NCI Projects Relevant to LINCS

Jennifer Couch, Ph.D.Daniela Gerhard, Ph.D.

National Cancer Institute, NIH

A Long History of Using Cell Lines to study disease: NCI 60

Cancer is Complicated

• Genes and Genetics• Complex Signaling Networks• Multiple Cellular Processes • Microenvironment• Host Systems• Environmental Factors• Population Dynamics

Time - Progression

His

tolo

gy V

aria

tion

Initiation Progression Metastasis Recurrence

Systems BiologyComputational

Modeling

Data & Information - Clinical, Biological, Epidemiological

Discovery and Knowledge- Basic and translational

Centers for Cancer

Systems Biology (CCSB)

The ICBP Approach

Initiation

Growth

Micro-environme

nt

Progression

Metastasis

Treatment

Regression

Recurrence

GrayCancer heterogeneity; molecular signaling and combinational therapyBreast

Lauffenburger Quantative modeling of critical cancer process; growth, migration, DNA repairBrain, lymphoma, Breast

Computer

simulations

Statistical mining

Game theory

Bayesian networks

Boolean Models

Markov chains

Differential

Equations

Popu

latio

n/

envi

ronm

ent

Org

anis

mTi

ssue

/ org

an

Mic

ro-

envi

ronm

ent

Cell

(com

pone

nts)

Inte

ract

ions

Mol

ecul

es

GolubIntegrate multiple data types to identify essential gene- new drug targetsLung Melanoma

HuangEpigenetic and microRNA examination of hormonal –chemo resistanceProstate, Ovarian Breast

PlevritisCell – cancer differentiation; single cell analysis; disease progression and resistanceLeukemia, Lymphoma

FriendCross disease analysis; Sage Bionetworks; network data integrationGlioma, Ovarian, colon, Liver, Meduloblastoma, Pancreas, Breast, ect

Sander3-D model cell-cell communication; microenvironment; variability of drug responseMelanoma, breast, panceas, glio, CLL

ClarkeHormone responsiveness; population exposure risk; GWAS; drug designBreast

WongTumor Stem cell, imaging, ME interations; protein cDNA arrayBreastHlatky

Tumor evolution; quanatative analysis; cellular population dynamic modelingProstate, Breast

QuarantaCellular model of tumor heterogeneity; cell phenotype measurements; multiscaleBreast, Lung

ICBP2:Centers for Cancer Systems Biology

Cancer Stage

Modeling Approach Physical Scale

CalifanoMultiscale analysis of Genomic and Cell Networks (MAGNet), B cell lymphoma, glioma

Glucose

GLUT4

G6P

F6P

F1-6BPGADP

1,3 BPG

3PG

2PG PEP Pyruvate

Glucose

HK2

PFK1

PK-M2

Acetyl-CoA

PDH

TCA-Cycle

AKT

AS160

AMPK

RTK

HIF1a

PDHK1

Lactate

LDH

c-Myc

CitrateAcetyl-

CoA

MalonylCoA

Fatty Acid Synthesis

ACL

ACC

FAS

LKB1AMPATP

GF

SHP2/SHC/SOS/

GRB2

Ras

MEK1,2

RSK

PI3K

PDK1

GSK3b

TSC2

MAPK1,2

mTOR

Rheb

p70S6K

S6

Raf

BAD

Fos Jun

CREB

AP1

ATF2RB

Glycolysis

Gluconeogenesis

Cellular respiration

Rac

Rho

cdc42

MEKK4

MEK4

MEKK6

MEK6

JNK/SAPK p38

C-Src

Stat3/5

Stat3/5

SGK3

GS

Glycogen synthesis

FOXO3A

ER

ER

CBX5

• Therapeutic perturbations of a breast cancer cell line panel show:

• Pathway function and mechanism of deregulation differ according to subtype

• Responses are subtype specific• Responses are not durable and

mechanisms are not understood

• Therapeutic responses depend on the chemical and mechanical microenvironment

Lawrence Berkeley Labs (Gray)Model Based Predictions of Responses to RTK Targeted

Therapies in Breast CancerOverall: Infer signaling network for cancer subtypes; model response to MAPK inhibitors, Her2 targeted therapies and P13K targeted therapies

Project 3 :HER-family signaling deregulated in breast cancer

ICBP: Efforts to Expand the Field• Community resources

– Software tools and models– Data sharing/Data portal– Biological (ICBP 43 cell lines)

• Educational /Outreach– Undergraduate opportunities– Junior PI training and mentoring– Curricula development

• Meetings/ workshop– PI meetings– Joint meetings (other consortia, etc.)– Junior Investigator meeting (yearly)– Mathematical Cancer Modeling workshop (2010, 2012)– National Cancer Systems Biology meeting/ AACR (March 2011)

ICBP 43

http://physics.cancer.gov/about/

Physical Sciences-Oncology Centers (PS-OCs) Network: Cell Line Project

• Selected human immortalized cell lines– MCF-10A: non-malignant breast epithelial cells– MDA-MB-231: metastatic breast cancer cells

• Scope of project was to have each PS-OC conduct their unique physical science measurements on the cell lines and share and cross-compare datasets

• Cell lines were propagated by one PS-OC Investigator and distributed to one site at each of the 12 PS-OCs along with detailed SOP for cell culture of each cell line

• Requirement to upload cell line data to a pilot data coordination site• Investigators presented results at Data Jamboree Meeting in January 2011• Cell line project data will serve as a pilot dataset for building the PS-OC

Data Coordination Center (DCC)• Currently a Network manuscript is in preparation describing the results of

the project

Cell Line Project Physical Science Measurements: Molecules to Cells

MIT

Genomics Proteomics

NU

ASU CornellCornell

• TEM and cryo-EM analysis of chromatin structure and distribution of chromatin-organizing proteins • Light scattering and TEM roughness nanoscale measurements• Deep sequencing-based nucleosome position analysis•Metaphase chromosome mechanics• Magnetic tweezer analysis of chromatin polymer elasticity•DNA replication dynamics analysis

Moffitt

MIT

UCB

JHU

• Cell CT• AFM

• Colony size/morphology in response to altered pH and pO2• Cell trace

• Cell mass measurements during cell cycle

• Gene expression analysis in response to changes in pO2, pH and metabolic load in 3D alginate

• FACS for adhesion receptors• Adhesion assay to E-selectin surfaces under shear stress• Aggregation with human leukocytes under flow

• Subcellular release assay to measure rate of de-adhesion • Flow chamber assay to measure cell-matrix adhesion strength

• Single-cell transcript counting

NU

• Proteasome processing analysis

Princeton• Cell morphology as a function of time and position on stressed landscapes

Princeton• Analysis of chromatin and mitochondria as a function of time and position on stressed landscapes•Genomic analysis of cells transported to and from stressed landscapes

Princeton

• Analysis of cell adhesion as a function of time and position on stressed landscapes under flow conditions

Scripps

Scripps • Nuclear dimensions (radius, aspect ratio)

• Cell dimensions (radius, aspect ratio)• Cortical tension• Cytoplasmic viscosity•Elasticity (E)• Cellular complexity/granularity

USC

• SILAC for MS-based proteomics

• ECM stiffness effects on cell polarity and spreading

Methodist

• In vitro nanoparticle •internatlization

Cell Mechanics/Morphology

Cell Surface/Adhesion

JHU• Cadherins via single- molecule force microscopy (2D/3D)• Ballistic Intracellular Nanorheology (BIN; 2D/3D)• Intracellular microrheology for cells in 2D & 3D

The Cancer Target Discovery and Development (CTD2) Network

LINCS Consortium Kick-Off Meeting October 28, 2011

Daniela S. Gerhard, Ph.D. Director, Office of Cancer Genomics

Large Projects Examples of NIH Investment in Genomic Research

• Therapeutically Applicable Research to Generate Effective Treatment (TARGET)

• The Cancer Genome Atlas (TCGA)• Cancer Genome Anatomy Project/Cancer Genome

Characterization Initiative (CGAP/CGCI)• Genome-wide association studies (GWAS) of common and

complex diseases and follow-up (~60/450 grants are cancer-related)

Data generated is made publicly available

• ~20% of NIH ARRA funded genomic projects

Molecular Characterization of Cancer Tissues is Essential but not Sufficient

• Each tumor has hundreds to thousands genomic alterations– Chromosomal changes: amplifications, deletions, translocations

– Epigenetic changes– Mutations

• Little is known about the cellular function of most genes, much less how sequence variants and mutations affect them– Distinguishing initiating vs. driver vs. passenger mutations

• Drivers are defined as genes involved in tumor maintenance • Evidence is accumulating that multiple subclones exist within a tumor and

their frequency varies between patients• As tumors evolve genes essential for survival may be different from those

that were necessary early on– Genomic alterations result in cancer within specific context

• Cell of origin• Other molecular alterations in genes that may have synergistic or

antagonistic impact

ARRA OpportunityQuestion:

Can a network be formed that would effectively address a current major scientific challenge: efficient transition from patient-based large multi-dimensional genomic data target validation small molecule modulators (therapy, not part of the initiative)

How to advantage the flood of genomic data and accelerate the transition to treatments of patients based on the genomic

profile of their cancer?

ARRA Cancer Target Discovery and Development (CTD2) Network Centers

• Broad Institute, Cambridge, MassachusettsPI: Stuart Schreiber, Ph.D.

• Cold Spring Harbor Laboratory, Long Island, New YorkPI: Scott Powers, Ph.D., co-PI: Scott Lowe, Ph.D.

• Columbia University, New York, New YorkPI: Andrea Califano, Ph.D.

• Dana-Farber Cancer Institute, Boston, MassachusettsPIs: William Hahn, M.D., Ph.D., L. Chin, M.D. and R. DePinho, M.D.

• University of Texas Southwestern Medical Center, Dallas, TexasPI: Michael Roth, Ph.D., co-PIs: M. White, Ph.D., J. Minna, M.D.

http://ocg.cancer.gov/programs/ctdd.asp

ARRA CTD2 Network• Each application included up to 3 mature projects

• Functional network formed rapidly

– Component centers share results “in real time” (pre-competitive)

• Established an ethos of data and resource sharing with scientific community upon validation

– IT WG developed file formats for data sharing compatible with Cancer Data Standards Registry and Repository (caDSR) within caBIG

• Enabled experiments, using new data generated by the molecular characterization projects to identify candidate targets, small molecule modulators and mechanisms: one example was ovarian cancer

Cancer Target Discovery and Development (CTD2) Network

cancer genomics small-molecule probesprobe acquired dependencies via proteins

probe acquired dependencies via RNA

probe acquired dependencies via network

analyses

determine relevance (STK33; TBK1)

determine relevance (STAT3; C/EBP in GBM)

(acquired dependency small-molecule probe kit in >400 genotyped cell lines)

cancer genomics-based mouse models

Decode the relationship of cancer genotype to acquired cancer dependencies and identify small molecules that target the dependencies (*Broad; *CSHL; *Columbia; *DFCI; *Dallas)

***

******

**

* **

small-molecule drugs *cancer patients

Summary of the CTD2 Network’s caOv Results

The power of the network: made rapid progress by sharing data, working together and taking advantage of complementary, non-overlapping expertise to carry out the experiments. Each Center contributed:

– Identified candidate signature to stratify patients into best and worst prognostic groups– Identified candidate targets for therapeutic development

• Confirmed a subset of candidates by in vitro and ex vivo experiments– Identified candidate small molecules for a subset of confirmed targets– Plan to generate mouse models for in vivo screening of other candidate genes within a

specific genetic context– Experiments are ongoing

Critical lesson: collaborative efforts to integrate several methods can yield exponential gains relative to the incremental gains achieved through improving any single method (united they are more than a sum of parts)

CTD2 Network Research Mission

• Shift current research paradigms in translation pathway of patient-derived multidimensional genetic data to the clinic and utilize novel concepts, approaches and methodologies

• Develop research that will exert a sustained influence on the

field

• Develop a pre-competitive culture to ensure sharing of data, methods (analytical, experimental) and reagents within the network and the scientific community at large

Goals for the New Network• Accelerate the translation of patient genomic data into clinical

application– Innovate the integration of computational mining large scale genomic

data analyses• Make tools available through web

– Identify and confirm new therapeutic target candidates– Identify and confirm novel modulators within specific cancer context

(cellular or mutational) in vitro (cell lines) or in vivo (cancer models)• Small, stereochemically “interesting“ molecules

– Use of novel organist chemistry – molecules more “natural products-like”– Mature molecules: optimize activity, structure activity relationship, systematic variation of

stereochemistry• siRNAs

– Multi-expertise team – Share models and reagents with the scientific community – Share data and methods with the scientific community through the web

• As genomic data become available from TARGET, TCGA etc.,: be nimble, flexible and open to new opportunities

The Cancer Target Discovery and Development (CTD2) Network

LINCS Consortium Kick-Off Meeting October 28, 2011

Daniela S. Gerhard, Ph.D. Director, Office of Cancer Genomics