Cancer Genome Atlas and Functional Systems Biology Wei Zhang, Ph.D. Professor Department of Pathology Director Cancer Genomics Core Laboratory M. D. Anderson.

Cancer Genome Atlas and Functional Systems

Biology

Wei Zhang, Ph.D.Professor

Department of PathologyDirector

Cancer Genomics Core LaboratoryM. D. Anderson Cancer Center

The 6th Chinese Conference on OncologyMay 21-23, 2010, Shanghai, China

Complexity of CancerComplexity of Cancer

• Cancers have heterogeneous etiology. One patient’s cancer is different from another patient’s cancer.

• Cancers have heterogeneous genetic defects.

• Cancers are results of combinations of multiple genetic and molecular alterations.

• Personalized medicine

• Targeted therapy

Complexity of Human GenomeComplexity of Human Genome

• 30-40,000 genes• 1-10 millions of Single Nucleotide Polymorphisms (SNPs)• 10-20 millions of proteins

• One gene• different spliced mRNAs

• different proteins• different modified forms of proteins

Genomics and Proteomics

Broad-scope, large-scale measurement of gene copy number, gene expression, gene methylation, and protein expression.

Data interpretation or signal processing in pursuit of biological understanding.

DNA RNA

Splice VariantsChIP

• protein/DNA interactions

• transcription• DNA

replication• DNA repair

• splice forms of specific genes

• downstream effects on translation

GX

• high sensitivity measurements of transcription

• correlate results with genomic data

mRNA

Agilent Technologies Microarray Portfolio…

CGH

• chromosomal aberrations

• gene copy number

Copy number

CH3

• methylation patterns

• downstream transcriptional effects

Methylation

Transcription Factors

mRNA isoforms

miRNA

• presence of microRNAs

• knockout analysis

• correlate results with transcription data

RNA interference

What is TCGA?What is TCGA?

• The Cancer Genome Atlas (TCGA). The first phase is a 3-year, 100 million pilot project of the National Cancer Institute (NCI) and the National Human Genome Research Institute (NHGRI) focusing on glioblastoma and ovarian cancer. The second phase will cover 25 major cancer types.

• TCGA Mission: Increase scientific understanding of the molecular basis of cancer and apply this information to improve our ability to diagnose, treat, and prevent cancer.

• TCGA Purpose: Develop a complete “atlas” of all genomic alterations involved in cancer.

TCGA Pilot Project MilestonesTCGA Pilot Project Milestones

Collect/Utilize tumor tissue samples and medical information from cancer patients during treatment.

Catalog and store samples at a centralized facility and send genetic material to research centers involved in the project.

Identify genomic changes associated with cancer in individual patients.

Identify genomic patterns associated with the disease, and use that information to inform cancer diagnosis, treatment, and prevention.

Make information available to scientists as it is produced, to speed treatment and prevention research and help doctors and patients make treatment decisions.

1 2 3 4 5

Graphics credit: The Washington Post, December 14, 2005

How TCGA FunctionsHow TCGA FunctionsData Management, Bioinformatics, and Computational Analysis (GDAC)

An integrated database providing access to all of the information generated by the TCGA pilot project

Technology Development

Throughout the pilot project, technology development will enable improvements to genomic analysis

Cancer Genome Characterization Centers

Technologies to investigate and characterize genes that may be associated with cancer

High-throughput sequencing of genes identified through cancer genome characterization centers

Genome Sequencing Centers

Human Cancer BiospecimenCore Resource

Centralized facility to catalog and store tumor samples, and distribute genetic material to TCGA research centers

Our GDAC Center

• Center for Systems Analysis of the Cancer Regulome

• Directors: Ilya Shmulevich (Institute for Systems Biology); Wei Zhang (M.D. Anderson Cancer Center)

• Bioinformatic researchers at MDACC: Da Yang, Yuexin Liu

• Focuses are on prognosis markers, systems understanding and functional validation

Copy Number Methylation

mRNA expression

BiomarkerSystematic Network

• Prediction Analysis of Microarray

• Top Scoring Pair Algorithm

• Co-occurrence Copy Number Alterations

• Bayesian Network

Tumor Subgroup

• Distinct Dosage Sensitive Expression Patterns

Consensus Copy Number Altered Regions

Survival Classification

> 3 yr survival

< 3 yr survival

Top 200 pairs

The Cancer Genome Atlas: Glioblastoma

The Cancer Genome Atlas Research Network, Nature, 2008

Statistical analysis of mutation significance identified Eight Genes as Significantly Mutated and P53 Mutation Is a Common Event in Primary Glioblastoma.

TCGA., “Comprehensive Genomic Characterization Defines Human Glioblastoma Genes And Core Pathways,” Nature, 455(23), 1061-1068, 2008

Genomic and transcriptional aberration analysis detected New Recurrent Focal Alterations such as Homozygous Deletions involving NF1 and PARK2, and Amplifications of AKT3.

TCGA., “Comprehensive Genomic Characterization Defines Human Glioblastoma Genes And Core Pathways,” Nature, 455(23), 1061-1068, 2008

ILK PI3K

Integrins RTK

IGFBP2

AKTP

PTEN

P

P

P

P

P P

P

Survival

GrowthMetabolism

Migration

Proliferation

Akt Cell Signaling

Kinase domain RDPH

308 473

Akt1

Kinase domain RDPH

309 474

Akt2

Kinase domain RDPH

305 472

Akt3

Chromosome location

14q32

19q13

1q44

Homology 75-84% 90-95% 73-79%

Akt Isoforms

Adapted from Cheng GZ et al. 2008.

Developmental Roles of Akt1/2/3

Akt1 Akt2

Akt3

Neuronal development

Glucosehomeostasis

Cellular growthAngiogenesis

Postnatal Survival

Embryonic Development and Survival

Whole body weight and size

Adapted from Gonzalez and McGraw. Cell Cycle. 2009.

Differential Roles of Akt Isoforms in Cancer

• Differential roles of Akt1 and Akt2 in breast cancer (Hutchinson et al. Can Res. 2004; Arboleda et al. Can Res. 2003)

• Akt2 predominant in ovarian cancer (Noske et al. Cancer Letters. 2007)

• Akt3 important in melanoma (Robertson. Can and Met Rev. 2005)

• Akt activation in glioma correlates with higher tumor grade (Wang et al. Lab Invest. 2004)

Differential Akt2 and Akt3 Levels in Oligodendroglioma

AKT1

N O/AO

AKT2

N O/AO

AKT3

N O/AO

Hypothesis

Akt3 is the dominant Akt isoform which preferentially induces Oligodendroglioma

progression

Is there a hierarchy in the ability of Akt isoforms to promote oligodendroglioma development and

progression?

Kristen Holmes

RCAS/tv-a Glial-specific Transgenic Mouse Model

Begemann, M., Uhrbom, L., Rajasekhar, V.K., Fuller, G.N., and E.C. Holland. 2004 Dissecting Gliomagenesis Using Glial-Specific Transgenic Mouse Models. In Zhang, W. and G.N. Fuller (Ed.) Genomic and Molecular Neuro-Oncology. Sudbury: Jones and Bartlett. p233-278

WHO Grade IIOligodendrogliomaNormal Brain

WHO Grade IIIAnaplastic Oligo

Histologic Criteria for Oligodendroglioma Progression

Akt3 Promotes Oligodendroglioma Progression

Tumor Penetrance

Anaplastic OligodendrogliomaGene Combination

PDGFB 81% (35/43) 11% (4/35)

PDGFB / Akt1 77% (42/57) 16% (7/42)

PDGFB / Akt2 39% (11/28) 9% (1/11)

PDGFB / Akt3 100% (35/35) 100% (35/35)

GFP 0% (0/22) 0% (0/22)

AKT1

N O/AO

AKT2

N O/AO

AKT3

N O/AO

PDGFB + Akt3PDGFB + Akt1 PDGFB + Akt2

Akt3 Promotes Oligodendroglioma Progression

Challenge

How do we better understand cancer systems?

Systems biology

Systems BiologySystems BiologySystems biology is an emerging field that aims at system-level understanding of biological systems.

Unlike molecular biology which focus on molecules, such as sequence of nucleotide acids and proteins, systems biology focus on systems that are composed of molecular components. Although systems are composed of matters, the essence of system lies in dynamics and it cannot be described merely by enumerating components of the system. At the same time, it is misleading to believe that only system structure, such as network topologies, is important without paying sufficient attention to diversities and functionalities of components. Both structure of the system and components play indispensable role forming symbiotic state of the system as a whole.

- H Kitano

Probabilistic Boolean network

1. Shmulevich I, Dougherty ER, Kim S, and Zhang W. Probabilistic Boolean network: a rule-based uncertainty model for gene regulatory networks. Bioinformatics 18:261-274, 2002.

2. Shmulevich I, Dougherty ER, and Zhang W. Gene perturbation and intervention in probabilistic Boolean network. Bioinformatics 18:1319-1331, 2002.

3. Shmulevich I, Lahdesmaki H, Dougherty ER, Astola J, Zhang W. Proc. Natl. Acad. Sci. USA 100 (16) 2003.

US Patent # 7,257,563 (Shmulevich, Dougherty, and Zhang)

• Such relationships should also be validated experimentally.

• The networks built from our models should provide valuable theoretical guidance to experiments.

Cancer tissues need nutrients. Gliomas are highly angiogenic.

Expression of VEGF is often elevated.

VEGF protein is secreted outside the cells and binds to its receptor on the

endothelial cells to promote their growth.

GRB2GRB2

FGF7FGF7

FSHRFSHR

PTK7PTK7

VEGFVEGF Member of fibroblast growth factor family

Follicle-stimulating hormone receptor

Tyrosine kinase receptor

•The protein products of all four genes are part of signal transduction pathways that involve surface tyrosine kinase receptors.

•These receptors, when activated, recruit a number of adaptor proteins to relay the signal to downstream molecules

•GRB2 is one of the most crucial adaptors that have been identified.

•GRB2 is also a target for cancer intervention because of its link to multiple growth factor signal transduction pathways.

GRB2GRB2

GNB2GNB2

•Molecular studies have demonstrated that activation of protein tyrosine kinase receptor-GRB-2 complex activates ras-MAP kinase-NFB pathway to complete the signal relay from outside the cells to the nucleus.

•GNB2 is a ras family member.

MAP kinase 1MAP kinase 1

c-relc-rel•GNB2 influences MAP

kinase 1, which in turn influences c-rel, an NFB component.

IGFBP-2 in Glioma Progression• Up-regulation of IGFBP2 is one of the most consistent and

distinctive gene expression changes in high-grade gliomas (Fuller et al., 1999)

IGFBP2 is a poor prognosis factor

Rembrandts Data TCGA Data

All gliomas Glioblastomas

IGFBP-2 Promotes Motility & Invasion

• IGFBP2 activates expression of invasion enhancing genes and promotes glioma invasion in vitro (Hua Wang et al., Cancer Res., 2003)

MMP2CD10

TIMP-1Fibronectin

Integrin 5Integrin 6

Vinculin

ILK-FAK-PI3K-AKT

Regulated matrix degradation

Guiding migration

Actin stress fiberCytoskeleton reorganization

Migration & survival

Thrombospondin 2

Filamin ABcl-2PUMAp21/WAF1XRCC2

TGF beta RinvasionBradykinin R B2

Centaurin

IGFBP2 Thrombin R

survival

Hua Wang, PhDFirst Prize poster Competition at MDACC Trainee Recognition Day

George Wang, M.D., Ph.D. Limei Hu, M.D., M.S.Resident at Mt Sinai Med Ctr.

IGFBP2 is an OncogeneProc Natl Acad Sci USA104(28):11736-41, 2007

First prize in 2007 Trainee Recognition Day at MDACC

American Legion Auxiliary Fellowship

NIH Training grant Pharmacoinformatics fellowship

Sarah Dunlap (now Sarah Smith)

+1-561

c-Myc AP2 NFB

NFNFBB

IGFBP2IGFBP2 •A review of the literature showed that Cazals et al. (1999) indeed demonstrated that NFB activated the IGFBP2 promoter in lung alveolar epithelial cells.

•Higher NFB activity in IGFBP2 overexpressing cells was also found.

NFNFBB

IGFBP2IGFBP2

TNFR2TNFR2

ILKILK•Our real-time PCR data showed that in stable IGFBP2-overexpressing cell lines, IGFBP2 indeed enhances ILK expression.

•In addition, IGFBP2 contains an RGD domain, implying its interaction with integrin molecules.

•ILK is in the integrin signal transduction pathway.

0

1

2

3

4

5

parental c1 c4 c5 c8

Rela

tive

Luci

fera

se A

ctiv

ity

OCT1

p65/p50

p50/p50

IGFBP2 c

lone

Paren

tal

IGFBP2 c

lone

Paren

tal

Non-specific

ILK is elevated in high-grade glioma and correlates with shorter survival

NH2 COOH

IGF binding domains

5’ TCCAGGGAGCCCCCACCATCCGGGGGGACCCCGAGTGTCATCTCTTCT 3’

R G D 306

GAA

E 306

D306E-IGFBP2 (RGE mutant)

Thyroglobulin type-1 motif(TG domain)

DXXDmotif RGD domain

RGE mutational substitution on IGFBP2

Integrin Linked Kinase

IKKα

IGFBP2

ILK

PI3K

AktP

PH

Receptor Tyrosine Kinase

NFĸBIĸB

IĸB

UU

U

NFĸB

Nucleus

RGD

Target genes

PNFĸB

Ligand

1

Ser 473

GSK3

P

Cyclin D

Proliferation

PIP3

Integrins

Integrin Binding is Required for IGFBP2-mediated Progression

n=500

5

10

15

20

25

30

35

40

45

50

RCAS Combination

Gra

de

III I

nc

ide

nc

e (

%)

n=42 n=32n=50

GFP

PDGFB

PDGFB IGFBP2

PDGFB IGFBP2(RGE)

n=22

IGFPB2 Drives Progression via ILK

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

RCAS Combination

Gra

de

III I

nc

ide

nc

e (

%)

n=42 n=50 n=28 n=26 n=22

PDGFB

PDGFB ILK

PDGFB IGFBP2

PDGFB ILK-KD

PDGFB ILK-KD IGFBP2

IGFBP2-ILK-AKT pathway

• Critical for cancer development and progression

• Opportunities for drug development

• Systems understanding to cancer and cancer therapeutics

• Predictive instead of reactive medicine

• TCGA is making a major impact on individualized medicine

Future Cancer Biology

Acknowledgment• NIH/NCI GDAC Center grant (Shmulevich/Zhang)• NIH/NCI RO1 CA098503 (Zhang/Fuller)• NIH/NCI NIH R01 CA141432 (Zhang/Fuller)• NIGMS/NIH R01 GM072855 (Shmulevich/Zhang)• Goldhirsh Foundation (Zhang) • James S McDonald Foundation (Zhang)• National Foundation for Cancer Research (Zhang/Hamilton)• NFCR Hope Fund (Zhang)• Anthony Bullock III Research fund (Zhang/Fuller/Sawaya)• The Oreffice Foundation (Zhang/Fuller/Sawaya)• Commonwealth Foundation for Cancer Research (Zhang/Trent)

• NIH/NCI NIH R01 CA098570 (Zhang/Pollock, completed)• NIH R21 GM070600 (Shmulevich/Zhang/Kauffman, completed)• Department of Defense (Zhang, completed)• Texas Higher Education Coordinating Board ARP and ATP grants (Zhang/Fuller, Zhang/Holland, completed)• RGK Foundation (Zhang, completed)

• NIH/NCI P30 CA016672-28 (CCSG)• Tobacco Settlement Fund• Kadoorie Foundation• Goodwin Fund