당뇨병에서의유전자연구 SNP 연구입문 GENE ( 유전자 ) 생명체에서유전의분자단위로, 생명체안에서기능을갖는 polypeptide 나 RNA chain 을 Coding 하는 DNA 나 RNA 2012 년 7 월 14 일연세대학교의과대학내과학교실강은석 DNA vs. RNA GENE ( 유전자 ) Five Carbon sugar; Ribose 1
Nucleotide DNA Double Strands Base Phosphate Five Carbon Sugar; Ribose 1953년4월제임스왓슨과프랜시스크릭 : DNA의구조가이중나선형이라는내용발표 9년뒤노벨상수상 Nucleotides & Base 1920-1958 1958 2
Chromosome ( 염색체 )and DNA Chromosome ( 염색체 ) Unsorted Human Chromosomes Male Karyotype Female Karyotype 44+XY OR 44+XX GENOME ( 유전체 ) = GENe + chromosome Chromosome( 염색체 ) 부위명명법 p 예 ) 12p23.1 12 = chromosome 12 p = short arm 유전체 = 유전자 + 염색체 q 2 = region 3 = band.1 = sub-band 3
Allele ( 대립유전자 ) Central Dogma ( 센트럴도그마 ) 같은염색체내의동일유전자의다른부위 Each locus on a chromosome has alternative versions of a gene called alleles. You inherit one allele from each parent. Human Genome ( 인간유전체 ) CODON ( 코돈 ) 염색체수 22x2+2=46 개 유전자수 : 32,000 개 염기서열수 : 30 억 Nucleotides 유전자의 <5% 만이 Active CODON ( 코돈 ): 유전정보의최소단위 1 Codon = 3 Nucleotide 1 Codon = 1 amino acid Number of Possible Codons: 4 (A,T,C,G) x 4 (A,T,C,G) x 4 (A,T,C,G) = 64 개 4
Genetic Code ( 유전암호 ) 아미노산의수 = 20 코돈의수 = 64 3 개 Codon (TAA, TGA, TAG): Stop Codon 61 개 Codon: 아미노산을 Coding ATG Codon: Methionine 을 Coding 하는 Start Codon Genetic Variations ( 유전자변이 ) Germline mutation, not somatic mutation Genetic Code 1 2 3 DNA Genetic Variations Copy Number Variation 1) STR (Short Tandem Repeat) = microsatellite 반복단위 2-7 base pair 2) VNTR (Variable Number of Tandem Repeat) = 반복단위 14-70 base pair Insertion/Deletion, Translocation 예 ) Frameshift SNP (Point mutation) Silent (Synonymous) mutation: No AA Change Missense (Non-Synonymous) mutation: AA Change Nonsense mutation = Stop Codon 5
SNP (Single Nucleotide Polymorphism) SNP ID Submitted SNP (SS) & Reference SNP (RS) SNP: DNA 염기서열에서한개의염기서열의차이를보이는유전적변화 Polymorphism: 1% 이상의빈도로존재하는 2 개의대립염기서열 (Bi-allelic) 변이 Common polymorphism (MAF > 5%) Rare polymorphism p (MAF = 1-5%) Mutation (MAF < 1%) Useful genetic marker No. of SNP: 2,365 만개 dbsnp build 131 (2011.5) PURINE PYRIMIDINE Classification of SNP by Location A 10% 10% 30% 30% 10% G 10% C T 6
MAF (Minor Allele Frequency) 대립유전자형의빈도가낮은것의비율 SNP 마커의유전적다양성을표현하는표준지표 MAF > 5% : Common polymorphism SNP 명명법 Nucleotide Numbering ATG Translation Start Site 기준 : ATG 의 A=+1 ATG codon 5` 앞염기서열 -1 No base 0 MAF 1-5% : Rare polymorphism MAF < 1% : Mutation SNP 명명법 g for genomic sequence; g85t>g c for cdna sequence; c85t>g m for mitochondrial sequence; m55t>g r for RNA sequence; r63u>a p for protein sequence; pr325w SNP 명명법 Intron; IVS (InterVening Sequence ) 앞쪽인트론 : 앞의 exon 끝번호 + exon 끝부터의염기수 77+1G (exon 번호알때 IVS1+1G) 뒤쪽인트론 : 뒤의 exon 시작번호 + exon 앞부터염기수 78-2A (exon 번호알때 IVS1-2A) 7
SNP 명명법 치환 Substitution 86A>G; 86번째염기가 A 에서 G로 IVS2-2A>C; intron 2 번의 -2 위치에서 A 가 C로 Haplotype 결실 Deletion 76_78delACT; 76번과 78 번사이에서 ACT결손 삽입 Insertion 76_77insT; 77i 76 번과 77 번사이에 T 삽입중복 Duplication 76_77dupCT; 76번과 77 번사이에 CT중복 같은 chromosome 내에서함께유전되는경향이있는인접한 SNP 들의집합 A or T G or C Haplotype A or T Locus #1 G or C Locus #2 Phase ambiguity Punnet square 8
Haplotyping Haplotyping 방법 Haplotype-Tagging SNP (htsnp htsnp) And HapMap (Haplotype Map) Molecular Haplotyping ( 분자생물학적방법 ) 가계도분석, 정확, 비용, 많은시간소요 In Silico Haplotyping ( 통계적방법 ) Haplotype Program Haploview (Broad Institute; Harvard and MIT) http://www.broadinstitute.org/scientific- community/science/programs/medical-and- population-genetics/haploview/haploview Haplotyper (Harvard University) http://www.people.fas.harvard.edu/~junliu/haplo/click.html PHASE2 (Univ. of Chicago, Metthew Stephens) http://www.stat.washington.edu/stephens/phasefaq.ht ml Recombination & Linkage SAS Genetics http://sas.com 9
Recombination ( 재조합 ) 감수분열과정에서염색체의일부가서로교환하는교차 (crossover) 가일어나많은대립유전자들이새로운조합을가지게되는것 Linkage LD 란두 loci 에서일어나는대립유전자들쌍의이론적인예측치와실제측정간의차이 (deviation) 를나타냄. Linkage ( 연관 ) 두개의 SNP 간의거리가매우가까우면 2개의 SNP는서로연관되어있어서다음세대에같이전달 Linkage Equilibrium ( 연쇄평형 ): 유전자들이독립적으로배합되어있는상태, 서로다른 locus에있는유전자의대립형질은서로독립적으로나타남. 따라서, haplotype 빈도는각대립형질빈도의곱. Linkage Disequilibrium ( 연쇄비평형 ): 유전자들이의존적으로배합되어있는상태 LD 를측정하는방법 D : Linkage disequilbrium coefficient D = P(AB) P(A)P(B) D=0 if LE. -0.25<D<+0.25 D : D/Dmax D =0 if LE D =1 1 if complete LD (no recombination) 0<D D <1 if variable LD with recombination r 2 : LD correlation coefficient r 2 =D 2 /p1p2q1q2 0<r 2 <1 r 2 =1 if perfect LD if r 2 >0.33, strong LD 10
Genotype / Allele Frequency 기초통계유전학 Genotype N AA 300 AB 500 BB 200 Total 1,000 Allele N A 300*2 + 500 1,100 B 200*2 + 500 900 Total 2,000 Allele, Genotype and Phenotype Phenotype Genotype Gene/Allele A형 B형 AB형 O형 M형 N형 MN형 AA, AO, OA BB, BO, OB AB OO MM NN MN A B O M N Hardy-Weinburg Equilbrium (Genotype 검증 ) 정의무작위교배를하는큰집단에서유전자와유전자형의빈도는세대를거듭하여도변하지않고평형을이루게된다. Godfrey Hardy 조건 1877~1947 1 교배는무작위적으로이루어져야한다 (Random mating). 2 돌연변이는생기지않는다 (no mutation). 3 이입과이출이없다 (no migration). 4 개체군에는선택이작용하지않는다 (no natural selection). 5 표본집단의크기가크다 (population size is infinite). Wilhelm Weinberg 1862-19371937 11
Hardy-Weinburg Equilibrium 검증방법 기대값과실제로관측한값의차이를 X2 (chi-square test 로검증 ) P-value <0.05 이면 deviated from HWE 질병유전자발굴방법 Hardy-Weinburg Equilbrium 에서 deviation 된경우는? Heterozygote Excess 다른생존율의차이에서기인 (differential survival) Genotyping error (nonspecific assay) Association vs. Linkage Study Homozygote Excess Population 문제일가능성 ( 이질적인집단의시료 ; population stratification) 치사유전자 (Null Allele) Genotyping error (nonspecific assay) 12
bject Number rs Su Population (Association Study) Polygenes Oligogenes Effect Size Family (Linkage Study) Major Genes 1. 실험설계 Association Study 단계 연구대상질환선정연구대상시료수결정정상과질병구분기준결정 2. 시료및데이터수집 유전체시료수집임상정보수집 3. 유전자형조사 (Genotyping) 후보유전자및후보 SNP 선정및유전형조사 4. 통계분석 HWE 검증, 단일 SNP 연관성분석, LD 구조분석, Haplotype 연관성분석 Multiple Comparison 보정 Association Study 의의미의미 1. 질병의직접적원인이되는경우 2. 질병원인유전자와 Linkage 된경우 Marker 3. 특정유전자형과질병이자연적으로선택된경우 4. 집단간의유전적조성차이에의한연관성 (population stratification) tifi ti 5. False Positive, Type 1 Error Association Study 방법 통계의유의성 (POWER) = 1- beta (type 2 error) Type 2 Error = False negative 통계적유의성 (POWER) 에영향을주는요소 1. 시료의수 절대적임. 2. 유전모델 dominant or recessive 3. Allele frequency 4. Relative Risk 임의로조절가능한것은시료의수 (Sample Size) 뿐임 13
Association Study 방법 Sample Size Association Study 재현성 (Replication) 통계적반복실험 (Statistical Replication) 사용된시료를나누어재분석다른시료에서동일한분석 MAF=5% 기능적반복실험 (Functional Replication) 기존연구결과가재현되지않을때의원인기존연구가 False Positive 재현연구의 False Negative 두연구에사용된시료에대한집단의차이 Allele Frequency Association Study Phenotype 정상과질병구분기준결정대부분의연관성연구에서는특정기준을중심으로질환자 (Case) 와정상인 (Control) 을구분함. 정상인중에는아직질병이발병하지않았지만이후에질병이발생할수있음. 예 ) 당뇨병의진단기준을공복혈당 126 mg/dl 로했을때, 혈당 125 와 127 의차이는? Association Study 개체선별법 집단을기초로한전향적조사, Cohort 연구 상대적위험도 Relative Risk, RR 병원을기초로한후향적조사 오즈비 Odds Ratio, OR 예 ) 정상인으로분류된 40 세남자, 10 년뒤에도정상일까? 14
Association Study 데이터의종류그룹의특성통계분석방법 범주형 Categorical 2 Groups 2 Groups (small group) 2 Groups (adjustment) X 2 test Fisher s exact test Logistic regression 연속형 2 Groups t-test Continuous 3 or more Groups 3 or more Groups (adjusment) ANOVA Multiple regression Genetic Background Familial Aggregation of Diabetes (both parents-offspring 40%) Twin Studies (70-90%) Genetic Syndromes Associated with Diabetes It is clear that T2D has a strong genetic component. 당뇨병에서의유전자연구 Story of Pima Indians ( ) prevalence of diabetes (54%) Arizona (6.3%) Mexico Arizona Pima Mexican Pima 15
Are genes responsible for Type 2 Diabetes and Obesity? Genetics of T2DM Geneticist s Nightmare Neel J. Diabetes mellitus: a geneticist's nightmare. In: Creutzfeldt W, Kobberling J, Neel JV, eds. The genetics of diabetes mellitus. Springer-Verlag Verlag, 1976:1-11. 11. Normal Age 20 Genetics of T2DM Genes Diabetogenes -Primary -Secondary Diabetes-related genes Insulin Resistance Decreased Insulin Secretion Environment Toxins Diet Activity 30 40 50 Type 2 DM 60 Natural History of Pre-type 2 Diabetes and the Impact of genes and the Environment in this Process. Is genetically heterogeneous Is almost certainly polygenic Strong gene/gene and gene/environmental interactions play important roles in development of T2D Common Gene, Common Disease and Complex Disease Phenotyping 당뇨병발병에관여하는유전적인원인은질병발생위험도가그리크지않으면서비교적흔한유전자들의변이일가능성이크다. 16
MODY (Maturity Onset Diabetes of the Young) Monogenic Causes Uncommon form of T2D (<5% of all T2D) Autosomal dominant inheritance Early onset of hyperglycemia; < 25 years Impairment in insulin secretion caused by a mutation in a different gene that is directly involved with beta cell function Slow onset of symptoms, Absence of obesity, No ketosis, No beta cell autoimmunity. Family-Based Linkage Analysis 장점 : Linkage analysis exhibits its maximal power in identifying loci implicated in rare (MODY, Neonatal Diabetes, ) Polygenic Causes 단점 : Loses precision for common forms of the diseases (T2DM,.) 17
연구방법 1. Candidate Gene Association Study 1) Functional lc Candidate Gene eg. PPARG, KCNJ1 2) Positional Candidate Gene eg. Calpain 10, TCF7L2 2. Genome Wide Association Study Candidate-Gene Association Study 장점 : Known Target Gene (may play a role in T2DM Pathogenesis) 단점 : Lack of Consistency replication issue Lack of Power Acceptance of low p value threshold (0.05) False Positive Contributed to Pathogenesis of Diabetes Successful target for anti diabetes medication PPARG P12A, KCNJ11 E23K Candidate Gene Association Study Positional Cloning Calpain 10 Gene Reported in a Mexican American population Not robustly replicated in other ethnic groups 18
Fine Mapping Linkage Analysis TCF7L2 Chromosome 10 region showed linkage to T2D Fine mapping TCF7L2 5 SNPS and 1 tetranucleotide repeat polymorphism (DG10S478) Replicated many ethnic groups ethnic groups GWAS is facilitated by Completion of Human Genome Project Completion of International HapMap Project Advance in Genotyping Technology Advance in Computer Technology Genome Wide Association Studies 3.9 million SNPs in 270 DNA samples from 4 different ethnic groups 90 Yoruba individuals (30 parent parent offspring trios) from Ibadan, Nigeria (YRI) 90 individuals (30 trios) of European descent from Utah (CEU) 45 Han Chinese individuals from Beijing (CHB) 45 Japanese individuals from Tokyo (JPT) Quantified by LD Tagging SNP 19
The First GWAS for T2D in 2007 Illumina HumanOmni5-Quad BeadChip Kit 4.3 million SNP loci Affymetrix Genome-Wide Human SNP Array 6.0 More than 906,600 SNPs French population p 661 T2DM and 614 Control 392,935 SNPs 4 new loci and confirm TCF7L2 Sladek R, et al. Nature 445: 881-885, 885, 2007 Successful GWAS Conditions The First GWAS for T2D 1. Sufficient sample size (at least 1,000 each of cases and controls) 2. P-value < 5 x 10-8 (Genome wide significance) 3. Confirmation of association by independent replication studies McCarthy MI, et al. Nature Reviews Genetics 9:356-369, 369, 2008 SLC30A8, HHEX, LOC387761, EXT2, TCF7L2 Sladek R, et al. Nature 445: 881-885, 885, 2007 20
The Second GWAS for T2D decode genetics, Iceland FUSION Finland United States Investigation of NIDDM SLC30A8, HHEX, CDKAL1 SLC30A8, HHEX, CDKAL1, IGFBP2, CDKN2A/B, PPARG P12A, KCNJ11 E23K Steinthorsdottir V, et al. Nat Genet 39: 770-775, 775, 2007 Science 316: 1341-1345. 1345. 2007 WTTCCC/UKT2D Wellcome Trust Case Control Consortium/ United Kingdom Type 2 Diabetes Genetics consortium DGI Diabetes Genetics Initiative of Broad Institute of Harvard and MIT, Lund University, and Novartis Institutes of BioMedical Research SLC30A8, HHEX, CDKAL1, IGFBP2, CDKN2A/B, PPARG P12A, KCNJ11 E23K Science 316: 1336-1341. 1341. 2007 SLC30A8, HHEX, CDKAL1, IGFBP2, CDKN2A/B, PPARG P12A, KCNJ11 E23K, FTO Science 316: 1331-1336. 1336. 2007 21
SLC30A8 Meta-analysis analysis DIAGRAM Plus DIAGRAM + (combined with diagram cohort) 2,426,886 SNPs 2,000 subjects of European origin 34,412 cases and 59,925 controls 12 new T2D association signals Association statistics from one of the 8 T2DM GWAS Nat Genet 42(7):579-589, 589, 2010 Meta-analysis analysis of Initial GWASs the Diabetes Genetics Replication And Meta-analysis analysis (DIAGRAM) Consortium BCL11A, ZBED3, KLF14, TP53INP1, CHCHD9, KCNQ1, CENTD2, HMGA2, HNF1A, ZFAND6, PRC1, DUSP9 (X-chromosomal association) 5 additional loci: JAZF1, CDC123-CAMK1D, CAMK1D, TSPAN8-LGR5, THADA, ADAMTS9, NOTCH2 4,549 cases and 5,579 controls 2.2 million SNPs Nat Genet 40: 638-645, 645, 2008 22
Identified Genome-Wide Associations GWAS for Glycemic Traits A polymorphism within the G6PC2 gene is associated with fasting plasma glucose levels. Science 320: 1085-1088, 1088, 2008 Variants in MTNR1B influence fasting glucose levels. Nat Genet 41: 77-81. 2009 A variant near MTNR1B is associated with increased fasting plasma glucose levels and type 2 diabetes risk. Nat Genet 41: 89-94. 94. 2009 Variations in the G6PC2/ABCB11 genomic region are associated with fasting glucose levels. J Clin Invest 118: 2620-2628. 2628. 2008 A common haplotype of the glucokinase gene alters fasting glucose and birth weight: association in six studies and population-genetics analyses. Am J Hum Genet 79: 991-1001. 1001. 2006 New genetic loci implicated in fasting glucose homeostasis and their impact on type 2 diabetes risk. Nat Genet 42: 105-116. 116. 2010 Genetic variation in GIPR influences the glucose and insulin responses to an oral glucose challenge. Nat Genet 42: 142-148. 148. 2010 Meta-Analyses of Glucose and Insulin- related traits Consortium (MAGIC) Examined 21 GWAS to identify loci associating with fasting glucose, fasting insulin, HOMA-β, and HOMA-IR 76,558 individuals from 34 additional cohorts, 9 new loci in or near ADCY5, MADD, CRY2, ADRA2A, FADS1, PROX1, SLC2A2, GLIS3, and C2CD4B were found to be associated with fasting gg glucose. Nat Genet 42: 105-116. 116. 2010 23
GWAS for Glycemic Traits GWAS for T2D in Japanese #2 Fasting glucose G6PC2, MTNR1B, GCK, ADCY5, MADD, CRY2, ADRA2A, FADS1, PROX1, SLC2A2, GLIS3, and C2CD4B Fasting insulin and HOMA-IR. GCKR, IGF1 2-hour postprandial glucose GIPR, ADCY5, GCKR, VPS13C, TCF7L2 Common GWAS genes of T2D and Glycemic traits MTNR1B, GCK, ADCY5, PROX1, DGKB-TMEM195, GCKR 187 T2D and 752 controls Nat Genet 42: 105-116. 116. 2010 Nat Genet 40: 1092-1097, 1097, 2008 GWAS for T2D in Japanese #1 GWAS for T2D in Japanese #3 KCNQ1 (potassium voltage-gated gated channel, KQT- like subfamily, member 1) to be a strong candidate for conferring susceptibility to type 2 diabetes 194 T2D and 1,558 controls 4,470 T2D and 3,071controls 459,359 SNPs stage 1, 4,470 cases and 3,071 controls stage 2, 2,886 cases and 3,087 controls stage 3, 3,622 cases and 2,356 controls UBE2E2 is not associated with T2D in Europeans Nat Genet 40: 1098-1102, 1102, 2008 Nat Genet 40: 864-869, 869, 2010 24
GWAS for T2D in KOREA? Current Limitations for GWAS 1. Considerable number of uncaptured SNPs No. of SNP: 2,365 만개 SNP Chip 4 백만 2. GWAS p-value may produce type 2 errors (false negative results) Meta-analysis analysis of 8 T2D GWAS (6,952 T2D, 11,865 controls) with a stage 2 in silico replication analysis (5,843 T2D, 4,574 controls) and a stage 3 de novo replication analysis (12,284 T2D, 13,172 controls). 3. Low frequency (MAF<1%) risk variants with large effects could be missed 8 new T2D loci reaching GW significance GLIS3, PEPD, FITM2-R3HDML- HNF4A, KCNK16, MAEA, GCC1-PAX4, PSMD6 and ZFAND3. Nat Genet 40: 864-869, 869, 2011 What have GWAS brought about so far? 1. Identified T2D loci are associated more frequently with beta cell function rather than insulin resistance ( Only GCKR, PPARG, FTO, KLF14 associated with HOMA-IR ) 2. Missing Heritability GWAS explain only 10%(-20%) of the known heritability in twin study. 3. Translation of T2D genetics into clinical Practice 3-1. Disease Prediction i and Prevention 3-2. identifying novel therapeutic targets Endo J 58: 723 739, 739, 2011 25