1. DNA methylation 2. Histone modifications 3. Chromatin structure 4. Imprinting 5. Non-coding RNAs Epigenetics
Topics
DNA " 5-Methylcytosine" Inheritance upon replication" O N NH2 N dr CH3 M" CG" GC" M" M" CG" GC" CG" GC" M" M" CG" GC" M" M" CG" GC" CG" GC" M" CG" GC" DNA replication" CG" GC" CG" GC" CG" GC" CG" GC" CG" GC" Methyltransferase" M" CG" GC" M" M" CG" GC" M" M" CG" GC" M" M" CG" GC" M" CG" GC" CG" GC" CG" GC" CG" GC"
CpG ", ON, OFF CpG CpG "
DNA
[Ushijima and Asada, Cancer Sci, 101:300, 2010]
: MGMT Unmethylated Methylated 40 glioma cases Alkylating agents Pol II Poor response Good response (From Esteller and Herman, Oncogene, 2004)
DNA (CIMP) " Total = 145 cases" N-myc! Amp (+)" = 23 CIMP(+) = 50" N-myc(-) & CIMP(+) " = 27" CIMP(-) = 95" 1.0" 0.8" 0.6" 0.4" 0.2" 0.0" 1.0" 0.8" 0.6" 0.4" 0.2" 0.0" 0" 2,000" 4,000" (Days)" 0" 1,000" 2,000" (Days)" Overall survival" "HR "P" CIMP "9.5 (3.2-28.1) "<0.0001" N-myc!12 (4.9-29) "<0.0001" " B to A "4.5 (1.3-16) "0.020" C to B "4.8 (1.7-14) "0.003" C to A "22 (6.8-69) "<0.0001" Disease-free survival" "HR "P" CIMP "5.4 (2.9-10) "<0.0001" N-myc!3.1 (1.6-6.0) "0.0007" " B to A "5.2 (2.6-11) "<0.0001" C to B "1.1 (0.53-2.3) "0.821" C to A "5.7 (2.6-12) "<0.0001" [Abe, Cancer Res, 65:828, 2005; Abe, Cancer Lett, 247:253, 2007]
Topics
Bisulfite treatment ( ) M M 5'-GCCCGTTGGTTTCCGGCGCAACGGCCGGGAGCCGT-3' Bisulfite treatment 5'-GUUCGTTGGTTTUCGGUGUAAUGGUUGGGAGUUGT-3' Before AEer Demethyla4ng agent Expression Methyla4on- sensi4ve restric4on enzyme Affinity M M 5'-GCCCGTTGGTTTCCGGCGCAACGGCCGGGAGCCGT-3' 3'-CGGGCAACCAAAGGCCGCGTTGCCGGCCCTCGGCA-5' M M HpaII M M 5'-GCCCGTTGGTTTCCGGCGCA 3'-CGGGCAACCAAAGGCCGCGT M M DNA fragmenta4on ACGGCCGGGAGCCGT-3' TGCCGGCCCTCGGCA-5' An4body MBD M M 5'-GCCCGTTGGTTTCCGGCGCAACGGC CGGGAGCCGT-3' 3'-CGGGCAACCAAAGGCCGCGTTGCCGGC CCTCGGCA-5' M M M M 5'-GCCCGTTGGTTTCCGGCGCA-3' 3'-CGGGCAACCAAAGGCCGCGT-5' M M
Bisulfite Step 1" Step 2" Step 3" Sulfonation" Hydrolytic deamination" Alkali desulfonation" NH 2 NH 2 O O N 3 5 HSO 3 -" +" HN H 2 O " NH 4 +" N OH -" N O 1 N dr OH -" O N dr SO 3 - O N dr SO 3 - HSO 3 -" O N dr Cytosine Cytosine sulfonate Uracil sulfonate Uracil Bisulfite:" NaHSO 3" M Bisulfite conversion" M PCR" C! C! C! C! U! T!
DNA (Beck et al. 2010 Yamashita )
Illumina HumanMethylation450 BeadChip 450K (450,000) probes 95% CpG islands 12 samples / slide 41,300 yen / sample human [From Illumina]
HumanMethylation450 BeadChip
DNA MeDIP-CGI microarray R 2 = 0.962 HumanMethylation450 R 2 = 0.993 Me value beta value [Yamashita et al. unpublished]
HOXA5 promoter Agilent CGI microarray Illumina HM450
Topics
+10 Beta
AGS: ASCL2 16 HSC39: ASCL2 628 HSC57: ASCL2 6808 KatoIII: ASCL2 1426 [Yamashita et al. manuscript in prepara4on]
CpG TSS200 CGI TSS1500 HSC39 cells TSS200 S-shore 1stExon [Yamashita et al. manuscript in prepara4on]
DNA (1) (ESCC) 3 MeDIP + Agilent Human CpG island array (24 ) 0.4 Me value, 0-1 52 25 Excel IF screening valida4on
Data of MeDIP- CGI microarray analysis in genomic regions around PAX6 CGIs Chr 11 (p13) ESCC N(- ) pool Differen4ally methylated region N(+)a pool N(+)b pool [Gyobu, Ann Surg Oncol, 18:1185, 2011]
Quan4ta4ve Methyla4on Analysis of Candidate CGIs in Screening Set and Construc4on of ROC Curve Methyla4on level (%) 14 12 10 8 6 4 2 0 CPLX2 N(- ) N(+) M 7 13 U 1 27 Methylated (M) Unmethylated (U) Specificity 0.0 0.2 0.4 0.6 0.8 1.0 p=0.0061 (Fisher s exact test) Threshold 2.0 % 0.0 0.2 0.4 0.6 0.8 1.0 1- sensi4vity 25 CGIs 7 CGIs ( p < 0.05)
Methylation level (%) 30 20 10 0 50 40 30 20 10 0 Screening set (N=48) PAX6 (Threshold 2.0 %) p=0.048 Validation set (N=48) p=0.033 ENST00000363328 (Threshold 8.3%) p=0.044 Node (-) Node (+) p=0.044 Node (-) Node (+) [Gyobu, Ann Surg Oncol, 18:1185, 2011] Combination marker PAX6 M U and or ENST U M C. marker (-) (+) Node metastasis C. marker (-) (+) (-) 10 5 (+) 7 74 Sensi4vity 91% Specificity 67% Accuracy 88%
ENST00000363328 PAX6 Disease free survival Unmethylated N=50 Methylated N=41 p=0.24 All samples Combina4on markers Nega4ve N=17 Methylated N=37 Unmethylated N=54 p=0.014 All samples Time (days) Disease free survival Posi4ve N=74 p=0.006 All samples Time (days) Analysis of Associa4on with Methyla4on Status and Prognosis
HM450 PAX6 promoter CGI microarray HM450 Differen4ally methylated region [Gyobu et al. 2011]
DNA (2) (GC) Illumina HumanMethyla4on450
MLN vs Primary GC without LNM Mean methyla4on Level 1.0 0.8 0.6 0.4 0.2 Metasta4c Lymph Nodes ( n = 3 ) 480,000 CpG sites 762 candidates CpG sites Ø Methyla4on level < 0.2 in all 3 GCs without LNM Ø Methyla4on level > 0.5 in all 3 Metasta4c lymph nodes 31 CpG sites 0 0 0.2 0.4 0.6 0.8 1.0 Primary GCs without LNM ( n = 3 ) Mean methyla4on Level
Methods for analysis of candidates 31 CpG sites CpG sites Culng Site by methyl cytosine sensi4ve restric4on enzyme, MspJI MSP primers PCR primers 10 CpG sites 21 CpG sites qmsp qpcr-mr
Validated one CpG site (cg06436185) Methylation level 100 80 60 40 20 Meta 0 cg06436185 28.8 (- ) (+) Primary PRKAG2 cg06436185 Exon 1 2 3 4 5 6 7 8-15 CpG island Sensi4vity 43% Specificity 85% [Shigematsu, Oncol Lett, 2012]
HM450 HumanMethylation450 R 2 = 0.993 beta value
HumanMethyla4on450 intensity intensity 1000 1 0.99 0.98 0.97 0.96 0.95 0.94 0.93 0.92 0.91 0.9 MCF7_demo S1 S12 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 Intensity Intensity500
Es4ma4on of the probe errors and their histogram Repeated measurements are available for 10 individuals. The within- sample varia4on is es4mated by use of "analysis of variance" (ANOVA). Quan'le Error 95% 0.074 96% 0.079 97% 0.086 98% 0.096 99% 0.117 99.5% 0.144 99.9% 0.222 [Rehnberg et al. manuscript in prepara4on]
20 AGS+5- aza- dc error diff cg14341177 0.044 0.894 cg01618928 0.064 0.709 cg03281139 0.071 0.698 cg19095568 0.060 0.683 cg09220326 0.049 0.678 cg16402006 0.050 0.677 cg04016485 0.046 0.657 cg20455854 0.044 0.649 cg09373676 0.041 0.643 cg20695611 0.052 0.642 cg18764836 NA 0.640 cg05255994 0.036 0.635 cg08635395 0.042 0.627 cg13321688 0.072 0.626 cg09312135 0.021 0.624 cg14251216 0.249 0.603 cg11481582 0.040 0.600 cg02927821 0.031 0.594 cg20519665 0.051 0.587 cg22950831 0.035 0.580 [Rehnberg et al. manuscript in prepara4on] SH- SY5Y+CDDP error diff cg06647360 0.033 0.884 cg26446133 0.031 0.868 cg16066505 0.370 0.790 cg25720930 0.455 0.679 cg00040566 0.045 0.628 cg07296841 0.290 0.622 cg18443450 NA 0.562 cg23476209 0.230 0.500 cg25376032 0.232 0.487 cg19871348 0.178 0.471 cg02076747 0.134 0.466 cg11822772 NA 0.458 cg05393828 0.243 0.446 cg18825597 0.231 0.433 cg13769609 0.072 0.433 cg06813250 0.159 0.432 cg04054303 0.226 0.427 cg07565150 0.193 0.423 cg06255004 0.106 0.415 cg03562350 0.291 0.410
Biologically DNA
Topics
Flexibility qualitative low Restrictionenzyme CpG any High Existence MSP Quantitative analysis 10% 5% Southern blot COBRA RE- Real- 4me PCR Bisulfite sequencing Pyrosequencing MassARRAY 1% quantitative MethyLight Real- 4me MSP
Check Point of MSP Primer location Quantity and quality of template PCR conditions
A B NFR Exon1 Transcription C D E Unmethylated CpG Methylated CpG NFR : nucleosomefree region CpG island F G ~
CpG NCBI Map Viewer CpG
Bisulfite ü Fragmenta4on of genomic DNA BamHI or Sonica4on Adequate fragmenta4on ü Our protocol 1. 3.1M sodium bisulfite, 6mM hydroqinone, ph 5.0 2. [95, 30 sec 50, 15 min] 15 cycles (x 50 16 hours) 3. Desulfona4on and purifica4on using Spin column (Zymo Research) ph High concentra4on bisulfite, high temperature Commercial kits (Invitrogen or Zymo)
[GENETYX-MAC: Search Restriction Map] Date : 2009.11.17 Filename : GATA2.gb Sequence Size : 2001 Sequence Position: 1-2001 MSP GATA2 [ Linear ] 500 1000 1500 2000 GpC (208) CpG (142) HpaII ( 19) Selected Enzymes With No Recognized Positions: BamHI (Top strand) 5'- T T A C G T G A A C A A A T A G C T G A G G G G C G G C C G G G C C A G A A C G -3' Bisulfite treatment Methylated DNA Unmethylated DNA M primer (forward) A T A A A T A G T T G A G G G G C G G T C 5'- t t a C G t g a a u a a a t a g u t g a g g g g C G g t C G g g u u a g a a C G U primer (forward) G A A T A A A T A G T T G A G G G G T G G T T 5'- t t a U G t g a a u a a a t a g u t g a g g g g U G g t U G g g u u a g a a U G -3' -3' CG site on primer 3' end =1, others = 1 to 4
PCR Template DNA Bisulfite DNA (ng, OD260) Copy numbers (in 1µL DNA solution) 1400 1200 1000 800 600 400 200 0 Treated Untreated 16000 14000 12000 10000 8000 6000 4000 2000 0 NOL4 NTRK2 Treated Untreated Treated Untreated 14,000 copies (50 ng) 700-1,700 copies (5-12%) Inevitable (Any kit!) 10 copies for DNA detec4on / 1 MSP reac4on = 200 copies (100 cells) before bisulfite treatment = 0.6 ng (human, mouse) / 1 MSP reac4on Our protocol: DNA 1 µg (40-80 PCRs) = 12-25 ng / 1 MSP
Annealing Temperature Primer for unmethylated DNA Methylated control Unmethylated control Annealing temperature 55 56 58 61 64 67 69 70 55 56 58 61 64 67 69 70 Gradient PCR system is useful.
Excessive Number of PCR Cycles PCR 40 cycles PLS30 Unmethylated control 1 1/10 1/100 1/1000 Methylated control 1 1/10 1/100 1/1000 M U M U M U M U M U M U M U M U M U 0.1 % of methylated DNA can be amplified. 0.1% methylated in cancer?? Minimum cycles to obtain visible bands with positive control samples 4 cycles were added for test samples (total of 30-36 cycles).
Before 5-Aza-2'- deoxycytidine" (5-aza-dC, DAC)" 5- Aza- dc HOH 2 C" HO" O" O" N" NH 2 " N" N" Re-expression by demethylation Protocol OD 260 nm 3 5-aza-dC 0.1 to 10 µm Day 0 1 2 3 4 120% 100% 80% 60% 40% Harvest Stability of 5-aza-dC in PBS After 20% 0% 4 C 37 C 0 1 2 3 4 incubation time (day)
5- Aza- dc Up-regulation by demethylation Up-regulation by something else cell growth (%) 100 80 60 40 20 5-aza-dC 5-aza-dC Not efficient Efficient Not efficient 0 0 1 2 3 4 4me (day) Low 0 High 5-aza-dC concentration
5- Aza- dc /β-actin 10-2 ANXA5 10-2 CAV1 10-3 SCRN1 8 x4.2 x1.2 4 x6.4 x2.3 x15 x32 2 6 3 4 2 1 2 1 0 5- aza- dc MSP M MSP U AGS KatoIII (- ) (+) (- ) (+) 0 AGS KatoIII (- ) (+) (- ) (+) 0 AGS KatoIII (- ) (+) (- ) (+) /β-actin 10-4 F2R IRS2 CTSL 10-4 10-4 10-3 5 4 x6.0 x37 x130 x1800 x20 x160 x320 x230 6 4 3 4 3 2 2 2 1 1 TFAP2C 5- aza- dc MSP M MSP U AGS KatoIII (- ) (+) (- ) (+) 0 AGS KatoIII (- ) (+) (- ) (+) 0 AGS KatoIII (- ) (+) (- ) (+) 0 AGS KatoIII (- ) (+) (- ) (+) [Nakajima et al. IJC 2009]
Quantification of DNA Methylation Reliability Methylation level as marker (Cancer diagnosis, risk) Number of molecules Low level methylation Precision Flexibility Ease of use Cost Bisulfite sequencing Combined bisulfite and restriction analysis (COBRA) Pyrosequencing MethyLight Real-time methylationspecific PCR (SYBR Green I) MSP
Principle of Real-time MSP Bisulfite treatment = Methyla4on level (%) # of M # of U + # of M x 100 Primer for " unmethylated" (universal) " Real- 4me PCR (SYBR Green)" Primer for " methylated " Fraction of cells with methylation" # of DNA molecules " Unmethylated " Methylated " 10 6" 10 4" 10 2"
Real-time MSP needs DNA [Quantification of high copy number of DNA is stable.] Amplifica4on curve 10 6 10 5 10 4 10 3 10 2 Signal (Log) 10 1 PCR cycles [PCR Template DNA is lost by bisulfite treatment.] DNA needed DNA before bisulfite treatment 1% precision 100 copies more than 2,000 copies (6 ng) 0.1% precision 1,000 copies more than 20,000 copies (60 ng)
Condi4on of Real- 4me MSP PCR with SYBR 32 cycles 40 cycles Methylated control M Primer " Unmethylated control 54 57 60 63 54 57 60 63 ( C) Methylated control U Primer " Unmethylated control 54 57 60 63 54 57 60 63 ( C) Confirma4on by real- 4me PCR Amplifica4on curve Mel4ng curve Calibra4on curve Signal (Log) 10 6 10 5 10 4 10 3 10 2 10 1 - d(signal) / d(temp) Ct value PCR cycles Temp. Copy number (Log) 0.999 PCR efficiency 98.3%
Quan4ta4ve Methyla4on Analysis Revealed Epigene4c Field for Canceriza4on 20" FLNc FLNc P < 0.05 Methyla4on level (%) 16" 12" 8" 4" P < 10-14 Methyla4on level (%) 7 4 2 P < 0.05 0" H. pylori (- ) (n = 56) H.pylori (+) (n = 98) 0 Healthy Single gastric Mul4ple volunteer (n = 6) cancer (n = 11) gastric cancer (n = 13) [Maekita, Clin Cancer Res, 2006] [Nakajima, CEBP, 2006] ü These findings showed the presence of an epigene4c field for canceriza4on in gastric mucosa.
Bisulfite Sequencing Bisulfite treatment PCR using universal primer A Cloning (10~ clones) Sequencing B Gold standard Methylation pattern
Me Cont. Un Cont. PCR Bias Bisulfite Sequencing
Number of Analyzed Clones in Bisulfite Sequencing 1 - p p Binomial distribution (p = 0.5, n = 10) 0.1% 1.0% 4.4% 11.7% 20.5% 24.6% 20.5% 11.7% 4.4% 1.0% 0.1% (From Sasaki et. al)
Number of DNA Molecules in Bisulfite Sequencing 100 copies for methyla4on payern / 1 reac4on = 2,000 copies (1000 cells) before bisulfite treatment = 6 ng (human, mouse) / 1 reac4on Nonspecific products of PCR Duplica4on of clones 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10
Acknowledgment National Cancer Center Research Institute" " Toshikazu Ushijima" Tohru Niwa" Hideyuki Takeshima" Yasuyuki Shigematsu" Ken Gyobu" Yasunori Matsuda" Kosuke Hosoya" Monali Lahoti" Satomi Takahashi"