BrainSeq: Neurogenomics to Drive Novel Target Discovery for Neuropsychiatric Disorders
2015; Cell Press; Volume: 88; Issue: 6 Linguagem: Inglês
10.1016/j.neuron.2015.10.047
ISSN1097-4199
AutoresChristian Schubert, Patricio O’Donnell, Jie Quan, Jens R. Wendland, Hualin Simon Xi, Ashley R. Winslow, Enrico Domenici, Laurent Essioux, Tony Kam-Thong, David Airey, John Calley, David Collier, Hong Wang, Brian J. Eastwood, Philip J. Ebert, Yushi Liu, Laura Nisenbaum, Cara Ruble, James Scherschel, Ryan M. Smith, Hui-Rong Qian, Kalpana Merchant, Michael Didriksen, Mitsuyuki Matsumoto, Takeshi Saito, Nicholas J. Brandon, A.J. Cross, Qi Wang, Husseini K. Manji, Hartmuth C. Kolb, Maura Furey, Wayne C. Drevets, Joo Heon Shin, Andrew E. Jaffe, Yankai Jia, Richard E. Straub, Amy Deep–Soboslay, Thomas M. Hyde, Joel E. Kleinman, Daniel R. Weinberger,
Tópico(s)Gene expression and cancer classification
ResumoWe outline an ambitious project to characterize the genetic and epigenetic regulation of multiple facets of transcription in distinct brain regions across the human lifespan in samples of major neuropsychiatric disorders and controls. Initially focused on schizophrenia and mood disorders, the goal of this consortium is to elucidate the underlying molecular mechanisms of genetic associations with the goal of identifying novel therapeutic targets. The consortium currently consists of seven pharmaceutical companies and a not-for-profit medical research institution working as a precompetitive team to generate and analyze publicly available archival brain genomic data related to neuropsychiatric illness. We outline an ambitious project to characterize the genetic and epigenetic regulation of multiple facets of transcription in distinct brain regions across the human lifespan in samples of major neuropsychiatric disorders and controls. Initially focused on schizophrenia and mood disorders, the goal of this consortium is to elucidate the underlying molecular mechanisms of genetic associations with the goal of identifying novel therapeutic targets. The consortium currently consists of seven pharmaceutical companies and a not-for-profit medical research institution working as a precompetitive team to generate and analyze publicly available archival brain genomic data related to neuropsychiatric illness. It has become a virtual cliché to opine that there is an unmet medical need for the development of novel medicines for psychiatric disorders, as novel treatment approaches have stalled for decades (Hyman, 2012Hyman S.E. Revolution stalled.Sci. Transl. Med. 2012; 4 (155cm11)Crossref PubMed Scopus (183) Google Scholar). Schizophrenia in particular is facing a crisis. There has been a complete lack of success in delivering drugs against novel therapeutic targets that show consistent efficacy in clinical trials, there has been only marginal success in treating the negative and cognitive domains that are not treated by current therapies, and there are lingering adverse effect profiles with these same compounds (Coyle, 2006Coyle J.T. Glutamate and schizophrenia: beyond the dopamine hypothesis.Cell. Mol. Neurobiol. 2006; 26: 365-384Crossref PubMed Scopus (735) Google Scholar). Developing the next generation of central nervous system (CNS) therapies for psychiatric disorders requires a deeper understanding of the underlying molecular mechanisms contributing to illness based on human data. It is generally believed that the application of human genetic findings to drug discovery may provide a rational entry point to better understand disease mechanisms at the phenotypic, molecular/cellular, and neurocircuitry levels. Recent successes in psychiatric genetics include the publication of 108 significantly associated common loci for schizophrenia by the Psychiatric Genomics Consortium (PGC) identified by genome-wide association analysis (GWAS) (Schizophrenia Working Group of the Psychiatric Genomics Consortium, 2014Schizophrenia Working Group of the Psychiatric Genomics ConsortiumBiological insights from 108 schizophrenia-associated genetic loci.Nature. 2014; 511: 421-427Crossref PubMed Scopus (5107) Google Scholar) and the identification of multiple rare de novo mutations associated with autism spectrum disorders (De Rubeis et al., 2014De Rubeis S. He X. Goldberg A.P. Poultney C.S. Samocha K. Cicek A.E. Kou Y. Liu L. Fromer M. Walker S. et al.DDD StudyHomozygosity Mapping Collaborative for AutismUK10K ConsortiumSynaptic, transcriptional and chromatin genes disrupted in autism.Nature. 2014; 515: 209-215Crossref PubMed Scopus (1630) Google Scholar, Iossifov et al., 2014Iossifov I. O’Roak B.J. Sanders S.J. Ronemus M. Krumm N. Levy D. Stessman H.A. Witherspoon K.T. Vives L. Patterson K.E. et al.The contribution of de novo coding mutations to autism spectrum disorder.Nature. 2014; 515: 216-221Crossref PubMed Scopus (1515) Google Scholar). Such advances instill optimism that new therapeutic hypotheses can be derived, and ultimately tested in the clinic, based on models that will be informed by an understanding of the molecular mechanism underlying the genetic association (Schubert et al., 2014Schubert C.R. Xi H.S. Wendland J.R. O’Donnell P. Translating human genetics into novel treatment targets for schizophrenia.Neuron. 2014; 84: 537-541Abstract Full Text Full Text PDF PubMed Scopus (23) Google Scholar). However, GWAS is a variant-based association method: it identifies often broad genomic loci containing multiple, sometimes hundreds, of correlated disease-associated variants and usually does not identify the disease-associated gene(s). More importantly, the association alone says little about the biological function or role of the locus in disease. GWAS results therefore lead to four important questions: (1) which variant(s) within the identified locus are responsible for the disease association, (2) which gene(s) do they act through, (3) is there a specific pathogenic transcript, and (4) do the functional consequences relate to disease-associated altered expression? The first step in translating these exciting genetic findings into transformative medicines is the mapping of SNP-to-gene-to-function, which is critical to link the information from genetic association to disease biology through the functional impact of implicated genetic variations. The availability of large transcriptional datasets from relevant case and control tissues is a critical component for such an analysis. As well as enhancing our general understanding of disease biology, this will aid in drug development by providing a clue to a pathogenic gene product and insight to molecular directionality of the associated product, thus initiating model building based on a molecular mechanism of risk to close in on potential new targets. Here, we discuss a new precompetitive initiative launched by the Lieber Institute for Brain Development (LIBD) with pharmaceutical industry partners (Astellas Pharma, AstraZeneca, Eli Lilly and Company, F. Hoffmann-La Roche, Johnson and Johnson, Lundbeck and Pfizer) to take advantage of the emerging and previously unprecedented genetic knowledge of psychiatric disorders and technical advances in the analysis of gene expression in brain tissue: BrainSeq, A Human Brain Genomics Consortium. Utilizing the growing brain tissue repository at the LIBD, with over 1,900 human postmortem neuropsychiatric disease and control samples available, the primary goal of BrainSeq is to generate and analyze spatial and temporal neurogenomics data (e.g., genotype, RNA sequence, and DNA methylation) and to establish a public database of these results. Short-term goals of the consortium are to determine how neurogenomics data can be used to elucidate the underlying molecular mechanisms of the genetic associations being discovered for psychiatric disorders, as well as the development of new bioinformatics tools and industry standards for scientific discovery in neurogenomics. The long-term goal is to develop new preclinical hypotheses that can drive the nomination of novel treatment targets. While initially focused on schizophrenia and mood disorders and recently identified common risk loci (Schizophrenia Working Group of the Psychiatric Genomics Consortium, 2014Schizophrenia Working Group of the Psychiatric Genomics ConsortiumBiological insights from 108 schizophrenia-associated genetic loci.Nature. 2014; 511: 421-427Crossref PubMed Scopus (5107) Google Scholar), the structure of BrainSeq and the ongoing analyses have broad applications and implications for other complex genetic disorders. The LIBD brain repository additionally includes brains of individuals with anxiety disorders diagnoses and with neurodegenerative disorders and is the largest collection of brains of individuals with post-traumatic stress disorder (PTSD). BrainSeq will involve the analysis of multiple brain regions on largely the same individuals across the human lifespan, from fetal life to old age, in a large sample of control brains and in samples from various psychiatric and neurological disorders. The brain specimens of neuropsychiatrically affected individuals available as part of this project are particularly informative in that they are relatively young (mean age 46 years) and have been ill for relatively short periods of time, thus reducing the impact of chronic illness-associated epiphenomena (see Table S2). The genetic architecture of complex disorders involves primarily sequence variation that influences gene regulation (e.g., transcript abundance, splicing, novel transcript architecture, translation efficiency) rather than protein sequence (Fu et al., 2013Fu W. O’Connor T.D. Akey J.M. Genetic architecture of quantitative traits and complex diseases.Curr. Opin. Genet. Dev. 2013; 23: 678-683Crossref PubMed Scopus (23) Google Scholar). Thus, the molecular identification of genetic regulation of the transcriptome is a critical step in understanding genetic mechanisms and discovering novel isoforms that may underlie disease mechanisms and that may be targetable for the development of new therapies. In other words, regardless of the specific molecular mechanism of the genetic variation (e.g., promoter regulation, splicing, microRNA, long non-coding RNA, epigenetic processes), we argue that if the variant is not a protein coding variation, the primary readout should ultimately manifest as an effect in the transcriptome (see Figure 1). Even for regulatory variants affecting translation, it is possible to elucidate their impact by measuring the transcriptome from purified translating polyribosomes (Sterne-Weiler et al., 2013Sterne-Weiler T. Martinez-Nunez R.T. Howard J.M. Cvitovik I. Katzman S. Tariq M.A. Pourmand N. Sanford J.R. Frac-seq reveals isoform-specific recruitment to polyribosomes.Genome Res. 2013; 23: 1615-1623Crossref PubMed Scopus (66) Google Scholar), a project that can be initiated in future stages of BrainSeq. As illustrated in Figure 2, the principle of gene-to-drug approaches is based on identifying a molecular mechanism in the transcriptome that accounts for the clinical association and then building cell and animal models based on the molecular species identified. This represents an approach to model building with a high level of construct validity, modeling the specific molecular pathology of illness and genetic risk, in contrast to traditional models based on overall gene up- or downregulation. The ideal convergence of molecular association would be the identification of a specific transcript associated both with the illness state and with genetic risk, and with the risk-associated genotype predicting the same directionality of expression difference between cases and controls. In such a case, knowing the directionality of the molecular mechanism underlying the clinical association will suggest how to target it pharmacologically (i.e., with antagonists or agonists). Recent data suggest that risk-associated loci for psychiatric disorders based on common variation influence expression of specific, often previously unannotated splice variants (Tao et al., 2014Tao R. Cousijn H. Jaffe A.E. Burnet P.W. Edwards F. Eastwood S.L. Shin J.H. Lane T.A. Walker M.A. Maher B.J. et al.Expression of ZNF804A in human brain and alterations in schizophrenia, bipolar disorder, and major depressive disorder: a novel transcript fetally regulated by the psychosis risk variant rs1344706.JAMA Psychiatry. 2014; 71: 1112-1120Crossref PubMed Scopus (90) Google Scholar), illustrating the importance of in-depth RNA characterization to identify potentially pathogenic transcripts, which may be of relatively low abundance. The risk-associated transcript and its expression differential is the specific construct for building preclinical models. Full-length transcript assembly is a challenge with current short read sequence technologies (Bradnam et al., 2013Bradnam K.R. Fass J.N. Alexandrov A. Baranay P. Bechner M. Birol I. Boisvert S. Chapman J.A. Chapuis G. Chikhi R. et al.Assemblathon 2: evaluating de novo methods of genome assembly in three vertebrate species.Gigascience. 2013; 2: 10Crossref PubMed Scopus (446) Google Scholar, Steijger et al., 2013Steijger T. Abril J.F. Engström P.G. Kokocinski F. Hubbard T.J. Guigó R. Harrow J. Bertone P. RGASP ConsortiumAssessment of transcript reconstruction methods for RNA-seq.Nat. Methods. 2013; 10: 1177-1184Crossref PubMed Scopus (448) Google Scholar). The BrainSeq RNA-seq protocol involves paired-end RNA sequencing of 100 base pair (bp) fragments. Though the BrainSeq depth of sequencing is considerable (approximately 100 million reads/sample), which allows identification of relatively rare transcripts, assembly is not straightforward. As outlined below, the BrainSeq analytic team will quantify reads that are assigned to genes, exons, and splice junctions. The latter may prove especially informative in identifying previously unannotated transcript fragments for more detailed investigation, including followup with targeted long-read technology. The LIBD Postmortem Human Brain Collection (HBC) has acquired informed consent from relatives of the deceased for 751 postmortem human brain samples (average of 21 brains per month) through the Office of the Chief Medical Examiner of the State of Maryland from September 2012 through early December 2015. Additionally, the collection has 1,213 postmortem human brain tissue samples acquired via material transfer agreements, including those from the National Institute of Mental Health (NIMH), the Eunice Kennedy Shriver National Institute of Child Health and Development (NICHD) Brain Bank, the Stanley Medical Research Institute, and The Johns Hopkins University. The BrainSeq consortium has committed initial resources for completion of four phases of RNA-seq processing and analysis, based on specific brain regions and clinical diagnoses. The BrainSeq Phase I tissue cohort consists of a diagnostic mix of 746 postmortem human brain samples from the dorsolateral prefrontal cortex. Phase II samples consist of 200 patients with schizophrenia and 300 controls from mid hippocampus. Phase III involves head of caudate tissue from a similar sample of patients with schizophrenia and controls. Phase IV consists of hippocampal and medial prefrontal cortex tissues from controls and patients with bipolar disorder and major depression, who are psychotropic drug-free at time of death. All brain donations are obtained by verbal, witnessed informed consent with the next-of-kin (see Table S1). The LIBD conducts retrospective clinical diagnostic reviews on every donation, which consist of: (1) a 34-item telephone screening with the next-of-kin on the day of donation conducted by a board-certified psychiatrist or board-certified neurologist to gather demographic, medical, medication, social, psychiatric symptoms, treatment, and other clinical history; (2) a macroscopic and microscopic neuropathological examination by a board-certified neuropathologist to screen for neuritic pathology and to exclude cases with evidence of cerebrovascular accidents, tumors, contusions, and neurodegenerative disorders; (3) autopsy information (cause, manner of death, contributing conditions to death, circumstances surrounding death) and any related history gathered through the medical examiner; and (4) a comprehensive toxicological analysis to screen for illicit and basic drugs and acute drug intoxication. Further, every case has a forensic postmortem toxicological analysis, to include volatiles (ethanol), and a comprehensive drug screening (illicit drugs including cocaine and metabolites, opiates and metabolites, phencyclidine [PCP], amphetamines, antihistamines, and other prescription drugs including antidepressants, antipsychotics, and mood stabilizing agents). If any testing is incomplete through the medical examiner, supplemental toxicology testing is ordered through National Medical Services, Inc. to screen for additional drugs on a case-by-case basis, as well as for cannabinoids and nicotine and cotinine levels. Additionally, for every psychiatric case, extensive medical records review, contact with treating physicians, and/or family informant interviews (LIBD psychological autopsy interview, the Structured Clinical Interview for DSM-IV Disorders [SCID], or Mini International Neuropsychiatric Interview [MINI]) are conducted, where applicable and available. A master’s-level clinician summarizes all of these data into a neuropsychiatric narrative summary, which is then reviewed and scored by two board-certified psychiatrists, to determine lifetime psychiatric diagnoses according to DSM-IV/V. Every case has a postmortem interval (in hours) calculated as the difference between the time the brain was frozen from the time the patient was pronounced or estimated to have last been alive if the death was unattended. Every brain has a pH measure (in cerebellar tissue) for determining basic tissue quality. RNA Integrity Number (RIN) is measured on specific brain regions as they are dissected and processed for specific research studies (e.g., dorsolateral prefrontal cortex [DLPFC] RIN, hippocampal RIN, caudate RIN, medial prefrontal cortex RIN, etc.), prior to the RNA-seq data described below. All samples selected for molecular genetic analyses are dissected from deep-frozen tissue in a standardized routine by the same individual, using a dental drill to minimize tissue injury and RNA degradation. The initial four phases of the BrainSeq work are focused on dorsolateral prefrontal cortex (Brodmann area [BA] 9, 46), mid-hippocampus, caudate nucleus head, and medial prefrontal cortex (BA 32), because these are brain regions consistently and repeatedly identified as functionally altered in patients with schizophrenia and mood disorders, and the functional connection between these regions also shows alterations in patients (Cocchi et al., 2014Cocchi L. Harding I.H. Lord A. Pantelis C. Yucel M. Zalesky A. Disruption of structure-function coupling in the schizophrenia connectome.Neuroimage Clin. 2014; 4: 779-787Crossref PubMed Scopus (87) Google Scholar, Rasetti et al., 2014Rasetti R. Mattay V.S. White M.G. Sambataro F. Podell J.E. Zoltick B. Chen Q. Berman K.F. Callicott J.H. Weinberger D.R. Altered hippocampal-parahippocampal function during stimulus encoding: a potential indicator of genetic liability for schizophrenia.JAMA Psychiatry. 2014; 71: 236-247Crossref PubMed Scopus (45) Google Scholar, Rasetti and Weinberger, 2011Rasetti R. Weinberger D.R. Intermediate phenotypes in psychiatric disorders.Curr. Opin. Genet. Dev. 2011; 21: 340-348Crossref PubMed Scopus (129) Google Scholar, Tan et al., 2006Tan H.Y. Sust S. Buckholtz J.W. Mattay V.S. Meyer-Lindenberg A. Egan M.F. Weinberger D.R. Callicott J.H. Dysfunctional prefrontal regional specialization and compensation in schizophrenia.Am. J. Psychiatry. 2006; 163: 1969-1977Crossref PubMed Scopus (163) Google Scholar). Moreover, recent work using in vivo functional imaging approaches has identified physiological alterations in these brain regions as being associated with increased genetic risk for these disorders in the healthy unaffected siblings of patients (Callicott et al., 2003Callicott J.H. Egan M.F. Mattay V.S. Bertolino A. Bone A.D. Verchinksi B. Weinberger D.R. Abnormal fMRI response of the dorsolateral prefrontal cortex in cognitively intact siblings of patients with schizophrenia.Am. J. Psychiatry. 2003; 160: 709-719Crossref PubMed Scopus (396) Google Scholar, Rasetti et al., 2011Rasetti R. Sambataro F. Chen Q. Callicott J.H. Mattay V.S. Weinberger D.R. Altered cortical network dynamics: a potential intermediate phenotype for schizophrenia and association with ZNF804A.Arch. Gen. Psychiatry. 2011; 68: 1207-1217Crossref PubMed Scopus (153) Google Scholar, Sambataro et al., 2013Sambataro F. Mattay V.S. Thurin K. Safrin M. Rasetti R. Blasi G. Callicott J.H. Weinberger D.R. Altered cerebral response during cognitive control: a potential indicator of genetic liability for schizophrenia.Neuropsychopharmacology. 2013; 38: 846-853Crossref PubMed Scopus (35) Google Scholar). The large-scale datasets generated by BrainSeq provide an unprecedented opportunity to interrogate the functional consequences of genetic risk variants on the transcriptome and the epigenome in the same subject samples (see Table S1). Imputation from observed microarray-based genotyping yields more complete genetic information on every sample. Briefly, all subjects were genotyped using cerebellar DNA with an Illumina genotyping microarray platform—the majority of genotyping utilized the Illumina Human1M-Duo BeadChip with subsets of samples genotyped using the Illumina HumanHap650Yv3 BeadChip, Omni2.5 BeadChip, or Omni5 BeadChip. Pre-imputation quality control steps, performed separately for each chip type, largely mirror recommended guidelines for pre-GWAS data processing (Anderson et al., 2010Anderson C.A. Pettersson F.H. Clarke G.M. Cardon L.R. Morris A.P. Zondervan K.T. Data quality control in genetic case-control association studies.Nat. Protoc. 2010; 5: 1564-1573Crossref PubMed Scopus (761) Google Scholar) and involve removing low-quality and/or rare variants. Cleaned observed data on each genotyping platform are phased into haplotypes using SHAPEIT (Delaneau et al., 2013Delaneau O. Howie B. Cox A.J. Zagury J.F. Marchini J. Haplotype estimation using sequencing reads.Am. J. Hum. Genet. 2013; 93: 687-696Abstract Full Text Full Text PDF PubMed Scopus (243) Google Scholar) and imputed to the 1000 Genomes Phase 3 variant set reference panel for the autosomal chromosomes and Phase 1 variant set for chromosome X, using Impute2 (Marchini et al., 2007Marchini J. Howie B. Myers S. McVean G. Donnelly P. A new multipoint method for genome-wide association studies by imputation of genotypes.Nat. Genet. 2007; 39: 906-913Crossref PubMed Scopus (1970) Google Scholar). Post-imputed genetic data are then merged across genotyping platforms, resulting in over 80 million genetic variants, consisting of single nucleotide variants (SNVs), insertions and deletions (indels), duplications and repeat elements, per individual. There are approximately 7.5 million common and high-quality variants (minor allele frequency, MAF, >5%; Hardy-Weinberg Equilibrium, HWE, p value > 10−6; and non-missing genotype data across >90% of the subjects), which can be correlated with transcriptome and epigenome data to identify the regulatory role of risk-associated genetic variation in the human brain. Following RNA extraction from relevant brain regions, RNA sequencing libraries are constructed following the manufacturer’s protocols using two different Illumina kits: polyA+ selection for DLPFC samples (see additional details in Jaffe et al., 2015Jaffe A.E. Shin J. Collado-Torres L. Leek J.T. Tao R. Li C. Gao Y. Jia Y. Maher B.J. Hyde T.M. et al.Developmental regulation of human cortex transcription and its clinical relevance at single base resolution.Nat. Neurosci. 2015; 18: 154-161Crossref PubMed Scopus (100) Google Scholar) and RiboZero Standard depletion for DLPFC, hippocampus, caudate nucleus, and medial prefrontal cortex samples. These RiboZero sequencing libraries are also prepared with ERCC spike-ins for data normalization (Risso et al., 2014Risso D. Ngai J. Speed T.P. Dudoit S. Normalization of RNA-seq data using factor analysis of control genes or samples.Nat. Biotechnol. 2014; 32: 896-902Crossref PubMed Scopus (894) Google Scholar) and also employ a strand-specific protocol. Sequencing libraries are barcoded to allow for multiple samples to be run across different flow cells and lanes to reduce potential biases induced by individual lanes, and samples resequenced using the Illumina HiSeq 2000 and HiSeq 3000 paired-end 100-bp reads targeting approximately 100 million reads per sample. The resulting sequencing reads are assessed using the FastQC tool, and the metrics are retained for each sample for downstream analysis. These sequencing reads are aligned to the genome using a splice-aware aligners TopHat2 (Kim et al., 2013Kim D. Pertea G. Trapnell C. Pimentel H. Kelley R. Salzberg S.L. TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions.Genome Biol. 2013; 14: R36Crossref PubMed Scopus (8798) Google Scholar), STAR (Dobin et al., 2013Dobin A. Davis C.A. Schlesinger F. Drenkow J. Zaleski C. Jha S. Batut P. Chaisson M. Gingeras T.R. STAR: ultrafast universal RNA-seq aligner.Bioinformatics. 2013; 29: 15-21Crossref PubMed Scopus (19483) Google Scholar), or GSNAP (Wu and Nacu, 2010Wu T.D. Nacu S. Fast and SNP-tolerant detection of complex variants and splicing in short reads.Bioinformatics. 2010; 26: 873-881Crossref PubMed Scopus (1372) Google Scholar), and expression levels for different features can be summarized for each sample, for example across annotated genes and exons (Liao et al., 2014Liao Y. Smyth G.K. Shi W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features.Bioinformatics. 2014; 30: 923-930Crossref PubMed Scopus (9082) Google Scholar), splice junctions, expressed sequences (Jaffe et al., 2015Jaffe A.E. Shin J. Collado-Torres L. Leek J.T. Tao R. Li C. Gao Y. Jia Y. Maher B.J. Hyde T.M. et al.Developmental regulation of human cortex transcription and its clinical relevance at single base resolution.Nat. Neurosci. 2015; 18: 154-161Crossref PubMed Scopus (100) Google Scholar), and transcripts (Frazee et al., 2015Frazee A.C. Pertea G. Jaffe A.E. Langmead B. Salzberg S.L. Leek J.T. Ballgown bridges the gap between transcriptome assembly and expression analysis.Nat. Biotechnol. 2015; 33: 243-246Crossref PubMed Scopus (429) Google Scholar, Trapnell et al., 2013Trapnell C. Hendrickson D.G. Sauvageau M. Goff L. Rinn J.L. Pachter L. Differential analysis of gene regulation at transcript resolution with RNA-seq.Nat. Biotechnol. 2013; 31: 46-53Crossref PubMed Scopus (2431) Google Scholar). Following DNA extraction from relevant brain regions, Illumina HumanMethylation450 BeadChip microarrays (Sandoval et al., 2011Sandoval J. Heyn H. Moran S. Serra-Musach J. Pujana M.A. Bibikova M. Esteller M. Validation of a DNA methylation microarray for 450,000 CpG sites in the human genome.Epigenetics. 2011; 6: 692-702Crossref PubMed Scopus (761) Google Scholar) are run on each sample following bisulfite conversion using the manufacturer’s protocols. Resulting image files (.idat) are processed and quality controlled using the minfi Bioconductor package (Aryee et al., 2014Aryee M.J. Jaffe A.E. Corrada-Bravo H. Ladd-Acosta C. Feinberg A.P. Hansen K.D. Irizarry R.A. Minfi: a flexible and comprehensive Bioconductor package for the analysis of Infinium DNA methylation microarrays.Bioinformatics. 2014; 30: 1363-1369Crossref PubMed Scopus (2101) Google Scholar), resulting in proportion DNA methylation levels for >485,000 CpG sites across the epigenome on each sample. Schizophrenia and bipolar disorder are common, debilitating, neuropsychiatric disorders with enormous public health costs. Although efficacious, current treatments are far from ideal, as many patients fail to respond, as about one third of patients with schizophrenia are treatment-resistant (Smith et al., 2009Smith T.E. Weston C.A. Lieberman J.A. Schizophrenia (maintenance treatment).BMJ Clin Evid. 2009; 2009: 1007PubMed Google Scholar). Of those who do respond, many only improve partially, leaving residual negative symptoms and cognitive deficits. In addition, adverse drug reaction and side effects remain major issues with both typical and atypical antipsychotics (Leucht et al., 2013Leucht S. Cipriani A. Spineli L. Mavridis D. Orey D. Richter F. Samara M. Barbui C. Engel R.R. Geddes J.R. et al.Comparative efficacy and tolerability of 15 antipsychotic drugs in schizophrenia: a multiple-treatments meta-analysis.Lancet. 2013; 382: 951-962Abstract Full Text Full Text PDF PubMed Scopus (1766) Google Scholar), leading to treatment failure because of lack of tolerability and poor compliance. Antipsychotics, especially atypicals such as clozapine, are multi-target drugs, i.e., they bind to multiple neurotransmitter receptors and other molecules. However, of all the molecular targets of antipsychotics, the occupancy of the dopamine D2 receptor remains a necessary condition for antipsychotic activity of all currently marketed drugs, despite attempts to develop alternative treatments (Ginovart and Kapur, 2012Ginovart N. Kapur S. Role of dopamine D(2) receptors for antipsychotic activity.Handbook Exp. Pharmacol. 2012; 212: 27-52Crossref PubMed Scopus (105) Google Scholar). The success of the PGC (Schizophrenia Working Group of the Psychiatric Genomics Consortium, 2014Schizophrenia Working Group of the Psychiatric Genomics ConsortiumBiological insights from 108 schizophrenia-associated genetic loci.Nature. 2014; 511: 421-427Crossref PubMed Scopus (5107) Google Scholar) in identifying 108 loci associated with schizophrenia provides the opportunity to refine, and move away from, prior simple neurotransmitter hypotheses of schizophrenia. Interestingly, the dopamine D2 receptor maps to one of the 108 loci, possibly supporting the promise of the GWAS approach in identifying drug targets. Of the many other genes within loci identified by the PGC, many are highly attractive and tractable drug targets, some of which have previously been the target of drug development programs. These include GRIN2A, which encodes the NMDA (N-methyl-d-aspartate) receptor subunit NR2A, a key mediator of synaptic plasticity (Yashiro and Philpot, 2008Yashiro K. Philpot B.D. Regulation of NMDA receptor subunit expression and its implications for LTD, LTP, and metaplasticity.Neuropharmacology. 2008; 55: 1081-1094Crossref PubMed Scopus (494) Google Scholar), and GRIA1, which encodes glutamate receptor 1 (GluR1;
Referência(s)