media guide t-bo_c2

This document contains an overview of the contents of the DVD-ROM packaged with The Biology of Cancer, 2nd Edition by Robert A. Weinberg. It also contains

transcripts of the movies and audio files on the DVD.

The DVD contains the following media for students and instructors:

• Figuresfromthebook(PowerPoint®andJPEGformats)

• SupplementarySidebars(PDF)

• Movies(QuickTime®andWMVformats)

• Mini-lecturesfromtheauthor(MP3format)

The media on this DVD, as well as future media updates, are also available to students and instructors at www.garlandscience.com.

SECOND EDITIONSECOND EDITION

SECOND EDITIONSECOND EDITION

SE

CO

ND

ED

ITIO

NS

EC

ON

DE

DIT

ION

CANCERRobert A. Weinberg

www.garlandscience.com

Roobbbberrtt AA

. WWWWeinbbberrgg

the biology of

CA

NC

ER

ttttttttthhhhhhhheeebiologyoofff

Thoroughly updated and incorporating the most important advances in the fast-growing field of cancer biology, The Biology of Cancer, Second Edition, maintains all of its hallmark features admired by students, instructors, researchers, and clinicians around the world.

The Biology of Cancer is a textbook for students studying the molecular and cellular bases of cancer at the undergraduate, graduate, and medical school levels. The principles of cancer biology are presented in an organized, cogent, and in-depth manner. The clarity of writing, supported by an extensive full-color art program and numerous pedagogical features, makes the book accessible and engaging. The information unfolds through the presentation of key experiments that give readers a sense of discovery and provide insights into the conceptual foundation underlying modern cancer biology.

Besides its value as a textbook, The Biology of Cancer is a useful reference for individuals working in biomedical laboratories and for clinical professionals.

Robert A. Weinberg is a founding member of the Whitehead Institute for Biomedical Research. He is the Daniel K. Ludwig Professor for Cancer Research and the American Cancer Society Research Professor at the Massachusetts Institute of Technology (MIT). Dr. Weinberg is an internationally recognized authority on the genetic basis of human cancer and was awarded the U.S. National Medal of Science in 1997.

DVD-ROM AND POSTER INCLUDED

• The enclosed DVD-ROM includes the book’s art program, a selection of movies with narration, audio files of mini-lectures by the author, Supplementary Sidebars, and a Media Guide.

• The revised “Pathways in Human Cancer” poster summarizes some of the key signaling pathways implicated in tumorigenesis and tumor progression in humans.

ISBN 9780815342199

9 780815 342199

The Biology of CancerSecond Edition

Media Guide

Media Guide for The Biology of Cancer, Second Edition2

Figures from the Book—The Art of The Biology of Cancer

F igures, tables, and micrographs from the book are located in the “Art of TBoC2” folder.Theyhavebeenpre-loaded intoPowerPointpresentations, one for each

chapter of the book. A separate folder contains individual versions of each figure, table, andmicrographinJPEGformat.Thefoldersarecalled“PowerPoint”and“JPEGs.”

Inadditiontoservingasatoolforcreatingvisualpresentationsforlectures,thePow-erPointfilesarealsousefulforprintingthefigures.Ifyouwishtoprintallthefiguresfromaparticularchapter,openthePowerPointpresentationforthechapter,selectthe“file/print” menu option, and print “All” the figures. If you wish to print only a select numberor justonefigure,choose theappropriate slidenumber in thePowerPointprinter options window, accessed through the menu described above. This process is roughlythesameonbothMacOSandWindowsoperatingsystems.

Youmayalsoprint the individual JPEGs fromtheJPEGarchiveusingAdobe®Pho-toshop®orsimilarimageeditingprograms.BothMacOSandWindowshavebuilt-inimageviewersthatalsoallowprinting.Pleaseconsulttheusermanualforthesepro-grams for further instructions.

The figures can also be imported into Microsoft Word® and other text editing pro-grams through the “insert/picture” menu option or by cutting-and-pasting the figure from one program to another. In Microsoft Word you can re-size the figure to match your document.

Media Guide for The Biology of Cancer, Second Edition 3

Supplementary Sidebars

SupplementarySidebar1.1 Each female cell can access information only from a single X chromosome

SupplementarySidebar1.2 Reproductive cloning demonstrates the extraordinary efficiency of the DNA repair apparatus

SupplementarySidebar1.3 The network of miRNA-controlled genes

SupplementarySidebar1.4 Knocking down gene expression with shRNAs and siRNAs

SupplementarySidebar1.5 Genecloningstrategies

SupplementarySidebar2.1 Commonly used histopathological techniques

SupplementarySidebar2.2 The complicated conventions for classifying and naming tumors

SupplementarySidebar3.1 IsSV40responsibleforthemesotheliomaplague?

SupplementarySidebar3.2 MaintenanceofKSHVandHPVgenomesinepisomalandchromosomalform

SupplementarySidebar3.3 Re-engineering the retrovirus genome for gene therapy

SupplementarySidebar3.4 Classic Kaposi’s sarcoma appears to be a familial disease

SupplementarySidebar3.5 ViruseslikeRSVhaveveryshortlives

SupplementarySidebar4.1 Endogenous viruses can explain tumor development in the absence of infectious viral spread

SupplementarySidebar4.2 BoveriandHansemannindependentlyhypothesizedgeneticabnormalityasthecauseofcancer cells’ malignant behavior

SupplementarySidebar4.3 SouthernandNorthernblotting

SupplementarySidebar4.4 Genesundergoamplificationforavarietyofreasons

SupplementarySidebar5.1 Makinganti-Srcantibodiespresentedamajorchallenge

SupplementarySidebar5.2 The protozoan roots of metazoan signaling

SupplementarySidebar5.3 Lateral interactions of cell surface receptors

SupplementarySidebar6.1 SystematicsurveysofphosphotyrosineandSH2interactions

SupplementarySidebar6.2 The complexities of understanding RTK signaling

SupplementarySidebar6.3 The rationale for multi-kinase signaling cascades

SupplementarySidebar6.4 Non-canonical Wnt signaling

SupplementarySidebar6.5 TheHippopathwayandcontrolofstemcellproliferation

SupplementarySidebar7.1 Heterozygosityinthehumangenepool

SupplementarySidebar7.2 Whichismorelikely—LOHorsecondary,independentmutations?

SupplementarySidebar7.3 The polymerase chain reaction makes it possible to genetically map tumor suppressor genes rapidly

SupplementarySidebar7.4 TheMSPreactionmakesitpossibletogaugemethylationstatusofpromoters

SupplementarySidebar7.5 Ubiquitylation tags cellular proteins for destruction in proteasomes

SupplementarySidebar7.6 Krebs cycle enzymes and cancer development

SupplementarySidebar7.7 Homologousrecombinationallowsrestructuringofthemousegermline

SupplementarySidebar8.1 The origins of embryonic stem cells

SupplementarySidebar8.2 Plasticityofthecellcycleclock

SupplementarySidebar8.3 Chromatinimmunoprecipitation(ChIP)

SupplementarySidebar8.4 SometumorsincreaseIdconcentrationsbyde-ubiquitylatingthem

SupplementarySidebar8.5 SpecifictargetingofcellcycleregulatorsbyE3ubiquitinligases

SupplementarySidebar8.6 The major puzzle surrounding the RB gene: retinoblastomas

SupplementarySidebar9.1 UV-Bradiation,HPV,andcutaneoussquamouscellcarcinomas

The Biology of Cancer contains a special feature called “Sidebars,” which consist of commentary that detours

slightly from the main thrust of the textual discussion. Often theseSidebarscontainanecdotesorelaborateonideaspre-sentedinthemaintext.InadditiontotheSidebarsthatappearin the text, the author has written the following additional

Sidebars, which can be accessed from the “SupplementarySidebars” folder on the DVD or from the Garland Sciencewebsite.TheseSupplementarySidebarsarecross-referencedthroughoutthetext.TheyareavailableinPDFformatforopti-mal viewing and printing.


SupplementarySidebar9.2 The TUNEL assay

SupplementarySidebar9.3 Dominant-negative functions of mutant p53 alleles: functional interactions between p53 and its p63 and p73 cousins

SupplementarySidebar9.4 Somemutantp53 alleles cause highly specific tumors

SupplementarySidebar9.5 Autophagy is critical to post-fertilization development

SupplementarySidebar10.1 TheuseoftheTRAPassayandadaptationsthereofpermitstherapidandquantitativeassessment of the levels of the catalytic activity of telomerase enzyme in eukaryotic cells

SupplementarySidebar11.1 Monoclonalantibodiesandfluorescence-activatedcellsorting(FACS)

SupplementarySidebar11.2 Howdoesmulti-steptumorprogressionactuallytakeplace?

SupplementarySidebar11.3 Symbiosisbetweendistinctsubpopulationswithinatumor

SupplementarySidebar11.4 Comparative genomic hybridization

SupplementarySidebar11.5 Arerodentcarcinogentestsreliableindicatorsofdangertohumans?

SupplementarySidebar11.6 Doessaccharincausecancer?

SupplementarySidebar11.7 Howdoesdietaffectcoloncancerincidence?

SupplementarySidebar12.1 Hematopoiesisasamodelfortheorganizationofmanykindsoftissues

SupplementarySidebar12.2 Stemcellpoolsmayexplaintheprotectiveeffectsofpregnancy

SupplementarySidebar12.3 The conserved-strand mechanism and protection of the stem cell genome

SupplementarySidebar12.4 Oxidation products in urine provide an estimate of the rate of ongoing damage to the cellular genome

SupplementarySidebar12.5 Howdoesredmeatcausecoloncancer?

SupplementarySidebar12.6 A convergence of bacterial, yeast, and human genetics led to the discovery of hereditary non-polyposis colon cancer genes

SupplementarySidebar12.7 Homology-directedrepair

SupplementarySidebar13.1 Localization of growth factors is important for proper heterotypic interactions

SupplementarySidebar13.2 Ongoing heterotypic signaling in carcinomas

SupplementarySidebar13.3 Certain highly advanced tumors provide exceptions to the generally observed dependence of carcinoma cells on stroma

SupplementarySidebar13.4 Myofibroblasts predict clinical progression of cancer

SupplementarySidebar13.5 A technique for separating stromal from epithelial cells

SupplementarySidebar13.6 Microvessel leakiness dooms many forms of anti-cancer therapy: optimizing anti-angiogenic treatments

SupplementarySidebar13.7 The temporary nature of vessel regression created by anti-angiogenesis therapy

SupplementarySidebar13.8 Kaposi’s sarcoma cells hold the record for the number of documented heterotypic signals they receive

SupplementarySidebar13.9 Effectsofananti-VEGF-Rmonoclonalantibodyonthegrowthofahumantumorxenograft

SupplementarySidebar13.10 Fibroblastsareheterogeneousandcanchangedynamicallyinresponsetosignals

SupplementarySidebar14.1 Visualization of the dynamics of pathfinding fibroblasts followed by squamous cell carcinoma cells

SupplementarySidebar14.2 Metastasizing cancer cells often take on hitchhikers while traveling through the blood

SupplementarySidebar14.3 Instruments for detecting circulating tumor cells

SupplementarySidebar14.4 Hiddenmicrometastasesarerevealedthroughorgantransplantation

SupplementarySidebar14.5 Wolves in sheep’s clothing: when carcinoma cells invade the stroma

SupplementarySidebar14.6 TGF-β works in conflicting ways during tumor progression

SupplementarySidebar14.7 Dynamics of EMT induction: the EMT may be controlled in some cancer cells exclusively by their own genomes

SupplementarySidebar14.8 An example of an EMT relatively late in embryonic development

SupplementarySidebar14.9 Relatively rapid metastatic dissemination of advanced primary tumor cells

SupplementarySidebar14.10 Our cells devote an enormous number of genes to regulating protein degradation

SupplementarySidebar14.11 Peritoneovenousshuntsprovidedramaticsupportfortheseedandsoilhypothesis


SupplementarySidebar14.12 Tooth extractions may occasionally become exceedingly painful

SupplementarySidebar14.13 Tumor stem cells further complicate our understanding of the metastatic process

SupplementarySidebar14.14 DoesDarwinianevolutionaccommodatemetastasis-specificalleles?

SupplementarySidebar15.1 Rearrangements of chromosomal DNA segments generate a vast array of antigen-binding domains in antibodies and T-cell receptors

SupplementarySidebar15.2 Virus-infected cells may not always be recognized by the immune system

SupplementarySidebar15.3 Bizarre tumors reveal how cancer cells can become infectious agents

SupplementarySidebar15.4 Mice have proven to be far more useful for tumor biologists than chickens

SupplementarySidebar15.5 AnHPVvaccineprotectsagainstmanycervicalcarcinomas

SupplementarySidebar15.6 Anunexpectedtypeofanti-p53reactivityisoftenfoundincancerpatients

SupplementarySidebar15.7 Immune recognition of tumors may be delayed until relatively late in tumor progression

SupplementarySidebar15.8 Someparaneoplasticsyndromesrevealdefectivetoleranceandoverlysuccessfulimmuneresponses to tumors

SupplementarySidebar15.9 TSTAscanariseasby-productsofchemicalandphysicalcarcinogenesis

SupplementarySidebar15.10 Aremelanomasmoreantigenicthanothertumors?

SupplementarySidebar15.11 StrategiesforcloninggenesencodingmelanomaTATAs

SupplementarySidebar15.12 Anti-CD47therapiesholdpromiseintreatinglymphomasandotherhematopoieticmalignancies

SupplementarySidebar15.13 Cancer cells may thwart extravasation by circulating T cells

SupplementarySidebar15.14 Herceptincanbemodifiedtopotentiatecancercellkilling

SupplementarySidebar15.15 Bone marrow transplantation and the treatment of hematopoietic malignancies

SupplementarySidebar15.16 Whole genome sequencing allows a new attack on tumor cells

SupplementarySidebar16.1 Moderncancertherapieshavehadonlyaminoreffectontheoveralldeathratefromthedisease

SupplementarySidebar16.2 Prostatecancersusuallydonotrequireaggressiveintervention—ataleoftwocountries

SupplementarySidebar16.3 Clinical practice and our understanding of disease pathogenesis have often been poorly aligned, leading to sub-optimal, often tragic outcomes

SupplementarySidebar16.4 The ability to assign tumors to specific disease subtypes is critical to the success of drug development

SupplementarySidebar16.5 p53 germ-line polymorphisms and somatic mutations can complicate the induction of apoptosis by drugs

SupplementarySidebar16.6 Ras function can be inhibited by interfering with the enzymes responsible for the maturation of the Ras protein

SupplementarySidebar16.7 Chemical synthesis, compound libraries, and high-throughput screening

SupplementarySidebar16.8 Evolution can generate huge collections of structurally similar proteins

SupplementarySidebar16.9 Large-scalescreenoftheinhibitoryeffectsofadrugonvariouskinases

SupplementarySidebar16.10 Epidermal growth factor receptor expression levels predict little about a tumor’s susceptibility to receptor antagonists

SupplementarySidebar16.11 Akt/PKBfunctioniscontrolledbymultipleupstreamsignals


Movies

The“Movies”foldercontainsthirty-onemovies,availableinbothQuickTimeandWMV formats, that will aid in understanding some of the proteins and processes

describedinthebook.TheQuickTimemovieswillprovidetheoptimalviewingexperi-ence.TheWMVformatisincludedbecauseQuickTimemovieswillnotworkinPower-PointforWindows,buttheWMVformattedmovieswillworkseamlessly.Eachmovieis accompanied by a voice-over narration. The movie table of contents, followed by the full text of each movie narration, is below:

Chapter 1: The Biology and Genetics of Cells and Organisms1.1 ReplicationI1.2 ReplicationII1.3 Translation1.4 Transcription

Chapter 2: The Nature of Cancer2.1 EmbryonicOriginsofTissues2.2 Mammary Cancer Cells 2.3 VisualizationofCancerI:Lymphoma

Chapter 3: Tumor Viruses3.1 ContactInhibition

Chapter 4: Cellular OncogenesNo Movies

Chapter 5: Growth Factors, Receptors, and Cancer5.1 CellularEffectsofEGFvs.HGF5.2 ActivationofKitReceptorSignalingbySCF5.3 EGFReceptorFamily5.4 IGFReceptorsandMonoclonalAntibodies

Chapter 6: Cytoplasmic Signaling Circuitry Programs Many of the Traits of Cancer6.1 RegulationofSignalingbytheSrcProtein6.2 SignalingbytheRasProtein6.3 EGFReceptorsandSignaling

Chapter 7: Tumor Suppressor Genes7.1 IntestinalCrypt

Chapter 8: pRb and Control of the Cell Cycle Clock8.1 AnimalCellDivision8.2 CDK2

Chapter 9: p53 and Apoptosis: Master Guardian and Executioner9.1 p53Structure9.2 Apoptosis

Chapter 10: Eternal Life: Cell Immortalization and Tumorigenesis10.1 TelomereReplication

Chapter 11: Multi-Step TumorigenesisNo Movies

Chapter 12: Maintenance of Genomic Integrity and the Development of Cancer12.1 DNARepairMechanisms


Chapter 13: Dialogue Replaces Monologue: Heterotypic Interactions and the Biology of Angiogenesis13.1 MechanismsEnablingAngiogenesis13.2 InteractionsofInnateImmuneCellswithaMammaryTumor

Chapter 14: Moving Out: Invasion and Metastasis14.1 AdhesionJunctions14.2 MechanismsofBrainMetastasisFormation14.3 VisualizationofCancerII:Metastasis

Chapter 15: Crowd Control: Tumor Immunology and Immunotherapy15.1 TheImmuneResponse15.2 AntigenDisplayandT-CellAttack

Chapter 16: The Rational Treatment of Cancer16.1 DrugExportbytheMulti-DrugResistancePump16.2 PI3K


Movie 1.1 Replication I

Using computer animation based on molecular research, we are able to picture how DNA is replicated in living cells. You are looking at an assembly line of amaz-

ing miniature biochemical machines that are pulling apart the DNA double helix and cranking out a copy of each strand. The DNA to be copied enters the production line from bottom left. The whirling blue molecular machine is called a helicase. It spins the DNA as fast as a jet engine as it unwinds the double helix into two strands. One strandiscopiedcontinuouslyandcanbeseenspoolingofftotheright.Thingsarenotso simple for the other strand because it must be copied backwards. It is drawn out repeatedly in loops and copied one section at a time. The end result is two new DNA molecules.

AnimationproducedforDNAInteractive(www.dnai.org)©2003HowardHughesMedicalInstitute,(www.hhmi.org).Allrightsreserved.

Movie 1.2 Replication II

In a replication fork, two DNA polymerases collaborate to copy the leading-strand template and the lagging-strand template DNA. In this picture, the DNA polymer-

ase that produces the lagging strand has just finished an Okazaki fragment.

The clamp that keeps the lower DNA polymerase attached to the lagging strand dis-sociates, and the DNA polymerase temporarily releases the lagging strand template DNA.

As the DNA helicase continues to unwind the parental DNA, the primase becomes activated and synthesizes a short RNA primer on the growing lagging strand. The DNA polymerase binds to the DNA again and becomes locked in by the clamp.

The polymerase uses the RNA primer to begin a short copy of the lagging-strand tem-plate DNA.The polymerase stalls when it reaches the RNA primer of the preceding Okazaki fragment, and the entire cycle repeats.

Animation:Sumanas,Inc.(www.sumanasinc.com)Music: Christopher Thorpe


Movie 1.3 Translation

When the mRNA is complete, it snakes out of the nucleus into the cytosol. Then in a dazzling display of choreography, all the components of a molecular machine

lock together around the RNA to form a miniature factory called a ribosome. It trans-lates the genetic information in the RNA into a string of amino acids that will become a protein. tRNA molecules‚ the green triangles‚ bring each amino acid to the ribosome. TheaminoacidsarethesmallredtipsattachedtothetRNAs.TherearedifferenttRNAsfor each of the twenty amino acids, each of them carrying a three-letter nucleotide code that is matched to the mRNA in the machine. Now we come to the heart of the process. Inside the ribosome, the mRNA is pulled through like a tape. The code for eachaminoacidisreadoff,threelettersatatime,andmatchedtothreecorrespond-ing letters on the tRNAs. When the right tRNA plugs in, the amino acid it carries is added to the growing protein chain. You are watching the process in real time. After a few seconds the assembled protein starts to emerge from the ribosome. Ribosomes can make any kind of protein. It just depends on what genetic message you feed in on the mRNA. In this case, the end product is hemoglobin. The cells in our bone marrow churn out a hundred trillion molecules of it per second! And as a result, our muscles, brain, and all the vital organs in our body receive the oxygen they need.

AnimationproducedforDNAInteractive(www.dnai.org)©2003HowardHughesMedicalInstitute(www.hhmi.org).Allrightsreserved.

Movie 1.4 Transcription

Transcription is the process by which DNA is copied into RNA in the first step of gene expression. It begins with a bundle of factors assembling at the start of a

gene, that is, a linear sequence of DNA instructions, shown here stretching away to the left.TheassembledfactorsincludeanRNApolymerase,thebluemolecule.Suddenly,RNA polymerase is let go, racing along the DNA to read the gene. As it unzips the dou-ble helix, it copies one of the two strands. The yellow chain snaking out of the top is the RNA, a copy of the genetic message. The nucleotide building blocks that are used to make the RNA enter through an intake hole in the polymerase. In the active site of the enzyme, they are then matched to the DNA, nucleotide by nucleotide, to copy the A’s,C’s,T’sandG’softhegene.TheonlydifferenceisthatintheRNAcopy,thymineis replaced with the closely related base uracil, commonly abbreviated “U.” You are watching this process, called transcription, in real time.

AnimationproducedforDNAInteractive(www.dnai.org)©2003HowardHughesMedicalInstitute(www.hhmi.org).Allrightsreserved.


Movie 2.1 Embryonic Origins of Tissues

Cells of this developing frog embryo rearrange in a dramatic ballet of orchestrated cell movements. In a continuous motion, cells from the outer layer of the embryo

sweep towards the vegetal pole and start invaginating, forming a deep cavity in the interior. The paths of the cells and the topology of these rearrangements are best seen inthisanimationofanembryothathasbeenslicedopen.Thedifferentcelllayersthatareformedinthiswayhaveverydifferentfates.Cellsthatlinethenewlyformedcavity,called the endoderm, develop into the lining of the gut and many internal organs such as liver, pancreas, and lung. Cells in the middle layer, called the mesoderm, give rise to muscle and connective tissue. Cells remaining on the outside, called the ectoderm, go on to form the outer layer of the skin, as well as the nervous system.

From“FromEggtoTadpole”JeremyPickett-HeapsandJuliannePickett-HeapsCytographics(www.cytographics.com)

Jeremy Pickett-HeapsUniversity of Melbourne

Movie 2.2 Mammary Cancer Cells

Normal human mammary epithelial cells can be suspended and grown in a gel-like medium. Under these conditions, they form structures that resemble the lit-

tle sacs of cells in the mammary gland, called alveoli, in which milk is made. Cells assemble into a well-organized, polarized epithelium that forms a closed sphere with an internal lumen.

In the mammary gland, this space would be connected to ducts, and cells would secrete milk into it.

By contrast, these human breast cancer cells grown under the same conditions divide aggressively and in an uncontrolled fashion. They are also more migratory and grow into disorganized clumps, which would form tumors in the body.

Originalscript:PeterWalter,HowardHughesMedicalInstitute,UniversityofCaliforniaatSanFrancisco

Mina J. Bissell, Karen Schmeichel, Hong Liu, and Tony Hansen Lawrence Berkeley Laboratories


Movie 2.3 Visualization of Cancer I: Lymphoma

Magnetic resonance imaging tomography enables the visualization of tumors. In this3-Dreconstruction,weseealymphomainlivingtissuethathasbeenartifi-

ciallycoloredred.Thelymphomawasgeneratedbycellsthatinitiallylackedp53func-tion;intheabsenseofp53expression,thelymphomagrewtobequitelarge.

Inthisexperiment,p53expressionwasreactivatedinthelymphomaatDay0.

18daysafterp53reactivation,wecanobservesignificantregressionofthetumor.Thisregressionillustratesthatp53-activatingsignalswerepresentinthelymphomacellsduring its initialgrowth,but failed, in theabsenceofp53expression, tohalt tumorgrowth.Thatis,withoutp53expression,thesesignalsalonecouldnottriggerapoptosisin the tumor cells.

However, once p53 was expressed in the tumor cells, it functioned powerfully toinduce tumor regression, presumably by causing widespread apoptosis in the lym-phoma cells.

Bythe28thdayofp53reactivation,thetumorcannotbeseenatall.

Andrea Ventura, Tyler Jacks, and David G. KirschTheDavidH.KochInstituteforIntegrative Cancer Research at MIT

Jan Grimm and Ralph Weissleder MassachusettsGeneralHospital

3.1 Contact Inhibition

WhennormalcellsareintroducedintoaPetridishatlownumbers,theybegintogrow and divide. As the cells begin to touch one another, they stop dividing. This

response is called contact inhibition. Once the cells fill up the bottom of the dish, all cell division ends, creating a state called confluence.

Contact inhibition ensures that the cells create a layer only one cell thick—a monol-ayer.

Thebehaviorofcancercellsisquitedifferent.Ifacancercellisseededamongnormalcells,allofthecellswillproliferateasbefore.However,onceconfluenceisreached,thenormal cells will stop multiplying while the cancer cells continue to divide, yielding a clump of cells often called a focus.

Contact inhibition can be demonstrated in vitro by removing cells from a confluent monolayer. In this experiment, cells are removed by scratching the monolayer with aneedle.Thesurvivingcellsattheedgeofthewoundnowdotwothings.First,theybegin to proliferate again, since they are no longer fully contact-inhibited. And sec-ond, they migrate into the empty area of the wound, attempting to fill it up.

Animation:Sumanas,Inc.(www.sumanasinc.com)

Sheryl Denker and Diane Barber University of California at SanFrancisco


Movie 5.1 Cellular Effects of EGF vs. HGF

EpidermalGrowthFactor,orEGF,hasbeenaddedtotheseliverepithelialprogeni-torcellsinculture.EGFbindsspecificallytotheEpidermalGrowthFactorRecep-

tor on the surface of the cells. This stimulates its tyrosine-kinase activity and a down-stream signaling cascade that promotes DNA synthesis and cell proliferation. Note how the cells cling to one another and form a clump of cells as they divide.

In thismovie, another growth factor,HepatocyteGrowth Factor, orHGF, has beenadded to liver epithelialprogenitor cells in culture.HGFbinds to thec-Met recep-toronthecellsurface,andsimilartoEGF,itstimulatestyrosine-kinaseactivityandadownstreamsignalingcascadethatpromotescellproliferation.However,unlikeEGF,HGFhasanadditionaleffectoncellsbypromotingcellmotility.Thiscausesthedivid-ingcellstoscatter.Becauseofthiseffect,HGFisalsoknownasScatterFactororSF.

ThedifferencesintheeffectsofEGFandHGFillustratethequitedifferentresponsesthat these two growth factors elicit from cells.

Andrea Bertotti and Paolo M. ComiglioInstitute for Cancer Research and Treatment, University of Torino, SchoolofMedicine

Movie 5.2 Activation of Kit Receptor Signaling by SCF

Intheabsenceofligand,agrowthfactorreceptorexistsinamonomeric(single-sub-unit)formembeddedintheplasmamembrane.Manygrowthfactorreceptors,such

as the Kit receptor shown here, have lateral mobility in the plane of the plasma mem-brane and are relatively free to wander back and forth across the surface of the cell.

Like many growth factor receptors, the Kit protein can be divided into 3 majordomains. The ectodomain, which sticks out of the surface of the cell and plays a key role in ligand binding; the transmembrane domain, which is extremely hydrophobic and spans the plasma membrane; and the cytoplasmic domain, which can initiate signaling inside the cell.

In the example of Kit, the signaling process begins when Kit encounters its ligand, called stem cell factor, or SCF. SCF is composed of two identical protein subunits.WhenpresentedwithSCF,Kitbindstooneofthetwoidenticalsubunits.

When the SCF/Kit complex encounters an unbound Kit monomer in the plasmamembrane,theunboundKitmonomercanbindtotheothersubunitoftheSCFlig-and, and form a dimeric receptor.

UpondimerizationoftheectodomainsdrivenbySCF,thecytoplasmicdomainsofthetwo Kit proteins are brought into close proximity.

The kinase domain of each Kit receptor then phosphorylates tyrosine residues present in the cytoplasmic domain of the other Kit receptor, a process called transphospho-rylation. Transphosphorylation triggers a cascade of signaling activity inside the cell thatcanaffectcellgrowthandproliferation.


Joseph SchlessingerYaleUniversitySchoolofMedicine


Movie 5.3 EGF Receptor Family

Epidermalgrowthfactorreceptors,alsocalledEGFreceptors,constituteafamilyoffour similarly structured receptor tyrosine kinases that interact with one another

to promote general cell growth and proliferation.

WhenEGFreceptorsbindwiththeirligands—suchasEGF,TGF-α,orNRG—theyformhomodimers or heterodimers, and activate their cytoplasmic domains. Once acti-vated, the cytoplasmic domains emit signals inside the cell that activate cytoplasmic signaling cascades that function, in turn, to promote cell growth and proliferation, inhibit apoptosis, and increase cell motility.

Incancercells,signalemissionbyEGFreceptorsbecomesderegulatedandenablesthem to promote tumor growth, resistance to chemotherapy, and tumor metastasis.

Intheabsenceofligand,normalEGFreceptorsexistinamonomericform,andmovelaterally in the plane of the plasma membrane. Even if they collide, they cannot form stable dimers.

WhenanEGFreceptorbindswithitsligand,itundergoesadramaticconformationalchange in its extracellular domain. This enables the altered extracellular domain of this receptor to bind to the extracellular domain of a second receptor molecule.

Dimerization brings the intracellular kinase domains of the two receptor molecules close together. The tyrosine kinases become activated, and each tyrosine kinase phos-phorylates the cytoplasmic tail of the other receptor molecule. This process is referred to as transphosphorylation.

The resulting phosphorylated tails then serve to recruit and activate a series of signal transducing proteins that in turn activate multiple downstream signaling cascades.


Mark SliwkowskiGenentech,Inc.


Movie 5.4 IGF Receptors and Monoclonal Antibodies

The surface of all cells is studded with a wide variety of receptor tyrosine kinases, amongthemthereceptorsforinsulinandinsulin-likegrowthfactors,calledIGF-1

andIGF-2.UponbindingtheirinsulinorIGFligands,thesetworeceptorstransducesignals into the cell, including those that encourage cell growth and prevent apoptosis.

ThekeymoleculesinvolvedintheIGFsystemare:thesolubleligandsIGF1andIGF2,aseriesofIGF-bindingproteins,thecellsurfaceIGFreceptors,type1andtype2,andthe insulin receptor.

Bothoftheligands,IGF1andIGF2,exerttheirbiologicalactivitiesthroughbindingtotype-1IGFreceptors.Theycanalsobindtotheinsulinreceptor,butwithabout1000times lower affinity.

Thetype-2 receptor isbelieved to functionasadecoy receptor, todraw IGF2awayfromthetype-1receptor.Itisconsideredtoactasatumorsuppressor,anditsexpres-sion is frequently lost during tumor progression.

Inaddition,aseriesofsixbindingproteinsassociatewithIGFsinplasmatoregulatetheamountoffreeIGFsthatareavailabletobindandactivatethetype-1IGFreceptor.

The type-1 IGF receptor tyrosinekinasecanbindeither IGF1or IGF2 ligands.Thisreceptor exists on the cell surface as a pre-formed dimer. It is composed of an extra-cellularalphasubunit,whichcontainstheIGFbindingsite,andabetasubunit,whichextends through the lipid bilayer of the cell membrane to the inside of the cell, where the tyrosine kinase enzyme is located.

The insulin receptor is structurally and functionally very similar to the type-1 IGFreceptor.Asaresult, the type-1 IGFreceptorand insulinreceptorcan formhybrid,heterodimericreceptorscontaininganIGFreceptorsubunitandaninsulinreceptorsubunit. Studies show that thesehybrid receptorspreferentially act like type-1 IGFreceptorsandbindIGFswithhighaffinity.

WhenIGFsbind,thereceptortyrosinekinaseactivityonthebetasubunitisactivated,causing tyrosine residues to become phosphorylated. As is the case with all other tyro-sine kinase receptors, these phosphotyrosine residues then recruit signaling proteins toturnonavarietyofdownstreamsignalingpathways,suchastheMAPkinasepro-liferationpathwayand theAKTsurvivalpathway.Therefore,bindingof IGFs to thetype1receptorprovidessignalstotumorcellsthatmayconferresistancetokillingbycytotoxic drugs.

Blockageofbindingtothetype-1receptorrepresentsonetherapeuticstrategytostopcellproliferationandaffecttumorcellsurvival.

Anti-type-1IGFreceptormonoclonalantibodiescanbegeneratedthatarehighlyspe-cificforbindingtothealphasubunitofthetype-1-IGFreceptor,therebyblockingitsfunction.

Monoclonal antibodies have been developed that have high selectivity and affinity forthetype-1IGFreceptor,anddonotrecognizeorbindtoinsulinreceptors.Duetotheselectivityforthetype-1IGFreceptors,theycanalsorecognizeandblockhybridreceptorsformedbetweentype-1IGFreceptorsandinsulinreceptors.TreatmentoftumorcellswithsuchanantibodycreatesadirectblockadeofIGFbindingtotype-1receptorsandinhibitsdownstreamsignaling.Italsocausestype-1receptormoleculesto be removed from the surface of tumor cells. Ultimately, it is hoped that clinical use ofIGFreceptor-blockingantibodiescanreducetumorcellgrowthandincreaseapop-tosis in tumors.

ImCloneSystems


Movie 6.1 Regulation of Signaling by the Src Protein

TheSrcproteinkinaseactsasamolecularsignalintegrationdevice.Srcconsistsoftwoclearlydemarcatedregulatorydomains:anSH3domainandanSH2domain.

The catalytic domain is structured much like other tyrosine kinases. In addition, a linker and the short C-terminal tail play regulatory roles.

To be an active protein kinase, the active site of the Src kinase domainmust firstbecome accessible to substrate proteins. In the inactive state, the active site is blocked by an activation loop.

The activation loop must be phosphorylated, most likely through the actions of anotherSRCkinasemolecule,notshownhere.Phosphorylationcausestheactivationloop to swing out of the way and rearrange so that other substrate proteins can bind and become phosphorylated.

SrciskeptinactivebyaninteractionofitsSH2domainwiththeC-terminaltailpeptide.A phosphorylated tyrosine on its C-terminal tail is buried in a deep binding pocket in theSH2domain.ToactivateSrc,thisphosphategroupisremovedbyspecificphos-phatases. Upon dephosphorylation of the tyrosine, the C-terminal tail is released from theSH2domain.

Inthiscartoonview,Srcisrepresentedinitsinactivestate.Thelinkersegment,showninred,connectstheSH2andSH3domainsontheonehand,andthekinasedomainon the other. The activation loop, shown in dark green, drapes across the catalytic sites and, in this configuration, blocks the catalytic domains access to the substrate of the kinase.

SeveralotherintramolecularinteractionsholdtheSrckinaseinaninactiveconfigura-tion.TheSH2domain,inblue,bindsaphosphotyrosineatresidueposition527.TheSH3domain,inlightgreen,bindsthelinkersegment.

When a PDGF receptor becomes activated by ligand binding, its C-terminal tailbecomes phosphorylated on tyrosine residues. One of these phosphotryrosines can berecognizedandboundbytheSH2domainofSrc.

TheSH2domain,whichuntilnowparticipatedinanintramolecularbinding,switchesand forms an intermolecular bridge by binding the phosphotryrosine on the C-termi-naltailofthePDGFreceptor.

TheSH3domainfollowssuit.Ittoobreaksitsintramolecularbindingandbindstoaproline-richsegmentonthePDGFreceptor.

TheseshiftsliberatethekinasedomainoftheSrcprotein.Inaseriesofconcertedreac-tions,thetyrosineresidueatposition527ofSrcisdephosphorylated,andtheactiva-tion loop becomes phosphorylated on a tyrosine residue.

This last alteration forces the activation loop to move out of the way so that it no longer obstructs the catalytic cleft of the Src enzyme.This allows Src to phosphorylate adiverse array of protein substrates on their tyrosine residues.

OriginalStoryboardbyPeterWalter,HowardHughesMedicalInstitute,UniversityofCaliforniaatSanFranciscoAnimation:Sumanas,Inc.(www.sumanasinc.com)


Movie 6.2 Signaling by the Ras Protein

TheRasproteinisarepresentativeexampleofthelargefamilyofGTPasesthatcanfunction as molecular switches. The nucleotide-binding site of Ras is formed by

several conserved protein loops that cluster at one end of the protein. In its inactive state,RasisboundtightlytoGDP.

Asamolecular switch,Ras can togglebetween twodifferent conformational statesdependingonwhetherGDPorGTPisbound.Tworegions,calledswitch1andswitch2, change conformation dramatically. The change in conformational state allows other proteinstodistinguishactiveRasfrominactiveRas.Active,GTP-boundRasbindsto,andactivates,downstreamtargetproteinsinthecellsignalingpathways,suchasRAF,PI3K,andGDS.

Aspace-fillingmodelshowsthattheconformationalchangesbetweentheGDPandGTPboundformsofRasspreadoverthewholesurfaceoftheprotein.Thetwoswitchregions move the most.

RashydrolyzesGTPtoswitchitselfoff;thatis,toconvertfromtheGTP-boundstatetotheGDP-boundstate.ThishydrolysisreactionrequirestheactionofaRasGTPase-activatingprotein,orRasGAPforshort.RasGAPbindstightlytoRasburyingtheboundGTP.It insertsanargininesidechain,sometimescalledanArgeninefinger,directlyinto the active site. The arginine, together with threonine and glutamine side chains ofRasitself,promotesthehydrolysisofGTP.RasthensitsinitsinactiveGDP-boundstate,awaitingastimulatorysignalthatwillcauseittoevictGDPandacquireGTP.

Molecular modeling and animation: Timothy Driscoll Originalscript:PeterWalter,HowardHughesMedicalInstitute,UniversityofCaliforniaatSanFranciscoAnimation:Sumanas,Inc.(www.sumanasinc.com)


Movie 6.3 EGF Receptors and Signaling

Whenepidermalgrowthfactor(orEGF)bindstoreceptorsontheplasmamem-brane,thereceptormoleculesstructurallyrearrangeanddimerize.EGFrecep-

tors are in a class of receptor tyrosine kinases, and the dimerization of these receptors results in the reciprocal activation of the two kinase domains. Once a tyrosine kinase is activated, it phosphorylates the C-terminal cytoplasmic tail of the other receptor at multipletyrosineresidues.Formanyreceptors,thistransphosphorylationisrecipro-cal, with each receptor molecule phosphorylating the other. As we will see, the phos-phate groups decorating the c-terminal tail enable the receptor to activate a series of downstream signaling events.

Through a series of intermediary signaling proteins, the activated receptor turns on aproteincalledRas.Rasactslikeabinaryswitch,andEGFsignalingtriggersRastoconvertfromitsGDP-boundinactivestatetoitsGTP-bound,activelysignalingstate.Ifmutated into a constitutively active form, Ras becomes oncogenic.

IthaslongbeenknownthattheactivationoftheEGFreceptorhasverysimilareffectson the cell as the activation of a Ras protein. This suggested the possibility of signaling between them.

ThediscoveryofSH2domainsprovided importantcluesofhow this signalingpro-ceeds. These domains, which are parts of larger proteins, enable proteins to recognize and bind phosphotryrosine residues displayed by yet other proteins.

TheSH2groupallowsrecognitionoftwochemicalstructures:first,thephosphotyro-sine itself; and second, the adjacently located amino acid residues. There are many phosphotyrosines attached to an activated growth factor receptor, and each of these receives its unique identity from its neighboring amino acid residues.

Such phosphate groups decorate the C-terminal tail of a growth factor receptor.Accordingly,amoleculelikeShc,whichhasanSH2group,canrecognizeandbindaphosphotyrosinegroupontheC-terminaltailoftheEGFreceptor.InthecaseofShc,the phosphotyrosine is followed by three other amino acids, yielding the sequence: tyrosine(Y),leucine(L),isoleucine(I),andproline(P).

LikeShc,thisthemerecursinanumberofotherproteins,whichalsocontaintheirownspecializedSH2domains.Eachrecognizesaparticularphosphotyrosinefollowedbythree distinct amino acid residues.

BecausecertainphosphotyrosinesattractmultipleSH2groups,theentirearrayofSH2containing proteins that can become associated with a receptor is quite elaborate and can be depicted like this.

Each of these associated proteins has its own distinct biochemical function, which is carriedoutbyotherdomainslinkedtoitsSH2domain.Forexample,JAK2isatyro-sinekinasethatbecomesactivatedafteritbindstothereceptortailviaitsSH2group.PLC-gisanenzymethatcleavesphospholipidsintheplasmamembrane.AndGrb2isa member of an important class of adaptor proteins whose sole function is to build a bridge between the receptor and a third protein.

FocusingonGrb2,wecanseethatSH2groupsarenottheonlywaybywhichproteinscanrecognizeandbindtooneanother.Forexample,SH3groupsrecognizeandbindproline-richsequencesonotherproteins.InthecaseofGrb2,itstwoSH3groupsrec-ognizeathirdproteincalledSOS.OnceSOSbecomesboundindirectlytothereceptor,itisbroughtincloseproximitytoRasproteinstetheredtotheplasmamembrane.SOSactsasaguaninenucleotideexchangefactor.SuchproteinsarecapableofinducingRAStoreleaseitsboundGDP,allowingGTPtojumpaboard.

ThesetwoformsofRasrepresentdifferentstatesofactivity.TheGDP-boundformisinactiveandtheGTP-boundformisactiveinsignaling.Theperiodofactivesignal-ing is quite short since other proteins interact with Ras and induce it to hydrolize its boundGTP.


While Ras is in its active state, it is able to interact with a variety of partner proteins that serve,inturn,toactivatedownstreamsignalingpathways.Forexample,Rascanbindto the serine threonine kinase called Raf. When Raf interacts with Ras, it becomes an active signal-emitting kinase.

An immediate downstream target of Raf is called MEK. Once MEK becomes activated through phosphorylation, it also becomes an active serine threonine kinase. Its most important substrate is called Erk. Once activated, Erk proceeds to phosphorylate a diverse group of proteins that favor cell proliferation.

Another important event initiated by Ras is the PI3 kinase cascade.Once again, akinase is being activated, but in this case the object of phosphorylation is not a pro-tein, but instead a phospholipid embedded in the plasma membrane.

The phospholipid molecule is constructed from two long hydrocarbon tails embedded in the plasma membrane, a glycerol, and an inositol sugar-like molecule. In the case of PI3K,itsimmediatesubstrateisaphosphotidylinositoltowhichtwophosphateshavebeenpreviouslyattached.PI3Kextendsthisprocessbyaddingyetanotherphosphateto the inositol ring.

This now creates a site to which a number of cytoplasmic proteins can attach. These proteins use another specialized domain to recognize and bind the triply phosphor-ylatedinositol.ThedomainiscalledaPHdomain.Oneofthemostimportantmol-eculesthatcontainsaPHdomainisakinasecalledAktorPkB.

Once Akt binds, it becomes phosphorylated and then becomes functionally activated. Once activated, Akt is released and can travel through the cytoplasm to activate mul-tiple substrates.

The ability of Akt to phosphorylate these other proteins, thereby altering them, has multipleeffectsonthecell.Forexample,byphosphorylatingBad,itreducesthelikeli-hood of apoptosis. By phosphorylating mTOR, it facilitates the physical growth of the cell.AndbyphosphorylatingGSK-3beta,itstimulatescellproliferation.

StoryboardandAnimation:Sumanas,Inc.(www.sumanasinc.com)


Movie 7.1 Intestinal Crypt

The lining of the small intestine, like the lining of most of the gut, is a single-layered epithelium.

Depending upon location in the gut, the epithelial cells, usually called enterocytes, help absorb nutrients from the lumen of the gut or absorb water from the intestinal contents. In the small intestine, the surface of the lining of the gut is increased enor-mously by thousands of villi that protrude into the lumen.

Hereweseeasinglevillusanditsinternalarchitecture.Thesurfaceofthevillusiscov-ered by a single layer of enterocytes, which extend down into the crypt below.

Stemcellsatthebottomofthecrypt,locatedbetweenpanethcells,divideandmakecopies of themselves, and also make transit-amplifying cells. The transit amplify-ing cells proliferate rapidly and move up the walls of the crypt. As the cells migrate upwardstheybegintodifferentiateintogobletcellsandenterocytes.Whilethecellsare moving up the sides of the villus, they carry out the essential functions of the small intestine, notably absorption of nutrients.

Whenthedifferentiatedcellsreachthetipofthevillus,theyundergoapoptosisandare shed into the lumen of the small intestine. The entire process of out-migration and cell death is completed in just three to four days.

The process of out-migration and rapid cell replacement is a defense mechanism against the development of colon cancer, since almost all epithelial cells, including those that have accidentally sustained mutations, are shed within days of their forma-tion. Therefore, the only mutations that can lead to the development of a cancer are those that are retained in the crypt. This dictates that such mutations must block the outmigration of mutant cells from the crypt.

The outmigration of transit amplifying cells from the bottom of the crypt depends on theproteincalledadenomatouspolyposiscoli,orsimplyAPC.Intheabsenceoffunc-tionalAPC,thiscontinuousoutmigrationisblocked,leadingtotheaccumulationoftransit amplifying cells in the crypt.

In thisanimationofanexperiment,APC loss isachieved throughan inducedgeneinactivation, which appears to mimic the mutation that initiates most gastrointestinal tumors.

APC is usually required to inactivate the intracellular protein calledβ-catenin; the inactivation of β-catenin permits the differentiation of the transit amplifying cellsandtheircontinuedoutmigrationfromthebaseofthecrypt.IntheabsenceofAPCfunction, β-catenin accumulates within the transit amplifying cells, which blocks both theiroutmigrationanddifferentiation.

Theaccumulatedtransitamplifyingcellsdonotthemselvesformacarcinoma.How-ever, they and their descendants can now accumulate additional mutations that will drive such cells progressively to become full-fledged carcinoma cells.

AnimationbyDigizyme,Inc.(www.digizyme.com)Models,Animation,Surfacing,Composite:EricKellerStoryboardandArtDirection:GaelMcGill©2009byHansClevers

Hans CleversHubrechtInstitute


Movie 8.1 Animal Cell Division

D ifferential interference contrast microscopy is used here to visualize mitoticevents in a lung cell grown in tissue culture. Individual chromosomes become

visible as the replicated chromatin starts to condense. The two chromatids in each chromosome remain paired as the chromosomes become aligned on the metaphase plate. The chromatids then separate and get pulled by the mitotic spindle into the two nascent daughter cells. The chromatin decondenses as the two new nuclei form and cytokinesis continues to constrict the remaining cytoplasmic bridge until the two daughter cells become separated.

Video reproduced from: The Journal of Cell Biology122:859–875,1993.©TheRockefellerUniversityPress

Edward D. (Ted) Salmon and Victoria SkeenUniversity of North Carolina at ChapelHill

Robert SkibbensLehigh University

Movie 8.2 CDK2

Like other kinases, Cyclin-dependent kinases, or Cdks for short, are crucial regula-tory proteins in the cell cycle. When activated, Cdks transfer phosphate groups

fromATPtoserineandthreoninesidechainsontargetedsubstrateproteins.Wheninactive, the active site of Cdks is sterically obstructed by a loop, often referred to as the activation loop.

As their name suggests, cyclin-dependent kinases are activated by cyclins. Cyclin binding to Cdk pulls the activation loop away from the active site and exposes the boundATP,allowingitaccesstotargetproteins.Thus,aCdkcanphosphorylatetargetproteinsonlywhenitisinacyclin–Cdkcomplex.

A third protein called a Cdk-activating kinase is required for full activation of a Cdk. This activating kinase adds a phosphate group to a crucial threonine in the activation loop,therebycompletingtheactivationofthecyclin–CDKcomplex.

Oligopeptidedomainsofsubstrateproteinsbindtotheactivesiteofthecyclin–Cdkcomplex so that the target serine or threonine side chains of the substrate are precisely positionedwithrespecttothegammaphosphateoftheboundATP.

Cdkinhibitorproteins,orCKIs,helpregulatetheriseandfallofcyclin–Cdkactivity.Someinhibitors—liketheoneshownhere—binddirectlyatthekinaseactivesiteandblockkinaseactivitybyinterferingwithATPbinding.Otherinhibitors—whicharenotshown here—bind near the active site and interfere with substrate binding.

Molecular modeling and animation: Timothy DriscollOriginalscript:PeterWalter,HowardHughesMedicalInstitute,UniversityofCaliforniaatSanFrancisco


Movie 9.1 p53 Structurep53isatumorsuppressorproteinthatpreventscellsfromdividinginappropriately.

Loss of p53 function is associated with many forms of cancer. In this image, p53polypeptideisshownbindingtoDNA,reflectingp53’sabilitytoactasatranscriptionfactor. In fact,whilenotshownhere,p53 isnormallya tetramerandacts to induceexpressionofsomegenesandrepressexpressionofothers.p53hasabetabarreladja-cent to its DNA-binding domain.

Thep53–DNAinterfaceiscomplex.Itinvolvesseveralloopsandahelixthatextendsfrom the β barrel core. Residues from one loop and the helix bind in the DNA major groove.Arginine248fromanotherloopmakesextensivecontactswiththeDNAback-bone and, indirectly through water molecules, with bases in the minor groove. Altera-tionsinarginine248andotherresiduesinvolvedinDNAbindingarecommonlyfoundinthep53proteinspresentinhumantumors.

Loop 2 does not bind to DNA directly but is essential for correctly positioning arginine 248ontheDNA.Threecysteinesandahistidinefrombothloop2andloop3cooperateto sequester a zinc ion, forming the rigid heart of a zinc-finger motif. Mutations that affectthezincfingerrepresentotherexamplesofhowDNAbindingbyp53iscompro-mised in human tumor cells.

Molecular modeling and animation: Timothy Driscoll Originalscript:PeterWalter,HowardHughesMedicalInstitute,UniversityofCaliforniaatSanFrancisco


9.2 Apoptosis

Apoptosis, a form of programmed cell death, has been induced in these cultured cells. Cell death is characterized by blebbing of the plasma membrane and frag-

mentationof thenuclei.Suddenly,cellsweakenattachment to thesubstratumthatthey have been growing on and shrivel up without lysing.

In the following movies we observe the process at higher magnification. The mecha-nism of apoptosis involves many tightly controlled steps, three of which are demon-stratedherebydifferentvisualizationtechniques.

One initial event is the sudden release of cytochrome c from mitochondria into the cytosol. This event has been visualized here using fluorescently labeled cytochrome c. Initially the greenish/yellow staining is restricted to a reticular pattern, which then suddenly disperses as the mitochondria release their content proteins into the cytosol.

At a later step, the lipid asymmetry of the plasma membrane breaks down. In normal cells, phosphatidyl serine is found only on the cytosolic side of the plasma membrane; but when cells undergo apoptosis, it becomes exposed on the outside of the cell. This event has been visualized here by adding a red fluorescent protein to the media, which specifically binds phosphatidyl serine head groups as they become exposed. In an intact organism, exposure of phosphatidyl serine on the cell surface labels the dead cell and its remnants so that they are rapidly consumed by other cells, such as mac-rophages.

Finally—althoughapoptosingcellsdon’tlyse—theirplasmamembranesdobecomepermeable to small molecules. This event has been visualized here by adding a dye to the media that fluoresces blue when it can enter cells and bind to DNA.

All three of these events can be observed in the same group of cells.

These epithelial cells express green fluorescent cadherin. They are grown at low den-sity,sothatisolatedcellscanbeobserved.Initially,labeledcadherinisdiffuselydis-tributed over the whole cell surface.

As cells crawl around and touch each other, cadherin becomes concentrated as it forms the adhesion junctions that link adjacent cells.

Eventually, as the cell density increases further, the cells become completely sur-rounded by neighbors and form a tightly packed sheet of epithelial cells.

PartI:©2007TheSakuraMotionPictureCompany.Allrightsreserved.Usedwithpermis-sion.

PartII:©2001J.C.GoldsteinandD.R.Green.Allrightsreserved.Usedwithpermission.

PartI:Shigekazu NagataKyoto University

PartII:Joshua C. GoldsteinTheGenomicsInstituteoftheNovartisResearchFoundation

Douglas R. GreenSt.JudeChildren’sResearchInstitute


Movie 10.1 Telomere Replication

The ends of linear chromosomes pose unique problems during DNA replication. BecauseDNApolymerasescanonlyelongatefromafree3ʹ hydroxyl group, the

replication machinery builds the lagging strand by a backstitching mechanism. RNA primersprovide3ʹ-hydroxyl groups at regular intervals along the lagging strand tem-plate.

Whereastheleadingstrandelongatescontinuouslyinthe5ʹ-to-3ʹ direction all the way to the end of the template, the lagging strand stops short of the end.

Even if a final RNA primer were built at the very end of the chromosome, the lagging strand would not be complete.

Thefinalprimerwouldprovidea3ʹ-OHgroup to synthesizeDNA,but theprimerswouldlaterneedtoberemoved.The3ʹ-hydroxyl groups on adjacent DNA fragments providestartingplacesforreplacingtheRNAwithDNA.However,attheendofthechromosomethereisno3ʹ-OHgroupavailabletoprimeDNAsynthesis.

Because of this inability to replicate the ends, chromosomes would progressively shorten during each replication cycle. This “end-replication” problem is solved by theenzymetelomerase.TheendsofchromosomescontainaG-richseriesofrepeatscalled a telomere. Telomerase recognizes the tip of an existing repeat sequence. Using an RNA template within the enzyme, telomerase elongates the parental strand in the 5ʹ-to-3ʹ direction, and adds additional repeats as it moves down the parental strand.

The lagging strand is then completed by DNA polymerase alpha, which carries a DNA primase as one of its subunits. In this way, the original information at the ends of lin-ear chromosomes is completely copied in the new DNA.



Movie 12.1 DNA Repair Mechanisms

Faults in DNA molecules can be harmful; so all organisms have DNA repair mechanismsthatcancorrectmostof them.Mutations(that is,changes inDNA

sequence)canappear inDNAeitherasmistakes inDNAreplication,which resultsin a mismatched nucleotide pair, or by the chemical or physical action of a mutagen. Other agents, known as clastogens, can cause breaks in DNA molecules that may, in turn, lead to mutations.

MostcellspossessfourdifferenttypesofDNArepairsystems.Directrepairandexci-sion repair systems fix DNA molecules carrying nucleotides damaged by mutagens. Mismatch repair corrects mismatched, but otherwise normal, nucleotides that result fromerrorsinreplication.Nonhomologousend-joining(NHEJ)isusedtomenddou-ble-strand breaks in DNA. There is also a fifth repair system called homology-directed repair, but it will not be covered here.

Direct repair is the simplest repair process and acts directly on damaged nucleotides, converting them back to their original structure. Only a few types of mutagen damage can be repaired directly. In this example, guanine acquires an alkyl group, which dis-rupts its normal pairing with cytosine.

In a future round of DNA replication, the alkylated guanine would pair with thym-ine, thereby propagating a DNA mutation. To fix this chemical damage, an enzyme transfers the alkyl group to itself, restoring the base to normal. Many organisms have enzymes that perform this function, such as the Ada enzyme in E.coliandMGMTinhumans.

In contrast to this simple, single-step repair, most types of DNA damage require sev-eral steps to fix. They can be repaired only by removing the damaged nucleotides and then filling in the gap. This process is called excision repair. Base excision repair is a type of excision repair mechanism and is used to repair relatively minor damage. The damaged base is excised from a nucleotide by a specific DNA glycosylase, which first flips the damaged base out of the helix and then cuts the β-N-glycosidic bond between thebaseandsugar,leavingabaselesssite,calledanAPsite.AnAPendonucleasecutsthephosphodiesterbondonthe5ʹsideoftheAPsite,andthentheother3ʹ side is cut, either by the endonuclease itself or by a phosphodiesterase. The resulting gap is filled in by a DNA polymerase and sealed by a DNA ligase.

To correct more extensive types of damage, such as those that cause helix distor-tions, cells use another type of excision repair, called nucleotide excision repair. The processisdifferentfrombaseexcisionrepairinthatitbeginswiththeremovalofanentire block of nucleotides rather than a base from a single nucleotide. A characteristic example is the short-patch process used by E.coli.

A complex of proteins, called the UvrAB trimer, scans DNA for damage. UvrA dissoci-ates once the site has been found and plays no further part in the repair process. UvrC now binds, forming an UvrBC dimer that cuts the polynucleotide on either side of the damagedsite,resultingina12-nucleotideexcision.UvrBcutsaphosphodiesterbond(often the5thdownstreamof thedamaged site) and thenUvrCcuts anotherbond(oftenthe8thupstreamofthedamagedsite).TheexcisedsegmentisthenremovedbyDNA helicase II, which breaks the hydrogen bonding holding the damaged segment in place. UvrC also detaches at this stage, but UvrB remains in place and bridges the gap produced by the excision. The gap is filled by DNA polymerase I. DNA ligase forms the last phosphodiester bond. In humans, a similar but more complex repair machinery fixes this type of damage.

The direct, base excision, and nucleotide excision repair mechanisms all act by searching for and correcting abnormal chemical structures inDNA.However, theycannot correct mismatches resulting from errors in DNA replication, because the mis-matched nucleotides are not abnormal in any way. A special type of excision repair, called mismatch repair, is used instead. Rather than detect an abnormal nucleotide, the mismatch repair system detects the absence of proper base pairing between the


parent and daughter strands. Before the mismatch repair system can excise part of the daughter strand and fill in the gap, it must first distinguish between the parent and daughter strands.

In E. coli, the twostrandsaredistinguishedby theirdiffering levelsofmethylation.The newly synthesized daughter strand is not methylated, but the older parent strand ismethylatedatGATC,CCAGG,andCCTGGsequences.Inthelong-patchmismatchrepair system of E. coli, a protein called MutS recognizes the mismatched nucle-otideswhileanotherprotein,calledMutH,distinguishesthetwostrandsbybindingto nearby unmethylated5ʹ-GATC-3ʹ sequences—that is, by binding to the daughter strand.MutHcutstheunmethylatedDNAatthemethylationsite.Startingatthiscutsite, a DNA helicase detaches a segment of the single strand. The detached single-stranded region is degraded by an exonuclease that follows the helicase and continues beyond the mismatch site. The gap is then filled in by DNA polymerase I and DNA ligase.Notethateukaryotesdon’thaveheavilymethylatedDNAandlikelyuseadiffer-ent mechanism to distinguish between the parent and daughter strands.

Adouble-strandedbreakinDNAispotentiallydevastatingtoacell.Suchbreakscanbe generated by exposure to ionizing radiation and some mutagens, and occasionally during DNA replication. They are repaired either by a process called nonhomologous end joining or homologous recombination, not illustrated here. In nonhomologous end-joining, a pair of proteins called Ku bind to broken DNA ends. The individual Ku proteins also have an affinity for one another, which brings them and the two bro-ken ends of the DNA molecule into proximity. The DNA fragments are joined back together by a DNA ligase.

Experimental studies indicate that restoring normal DNA structure is difficult to achieve. Often, the nucleotides flanking the double-stranded break are lost, resulting in loss of normal DNA sequence. Alternatively, if two chromosomes happen to be bro-ken,amisrepairresultinginhybridstructuresoccursrelativelyfrequently.Hence,thistype of repair is often prone to errors, in contrast to base excision repair, or mismatch repair, which restore normal DNA sequence with far greater fidelity.



13.1 Mechanisms Enabling Angiogenesis

As a normal part of growth and development, the body must generate new blood vessels to oxygenate the tissues. In a process called angiogenesis, new vessels

sprout from existing ones. In this movie, we see endothelial cells sprouting to form new branches from the aorta of a zebrafish embryo. Each sprout is initially formed by one or a few endothelial cells.

Angiogenesiscanbeinitiatedinanumberofdifferentways,suchaswhencellsdonotreceive enough oxygen, often due to inadequate access to the circulation. In response, the cells emit molecules that stimulate angiogenesis, notably vascular endothelial growthfactor,orVEGF.

When VEGFmolecules reach a nearby blood vessel, the vessel’s endothelial cellsbecome motile and form a “tip cell” that produces long extensions called filopodia, whichguidethedevelopmentofanewvesseltowardthecellsemittingVEGF.

As the tip cell moves toward its angiogenic stimulus, other endothelial cells behind it form a stalk. These stalk cells then start to hollow-out and form a tube.

Whentipcellsemergingfromdifferentbloodvesselsmeet,theymerge,andbloodcannow begin to flow through the new vessel. As the young vessel matures, its endothelial cells recruit pericytes, which closely resemble smooth muscle cells, to the vessel walls. The pericytes help to stabilize vessel structure.

Angiogenesis is also critical to the development andproliferation of tumors. Fromhypoxiaortheactionsofoncogenes,tumorcellsoftenemitelevatedlevelsofVEGF.

Increased VEGF causes nearby blood vessels to produce excessive numbers of tipcells, often resulting in malformed capillaries. The capillaries found in tumors are ten times more permeable than normal capillaries, a condition directly attributable to the excessivelevelsofVEGF.

Because blood cannot flow efficiently through malformed vessels, the tumor cells may remainhypoxicandcontinuetoemitVEGF,perpetuatingtheformationofadditionalmalformed vessels.

Significantly, it is believed that the poorly assembled vessel walls permit escapedtumor cells to enter the bloodstream more readily, where they can travel to other loca-tions in the body and start secondary tumors.

Peter CarmelietVesalius Research Center, Catholic University of Leuven


13.2 Interactions of Innate Immune Cells with a Mammary Tumor

This time-lapse microscopic video of the tissue of a live mouse shows cells of the immune system patrolling groups of mammary carcinoma cells. This tumor was

caused by the mouse mammary tumor virus polyoma middle T transgene. The tumor cells are stained in blue, while cells that have properties of macrophages and den-dritic cells are labeled in green. Macrophages are known to play diverse, if not oppos-ing,rolesintumorpathogenesis.Oneclassofmacrophages,calledM1macrophages,facilitate the attack on the tumor by the immune system. A second major class of mac-rophages, often called M2 macrophages, facilitate inflammatory responses and can actually accelerate tumorigenesis and malignant progression. As is apparent here, the cells of the immune system actively patrol the clusters of mammary tumor cells, ostensibly to acquire antigens for presentation to the immune system, and perhaps to releasecytokinesthatmayaffectsubsequentresponsesoftheimmunesystemtothepresence of the tumor.

Mikaela Egeblad and Zena Werb University of California, SanFrancisco

Movie 14.1 Adhesion Junctions

These epithelial cells express green fluorescent cadherin. They are grown at low density, so that isolated cells can be observed.

Initially,labeledcadherinisdiffuselydistributedoverthewholecellsurface.Ascellscrawl around and touch each other, cadherin becomes concentrated as it forms the adhesion junctions that link adjacent cells.

Eventually, as the cell density increases further, the cells become completely sur-rounded by neighbors and form a tightly packed sheet of epithelial cells.

Music: Christopher Thorpe Stephen J. SmithStanfordUniversitySchoolofMedicine

Cynthia AdamsFinchUniversityofHealthSciencesandChicagoMedicalSchool

Yih-Tai ChenCellomics, Inc.

W. James NelsonStanfordUniversitySchoolofMedicine


Movie 14.2 Mechanisms of Brain Metastasis Formation

Metastasis, or the formation of malignant growth at a site of the body far removed fromthelocationoftheprimarytumor, isresponsiblefor90%ofdeathsfrom

cancer. In this image we see multiple breast cancer metastases to the brain.

Multiphoton laser scanning microscopy, together with image processing software, can be used to image metastases to the brain of a live mouse. A transparent window permits visualization of the various steps of the brain metastatic cascade that we will observe in this movie.

Thebrainmetastaticcascadecanbebrokendownintofoursteps:(1)Initialarrestofcirculatingtumorcells innarrowmicrovessels;(2)escapeofarrestedcells fromthemicrovessel into thebrainparenchyma, theprocessof extravasation; (3) theestab-lishment of the recently extravasated cancer cell in a location close to a blood ves-sel,referredtoastheperivascularposition;(4)andfinally,twoalternativeroutesofmetastasis formation—either cooptive growth, in which cancer cells grow along exist-ing blood vessels, or angiogenic growth, in which cancer cells actively induce the for-mation of new blood vessels to nourish them. In the examples that follow, melanoma cells took on cooptive growth, whereas lung cancer cells were found to take on ang-iogenic growth.

Hereweseethefirststepinvolvingtheinitialarrestofacancercell,labeledred,inacapillary, labeled green. A second cell is later seen trapped above in an even smaller vessel.

The second step in the brain metastatic cascade is the extravasation of cancer cells from the lumen of the capillary. Cancer cells use several alternative strategies to extravasate. In this movie, we see a cluster of cancer cells extravasating, one after the other, out of vessel number one into the brain parenchyma.

The third step involves the establishment of perivascular positions, which seem to ensure the initial survival of the extravasated cells. This movie shows a tomographic image of cancer cells in perivascular positions.

Nonetheless, assumption of a perivascular position does not on its own guarantee long-term survival, as seen in these images. Although this micrometastasis grows in a perivascularpositionforthefirstninedays,itiscompletelygonebyDay14.

The fourth and final step in the brain metastatic cascade involves two alternative strat-egies—cooptive or angiogenic growth—depending on the type of cancer cell that has metastasized. As mentioned previously, lung cancer cells initiate angiogenic growth in order to nourish themselves.

In this series of images, we see how the potent angiogenic powers of lung carcinoma cells are able to provoke the florid outgrowth of microvessels that sustain the active growth of the tumor.

An alternative strategy for growth is exhibited by these melanoma cells, which have disseminatedtothebrainandassumedaperivascularposition.Hereweseehowtheirexpansion has depended on the cooptive strategy involving growth along the outside of blood vessels.

The multiphoton microscopy used to create these movies demonstrates the extraordi-nary advances in the powers of modern imaging technology to study cancer in living tissue.


Frank WinklerNeurology Clinic and National Center for Tumor Diseases & GermanCancerResearchCenter(DKFZ),UniversityofHeidelberg


Movie 14.3 Visualization of Cancer II: Metastasis

Metastasis has generally been a difficult process to visualize. In this model, high-resolution magnetic resonance imaging, or MRI, has been combined with tom-

ography techniques to visualize the metastasis of breast cancer cells to the brain of a mouse.

Before the injection of cancer cells into the circulation of the mouse, the cancer cells were labeled with a “dilutable” label, that is, one that loses half of its intensity each time a cell divides. This dilutable label is seen here in red.

Followinginitial injectionatDay1,manyhundredsof labeledcellsareapparent inthe brain. In the succeeding days, the number of these cells that retain full label, and therefore have not divided, is diminished.

At the beginning of the third week, small metastatic growths suddenly begin to appear in green. These green signals represent actively growing cancer cells. These green metastases continue to grow in size while the individual nongrowing, single-cell metastases continue to disappear.

Among other lessons, this experiment illustrates that hundreds if not thousands of micrometastatic single cells may be seeded initially in a target organ by cells arriving from theprimary tumorvia thecirculation.However,onlyaminute fractionof theinitially seeded cells succeed in spawning metastatic colonies.

Ann ChambersLondonHealthSciencesCentre

Movie 15.1 The Immune Response

An immune response involves events that unfold both locally, at the site of an infection, and at more distant sites, such as nearby lymph nodes. We can see the

integrationofthedifferentpartsoftheimmuneresponseifwefollowthecourseofatypical infection. Most pathogens are kept outside of the body by epithelial barriers, such as the epidermis, and are crossed only when there is an injury or tissue damage. After an injury, bacteria cross the epidermis and establish an infection in the under-lying tissue. Phagocytic cells in the tissues, such asmacrophages and neutrophils,engulf the pathogen. Dendritic cells are also phagocytic and are activated by binding pathogens to leave the site of infection and migrate to a lymph node. The migrating dendritic cells enter the lymphatic vessels and are collected in a draining lymph node. In the lymph node, T cells are activated by antigen presented by the dendritic cells, andinturnactivateBcellstosecreteantibody.EffectorTcellsandantibodymoleculesreturn to the circulation. They leave the circulation again at the site of infection, where inflammatory mediators have induced changes in the blood vessel endothelium. CD4Tcellsactivatemacrophagestobecomemorecytotoxic,whileantibodyrecruitscomplement to lyse bacteria directly and to opsonize them, enhancing their uptake byphagocytes. In thecaseofaviral infection,activatedCD8Tcellswouldkill anyinfected cells present.


Movie 15.2 Antigen Display and T-Cell Attack

When human cells are infected by viruses, the host cell ribosomes are exploited by the virus to translate their viral mRNA into viral proteins. In the normal course

of protein turnover in the cell, the viral protein, like many endogenous proteins, can be tagged by the attachment of a chain of Ubiquitin molecules, shown here in blue.

The poly-ubiquitin tagged protein is then led to a proteasome. The protein chain is then fed into the proteasome, which cleaves it into oligopeptides.

Peptidaseenzymespresentinthecytosolthendigestthemfurther.

On the cytoplasmic surface of the endoplasmic reticulum, some of these peptide frag-mentswillinteractwithTAPproteins,whicharespecializedpeptidetransportersthatpump them into the lumen of the endoplasmic reticulum.

Uponenteringthelumenoftheendoplasmicreticulum,thepeptidesencounterMHCclass I molecules. The peptides may then bind tightly to the peptide-binding groove of theMHCmolecules,aspecializedpartoftheMHCmolecule,shapedlikethepalmofa hand.

AmembranousvesiclecontainingtheMHCclassI–peptidecomplexisthenpinchedofftheERanddispatchedtotheinnersurfaceoftheplasmamembrane.Thevesiclethen fuses with the plasma membrane, and thereby exposes the contents of its lumen atthecellsurface.Inthisway,theMHCclassIandboundpeptidearedisplayedonthecell surface, making them available for surveillance by the immune system.

Cytotoxic T Lymphocytes, or CTLs, are an important arm of the immune system. A CTL, shown here in pink, may display on its surface an antigen-recognizing protein, specifically, a T cell receptor. The T cell receptor may recognize the oligopeptide being displayedbytheMHCclassIreceptor.TheCTLalsodependsonasecondcellsurfacemolecule,calledCD8,whichrecognizesandbindsallMHCclassImolecules.

This recognition and binding activates the cytotoxic T cell. The activated T cell first uses perforin to punch holes in the surface of the targeted cell, and then it injects pro-apoptotic enzymes, called granzymes, into the cytoplasm. Activation of the apoptotic cascade by granzymes eventually results in the death of the targeted, virus-infected cell.

Biointeractive,HowardHughesMedicalInstitutewww.hhmi.org/biointeractive©2013HHMI

Bruce D. WalkerHowardHughesMedicalInstitute,HarvardMedicalSchool


Movie 16.1 Drug Export by the Multi-Drug Resistance Pump

Transmembrane drug efflux pumps can remove chemotherapeutic drugs from inside a cancer cell, thereby reducing intracellular drug concentration, which

generatesresistancetothecytotoxiceffectsofthedrug.Certaindrugeffluxpumpscanextrude multiple types of drug molecules creating the state of multi-drug resistance, or MDR.

P-glycoprotein (Pgp) is themost prevalentMDR transporter and it is expressed atelevated levels in many kinds of cancer cells, especially those that have survived a chemotherapeutic treatment.

Thisspace-fillingmodelofP-glycoproteinshowsadrugmolecule (colored inpink)boundtothepump’sdrugbindingpocket(coloredinsilver).Suchadrugmoleculewill have entered the pump and its drug-binding pocket from the cytosol through the open portal seen at the bottom.

Energyisrequiredtopumpdrugsoutof thecell.P-glycoproteinisanATP-depend-ent transmembraneprotein, andATPmustbind to the interiornucleotide-bindingdomains in order to pump a drug or other toxins out of the cell.

Switchingtoaribbonrepresentationofthepump,wecanobservethedramaticcon-formationalchangeinthepumpcausedbyATPbinding.Drugsorothertoxinsenterthepump,andattachtothedrug-bindingpocket,whichresultsinanATP-dependentshift in the conformation of the entire pump. The drug is then extruded into to the extracellular space.

Rotating the structure 90degrees in the sameplane,we canobserve thepumpingaction, and the opening and closing of the cytosolic and extracellular ends of the pump.

In a third view, we observe the pump through its channel, looking down from the extracellular side.


Geoffrey ChangTheScrippsResearchInstitute


Movie 16.2 PI3K

Phosphatidylinositol-3kinase,orPI3K,isamultidomainenzyme.ThestructureofPI3K,asdeterminedbyX-raycrystallography,isshownhereasaribbondiagram.

EachdomainofPI3Kisassociatedwithdistinctfunctionsandcoloreddifferently.TheRas-binding domain is shown in red; this domain enables Ras to directly activate the PI3Kcatalyticdomain.Thecatalyticdomainiscoloredpurple.

This space-filling model shows the same protein but more closely resembles the actual 3-dimensionalstructure.

Zooming-into thecatalyticdomain,weenter thecatalyticcleftandobserveadrugmoleculethatinhibitsnormalPI3Kenzymefunction.TheinhibitormoleculeblocksthecatalyticsiteandpreventsPI3KfromaccessingitsusualATPsubstrate.Thissub-sequentlypreventsPI3Kfromphosphorylatingphosphatidylinositoldiphosphateanddisrupts downstream signaling.

Thisdrugmolecule,calledGDC-001,isshownhereasastickfigure...andnowasaspace-filling model. In this view, we can see the complementary three-dimensional structures of the drug molecule and the walls of the surrounding catalytic cleft.

The specificity of binding between the drug molecule and the catalytic cleft is achieved, in part, through the formation of hydrogen bonds between the drug molecule and the side-chains of amino acid residues lining the wall of the cleft.

Paul Workman and Rob L.M. van MontfortCancer Research UK


Mini-Lectures by Robert A. Weinberg

The author has recorded sixteen mini-lectures for students and instructors on a rangeoftopicscoveredinthebook.TheseareavailableinMP3formatandcanbe

transferred to a mobile device, as well as enjoyed on your computer. They are located in the “Mini-Lectures” folderon theDVDor canbe foundon theGarlandSciencewebsite. Below is a list of the Mini-Lectures, followed by transcripts of the recordings.

Mini-Lecture Table of Contents:01.Mini-Lecture:MutationsandtheOriginofCancer

02.Mini-Lecture:EpidemiologyandCancer

03.Mini-Lecture:CancerandReproduction

04.Mini-Lecture:GrowthFactors

05.Mini-Lecture:TumorSuppressorGenes

06.Mini-Lecture:p53andApoptosis

07.Mini-Lecture:CellSenescence

08.Mini-Lecture:CancerDiagnosis

09.Mini-Lecture:CancerStemCells

10.Mini-Lecture:InflammationandCancer

11.Mini-Lecture:HeterotypicCells

12.Mini-Lecture:MetastasisI

13.Mini-Lecture:MetastasisII

14.Mini-Lecture:ImmunologyandCancer

15.Mini-Lecture:CancerTherapies

16.Mini-Lecture:TheComingCancerEpidemic


01. Mini-Lecture: Mutations and the Origin of Cancer

An abiding theme in much of modern cancer research is the notion that many can-cer-causing agents, carcinogens, act through their ability to enter into the body’s

tissues, and to damage specific genes inside previously normal cells. In other words, that carcinogens can act as mutagens to mutate genes. And, in fact, we do know that a large number of carcinogenic agents are responsible for creating mutations inside cells. And through these mutations that they create, these carcinogens are able to elicit thediseaseofcancer.Stateddifferently,weknow,withoutanydoubt,thatcancercellsinvariably have mutated genomes.

Thisraisesthequestionof:whataretheagentsthatprovokemanyhumancancers?In fact, in the case of lung cancer there is a clear chain of causality. Cigarette smoke contains a large number of combustion products that are inhaled into the lungs, and these agents, these chemicals, are then converted via various metabolic enzymes into chemical compounds that are capable of interacting with and forming covalent bonds withtheDNA.SuchstructurallyalteredDNAmoleculesthenmaybereplicatedandyield ultimately altered DNA sequences, which we would call mutant genes. This raises the question of whether this can be generalized, or whether there are other sources of human cancer.

The fact of the matter is that lung cancer may be leading us astray because it may be the case that the great majority of human cancers are not traceable to specific muta-genicchemicalsthatenterintothebody.Possiblyit’sthecasethatmanyoftheagentsthat are carcinogens don’t act as mutagens but rather act through other mechanisms toprovokecancer.Forexample,weknowthatcertainkindsofchronicviralinfectionsare able to cause cancer in certain specific tissues. In East Asia for example, we know thatlife-longchronicHepatitisBviruscanleadtoanalmost100-foldincreasedriskoflivercancer.WeknowthattheHepatitisBvirusisnotdirectlymutatinggenesinsidethe liver; instead it’s creating a chronic inflammatory state that, in turn, seems to be responsible for the pathogenesis for the development of the liver cancers.

In a variety of other tissues we realize as well that areas of chronic inflammation have a greatly increased risk of ultimately spawning cancers. Then there is the striking and provocative discovery that certain anti-inflammatory drugs, including even simple householdaspirin,canbeveryeffectiveiftakenonadailybasisinreducingby30or40% the incidenceof certaincommonlyoccurringcancers, including forexample,coloncancerandperhapsevenbreast cancer.Theseeffectsof inflammationmightbe integrated into our rapidly evolving understanding of what the real agents are thatareresponsiblefortriggeringmanykindsofhumancancers.Simplyfocusingonmutagenic chemicals blinds us to these other agents, which may be more commonly involved in the etiology that is the causation of human cancers.


02. Mini-Lecture: Epidemiology and Cancer

The search for the origins of cancer has gone on for several centuries. Our first clues to what triggered cancer came in the late eighteenth century, with the discovery

bytheLondonphysicianPercivalPott,thatmenwho,intheiryouths,hadworkedaschimney sweeps came down with an unusually high rate of cancer of the scrotum, adisease thatwasotherwise relatively rare.He speculatedat the time that, in fact,this unusual cancer came from the fact that these individuals were exposed to a large amount of creosote and tars that were present in the floes of London chimneys. In fact, his discovery soon went to the Continent, where, within a decade, chimney sweeps made sure that after they swept chimneys they made sure they washed themselves within a day or two, unlike in England, where people washed themselves every week or so, whether or not they needed it. And so by the beginning decades of the nineteenth century there was a dramatic drop in the rate of scrotal cancer among the chimney sweeps of the Continent. This interesting tale represents the first documented instance where we can actually point to a mechanism of cancer that can be traceable ultimately to an external source, that is, some type of carcinogen entering into the body from the outside and provoking a disease, in this case a tumor, at a greatly elevated rate.

Today, two centuries later, we take for granted that many kinds of cancers are actually caused by external sources, external agents, which enter into the body, damage our tissues in one way or another, and provoke elevated risks of cancer. In the end, much of our conviction about these external agents for causing cancer comes from the sci-ence of epidemiology that surveys tumor incidence and tumor mortality in a variety of populations and attempts to correlate the risk of developing one or another kind of cancer with a specific lifestyle. Of course, such correlations do not prove causality, but sometimes correlations are so striking that they virtually represent a proof of causality. Forexample,alreadyin1950therewereindicationsintheUnitedStatesandBritainthatpeoplewhosmokedheavily,specificallymen,hadasmuchasa20-foldincreasedrisk of lung cancer. Of course, this did not prove that smoking actually causes lung can-cer, but it already sounded an alarm; and in the decades that followed there became experimental proofs of the notion that some of the substances in tobacco smoke are actually carcinogenic, that is to say, they are actually cancer-causing.

Todaythescienceofepidemiologyhasdevelopedinanumberofdifferentdirections,but it is clear already from the science of epidemiology that a number of the cancers thatweexperience in theWesternworldaredue inno smallpart todifferences inlifestyle:differences in smokinghabits, inexercise, inbodymass, and, very impor-tantly, in diet. Again, these are all correlations, but in some cases the correlations are sostrongastoremovemuchdoubtaboutcausation.Forexample,inmanypartsofAfricatherateofcoloncanceris1/20thasmuchasitisinEuropeandintheWest-ernHemisphere,specificallyintheUnitedStatesandCanada.Thesestaggeringdif-ferencescannotbeattributedtogeneticdifferencesinthepopulations.Thatistosay,American blacks that derive largely from western Africa and have therefore largely a African genetic heritage have rates of colon cancer that are comparable to the white population, and this proves tous that these staggeringdifferences in colon cancerincidenceandmortalityarenotduetogeneticsusceptibility.Similarly,untilrecentlytherateofdifferentkindsofdiet-inducedcancerswasdrasticallydifferentbetweenJapanandtheUnitedStates.However,whenJapanesemigratedtoHawaii,withinageneration their cancer rates closely approximated the rates of cancer that people of EuropeanorigindevelopedintheHawaiianIslands.Again,thisprovidesfuelfortheargumentthatthesestaggeringdifferencesincancerratesarenotduetogeneticsus-ceptibilitybutrathertodifferencesinlifestyle,inthiscase,largelyoneimagines,diet.

As a consequence, one can begin to assess what fraction of cancer in Western socie-ties isdue todifferences in lifestyle andwhatdifferencesaredue to the surround-ing environment. When dealing with environmental pollution and contamination of the food chain, most epidemiologists estimate that less than one percent of cancer rates are due to these sources. Instead, the great majority of the cancers in the West-ern world are due to either tobacco consumption or alternatively to various types of


diets.TobaccoconsumptionisresponsibleintheUnitedStates,forexample,forabout 30–40%ofcancerincidence.Atthesametime,oneimaginesanalmostequalamountofcausationtobeattributabletovariouskindsofdiet.In2007,forexample,theAmeri-canCancerSocietyestimatedthatasmanyas90,000ofthemorethan500,000deathsdue to cancer in that year were ascribable to the fact that individuals had a high body mass index, that is to say, they had a high weight ratio compared with the physical dimensions of their body. In that sense, body mass index, or obesity, is the second most important cause of cancer causation after tobacco usage. If we look at various types of commonly occurring cancers in the West, including, for example, colon can-cer, prostate cancer, and breast cancer, we see that diet appears to play a very impor-tant role in the causation of these diseases. And if we were to add up all these various causes,onemightimaginethatasmuchas70%ofcancerintheWestcouldbeavoidedif people were to alter their lifestyle. This in fact is a stunning revelation because it indi-cates that to the extent that we anticipate massive reductions in cancer mortality over the next generation, the bulk of those reductions will come not from the development of dramatic new cures to treat existing tumors, but rather from changes in lifestyle that prevent the appearance of cancer in the first place.

Discussions like these often provoke the question of how a person might be able to avoid the increased risk of cancer, and here one focuses on the assignable causes that have been identified to date. Tobacco smoke, either directly inhaled or secondhand smoke, is an obvious cause, being responsible for about 35% of cancer and obvi-ously avoidable in most cases. Diet seems to be an extraordinarily important cause of cancer. Increased body mass index, that is, increased obesity, is correlated directly withelevatedincidenceofavarietyofdifferentkindsofcancer,asmanyasadozeninmen and in women. And this obviously suggests the notion that it is good to stay slim and physically in good shape throughout one’s life, even though we don’t fully realize how slim physique actually reduces the risk of cancer incidence. Also, it is becoming increasingly apparent that certain elements of Western diet, such as large amounts of red meat, of high fat, and of meat cooked at high temperatures is also deleterious and is almost certainly responsible for the increased risks of colon cancer, and likely pros-tate cancer, that are observed in the West. Breast cancer incidence is also influenced by diet, but here one must also take into account the fact that the reproductive his-toryofawomanisalsoanimportantriskfactor.Forexample,awomanwhohashadchildrenstartingatarelativelyearlyage,saytheageof20,isatamuchlowerriskofeventually developing breast cancer than a woman who has her first child at the age of 30or35,orawomanwhoneverhashadanychildren.Intheend,anotherimportantand unavoidable factor comes from the genes that we inherit from our parents. This obviously does not fall under the purview of epidemiology, because here the playing field is tilted from the moment we are born, and here there is little we can do to avoid certain susceptibilities that we have inherited from our parents.


03. Mini-Lecture: Cancer and Reproduction

B reastcancerisofgreatconcerntomanywomen.LastyearintheUnitedStates,forexample,about40,000womendiedofthediseaseandperhapsfivetimesas

many were actually diagnosed with the disease. With that said, this raises the ques-tionof:whatarethecausesofbreastcancer?Weknow,forexample,thattheinternalhormonal environment of a woman is a strong determinant of risk of breast cancer; it is important to realize that because of nutrition and reproductive practices, the inter-nalhormonalenvironmentofwomenintheUnitedStates,forexample,haschangeddramatically over the last hundred and fifty years.

We know, for example, that the number of menstrual cycles through which a woman goes is approximately proportional to her ultimate cancer risk, breast cancer risk specifically. We know, for example, that the breast cancer risk of a woman is roughly proportional to the number of menstrual cycles she goes through in a lifetime. These cycleshave changeddramatically over thepast 150 years and therefore sohas theinternal hormonal environment of women. In the mid-nineteenth century, for exam-ple, a woman might at the age of sixteen begin cycling, at the age of eighteen become married and begin pregnancy, and between subsequent pregnancies, lactation, and breast feeding, might rarely if ever go through a full menstrual cycle for the next twenty or thirty years, and then she might go into menopause at the age of forty. These days agirlwillbeginmenstrualcycling,intheUnitedStatesatleast,attheageofasearlyas eleven or twelve because of increased nutrition, and she may not become preg-nant until the age of twenty-five or thirty and then have only one or two children. And because of improved nutrition she may continue to cycle until she’s in her late forties before she goes into menopause.

What this means is something quite astounding. An average eighteen year old Ameri-can girl may already have gone through as many menstrual cycles as her great, great grandmother went through in an entire lifetime, to give you one example of how the internal hormonal environment has changed dramatically.

In addition we know that once a woman goes through a pregnancy, part of the breast tissue is converted to a state where it is no longer susceptible to generating breast can-cers. And therefore, the earlier in life one goes through this state of bearing a child, sometimes called parody, the more protective pregnancy is. The longer one postpones parody, the greater the breast cancer risk; women who never have children have a cor-respondingly increased risk of breast cancer.

In saying all these things one can see how reproductive practices and nutrition have hadadramaticaffectontheriskofbreastcancerandthattheseeffectsdwarfanythatmight arise from the environment, that is, for example, from environmental pollution, which seem miniscule in comparison to the profound changes in the hormonal milieu existing inside a woman’s body.


04. Mini-Lecture: Growth Factors

One of the key insights into the peculiar biology of cancer cells has come over the past20yearsaswe’verealizedthatthereisanintimateconnectionbetweenthe

actions of oncogene proteins, sometimes called oncoproteins, and growth factor sig-naling. One can trace much of this logic back to a simple notion. In the context of a complex animal like ourselves, one cannot allow individual cells in one tissue or another to begin to proliferate on the basis of their own internally generated signals. Instead, in a well-structured tissue, each cell must consult its neighbors before it con-verges on the conclusion that it is time for it to grow. In the absence of such continuous consultation, individual cells here and there in a tissue would spontaneously begin to proliferate, multiply, maybe even yield large flocks of descendants, and very soon the architecturalintegrityofthetissuewouldsufferenormously.Consequently,eachofthe cells in a tissue lives in a condominium where all the neighbors are continuingly chatting with one another, discussing the appropriateness or the inappropriateness of one another beginning to proliferate.

We now realize therefore that growth factor signals are transferred from one cell to the other. These signals more often than not encourage proliferation, but there are also signals that discourage proliferation. And in all cases these signals are sensed on the surfaceofcellsthatreceivethesesignalsbyspecific,highlyspecializedreceptors(wecallthemgrowthfactorreceptors),whichsensethepresenceofgrowthfactorsintheextra cellular medium and then transfer the information of this growth factor and its presence into the cell interior, thereby informing the signaling molecules in the cell interior that there are signals in the extracellular space that suggest, and induce, and urge the cell to proliferate.

We now realize that this general scheme of growth factor signaling is often hijacked in cancer cells. In contrast to normal cells, which never undertake proliferation unless induced to do so by external signals, cancer cells generate their own internal signals that stimulate their proliferation and in that way become independent of their envi-ronments. We now realize that cancer cells are able to induce these growth stimula-torysignalsthroughavarietyofdifferentsignalingmechanisms.

Forexample,normallygrowthfactorsarereleasedbyonecellandimpingeonasec-ond cell, making the second cell dependent on the first. In the case of many cancer cells, in contrast, the cancer cells acquires the ability to release growth factors into its immediate extracellular environment, and once present there, these growth factors can then rebind to the surface of the cell that has just produced them.

This thereby allows the cell to stimulate its own proliferation, creating what is some-times called an autocrine signaling route. Clearly the fact that the cancer cell is now making its own growth factors makes it independent of growth factors that might orig-inateelsewhereinthetissueorinthebody.Similarly,thegrowthfactorreceptorsthatare present on the surface of cells and enable cells to sense the presence of growth factor molecules in the extracellular space can also be subverted in cancer cells. Thus, some cancer cells may express far too many copies of a growth factor receptor mol-ecule on their surface, and in that way induce its inappropriate signaling activation even in the absence of any of the growth factors that are normally required to induce this receptor to fire.

Sometimes the receptors displayed on the surface of a cancer cell are structurallyaltered in a way that causes the receptor to transmit signals into the cell interior, into the cytoplasm, that induce the cell to proliferate once again, even in the absence of the growth factor being present in the extracellular space. A third alternative strategy by which cancer cells can become growth factor independent derives from the numer-ous molecules that operate in the cell cytoplasm, and that are normally dedicated to receivingandprocessinggrowthstimulatorysignalsthatarereleasedbyreceptors.Forexample, the ras protein in normal cells is normally in a quiet state in which it does not signal, and it may persist in that state for an extended period of time.


However, in theevent thatagrowth factorreceptorbecomesactivated, thatgrowthfactor receptor will emit signals, which, through a series of intermediaries, are able to activate the ras protein, converting it from its quiescent state to an active signal-emitting state. And the ras protein thereafter is able to release yet other signals fur-ther into the cell that are capable of inducing the cell to commit itself to proliferate. However,incancercells,onceagainthissignalingpathwayissubvertedbecausethestructurally altered ras proteins that one can find in as many as one quarter of human cancers have undergone a profound functional change. Instead of releasing carefully measured parcels of growth stimulatory signals, which is the behavior of the normal ras protein in normal cells, the structurally abnormal ras oncoprotein found in can-cer cells releases a steady, unabated stream of growth stimulatory signals into the cell interior, thereby diluting the cell into thinking that it has experienced growth factors initsextracellularspace(when,inreality,nogrowthfactorsarepresentthereatall).Inallofthesecases,weseeexamplesofhowanindividualcellbecomescutofffromcommunication with its neighbors and begins to generate instead its own agenda of growth and proliferation-precisely the biological attributes that we ascribe to cancer cells.


05. Mini-Lecture: Tumor Suppressor Genes

Tumor suppressor genes have been intensively studied over the past quarter-cen-tury because they provide half of the genetic explanation for the origins of can-

cer. Thus one might imagine that in any well-balanced control system the signals that favor, for example, cell proliferation, must be counterbalanced by negative signals that act against it. Or, to put it in an automotive metaphor, the accelerator pedal needs to be counterbalanced by the braking system.

The fact is, there are multiple ways in which genes and their encoded proteins can prevent or block the outgrowth of cancer cells. One group of tumor suppressor genes blocks cell proliferation, or induces apoptosis, and thereby eliminates incipient can-cer cells. Another group of similarly acting genes minimizes the outgrowth of cancer cells by protecting the genome. This latter class of genes is sometimes called caretaker genes, to reflect the fact that these genes are responsible for maintaining the integ-rityofthegenome,andinthatwayreducingtheeffectivemutationrateandthustheprogression of cancers through the multiple steps that lead ultimately to the forma-tion of a malignant growth. In general, these DNA repair genes, these guardians of the genome, are not thought to be classic tumor suppressor genes, and therefore are discussedinaquitedifferentcontext,thatis,theissuessurroundingthemaintenanceof the genome.

If we focus on the genes that are widely agreed to be tumor suppressor genes, we realize that they have multiple alternative mechanisms of action. The classically studied retin-oblastoma gene makes a protein that prevents the progression of the cell through the G1 phase of the growth and division cycle. The retinoblastoma protein normally acts tointegrateavarietyofdifferentsignalsthatconvergeonthedecisionastowhetherit is appropriate for a cell to proceed to grow and to divide, or whether the cell should alternatively halt progression through the cell cycle and enter into the non-growing G0 state of the cell cycle. In the absence of the Rb protein, the retinoblastoma protein, thecellproceedswilly-nillythroughtheG1 phase of the growth cycle, independent of whether the conditions are suitable or propitious for cell proliferation.

The other frequently studied tumor suppressor gene is the p53 gene, which happens to beinactivatedinalmost50%ofallcancers.Thep53proteinisresponsibleforreceiv-ing and integrating a variety of signals that arise from the various systems implanted throughout the cell that are designed to monitor the cell’s well being, including the integrity of its genome, the availability of nucleotides, the state of the cellular metabo-lism,andsoforth.Inresponsetosomesignals,thep53proteinwillcallahalttocellproliferation,possiblyareversiblehalt.Inthepresenceofothersignals,p53proteinwilltriggerprogrammedcelldeath,orapoptosis.Througheithermechanism,thep53protein reduces the chance that an aberrantly functioning cell will be able to prolifer-ate and spawn the large number of descendants that together constitute a tumor mass.

Yetanotherfrequentlystudiedtumorsuppressorproteinisp10,whichactsasaphos-phatase and cleavesPIP3, phosphatidylinositol triphosphate. In sodoing, p10pre-vents the activation of the AKT anti-apoptotic kinase. And thus, by acting to prevent thisactivation,p10infactpreventscellsfromgrowingandsurvivingunderconditionswhere by all rights they should be induced to enter into apoptosis. In citing these three examples, I mean to indicate that there are a diverse array of biochemical mechanisms by which tumor suppresser genes and their encoded proteins are able to call a halt to cell proliferation, indeed, are able to weed out and actively eliminate aberrant cells. In the absence of these quality control mechanisms, such cells will be able to proliferate and ultimately evolve towards the highly malignant state that one encounters in the oncology clinic.


06. Mini-Lecture: p53 and Apoptosis

A s we’ve learned over the past 20 years, a key defender against cancer that isimplantedincellsthroughoutthebodyisthep53protein.It’sclassifiedasatumor

suppressorproteinbecauseifandwhenp53functionbecomescompromised,there’samuchgreater likelihood that the cell bearing this compromisedp53proteinmaybegin to proliferate in an abnormal fashion, leading, in the end to a large flock of prog-eny that grow and form a tumor.

Thep53proteininnormalcellsisrapidlymadeinlargeamountsandrapidlydegraded,leadingintheendtoaverylowsteadystate levelofp53proteins.However, incellsthathavesufferedvarioustypesofstress, includinganoxia(theabsenceofoxygen),DNAdamage,andimbalanceinthesignalingpathwaywithinthecell,p53levelsrisequicklysimplybecausep53protein,whileitcontinuestobemadeinlargeamounts,isnolongerdegradedatequalrates.Theresultingp53protein,onceitaccumulates,thenacts in the nucleus as a transcription factor, activating the expression of a large bank ofgenesthathavevariouseffectsoncellsdependingonthenatureoftheinsultsthatacellmaybesuffering.Forexample,ifacellhassufferedaminimalamountofgeneticdamage,p53levelsmayincreaseandinducetheexpressionofcellcycleinhibitorsthatprevent subsequent proliferation of a cell until the cell has succeeded in repairing the damage that was initially inflicted on its genome.

Oncethisdamagehasbeenrepaired,thenp53levelsdeclinetonormalandthecellproceeds to go through its normal growth and division cycle. Alternatively, the cell may suffer devastating damage to its genome, indeed damage that far exceeds itsability to repair all this damage. That leads, in principle, to the possibility of a highly mutated genome and once again the outgrowth of an aberrant cancer cell. In response to this situation, thep53protein, rather than temporarilyhalting cell proliferation,may decide through complex biochemical mechanisms instead to provoke the pro-gram of cell suicide, which is called apoptosis.

Apoptosis insures that once activated, it is able to eliminate all traces of a cell within an hour. The nucleus shrinks up, the plasma membrane in apoptosing cells undergoes profound remodeling, looking as if it’s rupturing, the DNA in the chromosomes is frag-mented and soon, the apoptotic cell breaks up into small clumps that are rapidly con-sumed either by macrophages or by its neighbors, thereby removing all traces of this cell.Thisservesasanextremelyeffectivemechanismfortissuestoremovepotentiallydangerous cells from their midst. We can realize situations where this may not operate properly.Forexample,inmorethanhalfofallhumantumors,perhapsallofthem,thep53pathwayisnotoperatingproperlyandundertheseconditions,eventhoughacellmaysuffercertainkindsofinsults,includinggenomicdamage,whichleadstomutantgenes,thatcellmaycontinuetosurvive.Ineffect,oncethep53proteinbecomesdam-aged and no longer able to signal, a cell becomes blind to many of its defects and its apoptotic machinery can no longer be activated, allowing this cell, and by extension, its descendants, to survive and spawn a large cohort of descendants that could even-tually form a tumor. The apoptotic program is interesting in its own right. We might expect, on the basis of our experience with other signal transduction circuits, that this pathway is also composed of a series of kinases, for example. But in fact it operates on verydifferentprinciples.

Much of the decision as to whether or not apoptosis should be activated is localized to the surface of the mitochondria and the cytoplasm. We’ve always thought of the mito-chondria as being the factories of energy production in the cell, but it now appears clear that in early evolution of eukaryotic cells, a totally unrelated function became inserted into the mitochondria; this unrelated function is the program of apoptosis. We now know that when apoptosis is initiated, one of the first events in many cases is that the mitochondrial membrane opens, releasing cytochrome c molecules into the soluble part of the cytoplasm that we call the cytosol.

Cytochrome c molecules have traditionally been studied in the context of cytochrome c transport inside the mitochondria, but now we learn of a totally unrelated function,


because when these molecules are leaked out into the cytoplasm, they associate with other molecules and activate the functioning of the whole series of intracellular pro-teases called caspases. These caspases cleave one another in a long sequence, and in the end the action of the caspases can be used to explain many aspects of the apop-totic program.


07. Mini-Lecture: Cell Senescence

The process, or the cellular state, termed senescence was first described more than40yearsagowhenitwasobservedthatwhencellsarepropagatedinculture,

that is in petri dishes and in incubators, they have a finite, apparently pre-ordained number of successive cell generations through which they can pass before they stop growing. After this number of cell generations, cells seem to enter into an essentially irreversible state of senescence, as it was termed. They become physically very large, they develop certain anomalies in the appearance of their nuclei, and can they can remain viable for many weeks thereafter, once again without ever again entering into the cell’s growth and division cycle.

The sources of cellular senescence have been strongly debated in recent decades. In some cases, there has been evidence that the shortening of telomeres at the end of chromosomes were responsible; but the weight of evidence increasingly suggests that cellular senescence comes from sub-optimal conditions of cell culture in the incuba-tor and from unbalanced signaling from oncogenes that might be active within cancer cells.Forexample,onehasroutinelypropagatedcellsintheincubatorat20or21%oxygen, which is the oxygen tension of the surrounding air. But in truth, the oxygen experienced by cells within our tissue is much closer to three or four percent, rather than the20% thatoneexperiencesroutinely in the incubator. Ifone indeed lowersthe oxygen tension to which cells are exposed in the petri dish to a more physiologi-cal level, then cells will pass through far more growth and division cycles before they enterintosenescence.Similarly,inthecaseofcertainepithelialcells,iftheyareprop-agated on their own in pure culture, they may senesce after six to eight generations. However,iftheyareprovidedwithastromalfeederlayerbeneaththem,whichsendscertain kinds of growth-stimulatory and survival signals to the epithelial cells, then epithelialcellsinculturecanoftendoublefor30,40,oreven50cellgenerations,againan indication that the number of successive cell generations through which cells pass in the tissue culture dish is often a reflection of sub-optimal conditions of culture.

Related to these processes are the actions of certain oncogenes. For example, theoncogene acting through the offices of its encoded oncoprotein, the Ras protein or Rasp21,isabletoinducesenescenceinprimarycellsifitisexpressedathighlevels.On the one hand, this might suggest that it is very difficult to transform cells with a ras oncogene, and that this senescence response is a reflection of a built-in or hard-wired defense that mammalian cells have to the overexpression or the excessive signaling released or emitted by oncoproteins. In fact, when the Ras oncoprotein is expressed at more physiologic levels, then cells may begin to exhibit some of the phenotypes of cell transformation without entering into senescence. Interestingly, the type of senes-cence into which cells enter in response to excessive ras oncogene signaling is very similar to the replicative senescence that is observed after extensive culture in vitro. These various forms of senescence begin to converge on the notion that this senes-cence state is a reflection of cumulative damage the cells have sustained, either in culture or being exposed to the signals released by oncogene proteins.

Senescencecanalsobeobservedin vivo, where pre-malignant legions, for example, the nevi that are precursors of melanomas, often show high numbers of senescent cells, indicating that senescence may be a physiological mechanism that is designed and hard-wired into our cells to prevent the outgrowth of premalignant cells, in other words, an anti-cancer defense mechanism. Even in this case, it appears likely that the senescence exhibited by these premalignant cells is also a reflection of excessive onco-gene signaling and cumulative damage sustained by cells that have experienced such signaling over a period of many days’ and many weeks’ time. Much of the damage seems to be focused on the cell’s genome, where there seem to be irreparable damage foci in the chromosomal DNA, whose presence somehow stabilizes the senescence state and somehow precludes such cells from ever reentering into the active growth and division cycle.


08. Mini-Lecture: Cancer Diagnosis

Over the past several decades we’ve developed increasingly powerful mechanisms and technologies for diagnosing human tumors. Indeed, now we can diagnose

tumors that are so small that in fact in previous years they would have eluded detec-tion by the human eye or most commonly used imaging technologies. This increased ability to detect small tumors is a double-edged sword, however. In principle, one might imagine that by detecting a small tumor, one can remove it and thereby prevent it from growing much larger and becoming life threatening. But there is a complica-tion because many small tumors will actually remain small for the life of the individual who is carrying them, and therefore with increased powers of diagnoses, one begins to pick up growths that quite possibly would have never become life threatening during the life of the patient who happens to carry these small tumors.

Onesaysthatthesedaysifonedoesautopsieson80-year-oldmen,independentofthecauseoftheirdeath,asmanyas80%ofthemhavedemonstrableprostatecancer,andyetweknowthatonly3and4%percentofmenwillactuallydiefromprostatecancer.This indicates that in a number of tissues as one gets older one accumulates growths which by certain criteria would be called cancer, but whose future growth, whose future properties, are really quite ambiguous. And therefore we are confronted with the dilemma of not knowing precisely how to respond to these recently diagnosed tumors.

A similar situation, perhaps not as dramatic, operates in the case of breast cancer. These days we are detecting very small breast tumors that previously would have gone undiagnosed, but because of various considerations, including malpractice concerns, these tumors are often treated quite aggressively even though it is clear from the epi-demiology of cancer that the majority of them would never become life threatening. The problem, of course, is then to discern how many actual cases of cancer there are in the population and how many of these are in part diagnostic bias, that is, we count them even though we imagine that many of them will never actually become clinically apparent.

Thesolutiontothis,ostensibly,istodevelopnewwaysofstratifyingdifferentkindsoftumors. By that I mean classifying recently diagnosed tumors into those that are likely to become malignant and aggressive and life threatening, and those tumors that are likely to remain indolent, that is, sitting there without any active growth.

The use of gene expression arrays may make this possible. Indeed, already it has had someeffectsinallowingrecentlydiagnosedbreastcancers,forexampleintheNeth-erlands, to be treated as indolent growths that will never become life threatening and never require, for example, surgery and aggressive chemotherapy. Women who have such small growths are therefore not treated in any aggressive clinical way, in contrast to those whose gene expression array analyses of their tumors indicates the possibility that some tumors might, within five or ten or fifteen years, eventually become malig-nant and perhaps metastatic, and therefore require active intervention.

All this says that with our increased powers of detecting small tumors, we’re now faced with a flood of growths whose actual optimal treatment remains quite ambiguous. We don’t know what to do with them in all cases, and in many cases, simply because of a doubt of their future growth, they’re treated aggressively, even though they probably don’t need to be.


09. Mini-Lecture: Cancer Stem Cells

The discovery of cancer stem cells, which has happened over the last five years or so, has really revolutionized our thinking about how tumors grow. If you would

have asked me ten, fifteen, or twenty years ago how can we understand the growth of a tumor, I would say that a tumor is simply a collection of cancer cells which grows exponentially. To be sure, one had to include in that description the notion of tumor-associated stroma, that is to say, the non-cancerous cells that are recruited into a tumor like a carcinoma, and are important for supplying the tumor with nutrients via the vasculature, and for providing other kinds of biological support. But in this case, in the instance of cancer stem cells, I’m talking about another dimension of complexity. Work in recent years indicates that both in hematopoietic tumors such as leukemias, as well as in solid tumors such as carcinomas, not all the cancer cells in each of these tumors is equivalent.

The fact of the matter is that if one plucks cancer cells out of, for example, a breast cancer, then divides them on the basis of cell surface markers that they happen to have, one can identify minority populations through the display of certain cell sur-face markers that can be introduced into a mouse, for example, a host mouse that is immuno-compromised, and these human breast cancer cells, in very low numbers, as few as two hundred, will seed a new tumor. In contrast to the behavior of this minority populationofcellstheremaybeamajoritypopulationthatincludesmorethan95%ofthetotalcancercellsofthetumor,andwhenoneimplants,forexample,20,000ofthose in a mouse, no tumor arises. This indicates that the minority population, far less than1%,hasaveryhighpercentageoftumorinitiatingcells,thatiscellsthatwhenintroduced into a mouse can trigger the growth of a new tumor, whereas the bulk pop-ulation of cancer cells contains very few of any of these tumor initiating cells.

We have to keep in mind here that the minority and the majority population are genet-ically identical. They’re all cancer cells and they’re all part of one genetic clone, how-everbiologically,they’rebehavingverydifferently,andthistypeofbehaviorsuggeststhe operations of a stem cell, that is to say a cell which upon division is able to gener-ate one daughter cell that once again becomes a stem cell thereby perpetuating the number of stem cells in a tissue, and a second daughter cell that undertakes a program of growth and division, ultimately spawning fifty or a hundred or even more descend-ants thatacquire thedifferentiatedcharacteristicsof a tissue,butat the same timegive up the option of ever dividing again. In this situation, the great bulk of the cancer cells in the tumor may actually be post-mitotic. They may lack self-renewal capability, unlike the stem cell which in principle is able to divide indefinitely each time generat-ing at least one daughter cell that itself becomes a stem cell.

If this kind of hierarchical organization is ultimately proven for a large number of can-cers,thishasimportantimplicationsforhowwediagnoseandtreatthisdisease.Forexample, in the instance of developing new kinds of anti-cancer therapeutics, we need now to begin to direct these drugs towards eliminating the cancer stem cells. Once one is able to eliminate the cancer stem cells, the tumor has lost its ability to regenerate itself, and therefore in principle, will ultimately wither away. In contrast, if one elimi-nates the bulk population of the cells within a tumor, that is the non-stem cells, the tumor as a whole may shrink, albeit temporarily, because the moment that the thera-peuticisremoved,thesurvivingcancerstemcells,whichmayhavebeenunaffectedby the chemotherapeutic treatment, may then begin the job of regenerating the tumor, and soon thereafter the tumor will have re-grown and become just as dangerous as it was prior to the chemo-therapeutic treatment.

We’ve only begun to appreciate the conceptual fallout from the discovery of these can-cer stem cells, initially in hematopoietic tumors such as leukemias, and most recently solid tumors. And it remains to be proven that such cancer stem cells inhabit virtually all human solid tumors, but this now becomes a possibility.


10. Mini-Lecture: Inflammation and Cancer

One of the most astounding discoveries in modern cancer research has been an epidemiological study of the rates of cancer among individuals who take non-

steroidalanti-inflammatories,oftencalledNSAIDs,suchas,forexample,ababyaspi-rin,everyyear.Suchindividualstakingababyaspirinoveraperiodoftentofifteenyears on a daily basis have significantly reduced rates of breast cancer, colon cancer, pancreatic cancer, often as much as twenty or thirty of forty percent below the rates that are typically associated with individuals of their age, their background, and their exposure to various environmental factors.

This suggests that inflammation in certain tissues throughout the body is actually conspiring, or contributing at least, to the formation of tumors. And indeed there is increasing evidence that chronically inflamed tissues represent sites of greatly increasedriskofvariouskindsoftumors.Possiblythemostdramaticexampleofthisis in the liver, where long term infection with hepatitis B virus or hepatitis C virus cre-ates a state of chronic inflammation, a reflection of the immune system’s attempts to rid the liver of these viruses, unsuccessful as they may be, and the resulting greatly increased risk of liver cancer, patocellular carcinoma. But similar situations seem to operateinavarietyofotherorgans.Peoplewhohavechronicgallstoneirritationofthegall bladder are at greatly increased risk for gallbladder cancer. Individuals who have chronic acid reflux in the esophagus are at greatly increased risk of esophageal cancer. Individuals who have chronic inflammation of the colon because of colitis and related conditions are at greatly increased risk of colon cancer.

Precisely how these inflammatory states actually predispose to the formation oftumorsisstillnotworkedoutingreatdetail.However,itisclearthatonecomponentofthe inflammatory state is the release by various kinds of immune cells of compounds called prostoglandins that are able to stimulate the growth of epithelial cells, often conferringonepithelialcellsmanyofthephenotypesoftransformedcells.Somehowthis chronic flux of irritating prostaglandins, along with other factors that are released in a chronically inflamed tissue, creates a biological microenvironment that greatly favors the expansion of clones of mutated initiated cells that can ultimately evolve into high grade malignancies.

We are just beginning to figure out how chronic inflammation contributes importantly to cancer risk, but it is clear that it does in profound ways. This also suggests, by the way, that if we could figure out ways of lowering the degree of inflammation in certain tissues, there would be a correspondingly greatly decreased incidence of disease in thosetissues.Andsincealoweringofincidentsisintheendfarmoreeffectivethantreating disease in terms of reducing mortality in a population, the notion of treating with anti-inflammatories in certain organs becomes a very attractive one.


11. Mini-Lecture: Heterotypic Cells

The most simplistic view of a tumor, such as a carcinoma, which is a tumor of epi-thelial cells, is that the tumor itself is only composed of a large number of these

cancer cells, a mass of a billion or ten billion cells that together represents the tumor. But in fact, to the extent that one examines carcinomas under the microscope, one discovers that they are complex tissues, indeed, as complex as the normal epithelial tissues from which they arise.

Indeed,thedrivingforcesincarcinomasarethecancerousepithelialcells.However,surrounding these epithelial cells are a large number of mesenchymal cells, cells of connected tissue origin, and from the immune system, which constitute together the stroma and are recruited into the tumor-associated stroma from various sources throughout the body. Many of the fibroblasts and myofibroblasts that form the bulk of the stroma are recruited from adjacent normal tissue. Many of the inflammatory cells that constitute arms of the immune system, and are involved not only in immune rejection but also inflammation, are recruited ultimately from the bone marrow via the circulation.

Together,asmanyas7or8distinctcelltypesconstitutethetumor-associatedstroma.Included among these are the aforementioned fibroblasts and myofibroblasts, lym-phocytes, neutrophils, macrophages, endothelial cells, and pericytes. Together, they represent an important source of functional support for the carcinoma cells. They can-not grow in a vacuum, and instead require various types of signals from the stroma for their support. The most obvious type of support comes from the endothelial cells and pericytes that form the tumor- associated neovasculature, that is to say, the blood vessels that are formed in the tumor’s stroma and that supply cancer cells with much needed nutrients and oxygen and that serve to evacuate metabolites, as well as carbon dioxide. But in addition to these cells, which form the tumor-associated microvascula-ture, there are a variety of other signals that come from the fibroblasts and myofibrob-lasts and that provide to the carcinoma cells mitogenic signals, as well as signals that prevent the carcinoma from entering apoptosis. As a consequence, we realize that this complex tumor-associated stroma has not been implanted there to confuse cancer biologists, but rather is recruited purposefully by the carcinoma cells because they absolutely require this stroma in order to survive and to prosper in the body of a can-cer patient.

As tumors become more and more malignant, the associated carcinoma cells may lessentheirdependencyonrecruitedstromalcellsfromthehost.However,ingeneral,even the most aggressive tumors still show a dependence on these recruited stromal cells for various types of cell physiologic support. Indeed, the inflammatory cells that are brought into the tumor-associated stroma and are often active transiently during the process of wound healing provide many of the signals that goad the cancer cells into ever-more rapid proliferation, and ultimately enable the cancer cells to evolve to higher states of malignancy. Accordingly, if we wish to understand the biology of the tumor as a whole, we must expand our gaze from just focusing on the carcinoma cells and their intracellular signaling pathways, and must instead now include in our vision the variety of mesenchymal cells that are present in the tumor stroma and that exchange signals with the carcinoma cells. This exchange of signals between one cell type and another is often called heterotypic signaling, and therefore heterotypic sig-nals are increasingly being factored into our view of the biochemistry and signaling physiology of cancer cells.


12. Mini-Lecuture: Metastasis I

I used to think that metastasis and invasiveness by cancer cells was an impossi-bly complex thing to study. After all, if you look at a highly malignant cell that has

acquired all these capabilities, not only can it move, not only can it degrade adjacent extra cellular matrix, but it is able to survive in the blood, it is able to establish itself in new tissue environments, and it changes many aspects of cell biology simply in order to achieve these ends.

Suchacomplicatedrepertoireofcellbiologicalbehaviorssuggestsenormouscom-plexity.Howcouldoneeverencompassitinasimpleandasingleconceptualscheme?The truth of the matter is that over the last five years, from a large number of laborato-ries, there has come increasing evidence that cancer cells acquire the ability to execute and to choreograph these complex steps by activating transcription factors that are normally used to program critical steps in early embryonic development. These tran-scription factors act in a very pleiotropic fashion. By that I mean that each of these transcription factors is able to simultaneously up-regulate a lot of genes, and therefore proteins, that enable a cell to change its shape, to acquire motility, to become invasive, and to execute many of the steps we believe are required in aggregate in order for a cell to leave a primary tumor and to take up residence in a distant tissue site.

This sequence of steps is often called the invasion metastasis cascade. It involves alto-gether: the acquisition of local invasiveness by a cancer cell; the invasion of the cell into blood vessels, which is often called intravasation; the transport of the cancer cells through the blood vessels to distant tissue sites; the escape of the cancer cells from the blood vessels, which is often called extravasation; and finally, the acquisition of the ability of the cancer cell to adapt to the local tissue environment in which it has landed and to begin to proliferate to form a growth which ultimately becomes visible at the macroscopic level, that is, with the naked eye, in other words, to create a life-threatening metastasis.

We know about many of these steps originally from the study of early embryogenesis. Forexample,duringtheprocessofgastrulation,cellsthat initially formintheecto-derm migrate into the embryo to form the mesoderm and the endoderm. This migra-tion entails a large number of changes in the cells that started out as ectodermal cells. These cells are initially epithelial. Cells in an epithelium normally form two-dimen-sional sheets, they are unable to move, they are tightly attached to their neighbors and as such, they also express a number of cytoskeleton proteins, such as keratins, that are characteristically expressed by all cells in an epithelium.

However,whenthesecellsleavetheectodermtomigrateintotheembryo,theymustdetach themselves from their epithelial neighbors; they must shut down epithelial gene expression programs, including the expression of keratins, and they become quite mesenchymal. That is to say, they take on many of the aspects of fibroblasts. They become migratory; they release proteases into the extra cellular space that enables them to degrade certain obstacles that may be in their path. As such these cells have changedfromanepithelialdifferentiationstatetoonethatisquitemesenchymal,andas such have undergone what is often called the epithelial mesenchymal transition, or the EMT.

Once cells migrate to the inside of the embryo, they may retain their mesenchymal traits, in which case they will become, for example, cells of the mesoderm. Alterna-tively, some of these cells may become endodermal and as such may once again reac-quire epithelial characteristics, in which case their initial conversion from an epithe-lial to a mesenchymal state was only reversible and could be rapidly changed back to an epithelial state once they reach their intended goal, which in this case is the inside of the embryo that forms the primitive gut.

There are strong analogies between this sequence of events and those that happen during the formation of metastases in human patients that bear for example carcino-mas.Carcinomasconstituteabout80%ofthetotaltumorburden;andtheseallarisefrom epithelial cells in various epithelial tissues including the lungs, the esophagus, the stomach, the liver, the uterus, and many other tissues as well.


The cells lining the glandular tissue in the breast are also epithelial, leading to breast cancer. In many cases we see that during the progression of a benign tumor to a malig-nant tumor in these various tissue sites, one initially has cancer cells that exhibit very many epithelial characteristics. These cells are still localized, they still may form ducts, for example, which are a characteristic of epithelial cells, and they have not yet become invasive.

However,astumorprogressionproceeds,thesecellsmayshedmanyoftheirepithe-lial characteristics and undergo an epithelial-mesenchymal transition, which enables them to become motile and invasive and begin the whole process that ultimately leads to their ability to form metastases in distant tissues. Thus they undergo an epithelial-mesenchymal transition, which may profoundly change their entire gene expression program. And the way they accomplish this, in many cases and perhaps all, is to reacti-vate early embryonic transcription factors, indeed, the same transcription factors that are responsible during early embryogenesis for programming critical steps such as gastrulation.

All of this raises the question of what causes carcinoma cells in the context of a pri-marytumortobegintoexpresstheseearlyembryonictranscriptionfactors.Herewebegin to believe that the environment of these cancer cells provides the signals that persuade these cells to turn on expression of these transcription factors. More par-ticularly, one can see that signals that are released by the stromal cells in the tumor, the non-malignant cells, impinge on the cancer cells, induce them to turn on these transcription factors, which then enable these cancer cells to acquire mesenchymal phenotypes and begin to invade and metastasize. Interestingly, when these cells land in a distant tissue site, they no longer experience the same mixture of signals from the stromal cells in that distant tissue site. Consequently, they no longer are able to main-tain the expression of these early embryonic transcription factors. This allows many of the metastasizing cells to shut down the expression of these transcription factors, we believe, and thereby revert from a mesenchymal to an epithelial state, in that way recapitulating the behavior of their ancestors in the core of the primary tumor from which they originated.


13. Mini Lecture: Metastasis II

The process of metastasis is actually quite complicated. Cancer cells within a pri-mary tumor must acquire the ability to invade into adjacent tissue, to enter into

blood vessels and lymphatic ducts, to travel through those vessels to distant sites in the body, to escape from those vessels, to invade into adjacent tissues, and perhaps most difficult is the last step of colonizing the tissues in which they have landed. That is, they begin to adapt to the tissue microenvironment, which is quite foreign from the microenvironment they experienced in the context of the primary tumor.

An interesting question is how cancer cells acquire the phenotypes of invasion and metastasis.Hereonethinksoftheprocessesthatprecededinvasivenessandmetasta-sis, that is, the processes that led to the formation of the primary tumor. We imagine that when a primary tumor forms, it forms as a consequence of a series of mutations that strike the genomes of evolving cancer cell populations, or at least pre-malignant populations. Each of these mutations, in a Darwinian sense, confers on the cell that happens to have sustained this mutation an increased survival advantage or prolifera-tive advantage, and consequently cells that have acquired one or another mutation begin to expand and increase their number within the context of the primary tumor. This suggests that most of the traits of cancer cells that inhabit a primary tumor have been accumulated as a consequence of the fact that they were selectively advanta-geous in a Darwinian sense during the growth of the primary tumor.

How, then,docancercellsacquire theability to invadeandmetastasize?Are thesetraitsalsoadvantageousascancercellsgrowwiththeconfinesofaprimarytumor?Infact,theanswertothatquestionishardlyobvious.Somemightarguethattheacqui-sition of metastatic traits has no advantageous consequences for the cell that hap-pens to have acquired this trait while it is living in the primary tumor. As such, it could be that the acquisition of invasiveness and metastatic potency may only be an unin-tendedsideeffectofthefactthatcancercellshaveacquiredcertainmutationswithinthe context of primary tumor growth that happen, in fact, to allow them to invade and metastasize.

Another point of view holds that once primary cancer cells form they recruit a so-called activated stroma, that is a collection of mesenchymal cells that are brought in fromthehostandthathelpthetumorasawholetogrow.Suchanactivatedstromahasclose resemblance to the stroma in a wound-healing site, and the cells that form this stroma release a series of signals that are known, at least in some biological contexts, to induce traits like motility and invasiveness in the cancer cells. As a consequence, the cells within a primary tumor may actually acquire the cell biological phenotypes thatenablethemtoleavetheprimarytumorandtomovetodistantsites.Hence,inthis kind of scenario, the acquisition of invasive and metastatic traits is really the con-sequence of the adaptation of cells in the primary tumor to the signals they are receiv-ing from the activated inflamed stroma, which provides many types of biological sup-port to the primary tumor, but also may influence in a profound way the biology of the cancer cells themselves.


14. Mini-Lecture: Immunology and Cancer

Throughout this book, we have seen a number of examples of protective mecha-nisms that evolution has implanted in cells throughout the body that help to

derailtheprogressoftumorformation.Forexample,thewholeprocessofcellsuicideis predicated on the notion that if the cell receives too many growth stimulatory sig-nals, the apoptosis or cell suicide program may be triggered, thereby eliminating a cell that is the potential progenitor of a flock of tumor cells. There are yet many other mechanisms that are hard-wired in the cells, which are guaranteed to stop cancer cells in their tracks, should they begin to proliferate. Of course these mechanisms are not foolproof. They are not fail-safe, because on frequent occasions, one does see cancers in the human population.

Nonetheless, these kinds of realizations provoke one to ask another kind of question. And that is, are there yet any other mechanisms that operate in the body in order to reducethelikelihoodofcancerappearing?Morespecifically,wecanposetheques-tion of whether the immune system also plays a role. This in fact is a very complicated issue, because much of what we know about the normal immune system stems from our study of the ways in which it protects us against various kinds of infectious agents. Thus, we become infected by a virus, or by a fungus, or a by a bacterium, and often we acquire a life long immunity which protects us from re-infection by that agent.

Onehaslearnedoverthepast40yearshowthisprotectionisachieved,andmuchofit depends on the ability of the immune system to discriminate between our own nor-mal cells, and the cells and proteins of infectious agents. Thus, we know of many situa-tions where foreign proteins brought into our body by viruses are readily detected. The virus particles that are expressing these proteins are rapidly eliminated by the immune system, which can readily distinguish these foreign proteins from the native proteins in our tissues.

On the other hand, the immune system becomes tolerant of the proteins that are made by our normal cells. To the extent it ever loses this tolerance, we have autoimmune diseases. These considerations serve as a useful background to highlight the possibil-ity that the immune system may be able to recognize cancer cells and eliminate them. That is, this raises the question of whether cancer cells are foreign enough or strange enough that they can readily be distinguished from the normal cells in our tissues, identified as such by certain arms of the immune system, attacked, and eliminated.

The truth is that such discrimination is extraordinarily difficult. If you look at the pro-teinsdisplayedbycancercells,99.9%ofthemareidenticaltotheproteinsmadebynormal cells. And therefore, these proteins offer the immune systemno clue as towhether cancer cells are worthy targets of destruction or should be ignored. In fact, in many cases we believe that incipient cancers are able to fly under the immunologi-cal radar and escape detection by the immune system, thereby being able to flourish. Still,therearecertaincluesexhibitedbycancercells,whicharenotexhibitedbynor-malcells,andsuggestthatthecancercellsare,inonewayoranother,abnormal.Forexample, we know of a class of cells that are part of the normal immune system called natural killer cells that are endowed with the ability to detect certain abnormal pro-teins present on the surface of cancer cells that are absent from the surfaces of normal cells. We believe that cells like natural killer, or NK cells as they are often called, are importantcomponentsoftheimmunedefensesagainstcancer.Perhapsthemostper-suasive evidence that the immune system plays a key role in defending us against can-cer comes from the realization that individuals who are immuno-compromised for a variety of reasons, including the fact that their immune system has been suppressed following organ transplants, show significantly increased levels of certain kinds of cancer, suggesting that their suppressed immune systems have allowed them to tol-erate engrafted tissues such as engrafted kidneys or livers. At the same time, these suppressed immune systems have permitted the outgrowth of cancers that otherwise would have been eliminated by normal healthy immune systems.


In fact, we are just beginning to learn how the immune system is able to distinguish normal cells from cancer cells and how, having made these distinctions, the immune system is able to specifically home in on cancer cells and to eradicate them, thereby reducing significantly the incidence of certain kinds of human cancers.


15. Mini-Lecture: Cancer Therapies

Modernmolecularcancerresearchreallybeganin1975and1976withthediscov-ery of the Srcproto-oncogene,andovertheensuing30years,wehavelearned

an enormous amount about the molecular mechanisms that create human cancers. Formanypeople,themotivationforlearningallthesefactsisnotsimplyintellectualcuriosity. Instead, they are much more interested in a far more practical outcome, which is: how can we exploit this recently acquired information to develop new ways oftreatingthediseaseofcancer?

The truth is that we have discovered a large number of proteins inside cells, specifi-cally inside cancer cells, that in theory should serve as useful targets for intervention when we are intent upon developing new chemicals that can shut down the growth of cancer cells, indeed even kill these cells.

Forexample,alowmolecular-weightcompoundmightbeabletogointothebody,move throughout the circulation, move into specific tissues, enter cells, and, once inside cells, it might be able to shut down the firing of a certain protein that is respon-sible for driving the proliferation of cancer cells or may even be responsible for block-ing cell death in these cells.

Once such an enzyme is blocked, the cancer cells may then activate their own cell suicide program, thereby triggering their own death. That, in the end, is the ideal. The question is, how can one make such compounds, and if one does make them, howcanoneensurethattheyareselective?BythatImeantosay,howcanweknowthat these compounds, once they enter into the body, will selectively strike down the cancercells,whileleavingthenormaltissuesunharmed?Thedifficultyofmanyexist-ing anti-cancer therapies, including many chemotherapies, comes from the fact that whilesomeofthemarehighlyeffectiveinshuttingdownthegrowthofcancercellsand indeed killing them, many of these chemotherapies also wreak havoc on normal tissues,andcreatehighandoftenunacceptablelevelsofsideeffecttoxicities.

Accordingly, one has attempted to craft new kinds of chemical molecules, new kinds of drugs, that can enter into cells and specifically shut down only those genes and pro-teins that are involved in the development of cancer without perturbing the metabo-lism of normal cells.

More often than not, it is much easier to shut down the firing of an existing enzyme that is active and signaling than it is to activate a previously latent or silent enzyme. All oftheseconsiderationshavefocusedeffortsondevelopingdrugsthatcan,forexam-ple, shut down the firing by certain tyrosine kinase enzymes that are responsible for drivingtheproliferationofavarietyofdifferentkindsofhumancarcinomas.Weknowthat the growth factor receptors at the cell surface of many of these carcinomas are fir-ing into the cell interior by using their tyrosine kinase domains. These tyrosine kinase domains are able to emit a diverse radiating streamof signals that affects awholeseries of distinct signaling pathways that lead, in turn, to the rapid proliferation and growth of cancer cells, and thus the growth of tumors.

The question is then: can we develop chemical compounds that are able to shut down the firing of certain growth factor receptors, while leaving normal cells and their growthfactorreceptorsuntouched?Andinrecentyears,onehashadsomesuccessesthere in developing highly selective drugs that are specific for interfering with one tyrosine kinase, but do not interfere with the ninety or so other tyrosine kinases that might be active within a cell.

Inthecaseofchronicmyelogenousleukemia,onehasadifferentexampleofwhereatyrosine kinase that is not part of a growth factor receptor is running amok, signaling constitutively, which is to say in an unabated fashion, and driving the growth of these leukemiccells.Hereonehashadenormoussuccessindevelopingacompoundthatisabletoshutdownthiskinasemolecule;whichinthiscaseiscalledABL(pronouncedable),withoutaffectingvirtuallyalltheotherkindsoftyrosinekinasesthatarepresentinside normal cells. And thus one achieves enormous selectivity in hitting the can-cercellswhilehavingrelativelyminimaleffectonthenormaltissueselsewhereinthebody.


16. Mini-Lecture: The Coming Cancer Epidemic

We’ve all read in the media about the increased numbers of various types of can-cer in the population. In fact, the only really accurate way of measuring cancer

incidence is to measure age-adjusted incidence. In other words, what is the risk of a 60-year-oldwomandyingthisyearofbreastcancercomparedtotheriskofa60-year-oldwomandyingfrombreastcancerintheyear1930,ratherthansimplymeasuringthe absolute number of cases in the population. This is important in no small part becausecancer incidencevariesdramaticallyatdifferent timesof life. Ifonemeas-ures cancer incidence by looking at the age-adjusted rate, one sees that many kinds of cancers that one has imagined are present in epidemic proportion really have not increased that much. Breast cancer, which was widely stated to be present in epidemic proportionsinourcountry,hadaratherflatrateofincidencefromabout1930toabout1995 in age-adjusted rates. Starting in 1995, therehasbeen an approximately 20%decrease in cancer-associated mortality, once again on an age-adjusted basis. Other kindsofcancershavealsogonedown.Stomachcancerhasgonedownbyafactoroffourorfive,coloncancerperhapsby20or25%becauseofcolonoscopy.Cervicalcan-cerhasgonedowndramaticallybyabout80%inWesternpopulationsbecauseoftheuseofthePaptesttodetectpremalignantgrowths.

Still,takingallthisintoaccount,andchangingpatternsoftobaccousage,itbecomesclear that the absolute number of cancers in the population is going to increase over the coming decade. This is because cancer is essentially a disease of the elderly. The riskofa70-year-oldmandevelopingcoloncancer isabout1000 timeshigher thanthat of a ten-year-old boy. This is true for a variety of other cancers with the exception, obviously, of childhood cancers, and among adults with breast cancer, which begins rather early in certain women. Taking all these factors together, one begins to realize that the number of cases in the population will increase simply because people are living long enough to develop cancer in their old age. Or, to put it another way, the number of cases of Alzheimer’s is also increasing. A disease like Alzheimer’s or cancer, whichwaspreviouslyrelativelyrare100yearsago,hasnowbecomequitecommonbecausepeoplearereachingtheageof80or90whenthesediseasesbecomeveryhighrisk- when these diseases begin to appear in large numbers.

If we look at the aging, therefore, of the Western population, the fact is that the number ofpeopleovertheageof65,whenmostadultcancersincreaseinincidence,isgoingto increase dramatically and therefore the absolute number of cases in any Western society will increase, even though the risk of any individual dying of cancer at a certain age may actually be going down.

AppleandQuickTimeareregisteredtrademarksofApple,Inc.MicrosoftWindows,MicrosoftWord,andPowerPointareregisteredtrademarksofMicrosoft,Inc.AdobeandAdobePhotoshopareregisteredtrademarksofAdobeSystemsIncorporated

media guide t-bo_c2

Documents