DNA Methylation in MDS/MPD/AML: Implications for application James G. Herman, M.D. Professor of Oncology Evelyn Grollman Glick Scholar The Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins
Disclosures Research Support: OncoMethylome Sciences, Celgene Scientific Advisory Board: OncoMethylome Sciences I will not be discussing off-label use.
Methylation changes in Neoplasia Normal 1 2 3 Cancer 1 2 3 When do changes in DNA methylation develop? Is the detection of DNA methylation of prognostic or predictive use? Are epigenetic changes targets for therapy?
Gene Specific Methylation in MDS:p15/CDKN2B M Lubbert, Leukemia 2003, 1762-64
p15/cdkn2b in MDS/AML: Increase with progression and use for predicting survival t-mds t-aml Christiansen, Leukemia, 2003, 1813
p15/cdkn2b in MDS/AML: Interaction with Important Genetic Abnormalities Christiansen, Leukemia, 2003, 1813
Why Has Methylation Been Inconsistent as a Prognostic Marker? No systematic analysis of methylation and prognostic, cytogenetic and molecular markers Gene selection and methods How do we choose the right genes? Does method effect use as marker? Is it presence of methylation or density (level) that affects prognosis? Confounders good and poor outcome groups as the basis for comparison, differences in terms of age, WBC, and cytogenetics.
Study Design- Primary AML Samples (Elizabeth Griffiths, Hetty Carraway) 207 1 o AML samples from the JHU tumor bank, 1998-2008 114 with normal karyotype: 73 diagnostic, 40 relapse specimens Methylation analysis by MSP was performed for: 11 known genes, 12 Wnt pathway inhibitor genes Samples annotated for: age, specimen type, karyotype, WBC at diagnosis, induction therapy, antecedent MDS/MPD, primary refractoriness, FLT3-ITD status and survival. NPM1 mut and FLT3-ITD status assessed for samples without information
Methylation in Cytogenetically Normal and Abnormal AML 90.0 80.0 70.0 60.0 Normal Any Karayotypic Abnormality % Methy la tio n 50.0 40.0 30.0 20.0 10.0 0.0 p15 C DH1 CEBP A* C TNN A1* SOCS1 Gene p73* CEBPd F H IT* E R* MLH 1 MGM T * Statistically different between two groups, p<0.04
Univariate Hazard Ratio for Death: Known MGMT genes (n=73) MLH1 ER FHIT CEBP p73 SOCS1 -Catenin CEBP ECAD p15 0.10 1.00 10.00 Univariate Hazard Ratio
Multivariate Hazard Ratio for Death (n=73) MGMT MLH1 ER FHIT CEBP p73 SOCS1 -Catenin CEBP Corrected for age, NPM1/FLT3-ITD status, antecedent MDS, and WBC at diagnosis. ECAD p15 0.10 1.00 10.00 Multivariate Hazard Ratio
Multivariate Hazard Ratio for Death by Gene: Wnt antagonists WIF1 RUNX3 SOX17 SFRP5 SFRP4 SFRP2 SFRP1 DKK2 DKK1 0.10 1.00 10.00 Multivariate Hazard Ratio n= 73 uniformly treated patients with normal karyotype AML, multivariate hazard ratio corrected for age, antecedent hematologic diagnosis, WBC, FLT3-ITD, NPM1 mut Status
Kaplan-Meier Survival Estimates In patients with Normal Karyotype Proportion Surviving 0.00 0.25 0.50 0.75 1.00 Log Rank p=0.04 0 20 40 60 analysis time Months n=73, p<0.05 sfrp2 = U sfrp2 = M
Why Has Methylation Been Inconsistent as a Prognostic Marker? No systematic analysis of methylation and prognostic, cytogenetic and molecular markers Gene selection and methods How do we choose the right genes? Does method effect use as marker? Is it presence of methylation or density (level) that affects prognosis? Confounders good and poor outcome groups as the basis for comparison, differences in terms of age, WBC, and cytogenetics.
DNA Methylation Predicts Survival and Response to Therapy in Patients With Myelodysplastic Syndromes Shen, JCO, 2009, 605
Quantitative DNA methylation predicts survival in adult acute myeloid leukemia Bullinger, Blood, 2010, 636
Quantitative DNA methylation predicts survival in adult acute myeloid leukemia Bullinger, Blood, 2010, 636
Epigenetic progression of MDS/MPD and AML Normal Marrow Low Risk MDS PV High Risk MDS Myelofibrosis AML ET Gene Methylation Gene A p15 SOCS1 CTNNA1 Gene B Methylation Density
HL60 HNT34 DNA methylation of CTNNA1 in AML cell lines (Ying Ye, Mike McDevitt) MSP U U KG1a M ML-1 U U937 U/M +40 MSP 1 MSP 2 Bisulphite Sequence 0: Transcriptional start site -303 0 +43 +149 +273 + 352
Chromatin immunoprecipitation (ChIP) at CTNNA1 promoter in AML cell lines 20 15 10 5 Fold Enrichment IgG H3K9Ac H3K4me2 H3K9me2 H3K27me3 H3K9Ac H3K4me2 H3K9me2 H3K27me3 IgG H3K9Ac H3K4me2 H3K9me2 H3K27me3 0 IgG IgG H3K9Ac H3K4me2 H3K9me2 H3K27me3 HNT34 HL60 KG1 KG1a CTNNA1 expression CTNNA1 MSP - - - +
DNA methylation and repression of CTNNA1 in primary AML A B NL H2O IVD p010 p052 p078 p151 p195 p209 p223 p239 p240 MW U M U M U M U M U M U M U M U M U M U M U M U M MW NL H2O IVD p002 p006 p045 p024 p027 p032 p033 p040 p068 U M U M U M U M U M U M U M U M U M U M U M U M MW p301 p338 p380 p383 p397 p432 p412 G571 G577 G1042 G1361 U M U M U M U M U M U M U M U M U M U M U M MW p153 p083 p088 p096 p128 p303 p119 p252 P390 P393 U M U M U M U M U M U M U M U M U M U M C P=0.01 D P=0.03 C1 CTNNA1/GAPDH 3 2.5 2 1.5 1 0.5 0 CTNNA1 Normal Patient group A Patient group B (n=7) (n=17) (n=9) MSP - - + C2 C9 C10-250 -200-150 -100-50 0 50
Chromatin immunoprecipitation (ChIP) of the CTNNA1 promoter: Repression is associated with inactive histone modification (Primary AML) CTNNA1/GAPDH 2.5 2 1.5 1 0.5 A CTNNA1 expression by Real Time PCR Fold Enrichment 1.4 1.2 1 0.8 0.6 0.4 0.2 B H3K9Ac 0 1 2 3 4 5 6 7 8 9 10 0 1 1 2 3 4 5 6 7 8 9 10 14 12 C H3K4me2 1.2 D H3K27me3 Fold Enrichment 10 8 6 4 2 Fold Enrichment 1 0.8 0.6 0.4 0.2 0 1 2 3 4 5 6 7 8 9 10 0 1 1 2 3 4 5 6 7 8 9 10
Model of progressive epigenetic inactivation Active chromatin High gene expression Bivalent Chromatin Modest gene expression Compacted chromatin Depressed expression/ Silencing Inactive chromatin Gene silencing H3K9Ac H3K4Me2 H3K9Me2 H3K27Me3 Methylated DNA
Acknowledgements AML prognosis Elizabeth Griffiths Hetty Carraway CTNNA1 studies Ying Ye Michael McDevitt Steve Gore Hetty Carraway Oliver Galm