`TWNOILVNFHLJOALIdOUd
`ANIDIGAWJOAAVAUAIT
`
`
`
`
`
`Relapse-FreeSurvival
`
`0
`
`.
`
`p 0.00001, Log-rank
`Genome & Co. v. Univ. of Chicago
`PGR2019-00002
`UNIV. CHICAGO EX. 2026
`
`
`
`
`
`
`
`
`
`
`
`IGaW
`
`16kinases
`
`does 1
`
`Poo 00 Molecular
`ric damm WMA LUN WH ETI E Subtype
`(aA0 {
`
`|
`
`Differentiation
`
`0
`
`Kinase
`Score
`
`‘
`
`pen
`
`|
`
`
`
`
`
`
`
`Cancer
`
`Research
`
`A Journal ofthe American Association for Cancer Research
`Publishing Information
`
`AACR Officers
`William N. Hait, President
`Raymond N. DuBois,
`President-Elect
`to AACR Subscriptionoffice. The
`Bayard D. Clarkson, ‘Treasurer
`address to whichall business
`Geoffrey M. Wahl, Past President
`communication, remittances (in
`United States currency orits
`MargaretFoti,
`equivalent), and subscription orders
`Chief Executive Officer and
`should be sent: Turpin Distribution,
`Managing Editor
`(North American Office), ‘The
`Publications Committee
`Bleachery’, 143 West Street,
`MichaelA.Caligiuri, Chairperson
`New Milford, Cl 06776 |[Phone:
`Ann F. Chambers
`(860) 350-0041; Fax: (860) 350-0039;
`Bayard D. Clarkson
`or E-mail: turpinNA@turpin-
`Yves A. De Clerck
`distribution.com|; or (UK Main
`Glen Dranoff
`Office), Pegasus Drive, Stratton
`RaymondN. DuBois
`Business Park, Biggleswade,
`D, Gary Gilliland
`Bedfordshire 8G18 8TQ, United
`Joe W. Gray
`Kingdom |Phone: (44) 0-1767-604-
`Susan Band Horwitz
`957; Fax: (44) 0-1767-601-640; or
`David J. Hunter
`E-mail: custserv@turpin-
`V. Craig Jordan
`distribution.com|]. In Japan, send
`Kenneth W. Kinzler
`orders and inquiries to: USACO
`Michelle M. Le Beau
`Corporation, usaco@usaco.co.jp;
`Leroy Fong Liu
`Swets Japan, info@jp.swets.com;
`Luis FE. Parada
`EBSCO Japan, Japan@ebsco.com,
`David R. Parkinson
`Individuals who are not AACR
`Ramon Parsons
`members may subscribe to Volume
`Louis M. Staudt
`68 (2008) of Cancer Research at the
`Steven R. Tannenbaum
`rate of $760 ULS./$900foreign.
`Cheryl L. Willman
`Cancer Research is only available
`to institutions as a combined
`George D. Yancopoulos
`subscription with Clinical Cancer
`Publishing and Editorial
`Research. (Foreign subscription cost
`Staff, Cancer Research
`includes cost of surface mail. For
`Publisher: Kathleen Case
`faster delivery, expedited mailing rates
`Editorial Director:
`are available by contacting the AACR
`J. Heather Cullen
`Publications department.) Canadian
`Asst. Director, Editorial Systems
`subscribers should add 7%GST.
`& Journal Manager:
`Changes of address notification
`Kelly A. Hadsell
`should be sent 60days in advance
`Production Manager:
`andinclude both old and new
`Morgan M. Robinson
`addresses. Membersubscribers should
`Production Editors:
`send address changes to: AACR
`Kristin D. Murray
`MemberServices, 615 Chestnut St,
`Jeremy Rosenberg
`17th Floor, Philadelphia, PA 19106-
`Staff Assistant:
`4404, Nonmember subscribers should
`Shira Carroll
`send changes ofaddress to: AACR
`Editorial Associates:
`Subscriptionoffice (aforementioned
`Crystal Cheepudom
`address). Copies of the journal which
`MikeWilliams
`are undeliverable because of address
`Editorial Assistants:
`changeswill be destroyed,
`Kathleen Jasion
`Kyle A. Overturf
`Claims for missing issues made
`Naima D. Stone
`within 90days oftheissue date
`Special Issues Editor:
`will be honored while supplies
`Diane M. Marinelli
`last. Institutions and nonmember
`Associate Director,
`individuals, contact AACR
`Electronic Publishing:
`Subscription office at the
`Michael Beveridge
`aforementioned address. AACR
`Electronic Publishing Specialist:
`members, contact AACR Member
`Mary Beth Cunney
`Services department. Phone: (215)
`Graphics Director:
`440-9300; Fax: (215) 440-9412;
`Paul Driscoll
`E-mail: membership@aacr.org,
`Advertising Production:
`Susan Moore
`Singleissues copies and copies of
`No responsibility is accepted by the
`back stock ofthe journal may be
`Marketing Director:
`Editors or by AACR, Inc. for the
`Subscription Information
`ordered from the AACRSubscription
`Jeremy Thompson
`opinions expressed by contributors 0!
`Subscriptions include online access.
`office (see “Subscription
`Circulation: Karola Rac
`for the content of the advertisements.
`Except for members of the AACR,all
`Information"). Prices are available
`Site Licensing: Robert Bergiven
`Copyright 2008 by the American
`subscriptions are payablein advance
`upon request.
`
` Association for Cancer Research, Inc.
`
`
`
`Cancer Research [ISSN 0008-5472
`* CNREA8| is published twicea
`month, onevolumeperyear, by the
`AACR,Inc. Periodicals postage paid
`at Philadelphia, PA, and additional
`mailing offices. POSTMASTER: Send
`address changes to: AACR, 615
`Chestnut St, 17th Floor, Philadelphia,
`PA 19106-4404. Printed on acid-free
`paper in the United States of America,
`Submission of Manuscripts
`Online submission of new
`manuscripts is available at http://ean.
`msubmit.net. Until January 31,
`2008, revisions of manuscripts with
`numbers lower than CAN-07-5000
`must be submittedat http://www.
`rapidreview.com/AACR2/author,
`html. Beginning February 1, 2008,
`revisionsof manuscripts with numbers
`lower than CAN-07-5000 must be
`uploaded as new submissions at
`http//can.msubmit.net and must
`includea copy ofthe previous
`decision letter and comments as
`supplemental data. Complete
`submission instructions andfurther
`information onthejournal's scope
`andpublicationpolicies are available
`in the Information for Authors at
`hup://cancerres.aacrjournals.org,
`and authors are strongly encouraged
`to review this information prior
`to submission.
`
`Manuscript Processing Fee
`Journal policy requires that a
`manuscript fee of $75 be assessed
`for each papertodefray the expenses
`incurred in theeditorial review
`process. This fee is nonrefundable
`andis levied regardless of the decision
`rendered on the paper. Upon
`completion ofthe online submission
`process, authors will be directed to a
`securesite for remitting payment via
`credit card, Authors whoare unable
`to pay by credit card at the time of
`submission must contact Ms. Jeri
`Williamsof the AACRFinanceoffice
`at jeri.williams@aacr.org to make
`other arrangements. Manuscripts will
`not enter the review process until the
`submissionfee has been paid.
`Page Charges
`Accepted manuscripts will be
`published with the understanding
`that the author(s) will pay a charge
`of $85 perprinted pagefor articles
`one toSix pages in length; $115 per
`additional page.
`
`Advertisements in
`Cancer Research
`
`Advertisement insertion orders and
`copy must be received approximately
`five weeksprior to the date ofthe
`issue in which the advertisement
`is to be published. Inquiries about
`advertising should be directedto:
`M. |. Mrvica Associates, Inc., 2 West
`‘Taunton Avenue, Berlin, NJ 08009;
`Phone: (856) 768-9360; Fax: (856)
`753-0064.
`
`Copyright and Permissions
`All authors whowish to publish
`in Cancer Research must formally
`transfer copyright to the American
`Association for Cancer Research,
`Inc., using the release formavailable
`al www.aacrsmartsubmit.org. It is
`understoodby this transferthat
`theauthors relinquish all exclusive
`rights of copyright ownership,
`including the rights of reproduction,
`derivation distribution,sale,
`anddisplay,
`
`Authors whopreparedtheirarticles
`as part oftheir official duties as
`employeesofthe U.S, Federal
`government are not requiredto
`transfer copyright to the AACR, Inc.,
`since thesearticles are considered to
`be in the public domain. However, it
`is necessary for these authors to sign
`the appropriatesection of the transfer
`form, In thecase ofarticles supported
`by federal grants or contracts,
`copyright transfer to AACR,Inc., is
`required. The federal government
`may retain a nonexclusive license to
`publish or republish such material.
`
`Further information onthejournal's
`Copyright and Permissions policy,
`including instructions on how to
`request permission to reproduce
`articles, or portionsofarticles,is
`available at http://www.aacr.org/
`home/scientists/publications-
`of-the-aacr/copyright-and-
`permissions-policy.aspx.
`
`CancerResearchis abstracted or indexed
`in Biological Abstracts, Biochemistry & .
`Biophysics Citation Index, Basic BIC ISIS,
`BIOSIS Previews (R) Database,
`€ shemical
`Abstracts, Index Medicus, MEDLINE,
`Current Contents/Clinical Medicine,
`Reference Update, and Science Citation
`Index (expanded).
`
`
`
`Cancer Research
`
`A Journal of the American Association for Cancer Research
`hasSnRR
`February 1, 2008 - Pages 627-956
`Volume 68 +» Number3
`
`
`Reviews
`
`HumanLeukocyte Antigen-G and Cancer Immunoediting.
`Mirjana Urosevic and Reinhard Dummer sessO27
`
`TBX3 Is Overexpressedin Breast Cancer and Represses
`p1448Fby Interacting with Histone Deacetylases. Will Yarosh,
`Tomasa Barrientos, Taraneh Esmailpour, Limin Lin,
`Philip M. Carpenter, Kathryn Osann, Hoda Anton-Culver,
`and Taosheng Huang.csssessssesssssssssssssssssssnesssesssesssssssnsssasssssssssnsussssnsensseOOD
`
`‘Targeting the Eukaryotic Translation Initiation Factor 4E for
`Cancer Therapy. Jeremy BR. Graff, Bruce W. Konicek,Julia H. Carter,
`
`
`and Eric G. MarcussOn..scssssssorsrsessssssnsrsssensseseerenstsacersnenssnsnenerssss
`ww31
`
`Mitochondrial Cytochrome B Gene Mutation Promotes Tumor
`Growth in Bladder Cancer, Santanu Dasgupta,
`Mohammad Obaidul Hoque, Sunil Upadhyay, and David Sidransky sess700
`
`Meeting Report
`
`Cell, Tumor, and Stem Cell Biology
`
`Meeting Report: Innovations in Prostate Cancer Research.
`Wadih Arap, Martin Trepel, Bruce R. Zetter, and Renata Pasqualintsss635
`
`Priority Reports
`
`Alterations in Gemin5 Expression Contribute to Alternative
`mRNASplicing Patterns and Tumor Cell Motility. Jong Heun Lee,
`Christine E. Horak, Chand Khanna, Zhaojing Meng, Li Rong Yu,
`Timothy D. Veenstra, and Patricia S. SUCCG sesssessesscseessesteeeseneeersensenneeneeOSD
`
`A First-Generation Multiplex Biomarker Analysis of Urine for
`the Early Detection of Prostate Cancer. Bharathi Laxman,
`David S. Morris, Jianjun Yu, Javed Siddiqui, Jie Cao, Rohit Mehra,
`Robert J. Lonigro, Alex ‘lsodikov, John T. Wei, Scott A, Tomlins,
`and Arul M, Chinmaiyariesseuessssessessesnesnessssssesseescenassssereorsnnarneenecnnsenneesee ODD
`
`Molecular Biology, Pathobiology, and Genetics
`
`In vive Switching of Human MelanomaCells between
`Proliferative and Invasive States. Keith $. Hoek, Ossia M. Eichholf,
`Natalie C. Schlegel, Udo Débbeling, Nikita Kobert, Leo Schaerer,
`Silvio Hemmi, and Reinhard Dummer sssssesesssseessssessneeeeseeneenenn senna OOO)
`
`Multiple Alternative Splicing Markers for Ovarian Cancer.
`Roscoe Klinek, Anne Bramard, Lyna Inkel, Genevieve Dufresne-Martin,
`Julien Gervais-Bird, Richard Madden, Eric R. Paquet, ChuShin Koh,
`Julian P. Venables, Panagiotis Prinos, Manuela Jilaveanu-Pelmus,
`RaymundWellinger, Claudine Rancourt, Benoit Chabot,
`and Sherif Abou Elelas.seccsue
`
`Modeling Genomic Diversity and Tumor Dependency in Malignant
`Melanoma.William M. Lin, Alissa C. Baker, Rameen Beroukhim,
`Wendy Winckler, Whei Feng, Jennifer M. Marmion,Elisabeth Laine,
`Heidi Greulich, Hsiuyi Tseng, Casey Gates, F. Stephen Hodi,
`Glenn Dranoff, William BR. Sellers, Roman K, Thomas,
`Matthew Meyerson, Todd R. Golub, Reinhard Dummer,
`Meenhard Herlyn, Gad Getz, and Levi A. Garraway ssessssseseessser reese O64
`
`Tristetraprolin Down-regulates Interleukin-8 and Vascular
`Endothelial Growth Factor in Malignant Glioma Cells,
`Esther Suswam, Yanyan Li, Xiaowen Zhang, G. Yancey Gillespie,
`Xuelin Li, John J. Shacka, Liang Lu, Lei Zheng, and Peter H. King.........674
`
`Spontaneous Squamous Cell Carcinoma Induced by the Somatic
`Inactivation of Retinoblastoma and Trp53 Tumor Suppressors.
`Ana Belén Martinez-Cruz, Mirentxu Santos, M. FernandaLara,
`CarmenSegrelles, Sergio Ruiz, Marta Moral, Corina Lorz,
`Ramon Garcia-Escudero, and Jestis M. ParaimiocscsO83
`
`EnhancedActivation of Epidermal Growth Factor Receptor
`Caused by Tumor-Derived E-Cadherin Mutations. Anja Bremm,
`Axel Walch, Margit Fuchs, Jorg Mages, Justus Duyster, Gisela Keller,
`Christine Hermannstidter, Karl-Friedrich Becker, Sandra Rauser,
`Rupert Langer, Claus Hann von Weyhern, Heinz Hofler,
`eeneeeees
`BOE Birgit: LAE sxsssecsscsescsessresssesavsensiencssosnseesisensén jaseavsensonnsesssseeedOF
`
`CCN3/Nephroblastoma Overexpressed Matricellular Protein
`Regulates Integrin Expression, Adhesion, and Dissemination
`in Melanoma.Viviana Vallacchi, Maria Daniotti, Francesca Ratti,
`Delia Di Stasi, Paola Deho, Annamaria DeFilippo, Gabrina Tragni,
`Andrea Balsari, Antonino Carbone, Licia Rivoltini, Giorgio Parmiani,
`NoureddineLazar, Bernard Perbal, and Monica Rodolfo veces? 15
`
`Modulation of Oncogenic Phenotype in Human GliomaCells by
`Cytomegalovirus IE1-Mediated Mitogenicity. Charles $. Cobbs,
`Liliana Soroceanu, Scott Denham, Wenyue Zhang,
`and Matthias: EH. Kraus sssessecsssosssnssescssssseesssecnsssnscssssssussecsustsessrecsssennsaneente 24
`
`Tumor Cell-Secreted Caveolin-1 Has Proangiogenic Activities in
`Prostate Cancer. Salahaldin A. Tahir, Guang Yang, Alexei A. Goltsovy,
`Masami Watanabe, Ken-ichi Tabata, Josephine Addai,
`El Moataz Abdel Fattah, Dov Kadmon, and Timothy C. Thompson 731
`
`The Tumor Suppressor LKB1 Regulates Lung Cancer Cell Polarity
`by Mediating cde42 Recruitment and Activity. Shumin Zhang,
`Katherine Schafer-Hales, Fadlo R. Khuri, Wei Zhou, Paula M. Vertino,
`ANAdamI, MarCus..cnssssseseessserssnsssenstearsesteenseetseessapseestensseseesassnesaeseeseeeee 40
`
`The HumanTrithorax Protein hASH2 Functions as an Oncoprotein.
`Juliane Lischer-Firzlaff, Isabella Gawlista, Jorg Vervoorts, Karsten Kapelle,
`Till Braunschweig, Gesa Walsemann, Chantal Rodgarkia-Schamberger,
`Henning Schuchlautz, Stephan Dreschers, Elisabeth Kremmer,
`Richard Lilischkis, Christa Cerni, Axel Wellmann,
`
`and Bernhard Liischer scsssssssesssssesneees
`Loss of Lkb1 Provokes Highly Invasive Endometrial
`Adenocarcinomas.Cristina M. Contreras, Sushma Gurumurthy,
`J. Marshall Haynie, LaneJ. Shirley, Esra A. Akbay, Shana N. Wingo,
`John O. Schorge, Russell R. Broaddus, Kwok-Kin Wong, Nabeel Bardeesy,
`and Diego H. CastrillonsssscssscsssssssssossssseossssssntssnsssnsssetsssissssnsesssessssversensssonsanessOD)
`
`Sixteen-Kinase Gene Expression Identifies Luminal Breast
`Cancers with Poor Prognosis. Pascal Finetti, Nathalie Cervera,
`Emmanuelle Charafe-Jauffret, Christian Chabannon, Colette Charpin,
`Max Chaffanet, Jocelyne Jacquemier, Patrice Viens, Daniel Birnbaum,
`eee 767
`and Francois Bertucci..
`
`Genotoxic Stress-Induced Expression of p53 and Apoptosis in
`Leukemic Clam Hemocytes with Cytoplasmically Sequestered p53.
`Stefanie Bottger, EmilyJerszyk, Ben Low, and Charles Walker sesso777
`
`
`
`
`
`Contents (Continued)
`
`BCL6 Represses Smad Signaling in Transforming Growth Factor-B
`Resistance. Degang Wang,Jianyin Long, Fangyan Dai, Min Liang,
`Xin-Hua Ferg, and Nia Livsssssessesssssesessesecscsseeeetonssesesseesrsacesssscsearesse OS
`
`Epitope Landscapein Breast and Colorectal Cancer. Neil H. Segal,
`D. Williams Parsons, Karl S. Peggs, Victor Velculescu, Ken W, Kinzler,
`Bert Vogelstein, and James P. AllisODsssBD
`
`CD43, but not P-Selectin Glycoprotein Ligand-1, Functions as an
`E-Selectin Counter-Receptor in Human Pre-B-Cell Leukemia NALL-1.
`Chizu Nonomura, Jiro Kikuchi, Nobutaka Kiyokawa, Hidenori Ozaki,
`Kanae Mitsunaga, Hidenobu Ando, Akiko Kanamori, Reiji Kannagi,
`Junichiro Fujimoto, Kazuo Muroi, Yusuke Furukawa, and
`Mitsuru Nakamura...
`sosvenyuivaansvunssnenvee OU
`
`ABCG2 Expression and Side Population Abundance Regulated by a
`‘Transforming Growth Factor $-Directed Epithelial-Mesenchymal
`Transition. Liqun Yin, Paola Castagnino, and Richard K. Assoian.........800
`
`Repression of B-Cell Linker (BLNK) and B-Cell Adaptor for
`Phosphoinositide 3-Kinase (BCAP) Is Important for Lymphocyte
`‘Transformationby Rel Proteins, Nupur Gupta,Jeffrey Delrow,
`AmarDrawid, Anirvan M, Sengupta, Gaofeng Fan,
`
`and Céline Gélinas ws
`dsinagendarnenesssbadvdissbuiueesbivadessieniitessieiOOS:
`
`The EGER-STAT3 Oncogenic Pathway Up-regulates the Emel
`Endonuclease to Reduce DNA Damageafter TopoisomeraseI
`Inhibition. Arnaud Vigneron, Erick Gamelin, and Olivier Coqueret.....815
`
`
` Experimental Therapeutics, Molecular Targets,
`
`and Chemical Biology
`
`‘Tamoxifen Resistance in Breast Tumors Is Driven by Growth Factor
`Receptor Signaling with Repression of Classic Estrogen Receptor
`Genomic Function, Suleiman Massarweh, C. Kent Osborne,
`ChadJ. Creighton, Lanfang Qin, Anna Tsimelzon, Shixia Huang,
`Heidi Weiss, Mothaffar Rimawi, and Rachel Schiff.B26
`
`Inhibition of Endoplasmic Reticulum Stress-Induced Apoptosis
`of Melanoma Cells by the ARC Protein, Li Hua Chen, Chen ChenJiang,
`Ralph Watts, Rick F. Thorne, Kelly A. Kiejda, Xu Dong Zhang,
`and Peter Hersey scessnsseessensecesersessssnessansnesaneseeanen
`
`Aggravated Endoplasmic Reticulum Stress as a Basis for Enhanced
`Glioblastoma Cell Killing by Bortezomib in Combination with
`Celecoxib or Its Non-Coxib Analogue, 2,5-Dimethyl-Celecoxib.
`Adel Kardosh, Encouse B. Golden, Peter Pyrko, Jasim Uddin,
`Florence M. Hofman, ThomasC. Chen, Stan G. Louie, Nicos A. Petasis,
` saseeereane neers:
`and Axel H. Schénthal.......
`sceegunseensearseersvereneee4d
`
`Tamoxifen Stimulates the Growth of Cyclin D1-Overexpressing
`Breast Cancer Cells by Promoting the Activationof Signal
`Transducer and Activator of Transcription 3, Yuki Ishii,
`seeeresneeeres:
`Samuel Waxman, and Doris Germain cscs
`
`
`Prostate StemCell Antigen Vaccination Induces a Long-term
`Protective Immune Response against Prostate Cancerin the
`Absence of Autoimmunity. Maria dela Luz Garcia-Hernandez,
`Andrew Gray, Bolyn Hubby, Otto J. Klinger, and W. Martin Kast..861
`
`Tumor-Induced Senescent T Cells with Suppressor Function;
`A Potential Form of Tumor Immune Evasion, Carolina L. Montes,
`Andrei 1. Chapoval, Jonas Nelson, Vbenosa Orhue, Xiaoyu Zhang,
`Dan H, Schulze, Scott E. Strome, and Brian BR. Gastman sssS70
`
`Systematic High-Content Proteomic Analysis Reveals Substantial
`Immunologic Changes in Colorectal Cancer. Uta Berndt,
`Lars Philipsen, Sebastian Bartsch, Bertram Wiedenmann,
`Daniel C, Baumgart, Marcus Hammerle, and Andreas Sturm...880
`
`Self-Tolerance Does Not Restrict the CD4+ T-Helper Response
`against the p53 Tumor Antigen. Marjolein M. Lauwen,
`Sander Zwaveling, Linda de Quartel, 5. Carmela Ferreira Mota,
`Janine A.C. Grashorn, Cornelis J.M. Melief, Sjoerd H. van der Burg,
`aNd Rienk OPIN ga ssecssscssecsssssssensersenssseetssnsessssensnssansenssneseasesesnescessnessssen OOD
`
`Induction of EBV-Latent Membrane Protein 1-Specific MHC
`Class [I-Restricted T-Cell Responses against Natural Killer
`LymphomaCells. Hiroya Kobayashi, Toshihiro Nagato, Miki Takahara,
`Keisuke Sato, Shoji Kimura, Naoko Aoki, Makoto Azumi,
`Masatoshi Tateno, Yasuaki Harabuchi, and Esteban Celis sssDOL
`
`Endocrinology
`
`DRS Receptor Mediates Anoikis in Human Colorectal Carcinoma
`Cell Lines. Luciana M. Laguinge, Raed N. Samara, Wenge Wang,
`Wafik S. El-Deiry, Georgia Corner, Leonard Augenlicht, Lopa Mishra,
`ANA J. Milburn Jessup cessccacssssssessorrsessencrsassonseessnessnsesssersessncseasssarssensseveneesTOD
`
`Clinical Research
`
`Cancer-Associated Stromal Fibroblasts Promote Pancreatic Tumor
`Progression. Rosa F. Hwang, Todd Moore, Thiruvengadam Arumugam,
`Vijaya Ramachandran, Keith D. Amos, ArmandoRivera, BaoanJi,
`Douglas B, Evans, and Craig D. LOgSdOM sssessessssesseseesessessensenssnseneeseeseesesses918
`
`Epidemiology
`
`Tumor Immunobiological Differences in Prostate Cancer between
`African-American and European-American Men.Tiffany A. Wallace,
`Robyn L. Prueitt, Ming Yi, Tiffany M. Howe, John W.Gillespie,
`Harris G. Yfantis, Robert M. Stephens, Neil E, Caporaso,
`
`Christopher A. Loffredo, and Stefan AmMbS.sscsssssenssssssesseseeesseneeseerseeeeneee27
`
`Prevention
`
`Direct Evidence for Epithelial-Mesenchymal Transitions in Breast
`Cancer. Anthony J. Trimboli, Koichi Fukino, Alain de Bruin, Guo Wei,
`Lei Shen, Stephan M. Tanner, Nicholas Creasap, Thomas J. Rosol,
`Michael L. Robinson, Charis Eng, Michael C. Ostrowski,
`ANd Gustavo LEONE secsconsssssersesarssvavasnssesssnieanssnsasasassssanssssosnercnseidverssncessansDDT
`
`Raf and MEK Protein Kinases Are Direct Molecular Targets
`for the Chemopreventive Effect of Quercetin, a Major Flavonol
`in Red Wine. Ki Won Lee, NamJoo Kang, Yong-Seok Heo,
`Evgeny A. Rogozin, Angelo Pugliese, Mun Kyung Hwang,
`G. Tim Bowden, Ann M.Bode, HyongJoo Lee, and Zigang Dong...........946
`
`Corrections
`
`Correction: SNAII Silencing and Inhibition of Breast
`Turrttor Growth ...ssssccsssesvevesssserrssessssnenecssizsnnasieieassnidats iiseteneinaassesesassiossenasc56
`
`Correction: IFN Signaling in Aggressive Fibromatosis........0000000..956
`
`Correction: RPS27L Modulates DNA Damage Response..........00.956
`
`
`
`
`This material may beprotectedby Copyright law (Title 17 U.S. Code)
`
`
`
`Epitope Landscapein Breast and Colorectal Cancer
`
`Neil H. Segal,’ D. Williams Parsons,’ Karl S. Peggs,”’ Victor Velculescu,'
`Ken W.Kinzler,' Bert Vogelstein,’ and James P. Allison™
`
`and ‘Immunology Program, Memorial Sloan-Ketlering Cancer
`'Departinent of Medicine, “Ludwig Center for Cancer Immunotheray
`and Therapeutics at
`Center, New York, New York: and ‘Ludwig Center for Cancer Genetics
`The Johns Hopkins Kimmel Cancer Center, Baltimore, Maryland
`
`Abstract
`
`individual cancers contain many mutant
`The finding that
`genes not present in normal tissues has prompted consider-
`able interest
`in the cancer epitope landscape. To further
`understand such effects, we applied in silico—-based epitope
`prediction algorithms and high throughput post hoc analysis
`to identify candidate tumor antigens. Analysis of 1,152
`peptides containing missense mutations previously identified
`in breast and colorectal cancer
`revealed that
`individual
`cancers accumulate on average ~10 and ~7 novel and
`unique HLA-A*0201 epitopes,
`respectively,
`including genes
`implicated in the neoplastic process. These data suggest
`that, with appropriate manipulation of the immune system,
`tumor cell destruction in situ may provide a polyvalent tumor
`vaccine without a requirement for knowledge of the targeted
`antigens. [Cancer Res 2008;68(3):889-92]
`
`Introduction
`
`Several classes of tumor antigens have been described and
`named according to their distribution in normal and neoplastic
`tissues. These include the shared differentiation antigens, such as
`melan-A/MARTI
`(1, 2) in melanoma; cancer testes or germ cell
`antigens, such as MAGE-1 (3) and NY-ESO-1 (4) in adult testes and
`diverse tumor types; and unique tumorantigens, which generally
`carry mutations, such as CDK4 in melanoma (5) and CASP-8 in
`head and neck cancer(6).
`the unique tumor antigens was
`The immunogenicity of
`recognized in several seminal studies including animal transplant
`models (7) and chemically or UV light-induced tumors (8, 9).
`They are of particular interest because they result from somatic
`mutations in individual tumors and are absent from normal tissues
`(10,
`11), providing antitumor specificity without anticipated
`deleterious autoimmunity, Somatic mutations
`can be classified
`as either “drivers” or “passengers”. Passenger mutations provide
`no positive or negative selective advantage to the tumorbut are
`retained by chance during cell division and clonal expansion.
`In contrast, driver mutations provide a selective advantage that
`promotes the tumorigenic process. The generation of mutationsis
`continuous due to the imperfect nature of DNA replication and
`repair. Thus, the generation of additional antigens during tumor
`progression, whether driver or passenger
`(12), provides a
`continuously renewable source ofantigen.
`Recent analyses of breast and colorectal cancers showed a
`remarkable number of somatic mutations in human cancer (13).
`
`Requests for reprints: James P, Allison, Ludwig Center of Cancer Immunotherapy,
`Memorial Sloan-Ketle
`r Center, 415 East 68th Street, Z7-1560, New York, NY
`
`10065. E-mail: allisonj@mske
`O2008 American A:
`
`wiation for Cancer Research,
`
`2.CAN-07-3095
`doi: 10.1158/0008-547
`
`Among >13,000 genes analyzed, a total of 1,307 somatic mutations
`were identified in 1! breast and 11 colorectal cancers. Approxi-
`mately 83% were missense mutations, 6% nonsense,
`and_
`the
`remainder were insertions, deletions, duplications, and changes
`in noncoding regions. When extrapolated to the whole genome,
`it was calculated that
`individual
`tumors harbored an average of
`~90 amino acid-altering (i.e. nonsynonymous) mutations. The
`kind of information available from such large-scale sequencing
`studies of individual tumors has not heretofore been available but
`clearly has implications for tumor immunity.
`In the current
`study, we designed an in silico approach to
`examine whether the mutations identified in Sjoblom et al. (15)
`have the potential to generate novel epitopes that might serve as
`targets for an immune response. Using epitope prediction
`algorithms and high throughput post hoc analysis, we found
`evidence to support
`the notion that
`the human tumorigenic
`process results in the generation of multiple immune targets.
`Individual breast and colorectal cancers accumulated an average of
`~10 and ~7 novel and unique HLA-A*0201 epitopes, respectively;
`several within genes that may be drivers. These resulls provide
`insights into the unique immuneprofiles of individual tumors with
`potential clinical relevance.
`
`Materials and Methods
`
`Epitope prediction. Peptide sequences corresponded to missense
`mutations identified during the discovery phase by Sjoblom et al, (13),
`
`flanked by up to 10 amino acids on either
`side. Coneatamers of these
`peptides were analyzed with several epitope prediction algorithms for HLA-
`A®°0201 binders. Major histocompatibility complex (MEIC)-1 antigenic
`peptide processing prediction (MAPPP; ref 14), developed at the Max-Plank
`Institute, facilitates the prediction of epitopes that can bind to MHC class |
`molecules based on a score calculated for each subsequence. Each amino
`acid at a specific position within a subsequenceis given a value that has
`been precalculated and stored in static matrices. The precalculation was
`done either by BIMAS (15) or SYFPEITHI (16). Depending on the algorithm
`selected, the values were then multiplied (BIMAS) or added (SYFPEITHE) to
`determine the score for the subsequence. Peptides qualified as posilive if
`they scored = 100 and 24, respectively (17, 18). RANKPEP (19) uses specific
`scoring matrices from sets of peptides known to bind to MIIC molecules as
`the predictor of MHC-peptide binding. Peptides qualified as positive if the
`percentage optimum was 250% or higher. NetMHC (20, 21) predicts
`peplide-MHC binding using artificial neural networks (ANN) and weight
`matrices. For ANN, used for HLA-A*0201 prediction, peptides scored
`positive if ICs, is =500.
`Post hoc analysis. First, we searched for unique epitopes within
`concatamers of wild-type and mutant peptides, Epitopes identified in the
`“wild-type concatamer” included both true wild-type epitopes andartifacts
`across the concatenation sites and were removed from further analysis.
`The “mutant concatamer” was then used to seareh for remaining epitopes.
`‘To ensure thal potential mutant epitopes did not span concatenation sites,
`the “mutant concatamer” used in this confirmatory phase included
`additional
`redundant characters spaced between peptides,
`thereby
`permitting confirmation of epitopes contained entirely within a mutant
`
`www.aacrjournals.org
`
`This material was copied
`
`Cancer Res 2008; 68: (3). February 1, 2008
`
`
`
`Cancer Research
`
`
`
`
`
`Table 1. HLA-A*0201 epitopes (SYFPEITHI)
`
`‘Total epitopes
`CRC; =1 epitope
`BC: =1 epitope
`
`Per sample: CRC
`: BC
`:CRC + BC
`
`241
`tH
`I
`
`Ave
`8.2
`13.6
`109
`
`SD
`47
`6.7
`63
`
`Min
`2
`8
`2
`
`Max
`17
`30)
`30
`
`2
`0
`0.56
`O01
`Per peptide: CRC
`t
`0
`0.42
`O15
`BC
`3
`0
`040
`0.13
`> CRC+BC
`
`
`Abbreviations: BC, breast cancers; CRC, colorectal cancers,
`
`
`peptide only. These were then annotated for specimen and tumor type as
`described in supplementary information in Sjoblom et. al. (13).
`Estimates of epitope frequency. The total number ofepitopes
`corresponding to each peptide, original specimen, and tumor type were
`calculated for the 11,721 genes that were successfully sequenced of the
`13,023 CCDS genes. As per Sjoblomet al. (13), we extrapolated this number
`to the total number of genes in the human genome (conservatively
`estimated at 18,203) by dividing the number ofidentified epitopes by 0.64,
`
`Results and Discussion
`
`A total of 1,152 peptides containing missense mutations,
`previously identified in breast and colorectal cancer (13) were
`concatenated into a single string of 23,924 amino acids and
`analyzed for potential MHC class
`| binders. Epitopes were
`predicted using several algorithms
`then applied to post hoc
`analysis. This analysis entailed a series of manual steps, predefined
`calculations, and macro-based algorithms to identify epitopes that
`are absent
`from corresponding wild-type sequences but present
`within the mutant peptide. We restricted our analysis to HLA-
`A*0201 9-merepitopes because this haplotype has been extensively
`
`studied andis represented in up to 27%of the population (22). All
`epitopes were confirmedto be unique 9-mers after BLASTanalysis,
`searching for “short, nearly exact matches” in the “nr” database,
`Two hundredandforty-one epitopes wereidentified using MAPPP
`(14)/SYPEITHE (16). On average, 8.2 and 13.6 epitopes were
`found per specimen in colorectal and breast cancers, respectively
`(Table 1). Each tumor contained a minimumof two and as manyas
`30 epitopes in a case of breast cancer(Fig, 1}.
`Next, we focused our attention on the 191 candidate cancer
`(CAN) genes. CANgenes wereidentified as containing mutationsin
`at
`least
`two independent
`tumors and were mutated at greater
`frequency than non-CAN genes whenadjustedforsize, nucleotide
`composition, and mutational spectra (13). We
`identified
`47 potential epitopes in CANgenes (Table 2). These epitopes were
`identified in 6 of 11 colorectal cancers and 7 of 11 breast cancers.
`Additional algorithms including NetMHC(20, 21), MAPPP (14)/
`BIMAS (15), and RANKPEP (19), were then used,
`identifying
`an averageof 182 epitopes. The majority of epitopes predicted by
`each of the different algorithms were identified by at
`least
`two
`algorithms. Epitopes predicted by SYFPEITHI, BIMAS, RANKPEP,
`and NetMHCoverlapped by 59%, 73%, 64%, and 75%, respectively.
`In sum, we found that individual colorectal and breast cancers
`accumulated an average of ~7 and ~ 10 novel and unique HLA-
`A*0201 epitopes, respectively, Approximately one new epitope was
`generated for every 10 mutations, and 45% of predicted epitopes
`were shownto be cleaved at their COOH terminal accordingto the
`PAProC proteasomeprediction algorithm(refs. 23, 24; Table 3),
`Note that these numbers are underestimates because other MHC
`molecules, not
`studied here, can present additional mutant
`peptides depending on the haplotype of individual patients.
`Because individual tumors potentially contain six distinct MHC
`class | molecules,
`including two loci each for HLA-A, HLA-B,
`and HLA-C,
`the estimated frequency of novel epitopes may be
`multiplied up to 6-fold. Thereby estimating that
`individual
`colorectal and breast cancers potentially accumulate up to ~40
`and ~ 60 novel MHCclassI restricted epitopes, respectively.
`On a more cautionary note, we have not yet shownthat these
`epitopes are actually expressedin tumorcells; we have studied only
`
`
`Epitopes per sample
`
`
`
` Figure 1. Distribution of HLA-A*0201
`
`epitopes by sample. Solid black
`bars, colorectal cancer samples;
`lined bars, breast cancer samples.
`
`i
`
`y]
`
`y]Z
`
`Co74
`
`Cag2
`
`Co108
`
`Mx22
`
`Mx27
`
`Mx30
`
`Mx32
`
`Mx38
`
`Mx41
`
`Mx42
`
`Mx43
`
`B7C E
`
`Bac
`
`Bic
`
`CancerRes 2008; 68: (3). February 1, 2008
`
`This material was copied
`
`www.aacrjournals.org
`
`
`
`Breast and Colorectal Cancer Epitopes
`
`for drivers among those studied. Mutations in drivers may be
`advantageous in that
`the tumor might not be able to “down-
`regulate” expression of the epitope if it
`is required for continued
`neoplastic growth (addicted), However, the exploitation of muta-
`tions in cancer
`immunotherapeutics is not dependent on the
`mutation being a driver. As long as the epitope is presented to the
`immune system appropriately,
`it can,
`in theory, stimulate an
`antitumor response. All of the multiple nonsynonymous mutations
`in cancers therefore potentially contribute to multiple targets
`for
`immune attack, providing the basis for
`the concept of
`combinatorial targeted immunotherapy.
`Several methods for exploiting these findings can be devised.
`Initially, they depend on sequencing each patient's cancer to some
`extent. Although this is too difficult and expensive to be done
`routinely at present, improvements in technology will soon make
`such sequencing feasible, at
`least for a limited number of genes
`most likely to harbor missense mutations. The simplest approach
`to exploit
`the resultant
`sequencing information would be to
`administer vaccines to individual patients made from their own
`tumor. Although this strategy has been attempted before in
`humans,it has often been difficult to judge the magnitude ofthe
`immune responseelicited because the antigens, if any, were nol
`known. Once potential antigens are known through sequencing,
`patients’ re