Integrating genetic and epigenetic variation in schizophrenia Jonathan Mill www.epigenomicslab.com 1
Integrated –omics approach: heterogeneous etiology, convergent molecular pathology Functional annotation of human brain genome Voineagu et al, Nature, 2011 Pidsley et al, Genome Biology, 2014 Human brain tissue Biomarkers Clinical cohorts Lunnon et al, Nature Neuroscience, 2014 Animal models Cell models (iPSC) Genome editing Single-cell profiling Genomics Epigenomics (5mC, 5hmC, histone modifications) Transcriptomics (GWAS) (RNA-seq) Function 2 Systems-level analyses
A suite of epigenetic modifications act to fine-tune genomic function Zhou et al (2011) Addi6onal DNA modifica6ons: 5hmC, 5fC, 5cC 3
Functional annotation of regulatory variation in the genome Roadmap Epigenomics Consortium, Nature, 2015 PsychENCODE Consortium, Nature Neuroscience, 2016 Regulatory elements Enhancers Generate hypotheses about function 4
Integrated genetic-epigenetic analysis of schizophrenia Multi-centre clinical cohorts SZ- Adult post- UCL (n = 675) Human fetal discordant mortem brain ABER (n = 971) brain (23 to 184 monozygotic (matched PFC, KCL (n = 801) days post- twin pairs striatum, CARDIFF (n = 950) conception) (n = 99 hippocampus, DUBLIN (n = 713) (n = 179) pairs) cerebellum) EDIN (n = 540) • Illumina 450K/EPIC DNA methylation data • Genotype data (imputed to 1000G phase 3) • Differences in DNA methylation associated with schizophrenia • Differences in DNA methylation associated with high polygenic burden for schizophrenia • Epigenetic consequences of genetic variants associated with schizophrenia 5
90 p=1.16e − 07 p=1.25e − 07 75 PFC 85 70 DNA methylation (%) DNA methylation (%) 65 80 60 75 55 50 70 Dopamine D2S 45 Receptor-mediated Control Schizophrenia Control Schizophrenia (n=23) (n=20) (n=23) (n=20) MAPK signaling cg00903099 (HTR5A) cg08171022 (PPFIA1) 65 20 p=2.4e − 07 p=2.85e − 07 60 Serotonin receptor DNA methylation (%) DNA methylation (%) 15 55 50 10 45 5 40 Axon guidance Control Schizophrenia Control Schizophrenia (n=23) (n=20) (n=23) (n=20) Pidsley et al, Genome Biology, 2014 6 Viana et al, in review
Schizophrenia-associated differentially methylated regions – consistent signals across different brain regions RPH3AL : calcium- dependent exocytosis 7 Viana et al, in review
PFC co-methylated modules associated with schizophrenia DPPB DNMTDA CAMKXD MICALI KIAAXDXO LOXLX HLA-L SLCXEAE KALRN PACSINX EPHBB CALMD GRMI SRRT UNCU6A PRKDC TMEMUBB;SAPSX CRHRX VPS3D PPPIRIC ABLIMX SHANKI RICHI MUCB PVRLX RGSXI LHPP VOPPX IGHMBPI CACNAXC INTSX TSNAREX KIRRELD GLTSCRX CPLXD;LMANXL BRD6 CPSFX Key for node colour TMEMXDID MEGFXX MMPXE RIMBPI XO -E XO -X6 XO -IO CTBPX X LRFNI Rbest p-value within 3kb of gene coding region in largest CUXI GWAS to date CSchizophrenia Working Group of the Psychiatric Genomics Consortium Nature IOX6) PITPNMI Pidsley et al, Genome Biology, 2014 ITPKA 8 GNAOX Viana et al, in review
Neur Neurodevelopmental origins of mental il odevelopmental origins of mental illness lness Schizophrenia Normal Brain Development Autism Mutations Polygenic variation Environmental insults Stochastic factors 9
Fetal brain tissue from 191 elective abortions DNA methyla6on DNA hydroxymethyla6on RNA-seq (Nick Bray) Gene6c varia6on Single-cell transcriptomics 179 human fetal brain samples profiled for DNA methylation 100 male, 79 females 23 to 184 days post-conception DPC calculated by Carnegie staging and fetal foot length 10
>28,000 DMPs (annotated to >5,000 genes) significantly associated with fetal brain development Hypermethylation Hypomethylation 11 http:epigenetics.iop.kcl.ac.uk/fetalbrain Spiers et al (2015)
Clustered regions of developmentally-coordinated DNA methylation in the human fetal brain. Coordinates (Genome build 37) 2660000 2670000 2680000 2690000 2700000 ● P<1E − 07 20 ● ● P<1E − 06 ● ● ● ● ● ● ● ● P<1E − 05 ● ● ● ● P>1E − 05 15 − log10 (p value) ● ● ● ● ● ● ● ● 10 ● ● ● ● ● ● ● 5 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0 ● ● ● ● ● ● ● ● ● 0.1 ● ● 0.0 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● Regression coefficient ● ● ● ● ● ● ● ● ● − 0.1 ● ● ● ● ● ● ● ● ● ● ● ● ● ● − 0.3 ● ● ● ● ● ● ● ● ● ● ● ● − 0.5 ● 5’ TTYH3 3’ 12
regions… The distribution and direction of fetal brain dDMPs is not equal across genomic Percentage significant probes (%) Downstream region CGI 0 20 40 60 80 100 Distal promoter CGI Distal promoter shore hypermethylated dDMPs CG-rich domains: enriched for 6.74E-53) hypomethylated dDMPs (P = Overall enrichment for Intergenic CGI Proximal promoter CGI rela6ve enrichment = 0.26, p <3.34E-92 Downstream shore Intergenic shore Proximal promoter shore Gene body CGI Intergenic non − CGI Gene body shore Gene body non − CGI Downstream region non − CGI rela6ve enrichment = 0.34, p <3.34E-92 Distal promoter non − CGI Intergenic shelf Gene body shelf Proximal promoter shelf rela6ve enrichment = 0.30, p < 3.34E-92 Distal promoter shelf Hypomethylated Hypermethylated Proximal promoter non − CGI Downstream shelf 13
Correlation with gene expression data from Brain Cloud resource (http://braincloud.jhmi.edu) BTG3 (cg14394939) FAM49A (7311) (cg06829760) 2.0 ● ● ● Expression (log2(sample/ref)) Expression (log2(sample/ref)) p = 8.68E − 05 ● ● ● ● ● ● 0.5 ● ● ● ● 1.5 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0.0 ● ● ● ● ● ● ● 1.0 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● − 0.5 ~30% of top- 0.5 ● ● ● ● ● ranked dDMPs ● ● p = 2.49E − 06 0.0 ● ● ● ● ● annotated to a − 0.50 − 0.46 − 0.42 − 0.38 − 0.50 − 0.46 − 0.42 − 0.38 gene that is Age Age dynamically- expressed KLHL35 (cg02313829) WIPF1 (cg18185980) during brain ● ● ● ● ● Expression (log2(sample/ref)) Expression (log2(sample/ref)) development p = 2.08E − 05 − 1.5 ● ● ● 0.5 ● ● ● ● ● ● ● ● ● ● ● ● − 2.0 ● ● ● ● ● ● ● ● ● ● ● ● ● 0.0 ● ● ● ● ● ● ● ● ● ● − 2.5 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● − 0.5 ● ● ● ● − 3.0 ● ● ● ● ● − 1.0 p = 3.75E − 03 ● ● ● ● ● − 0.50 − 0.46 − 0.42 − 0.38 − 0.50 − 0.46 − 0.42 − 0.38 14 Age Age
There are autosomal sex differences … cg03691818 (KRT77) cg12691488 ● ● ● ● P=7.59E − 69 ● ● P=1.66E − 80 ● ● ● ● ● ● ● DNA methylation (%) ● ● ● ● ● ● DNA methylation (%) ● ● ● ● 18 ● 40 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 30 ● ● ● ● ● ● ● ● ● ● ● 14 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 20 ● ● ● ● ● ● ● ● 10 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 8 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 10 ● ● ● ● ● 50 100 150 50 100 150 Days post − conception Days post − conception 15
There are autosomal sex differences and distinct sex-specific developmental trajectories in the human fetal brain methylome 16
There are distinct modules of co-methylated loci in the developing human brain… 17
Recommend
More recommend