Exome sequencing supports a de novo mutational paradigm for schizophrenia

Transcription

1 Supplementary Information Exome sequencing supports a de novo mutational paradigm for schizophrenia Bin Xu 1,2, J. Louw Roos 3, Phillip Dexheimer 4, Braden Boone 4, Brooks Plummer 4, Shawn Levy 4, Joseph A. Gogos 2,5,*, Maria Karayiorgou 1,* 1 Department of Psychiatry, Columbia University, New York, NY Department of Physiology & Cellular Biophysics, Columbia University, New York, NY Weskoppies Hospital & Department of Psychiatry, University of Pretoria, Pretoria, RSA HudsonAlpha Institute for Biotechnology, Huntsville, AL Department of Neuroscience, Columbia University, New York, NY *Correspondence should be addressed to Maria Karayiorgou (mk2758@columbia.edu) or Joseph A. Gogos (jag90@columbia.edu)

2 Supplementary Notes Sample characteristics. The case sample consisted of 33 males (62%) and 20 females (38%) and both their unaffected biological parents, all of Afrikaner origin. In this sample, 90% of the probands met strict diagnostic DSM-IV criteria for schizophrenia (n = 47), 5% for schizoaffective disorder, depressed type (n = 3), and 5% for schizoaffective disorder, bipolar type (n = 3). The control sample consisted of 14 males (64%) and 8 females (36%) and both their biological parents. All controls were also of Afrikaner origin (all 4 grandparents were Afrikaans-speaking), adults (past the age at risk for any major psychiatric disease) and were screened via a systematic questionnaire to exclude history of psychiatric illness in themselves and close first- or second-degree relatives. Phenotypic variables in the sporadic cases cohort. The entire case sample includes 1 patient with mental retardation (who did not have a de novo mutation), 10 patients with a history of learning disabilities (4 of whom had de novo mutations) and 9 patients with a history of developmental delays (5 of whom had de novo mutations). History of learning disabilities was recorded as positive if there had been a diagnosis of a learning disability, or clear history of being a slow learner, requiring remediation at school, or placement in a special class. History of developmental delay was recorded as positive when there was clear history of maturation lag or milestone delays prior to age 6 (i.e. delayed crawling, walking, speaking, etc.). There were no significant differences in the distribution of the sexes among the

3 cases carrying de novo mutations and non-carriers [19 males (70%) and 8 females (30%) in the subgroup carrying de novo mutations; 14 males (54%) and 12 females (46%) in the non-carrier subgroup, Fisher s exact test, P = 0.26]. Similarly, age at disease onset defined as the age at which full DSM-IV criteria for SCZ or SCZAFF disorder were met was similar among cases carrying de novo mutations (23 yrs) versus non-carriers (21.5 yrs). Moreover, comparison of available clinical data from the families carrying de novo events and the rest of the sporadic cohort cases did not reveal any exceptional features suggestive of a more severe clinical presentation or of elevated rates of developmental delays, learning disabilities or presence of mental retardation (Supplementary Table 2). In particular, in comparing variables that could reflect severity of the disease (i.e. number of hospitalizations and total time spent in illness) we found no differences (average number of hospitalizations at the time of recruitment in the study was 2.5 in the subgroup carrying de novo mutations versus 2.7 in the non-carrier subgroup; similarly, total time spent in illness (in months) prior to the study entry diagnostic interview was 107 versus 102, respectively (Supplementary Table 2). We also asked whether the number of de novo events per case may act additively or synergistically to determine disease severity or atypical developmental course but we found no differences between single and multiple de novo mutation carriers. In general, the significance of two or more validated events in a single proband in terms of disease risk, severity, or both, remains to be determined using considerably larger patient samples.

4 Clinical information for the patient carrying the de novo p.pro429arg mutation in DGCR2. At the time of recruitment in the study, patient was a 35-year-old unmarried male staying with his parents. He has had numerous admissions to the psychiatric hospital because he did not comply with his medication plan and kept on relapsing into psychosis. Early childhood was insignificant with birth and developmental milestones reported as normal. There was no history of learning difficulties or early strange behavior and he did socialize easily. It was after the age of 10 that socialization became poor and after age 15 his academic performance started to deteriorate. He struggled through the military and a number of jobs over the next 10 years of his life, and at age 26 he developed grandiose and persecutory delusions meeting full diagnostic criteria for schizophrenia. He was also objectively hallucinating (talking to himself). There have been several admissions to the psychiatric hospital since disease onset at 26. He still has delusions and no insight into his illness. Since onset, he has not been working and his functioning has steadily declined. Literature-based evidence for other conserved/damaging variants. The WDR11(also known as BRWD2) gene encodes for a protein that contains multiple WD-domain and interacts with EMX1, a homeodomain transcription factor involved in brain development 1. Interestingly, mutations in the WDR11 gene have been found in subjects with Kallmann syndrome, a rare condition often associated with psychiatric pathology 1. The PLCL2 gene is thought to modulate GABAergic transmission, as well as behavioral performance in transgenic mice 2,3. TRAK1 encodes for kinesin adaptor that appears to regulate mitochondrial transport along axons.

5 Interestingly, TRAK1 also interacts with GABA type A (GABA(A)) receptors and regulates their levels. Mice carrying a homozygous mutation in the gene (hyrt) show hypertonicity, which is likely caused by deficits in GABA-mediated motor neuron inhibition 4,5. The KLF12 gene that encodes for a transcription factor has been tentatively implicated in the pathogenesis of bipolar disorder via a convergent functional genomics approach 6. The RBB1CC1 (FIP200) gene encodes for a protein that interacts with signaling pathways to coordinately regulate cell growth, cell proliferation, apoptosis, and especially autophagy, as well as cell migration. It has also been identified as a tumor suppressor, which enhances retinoblastoma 1 gene expression in cancer cells 7,8. The LAMA2 gene encodes the alpha 2 chain, which constitutes one of the subunits of laminin 2 and 4. Homozygous mutations in LAMA2 lead predominantly but with variable penetrance to classical congenital muscular dystrophy characterized by clinical central nervous system involvement including white matter abnormalities, cognitive impairment, seizures and neuronal migration defects 9. De novo mutations in a related gene, LAMC3, were recently described in a autism spectrum disorders cohort 10.

6 Supplementary Figure 1: Flow diagram of sequence analysis pipeline. The flow diagram describes the procedure for sequence data cleanup, variant genotype calling and inheritance pattern determination. First, raw sequence reads were mapped to the human genome (build hg19) using BWA. Second, duplicates were removed, followed by data cleanup using GATK. Third, variant genotype calls were made by GATK Unified Genotyper and in/del genotype calls were made by Dindel. Fourth, low QC variants and variants overlapping with dbsnp132 were removed using a custom script. Fifth, the inheritance pattern of each variant was determined by comparing offspring and parental genotypes. Sixth, candidate variants were revalidated using the mpileup module of the SAMtools. Lastly, de novo variants were experimentally validated by Sanger sequencing.

7 Supplementary Figure 2: Identification and confirmation of a de novo SNV in a subject with SCZ. A. Identification of a de novo SNV in a subject with SCZ. IGV browser view of a de novo SNV in the DGCR2 gene. Top panel shows chromosome 22 with the mutation location indicated by a red box. Middle panels depict representative individual reads as well as the relative coverage per base pair for proband, mother and father. Colored coverage columns indicate the location of the SNV. The relative proportion of colors represents the ratio of variant reads observed. Bases that deviate from the reference sequence are highlighted in blue. Lower panel shows reference sequence and translated aminoacids. B. Sanger sequencing traces. Partial electropherograms surrounding the predicted mutation site validating the de novo occurrence for the candidate SNV depicted in (A).

8 Red arrow indicates that the proband, but not his parents, is a heterozygous SNV carrier at the predicted site. C. DGCR2 multi-species sequence alignment. The protein sequences of human, chimpanzee, mouse and zebrafish orthologues were compared. Black blocks indicate completely conserved residues; grey blocks indicate similar residues (defined by Boxshade default similarities); white blocks indicate variable residues. The position of the de novo SNV depicted in (A) is labeled with a red asterisk.

9 Supplementary Figure 3: Identification and confirmation of a de novo in/del in a subject with SCZ. A. Identification of a de novo in/del in a subject with SCZ. IGV browser view of a de novo in/del in the RB1CC1 gene. Top panel shows chromosome 8 with the mutation location indicated by a red box. Middle panels depict representative individual reads as well as the relative coverage per base pair for proband, mother and father. Lower coverage columns and dashes in the variant reads indicate the location of the in/del. Lower panel shows reference sequence and translated aminoacids. B. Sanger sequencing traces. Partial electropherograms surrounding the predicted mutation site validating the de novo occurrence for the candidate in/del depicted in (A). Red arrow indicates that the proband, but not his parents, has a deletion at the predicted mutation site.

10 Supplementary Figure 4: Sequencing coverage comparison. A. Average base pair reads at 1X, 4X, 8X, 20X and 30X coverage in case (blue) and control cohorts (red). B. The percentage of read coverage at each depth in case (blue) and control (red) cohorts. C. Average base pair reads at 1X, 4X, 8X, 20X and 30X coverage in cases with (blue) or without (red) de novo events. The error bars represent s.e.m.

11 Supplementary Table 1 De novo mutations identified in 22 control trios Gene Symbol Mutation Type NS vs S Chr:Pos Nucleotide change Aminoacid change Sex Trio ID LY75 SNV NS 2: TGT-gGT p.cys1511gly F C trio_054 KAZALD1 SNV S 10: CTG-tTG p.leu284leu F C trio_059 MYO5B SNV NS 18: CAG-CgG p.gln430arg F C trio_062 CPSF4 SNV NS 7: CGG-CaG p.arg144gln M C trio_068 DGKA SNV S 12: CCC-CCg p.pro724pro M C trio_069 SCARB1 SNV NS 12: CGC-CtC p.arg174leu F C trio_076 KIAA0430 SNV S 16: GGG-GGc p.gly231gly F C trio_076 MAP3K5 SNV Splice 6: T/C - M C trio_070 NS = Non-synonymous; S = Synonymous; C = Control

12 Supplementary Table 2 Phenotypic variables in cases with de novo mutations Sex Age Diagnosis Mental Retardation History of Learning Disabilities History of Developmental Delay Age at Disease Onset Education (yrs) # Hospitalizations History of Suicide Attempts # Suicide attempts Male 38 SCZ NO NO NO NO 0 MAGEC1 Male 31 SCZ NO NO NO NO 0 WDR11,TEKT5 Male 43 SCZ NO YES YES YES 2 THBS1 Male 31 SCZ NO NO NO YES 1 DPYD,CELF2 Female 36 SCZ NO YES NO NO 0 OR4C46 Male 35 SCZ NO NO YES U U PAG1 Gene(s) Affected Female 29 SCZ NO NO NO NO 0 UGT1A4,NPRL2,ADAMTS3,SPATA5 Male 17 SCZ NO NO NO 16 U 3 U U FAM3D Male 38 SCZ NO NO NO NO 0 RB1CC1 Male 32 SCZ NO NO YES NO 0 KLF12,TRRAP Male 20 SCZ NO NO NO NO 0 ADCY7 Female 26 SCZAFFdpr NO NO NO YES 6 GIF Male 33 SCZ NO NO NO NO 0 PLCL2,VPS35 Male 28 SCZ NO NO NO NO 0 PITPNM1 Male 23 SCZAFFdpr NO NO NO U U GPR153,RGS12 Male 23 SCZ NO NO YES NO 0 ESAM Male 34 SCZAFFdpr NO NO NO NO 0 PML Female 31 SCZ NO NO NO NO 0 ACOT6,SAP30BP Female 43 SCZAFF-bp NO NO NO YES 1 ZNF530,SLC26A8 Female 40 SCZ NO YES YES YES 1 MTOR,CCDC108,GPR115 Female 49 SCZ NO NO NO NO 0 TRAK1 Male 22 SCZ NO NO NO NO 0 COL3A1,FASTKD5 Male 35 SCZ NO NO NO NO 0 DGCR2 Male 20 SCZ NO NO NO U U LAMA2 Female 31 SCZAFF-bp NO NO NO NO 0 INPP5A Male 31 SCZ NO NO NO NO 0 SLC26A7 Male 24 SCZ NO NO NO NO 0 EDEM2 SCZ: Schizophrenia ; SCZAFF-dpr: Schizoaffective disorder, depressed subtype ;SCZAFF-bp: Schizoaffective disorder, bipolar subtype ;U: Unknown Note: Subject Age, # Hospitalizations, as well as History and Number of Suicide Attempts are all reported for the time of admission to the study.

13 Supplementary References 1. Kim, H.G. et al. WDR11, a WD protein that interacts with transcription factor EMX1, is mutated in idiopathic hypogonadotropic hypogonadism and Kallmann syndrome. Am J Hum Genet 87, (2010). 2. Mizokami, A. et al. Phospholipase C-related inactive protein is involved in trafficking of gamma2 subunit-containing GABA(A) receptors to the cell surface. J Neurosci 27, (2007). 3. Tomiyama, K. et al. Orofacial movements in phospholipase C-related catalytically inactive protein-1/2 double knockout mice: Effect of the GABAergic agent diazepam and the D(1) dopamine receptor agonist SKF Synapse 64, (2010). 4. Brickley, K. & Stephenson, F.A. Trafficking Kinesin Protein (TRAK)-mediated Transport of Mitochondria in Axons of Hippocampal Neurons. J Biol Chem 286, (2011). 5. Gilbert, S.L. et al. Trak1 mutation disrupts GABA(A) receptor homeostasis in hypertonic mice. Nat Genet 38, (2006). 6. Le-Niculescu, H. et al. Convergent functional genomics of genome-wide association data for bipolar disorder: comprehensive identification of candidate genes, pathways and mechanisms. Am J Med Genet B Neuropsychiatr Genet 150B, (2009). 7. Gan, B. & Guan, J.L. FIP200, a key signaling node to coordinately regulate various cellular processes. Cell Signal 20, (2008). 8. Liang, C.C., Wang, C., Peng, X., Gan, B. & Guan, J.L. Neural-specific deletion of FIP200 leads to cerebellar degeneration caused by increased neuronal death and axon degeneration. J Biol Chem 285, (2010). 9. Jones, K.J. et al. The expanding phenotype of laminin alpha2 chain (merosin) abnormalities: case series and review. J Med Genet 38, (2001). 10. O'Roak, B.J. et al. Exome sequencing in sporadic autism spectrum disorders identifies severe de novo mutations. Nat Genet (2011).