Top Banner
David W. Salzman Yale School of Medicine Department of Therapeutic Radiology Fallopian Mucosa NSCLC Papillary Serous Ovarian Tumor Dicer mAb (13D6) miRNA-Target Site SNPs as Predictors of Cancer Risk and Treatment Response
31

miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

Nov 28, 2014

Download

Health & Medicine

David Salzman

9th Annual miRNA Meeting: To
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

David W. SalzmanYale School of Medicine

Department of Therapeutic Radiology

David W. SalzmanYale School of Medicine

Department of Therapeutic Radiology

Fallopian Mucosa NSCLC Papillary Serous Ovarian TumorD

ice

r m

Ab

(1

3D

6)

miRNA-Target Site SNPs as Predictors of Cancer Risk and Treatment Response

miRNA-Target Site SNPs as Predictors of Cancer Risk and Treatment Response

Page 2: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

Need for new companion diagnostic markers in cancer therapyNeed for new companion diagnostic markers in cancer therapy

• Tumors contain a heterogeneous array of inherited and acquired mutations

• Best cure rates are achieved when specific drugs are used to target tumors with particular mutations (targeted therapeutics)

• Current strategies to identify targeted therapeutics rely heavily on the identification of tumor acquired mutations

• May only be represented in a small population of cells – elude identification• Short-term response is good, long-term response is poor due to drug resistance

Therefore: we need a new paradigm to identify companion diagnostics that do NOT rely on identifying tumor acquired mutations

but where do you find such mutations in the genome?

Page 3: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

microRNA-Target Sites:uncharted territory to identify disease biomarkers

microRNA-Target Sites:uncharted territory to identify disease biomarkers

• The 3’UTR of mRNAs contain cis-regulatory elements that regulate the nature and timing of gene expression in conjunction with a requisite trans-acting factor

• MicroRNAs are a class of non-coding, trans-acting RNAs that negatively regulate gene expression by binding to complementary elements in the 3’UTR of a target mRNA

5'5' 3'3'AAAAAAAAAAAA

ORFORF5’UTR5’UTR 3’UTR3’UTR

ProteinProtein

Page 4: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

miRNA-mRNA complementarity is required for target selectionmiRNA-mRNA complementarity is required for target selection

AGOAGO3'3' 5'5'

• Seed pairing = complementarity between nucleotides 2-7• ± complementarity of nucleotide 8• ± Adenosine residue opposite nucleotide 1• ± 3’ end complementarity

5'5' 3'3'AAAAAAAAAAAA

ORFORF5’UTR5’UTR 3’UTR3’UTR

ProteinProtein

3’ NNNNNNNNNNNNNNNNNNNNNN 5’ |||||||||| ||||||| 5’...NNNNNNNNNNNNNNNNNNNNNNNNNNNN...3’

3’ NNNNNNNNNNNNNNNNNNNNNN 5’ |||||||||| ||||||| 5’...NNNNNNNNNNNNNNNNNNNNNNNNNNNN...3’

Page 5: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

miRNA-mRNA complementarity is required for target selectionmiRNA-mRNA complementarity is required for target selection

AGOAGO3'3' 5'5'

5'5' 3'3'AAAAAAAAAAAA

ORFORF5’UTR5’UTR 3’UTR3’UTR

ProteinProtein

Translation InhibitionTranslation Inhibition

XX

• Seed pairing = complementarity between nucleotides 2-7• ± complementarity of nucleotide 8• ± Adenosine residue opposite nucleotide 1• ± 3’ end complementarity

Page 6: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

miRNA-mRNA complementarity is required for target selectionmiRNA-mRNA complementarity is required for target selection

AGOAGO3'3' 5'5'

• Centered pairing = complementarity between nucleotides (approx.) 6-16• Must include Watson-Crick base-pairing between nucleotides 9-12

5'5' 3'3'AAAAAAAAAAAA

ORFORF5’UTR5’UTR 3’UTR3’UTR

ProteinProtein

3’ NNNNNNNNNNNNNNNNNNNNNN 5’ |||||||||||||| 5’...NNNNNNNNNNNNNNNNNNNNNNNNNNNN...3’

3’ NNNNNNNNNNNNNNNNNNNNNN 5’ |||||||||||||| 5’...NNNNNNNNNNNNNNNNNNNNNNNNNNNN...3’

Page 7: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

miRNA-mRNA complementarity is required for target selectionmiRNA-mRNA complementarity is required for target selection

5'5'3'3'

AAAAAAAAAAAA

ORFORF5’UTR5’UTR

3’UTR3’UTR

ProteinProtein

XX mRNA cleavagemRNA cleavage

• Centered pairing = complementarity between nucleotides (approx.) 6-16• Must include Watson-Crick base-pairing between nucleotides 9-12

Page 8: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

Hypothesis: variants in microRNA-target sites can deregulate gene expression and result in cancer

Hypothesis: variants in microRNA-target sites can deregulate gene expression and result in cancer

XXSNPSNP

inhibit oncogene targeting

inhibit oncogene targeting

enhance (or lead to aberrant) targeting of

a tumor suppressor

enhance (or lead to aberrant) targeting of

a tumor suppressor

XXSNPSNP

over expression ofoncogene

over expression ofoncogene

under expression of tumor suppressor

under expression of tumor suppressor

Page 9: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

Identification of a germline let-7 target site variant (rs61764370)Identification of a germline let-7 target site variant (rs61764370)

5' 3'AAAAAAKRAS5’UTR 3’UTR

1 9 2 34 5 610

78

LCS6 Genotype Tumor and NAT Tumor

TT 35 24

TG/GG 8 7

n=74 (NCSLC patients)

G-allele present in (approx.) 20% of lung cancer patients (otherwise KRAS ORF WT)

Chin et al, Cancer Research (2008)

• let-7 targets the KRAS 3’UTR• 10 predicted let-7 complementary sites

Page 10: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

rs61764370 associates with cancer riskrs61764370 associates with cancer risk

Cancer Subtype (subgroup) GenotypeFold-increased risk

(OR, 95% CI; p-value) Reference

Lung Non-small cell lung cancer (NSCLC) TG/GG 2.3 (1.1-4.6; p=0.02) Chin et al, 2008

Ovarian Hereditary breast ovarian (HBOC) TG/GG 2.46 (1.14-5.29; p=0.02) Ratner et al, 2010

Breast Triple negative (ER-/PR-/Her2-) TG/GG 2.307 (1.261-4.219; p=0.0067) Paranjape et al, 2011

multivariate analysis – adjusted for age and race

• Over 40,000+ individuals studies worldwide• Represented in (approx.) 6% of world populations• More frequently associated with cancer in women• Associated with later onset for most patients

Page 11: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

rs61764370 associates with poor OS in HNSSCrs61764370 associates with poor OS in HNSSC

Christensen et al, Carcinogenesis (2009)

TT (wild type)TG/GG (variant)n=344

All sites (oral, pharyngeal, laryngeal)HR (CI 95%): 1.6 (1.0-2.5, p=0.20)

TT (wild type)TG/GG (variant)n=190

Oral cancerHR (CI 95%): 2.7 (1.4-5.3, p=0.06)

• rs61764370 is present in (approx.) 17.5% of this HNSSC cohort• Treatment not detailed

Page 12: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

rs61764370 associates with poor OS in ovarian cancerrs61764370 associates with poor OS in ovarian cancer

Ratner et al, Oncogene (2011)

Variable HR 95% CI p-value

KRAS mutation 1.671 1.087 - 2.568 0.0192

Age 1.025 1.002-1.049 0.0307

Stage 1.380 1.185-1.607 <0.0001

Grade 1.341 0.912-1.972 0.1360

Histology 0.970 0.900-1.045 0.4168

Center (Yale vs non-Yale) 1.868 1.438-2.427 <0.0001

TT (wild type)TG/GG (variant)n=279

Ovarian cancerHR (CI 95%): 1.671 (1.087-2.568, p=0.0192)

Page 13: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

rs61764370 confers platinum resistance in ovarian cancerrs61764370 confers platinum resistance in ovarian cancer

Adapted from Ratner et al, Oncogene (2011)

Genotype Univariate Multivariate

OR 95% CI p-value OR 95% CI p-value

Wild type (TT) (n=225) 1.00 1.00

Variant (TG/GG) (n=66) 2.45 1.08-5.53 0.0313 3.18 1.31-7.72 0.0106

Multivariate analysis: adjusted for age, stage, grade, histology, residual disease after cytoreductive surgery and treatment center

Cell LineGemcitabin

e Doxorubicin Topotecan

BG1 (TG variant) 30.4uM 307.5nM 161.8nM

CAOV3 (TT wild type) 2.2nM 75.9nM 30.8nM

p=<0.0001

p=<0.04 (TG variant)(TT wild type)

(TG variant/BRCA1 MT)

1st line therapy 2nd line therapy

Page 14: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

rs61764370 is sufficient to up-regulate KRAS gene expressionrs61764370 is sufficient to up-regulate KRAS gene expression

p=0.007

p=0.036TT TG

KRAS

Actin

TT TG

LungNormal

Lung1o Tumor

Page 15: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

rs61764370 positive tumors display unique transcription patternsrs61764370 positive tumors display unique transcription patterns

Expression SignatureTT vs TG/GGTNBC Tumor pK-S Test

NRAS up 0.02

BRCA mutant-like up 0.04

Luminal Progenitor up 0.04

MAPK Creighton up 0.06

PCA Estrogen down 0.04

Adapted from Paranjape et al, Lancet Oncology (2010)

The let-7 family of microRNAs is also consistently and significantly down-regulated in rs61764370 positive (NSCLC, TNBC, ovarian and HNSSC) tumors

Adapted from Ratner et al, Oncogene (2011)

TTTG/GGn=10, p=0.095

TTTG/GGn=10, p=0.095

TTTG/GGn=10, p=0.05

Page 16: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

Targeted therapeutics in the EGFR signaling pathwayTargeted therapeutics in the EGFR signaling pathway

EGFR

Proliferation

RASRAS

RAFRAF

MEKMEK

MAPKMAPK

anti-EGFRGefitinibErlotinib

Cetuximab

T790ML858R

G12DQ61L

V600Eanti-BRAFSorafenib

anti-MEKSelumetinibTrametinibAZD6244

Page 17: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

rs61764370 associates with poor OS in mCRC patients undergoing cetuximab-irinotecan salvage therapy

rs61764370 associates with poor OS in mCRC patients undergoing cetuximab-irinotecan salvage therapy

anti-EGFRCetuximab

TT (wild type) n=100TG/GG (variant) n=34p=0.001

Graziano et al, Pharmacogenomics J. (2010)

*Patient cohort was otherwise KRAS ORF WT and BRAF ORF WT

EGFR

Proliferation

RASRAS

RAFRAF

MEKMEK

MAPKMAPK

rs61764370

Page 18: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

The germline rs61764370 3’UTR variant phenocopies a tumor acquired KRAS ORF mutation

The germline rs61764370 3’UTR variant phenocopies a tumor acquired KRAS ORF mutation

KRASKRAS5' 3'

Tumor AcquiredKRAS ORF Mutation

X

X

Canceranti-EGFR Rx resistance

KRASKRAS5' 3'

Germ-lineKRAS rs61764370 T>G

X

X

Canceranti-EGFR Rx resistance

KRASKRAS5' 3'

WT KRAS

Normal anti-EGFR Rx sensitive

let-7 RISC let-7 RISClet-7 RISC

Page 19: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

Future directions

Clinicopathology

Hypotheses

Genotype

Summary and Current Work FlowSummary and Current Work Flow

rs61764370rs61764370

Increased cancer riskIncreased cancer risk Altered Responseto Therapy

Altered Responseto Therapy

rs61764370 positive cells will display

enhanced cancer-associated phenotypes

rs61764370 positive cells will display

enhanced cancer-associated phenotypes

We can selectively target rs61764370

positive cells

We can selectively target rs61764370

positive cells

rs61764370 is a predictive biomarker to direct cancer therapy

rs61764370 is a predictive biomarker to direct cancer therapy

Clinical trialsClinical trials

Isogenic cell lines

High throughput screening of FDA-

approved compounds

High throughput screening of FDA-

approved compounds

Cell biology: Transformation, growth, Mobility, Invation, EMT

Cell biology: Transformation, growth, Mobility, Invation, EMT

Page 20: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

Generation of Isogenic KRAST/T and KRAST/G Cell Lines(workflow)

Generation of Isogenic KRAST/T and KRAST/G Cell Lines(workflow)

Obtain cell line from NCI (KRAST/T)

Transfect with•zinc-finger plasmids (x2)•donor plasmid

A B C3 2 1

{

(3) zinc-fingers/nuclease(highly specific DNA binding)

bidentate nucleases(dsDNA cleavage)

KRAS 3’UTR

dsDNA cleavage

DNA repair

Homologousrecombination

Donor(mutant KRAS 3’UTR)

A B C3 2 1

DNA binding

Page 21: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

Generation of Isogenic KRAST/T and KRAST/G Cell Lines(workflow)

Generation of Isogenic KRAST/T and KRAST/G Cell Lines(workflow)

Obtain cell line from NCI (KRAST/T)

Transfect with•zinc-finger plasmids (x2)•donor plasmid

Single cell clone

Expand

1 2 3 4 5 6 7 8 9 10 11 12

A

B

C

D

E

F

G

H

1 2 3 4 5 6 7 8 9 10 11 12

A

B

C

D

E

F

G

H

X XXX X

Exclude wells w/ >1 cell

Expand

Extract gDNA

Page 22: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

Generation of Isogenic KRAST/T and KRAST/G Cell Lines(workflow)

Generation of Isogenic KRAST/T and KRAST/G Cell Lines(workflow)

Obtain cell line from NCI (KRAST/T)

Transfect with•zinc-finger plasmids (x2)•donor plasmid

Single cell clone

Expand

Screen for rs61764370 insertion into KRAS 3’UTR

ID KRAST/G

•sequence verify•expand•store

TaqMan genotype for rs61764370 (ID positive clones)

PCR amplify KRAS 3’UTR(1kb +/- donor sequence)

Topo clone PCR amplicon

TaqMan genotype bacterial colonies for rs61764370

Cell Line Cell type rs61764370 (T:G)

Cal27 (+ control) Lung  12:16

MCF10a-luc Parental Normal breast epithelial 12:0

MCF10a-luc Isogenic Normal breast epithelial 15:15

HCC1937 Parental TNBC (BRCA1-/-) 12:0

HCC1937 Isogenic TNBC (BRCA1-/-) 22:19

H1299 Parental Lung (P53-/-) 10:0

H1299 Isogenic Lung (P53-/-) 15:16

If KRAST/G then allele frequency = 50:50 (T:G)

14 cell lines failed to make isogenic pairs2isogenic cell line generation is ongoing process

anal

ysis

of

posi

tiona

l ins

ertio

n

Page 23: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

Sequence Verification of MCF10a Isogenic Cell LinesSequence Verification of MCF10a Isogenic Cell Lines

ReverseSequence

Reads

Allele-1 Allele-2

MCF10a WT (KRAS 3’UTRT/T)

ForwardSequence

Reads

‘T-allele’Allele-1

‘G-allele’Allele-2

MCF10a MT (KRAS 3’UTRT/G)

Allele-1 A T C

Allele-2 T TC

Allele-1 A T C

Allele-2 G TC

* * * *

** *

*

rs612587 rs61764370 rs2966 rs612587 rs61764370 rs2966

Page 24: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

MCF10a rs61764370 positive cells senesce in cultureand display a mesenchymal-like morphology

MCF10a rs61764370 positive cells senesce in cultureand display a mesenchymal-like morphology

Page 25: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

Knock-in of rs61764370 into MCF10a cells causedan epithelial-to-mesenchymal transition (EMT)

Knock-in of rs61764370 into MCF10a cells causedan epithelial-to-mesenchymal transition (EMT)

Page 26: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

Knock-in of rs61764370 into MCF10a caused a mild growth defectKnock-in of rs61764370 into MCF10a caused a mild growth defect

0

0.2

0.4

0.6

0.8

1

1.2

0 20000 40000 60000 80000 100000

Cell number

Absorbance

MCF10a WT MCF10a MT

Page 27: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

Sequence verification of HCC1937 isogenic cell linesSequence verification of HCC1937 isogenic cell lines

‘T-allele’Allele-1

‘G-allele’Allele-2

HCC1937 MT (KRAS 3’UTRT/G)

ReverseSequence

Reads

Allele-1 Allele-2

HCC1937 WT (KRAS 3’UTRT/T)

ForwardSequence

Reads

Allele-1 T C

Allele-2 T TC

T-Allele T C

G-Allele G TC

CC

* ** *

*

***

rs612587 rs61764370 rs2966 rs612587 rs61764370 rs2966

Page 28: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

HCC1937 rs61764370 positive cells display altered platting efficiency and cell growth

HCC1937 rs61764370 positive cells display altered platting efficiency and cell growth

2D

Nu

mb

er o

f co

lon

ies

HCC1937 WT

HCC1937 MT

HCC1937 WT

HCC1937 MT

100

3D

Nu

mb

er o

f co

lon

ies

HCC1937 WT

HCC1937 MT

HCC1937 WT HCC1937 MT

Page 29: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

MEK inhibitors target HCC1937 rs61764370 positive cellsMEK inhibitors target HCC1937 rs61764370 positive cells

anti-MEKSelumetinibTrametinibAZD6244

HCC1937 MT (TG)

HCC1937 WT (TT)

EGFR

Proliferation

RASRAS

RAFRAF

MEKMEK

MAPKMAPK

rs61764370

Page 30: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

BATTLE1: Biomarker-integrated Approaches of Targeted Therapy of Lung Cancer Elimination

BATTLE1: Biomarker-integrated Approaches of Targeted Therapy of Lung Cancer Elimination

0-wt (Event/N = 32/32)1-Variant (Event/N = 4/4)p=0.001

(Erlotinib)

0-wt (Event/N = 39/61)1-Variant (Event/N = 3/8)p=0.056

(Sorafenib)

anti-EGFRErlotinib

anti-BRAFSorafenib

EGFR

Proliferation

RASRAS

RAFRAF

MEKMEK

MAPKMAPK

rs61764370

Page 31: miRNA-Target Site SNPs as Predictors for Cancer Risk and Treatment Response

AcknowledgementsAcknowledgements

The Weidhaas Lab

CollaboratorsFrank Slack, Roy Herbst – Yale UniversityNicola Miller, Michael Kerin – National University of IrelandKim Smits, Manon van England - Maastricht UniversityJeffrey Weitzel – City of HopeRob Pilarski – OHSUKen Offit – MSKCCChristine Chung – JHMISabine Tejpar – LeuvenXifeng Wu, Hai Tran - MDACC

FundingNIH/NCI RO1NIH K08Yale Cancer CenterMary K. Ashe FoundationShanon FoundationRTOG Seed GrantsCT State Funding