze the activation of doxorubicin prodrugs.
These studies lay the basis for development of a new
class of therapeutic antibodies with both targeting and
drug-activating functions.
Despite the advent of high-throughput screening and
library technologies, the success rate for generation of
novel drugs from new chemical entities is depressingly
low. Ruben Abagyan and coworkers have developed a
novel computational method for “drug repurposing.” With
this method, they optimize leads to side-effect targets of
marketed drugs while eliminating activity at the original
target and maintaining acceptable bioavailability and
toxicity profiles. Because these repurposed compounds
are derived from safe and established drugs, they are
likely to enter clinical trials at lower cost than would
lead compounds based entirely on new chemical entities.
Molecular biology remains a field of enormous opportunity and excitement. The scientists in this department
are taking full advantage of powerful new technologies
THE SCRIPPS RESEARCH INSTITUTE
189
to advance our understanding of fundamental biological
processes at the molecular level. Their discoveries will
ultimately be translated into new advances in biotechnology and in medicine.
190 MOLECULAR BIOLOGY
2007
INVESTIGATORS’ R EPORTS
Structural Biology of Viral
Proteins, Molecular Assemblies,
and the Immune System
THE SCRIPPS RESEARCH INSTITUTE
Z13 bind to conserved and overlapping epitopes on
the membrane-proximal region on the viral envelope
protein gp41. We have determined multiple 4E10
structures in complex with peptide epitopes and have
tested the peptides for immunogenicity. In addition,
we have grafted this epitope into the gp120 of HIV-1
(Fig. 1) to investigate whether these engineered antigens can elicit HIV-neutralizing antibodies.
I.A. Wilson, R.L. Stanfield, J. Stevens, X. Zhu, M.A. Adams,
Y. An, K. Beis, C.H. Bell, D.A. Calarese, R.M.F. Cardoso,
J. Carlson, P.J. Carney, J.-W. Choe, S. Connelly, A.L. Corper,
T.A. Cross, X. Dai, E.W. Debler, W.L. Densley, D.C. Ekiert,
M.-A. Elsliger, S. Ferguson, B.W. Han, G.W. Han, M. Hong,
M.J. Jimenez-Dalmaroni, J.R. Mikolosko, R. Pejchal,
A. Schiefner, D.A. Shore, R.S. Stefanko, J.A. Vanhnasy,
P. Verdino, L. Xu, X. Xu, S.I. Yoon, D.M. Zajonc
e use x-ray crystallography to investigate the
structure and function of many different receptors in the innate and adaptive immune systems. We are also studying antigens from influenza
virus and HIV type 1 (HIV-1) to determine how these
viruses are neutralized by the immune system.
W
INFLUENZA VIRUS
F i g . 1 . HIV-1 vaccine design by insertion of the HIV-1 gp41
The 1918 influenza pandemic remains the most
devastating single infectious outbreak on record; at
least 40 million people died of the disease. Influenza
has 2 surface glycoproteins: hemagglutinin and neuraminidase. Hemagglutinin plays a key role in viral infection by binding to sialylated receptors on target cells,
resulting in viral entry and membrane fusion. Neuraminidase facilitates release of newly formed viral particles from the host cell, enabling spread of the virus
to neighboring cells. As part of a consortium funded
by the National Institute of Allergy and Infectious Diseases to understand the pathogenicity of the 1918
virus and the high mortality rate, we have determined
crystal structures of the H1 hemagglutinin from 1918
as well as the more recent H5N1 hemagglutinin from
2004 and the 1918 N1 neuraminidase to aid in design
of vaccines to prevent future influenza pandemics.
membrane-proximal region into gp120. The red helix represents
the 4E10 epitope that has been inserted into the V1/V2 region of
NEUTRALIZING ANTIBODIES TO HIV-1
Successful eradication of HIV-1 depends on an
effective vaccine. Unfortunately, this virus excels in
evading the immune system, thus hindering vaccine
development. By studying a handful of broadly neutralizing antibodies to HIV-1 in complex with their
viral antigens, we hope to elucidate vulnerable sites
for antibody neutralization. The antibodies 4E10 and
gp120 for testing as a potential immunogen for an HIV-1 vaccine.
The gp41 and gp120 viral envelope proteins form
trimeric spikes on the intact virus. After binding human
CD4 and chemokine receptors, the trimer undergoes
conformational changes that lead to fusion of the viral
and target cell membranes, initiating infection. We are
characterizing the structure of the trimer to understand
membrane fusion and why antibodies can only neutralize
intact envelope trimers.
Our research on HIV is done in collaboration with
D. Burton, Department of Immunology; P. Dawson,
Department of Cell Biology; C.-H. Wong, Department
of Chemistry; J.K. Scott, Simon Fraser University,
Burnaby, British Columbia; J. Moore, Weill Medical
College of Cornell University, New York, New York;
H. Katinger, R. Kunert, and G. Stiegler, University für
Bodenkultur, Vienna, Austria; R. Wyatt and P. Kwong,
Vaccine Research Center, National Institutes of Health,
Bethesda, Maryland; W. Olson, and K. Kang, Progenics Pharmaceuticals, Inc., Tarrytown, New York; the
National Institutes of Health, Bethesda, Maryland; and
the Neutralizing Antibody Consortium of the International AIDS Vaccine Initiative, New York, New York.
MOLECULAR BIOLOGY
2007
THE SCRIPPS RESEARCH INSTITUTE
191
G PROTEIN–COUPLED RECEPTORS
G protein–coupled receptors (GPCRs) are a large
family of integral membrane proteins central to numerous biological processes. Because of their involvement
in many diseases, GPCRs collectively account for more
than 50% of current chemotherapeutic targets. We are
developing recombinant expression systems for numerous chemokine receptor GPCRs. Although yield and
protein stability continue to challenge structural investigation of these proteins, early results in protein production appear promising.
CLASSICAL AND NONCLASSICAL MHC AND T-CELL
RECEPTOR SIGNALING
MHC class I molecules enable the immune system
to monitor cells for infection and cancer via presentation
of antigenic peptides on the cell surface. Efficient presentation of a diverse repertoire of peptides by MHC
class I molecules depends on the peptide loading complex, of which tapasin is an essential component. We
have expressed soluble tapasin in Escherichia coli for
structural studies to further elucidate the mechanism
of MHC class I peptide loading.
We have a long-term interest in the structural
aspects of MHCs and T-cell receptor (TCR) recognition.
Recent results include the crystal structure of rat MHC
class Ib BM1 in complex with peptide Qdm at 2.15 Å.
This structure reveals how BM1 differs from other MHC
homologs and hence defines the structural determinants
of natural killer cell regulation in innate immunity. We
are also studying the TCR KRN that recognizes the
MHC class II allele I-Ag7 in complex with a self-peptide derived from glucose-6-phosphate isomerase that
initiates rheumatoid arthritis in KRN transgenic mice.
Crystals of the complex composed of KRN, I-Ag7, and
the glucose-6-phosphate isomerase peptide have been
obtained, and structure determination is ongoing. To
provide insights into potential mechanisms of diabetogenesis, we are also investigating whether TCRs recognize an oxyanion hole in I-Ag7. To this end, a crystal
structure for the TCR HEL, isolated from nonobese
diabetic mice immunized against the peptide HEL9-27,
in complex with I-Ag7–HEL9-27 has been determined
(Fig. 2). This work is a collaboration with L. Teyton,
Department of Immunology.
The T-cell response also depends on the CD3 complex, which consists of 4 subunits (ε, δ, γ, and ζ) that
form pairs of dimers that assemble on the TCR to
facilitate transmembrane signal transduction. Although
structures of CD3εδ and CD3εγ are now available, the
F i g . 2 . Crystal structure of the HEL TCR–I-A g7–HEL 9-27 com-
plex. The TCR is shown in pink, the MHC in olive green, and the
peptide in yellow.
mechanism coupling TCR recognition of antigen to
T-cell activation has yet to be determined. Thus, we
have developed mutant TCRs with high affinity for
CD3 subunits for structural analysis of the intact TCRCD3 complex. We are also studying the TCR coreceptor CD8, an essential element in the cytotoxic T-cell
response to peptide antigen, to determine the molecular basis of the function of the CD8 coreceptor.
Mycobacterial phosphatidylinositol tetramannosides
(PIMs) are a major component of the mycobacterial
outer membrane leaflet and when presented by CD1d
stimulate invariant human Vα24 and mouse Vα14
natural killer T cells. In collaboration with W. Severn
and G. Painter, Industrial Research Ltd., Upper Hut,
New Zealand, we determined the crystal structure of
mouse CD1d in complex with synthetic PIM2, which
192 MOLECULAR BIOLOGY
2007
THE SCRIPPS RESEARCH INSTITUTE
represents the most complex headgroup of all CD1d
ligand structures to date (Fig. 3). Compared with other
CD1d ligands, PIM2 has an increased number of polar
interactions between its headgroup and CD1, but reduced
specificity for the diacylglycerol backbone.
F i g . 4 . Crystal structure of JAML in complex with the mitogenic
antibody 4E10. JAML has 2 immunoglobulin domains (D1 and D2)
arranged in a unique orientation for tandem immunoglobulin domains.
Fab complementarity-determining region loops (L1, L2, L3, H1,
H2, H3) contact primarily the D2 domain.
F i g . 3 . Comparison of the different ligands bound to mouse CD1d.
A, PIM2. B, Sulfatide. C, The highly potent α-galactosylceramide.
T H E I N N AT E I M M U N E S Y S T E M
The enigmatic γδ T cells link innate and adaptive
immunity. These cells carry out functions such as tumor
cell recognition, maintenance of tissue homeostasis,
and tissue repair and act as the first line of defense
against bacterial and viral infections. In collaboration
with W. Havran, Department of Immunology, we determined the crystal structures of JAML, a costimulatory
molecule specific for γδ T cells, alone and in complex
with a mitogenic antibody (Fig. 4). The structures
revealed a unique assembly of tandem immunoglobulin
domains. Conformational changes in the ectodomain
of JAML on activation may be the trigger for initiating
kinase signaling cascades, cytokine and chemokine production, and, ultimately, cell proliferation.
Toll-like receptors (TLRs) are cell-surface receptors
in humans that detect microbes invading through the
skin or intestinal mucosa. TLRs play a key role in initiating immune responses by recognizing a variety of
pathogen-associated molecular patterns, including components of bacterial cell walls and viral nucleic acids.
Since determining the structure of TLR3 , we are working on other human TLRs in complex with their ligands.
The nucleotide-binding oligomerization domains
(NODs) 1 and 2 are intracellular receptors that recognize bacterial peptidoglycans. NOD2 mutations have
been associated with inflammatory Crohn’s disease.
We aim to crystallize the putative binding leucine-rich
repeat domain of NOD2 and also the intracellular caspase-recruitment domain involved in signal transduction. The TLR and NOD studies are collaborations with
B. Beutler and R. Ulevitch, Department of Immunology.
N O N M A M M A L I A N I N N AT E A N D A D A P T I V E I M M U N I T Y
Sharks and other cartilaginous fish diverged from
mammals over 5 million years ago and are the most
ancient vertebrates with an adaptive immune system
that involves antibodies, TCRs, MHC molecules, and
recombination-activating genes. In collaboration with
M. Flajnik and H. Dooley, University of Maryland, Baltimore, we have determined structures for germ-line
and somatically mutated variable domains from a
MOLECULAR BIOLOGY
2007
nurse shark “new antigen receptor ” antibody in unliganded and antigen-complexed forms. These primitive
antibody binding domains share many features with
mammalian antibodies, such as flexibility of the H3
complementarity-determining region, and correlation
between affinity, increased contacts in the binding
site, and somatic mutations.
Furthermore, sequence comparisons indicate that
shark proteins Saac-UAA*01 and Sqac-UAA*NC1 are
members of the classical and nonclassical class I MHCs,
respectively. Currently, in collaboration with C. Dascher,
Mount Sinai School of Medicine, New York, New York,
we are optimizing expression of these proteins in baculovirus and Drosophila systems for crystallization trials to explore the proteins’ structure and function.
Jawless fish are thought to have an adaptive immune
system based not on antibodies, but on variable lymphocyte receptors (VLRs). VLRs are composed of variable
numbers of leucine-rich repeats that can be rearranged
in a combinatorial fashion to recognize a number of
diverse antigens. In collaboration with M.D. Cooper,
University of Alabama, Birmingham, we have produced
and are crystallizing several different constructs of VLR4,
VLR5, and VLR2913.
C ATA LY T I C A N T I B O D I E S
Antibodies can catalyze a myriad of interesting,
often difficult, enzymatic reactions. Currently, we are
working with D. Hilvert, ETH, Zurich, Switzerland, to
determine structures of the catalytic antibodies 34E4
and 13G5 that carry out proton transfer, one of the most
fundamental chemical reactions in chemistry and biology
(Fig. 5). With K.D. Janda and P. Wentworth, Department of Chemistry, we are studying antibodies 38H10
and 29G12, which catalyze the 1,3-dipolar cycloaddition reaction that is extremely useful for synthesis of
diverse chiral heterocyclic compounds and is not efficiently catalyzed by any natural enzyme.
F i g . 5 . Crystal structures of antibody 13G5 show how bifunctional catalysis of the proton-transfer reaction is realized by AspH35
as the general base and a water molecule as the acid catalyst.
THE SCRIPPS RESEARCH INSTITUTE
193
GREEN FLUORESCENT ANTIBODIES
Scientists in the laboratory of P.G. Schultz, Department of Chemistry, have generated antibodies to the
donor-acceptor substituted stilbene trans-4-N,N-dimethylamino-4′-cyanostilbene (DCS) that yield bright blue
to green fluorescence upon illumination. The relatively
long excitation and emission wavelengths of the antibody-stilbene complexes may make them useful for in
vitro and in vivo applications as fluorescent biosensors.
To investigate the photophysics of DCS in atomic detail,
we have determined the crystal structure of Fab 11G10
in complex with hapten at 2.75 Å. The charge distribution and aromatic ring systems in the combining site are
well suited to stabilize charge separation in the excited
state of DCS and thus provide a structural basis for emission at longer wavelengths than those associated with
other antibodies to stilbene.
THERAPEUTIC ANTIBODIES
Gram-negative bacteria can use N-acyl homoserine
lactones (AHLs) as signaling molecules in quorum sensing, a population density–dependent mechanism to coordinate gene expression. K.D. Janda, Department of
Chemistry, has used a lactam mimetic of AHL as an
immunogen to generate antibody RS2-1G9, which also
recognizes the naturally occurring AHL with high affinity. Because of this cross-reactivity, RS2-1G9 shows
remarkable inhibition of quorum-sensing signaling in
Pseudomonas aeruginosa, a common opportunistic
pathogen in humans. The crystal structure of Fab RS21G9 in complex with a lactam analog (Fig. 6) revealed
complete encapsulation of the polar lactam and can now
aid in further development of an antibody-based therapy against bacterial pathogens that interferes with
quorum sensing.
F i g . 6 . The crystal structure of the quorum-quenching RS2-1G9
Fab in complex with a lactam analog. The polar lactam moiety is
totally encapsulated in the antibody-combining site.
194 MOLECULAR BIOLOGY
2007
THE SCRIPPS RESEARCH INSTITUTE
N-formyl peptides, such as formyl-methionyl-leucylphenylalanine (fMLF), are potent chemotactic factors
for leukocytes. Unwanted inflammation should be suppressed by removing the N-formyl peptides through binding to antibodies. F. Tanaka, Department of Chemistry,
has developed antibody 6G5 that selectively binds fMLF.
Crystal structures of the Fab′ and its complex with fMLF
at 2.7 and 1.95 Å, respectively, revealed the mechanism of the discrimination between N-formyl and nonformylated peptides and now provide information for
further generation of humanized antibodies as antiinflammatory drugs.
Didonato, M., Krishna, S.S., Schwarzenbacher, R., et al. Crystal structure of
2-phosphosulfolactate phosphatase (ComB) from Clostridium acetobutylicum at 2.6 Å
resolution reveals a new fold with a novel active site. Proteins 65:771, 2006.
JOINT CENTER FOR STRUCTURAL GENOMICS
Kosloff, M., Han, G.W., Krishna, S.S., et al. Comparative structural analysis of a
novel glutathione S-transferase (ATU5508) from Agrobacterium tumefaciens at
2.0 Å resolution. Proteins 65:527, 2006.
The Joint Center for Structural Genomics is a large
consortium of scientists from Scripps Research; the
Stanford Synchrotron Radiation Laboratory; the University of California, San Diego; the Burnham Institute for
Medical Research; and the Genomics Institute of the
Novartis Research Foundation. The center is funded by
the Protein Structure Initiative of the National Institute
of General Medical Sciences. Its purpose is high-throughput structure determination of large families of protein
sequences with no structural representatives, biologically
important targets that are conserved as the central machinery of life; the complete proteome from Thermotoga
maritima; metagenomic and human gut microbiome
targets; and other targets suggested by the community.
To date, the members of the consortium have pioneered
many novel high-throughput methods and technologies
applicable to structural biology and have determined
more than 400 novel structures.
PUBLICATIONS
Binley, J.M., Ngo-Abdalla, S., Moore, P., Bobardt, M., Chatterji, U., Gallay, P., Burton, D.R., Wilson, I.A., Elder, J.H., de Parseval, A. Inhibition of HIV Env binding to
cellular receptors by monoclonal antibody 2G12 as probed by Fc-tagged gp120 [published correction appears in Retrovirology 4:23, 2007]. Retrovirology 3:39, 2006.
Cardoso, R.M., Brunel, F.M., Ferguson, S., Zwick, M., Burton, D.R., Dawson,
P.E., Wilson, I.A. Structural basis of enhanced binding of extended and helically
constrained peptide epitopes of the broadly neutralizing HIV-1 antibody 4E10. J.
Mol. Biol. 365:1533, 2007.
Cheng, T.Y., Relloso, M., Van Rhijn, I., Young, D.C., Besra, G.S., Briken, V.,
Zajonc, D.M., Wilson, I.A., Porcelli, S., Moody, D.B. Role of lipid trimming and
CD1 groove size in cellular antigen presentation. EMBO J. 25:2989, 2006.
Cooper, Z.D., Narasimhan, D., Sunahara, R.K., Mierzejewski, P., Jutkiewicz,
E.M., Larsen, N.A., Wilson, I.A., Landry, D.W., Woods, J.H. Rapid and robust
protection against cocaine-induced lethality in rats by the bacterial cocaine
esterase. Mol. Pharmacol. 70:1885, 2006.
Han, G.W., Sri Krishna, S., Schwarzenbacher, R., et al. Crystal structure of the
ApbE protein (TM1553) from Thermotoga maritima at 1.58 Å resolution. Proteins
64:1083, 2006.
Jiang, Z., Georgel, P., Li, C., Choe, J., Crozat, K., Rutschmann, S., Du, X., Bigby,
T., Mudd, S., Sovath, S., Wilson, I.A., Olson, A., Beutler, B. Details of Toll-like
receptor:adapter interaction revealed by germ-line mutagenesis. Proc. Natl. Acad.
Sci. U. S. A. 103:10961, 2006.
Kinjo, Y., Tupin, E., Wu, D., Fujio, M., Garcia-Navarro, R., Benhnia, M.R.,
Zajonc, D.M., Ben-Menachem, G., Ainge, G.D., Painter, G.F., Khurana, A.,
Hoebe, K., Behar, S.M., Beutler, B., Wilson, I.A., Tsuji, M., Sellati, T.J., Wong,
C.-H., Kronenberg, M. Natural killer T cells recognize diacylglycerol antigens from
pathogenic bacteria. Nat. Immunol. 7:978, 2006.
Law, M., Cardoso, R.M., Wilson, I.A., Burton, D.R. Antigenic and immunogenic
study of membrane-proximal external region-grafted gp120 antigens by a DNA
prime-protein boost immunization strategy. J. Virol. 81:4272, 2007.
Lazoura, E., Lodding, J., Farrugia, W., Ramsland, P.A., Stevens, J., Wilson, I.A.,
Pietersz, G.A., Apostolopoulos, V. Enhanced major histocompatibility complex
class I binding and immune responses through anchor modification of the noncanonical tumour-associated mucin 1-8 peptide. Immunology 119:306, 2006.
Mathews, I.I., Krishna, S.S., Schwarzenbacher, R., et al. Crystal structure of
phosphoribosylformyl-glycinamidine synthase II, PurS subunit (TM1244) from
Thermotoga maritima at 1.90 Å resolution. Proteins 65:249, 2006.
Muller, R., Debler, E.W., Steinmann, M., Seebeck, F.P., Wilson, I.A., Hilvert, D.
Bifunctional catalysis of proton transfer at an antibody active site. J. Am. Chem.
Soc. 129:460, 2007.
Nelson, J.D., Brunel, F.M., Jensen, R., Crooks, E.T., Cardoso, R.M., Wang, M.,
Hessell, A., Wilson, I.A., Binley, J.M., Dawson, P.E., Burton, D.R., Zwick, M.B.
An affinity-enhanced neutralizing antibody against the membrane-proximal external
region of human immunodeficiency virus type 1 gp41 recognizes an epitope
between those of 2F5 and 4E10. J. Virol. 81:4033, 2007.
Rudolph, M.G., Stanfield, R.L., Wilson, I.A. How TCRs bind MHCs, peptides, and
coreceptors. Annu. Rev. Immunol. 24:419, 2006.
Schwarzenbacher, R., McMullan, D., Krishna, S.S., et al. Crystal structure of a
glycerate kinase (TM1585) from Thermotoga maritima at 2.70 Å resolution reveals
a new fold. Proteins 65:243, 2006.
Stanfield, R.L., Dooley, H., Verdino, P., Flajnik, M.F., Wilson, I.A. Maturation of
shark single-domain (IgNAR) antibodies: evidence for induced-fit binding. J. Mol.
Biol. 367:358, 2007.
Stevens, J., Blixt, O., Paulson, J.C., Wilson, I.A. Glycan microarray technologies: tools
to survey host specificity of influenza viruses. Nat. Rev. Microbiol. 4:857, 2006.
Tian, F., Debler, E.W., Millar, D.P., Deniz, A.A., Wilson, I.A., Schultz, P.G. The
effects of antibodies on stilbene excited-state energetics. Angew. Chem. Int. Ed.
45:7763, 2006.
Weekes, D., Miller, M.D., Krishna, S.S., et al. Crystal structure of a transcription
regulator (TM1602) from Thermotoga maritima at 2.3 Å resolution. Proteins
67:247, 2007.
Debler, E.W., Kaufmann, G.F., Kirchdoerfer, R.N., Mee, J.M., Janda, K.D., Wilson,
I.A. Crystal structures of a quorum-quenching antibody. J. Mol. Biol. 368:1392, 2007.
Xu, L., Chong, Y., Hwang, I., D’Onofrio, A., Amore, K., Beardsley, G.P., Li, C.,
Olson, A.J., Boger, D.L., Wilson, I.A. Structure-based design, synthesis, evaluation
and crystal structures of transition state analogue inhibitors of inosine monophosphate cyclohydrolase. J. Biol. Chem. 282:13033, 2007.
DeMartino, J.K., Hwang, I., Xu, L., Wilson, I.A., Boger, D.L. Discovery of a potent,
nonpolyglutamatable inhibitor of glycinamide ribonucleotide transformylase. J. Med.
Chem. 49:2998, 2006.
Xu, Q., Krishna, S.S., McMullan, D., et al. Crystal structure of an ORFan protein
(TM1622) from Thermotoga maritima at 1.75 Å resolution reveals a fold similar to
the Ran-binding protein Mog1p. Proteins 65:777, 2006.
MOLECULAR BIOLOGY
2007
Xu, Q., Schwarzenbacher, R., Krishna, S.S., et al. Crystal structure of acireductone dioxygenase (ARD) from Mus musculus at 2.06 Å resolution. Proteins
64:808, 2006.
Zajonc, D.M., Ainge, G.D., Painter, G.F., Severn, W.B., Wilson, I.A. Structural
characterization of mycobacterial phosphatidylinositol mannoside binding to mouse
CD1d. J. Immunol. 177:4577, 2006.
Structure and Function of
Proteins as Molecular Machines
E.D. Getzoff, P. Aoto, A.S. Arvai, D.P. Barondeau,
R.M. Brudler, T. Cross, E.D. Garcin-Hosfield, C. Hitomi,
K. Hitomi, M.E. Pique, J.L. Tubbs, T.I. Wood
ur goal is to understand how proteins function
as molecular machines. We apply the tools of
structural, molecular, and computational biology to proteins of biological and biomedical interest,
especially proteins that work synergistically with coupled chromophores, metal ions, or other cofactors.
O
NITRIC OXIDE SYNTHASES
Nitric oxide synthase (NOS) enzymes synthesize
nitric oxide, a signal for vasodilation and neurotransmission at low levels and a defensive cytotoxin at higher
levels. Synthesis of nitric oxide by NOS requires calmodulin-orchestrated interactions between the catalytic,
heme-containing oxygenase module and the electronsupplying reductase module of the enzyme. Our x-ray
crystallographic structures of wild-type and mutant NOS
oxygenase dimers with substrate, intermediate, inhibitors, cofactors, and cofactor analogs, determined in
collaboration with J. Tainer, Department of Molecular
Biology, and D. Stuehr, the Cleveland Clinic, Cleveland,
Ohio, provide insights into the catalytic mechanism and
dimer stability.
The goals of our structure-based drug design projects are to selectively inhibit inducible NOS, to prevent
inflammatory disorders, or neuronal NOS, to prevent
migraines, while maintaining blood pressure regulation
by endothelial NOS. The nearly complete sequence and
structural conservation in the active sites of the 3 NOS
isozymes is a significant challenge in the design of
isozyme-specific inhibitors. Nevertheless, our latest
results indicate that plasticity of distant isozyme-specific
residues modulates conformational changes of invariant
residues in the substrate-binding site. These differences
in residue plasticity can be exploited to create inhibitors
that are 3000 times more selective for one isozyme than
for another despite binding-site conservation.
THE SCRIPPS RESEARCH INSTITUTE
195
Our structure of the neuronal NOS reductase reveals
new insights into the complex regulatory mechanisms
of this enzyme family. We integrated biochemical data
with our structures of dimeric NOS oxygenase, dimeric
NOS reductase, and calmodulin in complex with peptides derived from NOS to propose a model for the
assembled holoenzyme. We have obtained promising
results in support of this assembly model by using solution small-angle x-ray scattering, which can provide
molecular envelopes for macromolecules and macromolecular complexes in solution.
On the basis of our NOS research, we have proposed a moving-domain mechanism for controlling the
rate-limiting flow of electrons from the flavin cofactors
of NOS reductase to the catalytic NOS oxygenase heme.
Our assembly and mechanistic hypotheses also explain
the kinetics of regulatory site-specific phosphorylation
and dephosphorylation events, as defined by our collaborators G. Rameau, Johns Hopkins School of Medicine, Baltimore, Maryland, and E. Ziff , New York
University, New York City, that both activate and inactivate nitric oxide synthesis in vivo (Fig. 1).
PHOTOACTIVE PROTEINS AND CIRCADIAN CLOCKS
To understand in atomic detail how proteins translate sunlight into defined conformational changes for
biological functions, we are exploring the reaction mechanisms of the blue-light receptors photoactive yellow
protein (PYP), photolyase, and cryptochrome. PYP is
the prototype for the Per-Arnt-Sim domain proteins of
circadian clocks, whereas proteins of the photolyase/
cryptochrome family catalyze DNA repair or act in circadian clocks. To understand the protein photocycle of
PYP and propose a common mechanism for signaling
by Per-Arnt-Sim domains, we combined ultra-high-resolution and time-resolved crystallographic structures of
the PYP dark state and 2 photocycle intermediates with
site-directed mutagenesis; ultraviolet-visible spectroscopy; time-resolved Fourier transform infrared spectroscopy; deuterium-hydrogen exchange mass spectrometry,
in collaboration with V. Woods, University of California, San Diego; and quantum mechanical and electrostatic computational methods, in collaboration with
L. Noodleman, Department of Molecular Biology.
Cryptochrome flavoproteins are homologs of lightdependent DNA repair photolyases that function as
blue-light receptors in plants and as components of
circadian clocks in animals. We determined the first
crystallographic structure of a cryptochrome; the structure revealed commonalities with photolyases in DNA
196 MOLECULAR BIOLOGY
2007
THE SCRIPPS RESEARCH INSTITUTE
F i g . 1 . Structural mechanism for the temporal sequence of neuronal NOS (nNOS) phosphorylations.
binding and redox-dependent function but showed differences in active-site and interaction-surface features.
Recently, we showed that this cryptochrome binds the
same antenna cofactor found in a photolyase homolog
but uses different amino acid residues to form the
cofactor-binding site. Our new structures and spectroscopy, in collaboration with S. Weber, Freie Universität
Berlin in Germany, of cryptochromes and of photolyases
from 2 other branches of the photolyase/cryptochrome
family that repair cyclobutane pyrimidine dimers and
(6-4) photoproducts help us decipher the cryptic structure, function, and evolutionary relationships of these
fascinating redox-active proteins. Furthermore, the (6-4)
photolyase enzyme provides an excellent model for evaluating the functions of human cryptochrome (Fig. 2).
A simple, but functional, circadian clock can be
reconstituted in vitro from the 3 cyanobacterial proteins
KaiA, KaiB, and KaiC alone. Yet, the structure and
dynamics of the functional assembly are not understood. Our crystallographic, dynamical light scattering,
and small-angle x-ray scattering studies revealed that
KaiB self-assembles into a tetramer. We also study
clock proteins with PYP-like Per-Arnt-Sim domains that
bind to mammalian cryptochromes. Our goal is to determine the detailed chemistry and atomic structure of
these proteins, define their mechanisms of action and
F i g . 2 . The FAD binding sites in the crystallographic structure of
cryptochrome (left) and a homology model for (6-4) photolyase (right)
were used to design single-site mutants of the 2 histidine residues
evolved to function catalytically in DNA repair. We differentiated
distinct functions for these catalytic histidines (His354 and His358)
by examining pulsed electron nuclear double resonance spectra of
the wild-type and mutant proteins under different pH conditions.
interaction, and use our results to understand and regulate their biological function.
P O S T T R A N S L AT I O N A L M O D I F I C AT I O N C H E M I S T R Y
Green fluorescent protein (GFP) and the homologous
red fluorescent protein (RFP) spontaneously self-modify their polypeptide chains to form their characteristic
green and red fluorophores. We discovered that the
architecture of GFP and RFP promotes a remarkable
MOLECULAR BIOLOGY
2007
range of other posttranslational modification chemistry.
Our high-resolution crystallographic studies of GFP and
RFP intermediates in chromophore cyclization and oxidation provide a novel mechanism for the spontaneous
synthesis of this tripeptide fluorophore within the protein scaffold. Remarkably, the same protein architectural features that favor peptide cyclization can drive
peptide hydrolysis at 4 consecutive positions along the
polypeptide backbone and red-shift the spectral properties of the chromophore. Decarboxylation and cleavage of carbon-carbon bonds in designed GFP variants
further support a role for the GFP protein environment
in facilitating formation of radicals and 1-electron
chemistry. Together, our results elucidate the natural
mechanism of fluorophore formation and provide the
groundwork for the design of proteins with novel catalytic or reporter properties.
THE SCRIPPS RESEARCH INSTITUTE
197
S M A L L - A N G L E X - R AY S C AT T E R I N G I N S O L U T I O N
C O M B I N E D W I T H C R Y S TA L L O G R A P H Y A N D
C O M P U TAT I O N
Aided by our advanced synchrotron facility SIBLYS,
we are developing tools and technologies to combine
x-ray scattering in solution with x-ray crystallography
and computation. These paired x-ray techniques can
create the complete and accurate images of macromolecules in solution often required to address critical structural questions in biology. Small-angle x-ray scattering,
crystallography, and computation together allow multiscale modeling of and fundamental insights to allosteric
mechanisms, self-assemblies, and dynamic molecular
machines acting in diverse processes ranging from
eukaryotic DNA replication, recombination, and repair
to microbial membrane secretion and assembly systems (Fig. 1).
PUBLICATIONS
Barondeau, D.P., Kassmann, C.J., Tainer, J.A., Getzoff, E.D. The case of the missing ring: radical cleavage of a carbon-carbon bond and implications for GFP chromophore biosynthesis. J. Am. Chem. Soc. 129:3118, 2007.
Hill, N.J., Stotland, A., Solomon, M., Secrest, P., Getzoff, E., Sarvetnick, N.
Resistance of the target islet tissue to autoimmune destruction contributes to
genetic susceptibility in type 1 diabetes. Biol. Direct 2:5, 2007.
Panda, K., Haque, M.M., Garcin-Hosfield, E.D., Durra, D., Getzoff, E.D., Stuehr,
D.J. Surface charge interactions of the FMN module govern catalysis by nitric-oxide
synthase. J. Biol. Chem. 281:36819, 2006.
Rameau, G.A., Tukey, D.S., Garcin-Hosfield, E.D., Titcombe, R.F., Misra, C., Khatri, L., Getzoff, E.D., Ziff, E.B. Biphasic coupling of neuronal nitric oxide synthase
phosphorylation to the NMDA receptor regulates AMPA receptor trafficking and
neuronal cell death. J. Neurosci. 27:3445, 2007.
Schleicher, E., Hitomi, K., Kay, C.W., Getzoff, E.D., Todo, T., Weber, S. Electron
nuclear double resonance differentiates complementary roles for active site histidines in (6-4) photolyase. J. Biol. Chem. 282:4738, 2007.
Structural Biology of Molecular
Interactions and Design
J.A. Tainer, A.S. Arvai, D.P. Barondeau, B. R. Chapados,
T.H. Cross, L. Fan, C. Hitomi, K. Hitomi, J.J. Perry,
M.E. Pique, D.S. Shin, R.S. Williams, A. Yamagata
e focus on molecular mechanisms and relationships of protein regulators and effectors
of DNA damage responses, reactive oxygen
species, and pathogenesis. To help understand multidomain macromolecules with conformational changes
and functionally important flexibility in these processes,
we are combining solution methods with high-resolution structures and advanced computation.
W
F i g . 1 . Structures of the Gsp ATPase superfamily for microbial
secretion and assembly based on the results of a combination of
small-angle x-ray scattering and crystallography. A, Side view of the 2
alternating configurations of the ATPase GspE monomer bound to
the nonhydrolyzable ATP analog adenylylimidodiphosphate (AMPPNP)
in the hexameric structure determined by crystallography. B, Top
views of the crystal structure and 2 models proposed on the basis
of modifying subunits to adopt either all-open (brown) or all-closed
(green) conformations. C, In solution, the ATPase in excess AMPPNP
adopts a conformation most similar to the all-closed model, whereas a solution with excess ADP is best described as a mixture of all
the models. Adapted from Yamagata, A., Tainer, J.A. Hexameric
structures of the archaeal secretion ATPase GspE and implications
for a universal secretion mechanism. EMBO J. 26:878, 2007.
198 MOLECULAR BIOLOGY
2007
P R O T E I N M O D I F I C AT I O N S A N D F U N C T I O N
The number of protein-coding genes in the human
genome is much smaller than the number of proteins
found in human cells. This difference is due to the
increase in protein diversity caused by alternative splicing and posttranslational modification of proteins. We
are collaborating with E. Getzoff, Department of Molecular Biology, to understand the spontaneous cyclization
of the peptide backbone and the oxidation chemistry
that convert 3 amino acids into a fluorophore for the
family of green fluorescent proteins. We recently discovered how the structural chemistry of green fluorescent protein can accomplish many different types of
posttranslational modification, including the surprising
radical cleavage of a carbon-carbon bond.
E N Z Y M E S T H AT C O N T R O L R E A C T I V E O X Y G E N
SPECIES
Superoxide dismutases and nitric oxide synthases are
master regulators for reactive oxygen species involved in
injury, pathogenesis, aging, and degenerative diseases.
For human copper, zinc superoxide dismutase, we are
examining single-site mutations that cause the neurodegeneration in Lou Gehrig disease or familial amyotrophic
lateral sclerosis. We recently solved high-resolution structures that reveal a key role for the zinc ion in the structural defects associated with the disease. For nitric oxide
synthases, we are using our combined solution x-ray
scattering and crystallographic methods to examine electron transfer and regulatory mechanisms that control levels of nitric oxide, which acts as an important signal and
cytotoxin with implications for inflammatory and neurodegenerative diseases.
DNA REPAIR AND GENETIC EVOLUTION
The genetic information for heredity is encoded in
DNA molecules that are constantly under attack from
sunlight, ionizing radiation, and other environmental
carcinogens. Surprisingly, however, most DNA damage
is due to chemical reactions and free radicals that arise
from normal cellular metabolism and oxidative responses
to infectious disease. In fact, life is impossible without strong DNA repair responses. As a result, mutations
that cause defects in DNA repair systems may cause
cancer and degenerative diseases associated with aging.
However, selective inhibition of certain DNA repair pathways may offer new methods for cancer therapy.
We are solving structures of the O 6 -alkylguanineDNA alkyltransferase to aid in the design of inhibitors
for cancer chemotherapy. Our structures of several DNAbase repair enzymes are revealing the detailed structural
THE SCRIPPS RESEARCH INSTITUTE
chemistry of base-excision repair machinery. Our structures of the enzyme that cuts the DNA backbone for the
second step of base repair suggest how the enzyme
holds the DNA product and hands the product off to the
polymerase to replace the removed nucleotide.
We are also examining the interface exchange that
allows dynamic complexes to form on proliferating cell
nuclear antigen when it is loaded onto double-stranded
DNA. We found that it binds the DNA minor groove
and can cooperate with replication and repair enzymes
in their interactions with the adjacent DNA double helix
to process and rejoin DNA ends. To understand the
initial response to DNA double-strand breaks, we are
solving new structures of the Rad50 ABC ATPase and
the partner Mre11 nuclease with bound DNA that reveal
how this complex holds and processes DNA ends to
repair DNA double-strand breaks (Fig. 2). In general,
F i g 2 . Models for assembly of the Mre11-Rad50-Nbs1 complex
(MRN). Intracomplex zinc hook minimizes unproductive, intercomplex interactions in the absence of DNA. DNA binding (1) straightens the Rad50 coiled coils, a condition that favors intercomplex
tethering via Rad50 zinc hooks with extended and parallel coils
(2). K indicates kink regions that intersperse regions with strong
potential for forming coiled coils (cc).
we think that the structural biology of the proteins that
control reactive oxygen species and DNA repair may provide master keys to understanding of and therapeutic
interventions for brain abnormalities, cancer, and aging.
BACTERIAL PILI AND INFECTIOUS DISEASES
Type IV pili are essential virulence factors for many
bacterial pathogens. Pili play key roles in surface motility, adhesion, formation of microcolonies and biofilms,
natural transformation, and signaling. We are characterizing the system machinery of type IV pili, including
structures of type IV pilin subunits, assembly of the
subunits into pilus fibers, the pilus membrane protein
partners, and the assembly ATPases (Fig. 1). We are
MOLECULAR BIOLOGY
2007
developing an integrated understanding of assembly
and disassembly of type IV pili by a pistonlike pushpull mechanism revealed by our structures of the assembly machinery and assembled type IV pilus fibers. This
understanding suggests new approaches to drug and
vaccine design for bacterial pathogens.
PUBLICATIONS
Barondeau, D.P., Kassmann, C.J., Tainer, J.A., Getzoff E.D. The case of the missing ring: radical cleavage of a carbon-carbon bond and implications for GFP chromophore biosynthesis. J. Am. Chem. Soc. 129:3118, 2007.
Blaine, R., Roberts, B.R., Tainer, J.A., Getzoff, E.D., Malencik, D.A., Anderson,
S.R., Bomben, V.C., Meyers, K.R., Karplus, A., Beckman, J.S. Structural characterization of zinc-deficient human superoxide dismutase and implications for ALS.
J. Mol. Biol., in press.
Hitomi, K., Iwaia, S., Tainer, J.A. The intricate structural chemistry of base excision repair machinery: implications for DNA damage recognition, removal, and
repair. DNA Repair (Amst.) 6:410, 2007.
Ivanov, I., Chapados, B.R., McCammon, J.A., Tainer, J.A. Proliferating cell nuclear
antigen loaded onto double-stranded DNA: dynamics, minor groove interactions
and functional implications. Nucleic Acids Res. 34:6023, 2006.
Ivanov I., Tainer, J.A., McCammon, J.A. Unraveling the three-metal-ion catalytic
mechanism of the DNA repair enzyme endonuclease IV. Proc. Natl. Acad. Sci.
U. S. A. 104:1465, 2007.
Perry, J.J., Fan, L., Tainer, J.A. Developing master keys to brain pathology, cancer
and aging from the structural biology of proteins controlling reactive oxygen species
and DNA repair. Neuroscience 145:1280, 2007.
Putnam, C.D., Hammel, M., Hura, G.L., Tainer, J.A. Solution scattering (SAXS)
combined with crystallography and computation: defining accurate macromolecular
structures, conformations and assemblies in solution. Q. Rev. Biophys., in press.
Tsutakawa, S.E., Hura, G.L., Frankel, K.A., Cooper, P.K., Tainer, J.A. Structural
analysis of flexible proteins in solution by small angle x-ray scattering combined
with crystallography. J. Struct. Biol. 158:214, 2007.
Tubbs, J.L., Pegg, A.E., Tainer,J.A. DNA binding, nucleotide flipping, and the
helix-turn-helix motif in base repair by O6-alkylguanine-DNA alkyltransferase and
its implications for cancer chemotherapy. DNA Repair (Amst.), in press.
Vijayakumar, S., Chapados, B.R., Schmidt, K.H., Kolodner, R.D., Tainer, J.A.,
Tomkinson, A.E. The C-terminal domain of yeast PCNA is required for physical and
functional interactions with Cdc9 DNA ligase. Nucleic Acids Res. 35:1624, 2007.
Williams, R., Sengerova, B., Osborne, S., Syson, K., Ault, S., Kilgour, A., Chapados, B.R., Tainer, J.A., Sayers, J.R., Grasby, J.A. Comparison of the catalytic parameters and reaction specificities of a phage and an archaeal flap endonuclease. J.
Mol. Biol. 371:34, 2007.
Williams, R.S., Tainer, J.A. Learning our ABCs: Rad50 directs MRN repair functions via adenylate kinase activity from the conserved ATP binding cassette. Mol
Cell 25:789, 2007.
Williams, R.S., Williams, J.S., Tainer, J.A. Mre11-Rad50-Nbs1 is a keystone complex connecting DNA repair machinery, double-strand break signaling, and the
chromatin template. Biochem. Cell Biol. 85:509, 2007.
Yamagata, A., Tainer, J.A. Hexameric structures of the archaeal secretion ATPase
GspE and implications for a universal secretion mechanism. EMBO J. 26:878, 2007.
THE SCRIPPS RESEARCH INSTITUTE
199
Structural Biology of Integral
Membrane Proteins
G. Chang, S. Aller, Y. Chen, X. He, A. Karyakin, S. Lieu,
C.L. Reyes, P. Szewczyk, T. Tuan, A. Ward, J. Yu
ur studies of membrane proteins encompass 5
areas: (1) the molecular structural basis for lipid
and drug transport across the cell membrane
by multidrug resistance (MDR) transporters, (2) the
crystallography of mammalian MDR transporters, (3)
signal transduction by receptors, (4) the discovery and
design of potent MDR reversal agents, and (5) the development of a cell-free system capable of producing large
quantities of integral membrane proteins. We use several experimental methods, including detergent/lipid
protein biochemistry, 3-dimensional crystallization of
integral membrane proteins, fluorescence anisotropy,
calorimetry, protein x-ray crystallography, and functional analysis of transporters.
We are addressing the molecular basis of MDR in
the treatment of infectious disease and cancer. MDR
can be caused by drug efflux pumps imbedded in the
cell membrane. Through our structural studies on MDR
transporters, we delineated the molecular mechanics
for the transport of amphipathic substrate across the
cell membrane. We hope that these structures can be
used to design more potent inhibitors to be used synergistically with established chemotherapeutic agents.
We are combining chemistry and biology with structures
in collaboration with M.G. Finn, Department of Chemistry, and Q. Zhang, Department of Molecular Biology.
In collaboration with R.A. Milligan, Department of Cell
Biology, we are using electron cryomicroscopy to visualize the structures of our transporters.
We recently reanalyzed our x-ray diffraction data
for the lipid ATP-binding cassette transporter MsbA
and determined 4 structures trapped in different conformations: 2 with nucleotide bound and 2 with no
nucleotide. Comparisons of the 2 types of conformations revealed that MsbA has a flexible hinge formed
by extracellular loops 2 and 3. The hinge allows the
nucleotide-binding domains to disassociate while the
ATP-binding half sites remain facing each other. The
binding of nucleotide causes a packing rearrangement
of the transmembrane helices and changes the accessibility of the transporter from cytoplasmic (inward)
facing to extracellular (outward) facing. Interestingly,
O
200 MOLECULAR BIOLOGY
2007
the inward and outward openings are mediated by 2
different sets of transmembrane helix interactions.
We have also reanalyzed our x-ray data on the MDR
transporter EmrE. EmrE functions as a homodimer of a
small 4-transmembrane protein. The membrane insertion topology of the 2 monomers is controversial. EmrE
was reported to have a unique orientation in the membrane. Models based on electron microscopy and on
recent biochemical studies posit an antiparallel dimer.
The corrected structures in complex with a transport
substrate are similar to the electron microscopy structure and indicate an antiparallel orientation for the
monomers, supporting a “dual topology” model.
Previously, we determined the x-ray structure of
EmrD, an MDR transporter from the major facilitator
superfamily. EmrD expels amphipathic compounds
across the inner membrane of Escherichia coli. The
structure reveals an interior that is composed of hydrophobic residues, a finding consistent with the role of
EmrD in transporting amphipathic molecules. Two long
loops extend into the inner leaflet side of the cell membrane and may recognize and bind substrate directly
from the lipid bilayer. We propose that multisubstrate
specificity, binding, and transport are facilitated by these
loop regions and the internal cavity.
Structure and Function of
Membrane-Bound Enzymes
C.D. Stout, M. Yamaguchi, R. Akhouri, V.M.M. Luna,
A. Annalora, H.A. Heaslet, J. Chartron
e focus on the structure and function of membrane-bound enzymes and the development
of methods for crystallizing membrane proteins. We study the mechanism of transhydrogenase, a
mitochondrial respiratory enzyme complex that couples
proton translocation with hydride transfer. This enzyme
is essential for maintaining NADPH levels in mitochondria, and it plays a critical role in insulin secretion in
beta cells of the pancreas. We use x-ray crystallography, biochemical and spectroscopic methods, electron
microscopy studies in collaboration with M. Yeager,
Department of Cell Biology, and nuclear magnetic resonance in collaboration with J. Dyson, Department of
Molecular Biology. Currently, further progress in understanding the structure and function of transhydrogenase
is stymied by the lack of a 3-dimensional structure;
W
THE SCRIPPS RESEARCH INSTITUTE
therefore, our primary effort now is to obtain diffraction-quality crystals of the enzyme in its membranebound configuration.
In collaboration with J.A. Fee, Department of Molecular Biology, we are studying the mechanism of cytochrome ba3 oxidase, a homolog of the terminal enzyme
of respiration in mitochondria. We are using high-resolution crystal structures, mutagenesis, and spectroscopy
to visualize intermediates in the reduction of oxygen to
water, to define the pathways for protons and oxygen
into the active site of the oxidase (Fig. 1), and to understand the coupling of reduction potential to proton translocation across the membrane.
F i g . 1 . The pathway for oxygen molecules from the membrane
into the active site of cytochrome ba 3 oxidase, as identified by
pressurization of crystals with xenon gas. The enzyme contains a
forked, hydrophobic channel (green mesh) leading from the lipid
bilayer into the active site where oxygen binds between iron and
copper atoms and is reduced to water.
In collaboration with P. Dawson, Department of
Cell Biology, we are developing the assembly of synthetic peptides and phospholipids into discs that contain lipid bilayers. Nanodiscs composed of human
apolipoprotein A-I and phospholipids self-assemble
into discrete, water-soluble, bilayer-containing particles.
Integral membrane proteins incorporated into these
particles retain their enzymatic activity, are amenable
to biochemical assays, and may have superior properties for crystallization in the absence of detergents. Both
transhydrogenase and cytochrome ba 3 oxidase have
been incorporated into nanodiscs.
In collaboration with E.F. Johnson, Department of
Molecular Biology; J.R. Halpert and I. Pikuleva, University of Texas Medical Branch, Galveston, Texas; and
MOLECULAR BIOLOGY
2007
others, we are characterizing the structure and function
of mammalian cytochrome P450s. These membraneassociated enzymes are involved in the biosynthesis of
lipophilic hormones and specifically metabolize a wide
variety of exogenous compounds and drugs. High-resolution structures have been determined for the principal
drug-metabolizing microsomal P450s in liver and lung
in humans: 1A2, 2A6, 2A13, 2C8, 2C9, 2C19, 3A4,
and 2B4. For 2B4, 4 structures of the enzyme in
markedly different conformations provide insight to substrate binding and membrane insertion. A structure of
the brain-specific, cholesterol-metabolizing P450 CYP46
has been determined, and the first mitochondrial P450,
CYP24A1, has been crystallized (Fig. 2).
THE SCRIPPS RESEARCH INSTITUTE
201
PUBLICATIONS
Chartron, J., Shiau, C., Stout, C.D., Carroll, K.S. 3′-Phosphoadenosine-5′-phosphosulfate reductase in complex with thioredoxin: a structural snapshot in the catalytic cycle. Biochemistry 46:3942, 2007.
Heaslet, H., Lin, Y.-C., Tam, K., Torbett, B.E., Elder, J.E., Stout, C.D. Crystal
structure of an FIV/HIV chimeric protease complexed with the broad-based inhibitor, TL-3. Retrovirology 4:1, 2007.
Heaslet, H., Rosenfeld, R., Giffin, M., Lin, Y.-C., Tam, K., Torbett, B.E., Elder,
J.H., McRee, D.E., Stout, C.D. Conformational flexibility in the flap domains of
ligand-free HIV protease. Acta Crystallogr. D Biol. Crystallogr. 63(Pt. 8):866,
2007.
Sansen, S., Hsu, M.-H., Stout, C.D., Johnson, E.F. Structural insight into the
altered substrate specificity of human cytochrome P450 2A6 mutants. Arch.
Biochem. Biophys., in press.
Sansen, S., Yano, J.K., Reynald, R.L., Schoch, G.A., Griffin, K.J., Stout, C.D.,
Johnson, E.F. Adaptations for the oxidation of polycyclic aromatic hydrocarbons
exhibited by the structure of human P450 1A2. J. Biol. Chem. 282:14348, 2007.
Smith, B.D., Sanders, J.L., Porubsky, P.R., Lushington, G.H., Stout, C.D., Scott,
E.E. Structure of the human lung cytochrome P450 2A13. J. Biol. Chem.
282:17306, 2007.
F i g . 2 . Crystals of mitochondrial cytochrome P450 CYP24A1, the
enzyme responsible for stereospecific 6-step oxidation of 1,25-dihydroxy vitamin D3. The active form of vitamin D3 regulates calcium
and phosphorus homeostasis and also stimulates cellular differentiation while inhibiting proliferation. Thus, CYP24A1 is an attractive
target for the development of cancer therapeutics.
A major effort to determine the basis of HIV resistance to antiviral drugs is ongoing in collaboration with
A. Olson and J.E. Elder, Department of Molecular Biology; B.E. Torbett, Department of Molecular and Experimental Medicine; M.G. Finn, Department of Chemistry;
and D.E. McRee, ActiveSight, San Diego, California. One
aspect of this project entails determining the crystal
structures of mutant proteases from drug-resistant HIV
in complex with broad-spectrum inhibitors; another
is the use of fragment-based screening to discover
new classes of inhibitors as lead compounds for drug
discovery. Additional research projects involve crystal
structures and mechanistic studies of an iron-sulfur
enzyme, adenosine-5′-phosphosulfate reductase, and
its homolog, 3′-phosphoadenosine-5′-phosphosulfate
reductase, in collaboration with K.S. Carroll, University of Michigan, Ann Arbor, Michigan.
Effect of Surface Engineering
on the Crystallization Properties
of the Integral Membrane Protein
Cytochrome ba3 From Thermus
thermophilus
J.A. Fee, B. Liu, V.M. Luna, C.D. Stout, Y. Chen
e are using the cytochrome ba3 from Thermus
thermophilus to increase our understanding
of how cytochrome c oxidases function. Cytochrome c oxidases catalyze the following deceptively
simple reaction:
W
4 cytochrome c2+ + O2 + 8 H+in g 4 cytochrome
c3+ + 2 H2O + 4 H+out
where the subscripts in and out refer, respectively, to
the cytoplasm and the periplasmic space of prokaryotic
cells. The free energy of dioxygen reduction is thus captured as a proton gradient; the out side is positive and
the in side is negative-proton pumping.
Recently, we altered the amino acid sequence of
cytochrome ba 3 from T thermophilus and made the
unexpected finding that the modified protein readily
and rapidly crystallizes, with x-ray diffraction to approximately 2.6–2.5 Å. Our long-range goals are to obtain
structures of intermediates that occur during dioxygen
202 MOLECULAR BIOLOGY
2007
THE SCRIPPS RESEARCH INSTITUTE
reduction and to relate the structures to the important, unsolved problem of proton pumping. For these
experiments, we will trap intermediates in crystals of
cytochrome ba 3 at subzero temperatures.
Central to achieving these goals is our homologous
expression system for cytochrome ba3 that permits easy
purification of the enzyme in amounts needed for physicochemical experimentation. The recombinant enzyme
crystallizes in the P43212 space group with considerable difficulty, and the crystals are not suitable for the
kinds of studies we envision. Native protein molecules
disposed in the unit cell are shown in Figure 1A. Careful
F i g . 2 . Lattice contacts of the structure of wild-type (A) and double-
mutant (B, I-Lys258Arg and II-Glu4Gln) cytochrome ba3 oxidase. In
the wild-type enzyme, 2 salt bridges occur between the symmetry
mates I-Lys258–I-Glu510 and I-Glu510–I-Lys258; the hydrogen bonds
involving II-Glu4 and I-Lys258 are also evident. In the mutant, 2 possible new hydrogen bonds are found between II-Gln4 and symmetrical
II-Glu96 and between I-R258 and symmetrical II-Asn93. Subunit I
and subunit II in the original molecule are pink and yellow, respectively. For the symmetrical mate, they are green and blue, respectively.
F i g . 1 . Packing contacts of the wild-type (A) and double-mutant
(B, I-Lys258Arg and II-Glu4Gln) cytochrome ba3 oxidase. Native protein molecules can be viewed as “linear dimers” in the unit cell (see
text), whereas doubly mutated protein molecules are seen as “angled
dimers” forming contacts with one another in a springlike structure.
Yellow indicates subunit I; magenta, subunit II; and pink, subunit IIa.
The transmembrane parts are defined by the barrels of transmembrane
helices. The paucity of hydrophilic contacts is evident.
examination of the protein-protein contacts revealed a
possibility for a unique symmetry-related protein-protein
ineraction within the dimeric pair of molecules related
by a 2-fold rotation (left-to-right at top and bottom in
Fig. 1A). As shown in Figure 2A, the positively charged
residue I-Lys258 and the negatively charged residue
I-Glu510 make a symmetrical salt bridge across the
dimer interface. However, the negatively charged residue II-Glu4 is also in close contact with I-Lys258,
adding an unbalanced negative charge that would be
expected to weaken the primary interaction between
the symmetry mates. We hypothesized that mutagenic
conversion of I-Lys258 to I-Arg258 and of II-Glu4 to
II-Gln4 may result in 2 strong, intermolecular salt
bridges between I-Arg258 and I-Glu510 and eliminate the unbalancing charge of II-Glu4.
The double-mutant protein (containing I-Arg258
and II-Gln4), obtained by now-standard techniques of
molecular biology, was fully active as a cytochrome c
oxidase and crystallized in 2–3 days compared with
approximately 1 month for recombinant cytochrome
ba3. X-ray crystallographic studies yielded a new struc-
MOLECULAR BIOLOGY
2007
ture of the enzyme. Remarkably, crystallization of the
double-mutant occurred in a different space group,
P41212, with distinctively different protein-protein contacts (Figs. 1B and 2B). Thus, I-Arg258 no longer
interacts with I-Glu510; instead it forms a hydrogen
bond with the carbonyl atom of II-Asn93, and II-Gln4
may form a hydrogen bond to II-Glu96.
Although none of our predictions were fulfilled,
the double-mutant protein provides an easy source of
crystals for x-ray diffraction, thereby opening many
possibilities for novel mechanistic studies.
Developing Reagents for Studies
of Membrane Proteins
Q. Zhang, M.G. Finn,* W.-X. Hong, R.S. Roy
* Department of Chemistry, Scripps Research
ntegral membrane proteins tend to lose stability
and activity outside the membrane bilayer, a situation that makes their biophysical and biochemical
characterization difficult. We have developed new membrane-mimicking systems to stabilize integral membrane
proteins for structural and functional studies.
We designed and synthesized a type of facial amphiphile derived from cholic acid that features increased
facial amphiphilicity and a hydrophobic skeleton tailored
to mimic the properties of cholesterol components of
real lipids (Fig. 1). Our structurally unique facial amphi-
I
THE SCRIPPS RESEARCH INSTITUTE
203
Crystallization of a variety of targets is being attempted,
in collaboration with Drs. Chang and Stevens and
M. Yeager, Department of Cell Biology.
Last, in collaborative studies with K. Wüthrich and
colleagues, Department of Molecular Biology, we have
characterized new reagents useful for the solution nuclear
magnetic resonance examination of integral membrane
proteins. A library of newly synthesized detergents has
been evaluated for support on refolding of the β-barrel
membrane protein OmpX. One class of lipidlike zwitterionic detergents appears to be superior, producing structure-quality nuclear magnetic resonance spectra. In each
of these efforts, different examples of new amphiphiles
appear to be useful, emphasizing the need for more
diverse reagents to be developed for this important
branch of structural biology.
PUBLICATIONS
Bieschke, J., Zhang, Q., Bosco, D.A., Lerner, R.A., Powers, E.T., Wentworth, P.,
Jr., Kelly, J.W. Small molecule oxidation products trigger disease-associated protein
misfolding. Acc. Chem. Res. 39:611, 2006.
Stewart, C.R., Wilson, L.M., Zhang, Q., Pham, C.L.L., Waddington, L., Staples,
M.K., Stapleton, D., Kelly, J.W., Howlett, G.J. Oxidized cholesterol metabolites
found in human atherosclerotic lesions promote apolipoprotein C-II amyloid fibril
formation. Biochemistry 46:5552, 2007.
Zhang, Q., Ma, X., Ward, A., Hong, W.-X., Jaakola, V.-P., Stevens, R.C., Finn,
MG., Chang, G. Designing facial amphiphiles for the stabilization of integral membrane proteins. Angew. Chem. Int. Ed., in press.
Structural Neurobiology and
Development of Protein
Therapeutic Agents
R.C. Stevens, E.E. Abola, A.I. Alexandrov, H.M. Archer,
J.W. Arndt, G.A. Asmar-Rovira, R.R. Benoit, M.H. Bracey,
A. Brooun, Q. Chai, V.G. Cherezov, E. Chien, S. Daudenarde,
J. Dupuy, A. Gámez, M.T. Griffith, M.A. Hanson, V.-P. Jaakola,
F i g . 1 . New design of facial amphiphiles derived from cholic
J.S. Joseph, K. Masuda, M. Mileni, K. Moy, J. Ng, C. Roth,
K.S. Saikatendu, V. Subramanian, J. Velasquez, L. Wang,
acid and their proposed binding mode by integral membrane proteins. The membranous sector in deep blue is shielded by the flat
hydrophobic surface of facial amphiphiles.
B. Wu, Q. Zhao
philes have binding properties distinct from those of
the classical head-to-tail detergents, as evidenced by
the formation of smaller and stronger membrane protein complexes by the synthetic amphiphiles.
In collaboration with G. Chang and R.C. Stevens,
Department of Molecular Biology, we have shown that
the designed facial amphiphiles substantially stabilize
several well-known integral membrane protein systems.
n the past, out of frustration with the rate at which
information on structural biology was emerging, we
focused on developing new tools to change the field
by accelerating the rate of determination of protein structures. This endeavor included pioneering microliter
expression/purification for structural studies, nanovolume
crystallization, automated image collection, and robotics
for collecting synchrotron beamline data. These tech-
HIGH-THROUGHPUT STRUCTURAL BIOLOGY
I
204 MOLECULAR BIOLOGY
2007
nologies were initially tested by the team at the Joint
Center for Structural Genomics (http://www.jcsg.org)
in collaboration with I. Wilson, Department of Molecular
Biology, where the power of the new tools was demonstrated. To advance the technologies toward more challenging protein complexes and membrane proteins, in
collaboration with P. Kuhn, Department of Cell Biology, we have created 2 new technology-focused centers funded by the National Institutes of Health.
The first center is the Joint Center for Innovative Membrane Protein Technologies (http://jcimpt.scripps.edu).
Here, in collaboration with K. Wüthrich, Q. Zhang, and
G. Chang, Department of Molecular Biology; M.G. Finn,
Department of Chemistry; and P. Kuhn and M. Yeager,
Department of Cell Biology, we do research exclusively
on membrane proteins, including G protein–coupled
receptors. The second center is the Accelerated Technologies Center for Gene to 3D Structure (http://www
.atcg3d.org). Here we are doing collaborative work with
Dr. Kuhn and with researchers from deCODE biostructures, Bainbridge Island, Washington; Lyncean Technologies, Palo Alto, California; and the University of
Chicago, Chicago, Illinois.
In 2005, we showed that high-resolution electron
density maps and refined models can be obtained from
in situ diffraction of crystals grown in microcapillaries.
In 2008, the first laboratory-sized synchrotron will be
installed at Scripps Research. The synchrotron has
performance characteristics comparable to those of a
synchrotron beam in terms of intensity and tunability
and will enable us to use direct diffraction analysis of
ongoing in situ crystallization experiments to accelerate determination of macromolecular structures.
STRUCTURAL NEUROBIOLOGY
Although we have developed high-throughput methods to accelerate the determination of protein structures,
our primary interest is using these tools to study the
chemistry and biology of neurotransmission and of
diseases that affect neurons, particularly childhood
neurologic disorders. Our goals are to understand how
neuronal cells function on a molecular level and, on
the basis of that understanding, create new molecules
and materials that mimic neuronal signal transduction
and recognition.
BIOSYNTHESIS OF NEUROTRANSMITTERS
For neuronal signal transduction, the presynaptic
cell synthesizes neurotransmitters that then traverse
the synaptic cleft. We are using the high-throughput
methods to determine the inclusive structures of complete biochemical pathways. Specifically, we are inter-
THE SCRIPPS RESEARCH INSTITUTE
ested in determining the structures of all the enzymes
in the biosynthesis pathways of neurotransmitters in
order to understand the mechanistic details of each
individual enzymatic reaction at the atomic level. This
approach also allows us to determine the best path of
drug discovery for the biosynthesis of neurotransmitters.
T H E R A P E U T I C A G E N T S F O R T R E AT M E N T O F
CHILDHOOD PHENYLKETONURIA
In addition to the basic questions under investigation about neurotransmitter biosynthesis, recent clinical
studies suggest that some patients with the metabolic
disorder phenylketonuria are responsive to (6R)- L erythro-5,6,7,8-tetrahydrobiopterin, the natural cofactor
of phenylalanine hydroxylase. We are collaborating with
scientists at BioMarin Pharmaceuticals Inc., Novato,
California, to correlate how structure can be used to
predict which patients with phenylketonuria most likely
will respond to treatment with this cofactor. Phase 3
clinical trials for the treatment of phenylketonuria with
the cofactor, called Kuvan, have been completed.
For classical phenylketonuria, we are developing
an enzyme replacement therapeutic agent that is currently in preclinical development. The therapy is based
on administration of a modified form of phenylalanine
ammonia lyase discovered in our structural studies
(Fig. 1). Last, we are determining the structural basis
of diseases caused by several other enzymes involved
in the biosynthesis of neurotransmitters. Many of these
disorders are rare or occur during childhood.
BOTULINUM NEUROTOXINS
Clostridial neurotoxins are responsible for disrupting neurotransmission. They include tetanus toxin and
the 7 serotypes of botulinum toxin. We are determining the molecular events involved in the binding, pore
formation, translocation, and catalysis of botulinum
neurotoxin. Although botulinum toxin is most known
for its deadly effects, it is now being used therapeutically to treat involuntary muscle disorders such as cerebral palsy and neuromuscular dystonias. Previously,
we determined the structures of the 150-kD holotoxin
form, the holotoxin bound to antibodies, and the catalytic
domains of several serotypes (A, B, D, F, G). Recently,
we determined the structure of the cell-surface receptor toxin complex (Fig. 2). These structures are being
used to understand and redesign the toxin’s mechanism of action and to determine additional therapeutic
applications of the toxin.
CANNABINOID SIGNALING
In collaboration with B.F. Cravatt, Department of
Cell Biology, we solved the structure of fatty acid amide
MOLECULAR BIOLOGY
2007
F i g . 1 . Crystal structures of phenylalanine ammonia lyase (PAL)
tetramers. The structure of wild-type Rhodosporidium toruloides
PAL (RtPAL) and the Anabaena variabilis PAL (AvPAL) C503S/C563S
double mutant were determined at 1.6- and 2.2-Å resolution, respectively. A, Tetramer structures of RtPAL and AvPAL with each monomer
in a different color. B, The bottom view (90° rotation of the view in A)
of the same structures with the prosthetic group 4-methylideneimidazole-5-one in the 5 active sites as red sticks. These proteins
were engineered and chemically modified as potential once-a-week
injectable therapeutic agents for treatment of phenylketonuria.
hydrolase, a degradative integral membrane enzyme
responsible for setting intracellular levels of endocannabinoids, to 2.8 Å. Fatty acid amide hydrolase is intimately
associated with CNS signaling processes such as retrograde synaptic transmission, a process that is also modulated by the illicit substance 9∆-tetrahydrocannabinol.
With our knowledge of the 3-dimensional structure, we
are trying to understand how the enzyme works at a
basic level and how it might be the basis for potential
drug discovery.
THE SCRIPPS RESEARCH INSTITUTE
205
F i g . 2 . A, Model of binding of type B botulinum neurotoxin via
both Syt-II and ganglioside receptors at the presynaptic membrane.
The cytoplasm domain of synaptotagmin is presented as 2 C2
domains (light pink) bound with calcium ions (yellow spheres),
modeled by 1TJX and 1DQV. Reprinted from Chai, Q., Arndt, J.W.,
Dong, M., Tepp, W.H., Johnson, E.A., Chapman, E.R., Stevens, R.C.
Structural basis of cell surface receptor recognition by botulinum
neurotoxin B. Nature 444:1096, 2006. B, Overview of type A1
botulinum neurotoxin (yellow) in complex with the CR1 antibody
(antibody heavy and light chains in magenta and green, respectively).
Reprinted from Garcia-Rodriguez, C., Levy, R., Arndt, J.W., Forsyth,
C.M., Razai, A., Lou, J., Geren, I., Stevens, R.C., Marks, J.D. Molecular evolution of antibody cross-reactivity for two subtypes of type
A botulinum neurotoxin. Nat. Biotechnol. 25:107, 2007.
G PROTEIN–COUPLED RECEPTORS
We are also trying to determine the 3-dimensional
structure of G protein–coupled receptors. These receptors are the largest mammalian protein family known
and are key signaling molecules for neuronal signal
transmission, viral entry, vision, and smell. Using novel
technologies that we have developed, we hope to produce a critical breakthrough in the coming year.
Abola, E., Kuhn, P., Stevens, R.C. Miniaturization of the structural biology technologies: from expression to biophysical analyses. In: Structural Genomics on Membrane
Proteins. Lundstrom, K.H. (Ed.). CRC Press, Boca Raton, FL, 2006, p. 261.
PUBLICATIONS
Abola, E., Carlton, D.D., Kuhn, P., Stevens, R.C. Five years of increasing structural
biology throughput: a retrospective analysis. In: Structure-based Drug Discovery.
Jhoti, H., Leach, A. (Eds.). Springer, New York, 2007, p. 1.
Gámez, A., Wang, L., Sarkissian, C.N., Wendt, D., Fitzpatrick, P., Lemontt, J.F.,
Scriver, C.R., Stevens, R.C. Structure-based epitope and PEGylation sites mapping
of phenylalanine ammonia-lyase for enzyme substitution treatment of phenylketonuria Mol. Genet. Metab., in press.
Chai, Q., Arndt, J.W., Dong, M., Tepp, W.H., Johnson, E.A., Chapman, E.R.,
Stevens, R.C. Structural basis of cell surface receptor recognition by botulinum
neurotoxin B. Nature 444:1096, 2006.
DiDonato, M., Krishna, S.S., Schwarzenbacher, R., et al. Crystal structure of 2-phosphysulfolactate phosphatase (ComB) from Clostridium acetobutylicum at 2.6 Å resolution reveals a new fold with a novel active site. Proteins 65:771, 2006.
206 MOLECULAR BIOLOGY
2007
Garcia-Rodriguez, C., Levy, R., Arndt, J.W., Forsyth, C.M., Razai, A., Lou, J., Geren,
I., Stevens, R.C., Marks, J.D. Molecular evolution of antibody cross-reactivity for two
subtypes of type A botulinum neurotoxin. Nat. Biotechnol. 25:107, 2007.
Gerdts, C.J., Tereshko, V., Yadav, M.K., Dementieva, I., Collart, F., Joachimiak,
A., Stevens, R.C., Kuhn, P., Kossiakoff, A., Ismagilov, R.F. Time-controlled
microfluidic seeding in nL-volume droplets to separate nucleation and growth
stages of protein crystallization. Angew. Chem. Int. Ed. 45:8156, 2006.
Hanson, M.A., Brooun, A., Baker, K.A., Jaakola, V.-P., Roth, C., Chien, E.,
Alexandrov, A., Velasquez, J., Davis, L., Griffith, M., Moy, K., Ganser-Pornillos,
B., Kuhn, P., Ellis, S., Yeager, M., Stevens, R.C. Profiling of membrane protein
variants in a baculovirus system by coupling cell-surface detection with small-scale
parallel expression. Prot. Expr. Purif., in press.
Joseph, J.S., Saikatendu, K.S., Subramanian, V., Neuman, B.W., Buchmeier,
M.J., Stevens, R.C., Kuhn, P. Crystal structure of a monomeric form of severe
acute respiratory syndrome coronavirus endonuclease nsp15 suggests a role for
hexamerization as an allosteric switch. J. Virol. 81:6700, 2007.
Kosloff, M., Han, G.W., Krishna, S.S., et al. Comparative structural analysis of a
novel glutathione S-transferase (ATU5508) from Agrobacterium tumefaciens at
2.0 Å resolution. Proteins 65:527, 2006.
Saikatendu, K.S., Joseph, J.S., Subramanian, V., Neuman, B.W., Buchmeier, M.J.,
Stevens, R.C., Kuhn, P. Ribonucleocapsid formation of severe acute respiratory syndrome coronavirus through molecular action of the N-terminal domain of N protein.
J. Virol. 8:3913, 2007.
Stevens, R.C. SPINE Forward. Acta Crystallogr. D62:0-0, 2006.
Swaminathan, S., Stevens, R.C. Three-dimensional protein structures of light chains
of botulinum neurotoxin serotypes A, B, and E and tetanus neurotoxin. In: Treatments
From Toxins: The Therapeutic Potential of Clostridial Neurotoxins. Foster, K.A., Hambleton, P., Shone, C.C. (Eds.). CRC Press, Boca Raton, FL, 2007, p. 19.
Wang, L., Surendran, S., Michals-Matalon, K., Bhatia, G., Tanskley, S., Koch, R.,
Grady, J., Tyring, S.K., Stevens, R.C., Guttler, F., Matalon, R. Mutations in the regulatory domain of phenylalanine hydroxylase and response to tetrahydrobiopterin. Genet.
Test. 11:174, 2007.
Xu, Q., Krishna, S.S., McMullan, D., et al. Crystal structure of an ORFan protein
(TM1622) from Thermotoga maritima at 1.75 Å resolution reveals a fold similar to
the Ran-binding protein Mog1p. Proteins 65:777, 2006.
High-Throughput Approaches to
Protein Structure and Function
S.A. Lesley, L. Columbus, H. Johnson, S. Sudek,
T. Janaratne, M. Deller, D. Carlton, Y. Elias, T. Clayton,
T. Trout, A. Grzechnik
xamining protein structure and function is of primary importance for understanding the basic
biology of the cell and is a challenge because
of the constantly expanding wealth of genomic information. To address this challenge, we have established
high-throughput approaches for evaluating structural
and functional diversity of proteins as part of a structural
genomics effort with the Joint Center for Structural
Genomics. We use these same tools to characterize the
molecular basis of the specificity of enzyme substrates.
The goals of the Joint Center for Structural Genomics are to develop a high-throughput and cost-effective
E
THE SCRIPPS RESEARCH INSTITUTE
structure pipeline and to use the pipeline to determine
novel protein folds and explore protein structure-function relationships. We have used this approach in an
extensive study of the thermophilic bacterium Thermotoga maritima and for targets from mouse and more
than 100 other bacterial genomes. Our technologies
have enabled us to perform comprehensive structural
studies of these proteomes. To date, these efforts have
resulted in more than 450 novel protein structures from
the center. We have also used structural data and performed biochemical assays to validate predicted activities and determine function for numerous targets for
which no function could be predicted.
In order to understand how genome sequences are
related to the biology of an organism, correct annotation of gene function is essential. Many computational
approaches are available for assigning a putative gene
function on the basis of similarity to known activities.
However, as the evolutionary distances extend and the
number of putative homologs increases, the reliability of
such unvalidated predictions becomes suspect. Likewise, although enzyme classes and broad activities
can be predicted with some certainty, much less can
be inferred about substrate specificity. This specificity
defines the role of an enzyme at a functional level, and
understanding enzyme selectivity is essential to understanding the enzyme in the biology of an organism.
In collaboration with A. Osterman, the Burnham
Institute for Medical Research, La Jolla, California, and
B. Geierstanger, Genomics Institute of the Novartis
Research Foundation, San Diego, we are undertaking
an approach to such characterization that can be
applied generally and in a high-throughput fashion
to proteins with unknown or putative functions. The
approach consists of a bioinformatic platform to analyze gene relationships and propose putative functions
for testing, use of an existing high-throughput platform
to produce thousands of proteins in milligram quantities for analysis, ligand screening technologies combined with metabolite-focused compound libraries to
define specificity profiles of enzymes, and focused experimental validation of proposed functions of key target
genes and pathways.
PUBLICATIONS
Columbus, L., Lipfert, J., Klock, H., Millett, I., Doniach, S., Lesley, S.A. Expression, purification, and characterization of Thermotoga maritima membrane proteins
for structure determination. Protein Sci. 15:961, 2006.
Eshaghi, S., Niegowski, D., Kohl, A., Martinez Molina, D., Lesley, S.A., Nordlund, P.
Crystal structure of a divalent metal ion transporter CorA at 2.9 angstrom resolution
[published correction appears in Science 313:1389, 2006]. Science 313:354, 2006.
MOLECULAR BIOLOGY
2007
Mason, A., Agrawal, N., Washington, M.T., Lesley, S.A., Kohen, A. A lag-phase in
the reduction of flavin dependent thymidylate synthase (FDTS) revealed a mechanistic missing link. Chem. Commun. (Camb.) 1781, Issue 16, 2006.
Nuclear Magnetic Resonance
Spectroscopy, Chaperonins,
and Structural Genomics
K. Wüthrich, W. Augustyniak, A. Chatterjee, M. Geralt,
R. Horst, M. Johnson, B. Pedrini, W.J. Placzek, J.K. Rhee,
THE SCRIPPS RESEARCH INSTITUTE
207
on a fully deuterated background, this experiment
resulted in backbone resonance assignments and the
identification of conformational constraints for GroES
in the 472-kD SR1-GroES complex.
We currently are extending the research on the GroEsystem of E coli, which must be studied at temperatures
near 25°C, to the GroEL system and its substrate proteins in the extreme thermophile Thermus thermophilus,
which can be studied at 60°C or possibly even at higher
temperatures. This change led to a remarkable improvement of the spectral resolution (Fig. 1), and we are now
P. Serrano
e focus on 2 areas of research: chaperonins
and structural genomics. For both areas, we
use and develop methods of nuclear magnetic
resonance (NMR) spectroscopy.
W
CHAPERONINS
In collaboration with A. Horwich, Yale University
and Howard Hughes Medical Institute, New Haven, Connecticut, who is a visiting scientist at Scripps Research,
we are investigating structural and mechanistic aspects
of the function of GroE-type chaperonin systems in
Escherichia coli. This research concerns protein folding in healthy and diseased organisms and thus is
directly related to the currently extensively discussed
protein misfolding diseases.
We used transverse relaxation-optimized spectroscopy (TROSY) and cross-correlated relaxation-enhanced
polarization transfer (CRINEPT) to study complexes
composed of the stable isotope-labeled cochaperonin
GroES (72 kD) with the unlabeled chaperonin GroEL
(800 kD) or its single-ring variant SR1 (400 kD). We
found that informative [15N,1H]-correlation spectra can
be obtained for GroES in these complexes; the molecular weights were 872 kD for the GroES-GroEL complex
and 472 kD for the GroES-SR1 complex. We then started
work on sequence-specific resonance assignments for
GroES in these complexes, which are mandatory for
further, more detailed analysis. This research required
the development of new NMR approaches because
routine procedures are not sufficiently sensitive for these
large structures. We found that 3-dimensional [1H,1H]nuclear Overhauser effect–[15N, 1H]–CRINEPT–heteronuclear multiple quantum correlation spectroscopy could
be used to detect proton-proton nuclear Overhauser
effect connectivities for the 2 H, 15 N-labeled GroES in
1:1 complexes with the chaperonins. In combination
with uniform and residue-selective 15N labeling of GroES
F i g . 1 . Two-dimensional [15N,1H]-TROSY correlation spectrum of
T thermophilus GroEL recorded with a sample volume of 300 µL
on a Bruker AVANCE 800 spectrometer at 60°C and a 5-mm roomtemperature probehead. The overall measuring time was 2 hours.
working on the sequence-specific NMR assignment of
the 800-kD T thermophilus GroEL. Our results suggests
that previously inaccessible detailed information on
chaperonin systems may result from studies of the
T thermophilus proteins.
STRUCTURAL GENOMICS
In our studies in structural genomics, we participate
in the Joint Center for Structural Genomics, the Joint
Center for Innovative Membrane Protein Technologies,
and the consortium for Functional and Structural Proteomics Analysis of SARS-CoV–Related Proteins. Our
activities with the use of microcoil NMR equipment
combined with microexpression of proteins for quality
screening of recombinant proteins are illustrated here
with studies of membrane proteins.
We previously established a miniaturized pipeline
in which 1-dimensional 1H NMR spectroscopy is used
for automated screening of newly synthesized polypeptides for the presence of globular domains. Although this
approach is now in routine use for soluble proteins with
208 MOLECULAR BIOLOGY
molecular weights up to about 30 kD, initial experiments
with detergent-solubilized membrane proteins showed
that the use of this type of spectroscopy is limited by
the background signals from the nondeuterated detergents, which partially overlap with the protein signals.
Furthermore, it turned out that the high molecular weight
of the mixed protein-detergent micelles requires the
use of TROSY-based NMR experiments in combination
with uniform 2H,15N-labeling of the membrane proteins.
Using this approach and new detergents synthesized by
Q. Zhang, Department of Molecular Biology, we solubilized the E coli outer-membrane protein X (OmpX).
We then used the resulting samples to obtain [15N,1H]NMR correlation maps as diagnostic fingerprints of the
conformational state of OmpX reconstituted in mixed
micelles with these detergents. Figure 2 shows NMR
2007
THE SCRIPPS RESEARCH INSTITUTE
aforementioned technology, we expect to contribute to
a future widening of this bottleneck.
PUBLICATIONS
Almeida, M.S., Johnson, M.A., Wüthrich, K. NMR assignment of the SARS-CoV
protein nsp1. J. Biomol. NMR 36(Suppl.1):46, 2006.
Baker, K.A., Hilty, C., Peti, W., Prince, A., Pfaffinger, P.J., Wider, G., Wüthrich,
K., Choe, S. NMR-derived dynamic aspects of N-type inactivation of a Kv channel
suggest a transient interaction with the T1 domain. Biochemistry 45:1663, 2006.
Etezady-Esfarjani, T., Herrmann, T., Horst, R., Wüthrich, K. Automated protein
NMR structure determination in crude cell-extract. J. Biomol. NMR 34:3, 2006.
Etezady-Esfarjani, T., Placzek, W.J., Herrmann, T., Wüthrich, K. Solution structures
of the putative anti-σ-factor antagonist TM1442 from Thermotoga maritima in the free
and phosphorylated states. Magn. Reson. Chem. 44(Spec. No.):S61, 2006.
Horst, R., Wider, G., Fiaux, J., Bertelsen, E.B., Horwich, A.L., Wüthrich, K. Proton-proton Overhauser NMR spectroscopy with polypeptide chains in large structures. Proc. Natl. Acad. Sci. U. S. A. 103:15445, 2006.
Johnson, M.A., Peti, W., Herrmann, T., Wilson, I.A., Wüthrich, K. Solution structure of As11650, an acyl carrier protein from Anabaena sp. PCC 7120 with a variant phosphopantetheinylation-site. Protein Sci. 15:1030, 2006.
Placzek, W.J., Almeida, M.A., Wüthrich, K. NMR assignment of a human cancerrelated nucleoside triphosphatase. J. Biomol. NMR 36(Suppl. 1):59, 2006.
Serrano, P., Almeida, M.S., Johnson, M.A., Wüthrich, K. NMR assignment of the
protein nsp3a from SARS-CoV. J. Biomol. NMR 36(Suppl.1):45, 2006.
F i g . 2 . Two-dimensional [ 15N, 1H]-TROSY correlation spectra of
2H, 15N-labeled
OmpX reconstituted in different detergents: A, 138Fos. B, 179-Fos. C, 34-Fos. D, 185-Fos. The spectra were recorded
with sample volumes of 7 µL on a Bruker DRX-700 spectrometer
at 25°C with a 1-mm room-temperature microcoil probehead. The
overall measurement time per experiment was 9 hours.
spectra recorded with 7-µL samples of solutions containing OmpX in mixed micelles with the detergents
138-Fos and 179-Fos (panels A and B), where the
protein is uniformly well folded, and with 34-Fos and
185-Fos (panels C and D), where the protein forms
nonspecific aggregates. As a result of this study, we
have identified 2 new detergents with promising properties for the preparation of membrane proteins for
structural studies.
The preparation of integral membrane proteins for
structural biology and structural genomics is a bottleneck that has limited structure determinations by x-ray
crystallography or NMR spectroscopy to only about 100
structures, as compared with over 30,000 structures
of soluble proteins. By systematic screening of new
membrane protein–detergent combinations with the
Nuclear Magnetic Resonance of
3-Dimensional Structure and
Dynamics of Proteins in Solution
P.E. Wright, H.J. Dyson, M. Arai, R. Burge, J. Ferreon,
T.-H. Huang, B.A. Buck-Koehntop, M. Kostic, B. Lee,
C.W. Lee, M. Landes, M. Martinez-Yamout, T. Nishikawa,
K. Sugase, J. Wojciak, M. Zeeb, E. Manlapaz, L.L. Tennant,
J. Chung, D.A. Case, J. Gottesfeld, R. Evans,* M. Montminy*
* Salk Institute for Biological Studies, La Jolla, California
e use multidimensional nuclear magnetic
resonance (NMR) spectroscopy to investigate
the structures, dynamics, and interactions of
proteins in solution. Such studies are essential for understanding the mechanisms of action of these proteins
and for elucidating structure-function relationships. The
focus of our current research is protein-protein and
protein–nucleic acid interactions involved in the regulation of gene expression.
W
T R A N S C R I P T I O N FA C T O R – N U C L E I C A C I D C O M P L E X E S
NMR methods are being used to determine the
3-dimensional structures and intramolecular dynamics
of zinc finger motifs from several eukaryotic transcriptional regulatory proteins, both free and complexed with
MOLECULAR BIOLOGY
2007
target nucleic acid. Zinc fingers are among the most
abundant domains in eukaryotic genomes. They play
a central role in the regulation of gene expression at
both the transcriptional and the posttranscriptional level,
mediated through their interactions with DNA, RNA,
or protein components of the transcriptional machinery.
The C2H 2 zinc finger, first identified in transcription
factor IIIA (TFIIIA), is used by numerous transcription
factors to achieve sequence-specific recognition of DNA.
Growing evidence, however, indicates that some C 2H 2
zinc finger proteins control gene expression both through
their interactions with DNA regulatory elements and,
at the posttranscriptional level, through binding to RNA.
The best-characterized example of a C2H2 zinc finger
protein that binds specifically to both DNA and to RNA
is TFIIIA, which contains 9 zinc fingers. We showed
previously that different subsets of zinc fingers are
responsible for high-affinity binding of TFIIIA to DNA
(fingers 1–3) and to 5S RNA (fingers 4–6). To obtain
insights into the mechanism by which the TFIIIA zinc
fingers recognize both DNA and RNA, we have used
NMR methods to determine the structures of the complex formed by zf1-3 (a protein containing fingers
1–3) with DNA and by zf4-6 (a protein consisting of
fingers 4–6) with a fragment of 5S RNA.
Three-dimensional structures were determined previously for the complex of zf1-3 with the cognate 15-bp
oligonucleotide duplex. The structures contain several
novel features and reveal that prevailing models of DNA
recognition, which assume that zinc fingers are independent modules that contact bases through a limited
set of amino acids, are outmoded.
In addition to its role in binding to and regulating
the 5S RNA gene, TFIIIA also forms a complex with the
5S RNA transcript. NMR structures of the complex
formed by zinc fingers 4–6 with a truncated form of 5S
RNA have been completed and give important insights
into the structural basis for 5S RNA recognition. Finger 4 of the protein recognizes both the structure of
the RNA backbone and the specific bases in the loop
E motif of the RNA, in a classic lock-and-key interaction.
Fingers 5 and 6, with a single residue between them,
undergo mutual induced-fit folding with the loop A region
of the RNA, which is highly flexible in the absence of
the protein.
NMR studies of 2 alternate splice variants of the
Wilms tumor zinc finger protein (WT1) are in progress.
These proteins differ only through insertion of 3 additional amino acids (the tripeptide lysine-threonine-serine)
THE SCRIPPS RESEARCH INSTITUTE
209
in the linker between fingers 3 and 4, yet have marked
differences in their DNA-binding properties and subcellular localization. 15N relaxation measurements indicate
that the insertion increases the flexibility of the linker
between fingers 3 and 4 and abrogates binding of the
fourth zinc finger to its cognate site in the DNA major
groove, thereby modulating DNA-binding activity. X-ray
and NMR structures of the complexes of the WT1 zinc
fingers with 14- and 17-bp DNA oligonucleotides have
been determined. Zinc fingers 2–4 are inserted deeply
into the DNA major groove, making sequence-specific
contacts with bases. The structure provides insights
into the mechanism by which disease-causing mutations
in the zinc finger domain interfere with DNA binding. In
contrast to fingers 2–4, zinc finger 1 has mostly nonspecific interactions with the DNA. High-affinity DNA
binding is mediated by fingers 2–4; incorporation of
additional amino acids in the linker by alternate splicing disrupts the finger 4 interactions and abrogates
DNA binding (Fig. 1).
F i g . 1 . Ribbon diagram of the composite NMR/x-ray structures
of the complex of WT1 zinc fingers 1–4 and the 17-bp DNA duplex
in solution.
NMR structural studies of a complex of the 4 WT1
zinc fingers with an RNA aptamer are nearing completion. In contrast to DNA binding, the RNA interaction
is dominated by zinc fingers 1–3, which bind in the
widened major groove formed in the vicinity of a bulged
210 MOLECULAR BIOLOGY
2007
base. The interactions of zinc finger 4 with the RNA loop
make only a secondary contribution to binding affinity.
We have also determined the structure of a novel double-stranded RNA-binding zinc finger protein and have
commenced experiments to define the mechanism of
binding to adenovirus VA1 RNA.
We recently determined the structure of a novel
zinc finger protein termed Churchill that is involved
in regulation of neural induction during embryogenesis. At the time of its discovery, it was suggested that
the protein contained 2 zinc fingers of the C4 type and
functioned as a DNA-binding transcription factor. Our
NMR structure shows that far from containing canonical
C 4 zinc fingers, Churchill contains 3 bound zinc ions
in novel coordination sites, including an unusual binuclear zinc cluster, which jointly stabilize a single-layer
β-sheet. We showed further that Churchill does not bind
DNA and suggest that it may function in embryogenesis by mediating protein-protein interactions.
THE SCRIPPS RESEARCH INSTITUTE
experiments to elucidate the mechanism by which folding of the kinase-inducible activation domain of CREB
is coupled to binding to its KIX target domain. These
experiments revealed formation of an ensemble of transient and largely unfolded encounter complexes at multiple sites on the surface of KIX. The encounter complexes
are stabilized primarily by nonspecific hydrophobic contacts and evolve via an intermediate to the fully bound
state without dissociation from KIX. The C-terminal
helix of pKID is only partially folded in the intermediate
and becomes stabilized by intermolecular interactions
formed in the final bound state. Future applications of
our method will provide new understanding of the molecular mechanism by which intrinsically disordered proteins perform their diverse biological functions (Fig. 2).
PROTEIN-PROTEIN INTERACTIONS IN
T R A N S C R I P T I O N A L R E G U L AT I O N
Transcriptional regulation in eukaryotes relies on
protein-protein interactions between DNA-bound factors
and coactivators that, in turn, interact with the basal
transcription machinery. The transcriptional coactivator CREB-binding protein (CBP) and its homolog p300
play an essential role in cell growth, differentiation, and
development. Understanding the molecular mechanisms
by which CBP and p300 recognize their various target
proteins is of fundamental biomedical importance. CBP
and p300 have been implicated in diseases such as
leukemia, cancer, and mental retardation and are novel
targets for therapeutic intervention.
We previously determined the structure of the phosphorylated kinase-inducible activation domain (pKID) of
the transcription factor CREB bound to its target domain
(the KIX domain) in CBP. Ongoing work is directed
toward mapping the interactions between KIX and the
transcriptional activation domains of the proto-oncogene
c-Myb and of the mixed-lineage leukemia protein. The
solution structure of the ternary complex between KIX,
c-Myb, and the mixed-lineage leukemia protein has been
completed and provides insights into the structural basis
for the ability of the KIX domain to interact simultaneously and allosterically with 2 different effectors. Our
work has also provided new understanding of the thermodynamics of the coupled folding and binding processes involved in interaction of KIX with transcriptional
activation domains. We used R 2 relaxation dispersion
F i g . 2 . 15N R 2 relaxation dispersion profile for Arg124 of pKID
recorded at 800 MHz (filled circles) and 500 MHz (open circles).
Dispersion curves for 1 mM [ 15N]-pKID in the presence of 0.95,
1.00, 1.05, and 1.10 mM KIX are shown.
Recently, we determined the structure of the complex between the hypoxia-inducible factor Hif-1α and
the TAZ1 domain of CBP. The interaction between
Hif-1α and CBP/p300 is of major therapeutic interest
because of the central role Hif-1α plays in tumor progression and metastasis; disruption of this interaction
leads to attenuation of tumor growth. A protein named
CITED2 functions as a negative feedback regulator of the
hypoxic response by competing with Hif-1α for binding
to the TAZ1 domain of CBP. By determining the structure of the complex, we showed that the intrinsically
unstructured Hif-1α and CITED2 domains use partly
overlapping surfaces of the TAZ1 motif to achieve highaffinity binding and compete effectively with each other
for CBP/p300.
To further elucidate the molecular and structural
basis for CBP-dependent coordinated gene expression,
we have determined the solution structures of the com-
MOLECULAR BIOLOGY
2007
plexes formed by the transactivation domains of the
transcription factors STAT2 and STAT1 with CBP TAZ1
and TAZ2 domains, respectively. Despite the overall topological similarity of the CBP TAZ domains, the structures
reveal 2 very different modes of complex formation. Our
findings suggest that TAZ1 may bind activation domains
capable of contacting multiple surface grooves simultaneously in preference to smaller activation motifs
that are restricted to a single, contiguous binding surface. The latter mode of binding is sufficient for stable
complex formation with TAZ2. Binding of both STAT
activation domains involves coupled folding and binding processes.
We are continuing to map the multiplicity of interactions between CBP/p300 domains and their numerous biological targets. Our goal is to understand the
complex interplay of interactions that mediate key biological processes in health and disease.
PUBLICATIONS
Lee, B.M., Buck-Koehntop, B.A., Martinez-Yamout, M.A., Dyson, H.J., Wright,
P.E. Embryonic neural inducing factor Churchill is not a DNA-binding zinc finger
protein: solution structure reveals a solvent-exposed β-sheet and zinc binuclear
cluster. J. Mol. Biol., in press.
Stoll, R., Lee, B.M., Debler, E.W., Laity, J.H., Wilson, I.A., Dyson, H.J., Wright,
P.E. Structure of the Wilms tumor suppressor protein zinc finger domain bound to
DNA. J. Mol. Biol., in press.
Sugase, K., Dyson, H.J., Wright, P.E. Mechanism of coupled folding and binding
of an intrinsically unstructured protein. Nature 447:1021, 2007.
Sun, P., Yoshizuka, N., New, L., Moser, B.A., Li, Y., Liao, R., Xie, C., Chen, J.,
Deng, Q., Yamout, M., Dong, M.Q., Frangou, C.G., Yates, J.R. III, Wright, P.E.,
Han. J. PRAK is essential for ras-induced senescence and tumor suppression. Cell
128:295, 2007.
Folding of Proteins and
Protein Fragments
P.E. Wright, H.J. Dyson, D. Meinhold, C. Nishimura,
D. Felitsky, M. Kostic, S.J. Park, J. Chung, L.L. Tennant,
V. Bychkova,* T. Yamagaki
* Institute of Protein Research, Puschino, Russia
he molecular mechanism by which proteins fold
into their 3-dimensional structures remains one
of the most important unsolved problems in structural biology. Nuclear magnetic resonance (NMR) spectroscopy is uniquely suited to provide information on
the structure of transient intermediates formed during
protein folding. Previously, we used NMR methods to
show that many peptide fragments of proteins tend to
adopt folded conformations in water solution. The
T
THE SCRIPPS RESEARCH INSTITUTE
211
presence of transiently populated folded structures,
including reverse turns, helices, nascent helices, and
hydrophobic clusters, in water solutions of short peptides has important implications for initiation of protein
folding. Formation of elements of secondary structure
probably plays an important role in the initiation of
protein folding by reducing the number of conformations that must be explored by the polypeptide chain
and by directing subsequent folding pathways.
A P O M Y O G L O B I N F O L D I N G PAT H WAY
A major program in our laboratory is directed
toward a structural and mechanistic description of
the apomyoglobin folding pathway. Previously we used
quenched-flow pulse-labeling methods in conjunction
with 2-dimensional NMR spectroscopy to map the
kinetic folding pathway of the wild-type protein. With
these methods, we showed that an intermediate in
which the A, G, and H helices and part of the B helix
adopt hydrogen-bonded secondary structure is formed
within 6 milliseconds of the initiation of refolding.
Folding then proceeds by stabilization of additional
structure in the B helix and in the C and E helices.
We are using carefully selected myoglobin mutants
and both optical stopped-flow spectroscopy and NMR
methods to further probe the kinetic folding pathway.
For some of the mutants studied, the changes in amino
acid sequence resulted in changes in the folding pathway of the protein. These experiments are providing
novel insights into both the local and long-range interactions that stabilize the kinetic folding intermediate.
Of particular importance, long-range interactions have
been observed that indicate nativelike packing of some
of the helices in the kinetic molten globule intermediate. However, folding is impeded by local nonnative
helix packing; the H helix is translocated relative to
the G helix by a single helical turn, and folding cannot
proceed until this defect is repaired.
Apomyoglobin provides a unique opportunity for
detailed characterization of the structure and dynamics
of a protein-folding intermediate. Conditions were previously identified under which the apomyoglobin molten
globule intermediate is sufficiently stable for acquisition of multidimensional heteronuclear NMR spectra.
Analysis of 13C and other chemical shifts and measurements of polypeptide dynamics provided unprecedented
insights into the structure of this state.
The A, G, and H helices and part of the B helix are
folded and form the core of the molten globule. This
core is stabilized by relatively nonspecific hydrophobic
212 MOLECULAR BIOLOGY
2007
interactions that restrict the motions of the polypeptide
chain. Fluctuating helical structure is formed in regions
outside the core, although the amount of helix is low
and the chain retains considerable flexibility. The F
helix acts as a gate for heme binding and only adopts
stable structure in the fully folded holoprotein.
The acid-denatured (unfolded) state of apomyoglobin
is an excellent model for the fluctuating local interactions that lead to the transient formation of unstable
elements of secondary structure and local hydrophobic
clusters during the earliest stages of folding. NMR data
indicated substantial formation of helical secondary
structure in the acid-denatured state in regions that
form the A and H helices in the folded protein and also
revealed nonnative structure in the D and E helix regions.
Because the A and H regions adopt stabilized helical structure in the earliest detectable folding intermediate, these results lend strong support to folding models
in which spontaneous formation of local elements of
secondary structure plays a role in initiating formation
of the A-[B]-G-H molten globule folding intermediate.
In addition to formation of transient helical structure,
formation of local hydrophobic clusters has been
detected by using 15N relaxation measurements. Significantly, these clusters are formed in regions where
the average surface area buried upon folding is large.
In contrast to acid-denatured unfolded apomyoglobin,
the urea-denatured state is largely devoid of structure,
although residual hydrophobic interactions have been
detected by using relaxation measurements.
We measured residual dipolar couplings for unfolded
states of apomyoglobin by using partial alignment in
strained polyacrylamide gels. These data provide novel
insights into the structure and dynamics of the unfolded
polypeptide chain. We have shown that the residual
dipolar couplings arise from the well-known statistical
properties of flexible polypeptide chains. Residual dipolar couplings provide valuable insights into the dynamic
and conformational propensities of unfolded and partly
folded states of proteins and hold great promise for
charting the upper reaches of protein folding landscapes.
To probe long-range interactions in unfolded and
partially folded states of apomyoglobin, we introduced
spin-label probes at several sites throughout the polypeptide chain. These experiments led to the surprising
discovery that transient structures with nativelike longrange contacts between hydrophobic clusters exist
within the ensemble of conformations formed by the
acid-denatured state of apomyoglobin. They also indi-
THE SCRIPPS RESEARCH INSTITUTE
cated that the packing of helices in the molten globule
state is similar to that in the native folded protein.
The view of protein folding that results from our
work on apomyoglobin is one in which collapse of the
polypeptide chain to form increasingly compact states
leads to progressive accumulation of secondary structure and increasing restriction of fluctuations in the
polypeptide backbone. Chain flexibility is greatest at
the earliest stages of folding, when transient elements
of secondary structure and local hydrophobic clusters
are formed. As the folding protein becomes increasingly
compact, backbone motions become more restricted,
the hydrophobic core is formed and extended, and
nascent elements of secondary structure are progressively stabilized. The ordered tertiary structure characteristic of the native protein, with well-packed side
chains and relatively low-amplitude local dynamics,
appears to form rather late in folding.
We recently introduced a variation on the classic
quench-flow technique, which makes use of the capabilities of modern NMR spectrometers and heteronuclear NMR experiments, to study the proteins labeled
along the folding pathway in an unfolded state in an
aprotic organic solvent. This method allows detection
of many more amide proton probes than in the classic
method, which requires formation of the fully folded
protein and the measurement of the protein’s NMR
spectrum in water solution. This method is particularly
useful in documenting changes in the folding pathway
that result in the destabilization of parts of the protein
in the molten globule intermediate. We recently showed
that self-compensating mutations designed to change
the amino acid sequence such that the average area
buried upon folding is significantly changed while the
3-dimensional structure of the final folded state remains
the same. These studies showed that the average area
buried upon folding is an accurate predictor of those
parts of the apomyoglobin molecule that will fold first
and participate in the molten globule intermediate.
Quench-flow hydrogen exchange experiments performed
on a series of hydrophobic core mutants indicated that
the overall helix-packing topology of the kinetic folding
intermediate is like that of the native protein, despite
local nonnative interactions in packing of the G and H
helices (Fig. 1). Finally, using a rapid mixing device, we
have reduced the dead time of the kinetic refolding
experiments and have shown that a compact helical
intermediate is formed within 400 microseconds after
initiation of apomyoglobin refolding.
MOLECULAR BIOLOGY
2007
THE SCRIPPS RESEARCH INSTITUTE
213
CHAPERONE–COCHAPERONE–CLIENT PROTEIN
INTERACTIONS
F i g . 1 . Schematic representation of the amide proton occupancies in the kinetic intermediate state formed in the burst phase of
apomyoglobin folding (solid ribbons), mapped onto the structure of
fully folded myoglobin. Areas of the protein that are not folded until
later stages are shown as dotted lines.
FOLDING-UNFOLDING TRANSITIONS IN CELLULAR
M E TA B O L I S M
Many species of bacteria sense and respond to their
own population density by an intricate autoregulatory
mechanism known as quorum sensing; bacteria release
extracellular signal molecules, called autoinducers, for
cell-cell communication within and between bacterial
species. A number of bacteria appear to use quorum
sensing for regulation of gene expression in response
to fluctuations in cell population density. Processes
regulated in this way include symbiosis, virulence, competence, conjugation, production of antibiotics, motility, sporulation, and formation of biofilms.
We determined the 3-dimensional solution structure
of a complex composed of the N-terminal 171 residues
of the quorum-sensing protein SdiA of Escherichia coli
and an autoinducer molecule, N-octanoyl-1-homoserine
lactone (HSL). The SdiA-HSL system shows the “folding switch” behavior associated with quorum-sensing
factors produced by other bacterial species. In the
presence of HSL, SdiA is stable and folded and can be
produced in good yields from an E coli expression system. In the absence of the autoinducer, SdiA is expressed
into inclusion bodies. Samples of the SdiA-HSL complex
can be denatured but cannot be refolded in aqueous
buffers. The solution structure of the complex provides
a likely explanation for this behavior. The autoinducer
molecule is tightly bound in a deep pocket in the hydrophobic core and is bounded by specific hydrogen bonds
to the side chains of conserved residues. The autoinducer thus forms an integral part of the hydrophobic
core of the folded SdiA.
Understanding the role of unfolded states in cellular
processes will require an understanding of the structural
basis of their interactions, but unfolded proteins are
impossible to characterize structurally by x-ray crystallography, and spectroscopic methods of all kinds are
limited. Unfolded proteins must be explored under conditions that approximate their physiologic milieu: in
solution, at physiologic pHs and salt concentrations,
and in the presence of specific cofactors. Structural
insights will be obtained not only from the delineation
of 3-dimensional structures but also from the description of conformational ensembles and of the motions of
polypeptide chains under various conditions.
To gain new insights into the structural basis for
the ability of unfolded and partly folded proteins to
function in living systems, we study the interactions of
“client” proteins and cochaperones with a well-known
eukaryotic chaperone, Hsp90. Some of the protein components are much larger than have traditionally been
studied by using solution NMR. However, we have
designed a set of experiments that will allow us to draw
valid conclusions about the extent and role of disorder
in Hsp90 interactions. In particular, we will apply techniques recently developed in our laboratory for the analysis of hydrogen-deuterium exchange from unstable
partially folded proteins by trapping the 2 H-labeled
species in the aprotic solvent dimethyl sulfoxide. This
powerful new technique will be used to probe the structure, stability, and interactions of client proteins and
cochaperones with Hsp90.
Nuclear Magnetic Resonance
Studies of the Structure and
Dynamics of Enzymes
H.J. Dyson, P.E. Wright, S.H. Bae, G. Bhabha, D. Boehr,
P. Card, C. Cervantes,* G. Kroon, M. Martinez-Yamout,
S.C. Sue, L.M. Tuttle, L.L. Tennant, J. Chung, C.L. Brooks,
S.J. Benkovic,** A. Holmgren,*** E.A. Komives*
* University of California, San Diego, California
** Pennsylvania State University, University Park, Pennsylvania
*** Karolinska Institutet, Stockholm, Sweden
W
e use site-specific information on structure and
dynamics, obtained from nuclear magnetic
resonance (NMR), to further the understand-
214 MOLECULAR BIOLOGY
2007
THE SCRIPPS RESEARCH INSTITUTE
ing of protein function. We focus on the mechanism of
enzymes and the relationship between dynamics and
function in a number of medically important systems.
DYNAMICS IN ENZYME ACTION
Dynamic processes are implicit in the catalytic
function of all enzymes. We use state-of-the-art NMR
methods to elucidate the dynamic properties of several
enzymes. New methods have been developed for analysis
of NMR relaxation data for proteins that tumble anisotropically and for analysis of slow timescale motions.
Dihydrofolate reductase plays a central role in folate
metabolism and is the target enzyme for a number of
antibacterial and anticancer agents. 15N relaxation experiments on dihydrofolate reductase from Escherichia coli
revealed a rich diversity of backbone dynamical features for a broad range of timescales (picoseconds
to milliseconds).
A major focus is the characterization of all intermediates in the dihydrofolate reductase reaction cycle. We
have identified functionally important motions in loops
that control access to the active site of dihydrofolate
reductase. These motions differ in amplitude and
timescale depending on the presence of substrate and/or
cofactor in the active site. In addition, measurements of
the population distribution of aliphatic side-chain
rotamers provide evidence for coupled motion of activesite side chains that could enhance the catalytic process.
Most recently, we used relaxation dispersion measurements to obtain direct information on microsecond-millisecond timescale motions in dihydrofolate
reductase, allowing us to characterize the structures
of excited states involved in some of these catalysisrelevant processes. Each intermediate in the catalytic
cycle samples low-lying excited states whose conformations resemble the ground-state structures of the
preceding and following intermediates. Fluctuations
between these states occur on a timescale that is
directly relevant to the structural transitions involved
in progression through the catalytic cycle (Fig. 1).
Substrate and cofactor exchange occurs through
these excited substates. The maximum hydride transfer and steady-state turnover rates are governed by the
dynamics of transitions between the ground and excited
states of the intermediates. The modulation of the energy
landscape by the bound ligands funnels the enzyme
through its reaction cycle along a preferred kinetic path
(Fig. 1).
Dihydrofolate reductase is also the test system for
a series of experiments to address the question, If all
F i g . 1 . Schematic diagram of the energy landscape of dihydrofo-
late reductase catalysis. Ground-state (larger) and higher energy
(smaller) structures of each intermediate in the cycle, modeled on
published x-ray structures, are shown. For each intermediate in the
catalytic cycle, the higher energy conformations detected in the relaxation dispersion experiments resemble the ground-state conformations
of adjacent intermediates. Rate constants for the interconversion
between the complexes, measured by pre–steady-state enzyme kinetics
at 298 K, pH 6, are indicated with red arrows; the rates measured
in relaxation dispersion experiments are indicated with black arrows.
From Boehr et al., Science 313:1638, 2006. Reprinted with permission from AAAS.
of the chemistry goes on at the active site, what is the
purpose of the rest of the enzyme? We are using chimeric
mutants, synthesized by our collaborator S.J. Benkovic,
Pennsylvania State University, University Park, Pennsylvania, by using a library approach. The purpose of
these experiments is to test the hypothesis that local
variations in amino acid sequence, 3-dimensional structure, and polypeptide chain dynamics strongly influence
the local interactions that mediate enzyme catalysis
and may constitute the essential circumstance that
allows enzymes to achieve high turnover rates as well
as exquisite specificity in their reactions. A combination of NMR structure and dynamics measurements,
single-molecule fluorescence measurements, and analysis of the catalytic steps in these mutant proteins
provides new insights into the role of the protein in
enzyme catalysis.
S T R U C T U R E A N D D Y N A M I C S O F P R I O N VA R I A N T S
Onset of prion diseases is caused by conversion of
the cellular prion protein PrPC into an abnormally folded
isoform, PrPSc, that has the same primary structure as
PrPC but a totally different 3-dimensional conforma-
MOLECULAR BIOLOGY
2007
tion. The abnormally folded (“scrapie”) form of the
protein is associated with several diseases, including
scrapie in sheep, bovine spongiform encephalopathy
(mad cow disease), and human Creutzfelt-Jakob disease and other inherited prion diseases. We are gathering information on the mechanism of PrPSc formation
that can be obtained from structural and dynamic studies
of mutant prion proteins corresponding to inherited
prion diseases.
Individuals carrying familial mutations such as
P102L (P101L in our study) are more susceptible to
prion disease. On the other hand, sheep or humans
carrying Q167R and/or Q218K mutations are resistant
to scrapie and Creutzfelt-Jakob disease, respectively.
We are using the protease-resistant cores of wild-type
and mutant mouse prion proteins to study the structural and dynamic basis of PrPC-to-PrPSc conversion
in inherited prion diseases. The core is sufficient to
transmit infectivity.
DYNAMICS AND THE FUNCTION OF IκBα
It is becoming increasingly clear that the function
of many systems in living cells depends not only on
the structures of the components but also on the structures’ flexibility. Numerous examples exist in which
components of an important biological interaction are
unstructured or partly structured. In addition, even those
interacting molecules that can be classified as “folded”
have areas of mobility. Often, these areas are located
precisely in the active site of an enzyme or in the binding site of an interacting molecule.
A central molecular interaction in cellular control
is the interaction between the nuclear transcription factor NF-κB and its inhibitor IκBα. IκBα consists of a
series of ankyrin repeats, which appear to have differential mobility. Using hydrogen-deuterium exchange
and mass spectrometry, our collaborator, E.A. Komives,
University of California, San Diego, found that the second, third, and fourth ankyrin repeats of IκBα are well
folded, whereas the fifth and sixth repeats, apparently
with exactly the same structure, are highly dynamic.
These observations prompt a number of questions:
Are the motions inferred from the hydrogen-deuterium
mass spectrometry experiments also reflected in the
backbone and side-chain dynamics of the protein, as
measured by NMR relaxation? Are the motions still present in the IκBα–NF-κB complex? Are they necessary
for complex formation, so that if they are damped out,
for example, by site-directed mutagenesis at appropriate positions, is the formation of the complex disfavored?
THE SCRIPPS RESEARCH INSTITUTE
215
To answer these questions, we are doing a series of NMR
experiments on IκBα and its complexes with NF-κB.
Chaperonin-Mediated
Protein Folding
A.L. Horwich, E. Chapman, S.M. Johnson, E. Koculi
uring the past year, we have continued to investigate the GroEL-GroES chaperonin system. This
system assists the folding to native form of a
large number of newly translated proteins.
D
GroEL ACTION IN VIVO
In in vivo studies, we have isolated and characterized a new temperature-sensitive Escherichia coli GroEL
mutant. To our knowledge, this mutant, E461K, is the
most severe conditional mutant to date. The substitution at the ring-ring interface residue of GroEL abolishes the cooperative binding and hydrolysis of ATP
within rings and anticooperativity between them and
is strongly temperature sensitive for mediating folding
in vitro. In culture, E coli containing the mutant protein grow normally at 23°C, but upon a shift to 37°C,
their growth slows within 30 minutes and stops by
1–2 hours. During this time, translation continues,
but misfolding and aggregation occur, for example,
with the induced test protein ribulose-1,5-bisphosphate carboxylase/oxygenase.
More generally, as revealed by pulse-chase studies,
a broad collection of newly translated E coli proteins
are misfolded and aggregate in the mutant cells. These
abnormal proteins were identified by using proteolysis
and mass spectrometry in collaboration with J.R. Yates,
Department of Cell Biology, confirming the global effects
of E461K on newly translated proteins. The abnormal
proteins included many larger than 60 kD, too large
to be encapsulated in the GroEL-GroES cavity. In vitro,
2 of these larger proteins, MetE (88 kD) and aconitase
(92 kD), refolded to native form when GroEL alone was
present, suggesting that rapid interaction with the open
chaperonin ring can provide a form of folding assistance without encapsulation.
Thus, these studies suggest a potentially broader
role for GroEL in mediating folding, although we cannot entirely exclude a model in which the aggregation
of stringent cis cavity-requiring substrate proteins leads
to an avalanche of aggregation involving other newly
translated proteins. Direct cross-linking studies between
216 MOLECULAR BIOLOGY
2007
GroEL and substrate proteins in E coli cells in culture
are under way to resolve this question.
T O P O L O G Y O F S U B S T R AT E P R O T E I N B O U N D T O A N
OPEN GroEL RING
In other studies, we have investigated the topology
of substrate proteins bound to an open ring of GroEL.
In collaboration with H. Saibil, Birkbeck College, London, England, we used electron cryomicroscopy image
reconstructions of binary complexes of the substrate
protein malate dehydrogenase (MDH) in complex with
GroEL (Fig. 1). Additional studies of such binary com-
THE SCRIPPS RESEARCH INSTITUTE
ity (Fig. 1, left panels). Although parts of the substrate
protein were observed here in physical association with
the apical domains, other parts of the substrate not
detected by electron microscopy could populate any
location inside the central cavity, for example, down
into the equatorial zone, as shown by disulfide crosslinking studies in which cysteine substitutions were
“scanned” both inside the central cavity and across
the outside surfaces of GroEL (the outer surfaces did
not form cross-links with bound substrate).
C O N F O R M AT I O N A L C H A N G E S O F S U B S T R AT E
P R O T E I N D U R I N G F O L D I N G I N T H E C I S C AV I T Y
F i g . 1 . Electron cryomicroscopy reconstructions of binary complexes
of the substrate MDH (yellow, green, blue) with the chaperonin
GroEL (gray). End views, top panels; tilted views, middle panels;
cutaway side views, bottom panels.
plexes, as well as type II chaperonins, are under way in
collaboration with B. Carragher and C. Potter, Department of Cell Biology.
In the studies of MDH-GroEL complexes, we found
that the MDH subunit was bound to the apical domains
of the GroEL ring in various topologies. In one topology,
3 consecutive apical domains were contacted by substrate localized at the rim of the central cavity (Fig. 1,
right panels). In another topology, 3 consecutive domains
were occupied by substrate localized across the middle
part of the apical domains, where 2 α-helices (H and
I) were present whose hydrophobic side chains have
been implicated in binding by mutational studies (Fig. 1,
center panels). In a third topology, the substrate was
more deeply situated, interacting with consecutive apical domains via a surface that contains both helix I and
an underlying hydrophobic extended segment, with the
substrate pointing even more deeply into the central cav-
In other studies in vitro, we have probed the conformations of substrate proteins both during binding by
GroEL and during folding by GroEL-GroES. For example,
we have used formation of disulfide bonds in the secretory protein trypsinogen, which behaves as a stringent
(GroEL-GroES dependent) substrate during refolding in
vitro after dilution from urea and reductant. In these
studies, trypsinogen molecules were analyzed at various
times by using proteolysis and mass spectrometry. We
found that only native long-range disulfides are formed,
pinning the 2 β-barrels together at the top and bottom
of the active-site cleft, during folding of trypsinogen
inside GroEL-GroES; that is, only a nativelike global
topology is selected. However, shorter range nonnative
bonds occur early in the reaction, formed within the
2 β-barrel domains, but these are “corrected” to native
later during folding, suggesting a conformational annealing activity inside the cis cavity that occurs in the
absence of any further ATP binding or hydrolysis.
As another means of observing substrate proteins, in
collaboration with K. Wüthrich, Department of Molecular
Biology, we have been using nuclear magnetic resonance
to study another substrate protein: rhodanese. We are
examining behavior of this substrate bound to GroEL.
We have also been examining nuclear magnetic resonance spectra of the GroEL chaperonin itself.
PUBLICATIONS
Chapman, E., Farr, G.W., Usaite, R., Furtak, K., Fenton, W.A., Chaudhuri, T.K.,
Hondorp, E.R., Matthews, R.G., Wolf, S.G., Yates, J.R., Pypaert, M., Horwich, A.L.
Global aggregation of newly translated proteins in an Escherichia coli strain deficient
of the chaperonin GroEL. Proc. Natl. Acad. Sci. U. S. A. 103:15800, 2006.
Elad, N., Farr, G.W., Clare, D.K., Orlova, E.V., Horwich, A.L., Saibil, H.R. Topologies
of a substrate protein bound to the chaperonin GroEL. Mol. Cell 26:415, 2007.
Farr, G.W., Fenton, W.A., Horwich, A.L. Perturbed ATPase activity and not “close
confinement” of substrate in the cis cavity affects rates of folding by tail-multiplied
GroEL. Proc. Natl. Acad. Sci. U. S. A. 104:5342, 2007.
Park, E.S., Fenton, W.A., Horwich, A.L. Disulfide formation as probe of topology during folding in GroEL-GroES reveals correct formation of long-range bonds and editing of
incorrect short-ranges ones. Proc. Natl. Acad. Sci. U. S. A. 104:2145, 2007.
MOLECULAR BIOLOGY
2007
Chemical Regulation of
Gene Expression
J.M. Gottesfeld, R. Burnett, C.J. Chou, D. Herman, S. Ku,
S. Lefebvre, E. Soragni, S. Tsai,* M. Farkas,* P.B. Dervan,*
S.L. Perlman,** G. Coppola,** D. Geschwind,** M. Rai,***
M. Pandolfo***
* California Institute of Technology, Pasadena, California
** University of California, Los Angeles, California
*** Universite Libre de Bruxelles-Hospital Erasme, Brussels, Belgium
he ability to control gene expression at will has
been a longstanding goal in molecular biology
and human medicine. We focus on 2 classes of
small molecules that can alter gene expression in human
cells: (1) pyrrole-imidazole polyamides, molecules that
can be programmed by chemical synthesis to recognize
a wide range of DNA sequences, and (2) histone deacetylase inhibitors, compounds that alter the postsynthetic modification states of major chromosomal proteins
and thereby activate gene expression. Our recent efforts
to develop polyamides as therapeutic agents for human
cancer and novel histone deacetylase inhibitors that offer
promise in the treatment of neurodegenerative diseases
are summarized in the following sections.
T
B L O C K I N G C A N C E R C E L L P R O L I F E R AT I O N W I T H A
P O LYA M I D E - C H L O R A M B U C I L C O N J U G AT E
DNA alkylators, common agents used to treat cancer
in humans, act by damaging DNA. A DNA alkylator used
to treat a variety of lymphatic cancers is the nitrogen
mustard chlorambucil. Because chlorambucil alkylates
DNA at all available guanine residues in cellular DNA,
coupling of chlorambucil to a more sequence-selective
molecule, such as a polyamide, decreases the number of sites in the genome that are damaged and may
decrease unwanted side effects while retaining the ability of the compound to kill cancer cells.
We recently found that a specific polyamide-chlorambucil conjugate called 1R-Chl alters the morphology and
growth characteristics of multiple cancer cell lines in
culture and causes these cells to arrest in the G 2 /M
stage of the cell cycle. The compound blocks proliferation of various cancer cell lines in immunocompromised
mice, including cells derived from colon, prostate, and
lung cancers and from chronic myelogenous leukemia,
and no apparent toxic effects occur at doses required
for a therapeutic effect.
Using microarray analysis, we found that the gene
target of 1R-Chl is histone H4c, a member of the gene
THE SCRIPPS RESEARCH INSTITUTE
217
family that encodes a critical component of cellular
chromatin and a gene that is highly expressed in a
wide range of cancer cells. Reduction in histone H4
protein by polyamide treatment was confirmed in cells
treated with 1R-Chl, which caused chromatin decondensation. Small interfering RNAs to H4c mRNA also
have this effect, providing target validation for the effects
of 1R-Chl.
Another polyamide-chlorambucil conjugate that
targets the H4c gene elicits the same cellular response;
molecules of similar composition that do not target this
gene are ineffective. Using gene chip expression analysis, we found that only 156 genes are affected by 1RChl in K562 chronic myelogenous leukemia cells and
identified 2 genes of interest, which encode histones
H4c and H4k/j in this cancer cell type. Pathway analysis
suggests that DNA damage is the eventual outcome of
1R-Chl treatment, findings in agreement with a 2-hit
model for the action of 1R-Chl. 1R-Chl–induced DNA
damage is cell-type specific, indicating that downregulation of H4c and H4/j gene transcription is a prerequisite for eventual DNA damage and G 2 /M arrest of
K562 and other cancer cell lines.
These observations also explain why, unlike other
conventional DNA alkylators, 1R-Chl has few or no toxic
effects in cells in culture or in whole animals. The therapeutic potential of these molecules is being pursued
through additional animal and cellular models for human
cancer and proliferative diseases.
P O LYA M I D E S A S A C T I VAT O R S O F G E N E E X P R E S S I O N
The neurodegenerative disease Friedreich’s ataxia
is caused by gene silencing through expansion of
GAA•TTC triplet repeats in the first intron of a nuclear
gene that encodes the essential mitochondrial protein
frataxin. Normal frataxin alleles have 6–34 repeats,
whereas alleles from patients with Friedreich’s ataxia
have 66–1700 repeats. Longer repeats cause a more
profound deficiency in the protein frataxin and are associated with earlier onset and increased severity of the
disease. Two models have been proposed to account for
gene silencing by expanded GAA•TTC repeats: unusual
DNA structures and repressive heterochromatin.
Molecules that reverse formation of unusual DNA
structures and/or heterochromatin in the gene for frataxin most likely increase transcription through expanded
GAA•TTC repeats, thereby relieving the deficiency in
frataxin mRNA and protein in cells from patients with
Friedreich’s ataxia. We found that polyamides targeting GAA•TTC repeats partially alleviated repression of
218 MOLECULAR BIOLOGY
2007
the gene frataxin in a cell line derived from white blood
cells from a patient with Friedreich’s ataxia. These molecules also increased frataxin protein levels in these
cells, and microarray studies showed that a limited
number of genes in the human genome were affected
by polyamides targeting GAA•TTC repeat DNA.
We hypothesize that polyamides might act as a thermodynamic “sink” and lock GAA•TTC repeats into double-stranded B DNA. Such an event would disfavor
duplex unpairing, which is necessary for formation of the
unusual DNA structures associated with expanded triplet
repeats. Alternatively, polyamides may relieve heterochromatin-mediated repression by opening the chromatin
domain containing frataxin. To explore this latter hypothesis, we turned to another class of small molecules.
H I S T O N E D E A C E T Y L A S E I N H I B I T O R S T H AT
R E V E R S E F R ATA X I N S I L E N C I N G
We used antibodies to the various modification
states of the core histones and chromatin immunoprecipitation methods to examine the chromatin structure
of the gene for frataxin in normal cells and in cell lines
derived from patients with Friedreich’s ataxia. We found
that gene silencing at expanded frataxin alleles was
accompanied by hypoacetylation of histones H3 and H4
and methylation of histone H3 at lysine 9, consistent
with a heterochromatin-mediated repression mechanism.
These findings suggest that histone deacetylase
inhibitors, compounds that reverse heterochromatin,
might activate frataxin. We identified a commercial histone deacetylase inhibitor, BML-210, that partially
reverses silencing in the Friedreich’s ataxia cell line. On
the basis of the structure of this compound, we synthesized and assayed a series of derivatives of BML-210
and identified histone deacetylase inhibitors that reverse
frataxin silencing in primary lymphocytes from patients
with Friedreich’s ataxia. These molecules act directly on
the histones associated with frataxin, increasing acetylation at particular lysine residues on histones H3 and H4.
Of note, the histone deacetylase inhibitors cross
the blood-brain barrier and increase levels of frataxin
mRNA in the brain and other organs in a mouse model
of the human disease. Genome-wide microarray studies indicated that our compounds partially correct the
transcription pattern of genes affected by Friedreich’s
ataxia in the brain of these mice and in lymphocytes
from patients with the disease to the transcription pattern of healthy animals or individuals. We are investigating the pharmacokinetic and toxicity properties of
the compounds. We are also exploring the usefulness
THE SCRIPPS RESEARCH INSTITUTE
of these compounds in related neurodegenerative and
neuromuscular diseases, such as Huntington’s disease,
spinal muscular atrophy, and myotonic dystrophy.
PUBLICATIONS
Tsai, S.M., Farkas, M.E., Chou, C.J., Gottesfeld, J.M., Dervan, P.B. Unanticipated
differences between α- and γ-aminobutyric acid-linked hairpin polyamide-alkylator
conjugates. Nucleic Acids Res. 35:307, 2007.
Nucleic Acid Dynamics
D.P. Millar, J. Gill, G. Pljevaljc̆ić, S. Pond, J. Wang,
E.J.C. Van der Schans
he focus of our research is the assembly and
conformational dynamics of nucleic acid–based
macromolecular machines and assemblies. We
use single-molecule fluorescence methods to investigate a range of systems, including ribozymes, ribonucleoprotein complexes, and DNA polymerases. Our
studies reveal the dynamic structural rearrangements
that occur during the assembly and function of these
complex macromolecular machines.
T
RIBOZYMES
RNA conformation plays a central role in the mechanism of ribozyme catalysis. The hairpin ribozyme is a
small nucleolytic ribozyme that serves as a model system for studies of RNA folding and catalysis. The hairpin ribozyme consists of 2 internal loops, 1 of which
contains the scissile phosphodiester bond, displayed
on 2 arms of a 4-way multihelix junction.
To attain catalytic activity, the ribozyme must fold
into a compact conformation in which the 2 loops
become connected by a network of tertiary hydrogen
bonds. We monitor the formation of this docked structure by using fluorescence resonance energy transfer
(FRET) and ribozyme constructs labeled with donor and
acceptor dyes within the loop-bearing arms. By measuring FRET at the level of single ribozyme molecules, we
reveal subpopulations of compact and extended conformers that are not detected in ensemble experiments.
Using this approach, we found that the ribozyme
populates an intermediate state in which the 2 loops
are in proximity but tertiary interactions have yet to
form. This quasi-docked state forms rapidly (submillisecond timescale); however, the tertiary contacts
between the 2 loops are established through a slow
conformational search. The hairpin ribozyme is an
ideal system for exploring this fundamental mechanism of the formation of RNA tertiary structure.
MOLECULAR BIOLOGY
2007
R I B O N U C L E O P R O T E I N A S S E M B LY
The Rev protein from HIV type 1 is a key regulatory protein that controls the transition from early to
late patterns of viral gene expression. Rev binds to a
highly structured region within the viral mRNA, known
as the Rev response element (RRE), where it forms an
oligomeric ribonucleoprotein complex. The formation
of this complex inhibits splicing and facilitates export
of the unspliced viral RNA from the nucleus to the
cytoplasm. Because of its critical role in the viral life
cycle, the Rev-RRE complex provides a novel target
for the development of therapeutic drugs.
To dissect the mechanism of assembly of ribonucleoprotein complexes, we use single-molecule fluorescence
imaging methods to monitor the progressive formation of
oligomeric complexes of Rev on individual RRE molecules immobilized on a solid surface. We also use single-pair FRET to probe changes in the conformation of
the RRE during the assembly process. We are using the
results of these mechanistic studies to develop novel fluorescence-based methods for high-throughput screening
of libraries of chemical compounds. The new screening
tools are being used to identify small molecules that
block binding of Rev to the RRE or prevent the subsequent Rev-Rev oligomerization.
D N A P O LY M E R A S E S
DNA polymerases are remarkable for their ability
to synthesize DNA at rates approaching several hundred
base pairs per second while maintaining an extremely
low frequency of errors. To elucidate the origin of polymerase fidelity, we are using single-molecule fluorescence methods to examine the dynamic interactions
that occur between a DNA polymerase and its DNA and
nucleotide substrates. The FRET method is being used
to observe conformational transitions of the enzymeDNA complex that occur during selection and incorporation of an incoming nucleotide substrate.
Our results reveal that binding of a correct nucleotide substrate induces a slow conformational change
within the polymerase, causing the “fingers” subdomain
to close over the DNA primer terminus and incoming
nucleotide. Our studies are providing new insights into
the dynamic structural changes responsible for nucleotide recognition and selection by DNA polymerases.
The advantage of single-molecule observations is that
they eliminate the need to synchronize a population
of molecules, allowing these dynamic processes to be
observed directly.
THE SCRIPPS RESEARCH INSTITUTE
219
PUBLICATIONS
Bailey, M.F., Van der Schans, E.J.C., Millar, D.P. Dimerization of the Klenow fragment of Escherichia coli DNA polymerase I is linked to its mode of DNA binding.
Biochemistry, in press.
Grover, R.K., Pond, S.J., Cui, Q., Subramanian, P., Case, D.A., Millar, D.P., Wentworth, P. O-Glycoside orientation is an essential aspect of base J recognition by the
kinetoplastid DNA-binding protein JBP1. Angew. Chem. Int. Ed. 46:2839, 2007.
Tian, F., Debler, E.W., Millar, D.P., Deniz, A.A., Wilson, I.A., Schultz, P.G. The
effects of antibodies on stilbene excited-state energetics. Angew. Chem. Int. Ed.
45:7763, 2006.
Single-Molecule Biophysics:
Folding, Assembly, and Function
A.A. Deniz, S.Y. Berezhna, A.C.M. Ferreon, Y. Gambin,
E. Lemke, S. Mukhopadhyay
e develop and use state-of-the-art single-molecule fluorescence methods and high-sensitivity
ensemble methods to address key biological questions by probing multiple structures or reaction
pathways during the folding, assembly, and function of
biomolecules. These methods offer key advantages
over traditional measurements, allowing us to directly
observe the behavior of individual subpopulations in
mixtures of molecules and to measure the kinetics
of structural transitions of stochastic processes under
equilibrium conditions.
A major goal is to apply single-molecule methods
to studies of protein folding and aggregation. For example, partially folded or misfolded protein structures are
thought to play important cellular roles, and these
states can be studied by using single-molecule methods.
In this context, in collaboration with S.L. Lindquist,
Whitehead Institute for Biomedical Research, Cambridge, Massachusetts, we are examining the interplay
between folding and aggregation of Sup35, a yeast
translational termination factor whose activity is modulated by conversion to a prion form. By using singlemolecule fluorescence resonance energy transfer (FRET)
and subnanomolar concentrations of Sup35, we minimize potential artifacts due to protein aggregation.
Using single-molecule dual-color fluorescence coincidence analysis, we have directly tested that the protein
is monomeric at these concentrations. Our single-molecule FRET denaturation analysis of Sup35 indicated
that this protein does not have significant stable structure under native conditions. Additionally, an analysis
of signal fluctuations from a small number of Sup35
molecules (fluorescence correlation spectroscopy),
W
220 MOLECULAR BIOLOGY
2007
revealed rapid structural fluctuations on the nanosecond timescale. Overall, the results show that native
Sup35 is intrinsically disordered and populates a compact and rapidly fluctuating ensemble of structures,
behavior that may be key in its aggregation and biological function.
We are also studying the structural properties and
aggregation of α-synuclein, a protein implicated in the
pathogenesis of Parkinson’s disease and other neurodegenerative diseases. As a prelude to single-molecule
studies, we have carried out a detailed thermodynamic
study of the conformational properties of this protein
in isolation, as well as in the presence of the detergent
sodium dodecyl sulfate. Interestingly, sodium dodecyl
sulfate, which is often used to denature proteins, can
induce folding of α-synuclein. Our results indicate that
depending on its environment, α-synuclein can occupy
several different structural states. This structural plasticity could be important during the function of this
protein in the cell. We are now using single-molecule
methods to study in more detail both the monomeric
structures and the early stages of aggregation in Sup35
and α-synuclein.
Important events on the folding landscapes of proteins and other biomolecules occur on millisecond or
faster timescales. To facilitate understanding of such
fast folding processes, in collaboration with A. Groisman,
University of California, San Diego, we are developing
microfluidic mixing methods for both single-molecule
and ensemble experiments. In particular, microfluidic
mixing combined with single-molecule fluorescence
detection will provide more detailed insights into folding
landscapes of proteins and other biological molecules.
To better study folding processes and the assembly
and function of larger and multicomponent biological
complexes, we are developing multicolor single-molecule FRET methods. We are using these novel methods to study the assembly mechanisms of fragments
of the bacterial ribosome, in collaboration with J.R.
Williamson, Department of Molecular Biology.
Finally, we are also using multicolor fluorescence
imaging to study the pathways of nuclear and cytoplasmic RNA interference. Recently, in collaboration
with P.G. Schultz, Department of Chemistry, we probed
the relative localizations of short interfering RNA; a
key RNA interference protein, Ago2; and P-bodies, which
are RNA-processing cytoplasmic bodies that may be
involved in RNA interference. Our imaging and biochemical data indicate that P-bodies play only a minor
role in RNA interference. We are beginning single-par-
THE SCRIPPS RESEARCH INSTITUTE
ticle tracking experiments to probe RNA interference
mechanisms in greater detail.
PUBLICATIONS
Deniz, A.A., Mukhopadhyay, S., Lemke, E.L. Single-molecule biophysics: at the
interface of biology, physics, and chemistry. J. R. Soc. Interface, in press.
Ferreon, A.C.M., Deniz, A.A. α-Synuclein multistate folding thermodynamics:
implications for protein misfolding and aggregation. Biochemistry 46:4499, 2007.
Mukhopadhyay, S., Deniz, A.A. Fluorescence from diffusing single molecules illuminates biomolecular structure and dynamics. J. Fluoresc., in press.
Mukhopadhyay, S., Krishnan, R., Lemke, E.A., Lindquist, S., Deniz, A.A. A
natively unfolded yeast prion monomer adopts an ensemble of collapsed and
rapidly fluctuating structures. Proc. Natl. Acad. Sci. U. S. A. 104:2649, 2007.
Tian, F., Debler, E.W., Millar, D.P., Deniz, A.A., Wilson, I.A., Schultz, P.G. The
effects of antibodies on stilbene excited-state energetics. Angew. Chem. Int. Ed.
46:7763, 2006.
Computer Modeling of Proteins
and Nucleic Acids
D.A. Case, Y. Bomble, M. Crowley, Q. Cui, F. Dupradeau,*
S. Moon, V. Pelmenschikov, D. Shivakumar, T. Steinbrecher,
R.C. Walker, V. Wong, W. Zhang
* Université Jules Verne, Amiens, France
omputer simulations offer an exciting approach
to the study of many aspects of biochemical
interactions. We focus primarily on molecular
dynamics simulations (in which Newton’s equations of
motions are solved numerically) to model the solution
behavior of biomacromolecules. Recent applications
include detailed analyses of electrostatic interactions
in short peptides (folded and unfolded), proteins, and
oligonucleotides in solution.
In addition, molecular dynamics methods are useful in refining solution structures of proteins by using
constraints derived from nuclear magnetic resonance
(NMR) spectroscopy, and we continue to explore new
methods in this area. Our developments are incorporated
into the Amber molecular modeling package, designed
for large-scale biomolecular simulations, and into other
software, including Nucleic Acid Builder, for developing 3-dimensional models of unusual nucleic acid structures; SHIFTS, for analyzing chemical shifts in proteins
and nucleic acids; RNAmotif, for finding structural motifs
in genomic sequence databases; and DOCK, for placing inhibitors into enzyme active sites.
C
NMR AND THE STRUCTURE AND DYNAMICS OF
PROTEINS AND NUCLEIC ACIDS
Our overall goal is to extract the maximum amount
of information about biomolecular structure and dynam-
MOLECULAR BIOLOGY
2007
ics from NMR experiments. To this end, we are studying the use of direct refinement methods for determining biomolecular structures in solution, going beyond
distance constraints to generate closer connections
between calculated and observed spectra. We are also
using quantum chemistry to study chemical shifts and
spin-spin coupling constants. Other types of data, such
as chemical shift anisotropies, direct dipolar couplings
in partially oriented samples, and analysis of cross-correlated relaxation, are also being used to guide structure refinement.
In recent structural studies, we focused on the binding of small ligands to DNA and on models for chemically damaged DNA. For example, Figure 1 shows an
F i g . 1 . Structure of a part of DNA with a missing base. Left, Structure expected for an idealized B-form helix. Right, The actual structure,
as determined by NMR spectroscopy.
abasic site, a site in which the nucleic acid base has
been removed. The left side shows what is expected if
the regular, B-form helix is not modified. The right side
shows the NMR structure, illustrating marked structural modifications that could be recognized by DNA
repair enzymes.
NUCLEIC ACID MODELING
Another project centers on the development of novel
computer methods to construct models of “unusual”
nucleic acids that go beyond traditional helical motifs.
We are using these methods to study circular DNA,
small RNA fragments, and nucleosome core particles.
We continue to develop efficient computer implementations of continuum solvent methods to allow simplified
simulations that do not require a detailed description
of the solvent (water) molecules; this approach also
provides a useful way to study salt effects.
Recent efforts have made second derivatives of
these energies available, so that normal mode analyses of nucleic acids with dozens to hundreds of nucleotides can be analyzed and the predictions compared
THE SCRIPPS RESEARCH INSTITUTE
221
with those of simpler, elastic continuum models. These
efforts provide a new avenue for developing and testing low-resolution models that can be used for large
molecular assemblies. A key current application is to
nucleosome core particles.
D Y N A M I C S A N D E N E R G E T I C S O F N AT I V E A N D
N O N N AT I V E S TAT E S O F P R O T E I N S
Analysis methods similar to those described for
nucleic acids are also being used to estimate electrostatic and thermodynamic properties of proteins. A key
feature is the development of computational methods
that can be used to model pH and salt dependence of
complex conformational transitions such as unfolding
events. A second aspect of this work is a detailed interpretation of NMR results for proteins through molecular dynamics simulations and the construction of
models for molecular motion and disorder.
All of these modeling activities are based on molecular mechanics force fields, which provide estimates of
energies as a function of conformation. We continue to
work on improvements in force fields; recently, we
focused on adding aspects of electronic polarizability,
going beyond the usual fixed-charge models, and on
methods for handling arbitrary organic molecules that
might be considered potential inhibitors in drug discovery efforts. Overall, the new models should provide
a better picture of the noncovalent interactions between
peptide groups and the groups’ surroundings, leading
ultimately to more faithful simulations.
PUBLICATIONS
Chen, J., Dupradeau, F.-Y., Case, D.A., Turner, C.J., Stubbe, J. NMR structural
studies and molecular modeling of duplex DNA containing normal and 4′-oxidized
abasic sites. Biochemistry 46:3096, 2007.
Cui, Q., Tan, R.-K.Z., Harvey, S.C., Case, D.A. Low-resolution molecular dynamics
simulations of the 30S ribosomal subunit. Multiscale Model. Simul. 5:1248,
2006.
Grover, R.K., Pond, S.J., Cui, Q., Subramaniam, P., Case, D.A., Millar, D.P., Wentworth, P., Jr. O-Glycoside orientation is an essential aspect of base J recognition by the
kinetoplastid DNA-binding protein JBP1. Angew. Chem. Int. Ed. 46:2839, 2007.
Mongan, J., Simmerling, C., McCammon, J.A., Case, D.A., Onufriev, A. Generalized Born model with a simple, robust molecular volume correction. J. Chem. Theory Comput. 3:156, 2007.
Moon, S., Case, D.A. A new model for chemical shifts of amide hydrogens in proteins. J. Biomol. NMR 38:139, 2007.
Paesani, F., Zhang, W., Case, D.A., Cheatham, T.E. III, Voth, G.A. An accurate
and simple quantum model for liquid water. J. Chem. Phys. 125:184507, 2006.
Seabra, G.M., Walker, R.C., Elstner, M., Case, D.A., Roitberg, A.E. Implementation of the SCC-DFTB method for hybrid QM/MM simulations within the Amber
molecular dynamics package. J. Phys. Chem. A, in press.
Tang, S., Case, D.A. Vibrational averaging of chemical shifts anisotropies in model
peptides. J. Biomol. NMR, in press.
222 MOLECULAR BIOLOGY
2007
Quantum Chemical Analysis for
Redox-Active Metalloenzymes
and for Photochemistry
L. Noodleman, D.A. Case, W.-G. Han, V. Pelmenschikov,
J.A. Fee, L. Hunsicker-Wang,* T. Lovell,** T. Liu,***
M. Ullmann,**** D. Bashford*****
* Trinity University, San Antonio, Texas
** AstraZeneca R&D, Mölndal, Sweden
*** University of Maryland, College Park, Maryland
**** Universität Bayreuth, Bayreuth, Germany
***** St. Jude Children’s Research Hospital, Memphis, Tennessee
e are using a combination of modern quantum
chemistry (density functional theory, DFT) and
classical electrostatics to describe the energetics, reaction pathways, and spectroscopic properties
of metalloenzymes. In addition, we are analyzing other
systems that have novel catalytic, photochemical, or
photophysical properties.
Critical biosynthetic and regulatory processes may
involve catalytic transformations of fairly small molecules or groups by transition-metal centers. The ironmolybdenum cofactor center (Fig. 1) of nitrogenase
catalyzes the multielectron reduction of molecular nitrogen to 2 molecules of ammonia plus molecular hydrogen. We are continuing our work on the catalytic cycle
of this enzyme, following up on our earlier research on
the structure and oxidation state of the cofactor complex
in the “resting enzyme” before multielectron reduction
and nitrogen binding.
W
THE SCRIPPS RESEARCH INSTITUTE
On the basis of DFT calculated vs experimental
physical properties, including redox potentials, cluster
geometries, and Mössbauer isomer shifts, the core cluster has a (MoFe7S 9X) prismane active site, where the
central X most likely is nitride and the “resting cluster
oxidation state” is Mo(IV)Fe(II)4Fe(III)3. A central carbide is also a possibility, because no clear central ligand
hyperfine signal has yet been detected for the paramagnetic resting state. Further exploration of different
redox states, including bound alternative substrates and
inhibitors studied by using electron nuclear double
resonance and Mössbauer spectroscopies and compared with quantum chemical calculations, is under
way to illuminate both the active-site structure and the
reaction mechanism.
Methane monooxygenase catalyzes the oxidation of
methane to methanol. This reaction is extremely important in the biosphere because methane is a greenhouse
gas. Further, the monooxygenase can catalyze reactions
with alternative substrates, a characteristic that has
implications for detoxification of organic pollutants. We
have been examining various steps in the catalytic
cycle, comparing predictions based on DFT findings
with those of Mössbauer spectroscopy. We have focused
particularly on the critical intermediate Q (Fig. 2), which
F i g . 2 . The proposed active-site structure of methane monooxy-
genase intermediate Q.
F i g . 1 . The hyperfine signal from the interstitial X atom of the
nitrogenase FeMoco cofactor is small (Aiso ~0.3 MHZ, X = 14N
case studied) for the resting-state enzyme because of the symmetry
conditions. On the basis of DFT calculations, we propose 2 principal modes of FeMoco symmetry perturbation via 2-electron–2-proton reduction at the metal-sulfide cluster and ligand binding: electronic (Aiso ~6.8 MHZ) and structural (Aiso ~22.9 MHZ). A bound
allylic alcohol, derived from substrate propargyl alcohol plus 2 electrons plus 2 protons is shown.
performs the oxygen insertion reaction. The iron complex at the active site of methane monooxygenase
resembles the complex in ribonucleotide reductases
(see following); the resemblance is particularly close
for ribonucleotide reductases in some pathogenic bacteria, including Chlamydia, in which the tyrosine near
the iron-oxo dimer complex is replaced by phenylala-
MOLECULAR BIOLOGY
2007
nine. The altered mechanism of action of the ribonucleotide reductases in these pathogens is of great
interest for exploring feasible drug treatments.
Class I ribonucleotide reductases are aerobic
enzymes that catalyze the reduction of ribonucleotides
to deoxyribonucleotides, providing the required building
blocks for DNA replication and repair. These enzymes
are targets for anticancer, antiviral, and antibacterial
drugs. These ribonucleotide-to-deoxyribonucleotide reactions occur via a long-range radical (or proton-coupled
electron transfer) propagation mechanism initiated by
a fairly stable tyrosine radical, “the pilot light.” When
this pilot light goes out, the tyrosine radical is regenerated by a high-oxidation-state Fe(III)-Fe(IV)-oxo enzyme
intermediate called X. Using DFT and electrostatics
calculations in combination with analysis of Mössbauer,
electron nuclear double resonance, and magnetic circular
dichroism spectroscopies to search for a proper structural and electronic model for intermediate X, we have
zeroed in a promising structure. Now the pathways to
and from intermediate X can be properly examined.
In collaboration with K. Hahn, University of North
Carolina at Chapel Hill, we have examined the physical and spectroscopic properties of novel solvent-sensitive fluorescent dyes of interest for imaging live cells.
These dyes can act as biosensors for changes in conformational state in cell signaling proteins, with potential uses in high-throughput drug screening.
With J.A. Fee, Department of Molecular Biology,
we are exploring the mechanism of 4-electron reduction of molecular oxygen to 2 molecules of water by
cytochrome oxidase (complex IV of mitochondria and
the related cytochrome oxidase ba 3 from Thermus
thermophilus). The copper-iron-heme complex links
molecular oxygen reduction to proton pumping across
the mitochondrial membrane.
PUBLICATIONS
Han, W.-G., Noodleman, L. Structural model studies for the high-valent intermediate Q of methane monooxygenase from broken-symmetry density functional calculations. Inorg. Chim. Acta, in press.
THE SCRIPPS RESEARCH INSTITUTE
223
Theoretical and Computational
Molecular Biophysics
C.L. Brooks III, R. Armen, I. Borelli, D. Bostick, D. Braun,
L. Bu, J. Chen, M.F. Crowley, O. Guvench, R. Hills, W. Im,*
J. Khandogin, I. Khavrutskii, J. Knight, J. Lee, R. Manige,
M. Michino, A. Mitsutake,** H.D. Nguyen, S. Patel,***
V. Reddy, H.A. Scheraga,**** C. Shepard, I.F. Thorpe,
M.C. Tripp, R, Wheeler,***** K. Yoshimoto
* Kansas University, Lawrence, Kansas
** Kelo University, Tokyo, Japan
*** University of Delaware, Newark, Delaware
**** Cornell University, Ithaca, New York
***** University of Oklahoma, Norman, Oklahoma
e wish to understand the forces that determine
the structure of proteins, peptides, nucleic
acids, and complexes containing these molecules and the processes by which these structures are
adopted. To address these issues, we use statistical
mechanics, molecular simulation, statistical modeling,
and quantum chemistry.
Creating atomic-level models to simulate biophysical processes (e.g., protein folding or binding of a
ligand to a biological receptor) requires (1) development of new polarizable potential energy functions
that accurately represent the atomic interactions and
(2) use of quantum chemistry to aid in determining
the parameters for the models. Characterization of
thermodynamic and kinetic properties requires the
development and implementation of new theoretical
and computational approaches that connect averages
over atomistic descriptions to experimentally measurable thermodynamic and kinetic properties.
Interpreting experimental results at more microscopic levels is fueled by the development and investigation of theoretical models for the processes of interest.
Massive computational resources are needed to realize
these objectives, and this need motivates our efforts to
achieve the efficient use of new computer architectures,
including large supercomputers, Linux Beowulf clusters,
computational grids, and Internet-based volunteer supercomputers. Each of the objectives and techniques mentioned represents an ongoing area of development in our
research program in computational biophysics. The following are highlights of a few specific projects.
W
AMMONIUM TRANSPORT
Uncovering the means by which cells transport
ions and molecules across the lipid membrane via
224 MOLECULAR BIOLOGY
2007
channels and transporters will provide a deeper understanding of a range of diseases caused by the malfunction of these channels. The selective transport of
ammonium (NH4+) across biological membranes is a
homeostatic necessity in prokaryotic and eukaryotic
cells. In humans and other animals, ammonium is transported by members of the Rhesus (Rh) family of proteins. The family consists of 2 types of protein: erythroid
and nonerythroid. Erythroid Rh proteins are expressed
on the surface of erythrocytes, where they perform
immunogenic and structural roles. Nonerythroid Rh proteins are expressed in kidneys, liver, and testes, where
they aid in disposal of ammonium and regulation of pH.
Biophysical and structural characterization of a bacterial homolog of a human (kidney) ammonium channel,
AmtB from Escherichia coli, has revealed that selective
“sensing” of ammonium is achieved by forcing the deprotonation of ammonium and allowing only uncharged
ammonia (NH3) to traverse the channel via a narrow
hydrophobic lumen. The static x-ray structure alone,
however, does not clearly indicate where and how
along the pathway toward the cytoplasm ammonium
becomes deprotonated on the periplasmic end of the
channel and then reprotonated on the cytoplasmic end.
Using detailed molecular dynamics sampling techniques, we investigated and clarified the protonation control mechanism of AmtB. Our calculations reveal that
the “equivalence points” for deprotonation (and reprotonation) of ammonium occur at dehydrative phenylalanine landmarks along the transport pathway (Fig. 1).
At these landmarks, ammonium is able to form only 3
(or fewer) hydrogen bonds with the protein or water
molecules and has complete access to either periplasmic or cytoplasmic aqueous solution. According to
our simulation studies, AmtB indirectly controls ammonium (de)protonation by directly controlling its hydration, effectively exploiting the propensity of ammonium
to (de)protonate when only 3 hydrogen bonds are available. Ultimately, the only proton acceptor available at
the (de)protonation regions (green bars in Fig. 1) is water.
A S S E M B LY O F V I R U S PA R T I C L E S
The controlled assembly of viruslike particles is a
marked biomedical interest in the bionanotechnology
of vaccine design, gene therapy, and medical imaging.
For example, viruslike particles of human papillomavirus
are used in the vaccine for cervical cancer. Applications of viruslike particles require an understanding of
the principles that govern the spontaneous self-assembly of viral capsids.
THE SCRIPPS RESEARCH INSTITUTE
F i g . 1 . Transport of ammonium through the ammonium channel.
In the AmtB ammonium channel monomer, ammonium (Am1 and
Am5) and ammonia binding sites (Am2–Am4, the hydrophobic lumen)
derived from diffraction (red spheres) and molecular dynamics umbrella
sampling analysis (blue spheres) are shown (top, left). Ammonium
(de)protonation regions (green horizontal bars) coincide with dehydrative phenylalanine residues. Ammonium equivalence points (top,
right) along the transport pathway were determined by using free
energy calculations for ammonium (dashed line) and ammonia (solid
line) transport as a function of the respective positions of the 2 components along the transport axis. The equivalence points for deprotonation occur at the regions of intersection (green bars) of these free
energy curves. Close-up (bottom, right) shows bound ammonia at
each of the 3 binding positions within the hydrophobic lumen.
Spherical (icosahedral) viral capsids typically are
composed of multiple copies (e.g., 60, 180, 240) of
a single protein capsid and have a geometric shape
(Fig. 2). Using coarse-grained models for viral capsid
proteins and specialized molecular dynamics techniques,
we are exploring the nature of viral capsid assembly
as a function of temperature and protein concentration.
We found that the assembly of T = 1 icosahedral capsids occurs with high fidelity over only a small range
of temperatures and protein concentrations (Fig. 2, top).
Outside this range, particularly at low temperatures or
high protein concentrations, large enclosed “monster
particles” are produced. The dynamics of capsid assembly under optimal conditions is a nucleated process, with
few assembly intermediates of any great size.
The assembly of T = 3 capsids proceeds via a more
complex mechanism. Three protein species, corresponding to the capsid protein in quasi-equivalent environments in the assembled protein, spontaneously assemble
via a rich-phase diagram (Fig. 2, bottom). Under nearoptimal conditions for growth of T = 3 capsids, not only
icosahedral capsids but also oblate, angular, twisted
and tubular isometric forms occur. The presence and
MOLECULAR BIOLOGY
2007
THE SCRIPPS RESEARCH INSTITUTE
225
Chen, J., Brooks, C.L. III. Critical importance of length-scale dependence in implicit
modeling of hydrophobic interactions. J. Am. Chem. Soc. 129:2444, 2007.
Hills, R.D., Jr., Brooks, C.L. III. Hydrophobic cooperativity as a mechanism for
amyloid nucleation. J. Mol. Biol. 368:894, 2007.
Khandogin, J., Brooks, C.L. III. Molecular simulations of pH-mediated biological
processes. Annu. Rep. Comput. Chem., in press.
Khandogin, J., Brooks, C.L. III. Toward the accurate first-principles prediction of
ionization equilibria in proteins. Biochemistry 45:9363, 2006.
Khandogin, J., Chen, J., Brooks, C.L. III. Exploring atomistic details of pH-dependent peptide folding. Proc. Natl. Acad. Sci. U. S. A. 103:18546, 2006.
Khandogin, J., Raleigh, D.P., Brooks, C.L. III. Folding intermediate in the villin
headpiece domain arises from disruption of a N-terminal hydrogen-bonded network. J. Am. Chem. Soc. 129:3056, 2007.
Khavrutskii, I.V., Arora, K., Brooks, C.L. III. Harmonic Fourier beads method for
studying rare events on rugged energy surfaces. J. Chem. Phys. 125:174108, 2006.
Khavrutskii, I.V., Price, D.J., Lee, J., Brooks, C.L. III. Conformational change of
the methionine 20 loop of Escherichia coli dihydrofolate reductase modulates pKa
of the bound dihydrofolate. Protein Sci., in press.
F i g . 2 . Viral capsid assembly. The assembled capsids of the sim-
plest (T = 1) and next most complicated (T = 3) icosahedral viruses
are highly organized spherical particles composed of geometrically
regular copies of identical capsid proteins (top, left). For T = 1 particles, under optimal conditions for capsid growth, assembly occurs via
a kinetic mechanism that involves the stepwise addition of small
protein assemblies (i.e., monomers, dimers, and trimers), leading
to complete capsids. Under conditions of lower temperature, assembly is nucleated too rapidly, and partial growth leads to the combining
of many partial capsids to form monster particles. The assembly of
T = 3 capsids under near-optimal conditions yields a range of closed
capsid forms that are determined by the number of 5- to 6-fold symmetry dislocations that occur (top, right). Additionally, at lower temperatures, the assembly process yields quite open, spirallike monster
particles that strongly resemble species seen in electron microscope
imaging of viral assembly products for T = 3 and higher T-number
capsids (e.g., human papillomavirus).
abundance of these nonicosahedral closed structures
are controlled by the proclivity for 5- to 6-fold symmetric dislocations. Thus, by controlling the conformational equilibrium of the capsid proteins, we can shift
the nature and distribution of the assembled particles.
Nguyen, H.D., Reddy, V.S., Brooks, C.L. III. Deciphering the kinetic mechanism of
spontaneous self-assembly of icosahedral capsids. Nano Lett. 7:338, 2007.
Shepherd, C.M., Borelli, I.A., Lander, G., Natarajan, P., Siddavanahalli, V., Bajaj,
C., Johnson, J.E., Brooks, C.L. III, Reddy, V.S. VIPERdb: a relational database for
structural virology. Nucleic Acids Res. 34(Database Issue):D386, 2006.
Sherman, M.B., Guenther, R.H., Tama, F., Sit, T.L., Brooks, C.L. III, Mikhailov,
A.M., Orlova, E.V., Baker, T.S., Lommel, S.A. Removal of divalent cations induces
structural transitions in red clover necrotic mosaic virus, revealing a potential
mechanism for RNA release. J. Virol. 80:10395, 2006.
Tama, F., Ren, G., Brooks, C.L. III, Mitra, A.K. Model of the toxic complex of
anthrax: responsive conformational changes in both the lethal factor and the protective antigen heptamer. Protein Sci. 15:2190, 2006.
Thorpe, I.F., Brooks, C.L. III. Molecular evolution of affinity and flexibility in the
immune system. Proc. Natl. Acad. Sci. U. S. A., in press.
Zimmermann, J., Oakman, E.L., Thorpe, I.F., Shi, X., Abbyad, P., Brooks, C.L. III,
Boxer, S.G., Romesberg, F.E. Antibody evolution constrains conformational heterogeneity by tailoring protein dynamics. Proc. Natl. Acad. Sci., U.S.A. 103:13722,
2006.
Computation and Visualization
in Structural Biology
A.J. Olson, D.S. Goodsell, M.F. Sanner, S. Dallakyan,
PUBLICATIONS
Bostick, D.L., Brooks, C.L. III. Deprotonation by dehydration: the origin of ammonium sensing in the AmtB channel. PLoS Comput. Biol. 3:e22, 2007.
Bostick, D.L., Brooks, C.L. III. On the equivalence point for ammonium (de)protonation during its transport through the AmtB channel. Biophys. J., in press.
Bostick, D.L., Brooks, C.L. III. Selectivity in K+ channels is due to topological
control of the permeant ion’s coordinated state. Proc. Natl. Acad. Sci. U. S. A., in
press.
Bu, L., Im, W., Brooks, C.L. III. Membrane assembly of simple helix homo-oligomers studied via molecular dynamics simulations. Biophys. J. 92:854, 2007.
Chen, J., Brooks, C.L. III. Can molecular dynamics simulations provide high-resolution refinement of protein structure? Proteins 67:922, 2007.
A. Gillet, R. Harris, Y. Hu,* R. Huey, J. Huntoon, S. Karnati,
W. Lindstrom, G.M. Morris, A. Omelchenko, M. Pique,
B. Norledge, R. Rosenstein, M. Utsintong, G. Vareille,
Q. Zhang, Y. Zhao
* Tanabe Research Laboratories U.S.A., Inc., San Diego, California
n the Molecular Graphics Laboratory, we develop
novel computational methods to analyze, understand, and communicate the structure and interactions of complex biomolecular systems. In this past
year, we developed a new method for docking mole-
I
226 MOLECULAR BIOLOGY
2007
cules to flexible targets, showing its effectiveness with
an important target for AIDS therapy. Using our distributed computing resource FightAIDS@Home, we also
performed a massive virtual screen of potential drug
compounds that might affect this target. Within our
component-based visualization environment, we continue to develop 3-dimensional molecular models as a
tangible human-computer interface in educational and
research settings and methods for predicting biomolecular interactions, analyzing biomolecular structure
and function, and presenting the biomolecular world
in education and outreach.
THE SCRIPPS RESEARCH INSTITUTE
binding energies of a panel of 1800 ligands against 71
wild-type and mutant proteases (Fig. 1). This profile is
PROTEIN FLEXIBILITY IN DOCKING
Computational docking is an indispensable tool in
the development of new drugs. However, protein motion
and induced fit are major challenges in many medically
relevant systems, such as the flexible HIV protease that
is a major target for current AIDS therapy. To address
this challenge, we have developed a hierarchical and
multiresolution representation of the flexibility of biological macromolecules that can be used in computational simulations. This treelike structure enables the
computationally tractable encoding of a small subset
of a protein’s conformational subspace. We have developed a docking method, FLIPDock, in which Flexibility
Tree is used with our AutoDock empirical free energy
force field. With FLIPDock, we have reproduced a crossdocking experiment carried out earlier with AutoDock in
which 20 inhibitors of HIV protease I were docked
systematically into the 20 conformations of the receptor. We showed that by adding receptor flexibility, we
could increase the rate of successful cross docking from
72% to 98%.
I D E N T I F I C AT I O N O F S PA N N I N G M U TA N T S
AutoDock was the first docking code to be used in
a public, Internet-based, distributed computing project,
and we have continued to use this massive computing
resource to perform experiments that are beyond the
reach of traditional computing. We are currently using
the World Community Grid, a large Internet-distributed
computing project sponsored by IBM, to support FightAIDS@Home on more than 500,000 clients. Personal
computers are used by the program when the computers are not in use by their owners, providing an enormous, and largely untapped, computational resource.
We recently completed an analysis of results from
FightAIDS@Home; our goal was to identify a subset of
mutant protease structures that span the ligand-binding
diversity of the entire set. We analyzed the profile of
F i g . 1 . Top, Profile of binding energies from a large virtual
screen against a large collection of protease structures. Bottom,
Representative structures highlighted in a Sammon mapping.
like a fingerprint: proteases with similar characteristics
have similar profiles, but differences in the profiles
highlight significant structural differences between different proteases. For instance, some proteases may have
strong binding to large compounds, and others may prefer smaller compounds. Principal component analysis
was then used to simplify the data set and to identify
“spanning” proteases that represent the unique features
different subsets of proteases.
We identified 1 wild-type protease as the structure
that best characterizes the central tendency of the entire
protease set. We also identified 7 additional structures,
including 1 wild-type structure, 2 HIV type 2 protease
structures, and 4 single and multiple mutants, as structures that represent distinct subclasses and thus would
be our best choice for structures to use in future efforts
to design anti-HIV drugs.
TA N G I B L E I N T E R FA C E S I N S T R U C T U R A L B I O L O G Y
We have continued to develop autofabricated physical models (“solid printing”) of biological molecules
MOLECULAR BIOLOGY
2007
and their components and assemblies; our goal is to
use them in both research and education. We integrated
computer graphics and computation with these physical models by using augmented reality to create custom
interfaces to facilitate exploration and computation of
molecular interactions. We have begun to use a selfassisted protein-folding model to teach elements of protein structure and assembly to our graduate students. We
are continuing to develop the software that will enable
the control of interactive computations through manipulation of the tangible models. The responses to the models have been positive, and the research community is
beginning to see how a solid 3-dimensional model can
provide tangible, multimodal feedback that mouse, keyboard, and image behind a computer screen cannot.
Building on the success of our flexible, articulated
models of protein structure, we have begun work on
larger models that illustrate the dynamic characteristics
of biomolecules. We recently completed the design and
fabrication of a model of clathrin (Fig. 2). The model
THE SCRIPPS RESEARCH INSTITUTE
227
comprehensive thermodynamic model that allows incorporation of intramolecular energies into the predicted
free energy of binding. It also incorporates a chargebased method for evaluating desolvation that is designed
to use a typical set of atom types. The method has been
calibrated by using a set of 188 diverse protein-ligand
complexes of known structure and binding energy and
has been tested by using a set of 100 complexes of
ligands with retroviral proteases. AutoDock4 also allows
the explicit modeling of flexibility in selected side chains,
a characteristic that will greatly expand the systems
that can be studied.
AutoDockTools, the companion graphical user interface for AutoDock, has also been expanded and improved
for use with AutoDock4. In particular, AutoDockTools
now allows easy definition of flexible side chains in a
macromolecule. We have also developed new tools for
the analysis of docking results, including an advanced
tool for visualizing protein-ligand interactions, and
extensive tools for setting up and running large virtual
screens. We are currently developing tools to analyze
the results of virtual screens to create “probability maps”
that indicate the binding characteristics of a large panel
of small molecules.
C O M P O N E N T - B A S E D V I S U A L I Z AT I O N E N V I R O N M E N T S
F i g . 2 . A flexible physical model of clathrin.
is composed of articulated triskelion models that can
be assembled to form an entire clathrin coat. Magnets
are used to model the major point of interaction during the assembly.
A D VA N C E S I N C O M P U TAT I O N A L D O C K I N G
According to recent reports, the computer program
AutoDock is the world’s most widely cited docking
method, and we have continued to make it a central
tool for predicting biomolecular interaction. AutoDock4
was released in early 2007 and is available through
a convenient open source license. AutoDock4 includes
a new semiempirical free energy force field based on a
To facilitate the integration and interoperation of
computational models and techniques from a wide variety of scientific disciplines, we continue to expand our
component-based software environment. The environment is centered on Python, a high-level, object-oriented,
interpretive programming language. This approach allows
the compartmentalization and reuse of software components. Python provides a powerful “glue” for assembling computational components and, at the same time,
a flexible language for rapid prototyping and interactive
scripting of new applications.
We released version 1.4.4 of our software components in January 2007. This release contains substantial enhancements, including a completely rewritten
interface to APBS, a software package for the numerical
solution of the Poisson-Boltzmann equation, making it
easy to produce high-quality pictures of electrostatic
potentials on molecular surfaces. A new control panel
provides a high-level interface for rapidly displaying
molecular models in a variety of representations. This
new release is also distributed with installer programs for
computers running the Windows, Macintosh OS X, and
Linux operating systems. This release of the MGLTools
software suite also provides a new update mechanism,
228 MOLECULAR BIOLOGY
2007
allowing users to update their software interactively
over the Internet and granting them access to “bug”
fixes and enhancements on a nightly basis. This update
mechanism is safe, because the user retains the ability
to “roll back” and restore a previously installed version
of the software.
IDENTIFYING OPTIMAL BINDING SITES
We have develop a method, termed AutoLigand, that
can be used to identify and quantify optimal binding
sites, yielding a chemically detailed prediction of the
shape of a ligand within a predicted binding site. Using
a grid-based description of the binding interaction,
AutoLigand identifies a contiguous “envelope” of maximal affinity for the macromolecular structure (Fig. 3).
F i g . 3 . Several low-energy AutoLigand envelopes are found with
HIV protease, including the large envelope in the active site at the
center and 2 exosites on either side. The envelopes are colored
according to the optimal atom type at each point: gray for carbon,
red for oxygen, and blue for nitrogen.
The envelop is a contiguous region in space filled with
points that represent potential atom centers for atoms
in a ligand. In brief, with the AutoLigand method, an
affinity-potential grid around the macromolecular structure is calculated and then a flood-fill and site optimization process is used to identify the contiguous envelope
with the best summed interaction energy. We have
shown that the method is effective for identifying binding sites in proteins and for optimizing the binding of
drugs to protein targets.
PROTEIN-PROTEIN DOCKING WITH SURFDOCK
Protein-protein docking remains a major challenge
in predicting the structure of protein complexes. In our
protein-protein docking method, SurfDock, we recently
developed a blurred surface and a scoring term that
effectively conserve the shape complementarity of
protein-protein complexes under small conformational
changes. We are currently developing a novel electrostatic scoring term that uses the discrete charges on
the blurred surfaces. A coarse function and discrete
charges are generated to reproduce the electrostatic
energies by solving the nonlinear Poisson-Boltzmann
THE SCRIPPS RESEARCH INSTITUTE
equation by using the APBS software program. The
coarse function mimics the formula of either the Coulombic or the Debye-Hückel potential. We are testing different combinations of the coarse function and discrete
charges to determine the optimal method for calculating electrostatic interaction energies for a set of known
protein-protein complexes.
C O M P U TAT I O N A L M O D E L I N G O F E X T R A C E L L U L A R
INTERACTIONS OF TISSUE FACTOR
In collaboration with W. Ruf, Department of Immunology, we have continued to study protein interactions
in the blood coagulation pathway. Most recently, we
focused on protease-activated receptors (PARs), G protein–coupled receptors that convert an extracellular
proteolytic cleavage event into a transmembrane signal. We have built a homology structure of PAR-2 as
the basis for studying PAR-2 activation mechanisms,
and we have docked a series of peptides and chemical
compounds to the model to derive its activation mechanisms. The docking shows that the 2 PAR-2 activating
peptides have similar dominating binding modes and
that both bind to the same side of the PAR-2 extracellular domain with a similar interaction between arginine and glutamic acid. The docking results agree with
the experimental findings and have led to ideas for
several PAR-2 mutations to test hypotheses about
ligand binding.
Dr. Ruf and his group have also shown that disulfide isomerization by protein disulfide isomerase (PDI)
switches the complex composed of tissue factor and
coagulation factor VIIa (TF-VIIa) from coagulation to cell
signaling. To understand how the isomerase interacts
with the complex, we built the homology model of
human PDI and manually docked 3-dimensional physical models of PDI with the TF-VIIa complex. PDI is a
large, flexible molecule, so computer docking is extremely
difficult. However, 3-dimensional physical models
allowed effective exploration of possible binding modes.
The physical model of PDI has flexible joints between
the N- and C-terminal tails and the central domains,
and the model of the TF-VIIa complex has flexible joints
between the EGF-1 and Gla domains of VIIa and the
extracellular C-terminal strand of TF. With these highly
articulated physical models, we found a reasonable
binding mode that can use both catalytic sites of PDI
and both forms of an important disulfide. On the basis
of the binding mode, we suggested 6 mutations in TF
that will affect PDI binding but not binding of coagulation factor X.
MOLECULAR BIOLOGY
V I S U A L M E T H O D S F R O M AT O M S T O C E L L S
Understanding structural molecular biology is
essential to foster progress and critical decision making among students, policy makers, and the general
public. In the past year, we continued our long-standing commitment to science education and outreach
with a combination of presentations, popular and professional illustrations and animations, 3-dimensional
tangible models, and a presence on the World Wide
Web. In these projects, we use the diverse visualization
tools developed in the Molecular Graphics Laboratory
to disseminate results that range from atomic structure to cellular function.
We also continued several regular features that informally present molecular structure and function. The
“Molecule of the Month” at the Protein Data Bank
entered its eighth year of providing an accessible introduction to the central database of biomolecular structure. Each month, a new molecule is presented with a
description of the molecule’s structure, function, and
relevance to health and welfare (Fig. 4). Visitors are
2007
THE SCRIPPS RESEARCH INSTITUTE
229
of interest to clinical oncologists and provide a source
of continuing education for physicians; “Recognition
in Action,” a new series at the Journal of Molecular
Recognition; and work with the Nanoscale Informal
Science Network supported by the National Science
Foundation to develop new materials for presenting the
science of nanotechnology.
PUBLICATIONS
Chang, M., Lindstrom, W., Olson, A.J., Belew, R.K. Analysis of wild-type and
mutant structures via in silico docking against diverse ligand libraries. J. Chem.
Info. Model., in press.
Eubanks, L.M., Rogers, C.J., Beuscher, A.E., Koob, G.F., Olson, A.J., Dickerson,
T.J., Janda, K.D. A molecular link between the active component of marijuana and
Alzheimer’s disease pathology. Mol. Pharm. 3:773, 2006.
Goodsell, D.S. The molecular perspective: alcohol. Oncologist 11:1045, 2006.
Goodsell, D.S. The molecular perspective: estrogen sulfotransferase. Oncologist
11:418, 2006.
Goodsell, D.S. The molecular perspective: tissue factor. Oncologist 11:849, 2006.
Goodsell, D.S. Seeing the nanoscale. Nanotoday 1:44, August 2006.
Goodsell, D.S. Toll-like receptors. J. Mol. Recognit. 19:387, 2006.
Herman, T., Morris, J., Colton, S., Batiza, A., Patrick, M., Franzen, M., Goodsell,
D.S. Tactile teaching: exploring protein structure/function using physical models.
Biochem. Mol. Biol. Educ. 34:247, 2006.
Huey, R., Morris, G.M., Olson, A.J., Goodsell, D.S. A semi-empirical free energy
force field with charge-based desolvation. J. Comput. Chem. 28:1145, 2007.
Jiang, Z., Georgel, P., Li, C., Choe, J., Crozat, K., Rutschmann, S., Du, X., Bigby,
T., Mudd, S., Sovath, S., Wilson, I.A., Olson, A.J., Beutler, B. Details of Toll-like
receptor:adapter interaction revealed by germ-line mutagenesis. Proc. Natl. Acad.
Sci. U. S. A. 103:10961, 2006.
Morris, G.M., Huey, R., Olson, A.J. Using AutoDock for ligand-receptor docking.
Curr. Protoc. Bioinformatics, in press.
Rogers, J.P., Beuscher, A.E., Flajolet, M., McAvoy, T., Nairn, A.C., Olson, A.J.,
Greengard, P. Discovery of protein phosphatase 2C inhibitors by virtual screening.
J. Med. Chem. 49:1658, 2006.
Zhao, Y., Stoffler, D., Sanner, M. Hierarchical and multi-resolution representation
of protein flexibility. Bioinformatics 22:2768, 2006.
Structural Bioinformatics and
Computer-Aided Drug Discovery
R. Abagyan, G. Bottegoni, W. Bisson,* A. Cheltsov,*
K. Hyun,** J. Kovacs, I. Kufareva, G. Nicola, A. Saldanha
* Burnham Institute for Medical Research, La Jolla, California
F i g . 4 . Luciferase, the protein that creates light in fireflies, was
featured in the “Molecule of the Month.”
then given suggestions about how to begin their own
exploration of the structures in the data bank. Other
projects include “The Molecular Perspective,” articles
in the journal The Oncologist that present structures
** CrystalGenomics Inc., Seoul, Korea
urrently, the Protein Data Bank contains more
than 43,000 structures, providing a unique
opportunity for computational studies and rational design of therapeutic agents. We continue to focus
on developing and applying mathematical and computa-
C
230 MOLECULAR BIOLOGY
2007
tional methods of molecular modeling and function prediction. In the past year, using virtual ligand docking
and screening, we discovered novel inhibitors/antagonists of several therapeutically relevant protein targets.
M AT H E M AT I C A L M E T H O D S
Machine learning is widely used in protein annotation
and cheminformatics. However, its performance is often
hindered by computational bias caused by overrepresentation of particular subfamilies in the training sets. We
developed novel measures of prediction quality based on
inverse density weights assigned to data points. The
advantage of the new approach was demonstrated by
deriving a model in which a full set of known signal peptides was used to predict a single signal peptide.
Two-dimensional image alignment is a key step in
3-dimensional single-particle reconstruction. Recently,
we reported on “fast Bessel matching,” a real-space
correlation-based approach for performing the alignment. The problem is reduced to the calculation of a
single 3-dimensional fast Fourier transform.
C O M P U TAT I O N A L M E T H O D S
In our laboratory or in collaboration with researchers
at Structural Bioinformatics Group, Madrid, Spain, we
developed several automatic methods and tools for predicting protein flexibility, interactions, and large-complex
geometries. A method called PIER (Protein IntErface
Recognition; http://abagyan.scripps.edu/PIER/) is used
to predict interfaces on an isolated protein structure
and does not depend on evolutionary information. The
accuracy and efficiency of the method were proved on
a large and diverse benchmark.
Our previous work on protein flexibility was complemented with a Web server, DFprot (http://sbg.cib.csic.es/
Software/DFprot). The server identifies the hinge regions
of the uploaded structure by using normal-mode analysis.
To fill the gap between atomic resolution models of
protein domains and low-resolution density maps of
larger molecular assemblies, efficient fitting tools are
needed, ones that can handle partial and incomplete
data. Our novel method, ADP_EM (http://sbg.cib.csic.es/
Software/ADP_EM), combines a fast rotational search
based on spherical harmonics with a simple translational scanning. Use of the method produces accurate
fits in times ranging from seconds to a few minutes.
A P P L I C AT I O N S
Methods we have developed were successfully
applied to several proteins and complexes: G protein–
coupled receptors, complement system proteins (CD59),
gap junction channels, and a lipid-transfer protein.
THE SCRIPPS RESEARCH INSTITUTE
This work was done in collaboration with researchers
from the Mayo Clinic, Scottsdale, Arizona; Medical
University of South Carolina in Charleston; Memorial
Sloan-Kettering Cancer Center, New York City; University of Minnesota in Austin; and Cornell University,
Ithaca, New York.
Activation of G protein–coupled receptors is thought
to involve an agonist-induced conformational change.
We used the software docking program Internal Coordinates Mechanics to build a 3-dimensional model of
the agonist-bound cholecystokinin receptor. The model
helped us rationalize the experimental data on the interaction between the conserved lysine at position 187
and aspartic acid at position 5; these residues had no
influence on agonist binding but were crucial for agonist-stimulated signaling.
Using peptide screens, functional assays, and computer modeling/docking studies, we identified a 6-residue sequence of human complement C9, which spans
residues 365–371, as the primary recognition domain
for the membrane glycoprotein CD59. Our docking experiments confirmed that the C9-binding site on CD59 is
located at a hydrophobic pocket, as was putatively
indicated by our previous studies.
PIER prediction of the interaction propensity of
mammalian glycolipid-transfer protein in apo- and
ligand-bound forms was confirmed by the observed
oligomerization states: the apo form is a monomer,
whereas the sphingosine-bound complexes form crystallographic dimers. The increased propensity for protein-membrane interaction that occurs upon ligand
binding might underlie a unique mechanism of regulation of lipid transfer between proteins and membranes.
The internal coordinate sampling method was
extended to incorporate low-resolution electron microscopy data, symmetry, and requirements for a membrane environment. The obtained protocol was used in
collaboration with M. Yeager, Department of Cell Biology, to build a model of a large hexameric intermembrane gap junction channel.
INHIBITOR DISCOVERY AND DRUG REPURPOSING
This past year, we used virtual ligand docking and
screening not only to discover inhibitors but also to
study the side effects of marketed drugs and to find
new indications for the drugs. Using a rational structurebased search, we found that members of the phenothiazine family of antipsychotic agents act as antagonists
for the human androgen receptor. In collaboration with
groups from the University of Melbourne and Monash
MOLECULAR BIOLOGY
2007
University, Victoria, Australia; the Burnham Institute,
La Jolla, California; and Ohio University, Athens, Ohio;
we rationally “repurposed” one of the agents and arrived
at a potent new androgen receptor antagonist without
the original antipsychotic activity (Fig. 1).
THE SCRIPPS RESEARCH INSTITUTE
231
introduction of an allosteric ligand. This work was done
in collaboration with researchers at the University of
California, San Diego, and the Howard Hughes Medical Institute, La Jolla, California.
PUBLICATIONS
Altmann, S.M., Muryshev, A., Fossale, E., Maxwell, M.M., Norflus, F.N., Fox, J.,
Hersch, S.M., Young, A.B., MacDonald, M.E., Abagyan, R., Kazantsev, A.G. Discovery of bioactive small-molecule inhibitor of poly ADP-ribose polymerase: implications for energy-deficient cells. Chem. Biol. 13:765, 2006.
Bisson, W.H., Cheltsov, A.V., Bruey-Sedano, N., Lin, B., Chen, J., Goldberger, N.,
May, L.T., Christopoulos, A., Dalton, J.T., Sexton, P.M., Zhang, X.-K., Abagyan, R.
Discovery of antiandrogen activity of nonsteroidal scaffolds of marketed drugs.
Proc. Natl. Acad. Sci. U. S. A., in press.
Budagyan, L., Abagyan, R. Weighted quality estimates in machine learning. Bioinformatics 22:2597, 2006.
Dong, M., Ding, X.Q., Thomas, S.E., Gao, F., Lam, P.C., Abagyan, R., Miller, L.J.
Role of lysine 187 within the second extracellular loop of the type A cholecystokinin
receptor in agonist-induced activation: use of complementary charge-reversal mutagenesis to define a functionally important interdomain interaction. Biochemistry
46:4522, 2007.
Garzon, J.I., Kovacs, J., Abagyan, R., Chacon, P. ADP_EM: fast exhaustive multiresolution docking for high-throughput coverage. Bioinformatics 23:427, 2007.
Garzon, J.I., Kovacs, J., Abagyan, R., Chacon, P. DFprot: a web tool for predicting
local chain deformability. Bioinformatics 23:901, 2007.
Huang, Y., Qiao, F., Abagyan, R., Hazard, S., Tomlinson, S. Defining the CD59-C9
binding interaction. J. Biol. Chem. 281:27398, 2006.
Kovacs, J., Abagyan, R., Yeager, M. Fast Bessel matching. J. Comput. Theor.
Nanosci. 4:84, 2007.
Kovacs, J., Baker, K., Altenberg, G., Abagyan, R., Yeager, M. Molecular modeling
and mutagenesis of gap junction channels. Prog. Biophys. Mol. Biol., in press.
Kufareva I., Budagyan L., Raush E., Totrov M., Abagyan, R. PIER: protein interface recognition for structural proteomics. Proteins 67:400, 2007.
F i g . 1 . Model of phenothiazine derivative, D4, docked into human
androgen receptor. The native ligand, dihydrotestosterone, is shown
in thin purple sticks for comparison.
In collaboration with researchers from Harvard
Medical School, Boston, we discovered a compound
that inhibited the enzymatic activity of poly-(ADP-ribose)
polymerase 1 in vitro and prevented ATP loss and cell
death in a surrogate model of oxidative stress in vivo.
The neuroprotective role of the polymerase inhibitors was
confirmed in a model system of Huntington’s disease.
In another project, we targeted the enoyl-acyl carrier
protein reductase from a malaria parasite. Using structure-based virtual ligand screening, we identified 3 novel
compounds with low micromolar activity in enzymatic
and cell-based assays.
The high affinity of protein-protein interactions makes
them a difficult target for competitive small-molecule
inhibitors. We devised an assay principle for detecting
allosteric modulators of such interactions. In our assay
based on this principle, a competitive fluorescent peptide probe is used to assess the integrity of the complex of type Iα cAMP-dependent protein kinase upon
Malinina, L., Malakhova, M.L., Kanack, A.T., Lu, M., Abagyan, R., Brown, R.E.,
Patel, D.J. The liganding of glycolipid transfer protein is controlled by glycolipid
acyl structure. PLoS Biol. 4:e362, 2006.
Nicola, G., Smith, C.A., Lucumi, E., Kuo, M.R., Kavagyozov, L., Fidock, D.A.,
Sacchettini, J.C., Abagyan, R.A. Discovery of novel inhibitors targeting enoyl-acyl
carrier protein reductase in Plasmodium falciparum by structure-based virtual
screening. Biochem. Biophys. Res. Commun., in press.
Saldanha, S.A., Kaler, G., Cottam, H.B., Abagyan, R., Taylor, S.S. Assay principle
for modulators of protein-protein interactions and its application to non-ATP-competitive ligands targeting protein kinase A. Anal. Chem. 78:8265, 2006.
Mass Spectrometry:
Metabolomics and Imaging
G. Siuzdak, J. Apon, H.P. Benton, E. Go, K. Harris, L. Hoang,
E. Kalisiak, G. O’Maille, M. Sonderegger, S. Trauger,
W. Uritboonthai, E. Want, W. Webb, W. Wikoff, H. Morita,
D. Wong, A. Nordstrom, H.K. Woo, T. Northen, O. Yanes
M E TA B O L O M I C S
E
ndogenous metabolites, ubiquitous in biofluids,
tissues, and organisms of every kind, are crucial elements in understanding fundamental bio-
232 MOLECULAR BIOLOGY
2007
THE SCRIPPS RESEARCH INSTITUTE
chemistry, disease diagnosis, and drug toxicity. The
inherent advantage of monitoring small molecules
rather than proteins is the relative ease of quantitative
analysis of the molecules with mass spectrometry.
We are implementing novel mass spectrometry and
bioinformatics techniques (Fig. 1) to investigate the
F i g . 1 . A novel nonlinear approach that incorporates both ana-
lytical and bioinformatics technologies for quantitative comparisons
of metabolites.
profile of small-molecule metabolites. Our purposes
are to correlate metabolite activity with protein regulation and to develop metabolite analysis as a diagnostic method. Our ultimate goal is to create analytical
and chemical technologies and data analysis approaches
to identify and structurally characterize metabolites
of physiologic importance.
N A N O S T R U C T U R E - I N I T I AT O R M A S S S P E C T R O M E T R Y
A N A LY S I S A N D I M A G I N G
We are also using nanostructured clatherates to
facilitate vaporization and ionization of biomolecules
to develop ultra-high-sensitivity approaches in mass
spectrometry (Fig. 2). Using this technology, termed
nanostructure-initiator mass spectrometry, we can
analyze of a wide range of molecules with unprecedented sensitivity, in the yoctomole (10–24 moles)
range. The method is also being developed as a platform for biomolecular tissue imaging.
PUBLICATIONS
Chace, D.H., Barr, J.R., Duncan, M.W., Matern, D., Morris, M.R., Palmer-Toy, D.E.,
Rockwood, A.L., Siuzdak, G., Urbani, A., Yergey, A.L., Chan, Y.M. Mass Spectrometry in the Clinical Laboratory: General Principles and Guidance. Clinical and Laboratory Standards Institute, Wayne, PA, 2007. CSLI Document C50-P, Vol. 27, No. 3.
Cohen, L., Go, E.P., Siuzdak, G. Small-molecule desorption/ionization mass analysis. In: MALDI-MS: A Practical Guide to Instrumentation, Methods and Applications.
Hillenkamp, F., Peter-Katalinic, J. (Eds.). Wiley-VCH, New York, 2007, p. 299.
Go, E.P., Uritboonthai, W., Apon, J.A., Trauger, S.A., Nordstrom, A., O’Maille, G.,
Brittain, S., Peters, E.C., Siuzdak, G. Selective metabolite and peptide capture/mass
detection using fluorous affinity tags. J. Proteome Res. 6:1492, 2007.
Go, E.P., Wikoff, W.R., Shen, Z., O’Maille, G., Morita, H., Conrads, T.P., Nordstrom, A., Trauger, S.A., Uritboonthai, W., Lucas, D.A., Chan, K.C., Veenstra,
T.D., Lewicki, H., Oldstone, M.B., Schneemann, A., Siuzdak, G. Mass spectrometry reveals specific and global molecular transformations during viral infection. J.
Proteome Res. 5:2405, 2006.
F i g . 2 . Nanostructure-initiator mass spectrometry is a new plat-
form for mass spectrometry surface analysis and allows for sensitive biomolecule analysis from surfaces that also include array and
tissue imaging.
Mutch, D.M., O’Maille, G., Wikoff, W.R., Wiedmer, T., Sims, P.J., Siuzdak, G.
Mobilization of pro-inflammatory lipids in obese Plscr3-deficient mice. Genome
Biol. 8:R38, 2007.
Nordstrom, A., He, L., Siuzdak, G. Desorption/ionization on silicon (DIOS). In: Ionization Methods. Gross, M.L., Caprioli, R.M. (Eds.). Elsevier, New York, 2007,
p. 676. The Encyclopedia of Mass Spectrometry, Vol. 6. Gross, M.L., Caprioli,
R.M. (Eds. in Chief).
Pendyala, G., Want, E.J., Webb, W., Siuzdak, G., Fox, H.S. Biomarkers for neuroAIDS: the widening scope of metabolomics. J. Neuroimmune Pharmacol. 2:72, 2007.
Shen, Z., Want, E.J., Chen, W., Keating, W., Nussbaumer, W., Moore, R., Gentle,
T.M., Siuzdak, G. Sepsis plasma protein profiling with immunodepletion, threedimensional liquid chromatography tandem mass spectrometry, and spectrum
counting. J. Proteome Res. 5:3154, 2006.
Steenwyk, R.C., Hutzler, J.M., Sams, J., Shen, Z., Siuzdak, G. Atmospheric pressure
desorption/ionization on silicon ion trap mass spectrometry applied to the quantitation
of midazolam in rat plasma and determination of midazolam 1′-hydroxylation kinetics
in human liver microsomes. Rapid Commun. Mass Spectrom. 20:3717, 2006.
Want, E.J., Nordstrom, A., Morita, H., Siuzdak, G. From exogenous to endogenous: the inevitable imprint of mass spectrometry in metabolomics. J. Proteome
Res. 6:459, 2007.
Want, E.J., Smith, C.A., Qin, C., Van Horne, K.C., Siuzdak, G. Phospholipid capture combined with non-linear chromatographic correction for improved metabolite
profiling. Metabolomics 2:145, 2006.
MOLECULAR BIOLOGY
2007
RNA-Protein Interactions in
Germ-Line Development
J.R. Williamson, F. Agnelli, A. Beck, C. Beuck, A. Bunner,
A. Carmel, S. Edgcomb, E. Johnson, D. Kerkow, E. Kompfner,
S. Kwan, E. Menichelli, W. Ridgeway, G. Ring, H. Schultheisz,
E. Sperling, B. Szymczyna, J. Wu
NA-protein interactions regulate gene expression
at many different points, including transcription,
mRNA processing, and translation. Translational
regulation, whereby gene expression is modulated by
controlling the synthesis of protein from the mRNA, is
a complex processes involving many different protein
factors. Often in eukaryotic mRNAs, translational regulation is mediated through RNA sequences that are
immediately downstream of the actual coding region,
but the molecular mechanism for this regulation is
poorly understood.
We are studying a set of RNA-protein complexes
that mediate translational regulation of a particular
mRNA from the nematode Caenorhabditis elegans. The
tra-2 mRNA is a major control point for both somatic
sex determination and germ-line development in this
well-studied organism (Fig. 1A). Each adult nematode
THE SCRIPPS RESEARCH INSTITUTE
233
The regulation of tra-2 expression is controlled by
binding of proteins to the downstream noncoding region
of the tra-2 mRNA. This region has 2 important protein binding sites: the tra-2 retention element (TRE)
and the tra-2 gli element (TGE). Once tra-2 mRNA is
transcribed in the nucleus, a regulatory complex is
formed by binding of the nuclear retention factor NXF-2
to the TRE (Fig. 2A). This complex is retained in the
R
F i g . 1 . Germ-line development in the nemotode C elegans. A,
Light micrograph of a nematode. Scale bar = 0.1 mm. B, Diagram
of nematode anatomy, highlighting the gonads. C, Schematic of
germ-line development in the late larval stage when sperm is produced. D, Schematic of germ-line development in the adult animal
where eggs are produced.
has 2 gonads that account for 75% of the mass of the
animal (Fig. 1B). Caenorhabditis elegans is hermaphroditic and self-fertile because of a complex regulation
of germ-line development. Late in the larval stage, the
germ line produces sperm for a brief time that are stored
for later use (Fig. 1C). The germ line switches to production of eggs in the adult nematode (Fig. 1D). A key
gene in this developmental switch from sperm production to egg production is tra-2.
F i g . 2 . Regulation of tra-2 translation by RNA-protein complexes.
A, The tra-2 retention complex is formed in the nucleus by binding
of the protein NXF-2 to the TRE, preventing normal mRNA export. B,
Export of tra-2 mRNA from the nucleus to the cytoplasm is mediated
by binding of protein TRA-1 to the TRE. C, In the cytoplasm, expression of the protein TRA-2 is repressed by binding of GLD-1 and FOG-2
to the TGE element. D, After dissociation of the GLD-1–FOG-2
complex, the tra-2 mRNA can be translated to produce the TRA-2
required for normal production of eggs in the adult nematode.
nucleus, preventing translation of tra-2 mRNA. In
order for the tra-2 mRNA to be exported, NXF-2 must
be displaced by the protein TRA-1, which also binds
to the TRE (Fig. 2B). This complex is competent to be
exported into the cytoplasm, where the tra-2 mRNA is
subjected to another level of translational regulation. In
the cytoplasm, tra-2 mRNA can be bound by the proteins GLD-1 and FOG-2, which prevent translation
(Fig. 2C) By an unknown mechanism, this repressive
complex can be dissociated to release the tra-2 mRNA
for translation of the TRA-2 protein at the appropriate
time and place (Fig. 2D).
We are trying to understand the RNA-protein interactions between NXF-2, TRA-1, GLD-1, FOG-2, and the
tra-2 mRNA. We have expressed and purified these proteins and have developed biochemical assays for binding. Currently, we are determining the structures of these
interesting and important RNA-protein complexes. All
of these proteins have homologs in the human genome,
although the details of the function of the homologs can
be quite different than the function of the nemotode
proteins. These studies are important because they will
be models for interaction of these protein domains with
234 MOLECULAR BIOLOGY
2007
RNA that will be useful for many other systems. In addition, translational regulation by protein elements that
bind downstream of the coding region in mRNAs is
poorly understood. Our biochemical and structural
approach will offer new insights into both RNA-protein
recognition and regulation of translation.
PUBLICATIONS
Karnaukhov, A.V., Karnaukhova, E.V., Williamson, J.R. Numerical matrices method
for nonlinear system identification and description of dynamics of biochemical reaction networks. Biophys. J. 92:3549, 2007.
Leontis, N.B., Altman, R.B., Berman, H.M., Brenner, S.E., Brown, J.W., Engelke,
D.R., Harvey, S.C., Holbrook, S.R., Jossinet, F., Lewis, S.E., Major, F., Mathews,
D.H., Richardson, J.S., Williamson, J.R., Westhof, E. The RNA Ontology Consortium: an open invitation to the RNA community. RNA 12:533, 2006.
Szymczyna, B.R., Gan, L., Johnson, J.E., Williamson, J.R. Solution NMR studies
of the maturation intermediates of a 13 MDa viral capsid. J. Am. Chem. Soc.
129:7867, 2007.
Vallurupalli, P., Scott, L., Hennig, M., Williamson, J.R., Kay, L.E. New RNA labeling methods offer dramatic sensitivity enhancements in 2H NMR relaxation spectra. J. Am. Chem. Soc. 128:9346, 2006.
Development of the Genetic
Code and Its Connection to
Human Disease
P. Schimmel, X.-L. Yang, J. Bacher, K. Beebe, R. Belani,
E. Chong, Z. Druzina, P. Fanta, M. Guo, M. Hanan, M. Kapoor,
E. Merriman, M. Mock, C. Motta, L. Nangle, F. Otero,
R. Reddy, W. Waas, W. Xie, W. Zhang, Q. Zhou
e focus on a group of enzymes known as
aminoacyl tRNA synthetases. These enzymes
arose early in evolution, as proteins emerged
from a putative RNA world. The 20 synthetases (1 synthetase for each amino acid) make the connection
between the nucleotide triplets of the code imbedded
in tRNAs (anticodon triplets) and the cognate amino
acids. This connection is made through the aminoacylation reactions, in which alanine is attached to yield
tRNAAla, serine to yield tRNASer, valine to yield tRNAVal,
and so on. Remarkably, this simple set of 20 enzymes
is now understood to be connected to disease. This
connection occurs in at least 2 ways.
First, in all cells, from the simplest prokaryote to
the highest eukaryote, one amino acid is often confused for another. Examples are the confusion of serine for alanine and threonine for valine. The result is
the attachment of the wrong amino acid to a tRNA and
then incorporation of the amino acid at the wrong codon
W
THE SCRIPPS RESEARCH INSTITUTE
of an mRNA, such as the incorporation of serine at the
codon for alanine. The resulting mistranslation can
cause cell death. Even a small amount of mistranslation causes ataxia through the degeneration of Purkinje cells in the cerebellum.
However, in the development of the code, these
ancient proteins (tRNA synthetases) have acquired the
capacity for editing, that is, an activity to prevent the
incorporation of an amino acid at the wrong codon. This
activity clears mischarged tRNAs, such as Ser-tRNAAla
or Thr-tRNAVal, and thereby enforces the correct amino
acid–nucleotide triplet relationships of the genetic code.
At the same time, in mammals, even mild mutations in
the editing domain of a tRNA synthetase can lead to
disease. These mutations can be vertically transmitted
to progeny, which are also put into a diseased state.
(Stronger mutations are lethal and therefore are not
transmitted.) Most recent research has shown that
defects in editing are mutagenic in aging bacteria. These
observations have raised the possibility of whether editing defects have a role in diseases of aging, including
cancer, in humans.
We showed the consequences of defects in editing
in bacteria, mammalian cells, and mice. Our research
has included the development of a special sensor that
directly detects mistranslation in mammalian cells. With
this sensor, created by using a special construction with
green fluorescent protein, we can detect the confusion
of threonine for valine. This confusion occasionally
creates Thr-tRNA Val , so that threonine is inserted at
the codons for valine. The fluorescent protein in the
sensor is a mutated form that can generate a fluorescent signal only when threonine is mistakenly inserted
at a specific codon for valine (Fig. 1). Using this sys-
F i g . 1 . Top, Construction of a biosensor based on green fluores-
cent protein that detects mistaken insertion of threonine at codons
for valine. Bottom, Decrease in fluorescent signal when threonine is
replaced by valine.
MOLECULAR BIOLOGY
2007
tem, we were able to study how a defect in the editing activity of valyl-tRNA synthetase (which normally
clears Thr-tRNAVal) caused mistranslation that had
pathologic consequences.
The second connection of aminoacyl tRNA synthetases to disease is through their expanded functions. During their long evolution, the enzymes have
acquired other activities in cell signaling pathways.
These additional functions connect the synthetases
and translation to broad biological systems, such as
the pathways for angiogenesis, inflammation, and neurogenesis. Our recent findings support the idea that
these expanded functions were added in a stepwise
way, as the tree of life grew from a common ancestor
that split into the 3 great kingdoms: prokaryotes, archae,
and eukaryotes.
For eukaryotes, the building of biological systems
for the vasculature and nervous system, for example,
was paralleled by the increasing complexity of the synthetases. Understanding this process and how aminoacylation was adapted to the acquisition of new domains
and motifs for cell signaling is one of our central goals.
In addition, we are studying specific diseases that suggest alternative functions for tRNA synthetases (Table 1).
T a b l e 1 . Examples of expanded functions of human aminoacyl
tRNA synthetases
Aminoacyl tRNA synthetase
Expanded function
TyrRS
Inflammation and angiogenesis
TrpRS
Angiogenesis
Glu-ProRS
Inflammation
GlnRS
Antiapoptosis
LysRS
Regulation of transcription
Charcot-Marie-Tooth disease (CMT) is an example
of a tRNA-synthetase–related disease that suggests an
alternative function for glycyl-tRNA synthetase (GlyRS).
CMT is the most common heritable disease of the
peripheral nervous system. Human GlyRS and one of
its mutants, G526R, which is causally associated with
CMT, have been crystallized and their structures determined at 2.95 and 2.85 Å, respectively (Fig. 2). Altogether, at least 10 disease-causing mutant alleles of
GlyRS have been reported. These mutations are scattered broadly across the primary sequence and have
no apparent unifying connection. Mapping mutations
onto our structure of human GlyRS showed their proximity to the dimer interface. G526R has an overall struc-
THE SCRIPPS RESEARCH INSTITUTE
235
F i g . 2 . Crystal structure of human GlyRS with location of the
disease-causing mutation G526.
ture similar to that of the wild-type enzyme but, through
a long-range effect, has a greater dimer interface. Experimental analyses indicate that although the CMT phenotype did not correlate with aminoacylation activity,
most mutations affect dimer formation, either enhancing (e.g., G526R) or weakening the dimer. Remarkably,
all CMT mutant GlyRSs expressed in neuroblastoma
cells were defective in their distribution into neurites.
This defect may be connected in some way to a change
in the dimer interface, which itself may interact with
other specific partners in neurons.
PUBLICATIONS
Bacher, J, Schimmel, P. An editing-defective tRNA synthetase is mutagenic in aging
bacteria via the SOS response. Proc. Natl. Acad. Sci. U. S. A. 104:1907, 2007.
Bacher, J.M., Waas, W.F., Metzgar, D., de Crecy-Lagard, V., Schimmel, P. Genetic
code ambiguity affords a selective advantage to Acinetobacter baylyi. J. Bacteriol.,
in press.
Beebe, K., Waas, W., Druzina, Z., Guo, M., Schimmel, P. A universal plate format
for high-throughput assays that monitor multiple aminoacyl tRNA synthetase activities. Anal. Biochem., in press.
Nangle, L.A., Zhang, W., Xie, W., Yang, X.L., Schimmel, P. Charcot-Marie-Tooth
disease-associated mutant tRNA synthetases linked to altered dimer interface and
neurite distribution defect. Proc. Natl. Acad. Sci. U. S. A. 104:1123, 2007.
Schimmel, P., Yang, X.-L. Perfecting the genetic code with an RNP complex. Structure 14:1729, 2006.
Xie, W., Nangle, L.A., Zhang, W., Schimmel, P., Yang, X.-L. Long-range structural
effects of a Charcot-Marie-Tooth disease-causing mutation in human glycyl-tRNA
synthetase. Proc. Natl. Acad. Sci. U. S. A. 104:9976, 2007.
Xie, W., Schimmel, P., Yang, X.-L. Crystallization and preliminary x-ray analysis of
a native human tRNA synthetase whose allelic variants are associated with Charcot-Marie-Tooth disease. Acta Crystallograph. Sect. F Struct. Biol. Cryst. Commun.
62(Pt. 12):1243, 2006.
Yang, X.-L., Guo, M., Kapoor, M., Ewalt, K.L., Otero, F.J., Skene, R.J., McRee,
D.E., Schimmel, P. Functional and crystal structure analysis of active site adaptations
of a potent anti-angiogenic human tRNA synthetase. Structure 15:793, 2007.
236 MOLECULAR BIOLOGY
2007
Structure-Function Analysis of
Expanded Activities of Human
tRNA Synthetases
X.-L. Yang, P. Schimmel, R. Belani, E. Chong, P. Fanta,
M. Guo, M. Hanan, M. Kapoor, E. Merriman, M. Mock,
C. Motta, L. Nangle, F. Otero, R. Reddy, W. Waas, W. Xie,
W. Zhang, Q. Zhou
e focus on a subgroup of the components
of the translation apparatus that in some
instances have functions beyond translation.
This subgroup is known as the aminoacyl-tRNA synthetases, 20 enzymes (1 enzyme for each amino acid)
that catalyze the first step of protein synthesis by aminoacylation of tRNAs. The alternative functions of these
synthetases in humans suggest a broad connection of
translation to signal transduction pathways in angiogenesis, inflammation, and neurogenesis.
In mammalian cells, the synthetases are gathered
together with accessory factors into a large complex
known as the multisynthetase complex (MSC; Fig. 1).
W
THE SCRIPPS RESEARCH INSTITUTE
facilitate protein synthesis and, more importantly, how
it may allow the recruitment and the release of some or
all components of the MSC in response to various cell
signals. Mammalian aminoacyl-tRNA synthetases usually have acquired additional domains, which may provide a regulatory mechanism for noncanonical functions,
as shown for human tyrosyl-tRNA synthetase (TyrRS)
and tryptophanyl-tRNA synthetase (TrpRS).
Human TyrRS has a C-terminal appended domain
that is homologous to a known human cytokine, endothelial monocyte-activating polypeptide II. The C-terminal domain and the rest of the enzyme, mini-TyrRS, can
be released by extracellular proteases such as plasmin.
Interestingly, both mini-TyrRS and the C-terminal domain
are active in cell signaling. A glutamic acid–leucine–arginine (ELR) tripeptide motif is essential for the proangiogenic activity of mini-TyrRS. The full-length TyrRS
is inactive as a cytokine. We hypothesize that the steric
block of the critical ELR with the C-terminal domain
prevents elucidation of cytokine activity in native TyrRS.
This idea is being tested by mutating a conserved tyrosine (Y341) that in our crystal structure of mini-TyrRS
is tethered to the ELR tripeptide. This mutation could
potentially open up the ELR motif in the full-length
TyrRS structure and activate the cytokine function, consistent with the idea that the purpose of natural fragmentation is to relieve steric blocking of critical epitopes
(Fig. 2).
F i g . 2 . Potential mutational activation of the cytokine activity of
human TyrRS.
F i g . 1 . The multisynthetase complex formed by 9 human tRNA
synthetases and 3 accessory factors. Other tRNA synthetases are
thought to be loosely associated with the complex.
A subset of 9 synthetases are more tightly associated
in the complex than are others; 3 specific factors, p18,
p38, and p43, serve as scaffolding. This organization
into a complex may in some way facilitate protein synthesis. Additionally or alternatively, the complex itself
may be a reservoir or depository of unactivated cytokines.
We are working on understanding the structure and
the organization of the MSC, in relation to how it may
Human TrpRS, in contrast to TyrRS, has an N-terminal appended domain. Interestingly, it has an embedded
antiangiogenic activity upon removal of the N-terminal
domain. The activation occurs naturally both by alternative splicing and by proteolysis. We hypothesize that
the activation is achieved by exposure of the active site
of TrpRS, which allows the interaction with VE-cadherin as a receptor on endothelial cells. Removal of
the N-terminal domain can expose the tryptophan
binding site, which is closed in the full-length TrpRS
as shown in our crystal structure. We are testing this
hypothesis, which could link translation with angiogene-
MOLECULAR BIOLOGY
2007
THE SCRIPPS RESEARCH INSTITUTE
237
sis via the amino acid binding pocket. Structural and
functional analyses of human TrpRS have suggested a
new way to form an active site, a way that appears to
be an adaptation to accommodate the cytokine function of the enzyme. We think the adaptation creates
selective pressure to retain the expanded function.
PUBLICATIONS
Nangle, L.A., Zhang, W., Xie, W., Yang, X.L., Schimmel, P. Charcot-Marie-Tooth
disease-associated mutant tRNA synthetases linked to altered dimer interface and
neurite distribution defect. Proc. Natl. Acad. Sci. U. S. A. 104:11239, 2007.
Schimmel, P., Yang, X.-L. Perfecting the genetic code with an RNP complex. Structure 14:1729, 2006.
Xie, W., Nangle, L.A., Zhang, W., Schimmel, P., Yang, X.-L. Long-range structural
effects of a Charcot-Marie-Tooth disease-causing mutation in human glycyl-tRNA
synthetase. Proc. Natl. Acad. Sci. U. S. A. 104:9976, 2007.
Xie, W., Schimmel, P., Yang, X.-L. Crystallization and preliminary x-ray analysis of
a native human tRNA synthetase whose allelic variants are associated with Charcot- Marie-Tooth disease. Acta Crystallograph. Sect. F Struct. Biol. Cryst. Commun.
62(Pt. 12):1243, 2006.
Yang, X.-L., Guo, M., Kapoor, M., Ewalt, K.L., Otero, F.J., Skene, R.J., McRee, D.E.,
Schimmel, P. Functional and crystal structure analysis of active site adaptations of a
potent anti-angiogenic human tRNA synthetase. Structure 15:793, 2007.
Yang, X.-L., Otero, F.J. , Ewalt, K.L., Liu, J., Swairjo, M.A., Kohrer, C., RajBhandary, U.L., Skene, R.J., McRee, D.E., Schimmel, P. Two conformations of a
crystalline human tRNA synthetase-tRNA complex: implications for protein synthesis. EMBO J. 25:2919, 2006.
Mechanisms of RNA Assembly
and Catalysis
M.J. Fedor, J.W. Cottrell, C.P. Da Costa, S. Daudenarde,
E.M. Mahen, M. Roychowdhury-Saha
ur goal is to generate fundamental insights into
catalysis by RNA enzymes. Our results contribute
to the basic knowledge of RNA structure and
function in normal growth and development and provide
a framework for developing technical and therapeutic
applications involving RNAs as targets and reagents.
The well-characterized structure of the hairpin ribozyme provides a valuable framework for investigating
the contributions of individual active-site interactions to
RNA structure and function. The hairpin ribozyme catalyzes a reversible self-cleavage reaction in which nucleophilic attack of a ribose 2′-hydroxyl on an adjacent
phosphorus proceeds through a trigonal bipyramidal
transition state that leads to the formation of 2′,3′cyclic phosphate and 5′-hydroxyl termini (Fig. 1). A network of stacking and hydrogen-bonding interactions
align the reactive phosphate in the appropriate orientation for an SN2-type nucleophilic attack and orient
O
F i g . 1 . Chemical mechanism of RNA cleavage mediated by the
family of small catalytic RNAs that includes the hairpin ribozyme.
Cleavage of the phosphodiester bond occurs through an SN2-type
mechanism that involves in-line attack of the 2′ oxygen nucleophile
on the adjacent phosphorus to form a trigonal bipyramidal transition state. Breaking of the 5′ oxygen-phosphorus bond generates
products with 5′-hydroxyl and 2′,3′-cyclic phosphate termini.
nucleotide base functional groups near the reactive
phosphate to facilitate catalytic chemistry (Fig. 2).
G+1 is the first nucleotide on the 3′ side of the
reactive phosphodiester, so interactions with G+1 stabilize the ribozyme-product complex. G+1 position has
no direct contact with the reactive phosphodiester, but
interactions between G+1 and nucleotides in loop B
define the in-line trajectory of the reactive phosphodiester. Loss of G+1 virtually eliminates catalytic activity, highlighting the significant contribution of active-site
architecture in lowering the activation barrier to catalysis. Remarkably, the structure of the ribozyme-product
complex is virtually indistinguishable from the structure of the ribozyme complex with a transition-state
mimic, making it difficult to distinguish structural contributions to ground- and transition-state stability.
The purpose of our recent biochemical experiments
was to learn if seemingly identical interactions make
different contributions to the activation barrier to catalysis, the stability of ribozyme complexes in the ground
state, and the internal equilibrium between cleavage
and ligation. We found that all modifications of the
G+1 binding pocket inhibited ligation more than they did
cleavage. These results confirm our previous evidence
that the stability of tertiary structure is the major determinant of the balance between cleavage and ligation.
Substituting seemingly equivalent functional groups
sometimes had quite different functional consequences.
In one striking example, deletion of an active-site adenine, A38, increased the activation barrier to catalysis
by more than +5 kcal/mol but reduced the groundstate stability of the ribozyme-product complex by just
+1.6 kcal/mol. These quantitative functional studies
provide one of the most detailed views yet of structurefunction relationships within a ribozyme active site.
238 MOLECULAR BIOLOGY
2007
THE SCRIPPS RESEARCH INSTITUTE
Directed Evolution of
Nucleic Acid Enzymes
G.F. Joyce, S.E. Hamilton, D.P. Horning, T.A. Jackson,
B.J. Lam, B.M. Paegel, K.L. Petrie, S.B. Voytek
t has been 40 years since Spiegelman and colleagues
first demonstrated how RNA molecules can be
evolved in the test tube. Those early experiments
involved the replication of certain viral RNAs by the
corresponding viral replicase proteins. Since then, powerful methods have been developed for amplifying (and
mutating) almost any RNA, including RNAs with catalytic function. We have been developing methods for the
in vitro evolution of RNA and applying those methods
to the discovery of novel RNAs of biochemical and biomedical significance. In addition, we are studying the
processes of darwinian evolution itself, carried out at
the level of molecules rather than in whole organisms.
I
A C O N T I N U O U S LY E V O LV I N G R N A E N Z Y M E
F i g . 2 . Network of tertiary interactions formed with G+1 that
create the hairpin ribozyme active site. A, Three-dimensional structure
of the G+1 binding pocket in a ribozyme complex with a vanadate
mimic of the transition state with the vanadate indicated in yellow.
B, Three-dimensional structure of the G+1 binding pocket in a ribozyme complex with cleavage product RNA with the 2′,3′-cyclic phosphate terminus of the 5′ cleavage product instead of the vanadate
transition-state mimic. In order to help distinguish loop A and loop B
residues, G+1 and A–1 nucleotides in loop A are colored light blue,
and carbon atoms of C25, G36, and A38 nucleotides in loop B are
colored green. Tertiary interactions between the essential loop A and B
domains of the hairpin ribozyme define the architecture of the hairpin
ribozyme active site and fix the reacting groups in the in-line orientation appropriate for the SN2-type reaction mechanism. The interdomain interface is created by the extrusion of the G+1 nucleotide from
loop A into a binding pocket in loop B. C, Diagram of interdomain
hydrogen-bonding interactions formed with the G+1 nucleotide. Modifications of nucleotides that donate or accept interdomain hydrogen
bonds were designed to probe the functional significance of the
bonds. Reproduced with permission from Cottrell, J.W., Kuzmin,
Y.I., Fedor M.J. Functional analysis of hairpin ribozyme active site
architecture. J. Biol. Chem. 282:13498, 2007. Copyright © 2007
by the American Society for Biochemistry and Molecular Biology.
PUBLICATIONS
Da Costa, C.P., Fedor, M.J., Scott, L.G. 8-Azaguanine reporter of purine ionization
states in structured RNAs. J. Am. Chem. Soc. 129:3426, 2007.
Cottrell, J.W., Kuzmin, Y.I., Fedor M.J. Functional analysis of hairpin ribozyme
active site architecture. J. Biol. Chem. 282:13498, 2007.
Many RNA enzymes have been developed by using
in vitro evolution, but special attention is directed to
those that catalyze the RNA-templated joining of RNA.
This reaction is chemically equivalent to the reaction
carried out by an RNA polymerase protein. In some
instances, RNA enzymes with simple RNA-joining activity have been further evolved to function as RNA-dependent RNA polymerases. Ultimately, such RNA enzymes
might be evolved to catalyze the production of additional
copies of the RNA enzyme itself. Then the in vitro evolution of RNA could be self-sustaining, without the need
for any proteins.
As both a step toward and a model for the selfsustained evolution of RNA, we have devised a system
for the continuous in vitro evolution of RNA enzymes
that have RNA-joining activity. Any RNA molecule in the
population that performs the desired reaction becomes
amplified to produce “progeny” molecules, which then
have the opportunity to perform the reaction again. All
of the events of continuous evolution take place within
a common reaction mixture and occur repeatedly, so long
as an adequate supply of reaction materials is maintained. When these materials are exhausted, a small
aliquot of the mixture can be transferred to a new
reaction vessel that contains a fresh supply of reagents.
In this way, we have been able to propagate large populations of evolving RNA molecules for hundreds of successive “generations.”
Until recently, all continuous in vitro evolution experiments were done with descendants of a particular RNA
MOLECULAR BIOLOGY
2007
enzyme, the “class I” RNA ligase. Within the past year,
we established a second continuously evolving RNA
enzyme based on descendants of the “DSL” RNA ligase.
Achieving this result required many rounds of stepwise in
vitro evolution under stringent selection pressure. Using a
quench-flow apparatus, we selected molecules that could
perform the RNA-joining reaction in as little as 15 milliseconds. The resulting optimized variants were capable
of initiating continuous evolution, and once continuous
evolution began, it could be carried on indefinitely.
The continuously evolving population has been
carried through 80 successive transfers, maintained
against an overall dilution of 10207-fold. The molecules
can be maintained against such extraordinarily large
dilutions because they are amplified exponentially, so
long as they have the requisite catalytic properties.
Because of the acquisition of numerous mutations, RNA
enzymes isolated after the 80th transfer (Fig. 1) had
F i g . 1 . Secondary structure of an RNA enzyme capable of under-
going continuous in vitro evolution. The arrow indicates the site of
the RNA-joining reaction between an RNA substrate and the RNA
enzyme, which is the basis for selective amplification. A typical
RNA enzyme that was obtained after 80 transfers of continuous
evolution contains 24 mutations (highlighted in black) relative to
the RNA that was present at the start of the evolution procedure.
even better performance than did the enzymes used to
initiate continuous evolution. Now that we have 2 distinct “species” of continuously evolving RNA enzymes,
we can conduct in vitro evolution studies in which the
2 are made to operate within a common environment.
These studies will allow us to explore the possibility
of competition and cooperation among evolving molecular species.
EVOLUTION ON A CHIP
We recently developed a novel approach for the continuous evolution of RNA enzymes that uses microfluidic
chip technology. With this approach, evolution is car-
THE SCRIPPS RESEARCH INSTITUTE
239
ried out in an automated fashion under computer control, with continuous monitoring of the population size
and precise manipulation of the reaction mixture. We
have used the microfluidic device to conduct evolution
experiments, beginning with a population of billions of
RNA enzymes with RNA-joining activity and carrying out
many successive rounds of RNA catalysis and selective
amplification. The concentration of RNA is monitored
by using the intercalating dye thiazole orange and a
confocal laser fluorescence microscope. Whenever a
predetermined threshold concentration is reached, the
computer initiates a set of microvalve operations to
isolate part of the reaction mixture and combine it with
a fresh supply of reagents.
In one microfluidic evolution experiment, we carried
out 363 successive dilutions of 10-fold each, progressively reducing the concentration of substrate over time.
As the continuously evolving RNA enzymes adapted to
the reduced substrate concentration, they required less
time to achieve 10-fold amplification, a situation that
was reflected in a progressively reduced time between
successive dilutions (Fig. 2). We are using this system
F i g . 2 . Microfluidic-based evolution of an RNA enzyme with RNA-
joining activity. The RNA, which had been adapted to a substrate
concentration of 5 µM, was made to undergo 100 successive 10-fold
dilutions in the presence of a substrate concentration of only 1 µM.
Each dilution was triggered when the total concentration of RNA
enzyme reached about 350 nM.
for “evolution on a chip” to develop other evolved properties and to address fundamental questions about
macromolecular evolution, such as the role of genetic
diversity in escaping evolutionary bottlenecks.
PUBLICATIONS
Joyce, G.F. Forty years of in vitro evolution. Angew. Chem. Int. Ed., in press.
Joyce, G.F. A glimpse of biology’s first enzyme. Science 315:1507, 2007.
Paegel, B.M., Grover, W.H., Skelley, A.M., Mathies, R.A., Joyce, G.F. Microfluidic
serial dilution circuit. Anal. Chem. 78:7522, 2006.
240 MOLECULAR BIOLOGY
2007
THE SCRIPPS RESEARCH INSTITUTE
Studies at the Interface of
Molecular Biology, Chemistry,
and Medicine
C.F. Barbas III, K. Albertshofer, T. Bui, S. Eberhardy,
R.P. Fuller, C. Gersbach, B. Gonzalez, R. Gordley, J. Guo,
D.H. Kim, R.L. Lerner, W. Nomura, A. Onoda,
S.S.V. Ramasastry, M. Santa Marta, L.J. Schwimmer,
D. Shabat,* F. Tanaka, U. Tschulena, N. Utsumi, Y. Ye,
K.S. Yi, Y. Yuan, H. Zhang
* Tel Aviv University, Tel Aviv, Israel
e are concerned with problems in molecular
biology, chemistry, and medicine. Many of
our studies involve learning or improving
Nature’s strategies to prepare novel molecules that perform specific functional tasks, such as regulating a gene,
destroying cancer, or catalyzing a reaction with enzymelike efficiency. We hope to apply these novel insights,
technologies, methods and their products to provide solutions to human diseases, including cancer, HIV disease,
and genetic diseases.
W
D I R E C T I N G T H E E V O L U T I O N O F C ATA LY T I C F U N C T I O N
Using reactive immunization, we have developed
antibodies that catalyze aldol as well as retro-aldol
reactions of a wide variety of molecules. The catalytic
proficiency of the best of these antibodies is almost
10 14 , a value 1000 times that of the best catalytic
antibodies reported to date and overall the best of any
synthetic protein catalyst. We have shown the efficient
asymmetric synthesis and resolution of a variety of molecules, including tertiary and fluorinated aldols, and
have used these chiral synthons to synthesize natural
products (Fig. 1). These results highlight the potential
synthetic usefulness of catalytic antibodies as artificial
enzymes in addressing problems in organic chemistry
that are not solved by using natural enzymes or more
traditional synthetic methods.
Other advances in this area include the development
of the first peptide aldolase enzymes. By using both
design and selection, we have created small peptide
catalysts that recapitulate many of the kinetic features
of large enzyme catalysts. These smaller enzymes allow
us to address the relationship between the size of natural proteins and the proteins’ catalytic efficiency.
O R G A N O C ATA LY S I S : A B I O O R G A N I C A P P R O A C H T O
C ATA LY T I C A S Y M M E T R I C S Y N T H E S I S
To further explore the principles of catalysis, we
are studying amine catalysis as a function of catalytic
F i g . 1 . A variety of compounds synthesized with the world’s
first commercially available catalytic antibody, 38C2, produced at
Scripps Research.
scaffold. Using insights garnered from our studies of
aldolase antibodies, we determined the efficacy of simple
chiral amines and amino acids for catalysis of aldol
and related imine and enamine chemistries such as
Michael, Mannich, Knoevenagel, and Diels-Alder reactions. Although aldolase antibodies are superior catalysts in terms of the kinetic parameters, these more
simple catalysts are enabling us to quantify the significance of pocket sequestration in catalysis.
Furthermore, many of these catalysts are cheap,
environmentally friendly, and practical for large-scale
synthesis. With this approach, we showed the scope
and usefulness of the first efficient amine catalysts
of direct asymmetric aldol, Mannich, Diels-Alder, and
Michael reactions. The organocatalyst approach is a
direct outcome of our studies of catalytic antibodies
and provides an effective alternative to organometallic
reactions that use severe reaction conditions and oftentoxic catalysts.
We think that our discovery that simple naturally
occurring amino acids such as L-proline and other amines
can effectively catalyze a variety of enantioselective
intermolecular reactions will change the way many
reactions will be performed. As a testament to the mild
nature of this approach, we developed the first catalytic asymmetric aldol, Mannich, Michael, and fluorination reactions involving aldehydes as nucleophiles.
Previously, such reactions were considered out of the
reach of traditional synthetic methods.
In extensions of these concepts, we designed novel
amino acid derivatives that direct the stereochemical
MOLECULAR BIOLOGY
2007
outcome of reactions in ways not possible with proline
(Fig. 2). In other studies, we created the first asymmet-
F i g . 2 . A, Proline catalyst provides access to syn-Mannich and
anti-aldol products. B, Design of (3R,5R)-5-methyl-3-pyrrolidinecarboxylic acid allows efficient access to anti-Mannich products. C,
Design of a primary amine catalyst provides access to anti-Mannich
and syn-aldol products involving hydroxyketones.
ric small-molecule aldol catalysts that are highly effective with water and seawater as solvent. We think that
our results are also relevant to the prebiotic synthesis
of the molecules of life. For example, we have shown
that our amino acid strategy can be used to synthesize
carbohydrates directly, thereby providing a provocative prebiotic route to the sugars essential for life.
We hypothesize that organocatalysis with amino acids
plays an important role in the metabolism of living
organisms today.
ANTIBODY ENGINEERING: THERAPEUTIC
ANTIBODIES, IN AND OUT OF CELLS
We developed the first human antibody phage display libraries and the first synthetic antibodies and
methods for the in vitro evolution of antibody affinity.
The ability to manipulate large libraries of human antibodies and to evolve such antibodies in the laboratory
provides tremendous opportunities to develop new medicines. Laboratories and pharmaceutical companies
around the world now apply the phage display technology that we developed for antibody Fab fragments.
In our laboratory, we are targeting cancer and HIV disease. One of our antibodies, IgG1-b12, protects animals
against primary challenge with HIV type 1 (HIV-1) and
has been further studied by many researchers. We
improved this antibody by developing in vitro evolution
THE SCRIPPS RESEARCH INSTITUTE
241
strategies that enhanced its neutralization activity. By
coupling laboratory-evolved antibodies with potent toxins, we showed that immunotoxins can effectively kill
infected cells.
We are also developing genetic methods to halt
HIV by gene therapy. We created unique human antibodies that can be expressed inside cells to make the
cells resistant to HIV infection. In the future, these antibodies might be delivered to the stem cells of patients
infected with HIV-1, allowing the development of a
disease-free immune system that would obviate the
intense regimen of antiviral drugs now required to treat
HIV disease.
Using our increased understanding of antibodyantigen interactions, we extended our efforts in cancer
therapy and developed rapid methods for creating
human antibodies from antibodies derived from other
species. We produced human antibodies that should
enable us to selectively starve a variety of cancers by
inhibiting angiogenesis and antibodies that will be used
to deliver radioisotopes to colon cancers to destroy the
tumors. We hope that these antibodies will be used in
clinical trials done by our collaborators at the SloanKettering Cancer Center in New York City.
On the basis of our studies on HIV-1, we used intracellular expression of antibodies directed against angiogenic receptors to create a new gene-based approach
to cancer. Our studies indicate that this type of gene
therapy can be successfully applied to the treatment
of cancer.
T H E R A P E U T I C A P P L I C AT I O N S O F C ATA LY T I C
ANTIBODIES
The development of highly efficient catalytic antibodies opens the door to many practical applications.
One of the most fascinating is the use of such antibodies
in human therapy. We think that use of this strategy
can improve chemotherapeutic approaches to diseases
such as cancer and AIDS. Chemotherapeutic regimens
are typically limited by nonspecific toxic effects. To
address this problem, we developed a novel and broadly
applicable drug-masking chemistry that operates in
conjunction with our unique broad-scope catalytic antibodies. This masking chemistry is applicable to a wide
range of drugs because it is compatible with virtually
any heteroatom. We showed that generic drug-masking groups can be selectively removed by sequential
retro-aldol–retro-Michael reactions catalyzed by antibody 38C2 (Fig. 3). This reaction cascade is not catalyzed by any known natural enzyme.
242 MOLECULAR BIOLOGY
2007
THE SCRIPPS RESEARCH INSTITUTE
F i g . 3 . Targeting cancer and HIV with prodrugs activated by catalytic antibodies. A bifunctional antibody is shown targeting a cancer
cell for destruction. A nontoxic analog of doxorubicin, prodoxorubicin, is being activated by an aldolase antibody to the toxic form of
the drug.
Application of this masking chemistry to the anticancer drugs doxorubicin, camptothecin, and etoposide produced prodrugs with substantially reduced
toxicity. These prodrugs are selectively unmasked by
the catalytic antibody when the antibody is applied at
therapeutically relevant concentrations. The efficacy of
this approach has been shown in in vivo models of
cancer. Currently, we are developing more potent drugs
and novel antibodies that will allow us to target breast,
colon, and prostate cancers as well as cells infected
with HIV-1. On the basis of our preliminary findings,
we think that our approach can become a key tool in
selective chemotherapeutic strategies. To see a movie
illustrating this approach, visit http://www.scripps.edu/
mb/barbas/antibody/antibody.mov.
C H E M I C A L LY P R O G R A M M E D A N T I B O D I E S : T H E
ADVENT OF CHEMOBODIES
We think that combining the chemical diversity of
small synthetic molecules with the immunologic characteristics of antibody molecules will lead to therapeutic
agents with superior properties. Therefore, we developed
a conceptually new device that equips small synthetic
molecules with both the immunologic effector functions
and the long serum half-life of a generic antibody molecule. For a prototype, we developed a targeting device
based on the formation of a covalent bond of defined
stoichiometry between (1) a 1,3-diketone derivative of
an arginine–glycine–aspartic acid peptidomimetic that
targets the integrins αvβ3 and αvβ5 and (2) the reactive
lysine of aldolase antibody 38C2 (Fig. 4). The resulting complex spontaneously assembled in vitro and in
vivo, selectively retargeted antibody 38C2 to the surface of cells expressing the integrins αvβ3 and αvβ5,
dramatically increased the circulatory half-life of the
F i g . 4 . Designed small-molecule targeting agents (SCS-873 as
shown) program the specificity of the antibody 38C2 (A). The resulting
chemobodies (cp38C2, B) have characteristics that are often superior to those of either the small molecule or the antibody alone.
peptidomimetic, and effectively reduced tumor growth
in animal models of human Kaposi sarcoma, colon cancer, and melanoma.
ZINC FINGER GENE SWITCHES AND ENZYMES
The solutions to many diseases might be simply
turning genes on or off in a selective way or adding or
deleting genes. In order to accomplish all of these aims,
we are studying molecular recognition of DNA by zinc
finger proteins and methods of creating novel zinc finger DNA-binding proteins. We showed that proteins
that contain zinc fingers can be selected or designed
to recognize novel DNA sequences.
These studies are aiding the elucidation of rules
for sequence-specific recognition within this family of
proteins. We selected and designed specific zinc finger
domains that will constitute an alphabet of 64 domains
that will allow any DNA sequence to be bound selectively. The prospects for this “second genetic code” are
fascinating and may have a major impact on basic and
applied biology.
We showed the potential of this approach in multiple mammalian and plant cell lines and in whole
organisms. With the use of characterized modular zinc
finger domains, polydactyl proteins capable of recognizing an 18-nucleotide site can be rapidly constructed
(see www.zincfingertools.org). Our results suggest that
zinc finger proteins might be useful as genetic regulators
for a variety of human aliments and provide the basis
for a new strategy of gene therapy. Our goal is to develop
this class of therapeutic proteins to inhibit or enhance
MOLECULAR BIOLOGY
2007
the synthesis of proteins, providing a direct strategy
for fighting diseases of either somatic or viral origin.
We are also developing proteins that will inhibit
the growth of tumors and others that will inhibit the
expression of a protein known as CCR5, which is a
key to infection of human cells by HIV-1. We developed
an HIV-1–targeting transcription factor that strongly
suppresses HIV-1 replication and another transcription
factor that upregulates fetal hemoglobin as a treatment
for sickle cell anemia. More recently, we have focused
on evolving zinc finger enzymes that modify the genome.
These studies have led to the development of programmable zinc finger recombinases (Fig. 5) that promise
F i g . 5 . Model of a zinc finger recombinase with programmable
specificity created by using rational design and directed molecular
evolution.
to reshape the way scientists manipulate the genome
for study and therapy of disease.
PUBLICATIONS
Abraham, S., Guo, F., Li, L.-S., Rader, C., Liu, C., Barbas, C.F. III, Lerner, R.A.,
Sinha, S.C. Synthesis of the next-generation therapeutic antibodies that combine
cell targeting and antibody-catalyzed prodrug activation. Proc. Natl. Acad. Sci.
U. S. A. 104:5584, 2007.
Albertshofer, K., Thayumanavan, R., Utsumi, N., Tanaka, F., Barbas, C.F. III.
Amine-catalyzed Michael reactions of an aminoaldehyde derivative to nitroolefins.
Tetrahedron Lett. 48:693, 2007.
Chowdari, N.S., Ahmad, M., Albertshofer, K., Tanaka, F., Barbas, C.F. III. Expedient
synthesis of chiral 1,2- and 1,4-diamines: protecting group dependent regioselectivity
in direct organocatalytic asymmetric Mannich reactions. Org. Lett. 8:2839, 2006.
Doppalapudi, V.R., Tryder, N., Li, L., Aja, T., Griffith, D., Liao, F.F., Roxas, G.,
Ramprasad, M.P., Bradshaw, C., Barbas, C.F. III. Chemically programmed antibodies: endothelin receptor targeting CovX-Bodies. Bioorg. Med. Chem. Lett.
17:501, 2007.
Gordley, R.M., Smith, J.D., Graslund, T., Barbas, C.F. III. Evolution of programmable
zinc finger-recombinases with activity in human cells. J. Mol. Biol. 367:802, 2007.
Guo, F., Das, S., Mueller, B.M., Barbas, C.F. III, Lerner, R.A., Sinha, S.C. Breaking
the one antibody-one target axiom. Proc. Natl. Acad. Sci. U. S. A. 103:11009, 2006.
Mase, N., Watanabe, K., Yoda, H., Takabe, K., Tanaka, F., Barbas, C.F. III. Organocatalytic direct Michael reaction of ketones and aldehydes with β-nitrostyrene in brine.
J. Am. Chem. Soc. 128:4966, 2006.
Rader, C., Barbas, C.F. III. Synthetic antibodies prey on B cells. Blood 108:2889,
2006.
Ramasastry, S.S.V., Zhang, H., Tanaka, F., Barbas, C.F. III. Direct catalytic asymmetric synthesis of anti-1,2-amino alcohols and syn-1,2-diols through organocatalytic anti-Mannich and syn-aldol reactions. J. Am. Chem. Soc. 129:288, 2007.
THE SCRIPPS RESEARCH INSTITUTE
243
Segal, D.J., Crotty, J.W., Bhakta, M.S., Barbas, C.F. III, Horton, N.C. Structure
of Aart, a designed six-finger zinc finger peptide, bound to DNA. J. Mol. Biol.
363:405, 2006.
Shamis, M., Barbas, C.F. III, Shabat, D. A new visual screening assay for catalytic antibodies with retro-aldol retro-Michael activity. Bioorg. Med. Chem. Lett.
17:1172, 2006.
Sinha, S.C., Das, S., Li, L.-S., Lerner, R.A., Barbas, C.F. III. Preparation of integrin αvβ3-targeting Ab 38C2 constructs. Nat. Protoc. 2:449, 2007.
Tanaka, F., Barbas, C.F. III. Aldol and Mannich-type reactions. In: Enantioselective
Organocatalysis: Reactions and Experimental Procedures. Dalko, P.I. (Ed.). WileyVCH, Weinheim, Germany, 2007, p. 19.
Utsumi, N., Zhang, H., Tanaka, F., Barbas, C.F. III. A way to highly enantiomerically enriched aza-Morita-Baylis-Hillman-type products. Angew. Chem. Int. Ed.
46:1878, 2007.
Zhang, H., Mifsud, M., Tanaka, F., Barbas, C.F. III. 3-Pyrrolidinecarboxylic acid for
direct catalytic asymmetric anti-Mannich-type reactions of unmodified ketones. J.
Am. Chem. Soc. 128:9630, 2006.
Synthetic Enzymes, Catalytic
Antibodies, Ozone Scavengers in
Asthma, Molecular Motors, and
Biomolecular Computing
E. Keinan, C.H. Lo, N. Panda, O. Reany, C. Bauer,
Y. Weiss, N. Metanis, E. Kossoy, M. Soreni, R. Piran,
M. Sinha, I. Ben-Shir, T. Shekhter, T. Ratner, T. Mejuch,
E. Solel, S. Shoshani, R. Gershoni
e focus on synthetically modified enzymes,
applications of antibody catalysis, anticancer
and antiasthma agents, molecular rotary
motors, and biomolecular computation, as illustrated
in the following examples.
W
C ATA LY S I S W I T H S Y N T H E T I C A L LY M O D I F I E D
ENZYMES
Selenoenzymes have a central role in maintaining
cellular redox potential. These enzymes have selenylsulfide bonds in their active sites that catalyze the
reduction of peroxides, sulfoxides, and disulfides. The
selenol-disulfide exchange reaction is common to all of
these enzymes, and the active site of redox potential
reflects the ratio between the forward and the reverse
rates of this reaction. Preparation of enzymes containing selenocysteine is experimentally challenging. As a
result, little is known about the kinetic role of selenols
in enzyme active sites, and the redox potential of a
selenylsulfide or diselenide bond in a protein has not
been experimentally determined.
To fully evaluate the effects of selenocysteine on
oxidoreductase redox potential and kinetics, we chem-
244 MOLECULAR BIOLOGY
ically synthesized glutaredoxin 3 and all 3 selenocysteine variants of its conserved 11CXX14C active site
and determined their redox potentials. In particular,
the position of redox equilibrium between glutaredoxin
3(C11U-C14U) (–308 mV) and thioredoxin (–270 mV)
suggests a possible role for diselenide bonds in biological
systems. Kinetic analysis indicated that the lower redox
potentials of the selenocysteine variants are due primarily to the greater nucleophilicity of the active-site
selenium. The 102- to 104-fold increase in the rate
of thioredoxin reduction by the seleno-glutaredoxin 3
analogs indicates that oxidoreductases containing either
selenylsulfide or diselenide bonds can have physiologically compatible redox potentials and enhanced
reduction kinetics in comparison with their sulfide
counterparts. This research on synthetic enzymes is
being done in collaboration with P.E. Dawson, Department of Cell Biology.
C ATA LY T I C A N T I B O D I E S
Introduction of a herbicide-resistance trait in commercial plants is highly desirable because it allows
novel herbicide management options, particularly those
that allow the control of weed species closely related
to the crop and of other undesired plant species. We
recently showed that herbicide-resistant plants can be
engineered by designing both a herbicide and a catalytic antibody that destroys the herbicide within the
plants (Fig. 1). First, we developed a new carbamate
F i g . 1 . Influence of herbicide (4) on the rooting and development
of seedlings of F1 hybrids and control A thaliana plants. The control
plants are shown in A and C; the hybrid plant lines (F1) expressing
both light and heavy chains of catalytic antibody 38C2, in B and
D. Plantlets grown on medium without the herbicide are shown in
A and B; those grown with the herbicide are shown in C and D.
2007
THE SCRIPPS RESEARCH INSTITUTE
herbicide that can be catalytically destroyed by the
aldolase antibody 38C2. Then we separated expression
of the light chain and half of the heavy chain (Fab) of
the catalytic antibody in the endoplasmic reticulum of
2 plant lines of Arabidopsis thaliana. Finally, we used
cross-pollination of these 2 transgenic plants to produce
a herbicide-resistant F1 hybrid.
O Z O N E S C AV E N G E R S W I T H A N T I A S T H M A A C T I V I T Y
A new hypothesis we proposed for the mechanism
of asthmatic inflammation has led to an ozone-scavenging compound that prevents bronchial obstruction
in rats with asthma. Previously, scientists at Scripps
Research discovered that ozone can be generated not
only via the antibody-mediated water oxidation pathway but also by antibody-coated activated white blood
cells during inflammatory processes. This finding led
us to speculate that the pulmonary inflammation in
asthma might be caused by ozone production by white
blood cells in lungs and that inhalation of electronrich olefins, which are known ozone scavengers, might
have antiasthmatic effects. In experiments in rats,
inhalation of such a compound, limonene, caused a
significant improvement in signs of asthma. These
results could have consequences in the management
of asthma.
M O L E C U L A R R O TA R Y M O T O R S
The hypothesis that molecular-scale rotary motors
can be designed and constructed from synthetic components is based on the available structural information on the biological precedents, such as bacterial
flagellar motors and ATP synthase, which interconvert
chemical energy and coordinated mechanical motion.
Synthetic motors could offer considerable advantages in
the development of complex nanomachinery because
the motors can tolerate a more diverse range of conditions than biological machines can. One of the most
challenging design elements of molecular rotary motors
is the need for high-speed rotation, which requires low
“friction.” We suggest that in order to achieve minimal
friction, the members of the rotor-stator couple should
repel each other.
We propose a new approach to the design of frictionless molecular rotary motors in which a rotaxanetype architecture in which a macrocyclic cucurbituril
host serves as a stator and a rigid, polyyne guest serves
as a rotor. The feasibility of the key design element, the
repulsive interaction between these components, is
supported by molecular mechanics calculations with
model systems and has been experimentally confirmed
MOLECULAR BIOLOGY
2007
by microcalorimetry and by x-ray crystallography with
several synthetic host-guest complexes. The results all
suggest that the diyne rod floats at the center of the
macrocyclic host with no apparent van der Waals contacts between the rod and the host (Fig. 2).
THE SCRIPPS RESEARCH INSTITUTE
245
design of a 3-state–3-symbol automaton, thus increasing the number of syntactically distinct programs from
765 to 1 billion. We have also amplified the applicability of this design by using surface-anchored input
molecules and surface plasmon resonance technology
to monitor the computation steps in real time. This
technology allowed parallel computation and automatic,
real-time detection with DNA chips that have multiple
input molecules and can be used as pixel arrays for
image encryption.
PUBLICATIONS
Kossoy, E., Lavid, N., Soreni-Harari, M., Shoham, Y., Keinan, E. A programmable
bio-molecular computing machine with bacterial phenotype output. Chembiochem,
in press.
Lo, H.C., Han, H., D’Souza, L.J., Sinha, S.C., Keinan, E. Rhenium(VII) oxide catalyzed heteroacylative ring-opening dimerization of tetrahydrofuran. J. Am. Chem.
Soc. 129:1246, 2007.
Lo, H.C., Iron, M.A., Martin, J.M.L., Keinan, E. Proton walk in the aqueous platinum
complex [TpPtMeCO] via a sticky σ-methane ligand. Chem. Eur. J. 13:2812, 2007.
Metanis, N., Keinan, E., Dawson, P.E. Synthetic seleno-glutaredoxin 3 analogues
are highly reducing oxidoreductases with enhanced catalytic efficiency. J. Am.
Chem. Soc. 128:16684, 2006.
Weiss, Y., Rubin, B., Shulman, A., Ben Shir, I., Keinan, E., Wolf, S. Determination of plant resistance to carbamate herbicidal compounds inhibiting cell division
and early growth by seed and plantlets bioassays. Nat. Protoc. 1:2282, 2006.
F i g . 2 . Solid-state structure of inclusion complexes of cucurbituril with 1 (A, side view; B, top view), 2 (C), and 3 (D). Red indicates oxygen; blue, nitrogen; gray, carbon; brown, sulfur; and green,
chloride. Hydrogen atoms have been omitted for clarity. The com-
plexes and part of their environment are shown. The structures were
created by using the Oak Ridge Thermal Ellipsoid Plot computer
program for illustrating crystal structures.
Macromolecular Interactions:
Evolution, Engineering,
and Detection
BIOMOLECULAR COMPUTING DEVICES
Previously, we described the first nanoscale, programmable finite automaton with 2 symbols and 2 states
that computed autonomously. All of the components
of the device, including hardware, software, input, and
output, were biomolecules mixed together in solution.
The hardware consisted of a restriction nuclease and
a ligase; the software (transition rules) and the input
were double-stranded DNA oligomers. Computation was
carried out by processing the input molecule via repetitive cycles of restriction, hybridization, and ligation
reactions to produce a final-state output in the form of
a double-stranded DNA molecule.
Recently, we have taken the concept of molecular
computing a step further by constructing computing
devices in which the computation output is a specific
biological function rather than a specific molecule. In
addition, we markedly increased the levels of complexity and mathematical power of these automata by the
V.V. Smider, C.C. Liu, J. Mills, B. Hutchins, B. Leonard
e are involved in several projects related to
molecular recognition. These include understanding the evolution and biochemical characteristics of the human germ-line antibody repertoire,
chemically enhancing antibody properties, analyzing
RNA hairpin interactions that regulate DNA replication,
and developing new chemical technologies to detect
interactions between macromolecules.
W
DEVELOPMENT AND BIOCHEMICAL PROPERTIES OF
THE GERM-LINE IMMUNE REPERTOIRE
Some germ-line antibodies recognize certain carbohydrate antigens with high avidity, suggesting that
darwinian forces helped shape the antibody repertoire.
We have developed techniques to rapidly combinatorially pair heavy and light chains. Using these techniques,
we will be able to construct and analyze a large germline repertoire in a spatially addressed format. Libraries
246 MOLECULAR BIOLOGY
2007
will be produced and analyzed for binding to antigens
of common human pathogens to provide a repertoire
binding “snapshot.” Such data could lend insight into
the evolutionary forces behind the development of certain V regions and VH-VL combinations.
THE SCRIPPS RESEARCH INSTITUTE
ENHANCED AND NOVEL INTERACTIONS:
P.G. Schultz, Department of Chemistry, we are engineering other metalloantibodies and are incorporating
amino acids with unique side chains (unnatural amino
acids) into antibodies. This technology should allow
both genetic and chemical expansion of the antibody
repertoire for therapeutic and diagnostic applications.
ENGINEERED ANTIBODIES
R N A H A I R P I N S R E G U L AT I N G D N A R E P L I C AT I O N
Antibodies bind their antigens noncovalently. However, for certain diagnostic or therapeutic applications,
antibodies that irreversibly bind or modify their antigens would be useful. We recently succeeded in engineering a metal-dependent antibody that irreversibly
binds its antigen, TNF-α, perhaps through an exchange
inert cobalt complex (Fig. 1). In collaboration with
RNA molecules are now recognized as important
regulators of many cellular properties via RNA-RNA
interactions. One of the simplest and evolutionarily
ancient mechanisms of RNA-mediated regulation occurs
at plasmid origins of replication. We are applying molecular evolution techniques to these regions and using
simple copy number and compatibility assays to ascertain the regulatory and recognition features of these hairpins in Escherichia coli.
B I O M O L E C U L A R D E T E C T I O N : O X A L AT E E S T E R
CHEMIFLUORESCENCE
Sensitive methods such as fluorescence can be used
to detect multiple macromolecular interactions simultaneously (multiplex detection). An alternative approach
is detection via chemifluorescence, in which a fluorescent dye is activated through chemical transfer of energy.
We recently synthesized aqueous oxalate ester compounds capable of exciting fluorescent dyes conjugated
to biomolecules (Fig. 2). Unlike fluorescence, chemi-
F i g . 1 . An antibody scFv that binds its antigen irreversibly. Top,
Western blot shows a complex between an engineered scFv 21H9
and its TNF antigen at 44 kD (lanes 6 and 9) that only forms in the
presence of cobalt. Both scFv and TNF are in the complex, as indicated
by staining with antibody to TNF (lane 9a) and antibody to scFv
(lane 9b). The parental scFv RA72, which binds TNF noncovalently,
does not form the 44-kD complex (lanes 1–4). Bottom, Close-up
model of the 21H9 scFv (green) with its metal binding site (red) in
proximity to mapped linkage regions (pink) of TNF. The TNF trimer
is shown as an outset with linkage residues in cyan and pink.
F i g . 2 . A water-soluble oxalate ester can activate fluorescently
labeled protein. Top, The postulated reaction between an oxalate
ester and hydrogen peroxide to produce a dioxetane intermediate,
which can transfer energy to a fluorescent dye to produce carbon
dioxide and light emission at the dye’s characteristic wavelength.
Bottom, Samples of fluorescently labeled bovine serum albumin
(BSA-ROX, BSA-JOE, BSA-Cy3, and BSA-Cy5) adsorbed to a membrane and exposed to activated water-soluble oxalate ester emit
light of characteristic wavelengths.
MOLECULAR BIOLOGY
2007
fluorescence allows integration of the emission signal
over time, a situation that may result in highly sensitive detection of molecular interactions. Applications
include microarrays, blots, and solid-phase immunoassays in research and diagnostic settings.
PUBLICATIONS
Leonard, B., Sharma, V., Smider, V. Co-expression of antibody Fab heavy and light
chain genes from separate evolved compatible replicons in E. coli. J. Immunol.
Methods. 317:56, 2006.
Functional Characterization of
Enzymes via Combinatorial
Libraries
THE SCRIPPS RESEARCH INSTITUTE
247
In addition to using the PNA-encoded libraries to
discover new enzymes and enzyme functions, we have
also used the libraries as an ultra-high-throughput
screening platform for the identification of inhibitors
and substrates for individual enzymes. A library of more
than 500 inhibitors can be easily screened in just a
few drops of solution.
Another example of functional characterization of
protein activity is the profiling of the substrate specificity
of cysteine proteases, metacaspases, from Arabidopsis
thaliana in collaboration with F. Van Breusegem, Ghent
University, Ghent, Belgium. Using a substrate library of
approximately 160,000 fluorogenic substrates, we characterized the structure-activity relationship of the metacaspases and facilitated the identification of the first
reported endogenous plant protease-inhibitor interaction.
J.L. Harris, J. Alves
e are developing and using technologies based
on small-molecule protein modifiers to profile the active state of enzymes in various
biological environments.
In collaboration with N. Winssinger, Université Louis
Pasteur, Strasbourg, Germany, we have developed an
encoding strategy that uses peptide nucleic acid (PNA)
sequences. Encoding combinatorial libraries with PNA
tags allows not only for capture of synthetic history of
the library in the resulting molecule but also for spatial
deconvolution of the molecules on DNA microarrays.
The PNA tag is covalently cosynthesized with the smallmolecule compound that can interact with enzymes
in biological samples. With this approach, we can efficiently synthesize thousands of molecules in the same
time it would take to synthesize a single individual molecule. Although the resulting library of thousands of compounds is screened as a mixture, the specific molecules
within the library can be easily identified on a DNA
microarray by using the hybridization properties of the
encoded PNA tag.
We have developed PNA-encoded libraries to profile multiple enzyme classes, including cysteine proteases, serine proteases, and, most recently, kinases.
The screening platform has also been adapted to profile samples in parallel by using readily available laboratory equipment. These libraries have been applied to
various biological systems, including airborne allergens,
malaria, and viral samples, and have resulted in the
identification and characterization of enzymes within
those systems that may result in potential new therapies to combat the associated diseases.
W
PUBLICATIONS
Epple, R., Urbina, H.D., Russo, R., Liu, H., Mason, D., Bursulaya, B., Tumanut,
C., Li, J., Harris, J.L. Bicyclic carbamates as inhibitors of papain-like cathepsin
proteases. Bioorg. Med. Chem. Lett. 17:1254, 2007.
Urbina, H.D., Debaene, F., Jost, B., Bole-Feysot, C., Mason, D.E., Kuzmic, P.,
Harris, J.L., Winssinger, N. Self-assembled small-molecule microarrays for protease screening and profiling. Chembiochem 7:1790, 2006.
Vercammen, D., Belenghi, B., van de Cotte, B., Beunens, T., Gavigan, J.A., De
Rycke, R., Brackenier, A., Inze, D., Harris, J.L., Van Breusegem, F. Serpin1 of
Arabidopsis thaliana is a suicide inhibitor for metacaspase 9. J. Mol. Biol.
364:625, 2006.
Prodrug and Targeting Therapies
and Synthesis of Anticancer and
Antibacterial Agents
S.C. Sinha, R.A. Lerner, Z. Chen, S. De, Z.-Z. Huang
ne of our main research interests is antibody
catalysis, with emphasis on its applications in
organic synthesis and selective drug delivery.
In the past, using monoclonal aldolase antibodies 38C2
and 93F3, we synthesized numerous natural products,
including highly potent cytotoxic agents such as epothilones and their analogs. Currently, we are developing
these antibodies for use in selective cancer chemotherapy. We are using 2 related approaches. In the
first, we are using a small-molecule antagonist of the
integrin αvβ3 to redirect antibody 38C2 to cancer cells.
In the second, we are using antibody 38C2 or 93F3
to catalyze prodrug activation in an “antibody prodrug
therapy”; our aim with this method is to reduce indiscriminate toxic effects to normal cells. In addition, we
O
248 MOLECULAR BIOLOGY
2007
THE SCRIPPS RESEARCH INSTITUTE
are focusing on the synthesis of anticancer and antibacterial natural products and their analogs.
S E L E C T I V E C H E M O T H E R A P Y W I T H C ATA LY T I C
ALDOLASE ANTIBODIES
Using integrin αvβ3–targeting small molecules and
antibody 38C2, we have produced 2 types of 38C2
constructs: catalytic and noncatalytic (Fig. 1). The cata-
F i g . 2 . Structure of integrin-targeting compounds for the 38C2-
construct formation, doxorubicin prodrugs, the adjacent bis-tetrahydrofuran annonaceous acetogenins and their precursors, and
sorangiolide and its macrocyclic lactone intermediates.
ment of Molecular Biology, and B. Felding-Habermann,
Department of Molecular and Experimental Medicine.
F i g . 1 . Schematic drawing of the integrin αvβ3–targeting noncatalytic and catalytic antibody constructs. Abbreviations: Ab, antibody; TA, targeting agent.
lytic construct is a classical conjugate of small molecules
to 38C2 through the surface lysine residues or the sulfide groups obtained by reduction of the disulfide bridge
in the antibody hinge region. In the noncatalytic construct, the compound conjugates in the 38C2 binding
sites and redirects the antibody to cells. Using mass
spectrometry, we found that both constructs have approximately 2 molecules of a small molecule. In flow cytometry assays, both constructs bound cells expressing
integrin αvβ3 with high affinity.
Because integrin αvβ3 is expressed on numerous
primary and metastatic tumor cells as well as in the
vasculature of the tumors, the resulting antibody constructs are expected to be highly useful in cancer therapy. Using MDA-MB-231 cells in a mouse model of
breast cancer metastasis, we evaluated the efficacy of
the noncatalytic constructs. The animals treated with
these constructs had fewer metastases than those treated
with the small molecule alone or the control group did.
Similarly, the catalytic 38C2 construct (III in Fig. 1)
activated doxorubicin prodrugs and caused cytotoxic
effects in MDA-MB-231 breast cancer cells in vitro. For
the studies with the catalytic construct, we also designed
a series of doxorubicin prodrugs (Fig. 2) and evaluated
them in vitro. Our selective chemotherapy studies are
carried out in collaboration with C.F. Barbas, Depart-
SYNTHESIS OF ANTICANCER AND ANTIBACTERIAL
AGENTS
During the past year, we mainly focused on the
synthesis of the anticancer adjacent bis-tetrahydrofuran annonaceous acetogenins and total synthesis of
the antibacterial agent sorangiolide (Fig. 2). Annonaceous acetogenins are highly cytotoxic compounds, and
each acetogenin can have as many as 64 stereoisomers
because of the bis-tetrahydrofuran fragment alone, which
is flanked by a hydroxy group on each side. Understanding the structure-activity relationship of these compounds and carrying out comprehensive biological studies
require total synthesis of all 64 stereoisomers of each
acetogenin. These syntheses have been the focus of
research for several years not only by us but also by
many others. However, despite a continued effort, synthesis of all 64 stereoisomers of asimicin or bullatacin,
which are among the most cytotoxic adjacent bis-tetrahydrofuran acetogenins, has not yet been achieved. We
have now prepared 10 bis-tetrahydrofuran precursors
that will yield all 64 asimicin stereoisomers.
Sorangiolides A and B are 18-membered macrocyclic
lactones isolated from Sorangium cellulosum strain
So ce12. They have weak antibiotic activity against
gram-positive bacteria (e.g., Staphylococcus aureus).
Although the structure of these compounds was confirmed by an x-ray analysis of sorangiolide A, no synthetic studies have been reported. Moreover, the relative
or absolute stereochemistry of the hydroxy group in
MOLECULAR BIOLOGY
sorangiolide B at C-6 is yet to be determined. Therefore, to synthesize the naturally occurring sorangiolides
and their analogs for biological evaluations, we designed
and produced advanced macrocyclic precursors (Fig. 2),
which will be converted to the target compounds and
their analogs.
PUBLICATIONS
Abraham, S., Guo, F., Li, L.-S., Rader, C., Liu, C., Barbas, C.F. III, Lerner, R.A.
Sinha, S.C. Synthesis of the next-generation therapeutic antibodies that combine
cell targeting and antibody-catalyzed prodrug activation. Proc. Natl. Acad. Sci.
U. S. A. 104:5584, 2007.
Das, S., Abraham, S., Sinha, S.C. Studies toward the total synthesis of sorangiolides and their analogs: a convergent stereoselective synthesis of the macrocyclic
lactone precursors. Org. Lett. 9:2273, 2007.
Lo, H.C., Han, H., D’Souza, L.J., Sinha, S.C., Keinan, E. Rhenium(VII) oxide-catalyzed heteroacylative ring-opening dimerization of tetrahydrofuran. J. Am. Chem.
Soc. 129:1246, 2007.
Sinha, S.C., Das, S., Li, L.-S., Lerner, R.A., Barbas, C.F. III. Preparation of integrin αvβ3-targeting Ab 38C2 constructs. Nat. Protoc. 2:449, 2007.
Chemical Transformations of
Small Molecules and Proteins
and Reactions Between Them
F. Tanaka, K. Albertshofer, L. Asawapornmongkol,
C.F. Barbas III, R.P. Fuller, H.-M. Guo, M. Minakawa,
S.S.V. Ramasastry, N. Utsumi, H. Zhang
e create and develop molecules and methods
that contribute to basic biomedical research
and to the development of drug candidates
and therapeutic strategies.
One aspect of our research is the development of
new strategies and methods for labeling of proteins with
small molecules at certain positions. We are creating
protein and peptide tags that selectively form covalent
bonds with designed synthetic molecules; these tags
are generated by using reaction-based selections from
combinatorial libraries of proteins and peptides. When
the tag is fused to a protein of interest, the fusion
protein can be modified at the tag sequence through
the reaction with the designed synthetic compound.
Designed compounds can be conjugated with fluorescent molecules and drug molecules by chemical synthesis; therefore, through use of this labeling system,
a variety of molecules can be covalently and selectively
attached to a protein of interest.
We are also developing synthetic labeling molecules that selectively and covalently react with target
W
2007
THE SCRIPPS RESEARCH INSTITUTE
249
proteins without preengineering of proteins and without need for fusion to a tag sequence. In this strategy,
labeling molecules contain a moiety that selectively
and covalently reacts with a specific amino acid residue (e.g., tyrosine) and a moiety that noncovalently
interacts with a target protein. Specificity for the target protein is imparted by noncovalent interaction. With
this system, naturally occurring proteins of interest
can be selectively tagged with synthetic molecules in
natural environments, including cells and animals.
When a labeling molecule contains a moiety that
binds to the surface of a protein, the protein can be
labeled without affecting the protein’s function. On the
other hand, when the specificity of a labeling molecule
is imparted through a ligand that binds to the active
site of an enzyme, the resulting molecule can be an
efficient, selective, irreversible inhibitor of the enzyme.
For example, we are developing labeling molecules
and irreversible inhibitors for carbonic anhydrases.
These molecules will be useful for study of carbonic
anhydrases and may also lead to the development of
efficient inhibitors of these enzymes for treatment of
diseases, including cancer.
We are also developing synthetic methods for concise access to functionalized molecules in regioselective,
diastereoselective, and enantioselective fashions. These
methods are useful for syntheses of bioactive molecules,
ligands of proteins, and candidates of these molecules.
Overall, our research provides insight into the construction of functional molecules and into biological
functions of naturally occurring molecules such as proteins and enzymes.
PUBLICATIONS
Albertshofer, K., Thayumanavan, R., Utsumi, N., Tanaka, F., Barbas, C.F. III.
Amine-catalyzed Michael reactions of an aminoaldehyde derivative to nitroolefins.
Tetrahedron Lett. 48:693, 2007.
Chowdari, N.S., Ahmad, M., Albertshofer, K., Tanaka, F., Barbas, C.F. III. Expedient
synthesis of chiral 1,2- and 1,4-diamines: protecting group dependent regioselectivity
in direct organocatalytic asymmetric Mannich reactions. Org. Lett. 8:2839, 2006.
Mase, N., Watanabe, K., Yoda, H., Takabe, K., Tanaka, F., Barbas, C.F. III.
Organocatalytic direct Michael reaction of ketones and aldehydes with β-nitrostyrene in brine [published addition/correction appears in J. Am. Chem. Soc.
128:17153, 2006]. J. Am. Chem. Soc. 128:4966, 2006.
Ramasastry, S.S.V., Zhang, H., Tanaka, F., Barbas, C.F. III. Direct catalytic asymmetric synthesis of anti-1,2-amino alcohols and syn-1,2-diols through organocatalytic anti-Mannich and syn-aldol reactions. J. Am. Chem. Soc. 129:288, 2007.
Tanaka, F., Barbas, C.F. III. Aldol and Mannich-type reactions. In: Enantioselective
Organocatalysis: Reactions and Experimental Procedures. Dalko, P.I. (Ed.). WileyVCH, Weinheim, Germany, 2007, p. 19.
Tanaka, F., Fuller, R. Control of function of a small peptide by a protein. Bioorg.
Med. Chem. Lett. 16:4059, 2006.
250 MOLECULAR BIOLOGY
2007
Tanaka, F., Fuller, R., Asawapornmongkol, L., Warsinke, A., Gobuty, S., Barbas,
C.F. III. Development of a small peptide tag for covalent labeling of proteins. Bioconjug. Chem. 18:1318, 2007.
Tanaka, F., Jones, T., Kubitz, D., Lerner, R.A. Anti-formyl peptide antibodies.
Bioorg. Med. Chem. Lett. 17:1943, 2007.
Utsumi, N., Zhang, H., Tanaka, F., Barbas, C.F. III. A way to highly enantiomerically enriched aza-Morita-Baylis-Hillman-type products. Angew. Chem. Int. Ed.
46:1878, 2007.
Zhang, H., Mifsud, M., Tanaka, F., Barbas, C.F. III. 3-Pyrrolidinecarboxylic acid for
direct catalytic asymmetric anti-Mannich-type reactions of unmodified ketones. J.
Am. Chem. Soc. 128:9630, 2006.
Structure, Function, and
Applications of Virus Particles
J.E. Johnson, M. Banerjee, Z. Chen, I. Gertsman, R. Huang,
R. Khayat, G. Lander, J. Lanman, K.K. Lee, T. Matsui,
P. Natarajan, A. Odegard, J. Speir, R. Taurog
e investigate model virus systems that provide insights for understanding viral assembly,
maturation, entry, localization, and replication. We have also developed viruses as reagents for
applications in nanomedicine, chemistry, and biology.
We investigate viruses that infect bacteria, insects,
plants, and the extreme thermophile Sulfolobus. These
viruses have genomes of single-stranded RNA, and
double-stranded DNA.
We use a variety of physical methods to investigate structure-function relationships, including singlecrystal x-ray diffraction, static and time-resolved solution
x-ray diffraction, electron cryomicroscopy and image
reconstruction, mass spectrometry, structure-based
computational analyses, and methods associated with
thermodynamic characterization of virus particles and
their transitions. Biological methods we use include
the genetic engineering of viral genes and their expression in Escherichia coli, mammalian cells, insect cells,
and yeast and the characterization of these gene products by physical methods. For cytologic studies of viral
entry and infection, we use fluorescence and electron
microscopy and particles assembled in heterologous
expression systems. Our studies depend on extensive
consultations and collaborations with others at Scripps
Research, including groups led by B. Carragher, M.G.
Finn, M. Manchester, D.R. Millar, C. Potter, V. Reddy,
A. Schneemann, G. Siuzdak, J.R. Williamson, and
M.J. Yeager, and a variety of groups outside of Scripps.
W
DOUBLE-STRANDED DNA VIRUSES
HK97 is a double-stranded DNA virus similar to
bacteriophage λ. It undergoes a remarkable morpho-
THE SCRIPPS RESEARCH INSTITUTE
genesis in its assembly and maturation, and this process can be recapitulated in vitro. We determined the
atomic resolution structure of the 650-Å mature, head II
particle and discovered the mechanism used to concatenate the subunits of the particle into a chain-mail fabric
similar to that seen in armor of medieval knights. In the
past year, we focused on the dynamics of maturation.
Prohead II is a 500-Å metastable intermediate at
pH 7 that can be induced to begin maturation by lowering the pH to 4. Solution x-ray scattering and single-molecule fluorescence showed that the initial transition to
a particle of about 560 Å occurs as a highly cooperative, stochastic event with no detectable intermediates
that takes place in less than 1 second for an individual
particle. A quorum of cross-links must form in this particle to generate the second expansion intermediate (about
650 Å), which also forms cooperatively with no detectable intermediates. At pH 4, formation of cross-links
continues, with 360 formed per particle. The late stage
of maturation is a classic Brownian ratchet in which
pentameric subunits fluctuate like a piston through a
radial trajectory of 15 Å and are trapped at the top of
the trajectory by formation of the covalent cross-link.
If the cross-link can not form, the maturation stops
with the pentamers still sampling the trajectory.
Bacteriophage P22 is the prototype of the Podoviridae that are characterized by a T = 7 capsid with a
short tail structure incorporated into a unique 5-fold
vertex. We determined an asymmetric reconstruction
of this particle that revealed spooled DNA, the dodecameric portal, and the location of the 9 gene products known to be in the particle. Recently, structures
of bacteriophage λ were determined at subnanometer
resolution by electron cryomicroscopy. These structures showed that the fold of the capsid protein is the
same as that of the HK97 subunit.
Sulfolobus turreted icosahedral virus is an archaeal
virus isolated from Sulfolobus, which grows in the
acidic hot sulfur springs (pH 2–4, 72°C–92°C) in Yellowstone National Park. An electron cryomicroscopy
reconstruction of the virus showed that the capsid
has pseudo T = 31 quasi symmetry and is 1000 Å
in diameter, including the pentons. The x-ray structure of the major capsid protein of the virus revealed
a fold nearly identical to the folds of the major capsid
proteins of the eukaryotic adenoviruses and PRD-1, a
virus that infects bacteria. These findings indicate a
viral phylogeny that spans the 3 domains of life. Difference electron density maps in which the x-ray model
MOLECULAR BIOLOGY
2007
is subtracted from the electron cryomicroscopy density
clearly show an internal membrane in which the capsid proteins are anchored.
THE SCRIPPS RESEARCH INSTITUTE
251
Design and Informatics in
Structural Virology
SINGLE-STRANDED RNA VIRUSES
Flock House virus is a T = 3, single-stranded RNA
virus that infects Drosophila. We are studying viral entry
and early expression and assembly of the capsid protein. Recently, studies on viral entry indicated the presence of an “eluted” particle early in infection that has
initiated its disassembly program but is then eluted back
into the medium. We did a phenotypic characterization
of the particles, and we are using electron cryomicroscopy to study them. For studies on the expression and
assembly of the capsid protein, we are using tetra-cysteine tags inserted genetically in the capsid protein that
allow the freshly made proteins to be optically visualized
with a fluorophore and in the electron microscope with
photoconversion of the fluorophore. Recently, high-pressure freezing of infected cells revealed exceptionally
detailed features of viral entry and regions of replication
within the cell. Tomographs prepared with the micrographs show that translation of the RNA encoding the
capsid protein and the assembly of virions takes place
within chambers created by remodeled mitochondria.
Refined atomic models of tetravirus structures and
structure-based mutagenesis combined with highly sensitive assays for defining phenotypes have revealed the
electrostatic principles of maturation for the T = 4
tetraviruses.
PUBLICATIONS
Cheung, C.L., Chung, S.W., Chatterji, A., Lin, T., Johnson, J.E., Hok, S., Perkins,
J., De Yoreo, J.J. Physical controls on directed virus assembly at nanoscale chemical templates. J. Am. Chem. Soc. 128:10801, 2006.
Gan, L., Speir, J.A., Conway, J.F., Lander, G., Cheng, N., Firek, B.A., Hendrix,
R.W., Duda, R.L., Liljas, L., Johnson, J.E. Capsid conformational sampling in
HK97 maturation visualized by x-ray crystallography and cryo-EM. Structure
14:1655, 2006.
Johnson, J.E., Chiu, W. DNA packaging and delivery machines in tailed bacteriophages. Curr. Opin. Struct. Biol. 17:237, 2007.
Maia, L.F., Soares, M.R., Valente, A.P., Almeida, F.C., Oliveira, A.C., Gomes,
A.M., Freitas, M.S., Schneemann, A., Johnson, J.E., Silva, J.L. Structure of a
membrane-binding domain from a non-enveloped animal virus: insights into the
mechanism of membrane permeability and cellular entry. J. Biol. Chem.
281:29278, 2006.
Martin, B.D., Soto, C.M., Blum, A.S., Sapsford, K.E., Whitley, J.L., Johnson, J.E.,
Chatterji, A., Ratna, B.R. An engineered virus as a bright fluorescent tag and scaffold
for cargo proteins: capture and transport by gliding microtubules. J. Nanosci. Nanotechnol. 6:2451, 2006.
Poliakov, A., van Duijn, E., Lander, G., Fu, C.Y., Johnson, J.E., Prevelige, P.E.,
Jr., Heck, A.J. Macromolecular mass spectrometry and electron microscopy as
complementary tools for investigation of the heterogeneity of bacteriophage portal
assemblies. J. Struct. Biol. 157:371, 2007.
V.S. Reddy, G.V. Subbarao, S. Venkataraman, M. Tripp,
P. Singh, R. Mannige, I. Borelli, J. Loo, S. Kumar
e are interested in identifying and understanding the structural underpinnings and
requirements for formation and function of
viral capsids. We use this information to design novel
protein shells that polyvalently display multiple copies
of peptides or proteins of interest. We use structural,
computational, bioinformatics, and genetic methods.
Viruses are highly evolved macromolecular assemblages that perform a variety of functions during their
life cycle, including self-assembly into uniform capsids,
selective packaging of the genome, binding to host cells,
and delivery of the genetic material to the targeted cells.
Simple viruses, such as nonenveloped viruses, form
capsids with homogeneous composition and quaternary architecture. Hence, these viruses are useful for
structural and functional analyses.
In collaboration with G.R. Nemerow, Department
of Immunology, we are using x-ray crystallography to
determine the structure of the entire human adenovirus particle, currently at 7.5-Å resolution. Acquisition
of high-resolution data is under way. We recently determined the structure of Seneca Valley virus, which belongs
to a new genus (Senecavirus) of the Picornaviridae
family. Senecaviruses are of particular interest because
they are selectively pathogenic to cancer cells. This
research was done in collaboration with scientists at
Neotropix, Inc., Malvern, Pennsylvania.
We continue to maintain and expand the virus
structural database, namely VIPERdb (http://viperdb
.scripps.edu), a Web portal for structures and associated structural properties of viral capsids. The capsid
structures in the database were analyzed in terms of
protein-protein interactions, contacting residue pairs,
association energies, individual residue contributions,
and surface characteristics by using computational methods. The results of the analysis are stored in the database. VIPERdb is being developed and maintained as
part of the Multiscale Modeling Tools for Structural Biology, a National Institutes of Health research resource
headed by C.L. Brooks, Department of Molecular Biology.
We are also actively involved in generating novel
vaccines against cytotoxins such as ricin and against
pathogens by expressing antigenic regions of patho-
W
252 MOLECULAR BIOLOGY
2007
genic molecules on the surfaces of viral capsids. Currently, tomato bushy stunt virus–like capsids are our
display platform of choice. A unique subunit fold of
the viral subunit enables the attachment of peptides
or proteins of interest to the C terminus in a linear
fashion and their display on the capsid surface. Such
reagents act as multivalent decoys of the pathogenic
molecules and so can be used as potential vaccines.
PUBLICATIONS
Kumar, S., Ochoa, W., Kobayashi, S., Reddy, V.S. Presence of a surface-exposed
loop facilitates trypsinization of particles of Sinsiro virus, a genogroup II.3 norovirus.
J. Virol. 81:1119, 2007.
Nguyen, H.D., Reddy, V.S., Brooks, C.L. III. Deciphering the kinetic mechanism of
spontaneous self-assembly of icosahedral capsids. Nano Lett. 7:338, 2007.
Biology and Applications of
Icosahedral Viral Capsids
A. Schneemann, B. Groschel, D.J. Manayani, D. Marshall,
J.E. Petrillo, M.E. Siladi, P.A. Venter
oat proteins of nonenveloped, icosahedral viruses
perform multiple functions during the course of
viral infection, including capsid assembly, specific encapsidation of the viral genome, binding to a
cellular receptor, and uncoating. In some viruses, a
single type of protein is sufficient to carry out these
functions; we are interested in the determinants that
endow a polypeptide chain with such versatility. We
seek to harness this versatility for novel applications
of viruses in biotechnology and nanotechnology.
We focus on a structurally and genetically well-characterized virus family, the T = 3 nodaviruses. Nodaviruses are composed of 180 copies of a single coat
protein and 2 strands of positive-sense RNA. Currently, we are elucidating the mechanism by which the
2 genomic RNAs are packaged into a single virion. Our
long-term goal is to develop nodaviruses as RNA packaging and delivery vectors. Our data indicate that the
2 viral RNAs are recognized separately, but it is not
yet known whether packaging occurs sequentially and
whether one or more coat protein subunits are involved
in this process. Interestingly, we found that RNA genome
packaging is coupled to genome replication, suggesting potential approaches for packaging of foreign RNAs.
Specific packaging of the viral genome also requires
coat protein translated from newly synthesized viral
RNA. Thus, genome replication, RNA packaging, and
C
THE SCRIPPS RESEARCH INSTITUTE
viral assembly are tightly coupled processes. The coupling of these processes may be a safety mechanism
for the virus to ensure efficient and accurate formation
of progeny virions in infected cells.
We are also investigating the mechanism by which
nodaviral protein B2 suppresses RNA silencing in
infected cells. We are identifying the double-stranded
RNAs that serve as substrates for B2 binding during
nodavirus infection, and we are correlating these results
with data from confocal microscopy and immunoprecipitation studies. We have identified amino acid residues in B2 that are critical for the protein’s function
as an RNA-binding protein; preliminary data suggest
that B2 may have a second function during the viral
replication cycle that is unrelated to its function as a
suppressor of RNA silencing.
We are also collaborating with several investigators at Scripps Research, the Salk Institute, La Jolla,
California, and Harvard University, Boston, to develop
nodaviruses as platforms for delivery of anthrax antitoxins. To this end, we are using particles to display
the VWA domain of capillary morphogenesis protein 2,
the cellular receptor for anthrax toxin, in a multivalent
fashion on the surface of the virion. Two insertion sites
yielding different patterns of 180 copies of the VWA
domain were selected on the basis of computational
modeling of the high-resolution crystal structure of Flock
House virus, an insect nodavirus. The resulting chimeric
viruslike particles functioned as a potent anthrax antitoxin in cell culture and protected rats from challenge
with lethal toxin. This research is important because it
shows that protein domains containing more than 150
amino acids can be displayed on Flock House virus in
a biologically functional form, suggesting numerous
additional applications. Moreover, chimeric particles decorated with anthrax protective antigen elicited a potent
neutralizing antibody response against the antigen that
protected rats from challenge with lethal toxin 4 weeks
after a single immunization without adjuvants. This
chimeric particle platform is a dually acting reagent for
the treatment of and protection against anthrax.
PUBLICATIONS
Go, E.P., Wikoff, W.R., Shen, Z., O’Maille, G., Morita, H., Conrads, T.P., Nordstrom, A., Trauger, S.A., Uritboonthai, W., Lucas, D.A., Chan, K.C., Veenstra,
T.D., Lewicki, H., Oldstone, M.B., Schneemann, A., Siuzdak, G. Mass spectrometry reveals specific and global molecular transformations during viral infection. J.
Proteome Res. 5:2405, 2006.
Maia, L.F., Soares, M.R., Valente, A.P., Almeida, F.C., Oliveira, A.C., Gomes, A.M.,
Freitas, M.., Schneemann, A., Johnson, J.E., Silva, J.L. Structure of a membranebinding domain from a non-enveloped animal virus: insights into the mechanism of
membrane permeability and cellular entry. J. Biol. Chem. 281:29278, 2006.
MOLECULAR BIOLOGY
2007
THE SCRIPPS RESEARCH INSTITUTE
253
ur research involves the molecular characterization of retroviruses and the development of ways
to interfere with the retroviral life cycle. In particular, we use feline immunodeficiency virus (FIV) for
the study of lentivirus infections. FIV causes an AIDSlike syndrome in domestic cats and has structural and
functional similarities to HIV, the cause of AIDS in
humans. Thus, developing ways to interfere with FIV
infection may result in useful treatments for infections in
both cats and humans. Our primary interests continue
to be the molecular characterization of receptor interactions and the molecular basis for the development of
drug resistance in the critical aspartic protease encoded
as part of the enzyme cassette of all retroviruses.
tionary conservation of the mechanism of infection by
FIV and HIV, even though these 2 lentiviruses use different binding receptors. Results to date in both lentivirus
systems support the notion that this mechanism protects
certain epitopes on the surface glycoprotein from immune
surveillance until the moment of virus binding and
entry into the cell. We used a panel of neutralizing monoclonal antibodies to map the region of gp95 in which
these CD134-dependent neutralizing epitopes reside.
In more recent studies, we used synthetic peptides
containing the antibody-reactive region to map the
epitopes recognized by the neutralizing antibodies. A
cluster of these epitopes resides in the variable loop 3
of FIV gp95, consistent with the notion that this region
is central to CD134 receptor binding. We previously
mapped regions of feline CD134 receptor involved in
interaction with gp95 by using chimeric proteins consisting of feline and human CD134 (the human homolog does not bind FIV glycoprotein) and site-directed
mutagenesis. During the past year, we did similar mapping studies on gp95 to further define regions critical
to CD134 and CXCR4 binding. So far, the results are
consistent with the idea that variable loop 3 is the
contact region for binding to both receptors. Cocrystallization studies are in progress to determine the structure of the region surrounding the antibody-binding sites.
We are also using these antibodies to develop specific
agents that interfere with receptor binding and may be
useful as therapeutic agents.
RECEPTOR STUDIES
P R O T E A S E D R U G R E S I S TA N C E
Like acute strains of HIV, FIV uses the chemokine
receptor CXCR4 to enter the target cell, the CD4 +
T lymphocyte. However, both HIV and FIV have other
primary binding receptors that bind the virus as a prelude to interaction with the entry receptor. We think
that these other receptors increase the effective local
concentration of the incoming virus and alter the conformation of the surface glycoprotein to increase the
binding affinity of CXCR4. Whereas HIV uses the cellsurface protein CD4 as a primary binding receptor, FIV
uses the activation antigen CD134. CD134 is expressed
on activated CD4+ T cells, a finding that explains why
FIV can infect and kill CD4+ T cells, even though the
virus does not bind CD4.
We have shown that interaction of the FIV surface
glycoprotein gp95 with CD134 causes a conformational
change in gp95. This conformational change increases
the affinity of gp95 for CXCR4; similar changes occur
when HIV gp120 binds CD4. Thus, there is an evolu-
The aspartic protease of lentiviruses is responsible
for processing the viral Gag and Pol polyproteins that
must occur at the proper time and in the proper
sequence in order to generate infectious virus. Drugs
against the HIV protease are key components of highly
active antiretroviral therapy, a treatment regimen used
successfully to treat, but not cure, patients infected with
HIV. Both FIV and HIV encode an aspartic protease and
although the FIV and HIV proteases are structurally similar, the 2 enzymes have unique sequence-cleavage
properties. We have used the parallels and differences
between FIV and HIV proteases to better understand the
molecular determinants that govern substrate/inhibitor
selectivity. We hope that our results will define the
limits of plasticity of the 2 enzymes and lead to insights
into the development of drug resistance.
As reported previously, we showed that the number
of amino acid residues involved in the sensitivity of
the proteases to drugs is limited, and we can markedly
Thiéry, R., Cozien, J., Cabon, J., Lamour, F., Baud, M., Schneemann, A. Induction of a protective immune response against viral nervous necrosis in the European sea bass Dicentrarchus labrax by using betanodavirus virus-like particles. J.
Virol. 80:10201, 2006.
Venter, P.A., Schneemann, A. Assembly of two independent populations of Flock
House virus particles with distinct RNA packaging characteristics in the same cell.
J. Virol. 81:613, 2007.
Venter, P.A., Schneemann, A. Nodaviridae. In: Encyclopedia of Virology, 3rd ed.
Mahy, B.W.J., van Regenmortel, M. (Eds.). Academic Press, San Diego, in press.
Molecular Biology of Retroviruses
J.H. Elder, Y.C. Lin, M. Sundstrom, M. Giffin,* C.D. Stout,
B.E. Torbett,* J. Gatchalian, I.A. Kim
* Department of Molecular and Experimental Medicine, Scripps Research
O
254 MOLECULAR BIOLOGY
2007
change the sensitivity of the FIV protease to be more
like that of the HIV protease by changing as few as 4
amino acids around the active site. However, changing
the substrate-cleavage specificity requires substantially
more changes. These findings explain how the virus,
when an infection is treated with a drug, can mutate
to avoid the drug but retain sufficient substrate-cleavage
specificity to allow proper Gag/Pol processing and generation of infectious virus. Critical to this process is
maintaining the proper order of site cleavage in Gag/Pol,
and changes in this order result in generation of noninfectious virus. We think that altering the order of cleavage, in addition to blocking protease activity, may be
useful as a novel intervention strategy.
PUBLICATIONS
Heaslet, H., Lin, Y.C., Tam, K., Torbett, B.E., Elder, J.H., Stout, C.D. Crystal
structure of an FIV/HIV chimeric protease complexed with the broad-based inhibitor, TL-3. Retrovirology 4:1, 2007.
Lin, Y.C., Brik, A., de Parseval, A., Tam, K., Torbett, B.E., Wong, C.H., Elder,
J.H. Altered Gag polyprotein cleavage specificity of feline immunodeficiency
virus/human immunodeficiency virus mutant proteases as demonstrated in a cellbased expression system. J. Virol. 80:7832, 2006.
Manuell, A.L., Beligni, M.V., Elder, J.H., Siefker, D.T., Tran, M., Weber, A.,
McDonald, T.L., Mayfield, S.P. Robust expression of a bioactive mammalian protein in Chlamydomonas chloroplast. Plant Biotechnol. J., 5:402, 2007.
Whiting, M., Tripp, J.C., Lin, Y.C., Lindstrom, W., Olson, A.J., Elder, J.H., Sharpless, K.B., Fokin, V.V. Rapid discovery and structure-activity profiling of novel
inhibitors of human immunodeficiency virus type 1 protease enabled by the copper(I)-catalyzed synthesis of 1,2,3-triazoles and their further functionalization. J.
Med. Chem. 49:7697, 2006.
THE SCRIPPS RESEARCH INSTITUTE
tethered to a reporter or sensitizer, are designed to
bind specifically to the active-site channel of a given
enzyme. Wires specific for the substrate- or cofactorbinding sites of a given enzyme will be useful tools for
inhibitor discovery, phototriggered enzyme turnover, and
molecular evolution strategies.
We are developing a series of tethered substrates
with variable linkers and affinity tags as reporters of
binding interactions within the substrate access channel of P450cam. In the past, we showed that these
P450 wires induce a range of conformational changes
in the helices adjacent to the enzyme’s active site. On
the basis of these findings, we have designed a new
generation of P450 wires, and we have recently solved
the crystal structure of P450cam bound to a wire containing both substrate and a biotin affinity tag. In addition, we have successfully showed phage display of
P450cam as a C-terminal fusion to the g3 phage protein. These results will enable an unprecedented structural view of phage display evolution experiments with
P450.
We have also made significant progress in designing photoactive probes specific for the pterin site of
murine inducible NOS. Addressing the pterin site through
a molecular wire may allow photochemical triggering of
enzyme turnover and help define the role of the cofactor in catalysis. We have synthesized the series of
Ru(II)-pterin wires (Fig. 1) and have shown that at least
Metalloenzyme Engineering
D.B. Goodin, C.D. Stout, E.C. Glazer, R.F. Wilson,
A. Annalora, S. Vetter, A.-M. Hays
he purposes of our research are to understand
the diversity of metalloenzyme catalysts and to
develop methods for directed evolution of enzymes
with novel function. Our goal is to gain sufficient control over substrate-enzyme interactions and the subsequent oxidative chemistry catalyzed by hypervalent
heme cofactors to allow directed evolution of catalysts
capable of regiospecific and stereospecific oxidation
of a given target substrate. We use a number of techniques in structural biology and spectroscopy and strategies of rational protein redesign and molecular evolution.
In the past year, we focused our efforts on developing synthetic molecular wires to probe the active
site and function of P450cam and nitric oxide synthase
(NOS). These wires, which consist of substrate analogs
T
F i g . 1 . Pterin-Ru(II) wires synthesized as probes of the pterin
site of NOS.
one of the wires binds to murine inducible NOS. Binding was confirmed by heme-induced quenching of either
pterin or Ru(II) fluorescence upon interaction with the
heme domain of inducible NOS, and this quenching was
reversed by binding of the natural pterin cofactor. Timeresolved emission experiments showed bound and free
forms of the wire and allowed estimation of the affinity
MOLECULAR BIOLOGY
2007
and the distance between the Ru(II) and heme. Our findings are consistent with the modeled geometry of the
wire within the pterin-binding site (Fig. 2). Ongoing
F i g . 2 . Model of a pterin-Ru(II) wire docked into the structure of
inducible NOSheme.
studies of the structure and phototriggered redox behavior of these wires bound to inducible NOS will provide
a new approach to probing the mechanism of this important enzyme.
PUBLICATIONS
Contakes, S.M., Nguyen, Y.H.L., Gray, H.B., Glazer, E.C., Hays, A.-M.A., Goodin,
D.B. Conjugates of heme-thiolate enzymes with photoactive metal-diimine wires. In:
Structure and Bonding. Yam, V.W.W. (Ed.). Springer, New York, in press.
Control of Cell Division
THE SCRIPPS RESEARCH INSTITUTE
255
this protein kinase and its analogs are ubiquitous in
eukaryotic cells and central to a number of aspects of
control of cell-cycle progression.
One area of interest is regulation of cellular morphogenesis by Cdk1. The activity of Cdk1 driven by mitotic
cyclins modulates polarized growth in yeast cells. Specifically, these activities depolarize growth by altering
the actin cytoskeleton. We found that several proteins
that modulate actin structure are targeted by Cdk1, and
we are investigating how these phosphorylation events
control actin depolarization and cell shape.
While investigating mitosis in yeast, we found that
Cks1, a small Cdk1-associated protein, appears to regulate the proteasome. Proteasomes are complex proteases
that target ubiquitylated proteins, including important
cell-cycle regulatory proteins. Surprisingly, we found
that Cks1 regulates a nonproteolytic function of proteasomes, the transcriptional activation of CDC20. Specifically, Cks1 is required to recruit proteasomes to the
gene CDC20 for efficient transcriptional elongation. Our
investigations of CDC20 have led to the conclusion that
Cks1 is required for recruitment of proteasomes to and
transcriptional elongation of many other genes as well.
Currently, we are elucidating the mechanism whereby
Cks1 recruits proteasomes and facilitates transcriptional
elongation. Our most recent results suggest that Cks1
and proteasomes in conjunction with Cdk1 mediate
remodeling of chromatin by removing nucleosomes.
CONTROL IN MAMMALIAN CELLS
S.I. Reed, C. Baskerville, L.-C. Chuang, M. Henze, J. Keck,
V. Liberal, K. Luo, B. Olson, S. Ekholm-Reed, S. Rudyak,
F. van Drogen, J. Wohlschlegel
iological processes of great complexity can be
approached by beginning with a systematic
genetic analysis in which the relevant components are first identified and the consequences of selectively eliminating the components via mutation are
investigated. We have used yeast, which is uniquely
tractable to this type of analysis, to investigate control
of cell division. It has been apparent for some time
that the most central cellular processes throughout the
eukaryotic phylogeny are highly conserved in terms of
both the regulatory mechanisms used and the proteins
involved. Thus, it has been possible in many instances
to generalize from yeast cells to human cells.
B
CONTROL IN YEAST
We have focused on the role and regulation of the
Cdc28 protein kinase (Cdk1). Initially identified by
means of a mutational analysis of the yeast cell cycle,
We showed previously that the human homologs
of the Cdc28 protein kinase are so highly conserved,
structurally and functionally, relative to the yeast protein kinase, that they can function and be regulated
properly in a yeast cell. Analyzing control of the cell
cycle in mammalian cells, we produced evidence for
the existence of regulatory schemes, similar to those
elucidated in yeast, that use networks of both positive
and negative regulators.
A principal research focus is the positive regulator
of Cdk2, cyclin E. Cyclin E is often overexpressed and/or
deregulated in human cancers. Using a tissue culture
model, we showed that deregulation of cyclin E confers genomic instability, probably explaining the link to
carcinogenesis. The observation that deregulation of
cyclin E confers genomic instability has led us to hypothesize a mechanism of cyclin E–mediated carcinogenesis based on accelerated loss of heterozygosity at tumor
suppressor loci. We are testing this hypothesis in transgenic mouse models. We showed that a cyclin E trans-
256 MOLECULAR BIOLOGY
2007
gene expressed in mammary epithelium significantly
increases loss of heterozygosity at the p53 locus, leading to enhanced mammary carcinogenesis. We are
extending these investigations by using mouse prostate,
testis, and skin models.
In an attempt to understand cyclin E–mediated
genomic instability, we are investigating how deregulation of cyclin E affects both S phase and mitosis. Recent
data suggest that deregulation of cyclin E impairs DNA
replication by interfering with assembly of the prereplication complex. Cyclin E deregulation also impairs the
transition from metaphase to anaphase by promoting
the accumulation of inhibitors of anaphase.
Our interest in cyclin E deregulation in cancer led
us to investigate the pathway for turnover of cyclin E.
We showed that phosphorylation-dependent proteolysis of cyclin E depends on a protein-ubiquitin ligase
known as SCFhCdc4. The F-box protein hCdc4 is the
specificity factor that targets phosphorylated cyclin E.
We are investigating how ubiquitylation of cyclin E is
coordinated with other processes required for its degradation, including prolyl isomerization. We are also
investigating SCFhCdc4 ubiquitylation of other important cellular proteins.
Recently, we began determining the role of SCFhCdc4
in neurodegenerative disease. We found that parkin, a
protein often mutated in hereditary Parkinson’s disease,
regulates the stability of hCdc4, possibly leading to neuropathologic changes. We discovered an SCFhCdc4 substrate, peroxisome proliferator–activated receptor γ
coactivator-1, that may be the effector of hCdc4 deregulation in Parkinson’s disease. In addition, we showed
that SCFhCdc4 regulates the turnover of presenilins in the
brain, proteins strongly implicated in Alzheimer’s disease.
Another area of interest is the role of Cks proteins
in mammals, complementing our research in yeast.
Mammals express 2 paralogs of yeast Cks1, known as
Cks1 and Cks2. Experiments in mice lacking the gene
for Cks1 and Cks2 revealed that each paralog has a
specialized function. Cks1 is required as a cofactor for
Skp2-mediated ubiquitylation and turnover of inhibitors
p21, p27, and p130. Cks2 is required for the transition from metaphase to anaphase in both male and
female meiosis I. Nevertheless, mice nullizygous at
the individual loci are viable. However, doubly nullizygous mice have not been observed because embryos
die at the morula stage, a finding consistent with an
essential redundant function. We found that this function probably is involved in regulation of transcription
and linked to chromatin remodeling, as in yeast.
THE SCRIPPS RESEARCH INSTITUTE
PUBLICATIONS
Keck, J.M., Summers, M.K., Tedesco, D., Ekholm-Reed, S., Chuang, L.-C., Jackson, P.K., Reed S.I. Cyclin E overexpression impairs progression through mitosis by
inhibiting APCCdh1. J. Cell Biol., in press.
Keller, U.B., Old, J.B., Dorsey, F.C., Nilsson, J.A., Nilsson, L., Maclean, K.H.,
Chung, L., Yang, C., Spruck, C., Boyd, K., Reed S.I., Cleveland, J.L. Myc targets
Cks1 to provoke the suppression of p27Kip1, proliferation and lymphomagenesis.
EMBO J. 26:2562, 2007.
Smith, A.P.L., Henze, M., Lee, J.A., Osborn, K.G., Keck, J., Tedesco, D., Bortner,
D.M., Rosenberg, M.P., Reed, S.I. Deregulated cyclin E promotes p53 loss of heterozygosity and tumorigenesis in the mouse mammary gland. Oncogene 25:7245, 2006.
Control of Gene Expression
During the Cell Cycle and in
Response to Environmental
Stimuli
C. Wittenberg, R.A.M. de Bruin, M. Guaderrama,
T.I. Kalashnikova
he dynamism and plasticity of biological systems
depend on the capacity to rapidly alter the abundance and activities of cellular constituents. That
capacity depends, in large part, on the ability to rapidly
modulate gene expression. Recently, we have focused
on the mechanisms by which cells exert control over
gene expression to regulate cell proliferation and to
respond to changes in environmental conditions.
T
R E G U L AT I O N O F C E L L P R O L I F E R AT I O N
In most cells, commitment to a new round of cell
division during the G1 phase of the cell cycle is accompanied by the activation of a large family of genes that
encode activities involved in the duplication and segregation of cellular components. G1-specific genes also
encode regulatory factors that promote subsequent
cell-cycle events. In the budding yeast Saccharomyces
cerevisiae, G1-specific genes are regulated by either
the SBF or the MBF transcription factor. In collaboration with J.R. Yates, Department of Cell Biology, we
used mass spectrometry–based multidimensional protein identification technology to identify novel regulators of these transcription factors.
SBF acts as a transcriptional activator and promotes
expression of its targets specifically during the G1 interval. We established that promoter-bound SBF associates with the Whi5 repressor during early G 1 phase
and that Whi5 is inactivated via phosphorylation by a
G 1 -specific cyclin-dependent protein kinase, thereby
activating transcription (Fig. 1). This regulation is analo-
MOLECULAR BIOLOGY
2007
F i g . 1 . Model for G 1-specific transcriptional regulation in yeast
and humans. The recent identification of novel regulators of G1-specific transcription in yeast confirms the striking functional homology
between the apparently unrelated transcriptional regulatory networks
in yeast and humans. In budding yeast, the SBF transcription factor is repressed during early G 1 phase by the transcriptional inhibitor Whi5. Phosphorylation of Whi5 by a cyclin-dependent kinase
promotes its release from SBF, leading to activation of G1-specific
transcription. In mammalian cells, the tumor suppressor Rb represses
E2F-dependent transcription in early G1 phase. Rb, like Whi5, is
inactivated by phosphorylation by a cyclin-dependent kinase, initiating the G1-to-S transition. A second class of G1-specific genes,
regulated by MBF, is inactivated upon exit from the G1 phase. The
corepressor Nrm1, which is conserved between the distantly related
budding and fission yeasts, accumulates as cells progress into late
G1 phase and binds to MBF at target promoters, thereby repressing
transcription. Similarly, in mammalian cells, the expression of some
G1-specific genes is inactivated by E2F4 in collaboration with p107,
an E2F1 target. Phosphorylation of Nrm1 via the DNA replication
checkpoint in response to replication stress results in dissociation
of Nrm1 from MBF-regulated promoters and derepression of MBF
targets. It remains to be established whether the DNA structure
checkpoints regulate G1-specific transcription via corepressors of
E2F in mammalian cells.
gous to the regulation of E2F by the tumor suppressor
Rb in humans.
In contrast, MBF acts primarily as a transcriptional
repressor and limits transcription of target genes to the G1
phase. We identified Nrm1, a novel MBF-associated corepressor. When expressed as an MBF target during late
G1 phase, Nrm1 associates with MBF at target promoters
and represses expression as cells enter S phase (Fig. 1).
Similarly, the Nrm1 homolog, SpNrm1, in the fission
yeast Schizosaccharomyces pombe regulates MBF, the
sole G 1-specific transcription factor in this yeast.
In addition to being regulated during the cell cycle,
the G1-specific transcriptional machinery is regulated
by checkpoints that monitor the integrity of cellular
structures and processes. When replication forks are
stalled during S phase in S pombe, repression of MBFregulated transcription is disrupted and expression of
MBF target genes is induced. We have shown that
THE SCRIPPS RESEARCH INSTITUTE
257
SpNrm1 is a direct target for phosphorylation by the
checkpoint kinase Cds1, a homolog of human Chk2,
and that its phosphorylation leads to dissociation from
MBF and activation of gene expression (Fig. 1). That
response requires the ATM/ATR-like checkpoint kinase
Rad3. We recently found that derepression of MBF target
genes in S cerevisiae also occurs via regulation of Nrm1
in response to activation of the DNA replication checkpoint. A similar regulatory cascade may control G1-specific genes via E2F family members in animal cells
and thereby play a crucial role in the maintenance of
genome integrity.
A D A P TAT I O N T O E N V I R O N M E N TA L S T I M U L I
Remodeling of the gene expression program also
occurs in cells adapting to environmental changes. We
have studied the regulation of the HXT genes, which
encode hexose permeases. Those genes are induced
by growth on glucose and repressed on most other carbon sources. Extracellular glucose interacts with cellsurface receptors that initiate a signaling cascade that
culminates with the activation of HXT gene transcription. We have shown that signaling leads to the phosphorylation-dependent destruction of the transcriptional
corepressor Mth1 by the E3 ubiquitin ligase SCF Grr1 .
Destruction of Mth1 leads to phosphoryation of the
transcriptional repressor Rgt1 and its dissociation from
HXT gene promoters. Conversely, repression of HXT
gene expression requires Mth1 and is associated with
Rgt1 dephosphorylation. In collaboration with T. Hunter,
Salk Institute for Biological Studies, La Jolla, California,
we recently showed that a type 2A protein phosphatase
complex, Pph3/Psy2, which associates with Mth1, promotes Rgt1 dephosphoryation. Consequently, Pph3/Psy2
is important for the reestablisment of repression of
HXT genes when environmental glucose is limiting.
Interestingly, SCFGrr1, the E3 ubiquitin ligase required
for destruction of Mth1 during HXT gene induction, is
also important for destruction of G1 cyclins, critical regulators of cell-cycle initiation. We are investigating the
basis for discrimination between targets by SCFGrr1. We
found that basic residues in the leucine-rich repeat and
parts of the C terminus of the F-box protein Grr1 are
important for recognition of phosphorylated substrates.
Identification of additional Grr1 substrates, in collaboration with Dr. Yates, and additional characterization
of Grr1 will facilitate those studies.
PUBLICATIONS
Limbo, O., Chahwan, C., Yamada, Y., de Bruin, R.A.M., Wittenberg, C., Russell, P.
Ctp1 is a cell cycle-regulated protein that functions with Mre11 complex to control
double-strand break repair by homologous recombination. Mol. Cell, in press.
258 MOLECULAR BIOLOGY
2007
Cell-Cycle Checkpoints, DNA
Damage, and Cytotoxic Stress
Responses
P. Russell, A. Amar, C. Dovey, L.-L. Du, P. Kennedy,
O. Limbo, V. Martin, S. Rozenzhak, J. Williams, Y. Yamada
NA damage and cytotoxic stress elicit cellular
responses that are highly conserved throughout
eukaryotic evolution. Consequently, studies of
genetically tractable microorganisms such as the fission yeast Schizosaccharomyces pombe can provide a
useful framework for the design and interpretation of
experiments performed with more complex multicellular organisms. We use fission yeast to study cell-cycle
checkpoints, DNA repair, and stress response mechanisms. Defects in these mechanisms underlie a number of human diseases, including cancer.
D
CHECKPOINTS AND DNA DAMAGE RESPONSES
DNA double-strand breaks are among the most lethal
and genome-destabilizing type of DNA damage. They
can arise from exogenous sources, such as ionizing radiation, or through errors involving DNA replication. Rapid
and accurate repair of double-strand breaks is essential
for preserving genome integrity. The mechanism used
to repair the breaks depends on the circumstances in
which the DNA damage occurs.
Double-strand breaks that arise in the postreplicative G2 phase of a haploid organism are most effectively
repaired by homologous recombination, an error-free
mechanism that uses the undamaged sister chromatid
as a template for repair of the broken chromosome. In
contrast, when double-strand breaks occur in the prereplicative G1 phase, the sister chromatid is unavailable
as a template for DNA repair. In this circumstance,
the only option for repair of the breaks in unique DNA
sequences is nonhomologous end joining. This errorprone mechanism, which was originally discovered in
mammalian cells, is broadly conserved among eukaryotes, including the budding yeast Saccharomyces
cerevisiae and the fission yeast S pombe.
Recently, we have been identifying novel DNA repair
proteins in fission yeast. An example is the protein Xlf1,
which we found through its sequence similarity to the
human protein Cernunnos. Patients with defects in
Cernunnos have a set of phenotypes, including microcephaly and immunodeficiency, that closely resemble the
characteristics of patients with defects in DNA ligase
THE SCRIPPS RESEARCH INSTITUTE
IV, which is required for nonhomologous end joining.
We found that Xlf1 is required for nonhomologous end
joining in fission yeast. Notably, cells lacking Xlf1 cannot survive exposure to ionizing radiation during G1
phase. Xlf1 is 1 of 4 proteins required for nonhomologous end joining in fission yeast.
Preservation of genome integrity in eukaryotic organisms also depends on telomeres, specialized chromatin
structures that compose the ends of linear chromosomes. Telomeres cap and protect chromosome ends,
distinguishing the ends from double-strand breaks elsewhere in the genome. Failure to properly protect chromosome ends leads to chromosome end-to-end fusions
and other genome rearrangements, leading to forms of
genomic instability typically associated with cancer. In
the past year, we discovered 2 novel telomere proteins
in fission yeast. Our interest in these proteins began
when we discovered that they have predicted OB-fold
domains, which are a compact structural motif found
in a variety of proteins that interact with single-stranded
DNA. We found that the proteins form a heterodimeric
complex that binds to DNA ends. Bioinformatic studies revealed that the proteins are distant homologs of
the proteins Stn1 and Ten1 that occur in budding yeast;
Stn1 and Ten1 were heretofore thought to be restricted
to closely related species of budding yeast.
CYTOTOXIC STRESS RESPONSE
Production of reactive oxygen species (ROS), such
as hydroxyl radicals, superoxide anions, and hydrogen
peroxide, is a normal byproduct of aerobic metabolism
in all eukaryotic organisms. Elevation of intracellular
ROS can also arise through exposure to environmental
toxicants, such as heavy metals or metalloids (e.g., cadmium and arsenic) and some pesticides. Oxidative stress
in the form of ROS can be highly toxic, causing damage to proteins, lipids, and nucleic acids. Indeed, the
cumulative effects of exposure to ROS are thought to
be a causative factor in many of the most widespread
and debilitating human diseases, such as atherosclerosis, Alzheimer’s disease, Parkinson’s disease, and
cancer, and in the aging process itself. Consequently,
all eukaryotic organisms have multiple cellular mechanisms to prevent the excessive accumulation of ROS
and protect against their harmful effects. Antioxidant
defense mechanisms include use of both nonenzymatic
molecules such as glutathione and several vitamins
and ROS scavenger enzymes such as superoxide dismutase, catalase, and glutathione peroxidase.
In the past year, we made the surprising discovery
that the cellular response to oxidative stress substan-
MOLECULAR BIOLOGY
2007
tially depends on Upf1, a component of the nonsensemediated messenger RNA decay system. Whole genome
expression profiling studies showed that Upf1 controls
the expression of more than 100 genes that are transcriptionally induced in response to oxidative stress;
most of these are also controlled by the transcription
factor Atf1. The unexpected connection between a factor in nonsense-mediated messenger RNA decay and
the oxidative stress response in fission yeast may provide important new clues about the physiologic function of nonsense-mediated messenger RNA decay in
other species.
PUBLICATIONS
Cavero, S., Chahwan, C., Russell, P. Xlf1 is required for DNA repair by nonhomologous end joining in Schizosaccharomyces pombe. Genetics 175:963, 2006.
Martin, V., Du, L.-L., Rozenzhak, S., Russell, P. Protection of telomeres by a conserved Stn1-Ten1 complex. Proc. Natl. Acad. Sci. U. S. A., in press.
DNA Damage Responses in
Human Cells
C.H. McGowan, M. Duquette, E. Langley, J. Scorah,
D. Slavin, E. Taylor
omplex multiceullar organisms, such as humans,
contain large numbers of mitotically competent
cells that are capable of renewal, repair, and,
to some extent, regeneration. The advantages of being
able to replace damaged or aged cells are offset by
the inherent susceptibility of mitotic cells to mutate
and become cancerous. DNA is inherently vulnerable
to many sorts of chemical and physical modifications;
thus, as cells duplicate and divide, they can acquire
mutations. Both spontaneous and induced DNA damage must be repaired with minimal changes if growth,
renewal, and repair are to be successful. Our overall
objective is to understand how mammalian cells protect
themselves from DNA damage and thus from cancer.
We are especially interested in understanding
basic cellular responses to clinically relevant agents.
Eukaryotic cells have evolved with a complex network
of DNA repair processes and cell-cycle checkpoint
responses to ensure that damaged DNA is repaired
before it is replicated and becomes fixed in the genome.
These pathways are highly conserved throughout evolution, and much information about human responses
to DNA damage has been gained from studies of simple, genetically tractable organisms such as yeast. We
C
THE SCRIPPS RESEARCH INSTITUTE
259
use a combination of molecular, cellular, and genetic
techniques to determine how these pathways operate
in human cells.
Checkpoints control the order and timing of events
in the cell cycle; they ensure that independent processes
are appropriately coupled. In addition, checkpoints
promote the use of the most appropriate repair pathway. We used genetic models to identify 2 checkpoint
kinases in humans that limit progression of the cell
cycle when DNA is damaged. We are studying the function of both of these kinases and of a number of their
substrates. One of the kinases, Chk2, is activated in
response to DNA damage. Chk2 physically interacts
with Mus81-Eme1, a conserved DNA repair protein
that has homology to the xeroderma pigmentosum F
family of endonucleases. Xeroderma pigmentosum is
a cancer-prone disorder that results from a failure to
appropriately repair damaged DNA.
Biochemical analysis indicates that Mus81-Eme1
has endonuclease activity against structure-specific DNA
substrates, including Holliday junctions. Enzymatic
analysis, immunofluorescence studies, and the use of
RNA interference have all contributed to the conclusion
that Mus81-Eme1 is required for recombination repair
in human cells. We are also using gene targeting to
study the function of the Mus81-Eme1 endonuclease
in mice. Inactivation of Mus81 in mice increases genomic instability and sensitivity to DNA damage but does
not promote tumorigenesis. In addition, we have shown
that Mus81-Eme1 is specifically required for survival
after exposure to cisplatin, mitomycin C, and other commonly used anticancer drugs. As a point of interaction
between checkpoint control and DNA repair, the relationship between Mus8-Eme1 and Chk2 probably provides
information critical to understanding the response to
DNA damage as a whole.
Anticancer therapy is largely based on the use of
genotoxic agents that damage DNA and thus kill dividing cells. Coordination of cell-cycle checkpoints and
DNA repair is especially important when unusually high
amounts of DNA damage occur after radiation or genotoxic chemotherapy. Hence, a detailed understanding
of cellular responses to DNA damage is essential in
understanding both the development and the treatment
of disease in humans.
PUBLICATIONS
Gueven, N., Becherel, O.J., Howe, O., Chen, P., Haince, J.F., Ouellet, M.E.,
Poirier, G.G., Waterhouse, N., Fusser, M., Epe, B., de Murcia, J.M., de Murcia,
G., McGowan, C.H., Parton, R., Mothersill, C., Grattan-Smith, P., Lavin, M. A
novel form of ataxia oculomotor apraxia characterized by oxidative stress and apoptosis resistance. Cell Death Differ. 14:1149, 2007.
260 MOLECULAR BIOLOGY
2007
DNA Repair and the
Maintenance of Genomic
Stability
M.N. Boddy, S. Pebernard, J. Prudden
NA repair pathways have evolved to protect the
genome from ever-present genotoxic agents.
Highlighting the importance of the pathways,
defects in DNA repair mechanisms strongly predispose
the host to cancer and to neurologic and developmental
disorders. The DNA repair systems we study in fission
yeast are evolutionarily conserved, and therefore our
studies provide a valuable framework for understanding
genome maintenance in human cells. Although many
DNA repair mechanisms have been described, information on how they are coordinated with necessary
changes in chromatin structure is limited.
D
THE SMC5-SMC6 COMPLEX
In collaboration with J.R. Yates, Department of Cell
Biology, we purified the structural maintenance of chromosomes (SMC) complex Smc5-Smc6 to identify its
core components. The holocomplex consists of the
Smc5-Smc6 heterodimer and 6 additional non-SMC
elements, Nse1–Nse6 (Fig. 1A). We showed that Smc5-
F i g . 1 . Smc5-Smc6 holocomplex and STUbLs. A, Nse1, Nse3,
and Nse4 form a stable heterotrimer that associates with Smc5.
Nse2 interacts directly with Smc5 in the absence of the other Nse
proteins. Smc6 interacts directly with Smc5 but with none of the
other components. Nse5 and Nse6 form a stable heterodimer that
also binds directly to Smc5. Double-headed arrows indicate interactions between subcomplexes. Nse5 and Nse6 may recruit the
holocomplex to stalled replication forks and certain DNA damage
sites. B, A target protein is sumoylated, an event that leads to recruitment of STUbLs via interactions with SUMO and the substrate. For
Rad60 or NIP45, the STUbLs interact via the SUMO-like domains.
STUbLs then ubiquitinate the substrate, promoting regulation via
altered localization, protein interaction, or degradation.
THE SCRIPPS RESEARCH INSTITUTE
Smc6 prevents the deleterious engagement of an ordinarily beneficial DNA repair pathway called homologous
recombination. Smc5-Smc6 either prevents initiation
of homologous recombination or separates physically
linked chromosomes that arise late in this process.
Spontaneous DNA damage in Smc5-Smc6 mutant cells
is due to the attempted separation of chromosomes
into daughter cells while the chromosomes are still
physically linked. Such defective chromosome separation
in humans could result in cancer and other diseases.
A N U N P R E C E D E N T E D S U M O - TA R G E T E D U B I Q U I T I N
LIGASE
The covalent attachment of ubiquitin and the small
ubiquitin-like protein SUMO to target proteins plays key
roles in genome stability; each of the 2 moieties (i.e.,
ubiquitin and SUMO) has physiologically distinct effects
on the function of target proteins. We have identified
the SUMO-targeted ubiquitin ligase (STUbL) family,
which provides a novel and unanticipated regulatory link
between the ubiquitination and sumoylation pathways.
Members of the STUbL family include Slx8-Rfp1 in
fission yeast, RNF4 in humans, MIP1 in slime molds,
and SLX5/8 in budding yeast. STUbLs are recruited
to sumoylated proteins and proteins containing SUMOlike domains to mediate the ubiquitination and regulation of these proteins (Fig. 1B).
Cells with mutations in Slx8-Rfp1 accumulate sumoylated proteins, display genomic instability, and are hypersensitive to genotoxic stress. These Slx8-Rfp1 mutant
phenotypes are suppressed by concomitant deletion of
the major SUMO ligase Pli1, demonstrating the specificity of STUbLs as regulators of sumoylated proteins.
Expression of human RNF4 restores homeostasis of the
SUMO pathway in fission yeast that lack Slx8-Rfp1,
underscoring the evolutionary functional conservation
of STUbLs. The DNA repair factor Rad60 (an accessory
factor of the Smc5-Smc6 complex) and its human homolog NIP45, which contain SUMO-like domains, are
candidate STUbL targets. Consistently, mutations in
Rad60 and Slx8-Rfp1 cause similar DNA repair defects.
MOLECULAR BIOLOGY
2007
THE SCRIPPS RESEARCH INSTITUTE
261
Signal Transduction Pathways
Mediating Cellular Responses
to Oncogenic Mutations
P. Sun, C. Kannemeier, J. Kwong, R. Liao, A. Seit-Nebi,
N. Yoshizuka
evelopment of cancer is a result of multiple oncogenic genetic alterations, including activation
of oncogenes and inactivation of tumor suppressors. Although these oncogenic mutations contribute to tumorigenic phenotypes, normal cells can
respond to oncogenic changes by initiating tumor-suppressing defense mechanisms such as apoptosis and
premature senescence (a stable form of growth arrest).
As a result, tumor development requires additional mutations that compromise these antioncogenic responses.
Our main interests are to delineate the signal transduction pathways that mediate these tumor-suppressing
responses and to determine how these responses are
evaded during cancer development. Currently, we are
focusing on 2 well-known oncogenes: ras and mdm2.
The oncogene ras encodes a family of small GTPbinding proteins that are often activated in human
tumors and contribute to tumor development. In normal cells, however, the initial response to ras activation is premature senescence. Recent studies have
shown that like apoptosis, oncogene-induced senescence is a bona fide tumor-suppressing mechanism in
vivo that must be compromised in order for cancer to
develop. However, the signaling pathways responsible
for this important antitumorigenic response are poorly
understood. We have shown that ras induces senescence through sequential activation of 2 MAP kinase
pathways (Fig. 1). Initially, ras activates the MAP kinase
kinase (MEK)–extracellular signal–regulated kinase (ERK)
pathway. Sustained activation of MEK-ERK turns on the
stress-induced p38 pathway, which subsequently causes
senescence. These studies have revealed a novel, tumorsuppressing function of p38, in addition to its known
roles in inflammation and stress responses.
In other studies, we identified additional signaling
components that mediate ras-induced senescence
(Fig. 1), including p38-regulated/activated protein kinase
(PRAK). PRAK is a p38 MAP kinase substrate whose
physiologic functions are poorly understood. We found
that PRAK mediates senescence upon activation by p38
in response to oncogenic ras. In mice, PRAK deficiency
D
F i g . 1 . Signal transduction pathways mediating oncogene-induced
premature senescence. Oncogenic ras induces sequential activation of
the tumorigenic Raf-MEK-ERK MAP kinase pathway and the stressinduced MKK3/6-p38-PRAK MAP kinase pathway. Activation of the
p38 pathway may be mediated by increased intracellular levels of
reactive oxygen species (ROS) induced by the ras-Raf-MEK-ERK
signaling cascade. Activated components of the p38 pathway phosphorylate multiple residues on p53, including S33 and S46 (by p38),
S37 (by PRAK), and others, leading to increased transcriptional activity of p53 and induction of p21WAF1, a transcriptional target of p53.
Active p38 also induces the expression of p16INK4A and p14/p19ARF,
which, together with the p53-p21 cascade, cause premature senescence that serves as a tumor-suppressing defense mechanism. All the
oncogenic events are shown in red; the tumor-suppressing events are
in blue.
enhances skin carcinogenesis induced by the environmental mutagen 7,12-dimethylbenz(a)anthracene, coinciding with compromised induction of senescence. In
primary cells, inactivation of PRAK prevents senescence
and promotes oncogenic transformation. Moreover, PRAK
activates p53 by direct phosphorylation at Ser37 of p53
(Fig. 1). We propose that phosphorylation of p53 by
PRAK after activation of p38 MAP kinase by ras plays an
important role in ras-induced senescence and tumor suppression. Experiments are under way to discover additional signaling components that regulate the induction of
senescence and to determine whether the p38 pathway
is disarmed during cancer development in humans.
Another focus of our research is mdm2, an oncogene that can mediate transformation primarily through
inactivation of the tumor suppressor protein p53. Previously, we found that MDM2, the protein encoded by
mdm2, confers resistance to cell-cycle arrest induced
by transforming growth factor β (TGF-β), a growthinhibitory cytokine. In studies on the molecular mechanism that underlies MDM2-mediated resistance to
TGF-β, we found that MDM2 makes cells refractory to
the cytokine by overcoming a TGF-β–induced arrest of
262 MOLECULAR BIOLOGY
2007
the cell cycle in phase G1. Because the TGF-β–resistant phenotype is reversible upon removal of MDM2,
MDM2 probably confers resistance to TGF-β by directly
targeting the cellular machinery involved in the growth
inhibition mediated by the cytokine.
In both mink lung epithelial cells and human mammary epithelial cells, 3 elements were required for
MDM2-mediated resistance to TGF-β. One element was
the C-terminal half of the p53 binding domain, which
at least partially retained p53 binding and inhibitory
activity. Second, the ability of MDM2 to mediate TGF-β
resistance was disrupted by mutation of the nuclear
localization signal but was restored by coexpression of
MDMX, a relative of MDM2 and another p53 regulator.
Finally, mutations of the zinc coordination residues of
the RING finger domain abrogated TGF-β resistance but
not the ability of MDM2 to inhibit p53 activity or to
bind MDMX. These data suggest that RING finger–mediated p53 inhibition and MDMX interaction are not sufficient to cause TGF-β resistance and imply a crucial
role for the E3 ubiquitin ligase activity of this domain
in MDM2-mediated TGF-β resistance.
PUBLICATIONS
Kannemeier, C., Liao, R., Sun, P. The RING finger domain of MDM2 is essential
for MDM2-mediated TGF-β resistance. Mol. Biol. Cell 18:2367, 2007.
Sun, P., Yoshizuka, N., New, L., Moser, B.A., Li, Y., Liao, R., Xie, C., Chen, J.,
Deng, Q., Yamout, M., Dong, M.-Q., Frangou, C.G., Yates, J.R. III, Wright, P.E.,
Han, J. PRAK is essential for ras-induced senescence and tumor suppression. Cell
128:295, 2007.
Genetic Modifiers of Behavioral
Despair as Targets for New
Antidepressants
J.G. Sutcliffe, P.B. Hedlund, F.E. Bloom,* B.S. Hilbush*
* ModGene, L.L.C., La Jolla, California
ouse models have been developed that have
high value for predicting the antidepressant
activity of a drug. The 2 models used most
often are the forced-swim test and the tail-suspension
test. In both tests, a state of behavioral despair is created in which mice cease to struggle and become immobile when confronted with an adverse situation. Known
human antidepressants increase the length of time the
mice struggle, thus decreasing the immobility time.
Significant behavioral differences exist among inbred
mouse strains in these tests. For example, during the
M
THE SCRIPPS RESEARCH INSTITUTE
last 4 minutes of a 6-minute forced swim test, unmedicated, inbred C57BL/6J mice spent 152 seconds of
the possible 240 seconds (63%) immobile. In contrast,
inbred DBA/2J mice spent only 47 seconds (20%) immobile. Thus, genetic differences between the 2 mouse
strains influence behaviors in the test. In collaboration with researchers at ModGene, L.L.C., we used
the differences in the responses of the 2 strains to
identify genes whose activities contribute to relative
basal despair status.
We investigated the hypothesis that the differences
between the 2 inbred mouse lines that result in different
baseline immobility times in the forced-swim test are
due to cumulative quantitative differences in the activities of several modifier genes, called quantitative trait
loci (QTLs), that could be placed on the genetic map.
We further investigated the hypothesis that at least some
of the strain differences occur because of differences in
the amount of mRNA that accumulates from a single
gene in each of the mapped chromosomal regions. We
measured baseline immobility times of C57BL/6J mice,
DBA/2J mice, and 27 strains of recombinant inbred
mice that were produced from C57 x DBA matings (10
mice of each strain for each sex). Each strain had a
characteristic immobility time, ranging from 5 seconds
to 165 seconds; this finding indicated that several genes
must be contributing to the strain differences because
the range was greater than the difference between the
parental strains. We also found differences between
males and females within strains.
We correlated these data with the haplotype data
from the mouse strains and detected QTL genes associated with despair on chromosome 4 in all mice and additionally on chromosomes 11 and 13 in female mice and
chromosome 18 in male mice. These results indicated
some sexual dimorphism in determinants of this behavioral despair, as is known for depression in humans.
We then correlated the inheritance of each QTL with
the concentration of each of more than 30,000 mRNAs
in the brains of the recombinant inbred mice. We identified a 100% correlation between inheritance of each
QTL and the heritability of the amount of mRNA that
accumulated from a single chromosome 4 gene, a single chromosome 11 gene, and a single chromosome 13
gene. No chromosome 18 gene correlation was detected.
In analogous studies with the tail-suspension test, we
detected the same chromosome 4 genes in both males
and females and the same chromosome 11 and 13
genes in females only, suggesting that these 3 genes are
MOLECULAR BIOLOGY
2007
directly related to the despair behavior rather than to
the ability to perform in one of the behavioral tests.
The identities of the genes responsible for the quantitative traits provide a powerful point of departure for
the development of new pharmaceutical agents to treat
depression because the studies in which the genes
were detected provide evidence that differences in the
activities of the protein products of the genes directly
contribute to differences in phenotype. Thus, a drug
that altered the activity of a protein encoded by a specific gene in the beneficial direction (either inhibition
or augmentation), and did not have other deleterious
side effects, would be a suitable candidate to test for
antidepressant activity.
The chromosome 4 gene encodes a previously
known protein (never associated with brain disorders)
whose activity is regulated by phosphorylation by a specific protein kinase. Because the concentration of mRNA
for the protein is higher in C57BL/6J mice (the mice
that have the greater despair) than in DBA mice, we
sought a way to reduce the activity of the protein and
make the C57BL/6J mice behave more like the DBA/2J
mice. A compound that inhibits the activity of the specific kinase in cultured tumor cells has recently been
synthesized. When the compound was administered in a
very low dose to C57BL/6J mice, their immobility was
reduced to 78 seconds, indicative of an antidepressantlike effect. A 20-fold higher dose reduced immobility to
0 seconds. The compound also reduced immobility
when administered to DBA mice and had analogous
effects on tail-suspension immobility times. These data
suggest that the compound can be considered a lead
compound for testing as an antidepressant.
PUBLICATIONS
de Lecea, L., Sutcliffe, J.G. The Hypocretins (Orexins). In: Handbook of Biologically Active Peptides. Kastin, A. (Ed.). Academic Press, San Diego, 2006, p. 721.
Desplats, P.A., Denny, C.A., Kass, K.E., Gilmartin, T., Head, S.R., Sutcliffe, J.G.,
Seyfried, T.N., Thomas, E.A. Glycolipid and ganglioside metabolism imbalances In
Huntington’s disease. Neurobiol. Dis., in press.
Hedlund, P.B., Sutcliffe, J.G. The 5-HT7 receptor influences stereotypic behavior in
a model of obsessive-compulsive disorder. Neurosci. Lett. 414:247, 2007.
Semenova, S., Geyer, M.A., Sutcliffe, J.G., Markou, A., Hedlund, P.B. Inactivation
of the 5-HT7 receptor partially blocks phencyclidine-induced disruption of prepulse
inhibition. Biol. Psychiatry, in press.
Sutcliffe, J.G., de Lecea, L. The hypocretin/orexin system. In: Handbook of Contemporary Neuropharmacology. Sibley, D., et al. (Eds.). Wiley-Interscience, Hoboken, NJ, 2007, Vol. 3, p. 125.
THE SCRIPPS RESEARCH INSTITUTE
263
Molecular Neurobiology of
CNS Disorders
E.A. Thomas, J.G. Sutcliffe, P.A. Desplats, E.L. Woodward
DIVERSITY OF GENE EXPRESSION IN
SCHIZOPHRENIA
chizophrenia is a heterogeneous psychiatric disorder with variable signs and symptoms that
change during the course of the illness. Additionally, the outcome after initial diagnosis varies widely.
We have studied the molecular differences that occur
during progression of schizophrenia and those that may
be associated with its heterogeneity. We generated
genome-wide RNA expression profiles from tissue samples
obtained at autopsy from the prefrontal cortex of 32
patients who had schizophrenia for 2 to 53 years and
from the prefrontal cortex of 32 age- and sex-matched
controls (persons without schizophrenia). We found that
patients with acute and chronic illness had major differences in gene expression profiles. Acute illness was
associated with dramatic changes in gene expression,
whereas chronic illness was associated with a stabilization of aberrant gene expression. Currently, we are
focusing on genes correlated with disease progression
and those regulated by antipsychotic drug treatment.
Further, we have detected modules of coexpressed
genes corresponding to subgroups of patients from all
cohorts. We have delineated 4 biologically relevant subgroups of schizophrenia defined by different gene expression profiles; each subgroup is associated with distinct
physiologic pathways. In particular, we have focused on
alterations of genes associated with glycosphingolipid
metabolism and myelination in the pathophysiologic
changes associated with psychiatric disease. The distinct molecular portraits we identified might represent
bona fide “subtypes” of schizophrenia. Each subtype
might be associated with different signs and symptoms and have a different pattern of disease progression and outcome.
S
T R A N S C R I P T I O N A L D Y S R E G U L AT I O N I N
H U N T I N G T O N ’ S D I S E A S E : S T R I ATA L S P E C I F I C I T Y
Much evidence supports a role for transcriptional
dysregulation in Huntington’s disease. Of particular
interest is how these disturbances may be specifically
manifested in the striatum, the area of the brain with
the most striking neuropathologic deficits in this disease. Currently, we are investigating transcriptional
regulatory proteins that have highly enriched expres-
264 MOLECULAR BIOLOGY
2007
sion in the striatum and the roles of the proteins in the
pathologic changes associated with Huntington’s disease. We have focused on 2 particular transcription factors, Bcl11b and Foxp1, which we hypothesize play
important roles in striatal gene expression. Both of
these factors have decreased expression in mouse models
of Huntington’s disease and in the caudate of patients
with Huntington’s disease and can interact with the
protein huntingtin. Sequestration of these factors into
nuclear aggregates in Huntington’s disease resulting in
loss of function may contribute to specific dysregulation
of striatal gene expression. This mechanism may explain,
in part, the specificity of the pathologic changes associated with Huntington’s disease.
PUBLICATIONS
Dean, B., Keriakous, B., Scarr, E., Thomas, E.A. Gene expression profiling in
Brodmann’s area 46 from subjects with schizophrenia. Aust. N. Z. J. Psychiatry
41:308, 2007.
Desplats, P.A., Denny, C.A., Kass, K.E., Gilmartin, T., Head, S.R., Sutcliffe, J.G.,
Seyfried, T.N., Thomas, E.A. Glycolipid and ganglioside metabolism imbalances In
Huntington’s disease. Neurobiol. Dis., in press.
Narayan, S., Kass, K.E., Thomas, E.A. Chronic haloperidol treatment results in a
decrease in the expression of myelin/oligodendrocyte-related genes in the mouse
brain. J. Neurosci. Res. 85:757, 2007.
Sundram, S., Scarr, E., Digney, A., Thomas, E.A., Dean, B. Plasma apolipoprotein E
is decreased in schizophrenia spectrum and bipolar disorder. Psychiatry Res., in press.
Thomas, E.A. Molecular profiling of antipsychotic drug function: convergent mechanisms in the pathology and treatment of psychiatric disorders. Mol. Neurobiol.
34:109, 2006.
Thomas, E.A., Yao, J.K. Clozapine specifically alters the arachidonic acid pathway
in mice lacking apolipoprotein D. Schizophr. Res. 89:147, 2007.
The 5-HT7 Receptor in
Neuropsychiatric Disorders
P.B. Hedlund, P.E. Danielson, S. Semenova, M.A. Geyer,
A. Markou, J.G. Sutcliffe
nterest in the serotonin 5-HT7 receptor as a putative
target in neuropsychiatric disorders has been growing continually. The interest was prompted by the
finding that several classes of drugs used to treat disorders such as depression and schizophrenia have high
affinity for the 5-HT7 receptor. We have established evidence that supports a role for this receptor in depression, obsessive-compulsive disorder, and schizophrenia.
I
DEPRESSION
The forced swim test and the tail suspension test
are animal models of behavioral despair that have high
value for predicting the antidepressant efficacy of drugs.
THE SCRIPPS RESEARCH INSTITUTE
The tests can also be used to characterize animals in
which genes have been deleted. In both of these tests,
we showed that mice lacking the 5-HT7 receptor have
a behavioral profile similar to that of mice treated with
antidepressants. We replicated these findings by using
a compound that acts as a selective antagonist at the
5-HT7 receptor. Thus, both blockade and inactivation
of the receptor yield the same result. New evidence
suggests that the effects of antidepressants acting as
selective serotonin reuptake inhibitors are potentiated
by 5-HT 7 receptor antagonists. Taken together, our
results suggest an important role for the 5-HT7 receptor in depression, and antagonists at this receptor should
be evaluated as treatment for depression, either independently or in combination with existing antidepressants.
OBSESSIVE-COMPULSIVE DISORDER
Obsessive-compulsive disorder is commonly treated
with antidepressants, and thus is related to depression. In an animal model of obsessive-compulsive disorder (marble burying), we showed that blockade or
inactivation of the 5-HT7 receptor results in less compulsive behavior. These findings further support the
hypothesis that the 5-HT7 receptor is an important
alternative or supplemental target for antidepressants.
SCHIZOPHRENIA
Prepulse inhibition (PPI) of the acoustic startle reflex
is a well-characterized model of schizophrenia. The
model is especially relevant because similar responses
can be observed in patients with schizophrenia. We
showed that PPI per se is not altered in mice lacking
the 5-HT7 receptor, but that when PPI is disrupted by
phencyclidine, these mice are significantly less affected
than are mice that have the receptor. Phencyclidineinduced disruption involves a glutamatergic component
of PPI that is relevant for the action of atypical antipsychotic agents such as clozapine. Clozapine is a drug
with relatively high affinity for the 5-HT7 receptor.
PUBLICATIONS
de Lecea, L., Sutcliffe, J.G. The hypocretins (orexins). In: Handbook of Biologically
Active Peptides. Kastin, A. (Ed.). Elsevier, Philadelphia, 2006, p. 721.
Hedlund, P.B., Sutcliffe, J.G. The 5-HT7 receptor influences stereotypic behavior in
a model of obsessive-compulsive disorder. Neurosci. Lett. 414:247, 2007.
Landry, E.S., Lapointe, N.P., Rouillard, C., Levesque, D., Hedlund, P.B., Guertin,
P.A. Contribution of spinal 5-HT1A and 5-HT7 receptors to locomotor-like movement induced by 8-OH-DPAT in spinal cord-transected mice. Eur. J. Neurosci.
24:535, 2006.
Semenova, S., Geyer, M.A., Sutcliffe, J.G., Markou, A., Hedlund, P.B. Inactivation
of the 5-HT7 receptor partially blocks PCP-induced disruption of prepulse inhibition. Biol. Psychiatry, in press.
Sutcliffe, J.G., de Lecea, L. The hypocretin/orexin system. In: Handbook of Contemporary Neuropharmacology. Sibley, D., et al. (Eds.). Wiley-Interscience, Hoboken, NJ, 2007, Vol. 3, p. 125.
MOLECULAR BIOLOGY
2007
THE SCRIPPS RESEARCH INSTITUTE
265
Lysophospholipid Signaling and
Neural Aneuploidy
J. Chun, B. Anliker, J. Choi, A. Dubin, S. Gardell, D. Herr,
G. Kennedy, M. Kingsbury, C. Lee, D. Lin, M. Lu, T. Mutoh,
K. Noguchi, C. Paczkowski, S. Peterson, R. Rivera, W. Westra,
X. Ye, Y. Yung, L. Zhu
n the past year, several discoveries have led to new
directions in our long-term projects. In our studies
on lysophospholipid signaling, we discovered a new
biological role for sphingosine 1-phosphate (S1P) in
hearing and vestibular function. We also identified and
characterized a novel lysophosphatidic acid (LPA) receptor termed LPA5 and found new signaling properties of
another LPA receptor, LPA4. We continue to identify signaling pathways and possible new lysophospholipid
receptors; our aim is to understand key biological functions of lysophospholipid signaling.
We made several technological advancements in
our studies of neural aneuploidy, paving the way for a
deeper understanding of the functional consequences
of aneuploidy in the brain. While continuing to characterize the extent and regional variations of aneuploidy
in the brains of healthy humans and normal mice, we
are laying the groundwork to determine potential links
between neural aneuploidy and human brain disorders.
I
LY S O P H O S P H O L I P I D S I G N A L I N G
Lysophospholipids are classically known as metabolites in the biosynthesis of cell membranes; however,
they also have extracellular effects. Researchers in our
laboratory discovered the first lysophopholipid receptor,
now known as LPA1, and showed that most of the extracellular effects could be mediated via cell-surface receptors (of the G protein–coupled receptor superfamily).
We have continued to explore the cellular and physiologic functions of receptor-mediated lysophospholipid
signaling, primarily by generating and examining mutant
mice that lack lysophospholipid receptors. We currently
have multiple lines of these mutant “knockout” mice
that are missing one or more lysophospholipid receptors.
Identifying degenerative hearing loss in S1P2/S1P3
double knockout mice (2 inoperable genes) adds another
medically relevant indication to a growing list that
reflects the potential medical relevance of lysophospholipid signaling. In mice lacking the S1P2 and S1P3
receptors, auditory and vestibular problems develop
gradually, resulting in vestibular (related to balance)
dysfunction and complete hearing loss by early adulthood (Fig. 1A). We found that these problems are due
F i g . 1 . Auditory defects in S1P 2 /S1P 3 double knockout mice.
A, The acoustic startle response, a measure of hearing function,
is nearly absent in mice lacking both copies of S1P 2 , indicating
that these mice are deaf. B, Histologic examination of cochlea
from wild-type (left) and S1P2/S1P3 double knockout (right) adult
mice reveals loss and disorganization of inner hair cells.
to neurodegeneration of the hair cells that normally
transduce auditory and vestibular signals to the brain
(Fig. 1B). This result suggests that S1P receptors contribute to the survival of normal hair cells and may
provide a new molecular target for preventing neurodegenerative hearing loss.
NEURAL ANEUPLOIDY
Cells in the brain can be genomically nonidentical
because they have lost or gained chromosomes, a condition termed aneuploidy. Since discovering this new aspect
of brain organization, we have continued to extend our
knowledge. We are developing new methods to investigate the anatomic and functional importance of aneuploidy. In the past year, using optimized fluorescent in
situ hybridization, we were able to analyze multiple
chromosomes in individual postmitotic cells, providing
a clearer look at the chromosomal complement present
within single neurons (Fig. 2). In addition, we are devel-
F i g . 2 . Multiple sequential hybridizations with fluorescently labeled
DNA probes. In this image, a nucleus from a single postmitotic
mouse neuron was probed for 10 different chromosome (Ch) pairs
(mice have a total of 20 chromosome pairs). Three copies can be
detected for chromosomes 8 and 14.
266 MOLECULAR BIOLOGY
2007
THE SCRIPPS RESEARCH INSTITUTE
oping and optimizing techniques with increased throughput to determine the range of chromosomal aneuploidies
in larger populations of cells. We continue to investigate
how neural aneuploidy affects brain function at both
cellular and system-wide levels, including the possible
contribution of aneuploidy to neuropsychiatric disorders.
Chemical Glycobiology
PUBLICATIONS
Birgbauer, E., Chun, J. New developments in the biological functions of lysophospholipids. Cell. Mol. Life Sci. 63:2695, 2006.
P. Sobieszczuk, L. Stewart, H. Tian, Y. Zeng
Chao, C., Herr, D., Chun, J., Xu, Y. Ser18 and Ser23 phosphorylation is required
for p53-dependent apoptosis and tumor suppression. EMBO J. 25:2615, 2006.
Chun, J., Rosen, H. Lysophospholipid receptors as potential drug targets in tissue
transplantation and autoimmune diseases. Cur. Pharm. Des. 12:161, 2006.
Fukushima, N., Shano, S., Moriyama, R., Chun, J. Lysophosphatidic acid stimulates neuronal differentiation of cortical neuroblasts through the LPA1-Gi/o pathway.
Neurochem. Int. 50:302, 2007.
Gardell, S.E., Dubin, A.E., Chun, J. Emerging medicinal roles for lysophospholipid
signaling. Trends Mol. Med. 12:65, 2006.
J.C. Paulson, O. Blixt, L.K. Allin, H. Andersson-Sand,
O. Berger, O.V. Bohorov, J. Busch, R. Chakravarthy,
W. Chen, G. Completo, H. Fang, D. Lebus, L. Liao, X. Liu,
B. Ma, R. McBride, C. Nycholat, M. O’Reilly, N. Razi,
e investigate the roles of glycan-binding proteins that mediate cellular processes central
to immunoregulation and human disease. We
work at the interface of biology and chemistry to understand how the interaction of glycan-binding proteins with
their ligands mediates cell-cell interactions, endocytosis,
and cell signaling. Our multidisciplinary approach is complemented by a diverse group of chemists, biochemists,
cell biologists, and molecular biologists.
W
BIOLOGICAL ROLES OF SIGLECS
Herr, D.R., Grillet, N., Schwander, M., Rivera, R., Muller, U., Chun, J. Sphingosine 1-phosphate (S1P) signaling is required for maintenance of hair cells mainly
via activation of S1P2. J. Neurosci. 27:1474, 2007.
Inoue, M., Yamaguchi, A., Kawakami, M., Chun, J., Ueda, H. Loss of spinal substance P pain transmission under the condition of LPA1 receptor-mediated neuropathic pain. Mol. Pain 2:25, 2006.
Kingsbury, M.A., Yung, Y.C., Peterson, S.E., Westra, J.W., Chun, J. Aneuploidy in
the normal and diseased brain. Cell. Mol. Life Sci. 63:2626, 2006.
Lee, C.W., Rivera, R., Dubin, A.E., Chun, J. LPA4/GPR23 is a lysophosphatidic
acid (LPA) receptor utilizing Gs-, Gq/Gi-mediated calcium signaling and G12/13mediated Rho activation. J. Biol. Chem. 282:4130, 2007.
Lee, C.-W., Rivera, R., Gardell, S., Dubin, A.E., Chun, J. GPR92 as a new
G12/13- and Gq-coupled lysophosphatidic acid receptor that increases cAMP, LPA5.
J. Biol. Chem. 281:23589, 2006.
Rehen, S.K., Kingsbury, M.A., Almeida, B.S.V., Herr, D., Peterson, S., Chun, J. A
new method of embryonic culture for assessing global changes in brain organization. J. Neurosci. Methods 158:100, 2006.
Rivera, R., Chun, J. Biological effects of lysophospholipids. Rev. Physiol. Biochem.
Pharmacol. Published online August 10, 2006. doi:10.1007/112_0507.
Theilmeier, G., Schmidt, C., Herrmann, J., Keul, P., Schäfers, M., Herrgott, I., Mersmann, J., Larmann, J., Hermann, S., Stypmann, J., Schober, O., Hildebrand, R.,
Schulz, R., Heusch, G., Haude, M., von Wnuck Lipinski, K., Herzog, C., Schmitz,
M., Erbel, R., Chun, J., Levkau, B. High-density lipoproteins and their constituent
sphingosine-1-phosphate directly protect the heart against ischemia/reperfusion injury
in vivo via the S1P3 lysophospholipid receptor. Circulation 114:1403, 2006.
The siglecs are a family of 13 sialic acid–binding
proteins that function as cell-signaling coreceptors.
They are expressed on glial cells and on a variety of
leukocytes that mediate acquired and innate immune
functions, including B cells, eosinophils, macrophages,
dendritic cells, and natural killer cells. Siglecs are a
subfamily of the immunoglobulin superfamily that have
a unique N-terminal Ig domain that confers the ability
to bind sialic acid–containing carbohydrate groups
(sialosides) of glycoproteins and glycolipids. The cytoplasmic domains of most siglecs contain tyrosine-based
inhibitory motifs characteristic of accessory proteins
that regulate transmembrane signaling and endocytosis
of cell-surface receptor proteins. The diverse specificity for their sialoside ligands and their variable cytoplasmic regulatory elements provide siglecs with attributes
for unique roles in the cell-surface biology of each cell
that expresses them.
The best understood siglec is CD22 (siglec-2), an
accessory molecule of the B-cell receptor (BCR) complex that has both positive and negative effects on receptor signaling. The carbohydrate ligand recognized by
CD22 is the sequence Siaα2-6Galβ1-4GlcNAc found
on both neighboring glycoproteins of both B cells (cis
ligands) and on cells that interact with B cells (e.g.,
T cells, trans ligands). Interactions of CD22 with cis
or trans ligands regulate aspects of B-cell activation,
proliferation, and development.
We found that CD22 is predominately associated
with clathrin-coated pits in resting B cells, whereas BCRs
are minimally associated with clathrin domains. Mice
MOLECULAR BIOLOGY
2007
deficient in the ligand for CD22 have greater colocalization of CD22 and the BCR in fused raft-clathrin domains
than do mice that have the ligand, accounting for the
immunosuppression in deficient mice. In wild-type mice,
after antigen activation, the BCR is endocytosed via raftclathrin domains, a logical site for the dampening of
B-cell signaling by CD22. In resting cells, CD22 undergoes constitutive endocytosis, which can result in internalization of high-affinity ligands of CD22 (Fig. 1).
F i g . 1 . Relationship between microdomain localization of the
BCR and CD22 (siglec-2), a regulator of BCR signaling that binds
glycan ligands.
A major barrier to studying the ligand-binding properties of siglecs and their role in siglec biology is the
difficulty in creating synthetic probes that compete with
endogenous (cis) ligands. Even highly multivalent polymers containing the natural glycan ligand will not bind
to cell-surface siglecs unless cis ligands are first destroyed.
However, we found that bifunctional molecules containing a high-affinity ligand of CD22 coupled to the
antigen NP will dock an anti-NP IgM to CD22 on the
surface of B cells. In effect, the IgM acts as a decavalent protein scaffold that promotes spontaneous assembly of an immune complex on the surface of B cells
driven by the bifunctional ligand of CD22 (Fig. 2). Once
THE SCRIPPS RESEARCH INSTITUTE
267
formed, IgM activates complement-mediated killing of
the cell, suggesting a therapeutic approach for treatment of B-cell leukemia.
Another siglec of current interest is myelin-associated
glycoprotein (siglec-4). This siglec is expressed on glial
cells and recognizes the sialoside Siaα-3Galβ1-3(Siaα26)GalNAc-R found on O-linked glycans of glycoproteins and glycolipids. Functionally, myelin-associated
glycoprotein stabilizes interactions between glial cells
and axons essential for normal organization of myelin
and inhibits axonal regeneration, which is currently a
target for pharmaceutical intervention to promote nerve
regeneration. We have developed potent inhibitors of
myelin-associated glycoprotein that reverse its ability
to block growth of axons, and in collaborative studies
with R. Schnaar, Johns Hopkins University, Baltimore,
Maryland, we are investigating the potential of the inhibitors to promote nerve growth in vivo.
With these successes, we have embarked on a
major effort to identify high-affinity ligand analogs of
each siglec to produce ligand-based tools to investigate the biological roles of the siglecs in innate and
adaptive immunity. A major focus is investigation of
multivalent protein scaffolds that provide favorable
geometry for presentation of bifunctional ligands to
cell-surface siglecs. In this regard, we are collaborating with M.G. Finn, Department of Chemistry, on the
use of viral capsids that can be functionalized to carry
variable numbers of synthetic ligands.
S I A L O S I D E A N A L O G G LY C A N A R R AY S
We have developed a robotically printed glycan array
that displays sialoside analogs to assess the affinity of
siglecs for unnatural substituents at the C-9 and C-5
positions of sialic acids. Even in the initial experiments
with 65 acyl substituents at the C-9 position of sialic
acid, the method was a powerful one for identifying
substituents that increase the affinity of the natural
ligand for siglecs by 100-fold or more (Fig. 3). In collaboration with K.B. Sharpless, Department of Chemistry, we have created another 80 analogs by using
click chemistry to couple a library of alkynes to sialosides containing 9-azido-N-acetyl-neuraminic acid.
Results from the array can be rapidly assimilated into
the synthesis of high-affinity ligands and ligand-based
probes of the corresponding siglec by using our flexible
chemoenzymatic synthesis strategies.
C O N S O R T I U M F O R F U N C T I O N A L G LY C O M I C S
F i g . 2 . Bifunctional ligands of CD22 mediate binding of IgM to
CD22 on the surface of B cells.
Members of our laboratory also staff 2 scientific
cores for the Consortium for Functional Glycomics,
268 MOLECULAR BIOLOGY
2007
THE SCRIPPS RESEARCH INSTITUTE
influenza and related H1 avian influenza viruses and
the more recent avian influenza virus (H5N1) to identify mutations required to switch specificity from avian
receptors to human-type receptors.
PUBLICATIONS
Alvarez, R.A., Blixt, O. Identification of ligand specificities for glycan-binding proteins using glycan arrays. Methods Enzymol. 415:292, 2006.
Blixt, O., Hoffmann, J., Svenson, S., Norberg, T. Pathogen-specific carbohydrate
antigen microarrays: a chip for detection of Salmonella O-antigen-specific antibodies. Glycoconjugate, in press.
Blixt, O. Razi, N. Chemoenzymatic synthesis of glycan libraries. Methods Enzymol.
415:137, 2006.
Bohorov, O., Anderson-Sand, H., Hoffmann, J. Blixt, O. Arraying glycomics: a
novel bi-functional spacer for one-step microscale derivatization of free reducing
glycans. Glycobiology 16:21C, 2006.
F i g . 3 . Sialoside analog glycan microarray reveals high-affinity
ligands for CD22. A, Sialoside ligands of CD22 with amino-terminated linkers are printed on N-hydroxyl succinimide (NHS)–activated glass slides, resulting in a covalent amide bond. B, The natural
ligand (3) with various substituents (1, 2, 4, 6) and a nonligand
control (5) are printed in 10 replicates at 10 twofold diluted printing concentrations. Overlay with a fluorescence-labeled CD22-Ig
chimera reveals the increased binding to various substituents compared with the natural ligand.
organized to elucidate the mechanisms by which glycanbinding proteins mediate cell communication (http://www
.functionalglycomics.org/). Scientists in the Mouse Transgenics Core, led by B. Ma, have created 8 novel mouse
strains from C57Bl/6 embryonic stem cells that are
deficient in genes for key glycan-binding proteins that
affect immune function. Scientists in the Glycan Array
Synthesis Core, led by O. Blixt, have produced a library
of synthetic glycans by chemoenzymatic synthesis for
use in numerous applications. In addition, scientists
in the Scripps DNA Microarray Core, led by S. Head,
designed and conducted investigator-initiated analysis
with a custom-based microarray with genes of relevance for the consortium.
A major achievement by staff in the Glycan Array
Synthesis Core is the development of the world’s largest
glycan microarray, which currently has more than 300
unique structures, mostly synthetic glycans produced
by chemoenzymatic synthesis. Now produced in collaboration with the DNA Microarray Core, the microarray
is widely used by investigators around the world to
assess the specificity of glycan-binding proteins that
mediate a broad scope of biological interactions. In an
exemplary collaboration with I.A. Wilson and J. Stevens,
Department of Molecular Biology, this array was used
to investigate the specificity of the 1918 pandemic
Crocker, P., Paulson, J.C., Varki, A. Siglecs and their roles in the immune system.
Nat. Rev. Immunol. 4:255, 2007.
Glaser, L., Conenello, G., Paulson, J.C., Palese, P. Effective replication of human influenza viruses in mice lacking a major α2-6 sialyltransferase. Virus Res. 126:9, 2007.
Han, S., Collins, B.E., Paulson, J.C. Synthesis of 9-substituted sialic acids as
probes for CD22-ligand Interactions on B cells. In: Frontiers in Modern Carbohydrate Chemistry. Demchenko, A.V. (Ed.). Oxford University Press, New York, 2007,
p. 2. ACS Symposium Series 960.
Taniguchi, N., Paulson, J.C. Frontiers in Glycomics: Bioinformatics and Biomarkers
in Disease, September 11-13, 2006, Natcher Conference Center, NIH Campus,
Bethesda, MD, USA. Proteomics 7:1360, 2007.
Tateno, H., Li, H., Schur, M.J., Bovin, N., Crocker, P.R., Wakarchuk, W.E., Paulson, J.C. Distinct endocytic mechanisms of CD22 (siglec-2) and siglec-F reflect
roles in cell signaling and innate immunity. Mol. Cell. Biol. 27:5699, 2007.