Identification of Gene Ontologies Linked to Prefrontal–Hippocampal

Retraction and Correction

RETRACTION CORRECTION

SYSTEMS BIOLOGY CELL BIOLOGY Retraction for “Identification of gene ontologies linked to pre- Correction for “STAT1-induced ASPP2 transcription identifies a frontal–hippocampal functional coupling in the human brain,” by link between neuroinflammation, cell polarity, and tumor sup- Luanna Dixson, Henrik Walter, Michael Schneider, Susanne Erk, pression,” by Casmir Turnquist, Yihua Wang, David T. Severson, Axel Schäfer, Leila Haddad, Oliver Grimm, Manuel Mattheisen, Shan Zhong, Bin Sun, Jingyi Ma, Stefan N. Constaninescu, Olaf Markus M. Nöthen, Sven Cichon, Stephanie H. Witt, Marcella Ansorge, Helen B. Stolp, Zoltán Molnár, Francis G. Szele, and Rietschel, Sebastian Mohnke, Nina Seiferth, Andreas Heinz, Heike Xin Lu, which appeared in issue 27, July 8, 2014, of Proc Natl Tost, and Andreas Meyer-Lindenberg, which appeared in issue 26, Acad Sci USA (111:9834–9839; first published June 23, 2014; July 1, 2014, of Proc Natl Acad Sci USA (111:9657–9662; first 10.1073/pnas.1407898111). published June 16, 2014; 10.1073/pnas.1404082111). The authors note that the author name Constaninescu should The authors wish to note the following: “In this paper we re- instead appear as Constantinescu. The corrected author line ap- port an association of the ‘synapse organization and biogenesis’ pears below. The online version has been corrected. gene set with a neuroimaging phenotype, using gene set enrichment methodology. The methods and results of the paper, as Casmir Turnquist, Yihua Wang, David T. Severson, described, have been conducted after consultation with experts Shan Zhong, Bin Sun, Jingyi Ma, Stefan N. Constantinescu, in the field and support this conclusion. However, a potential Olaf Ansorge, Helen B. Stolp, Zoltán Molnár, confound relating to statistical inference has been brought to our Francis G. Szele, and Xin Lu attention that arises from the fact that several clustered genes, all of which are included in this gene set, have been tagged by the www.pnas.org/cgi/doi/10.1073/pnas.1415682111 same SNP. This problem, which concerns only a small fraction of our tested gene sets (unfortunately including our top finding), belongs to a known category of potential pitfalls in gene set association analyses, and we are sorry that this problem was not detected earlier. Our reanalyses suggest that if adjustments for this confound are applied, the results for our top finding no longer reach experiment-wide significance. Therefore, we feel that the presented findings are not currently sufficiently robust to provide definitive support for the conclusions of our paper, and that an extensive reanalysis of the data is required. The authors have therefore unanimously decided to retract this paper at this time.”

Luanna Dixson Henrik Walter Michael Schneider Susanne Erk Axel Schäfer Leila Haddad Oliver Grimm Manuel Mattheisen Markus M. Nöthen Sven Cichon Stephanie H. Witt Marcella Rietschel Sebastian Mohnke Nina Seiferth Andreas Heinz Heike Tost Andreas Meyer-Lindenberg

www.pnas.org/cgi/doi/10.1073/pnas.1414905111

13582 | PNAS | September 16, 2014 | vol. 111 | no. 37 www.pnas.org Downloaded by guest on September 30, 2021 Identification of gene ontologies linked to prefrontal– hippocampal functional coupling in the human brain

Luanna Dixsona,1, Henrik Walterb,1, Michael Schneidera, Susanne Erkb, Axel Schäfera, Leila Haddada, Oliver Grimma, Manuel Mattheisenc,d,e, Markus M. Nöthenc,d,e, Sven Cichonc,d, Stephanie H. Wittf, Marcella Rietschelf, Sebastian Mohnkeb, Nina Seiferthb, Andreas Heinzb, Heike Tosta,1, and Andreas Meyer-Lindenberga,1,2

Departments of aPsychiatry and Psychotherapy and fGenetic Epidemiology in Psychiatry, Central Institute of Mental Health, Medical Faculty Mannheim, University of Heidelberg, 68159 Mannheim, Germany; bDepartment of Psychiatry, Division of Mind and Brain Research, Charité Campus Mitte, 10117 Berlin, Germany; cDepartment of Genomics, Life and Brain Center, dInstitute of Human Genetics, and eInstitute for Genomic Mathematics, University of Bonn, 53127 Bonn, Germany

Edited by Robert Desimone, Massachusetts Institute of Technology, Cambridge, MA, and approved May 16, 2014 (received for review March 5, 2014)

Functional interactions between the dorsolateral prefrontal cor- Previous studies have characterized abnormal prefrontal– tex and hippocampus during working memory have been stud- hippocampal interactions in subjects with genetic risk factors for ied extensively as an intermediate phenotype for schizophrenia. schizophrenia (4, 9, 10, 16). In particular, genome-wide associ- Coupling abnormalities have been found in patients, their un- ation studies (GWAS) have become a standard approach for affected siblings, and carriers of common genetic variants associ- identifying common variants that may contribute to risk pheno- ated with schizophrenia, but the global genetic architecture of this types in structural and functional neuroimaging data (10, 16, 17). imaging phenotype is unclear. To achieve genome-wide hypothesis- However, although this approach has been effective in identify- free identification of genes and pathways associated with pre- ing genetic risk variants for imaging phenotypes, post hoc in- frontal–hippocampal interactions, we combined gene set enrich- terpretation of results is challenging. Detected risk variants often ment analysis with whole-genome genotyping and functional fall within intronic sequences, where a lack of prior knowledge magnetic resonance imaging data from 269 healthy German volun- on functionality hinders a mechanistic explanation of how they teers. We found significant enrichment of the synapse organization impact brain function (18). and biogenesis gene set. This gene set included known schizophre- Increasing evidence suggests that common genetic risk variants NRCAM for psychiatric disorders are not distributed randomly but rather SYSTEMS BIOLOGY nia risk genes, such as neural cell adhesion molecule ( )and – calcium channel, voltage-dependent, beta 2 subunit (CACNB2), as lie among sets of genes with overlapping functions (19 22). Gene well as genes with well-defined roles in neurodevelopmental and set enrichment analysis (GSEA) is a data analytical approach that plasticity processes that are dysfunctional in schizophrenia and have leverages a priori knowledge to gain insight into the biological functions of genes and pathways in the analysis of genetic data (23, mechanistic links to prefrontal–hippocampal functional interactions. 24). This approach relies on analysis of sets of genes grouped by Our results demonstrate a readily generalizable approach that can common biological characteristics, such as a shared role in par- be used to identify the neurogenetic basis of systems-level pheno- ticular molecular functions or metabolic pathways. GSEA can then types. Moreover, our findings identify gene sets in which genetic be used to test whether genes that are more strongly associated with variation may contribute to disease risk through altered prefrontal– hippocampal functional interactions and suggest a link to both on- going and developmental synaptic plasticity. Significance functional connectivity | GSEA | endophenotype | genetic risk variants This study combines neuroimaging and whole-genome genotyping techniques with a gene set enrichment analysis to unravel the genetic basis of a well-validated intermediate maging genetics is widely used to identify neural circuits linked phenotype for schizophrenia, dorsolateral prefrontal cortex– to genetic risk for heritable neuropsychiatric disorders, such as I hippocampal connectivity. We found significant enrichment of schizophrenia, autism, or bipolar disorder (1). A well-established genes with roles in synaptic plasticity and neurodevelopment imaging genetics phenotype is functional connectivity between the that are consistent with the neurobiological basis of pre- right dorsolateral prefrontal cortex (DLPFC) and the left hippo- – – frontal hippocampal interactions in schizophrenia. We further campus (HC) during working memory (WM) performance (2 4). provide additional independent evidence for the intermediate Specifically, impaired interaction of the HC and prefrontal cortex phenotype concept and present a readily generalizable ap- (PFC) has been proposed as a core abnormality during neuro- proach for a biologically driven analysis of imaging and development in schizophrenia. The hippocampus provides input to genetic data. the DLPFC through long-range glutamatergic connections, which have been linked to the glutamate hypothesis of the illness. Author contributions: L.D., H.W., M.S., M.M.N., M.R., A.H., H.T., and A.M.-L. designed Moreover, selective lesions of the hippocampus in primates and research; L.D., H.W., M.S., S.E., A.S., L.H., O.G., M.M., M.M.N., S.C., S.H.W., M.R., S.M., rodents have been shown to result in postpubescent changes in N.S., A.H., H.T., and A.M.-L. performed research; L.D., H.W., M.S., M.M., S.C., S.H.W., H.T., and A.M.-L. analyzed data; and L.D., M.S., H.T., and A.M.-L. wrote the paper. prefrontal regions that are consistent with neuropathological find- Conflict of interest statement: A.M.-L. has received consultant fees and travel expenses ings in schizophrenic patients (5, 6). Brain physiology during WM from Alexza Pharmaceuticals, AstraZeneca, Bristol-Myers Squibb, Defined Health, Deci- performance is highly heritable (7), and anomalies of prefrontal– sion Resources, Desitin Arzneimittel, Elsevier, F. Hoffmann–La Roche, Gerson Lehrman hippocampal functional coupling during WM have been identified Group, Grupo Ferrer, Les Laboratoires Servier, Lilly Deutschland, Lundbeck Foundation, in schizophrenia patients (1, 2, 4, 8), their unaffected first-grade Outcome Sciences, Outcome Europe, PriceSpective, and Roche Pharma and has received speaker’s fees from Abbott, AstraZeneca, BASF, Bristol-Myers Squibb, GlaxoSmithKline, relatives (4), healthy carriers of genome-wide supported schizo- Janssen-Cilag, Lundbeck, Pfizer Pharma, and Servier Deutschland. No other disclosures phrenia risk variants and subjects at risk (4, 9–12), and in genetic were reported. animal models of the disorder (13). These studies provide strong This article is a PNAS Direct Submission. support for a role of this neural systems-level phenotype in 1L.D., H.W., H.T., and A.M.-L. contributed equally to this work. schizophrenia pathophysiology and correspond well to current 2To whom correspondence should be addressed. E-mail: [email protected]. “ theories that conceptualize the illness as a brain disconnection This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. syndrome” rooted in disturbed synaptic plasticity processes (14, 15). 1073/pnas.1404082111/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1404082111 PNAS Early Edition | 1of6 a phenotype of interest tend to significantly aggregate within spe- biogenesis category is also modest in size, with 23 genes, in- cific biologically based “gene sets.” As an adjunct to established dicating that it maps to a lower, “more specialized” level of the GWA studies and candidate gene approaches, GSEA has success- gene ontology graph. Further details about this significant category, fully identified genes sets with established risk genes for complex including HUGO Gene Nomenclature Committee gene symbols diseases such as lung cancer, Parkinson’s disease, and psychiatric (30), full gene names, gene association P values, position of each disorders, yielding insight into plausible biological processes and gene on the ranked gene list, and gene enrichment scores, are molecular mechanisms warranting further investigation (24–26). showninTable1. Although in principle the same strategy can be applied to other quantitative risk-associated phenotypes (27), no prior Discussion study has attempted to identify shared biological pathways GSEA analysis of the biological process ontology revealed a sig- linked to individual variation in DLPFC–HC functional coupling nificant enrichment of genes encoding proteins integral to the through a combination of GSEA, whole-genome genotype data, formation, maintenance, and function of synapses in the brain. and neuroimaging. Here we used GSEA to test the association of Genes within this category have been associated with schizo- ontology-based gene sets derived from common genetic variants phrenia risk in previous studies and impact several downstream with prefrontal–hippocampal interactions in 269 healthy volun- processes and signaling pathways, including cellular adhesion and teers who performed the n-back WM task during functional trans synaptic signaling processes [protocadherin genes (PCDHs), magnetic resonance imaging (fMRI), a well-established paradigm NLGN1, neural cell adhesion molecule gene (NRCAM), agrin to challenge DLPFC–HC interactions. Given the reviewed evi- gene (AGRN)], organization and function of synaptic cytomatrix dence (14, 15), we hypothesized that we would identify gene sets and scaffold complexes [AGRN, collagen type IV alpha 4 gene linked to developmental plasticity and synaptic neurotransmis- (COL4A4)], calcium signaling [voltage-dependent calcium chan- sion, including previously identified risk genes for schizophrenia. nelbeta2subunitgene(CACNB2), protocadherin beta 10 gene (PCDH10)], and ion channel and neurotransmitter function Results [CACNB2, nicotinic cholinergic receptor alpha 1 subunit gene Consistent with previous studies (2, 9, 28, 29), healthy subjects (CHRNA1), acetylcholinesterase gene (ACHE)]. Therefore, our showed robust activation increases in the 2-back condition relative findings both shed light on the underlying genetic architecture of to the 0-back condition in cortical areas linked to WM performance, the systems-level phenotype we studied and suggest ways in which including the right DLPFC (Fig. 1A). Additionally, our functional abnormalities in the interaction between the PFC and HC may connectivity analysis with the right DLPFC seed revealed a relatively be linked to genetic risk for schizophrenia. uniform negative connectivity with the left hippocampus (Fig. 1B). The establishment and maturation of appropriate synaptic GSEA of genome-wide contributions to the DLPFC–HC cou- connections is crucial for neural circuit development and plas- pling phenotype revealed significant enrichment of the “synapse ticity in the developing and mature brain. There is persuasive organization and biogenesis” category within the biological pro- evidence linking dysfunction at the synapse level to schizo- cess ontology. Genes that contributed to the core enrichment phrenia pathophysiology (14, 31–34). Alterations to synaptic signal of this gene set included genes previously implicated in microcircuitry within the DLPFC of patients with schizophrenia schizophrenia pathophysiology, such as neuroligin 1 (NLGN1) include reduced excitatory inputs at layer 3 pyramidal neurons, (Table 1). This category was significant after correction for increases in neuronal density, and altered expression of synaptic comparison of multiple gene sets using the false discovery rate proteins (31, 35). In parallel, altered functional plasticity pro- (FDR q value = 0.04) and after conservative family-wise error cesses, such as long-term potentiation, which form the molecular correction (FWE P value = 0.04). The synapse organization and basis of learning and memory, are disrupted in animal genetic and behavioral models of schizophrenia (14, 31). Our data add to a growing body of genetic, structural, and functional evidence implicating perturbed synaptic plasticity processes in altered prefrontal network dynamics observed in schizophrenic patients (3, 34). A study of cortical thickness in patients with schizophrenia found decreases in the structural connectivity of the left and right DLPFC to be correlated with poor WM performance (36). Additionally, real-time transcranial magnetic stimulation techniques, which allow researchers to indirectly study synaptic plasticity in vivo (37, 38), have shown that inducing synaptic plasticity in the right DLPFC leads to altered prefrontal–hippocampal functional interactions during WM (3). These findings are in keeping with the “two hit” hypothesis of schizophrenia, whereby early environmental and genetic factors confer vulnerability of neural circuitry to adverse events during adolescence, such as excessive synaptic pruning or impaired plasticity (14, 39). Di- minished stabilization of these circuits is thought to contribute to enduring functional and structural alterations in the underlying neurocircuitry, leading to the altered experience-dependent plasticity and connectivity features that are associated with the cognitive and behavioral symptoms of schizophrenia. Finally, our data are highly consistent with a growing body of pathway ana- > lytic analyses that provide support for a role of synaptic gene Fig. 1. (A) WM-related BOLD activation (2-back 0-back contrast) in 269 groups in the pathophysiology of schizophrenia (22, 40–42). healthy controls for the n-back task. Functional maps (thresholded at P < 0.001, t = 3.12, uncorrected) are projected on a rendered MNI template and Adhesion. shown from a lateral (Left) and top (Right) view for presentation purposes. Among the genes included in the synapse organization Color bar represents t values. (B) Functional connectivity between right and biogenesis category are several cell adhesion genes, from the DLPFC seed and left HC during WM. Mean connectivity with right DLPFC protocadherin, neuroligin, and neural adhesion gene families. seed for 269 healthy volunteers within left HC mask is shown in sagittal Synapses in the central nervous system are formed through a series (Left) and axial (Right) sections, using an MNI structural template for pre- of reciprocal adhesion events between axons and corresponding sentation purposes. Color bar represents t values. Map coordinates refer to dendrites that regulate contact initiation, synapse formation, mat- the standard space as defined by the MNI. uration, and functional plasticity. Cell adhesion genes are gaining

2of6 | www.pnas.org/cgi/doi/10.1073/pnas.1404082111 Dixson et al. Table 1. Genes belonging to the significantly enriched “Synapse Organization and Biosynthesis” category from the Biological Process ontology Symbol Gene name Running ES Rank in gene list P

NLGN1 Neuroligin 1 −0.0073 1714 .0017 KLK8 Kallikrein-related peptidase 8 0.5455 4355 .0090 CACNB2 Calcium channel, voltage-dependent, β 2 subunit 0.3865 7188 .0106 PCDHB9 Protocadherin β 9 0.1670 2929 .0114 PCDHB10 Protocadherin β 10 0.2215 2930 .0114 PCDHB11 Protocadherin β 11 0.2760 2931 .0114 PCDHB6 Protocadherin β 6 0.3825 2994 .0114 PCDHB5 Protocadherin β 5 0.4304 3164 .0114 PCDHB16 Protocadherin β 16 0.4828 3209 .0114 PCDHB14 Protocadherin β 14 0.5203 3633 .0114 PCDHB13 Protocadherin β 13 0.6189 5135 .0114 GHRL Ghrelin/obestatin prepropeptide 0.5191 7091 .0208 NRCAM Neural cell adhesion molecule 1 0.1472 8031 .0306 PCDHB4 Protocadherin β 4 0.3281 2993 .0310 COL4A4 Collagen, type IV, α 4 0.1191 2749 .0343 POU4F1 POU class 4 homeobox 1 0.1113 8052 .0724 AGRN Agrin 0.4915 7183 .1080 ACHE ACHE 0.0319 8264 .1222 UBB Ubiquitin B 0.3756 8003 .1372 CHRNA1 Cholinergic receptor, nicotinic, α 1 0.6320 6085 .2309 NRD1 Nardilysin 0.5959 4396 .2723 PCDHB2 Protocadherin β 2 0.0125 2663 .3875 PCDHB3 Protocadherin β 3 0.0674 2664 .3875

HUGO gene symbol and full gene name are displayed in the first and second columns (from left). The SYSTEMS BIOLOGY running ES and position of each gene in the ranked gene list are shown in the third and fourth columns. Association P values for each gene are given in the final column. increasing attention within the schizophrenia research community, produce schizophrenia-like symptoms in healthy subjects, disrupt not least because of their importance in neurodevelopmental WM function in rats, and impair WM performance when admin- events important in circuit formation, such as neurite outgrowth isteredtotheDLPFCinmonkeys(58, 59). Lastly, a study has linked and synaptogenesis, but also for their roles in synaptic signaling NRCAM to the dopamine system, demonstrating that NRCAM and neurotransmission processes linked to the pathophysiology effects D2 receptor signaling by modulating the rate of D2 receptor of the disease (39, 43). internalization (60).

NRCAM. A notable gene in this category is the neuronal cell ad- Neuroligin. Another prominent adhesion gene in these categories hesion gene, NRCAM. NRCAM belongs to the cell adhesion is the NLGN1 gene. NLGN1 is a presynaptic adhesion protein molecule (CAM) family of adhesion genes and is located in the with a critical role in synapse formation. There is increasing middle of a genomic region strongly implicated in schizophrenia interest in the neuroligin family in schizophrenia because neu- etiology and has been associated with schizophrenia in a Korean roligins form transsynaptic complexes with schizophrenia- population (44, 45). Altered NRCAM levels have been reported associated neurexin (NRX) proteins (61, 62). NLGN–NRX in HC, DLPFC, and amygdala and in the cerebrospinal fluid of – NRCAM complexes are known to be important in brain development, and patients with schizophrenia (46 48). polymorphisms genetic variation in these genes has been associated with au- have also been associated with variation in neurocognitive scores – – CAM tism and schizophrenia (63 66). The NLGN1 NRX complex in patients with schizophrenia (49). As a group, genes have has been found to specifically localize to glutamatergic synapses, also been associated with schizophrenia in a pathway analytical where it helps stabilize the synapse and recruit additional syn- study (43). Transgenic mice lacking NRCAM isoforms show defi- aptic proteins involved in synapse structure and function. Several cits in learning and long-term potentiation (LTP) in the hip- lines of evidence suggest an important role for NLGN1 in pocampus and in prepulse inhibition responses, whereas mice overexpressing the extracellular domain of NRCAM exhibit WM modulation of the NMDA-type glutamate receptor (61, 67, 68). deficits and impaired plasticity in prefrontal regions (46, 50–53). In mice NLGN1 regulates the synaptic abundance of NMDA- Of particular interest is the polysialated form of NRCAM type glutamate receptors (61). Overexpression of NLGN1 in (PSA-NRCAM), which is expressed specifically within inhibitory mouse hippocampus results in increased inhibitory input and GABAergic interneurons. NRCAM-PSA–expressing interneurons increases in glutamatergic neurotransmission (61, 68). This – in mice show reduced dendritic spine numbers, decreased arbori- finding is in good agreement with the excitation inhibition hy- zation, and changes in the synaptic connectivity (54). A study of pothesis of schizophrenia, which posits that the NMDA receptor cultured hippocampal neurons showed that PSA-NRCAM is in the cortex functions as a kind of “excitatory sensor” and that required for N-methyl-D-aspartate (NMDA) receptor-dependent decreased NMDA activity, particularly within interneurons, can LTP and acts as an antagonist at N2RB-subunit–bearing NMDA lead to alteration of systems-level neural dynamics (69). receptors, preventing glutamate-induced cell death (55, 56). The Finally, there is evidence suggesting that neuroligin genes im- role of NRCAM at NMDA receptors is of particular interest given pact a number of other genes and associated pathways linked to that NMDA receptor hypofunction in the prefrontal cortex is one schizophrenia pathophysiology, including the neuregulin receptor of the leading schizophrenia hypotheses and has been linked to the signaling compound and the discs, large homolog 4 (DLG4) cognitive symptoms and oscillatory disturbances associated with molecule. However, further investigation of these downstream the disease (57). NMDA receptor antagonists have been found to gene targets is needed to elucidate underlying neurobiological

Dixson et al. PNAS Early Edition | 3of6 mechanisms contributing to modified prefrontal–hippocampal study supports a role for genes involved in synapse organization and connectivity (62). function in prefrontal–hippocampal interactions. Several genes in this pathway are highly consistent with the reported risk architec- CACNB2. Also part of the synapse organization and biogenesis ture of schizophrenia, and we provide independent evidence for the gene set is the voltage-dependent calcium ion channel family intermediate phenotype concept. In addition, we identify genes that member CACNB2. Voltage-gated calcium ion channels are are strongly implicated in the schizophrenia pathophysiology but distributed widely throughout the brain and are critical for me- have not been identified by classic genetic association studies. Our + diating intracellular Ca2 influx in response to action potentials findings highlight the value of this biologically driven method in at the synapse, and have an important role in NMDA receptor- determining the mechanisms and genes underlying intermediate independent synaptic plasticity processes (70). CACNB2 reached systems-level neural phenotypes linked to psychiatric disorders genome-wide significance in two large meta-analyses of schizo- and should be generalized to other phenotypes and disorders of phrenia GWAS, and also in bipolar disorder, which is known to interest. We expect that a more pathway-oriented approach will share a considerable genetic overlap with schizophrenia (71–73). In also be helpful in linking up systems-level observations in humans addition to these specific observations, our findings agree with with drug development in psychiatric disorders. a growing body of genetic, structural, functional, and brain expression evidence suggesting an involvement of calcium-dependent Materials and Methods regulatory processes in prefrontal–hippocampal network plas- Subjects. We studied 269 healthy German volunteers with parents and ticity (74, 75) and the pathophysiology of schizophrenia (16, 76). grandparents of European origin from the population of the cities of Mannheim, Berlin, and Bonn [mean (±SD) age 32.93 ± 9.81 y, 135 female; details in Table UBB. An interesting gene in this gene set is the ubiquitin gene. S1]. Exclusion criteria included the presence of a lifetime history of psychiatric, Ubiquitination processes have been closely linked to the as- neurological, or significant general medical illness, pregnancy, a history of head sembly, connectivity, and function of the synapse, including the trauma, and current alcohol or drug abuse. None of the volunteers had a first- turnover of pre- and postsynaptic proteins. Abnormalities of degree relative with a psychiatric disorder or received psychotropic pharma- ubiquitin gene expression have been reported in blood cells and cological treatment. All participants provided written, informed consent for the HC, PFC, and temporal gyrus of patients with schizophrenia a protocol approved by the Ethics Committee of the University of Heidelberg. (77–79). Gene sets related to the ubiquitin proteasome system Genotyping. DNA was extracted from white blood cells according to standard have also been identified in two pathway analytical studies of methods, and subjects were genotyped using HumanHap 610 and 660w Quad schizophrenia, providing further evidence for the involvement of BeadChips (Illumina Inc). A standard quality control protocol was applied to the ubiquitination processes in disease etiology (80). This is the data: minor allele frequency >0.03, individual call rate >0.98 and ge- consistent with a growing number of studies suggesting that notype call rate >0.98, and Hardy-Weinberg equilibrium P > 0.001. This ubiquitin genes may function as upstream factors impacting the resulted in a total of 486,036 SNPs for subsequent association analysis with disturbed synaptic plasticity process reported in schizophrenia functional MRI (fMRI) data. (81). Expression of genes of the ubiquitin proteasome system has been linked to positive symptom scores for schizophrenia and the Image Acquisition. Whole-brain blood oxygenation level-dependent (BOLD) dsyregulation of the ubiquitin-proteasome system has been fMRI was performed on three identical 3-Tesla MRI scanners at the Central linked to psychosis in two independent samples (80). Institute of Mental Health in Mannheim and the medical faculties of the universities of Bonn and Berlin (Siemens Trio). All three sites used identical data GHRL. Ghrelin is a metabolic peptide associated with the regu- acquisition protocols. Functional data were acquired using echo planar imaging lation of appetite and food intake; ghrelin levels in the brain sequences with the following specifications: 28 axial slices, 4-mm slice thickness, have been found to change with atypical antipsychotic treatment 1-mm gap, time to repetition/time to echo 2,000/30 ms, 80° flip angle, 192 mm × (82). Interestingly, several lines of evidence have linked ghrelin 192 mm field of view, and 64 × 64 matrix. To ensure comparable scanner to WM function (83). Infusion of ghrelin in rats to the HC and magnet stability over time, quality assurance (QA) measurements were dentate gyrus regions in adult rats promoted synaptogenesis and conducted following an established multicenter QA protocol at all sites (87). was associated with enhanced spatial WM function (84). This is fMRI WM Paradigm. – consistent with a study in healthy older subjects in which serum DLPFC function and DLPFC HC functional connectivity ghrelin was linked to poorer WM function (85). Ghrelin has also during WM was challenged with the n-back task, a well-established and reliable paradigm used extensively in imaging genetics (9, 29, 88). Briefly, been associated with the synaptic accumulation of the AMPA subjects viewed a series of digits (1–4) presented sequentially for 500 ms glutamate receptors and increases in LTP, a form of synaptic (interstimulus interval 1,500 ms). One number in each frame was highlighted plasticity that is thought to underlie learning and memory (86). to represent the target number to be maintained in memory. As the sequence progressed, the subject indicated which number corresponded to the Limitations. Although the use of biologically based gene sets as highlighted number (via a button press) shown in the current frame (0-back, units of analysis has several advantages, the GSEA of whole- control condition) or two frames previously (2-back, experimental condi- genome neuroimaging data are not without limitations. First, tion). The task is presented in eight blocks of 30 s, with alternating epochs of although ontology-based gene sets offer the most complete cat- 0-back and 2-back conditions, giving a total run length of 4 min 16 s or 128 egorization of gene properties to date, we still possess a limited whole-brain scans. To allow task familiarization and to minimize perfor- understanding of gene functions and pathways involved in nor- mance effects in the scanner, participants were instructed during test ver- mal and pathological brain function. Consequently, the gene sets sions of the paradigm offline before the scan. used here will evolve over time as new information is incor- porated, with potential effects on the reported findings. Imaging Processing and Connectivity Analysis. Image processing and statistical Lastly, although our data are consistent with prior genetic as- analyses were conducted using statistical parametric mapping methods sociation studies in schizophrenia and the neurobiology of plastic- (SPM8; www.fil.ion.ucl.ac.uk/spm/software/spm8/). Images were realigned to ity-related neural functions, our inferences on the pathophysiology the first image, slice-time corrected, and spatially normalized into a standard stereotactic space [Montreal Neurological Institute (MNI) template] with of schizophrenia rest on prior evidence linking the examined con- × × nectivity phenotype to disease risk and do not follow directly from volume units (voxels) of 3 3 3 mm and smoothed with a 9-mm FWHM Gaussian filter. Our functional connectivity analysis used a seeded connec- our data. tivity approach that closely follows a well-established approach for this risk Conclusion phenotype previously used by our group and others (4, 10). First, individual first eigenvariates of the seed time series were extracted A major challenge in schizophrenia research has been identify- from 6-mm spheres centered on the most significantly activated voxel in the ing and characterizing the complex neurobiological effects of mul- 2-back > 0-back activation contrast in the right DLPFC (Brodmann areas 46 tiple conjoined genetic variants contributing to disease risk. and lateral 9) as defined by the Wake Forest University PickAtlas (Fig. 1A Leveraging the combined power of neuroimaging and GWAS, this provides an illustration of the 2-back > 0-back activation patterns at the

4of6 | www.pnas.org/cgi/doi/10.1073/pnas.1404082111 Dixson et al. group level). The use of functionally defined individual seed regions is GSEA. GSEA tests whether particular biologically derived genes sets are over- thought to improve the sensitivity of the analysis to detect task-related represented among the loci that have been shown to be most associated effects on functional connectivity. To counter for the contribution of task- with a phenotype (24). Here we used the adapted version of the weighted related coactivation to functional connectivity measures, we further ad- Kolmogorov-Smirnov statistic following published procedures (24). In brief, justed the seed time-series for task-related variance. The time series of our genes are sorted into a ranked list by association P values and are assigned first-level functional connectivity model were high-pass filtered (128 s) and to gene sets. For each gene set an enrichment score (ES) is then calculated by adjusted for a global signal (first eigenvariate over the whole brain). To running down the ranked gene list, increasing an aggregate score when account for additional sources of variance, such as physiological noise and a gene from the gene set is met, and decreasing it if not. The ES is thus the movement, first eigenvariates from anatomical masks of the cerebrospinal maximum deviation from zero on the ranked list and reflects the degree to fluid and white matter were entered along with movement covariates into which a gene set correlates with the phenotype. The significance of each ES a whole-brain general linear model with the seed time-series as the cova- is estimated by an empirical permutation procedure, whereby gene associ- riate of interest. ation scores are shuffled across gene sets, and each ES is recalculated relative to this null distribution. The ES is then normalized to allow comparison of Statistical Inference. We analyzed the association between SNPs and DLPFC– differently sized gene sets and to correct for the comparison of multiple HC functional connectivity via the following steps. First-level β-images from gene sets using the FDR. Any gene set with an FDR q ≤ 0.05 was considered our contrast of interest (mean functional connectivity with right DLPFC seed) to be significant. were entered into a second-level multiple regression model in SPM, with age, sex, and site as covariates, and an additional four multidimensional GSEA Software and Analysis Setup. All available genes were assigned to gene scaling components to account for genetic stratification of the sample. To sets according to gene ontology category definitions provided by the Mo- capture the individual subject’s strength of functional connectivity in a sin- lecular Signatures Database (24). The gene ontology project categorizes gle parameter, we extracted the mean residual value within a standard genes into biologically related gene sets for three main domains (the mo- anatomical left hippocampus mask (Wake Forest University PickAtlas) from lecular function, biological process, and cellular component ontologies) on this model. Finally, the association of SNPs with our single-parameter mea- the basis of experimental and computational evidence gleaned from bio- sure of DLPFC–HC functional connectivity was obtained by coding the informatics databases (93). Given our a priori interest in the cellular and number of minor alleles per individual and SNP (additive model) and per- molecular events involved in schizophrenia, we assign genes to genes sets forming a genome-wide linear association analysis in Plink (89), resulting in from the biological process gene ontology. To avoid overinflating enrich- a corresponding P value for each SNP. ment scores by abnormally large or small gene sets, only gene sets with 15– 200 genes were analyzed. This step generated 346 out of a total of 825 gene Mapping SNPs to Genes. To map individual SNPs to genes and derive a gene- sets for further analysis by GSEA. To omit genes with weak evidence for level summary score, we used the LDsnpR package, previously implemented inclusion in a particular gene ontology category, we further excluded any by Ersland et al. (26, 90). The LDsnpR package was used to assign SNPs to genes that were annotated to gene ontology categories on the basis of SYSTEMS BIOLOGY genes using chromosomal position and linkage disequilibrium information electronic evidence alone (i.e., noncurated comparison of sequence similar- from Human Emsembl 66 release and the CEU (Utah residents with ancestry ity). Using these gene sets we then performed a GSEA using freely available from northern and western Europe) sample from HapMap Phase II. SNPs GSEA software developed by the Broad Institute (94). were mapped to gene bins if they were (i) located within the specified gene boundary, (ii) within an additional 20-kb window to account for proximal ACKNOWLEDGMENTS. This study was supported by the German Federal regulatory elements, and (iii) if they were in linkage disequilibrium with Ministry of Education and Research (BMBF) through the Bernstein Center for another SNP that fell within these boundaries (pairwise linkage disequilib- Computational Neuroscience (BCCN) Heidelberg/Mannheim (Grant 01GQ1003B subproject C7 to A.M.-L.) and the Integrated Genome Research Network rium threshold <0.8) (21, 25, 91). Once SNPs were mapped to gene bins the NGFNplus MooDS (Grant 01GS08147 to A.M.-L. and M.R., Grant 01GS08144 minimum SNP association P value per gene was extracted, and a gene score to M.M.N., S.C. and H.W., and Grant 01GS08148 to A.H.). A.M.-L. received was produced using a modified Sidak test as implemented in LDsnpR (92). also funding through a National Alliance for Research on Schizophrenia and This method has been shown to successfully control any length bias that may Depression Distinguished Investigator Award. H.T. receives grant support arise from large genes that bear a considerable number of SNPs (90). from the BMBF (Grant 01GQ1102).

1. Meyer-Lindenberg A (2010) From maps to mechanisms through neuroimaging of 14. Stephan KE, Baldeweg T, Friston KJ (2006) Synaptic plasticity and dysconnection in schizophrenia. Nature 468(7321):194–202. schizophrenia. Biol Psychiatry 59(10):929–939. 2. Meyer-Lindenberg AS, et al. (2005) Regionally specific disturbance of dorsolateral 15. Meyer-Lindenberg A (2011) Neuroimaging and the question of neurodegeneration in prefrontal-hippocampal functional connectivity in schizophrenia. Arch Gen Psychiatry schizophrenia. Prog Neurobiol 95(4):514–516. 62(4):379–386. 16. Erk S, et al. (2010) Brain function in carriers of a genome-wide supported bipolar 3. Bilek E, et al. (2013) Application of high-frequency repetitive transcranial magnetic disorder variant. Arch Gen Psychiatry 67(8):803–811. stimulation to the DLPFC alters human prefrontal-hippocampal functional interaction. 17. Walter H, et al. (2011) Effects of a genome-wide supported psychosis risk variant on – J Neurosci 33(16):7050–7056. neural activation during a theory-of-mind task. Mol Psychiatry 16(4):462 470. ’ 4. Rasetti R, et al. (2011) Altered cortical network dynamics: A potential intermediate 18. Sullivan PF, Daly MJ, O Donovan M (2012) Genetic architectures of psychiatric dis- – phenotype for schizophrenia and association with ZNF804A. Arch Gen Psychiatry orders: The emerging picture and its implications. Nat Rev Genet 13(8):537 551. 19. Jia P, Wang L, Meltzer HY, Zhao Z (2010) Common variants conferring risk of schizo- 68(12):1207–1217. – 5. Weinberger DR (1987) Implications of normal brain development for the pathogen- phrenia: A pathway analysis of GWAS data. Schizophr Res 122(1-3):38 42. 20. Askland K, Read C, Moore J (2009) Pathways-based analyses of whole-genome asso- esis of schizophrenia. Arch Gen Psychiatry 44(7):660–669. ciation study data in bipolar disorder reveal genes mediating ion channel activity and 6. Bertolino A, et al. (1997) Altered development of prefrontal neurons in rhesus synaptic neurotransmission. Hum Genet 125(1):63–79. monkeys with neonatal mesial temporo-limbic lesions: A proton magnetic resonance 21. Holmans P, et al.; Wellcome Trust Case-Control Consortium (2009) Gene ontology spectroscopic imaging study. Cereb Cortex 7(8):740–748. analysis of GWA study data sets provides insights into the biology of bipolar disorder. 7. Blokland GA, et al. (2011) Heritability of working memory brain activation. J Neurosci Am J Hum Genet 85(1):13–24. 31(30):10882–10890. 22. Lips ES, et al.; International Schizophrenia Consortium (2012) Functional gene group 8. Wolf RC, et al. (2009) Temporally anticorrelated brain networks during working analysis identifies synaptic gene groups as risk factor for schizophrenia. Mol Psychi- memory performance reveal aberrant prefrontal and hippocampal connectivity in pa- atry 17(10):996–1006. – tients with schizophrenia. Prog Neuropsychopharmacol Biol Psychiatry 33(8):1464 1473. 23. Mootha VK, et al. (2003) PGC-1alpha-responsive genes involved in oxidative phosphor- 9. Esslinger C, et al. (2009) Neural mechanisms of a genome-wide supported psychosis ylation are coordinately downregulated in human diabetes. Nat Genet 34(3):267–273. variant. Science 324(5927):605. 24. Subramanian A, et al. (2005) Gene set enrichment analysis: a knowledge-based approach 10. Esslinger C, et al. (2011) Cognitive state and connectivity effects of the genome-wide for interpreting genome-wide expression profiles. Proc Natl Acad Sci USA 102(43): – significant psychosis variant in ZNF804A. Neuroimage 54(3):2514 2523. 15545–15550. 11. Paulus FM, et al. (2014) Association of rs1006737 in CACNA1C with alterations in prefrontal 25. Wang K, Li M, Bucan M (2007) Pathway-based approaches for analysis of genome- – activation and fronto-hippocampal connectivity. Hum Brain Mapp 35(4):1190 1200. wide association studies. Am J Hum Genet 81(6):1278–1283. 12. Crossley NA, et al. (2009) Superior temporal lobe dysfunction and frontotemporal 26. Ersland KM, et al. (2012) Gene-based analysis of regionally enriched cortical genes in dysconnectivity in subjects at risk of psychosis and in first-episode psychosis. Hum GWAS data sets of cognitive traits and psychiatric disorders. PLoS ONE 7(2):e31687. Brain Mapp 30(12):4129–4137. 27. Potkin SG, et al. (2010) Identifying gene regulatory networks in schizophrenia. Neuro- 13. Sigurdsson T, Stark KL, Karayiorgou M, Gogos JA, Gordon JA (2010) Impaired hip- image 53(3):839–847. pocampal-prefrontal synchrony in a genetic mouse model of schizophrenia. Nature 28. Meyer-Lindenberg A, et al. (2001) Evidence for abnormal cortical functional connectivity 464(7289):763–767. during working memory in schizophrenia. Am J Psychiatry 158(11):1809–1817.

Dixson et al. PNAS Early Edition | 5of6 29. Callicott JH, et al. (2003) Abnormal fMRI response of the dorsolateral prefrontal cortex in 61. Budreck EC, et al. (2013) Neuroligin-1 controls synaptic abundance of NMDA-type cognitively intact siblings of patients with schizophrenia. Am J Psychiatry 160(4):709–719. glutamate receptors through extracellular coupling. Proc Natl Acad Sci USA 110(2): 30. Seal RL, Gordon SM, Lush MJ, Wright MW, Bruford EA (2011) genenames.org: The 725–730. HGNC resources in 2011. Nucleic Acids Res 39(Database Issue):D514–D519. 62. Bang ML, Owczarek S (2013) A matter of balance: Role of neurexin and neuroligin at 31. Faludi G, Mirnics K (2011) Synaptic changes in the brain of subjects with schizo- the synapse. Neurochem Res 38(6):1174–1189. phrenia. Int J Dev Neurosci 29(3):305–309. 63. Rujescu D, et al.; GROUP Investigators (2009) Disruption of the neurexin 1 gene is 32. Seshadri S, Zeledon M, Sawa A (2013) Synapse-specific contributions in the cortical associated with schizophrenia. Hum Mol Genet 18(5):988–996. pathology of schizophrenia. Neurobiol Dis 53:26–35. 64. Mühleisen TW, et al. (2011) Resequencing and follow-up of neurexin 1 (NRXN1) in 33. Voineskos D, Rogasch NC, Rajji TK, Fitzgerald PB, Daskalakis ZJ (2013) A review of evi- schizophrenia patients. Schizophr Res 127(1-3):35–40. dence linking disrupted neural plasticity to schizophrenia. Can J Psychiatry 58(2):86–92. 65. Kirov G, et al. (2009) Neurexin 1 (NRXN1) deletions in schizophrenia. Schizophr Bull 34. Goto Y, Yang CR, Otani S (2010) Functional and dysfunctional synaptic plasticity in 35(5):851–854. prefrontal cortex: Roles in psychiatric disorders. Biol Psychiatry 67(3):199–207. 66. Sun C, et al. (2011) Identification and functional characterization of rare mutations of 35. Lewis DA (2009) Neuroplasticity of excitatory and inhibitory cortical circuits in the neuroligin-2 gene (NLGN2) associated with schizophrenia. Hum Mol Genet 20(15): – schizophrenia. Dialogues Clin Neurosci 11(3):269–280. 3042 3051. 36. Wheeler AL, et al. (2013) Disrupted prefrontal interhemispheric structural coupling in 67. Peixoto RT, et al. (2012) Transsynaptic signaling by activity-dependent cleavage of – schizophrenia related to working memory performance. Schizophr Bull, in press. neuroligin-1. Neuron 76(2):396 409. 37. Hasan A, et al. (2011) Dysfunctional long-term potentiation-like plasticity in schizo- 68. Prange O, Wong TP, Gerrow K, Wang YT, El-Husseini A (2004) A balance between phrenia revealed by transcranial direct current stimulation. Behav Brain Res 224(1):15–22. excitatory and inhibitory synapses is controlled by PSD-95 and neuroligin. Proc Natl – 38. Rogasch NC, Daskalakis ZJ, Fitzgerald PB (2014) Cortical inhibition, excitation, and Acad Sci USA 101(38):13915 13920. 69. Kehrer C, Maziashvili N, Dugladze T, Gloveli T (2008) Altered excitatory-inhibitory connectivity in schizophrenia: A review of insights from transcranial magnetic stim- balance in the NMDA-hypofunction model of schizophrenia. Front Mol Neurosci 1:6. ulation. Schizophr Bull 40(3):685–696. 70. Moosmang S, et al. (2005) Role of hippocampal Cav1.2 Ca2+ channels in NMDA re- 39. Maynard TM, Sikich L, Lieberman JA, LaMantia AS (2001) Neural development, cell- ceptor-independent synaptic plasticity and spatial memory. J Neurosci 25(43): cell signaling, and the “two-hit” hypothesis of schizophrenia. Schizophr Bull 27(3): 9883–9892. 457–476. 71. Moskvina V, et al.; Wellcome Trust Case Control Consortium (2009) Gene-wide 40. Gilman SR, et al. (2011) Rare de novo variants associated with autism implicate a large analyses of genome-wide association data sets: evidence for multiple common risk functional network of genes involved in formation and function of synapses. Neuron alleles for schizophrenia and bipolar disorder and for overlap in genetic risk. Mol 70(5):898–907. Psychiatry 14(3):252–260. 41. Fromer M, et al. (2014) De novo mutations in schizophrenia implicate synaptic networks. 72. Sklar P, et al. (2008) Whole-genome association study of bipolar disorder. Mol Psy- – Nature 506(7487):179 184. chiatry 13(6):558–569. 42. Guilmatre A, et al. (2009) Recurrent rearrangements in synaptic and neuro- 73. Ripke S, et al.; Multicenter Genetic Studies of Schizophrenia Consortium; Psychosis developmental genes and shared biologic pathways in schizophrenia, autism, and Endophenotypes International Consortium; Wellcome Trust Case Control Consortium – mental retardation. Arch Gen Psychiatry 66(9):947 956. 2 (2013) Genome-wide association analysis identifies 13 new risk loci for schizo- ’ 43. O Dushlaine C, et al.; International Schizophrenia Consortium (2011) Molecular phrenia. Nat Genet 45(10):1150–1159. pathways involved in neuronal cell adhesion and membrane scaffolding contribute to 74. Zoladz PR, et al. (2012) Differential expression of molecular markers of synaptic – schizophrenia and bipolar disorder susceptibility. Mol Psychiatry 16(3):286 292. plasticity in the hippocampus, prefrontal cortex, and amygdala in response to 44. Lewis CM, et al. (2003) Genome scan meta-analysis of schizophrenia and bipolar spatial learning, predator exposure, and stress-induced amnesia. Hippocampus 22(3): disorder, part II: Schizophrenia. Am J Hum Genet 73(1):34–48. 577–589. 45. Kim HJ, et al. (2009) Association between neuronal cell adhesion molecule (NRCAM) 75. Spitzer NC (2006) Electrical activity in early neuronal development. Nature 444(7120): single nucleotide polymorphisms and schizophrenia in a Korean population. Psychi- 707–712. atry Clin Neurosci 63(1):123–124. 76. Bigos KL, et al. (2010) Genetic variation in CACNA1C affects brain circuitries related to 46. Brennaman LH, et al. (2011) Transgenic mice overexpressing the extracellular domain mental illness. Arch Gen Psychiatry 67(9):939–945. of NCAM are impaired in working memory and cortical plasticity. Neurobiol Dis 43(2): 77. Altar CA, et al. (2005) Deficient hippocampal neuron expression of proteasome, 372–378. ubiquitin, and mitochondrial genes in multiple schizophrenia cohorts. Biol Psychiatry 47. Gilabert-Juan J, et al. (2012) Alterations in the expression of PSA-NCAM and synaptic 58(2):85–96. proteins in the dorsolateral prefrontal cortex of psychiatric disorder patients. Neu- 78. Aston C, Jiang L, Sokolov BP (2004) Microarray analysis of postmortem temporal rosci Lett 530(1):97–102. cortex from patients with schizophrenia. J Neurosci Res 77(6):858–866. 48. Poltorak M, et al. (1996) Increased neural cell adhesion molecule in the CSF of patients 79. Middleton FA, Mirnics K, Pierri JN, Lewis DA, Levitt P (2002) Gene expression profiling with mood disorder. J Neurochem 66(4):1532–1538. reveals alterations of specific metabolic pathways in schizophrenia. J Neurosci 22(7): 49. Sullivan PF, et al. (2007) NCAM1 and neurocognition in schizophrenia. Biol Psychiatry 2718–2729. 61(7):902–910. 80. Bousman CA, et al. (2010) Preliminary evidence of ubiquitin proteasome system 50. Pillai-Nair N, et al. (2005) Neural cell adhesion molecule-secreting transgenic mice dysregulation in schizophrenia and bipolar disorder: Convergent pathway analysis display abnormalities in GABAergic interneurons and alterations in behavior. J Neurosci findings from two independent samples. Am J Med Genet B Neuropsychiatr Genet – 25(18):4659–4671. 153B(2):494 502. 51. Stork O, Welzl H, Cremer H, Schachner M (1997) Increased intermale aggression and 81. Rubio MD, Wood K, Haroutunian V, Meador-Woodruff JH (2013) Dysfunction of neuroendocrine response in mice deficient for the neural cell adhesion molecule the ubiquitin proteasome and ubiquitin-like systems in schizophrenia. Neuro- – (NCAM). Eur J Neurosci 9(6):1117–1125. psychopharmacology 38(10):1910 1920. 52. Bukalo O, et al. (2004) Conditional ablation of the neural cell adhesion molecule 82. Sentissi O, Epelbaum J, Olié JP, Poirier MF (2008) Leptin and ghrelin levels in patients reduces precision of spatial learning, long-term potentiation, and depression in the with schizophrenia during different antipsychotics treatment: A review. Schizophr – CA1 subfield of mouse hippocampus. J Neurosci 24(7):1565–1577. Bull 34(6):1189 1199. 53. Cremer H, et al. (1994) Inactivation of the N-CAM gene in mice results in size re- 83. Diano S, et al. (2006) Ghrelin controls hippocampal spine synapse density and memory performance. Nat Neurosci 9(3):381–388. duction of the olfactory bulb and deficits in spatial learning. Nature 367(6462): 84. Li E, et al. (2013) Ghrelin directly stimulates adult hippocampal neurogenesis: Im- 455–459. plications for learning and memory. Endocr J 60(6):781–789. 54. Gómez-Climent MA, et al. (2011) The polysialylated form of the neural cell adhesion 85. Spitznagel MB, et al. (2010) Serum ghrelin is inversely associated with cognitive molecule (PSA-NCAM) is expressed in a subpopulation of mature cortical interneurons function in a sample of non-demented elderly. Psychiatry Clin Neurosci 64(6):608–611. characterized by reduced structural features and connectivity. Cereb Cortex 21(5): 86. Ribeiro LF, et al. (2014) Ghrelin triggers the synaptic incorporation of AMPA receptors 1028–1041. in the hippocampus. Proc Natl Acad Sci USA 111(1):E149–E158. 55. Hammond MS, et al. (2006) Neural cell adhesion molecule-associated polysialic acid 87. Friedman L, Glover GH (2006) Report on a multicenter fMRI quality assurance pro- inhibits NR2B-containing N-methyl-D-aspartate receptors and prevents glutamate- tocol. J Magn Reson Imaging 23(6):827–839. – induced cell death. J Biol Chem 281(46):34859 34869. 88. Plichta MM, et al. (2012) Test-retest reliability of evoked BOLD signals from a cogni- 56. Kochlamazashvili G, et al. (2010) Neural cell adhesion molecule-associated polysialic tive-emotive fMRI test battery. Neuroimage 60(3):1746–1758. acid regulates synaptic plasticity and learning by restraining the signaling through 89. Purcell S, et al. (2007) PLINK: A tool set for whole-genome association and pop- – GluN2B-containing NMDA receptors. J Neurosci 30(11):4171 4183. ulation-based linkage analyses. Am J Hum Genet 81(3):559–575. 57. Marín O (2012) Interneuron dysfunction in psychiatric disorders. Nat Rev Neurosci 90. Christoforou A, et al. (2012) Linkage-disequilibrium-based binning affects the in- – 13(2):107 120. terpretation of GWASs. Am J Hum Genet 90(4):727–733. 58. Verma A, Moghaddam B (1996) NMDA receptor antagonists impair prefrontal cortex 91. Huang H, Chanda P, Alonso A, Bader JS, Arking DE (2011) Gene-based tests of asso- function as assessed via spatial delayed alternation performance in rats: Modulation ciation. PLoS Genet 7(7):e1002177. by dopamine. J Neurosci 16(1):373–379. 92. Saccone SF, et al. (2007) Cholinergic nicotinic receptor genes implicated in a nicotine 59. Dudkin KN, Chueva IV, Arinbasarov MU, Bobkova NV (2001) [Effect of a dopamine dependence association study targeting 348 candidate genes with 3713 SNPs. Hum agonist on working memory in monkeys]. Ross Fiziol Zh Im I M Sechenova 87(12): Mol Genet 16(1):36–49. 1579–1594. 93. Ashburner M, et al.; The Gene Ontology Consortium (2000) Gene ontology: Tool for 60. Xiao MF, et al. (2009) Neural cell adhesion molecule modulates dopaminergic sig- the unification of biology. Nat Genet 25(1):25–29. naling and behavior by regulating dopamine D2 receptor internalization. J Neurosci 94. Subramanian A, Kuehn H, Gould J, Tamayo P, Mesirov JP (2007) GSEA-P: A desktop 29(47):14752–14763. application for Gene Set Enrichment Analysis. Bioinformatics 23(23):3251–3253.

6of6 | www.pnas.org/cgi/doi/10.1073/pnas.1404082111 Dixson et al.