Karen F. Berman - US grants

Through our ongoing integrative neuroimaging studies of a unique and growing cohort of medication-free patients with schizophrenia, our group continues to make progress toward its goals of understanding the nature, molecular underpinnings, underlying neurochemistry, and clinical correlates of neural systems-level dysfunction in this devastating illness. As we describe in extensive detail in Eisenberg and Berman (Neuropsychopharmacology, 2010), critical disturbances in cognitive control neural circuitry in schizophrenia not only serve as sources of marked disability in affected individuals, but also provide a valuable phenotype for testing hypotheses regarding how genes implicated in schizophrenia might contribute risk. For example, by measuring regional cerebral blood flow during the N-back continuous working memory task, we have re-confirmed an aberrant prefrontal activation pattern even in patients who perform relatively well on the task and further demonstrated profoundly aberrant connectivity in prefrontal and medial temporal lobe regions, which showed strong ability to discriminate between healthy and ill participants. This latter finding was prospectively validated in two additional data sets, suggesting that disturbances in the prefrontal-limbic functional axis may be an illness trait marker. We now have extended this work even further, reporting on a unique gene-diagnosis interaction operating on regional cerebral blood flow involving the gene coding for catechol-O-methyltransferase, COMT, which harbors common variation that is weakly but consistently associated with schizophrenia risk and strongly implicated in both prefrontal and limbic functioning during executive and affective challenge, respectively, in healthy individuals. In particular, we have identified that even at rest there exists in patients with schizophrenia an inverse relationship between dorsolateral prefrontal cortical and medial temporal lobe blood flow, which is mediated by COMT genotype. This is an effect not seen in healthy study participants and suggests an important intersection between genetically determined cortical dopaminergic tone and fundamental biases in baseline prefrontal-limbic neural network activity in patients suffering with schizophrenia. This study therefore elucidates a mechanistic explanation for variation in characteristic resting-state neural abnormalities previously identified in schizophrenia. In parallel with advancing our neuroimaging genetic efforts, we have now made a substantial expansion to our cross-modal neuroimaging studies in schizophrenia to include comprehensive positron emission tomographic (PET) assessment of the dopaminergic synapse. Previously, we have demonstrated the power of cross-modal approaches by determining both presynaptic dopamine (measured by 18F-DOPA PET) and executive function related neural activation (measured by 15O-water PET during the Wisconsin Card Sorting Task) in patients and healthy controls to show not only exaggerated striatal FDOPA uptake and impaired prefrontal activation in patients, but also, and more importantly, a highly significant negative correlation between task-related prefrontal activation and striatal dopamine uptake in patients but not controls. This study provided a crucial basis for the coexistence of two key pathophysiological hallmarks of schizophrenia.

We previously used multimodal neuroimaging to define three fundamental aspects of the brain phenotype in WS that are related to clinical features: 1) Underlying the syndromes cognitive hallmark, visuospatial construction impairment, is a neurostructural anomaly (decreased gray matter volume) and adjacent hypofunction in the parietal sulcus region of the dorsal visual processing stream. 2) Hippocampal abnormalities in regional cerebral blood flow, neurofunctional activation, and N-acetyl aspartate concentration (measured in vivo with MR spectroscopy), as well as subtle structural changes also contribute to these visuospatial construction problems. 3) Underlying the syndromes hallmark social cognition features are structural and functional abnormalities in the orbitofrontal cortex, an important social and affect regulatory region that participates in a frontoamygdalar regulatory network found to be dysfunctional in WS. Because these features were defined in extremely rare persons with WS and normal IQs, allowing us to compare WS individuals to IQ-matched healthy controls and thus obviating an important potential confound, these brain phenotypes are likely proximal to the genetic core of the syndrome.[unreadable] [unreadable] We also explored whether the wiring of the human brain is genetically influenced. In this work, we used diffusion tensor imaging (DTI), a powerful, recently developed MRI technique that allows identification of white matter architecture invisible to conventional imaging, to study extremely rare individuals with Williams Syndrome (WS). This highly uncommon genetic disorder affords a unique opportunity to study genetic regulation of white matter development such as: the regulation of cytoskeletal dynamics in neurons, and neuronal migration and targeting. Because these same participants had been studied with other imaging modalities, we were able to tailor our analysis to specific areas of grey matter structural and functional abnormality previously identified in these individuals, thus permitting a particularly incisive investigation and affording increased power to identify genetically determined effects on white matter. We found that the missing genes confer the unique cognitive and social phenotype of the syndrome by affecting the integrity of long-range, white matter connections between cortical areas, thus affecting the coordination of large ensembles of neurons. Specifically, we showed for the first time that fibers found in white matter immediately underlying gray matter regions previously shown to be abnormal, were oriented differently, gave origin to aberrant posterior tracts, and showed altered lateralization patterns in individuals with WS. The identified overall reduction of water diffusion in the brain of WS additionally revealed microscopic alterations of tissue structure. Moreover, this sample was characterized by the presence of excess longitudinal bundles above the corpus callosum and the absence of an anterior commissure in some WS cases. Given that these genes are missing from the time of conception, this study offered insights into the genetics of neural development, a largely unexplored territory. From these data, we advanced the hypothesis that one or more of the affected genes in WS control development of fibers in the final stages of development and that these fibers, normally growing in a right to left orientation, are deviated longitudinally. This report was the first delineation of white matter structural abnormalities in WS and, because these alterations in brain connectivity occurred in the context of clear clinical phenomena, they provided data indicating that the axonal tracts where abnormalities were found may be critically involved in the cognitive and social functions specifically affected in WS subjects. Our observations linked the genes in the microdeleted region of chromosome 7 to the development of long-range connectivity in the brain, and engendered a hypothesis on the mechanism and timing of action of these genes that could guide future investigations in post-mortem tissue and animal models of WS, in particular, and of white matter development, in general.[unreadable] [unreadable] In another study of WS we extended our knowledge regarding the well characterized hypersocial personality and prominent visuospatial construction impairments, building on our previous findings of functional and structural abnormalities in the hippocampus formation (HF), prefrontal regions, and the dorsal visual stream. The visual stream is divided into two processing steams: a dorsal stream which processes spatial information and a ventral stream which subserves object processing. The hallmark cognitive impairment in WS is in visuospatial construction, the ability to visualize an object (or picture) as a set of parts and construct a replica from those parts. This impairment is characterized neurophysiologically by poor performance on tests of block design or pattern construction. This led to the hypothesis that dorsal, but not the ventral stream function is compromised. Although aberrant ventral stream activation has not been found, object-related visual information that is processed in the ventral stream is a critical source of input into these abnormal regions. This study examined the neural interactions of ventral stream areas in WS using a passive face- and house-viewing fMRI paradigm. During house-viewing, significant activation differences were observed between participants with WS and a matched control group in the brain region, intraparietal sulcus (IPS) (which processes aspects of the spatial environment). Abnormal functional connectivity was found between parahippocampal gyrus (PPA) (place-processing area) and parietal cortex, and between fusiform gyrus (FFA) (face-processing area) and a network of brain regions including amygdala (fear processing area) and portions of prefrontal cortex. These results indicated that abnormal upstream visual object processing may contribute to the complex cognitive/behavioral phenotype in WS, and provided a systems-level characterization of genetically-mediated abnormalities of neural interactions. We have also employed fMRI-based retinotopic mapping and generated cortical surface models from high-resolution structural MRI to map the size and neuroanatomical changes of the primary visual cortex.

In the service of its core scientific agenda, the Section on Integrative Neuroimaging (SoIN) has continued its long-standing commitment to characterizing the neurochemical, neurogenetic, and neuropsychological contributions to neural systems relevant to mental illness. The Section has made particular progress in developing an unprecedented multimodal positron emission tomography dataset that will soon be able to answer fundamental questions about dopamine pre- and post-synaptic function in a more comprehensive way than previously possible. Enormous efforts toward data acquisition in the past year have resulted in D1-like dopamine receptor, D2-like dopamine receptor, and presynaptic dopamine synthesis whole brain measurements collected in the same, painstakingly screened healthy individuals, which only now are beginning to allow for novel analyses synthesizing these disparate but interrelated indices of dopamine functioning. We expect that insights from these experiments will allow greater perspective on the piecemeal clues gathered to date about this vital neurotransmitter systems role in supporting cognitive functions in health and in psychiatric disorders. Interim achievements have been wide-ranging and include elaboration on dopaminergic mechanisms underlying attentional control and emotional salience processing as well as discovery of novel gene-gene interactions on mnemonic related activity in the hippocampus. A distinguished lineage of experimentation has established the importance of dopamine in cognitive functions that are disrupted in neuropsychiatric illness. We and others have previously shown that genetic variation in the gene for catechol-O-methyl transferase (COMT), an enzyme responsible for regulating cortical dopamine concentrations, is an important predictor of the efficient function of frontal neural systems recruited during cognitive challenge. Extending this work, we now have shown that in healthy individuals, pharmacological inhibition of COMT at doses that do not impact gross behavioral measures of attention are able to significantly decrease neural responses to an attention control task (Magalona et al., 2013). Additionally, we have discovered specific links between dopamine system operations and the neurophysiology of facial expression processing, as reported this year in Molecular Psychiatry (Jabbi et al., 2013). Using a highly multimodal, integrative neuroimaging approach that levies advantages of positron emission tomography (PET), magnetoencephalography (MEG) and functional magnetic resonance imaging (fMRI), this study was able to tackle long-standing lack of knowledge about how emotional facial expressions are processed in the brain. This is a ripe area of research because the neural mechanisms contributing to altered processing of social cues in neuropsychiatric conditions such as schizophrenia, autism and Williams syndrome remain mysterious, preventing development of targeted biological therapies. In this study, we first characterized functional correlates of facial expression viewing with unprecedented detail by measuring sustained blood-oxygenation level-dependent (BOLD) signal (providing excellent spatial resolution; from fMRI) and transient gamma-band activity (GBA; providing excellent temporal resolution; from MEG) responses to environmentally valid, dynamic emotional cues. We then progressed to demonstrate for the first time that midbrain presynaptic dopamine synthesis (from PET) predicts these dynamic signals in regions known to code perceptual, mnemonic and experiential aspects of emotional stimuli. Other recent work has focused on defining the impact of dopamine-relevant gene variants associated with schizophrenia on dopamine synthesis and has generated new hypotheses about how sequelae of common genetic variation intersects with the biology of mental illness. This includes several epistatic hypotheses that, as discussed in Eisenberg et al (2013), have become increasingly critical. In fact, one of the major advances we have made this year has been to demonstrate with fMRI an interaction between SLC12A2 and DISC1 genetic markers on hippocampal activation, representing the first in vivo human experiments to confirm previously observed interactions of the same genes on hippocampal neuronal development. (Callicott et al., 2013) In summary, the Section on Integrative Neuroimaging has made remarkable progress toward advancing its long-range, central research aims, and, by virtue of these efforts particularly the crucial ongoing multimodal data collection is well poised for progress in the coming year.

Researchers in this group identify a potential therapeutic target for the treatment of Schizophrenia, a debilitating disorder affecting approximately 1% of the population. This year the neurobiology group published data describing genetically regulated signaling pathways involving NRG1-ErbB4 and the PI3K enzyme, p110 all of which are associated with risk for schizophrenia. Law et al (PNAS 2012) show that pharmacological inhibition of p110 blocks behavioral effects of amphetamine in a mouse model of psychosis and reverses schizophrenia-like phenotypes in a neurodevelopmental rat lesion model. The p110 inhibitor, IC87114, has been shown to increase phosphorylation in another SZ risk gene, AKT1 in the brain of treated mice which is consistent with other antipsychotic-like molecules and suggests a mechanism of action. NRG1 and ErbB4 are known to be critical in neurodevelopment, brain plasticity. Previous reports from our lab and others have shown genetic mutations in NRG1 to be associated with risk for SZ, likewise genetic variation and structural microdeletions in ErbB4 also impacts risk for illness. While NRG1 and ErbB4 null mice show behavioral patterns consistent with other SZ mouse models, human postmortem SZ brains also show an increase in NRG1 and ErbB4 expression. Genetic variations in these genes affects human brain structure and function, but the mechanisms of how these changes turns into illness remain unknown. Enzyme p110, also related to SZ can act in concert with NRG1/ErbB4 pathways downstream and has been shown in studies of lymphoblasts from patients, where there is an increase in enzyme levels. Additionally, in human SZ brain there is an increase in enzyme expression. It appears NRG1 and ErbB4 could be potential therapeutic targets however they play roles in cell physiology making them unlikely candidates for targeting. The better choice, P110 which operates downstream of NRG1/ErbB4 may prove to be targetable and provide optimum therapeutic potential. Investigators in the Genetics and Bioinformatics Core Laboratory continues to identify novel SZ susceptibility genes and characterize their mechanism of action in both normal and diseased states. Our clinical, postmortem DNA and phenotype datasets are organized for efficient analysis using web-based family transmission and case-control methods. Genetic variants, genotypes, and statistical genetics results are shared with various phenotyping groups, including investigators in the Clinical Neuropsychology, and Neuroimaging Core Lab, investigators in the postmortem section and in our other research labs investigating risk genes and their biological impact. We select and prioritize functional and positional candidate genes based on the literature, in silico searches of interacting protein networks, and on new findings from ongoing collaborations. We also continue to identify and genotype variations in existing candidate genes and tests them for association with SZ, intermediate phenotypes from the Clinical Brain Disorders Branch, and expression phenotypes in human postmortem brain, cellular and animal model systems. This past year Zhang et al (Biol Psych 2011) performed association studies in 4 cohorts of European ancestry of a newly identified SNP (rs7597593) in ZNF804A, a previously described risk gene for SZ. We measured the SNP effect on mRNA expression using postmortem human brain. Since GWAS are generally used to identify common genetic variations in common diseases but less successful for identifying genetic variants in complex illnesses, like SZ, our study provides supportive evidence of an association of rs7597593 with risk for SZ that is also female-driven. A trend of sex-SNP interaction is seen in both, the combined 4 samples and US Gain cohort, the largest of all the samples. Risk association was seen at the level of clinical risk and in postmortem brain mRNA expression. Association and the sex-driven effect on risk were observed in 3 of the 4 cohorts (German, Scottish and US GAIN) individually as well as in the combined 4 case-control cohort sample, but statistical significance is not seen in the CBDB US cohort, whose limitation was more than likely sample size. To date, the function of the ZNF804A gene remains unknown. The results of the mRNA expression in postmortem brain and the sex-driven association of ZNF804A suggest a molecular mechanism between sex and rs7597593 on risk. Based on this study we are unable to ascribe causation to these genetic associations therefore additional studies of gene-gene interactions may help reveal the mechanism through which ZNF804A genetic variants affect risk for disease. Another group is our Transgenic Mouse and Cellular Models Lab, which translates human genetic mutations into genetic mouse models as an important strategy to study the pathogenesis of schizophrenia, identify potential drug targets, and tests new drugs for antipsychotic treatments. It is certainly impossible to capture the full spectrum of schizophrenia symptoms in animal models and as mentioned earlier in the Law, PNAS article rodent models have been successful in reproducing several schizophrenia-like behaviors and uncovering the roles of specific genes in dopamine and glutamine neurotransmission systems in mediating schizophrenia-like behaviors. Discoveries of susceptibility genes for schizophrenia and targeting cognitive dysfunction as a core feature of the disorder, provides the opportunity to develop and test newer genetic mouse models based on susceptibility. Although genetic mouse models based on genetic susceptibility are relatively new, we continue to study the roles of susceptibility genes in cognitive processing, neuronal function, and signal transduction in the brain during development. Examining candidate risk genes interactions with environmental factors, will most likely give us a better understanding of the molecular mechanisms of the pathophysiology of schizophrenia, reveal the molecular basis of normal cognitive function and human brain development, and guide us to novel antipsychotic therapies. Lastly, we look at how gene COMT relate to the biology and potential treatment of schizophrenia. The Transgenic Mouse and Cellular Models Lab explored the orientation and cellular distribution of Membrane-bound COMT. As been previously noted, COMT is a schizophrenia risk gene and a key enzyme for inactivating and metabolizing catechols, like dopamine, and plays a role in cognition, arousal, pain sensitivity and stress reactivity in humans and animal models. There are two forms of COMT, soluble (S) and membrane-bound (MB). In brain, MB is prevalent, but its neural cellular distribution and orientation are unclear. Chen et al (J Biol Chem 2011) show that MB is located in the neuron cell body, axons and dendrites in rat brain in addition MB orientation has the C-terminal catalytic domain in the extracellular space. This suggests MB has the capability to inactivate synaptic and extrasynaptic dopamine on the surface of pre-and postsynaptic neurons. We also show that the COMT inhibitor, tolcapone induces cell death via apoptosis and its cytotoxicity is dose dependent and correlated with COMT val/met genotype in human lymphoblasts. These data show that inhibitors impermeable to cell membrane in brain can be developed and for those who show drug sensitivity (COMT val/val genotype), use of low doses on a specific genetic background may ameliorate toxic effects of the drug.

Following are descriptions of our most significant work related to this project over the past year. Paragraphs 1-3 cover work that falls within the phenomics area described in our goals and objectives. Paragraphs 2 and 4 address heterogeneity reduction and subgrouping themes also described in the goals and objectives. Paragraphs 5-7 address other related research activities in connection with this project. (1) Earlier work by the group summarized the Branch's cognitive data using six cognitive domain composite variables (for verbal memory, visual memory, n-back working memory, processing speed, card sorting, and span working memory) and a higher order factor reflecting general cognitive ability, also called g. A recent application of these composite scores has been to explore genome-wide genetic associations in the CBDB samples. We identified an exciting and novel association between our general cognitive composite, g, and a genetic variant related to sodium channel biology that helps to explain the cognitive impairment in our sample of people with schizophrenia and in their unaffected siblings. This association would not have been detected without the cognitive data aggregation strategy and composite scores developed by the Neuropsychology group. The association is also supported by analyses of genotype effects in fMRI data and analyses of gene transcript expression in post-mortem brain tissue. These findings have been presented at genetics and psychiatry conferences and a comprehensive manuscript is under review at JAMA Psychiatry (Dickinson et al., 2014. (2)As noted in our goals and objectives, illness heterogeneity is a major challenge in schizophrenia research. Many kinds of studies are handicapped by a focus on broad, undifferentiated diagnosis. Using different strategies to identify robust, more homogeneous behavioral/clinical subgroups offers one means for disentangling illness heterogeneity. The large, comprehensively assessed Branch samples are well-suited for such analyses and the phenomics work already described helps create a foundation for subgrouping analyses. In Cole et al. (2012), we derived developmental trajectory subtypes based on academic and social adjustment during childhood and adolescence. With more detailed developmental data and a larger schizophrenia sample, we are conducting analyses to refine and extend the earlier analyses. Other current work addressing illness heterogeneity is seeking to extend these subgrouping strategies, simultaneously employing data from complementary data-streams (e.g., pre-onset development with post-onset symptoms and treatment response). The ultimate goal is to use robust subgrouping schemes to find genetic, brain structure, and other biological associations within or between subgroups that cannot be detected in analyses done at the level of broad diagnosis. (3) For example, there is considerable interest in the field in using cognitive variables to define psychosis subgroups (e.g., Dickinson, 2014). One current project is combining information about current cognitive performance (i.e., current IQ) with indexes of cognitive performance prior to illness onset to define cognitive developmental trajectory subgroups. Statistical clustering analyses have identified a three-subgroup scheme comprising an early impairment subgroup, an emerging cognitive impairment subgroup, and a preserved/limited impairment subgroup. Current analyses are focused on whether this subgrouping scheme provides leverage, through reduced heterogeneity, in analyses of genetics and neuroimaging data. (4) The group's work extends beyond cognition to other sorts of behavioral data. Earlier work (e.g., Wallwork et al., 2012) described our validation of a five-dimension structure that better reflects psychotic symptom data from the Positive and Negative Syndrome Scale (PANSS) than the three dimensions originally proposed for that scale. These analyses supported construction of new PANSS composite scores for positive, negative, agitated, concrete/disorganized symptoms, and general distress. These composites are now in use in CBDB neuroimaging and genetics studies. As a further example of efforts to address illness heterogeneity, ongoing analyses using are examining how the PANSS composites may be helpful in identifying illness subgroups (in particular, a high negative symptom/low distress or deficit syndrome subgroup). Other current work is examining how PANSS data relates to cognitive and brain structure data in our schizophrenia sample and in the unaffected siblings of these cases (some of the siblings show sub-clinical levels of symptomatology). (5) Other lines of work are examining the dimensions that underlie typical and abnormal personality in the CBDB data. Leading theories of personality posit five dimensions (neuroticism, extraversion, openness to experience, conscientiousness, and agreeableness). Using this model as a starting point, we have elaborated and are refining new dimensional models for the Tri-dimensional Personality Questionnaire (TPQ) and for the SCID-II Personality Questionnaire (SCID-II). The TPQ targets personality in the non-clinical range. The SCID-II is used to assess disordered personality symptomatology. All of this work has been presented at scientific conferences and a paper on the SCID-II analyses is nearing completion (6)Another part of our phenomics work seeks to estimate the degree to which different variables are heritable - that is, under genetic control. Recent efforts have focused on a simple neuromotor index from our neurological examination of CTNB protocol participants, involving the sequencing of hand movements (complex motor sequencing or CMS). CMS shows a clear association with schizophrenia and increased risk for schizophrenia, trait-like characteristics, substantial familiality/heritability (at a level similar to levels reported for individual cognitive measures), strong association with general cognitive ability, which is among the most widely-used behavioral intermediate phenotypes in schizophrenia genetics studies, directionally consistent association with the sodium channel genetic marker previously shown to associate with cognition in this sample (see (1)). (7)We have recently completed an update of earlier quantitative reviews of cognitive impairment in schizophrenia, showing that the cognitive impairment seen in people with schizophrenia has been consistent in magnitude and pattern over the past 30 years, and across different geographic regions around the world (i.e., North America, Europe and Asia). This work has been presented at conferences, published as a chapter in an edited collection (Dickinson et al., 2013), and as a journal article (Schaefer et al., 2013)

The past year has seen a number of accomplishments. Several studies in patients with schizophrenia have been completed. First, in medication-free inpatients with schizophrenia we have shown that a genetic alteration in the BDNF gene, which affects release of brain-derived neurothrophin factor in the brain, impacts hippocampal function differentially in patients off antipsychotic medication compared to controls. Specifically, in patients, the Met allele was associated with decreased hippocampal regional cerebral blood flow (rCBF) whereas in healthy individuals, this genetic alteration was associated with increased hippocampal rCBF. Next, recognizing that the human brain is a complex biological system composed of many interacting subsystems with widespread functional networks that operate on a millisecond timescale we used magnetoencephalography (MEG) to identify functional networks of the so-called resting human brain in health and schizophrenia using two graphical tools, small worldness and degree distribution. The goal of this study was to determine whether functionally connected brain networks show non-random features in healthy controls and/or patients with schizophrenia on antipsychotics. Results showed that functional brain networks presented small world topologies and non-random degree distributions in both controls and patients with schizophrenia on antipsychotics. Additionally, we sought to understand how alterations in brain dopamine in schizophrenia and Parkinsons disease (PD) affect learning. Because (1) both types of patients share dysfunctions of dopaminergic neurotransmission, (2) previous studies of probabilistic association learning (PAL) in patients with schizophrenia on treatment with antipsychotics provided conflicting evidence for normal or abnormal probabilistic learning, and (3)studies of patients with Parkinson's receiving dopamine-augmenting drugs similarly report conflicting evidence for early PAL that elicits frontal-striatal activity, we examined the effects of dopamine-altering drugs on PAL. Patients with schizophrenia were studied before and after withdrawal from antipsychotic dopamine blockers, and PD patients were studied following treatment with dopamine-augmenting agents. We found that both blockade of dopamine receptors with antipsychotics in schizophrenia and the depletion of presynaptic dopamine that causes PD resulted in impairments of PAL. However, when dopamine receptor signaling was enhanced by removing the dopamine blocker in schizophrenia (but not by restoring the dopamine signal by replacement with dopamine augmenters in PD), some aspects of probabilistic association learning improved. We also explored the administration of tolcapone, an agent that inhibits catechol-O-methyltransferase (COMT), an enzyme that breaks down dopamine, because dopamine dysregulation is a core biological feature of schizophrenia, and because the gene that codes for COMT harbors a common variation that has been associated with schizophrenia risk and is implicated in both prefrontal and limbic functioning. Individuals who possess the high activity enzyme and have less DA in the frontal cortex have previously been found to perform worse than individuals who possess the low activity enzyme on tests of attention, concentration and memory and tolcapone effects on cognition are modulated by this variation. We have now explored the effect of tolcapone on cortical information processing during attention in healthy volunteers, hypothesizing that the drug would enhance efficiency of information processing in the prefrontal cortex. This randomized, double-blind, placebo-controlled, crossover study showed that tolcapone selectively improves the efficiency of information processing in the cortex of healthy volunteers. Our continued search for genes that affect cortical function and increase risk for developing schizophrenia led to the discovery of a variant of a potassium channel gene, KCNH2, that produces a form of the potassium channel that is specifically present in the brain, in contrast to a similar family of potassium channels present in the heart that interact with antipsychotics and can be associated with changes in the heart rhythm. The gene that produces this form of the potassium channel confers risk for schizophrenia and is highly expressed in the brains of patients. We examined drug-response data from two groups of patients: (1) those from the NIMH inpatient program in which patients are studied both while off medications for four weeks and while on antipsychotics for four weeks, and (2) those from the long-term, outpatient antipsychotic treatment (CATIE) study. In both groups of patients, antipsychotic treatment produced a significant improvement in symptoms, but patients who had the TT type of the KCHNH2 gene preferentially showed this response. Patients with this genotype were also the least likely to discontinue their antipsychotic treatment. This study may open new avenues for the development of new therapeutic strategies. Our work translating human genetic markers that are associated with risk for neuropsychiatric disorders into genetic mouse models has also produced several important advances. For example, in Carr (Behav Brain Res 2013) a mouse model lacking the DTNBP1 gene had learning and memory deficits from changes in PFC circuitry; they required more practice trials to reach a criterion. To study the effects of mutations in the risk gene, COMT, on brain functions, our group developed a knock-in mutant mouse carrying the human form of the gene. The mutant mice showed higher level of anxiety, which suggests that COMT is a genetic factor for emotional control. Since COMT mutation mainly affects the dopamine level in the prefrontal cortex, prefrontal dopamine might play an important role in anxiety. Additionally, since some cognitive functions differ in males and females Papaleo et al (PNAS 2012) tested in mice the interaction of these COMT mutations and sex, on environmental manipulations of cognitive functions such as attention, impulsivity, compulsivity, motivation, and rule-reversal learning. This work demonstrated a series of complex sex- and COMT-related effects and their interactions with environmental factors to influence specific executive cognitive domains. For example, changes in mild stress negatively affected cognitive performance in males, particularly those without the COMT gene, but not females. In contrast amphetamine treatment produced small sex-COMT genotype and sex-treatment interactions for compulsive behavior. Interestingly, females improved performance after repeated testing, worked harder and outperformed males. Since this mouse model mimics the human genotype, studies such as these may be useful for genotype-specific testing of medications such as COMT inhibitors for treatment of brain disorders including Parkinsons disease and schizophrenia.

The group has continued to make strides conducting imaging-genetic studies to explore basic molecular biology and genetics of human information processing in healthy controls and patients with schizophrenia. Interesting findings have emerged from these studies. 1) Using fMRI during working memory, the group demonstrated that altered DLPFC connectivity is found in both patients with schizophrenia as well as healthy siblings and therefore is a familial and likely heritable feature of genetic risk for schizophrenia. The group also discovered that this novel intermediate phenotype is independent of the well documented altered dorsolateral prefrontal cortex engagement in patients with schizophrenia and siblings (viz. Rasetti et al. AJP 2011). These findings add to evidence that distributed network-based neurointegrative deficits manifested as disrupted network dynamics reflect genetic risk mechanisms for schizophrenia. As an illustration that some risk genes may map onto some intermediate phenotypes but not others, the group demonstrated that ZNF804A, a putative susceptibility gene for psychosis, impacts differentially on the above neuroimaging intermediate phenotypes 1) has no effect on DLPFC engagement during working memory, and 2) modulates DLPFC-HF coupling during WM (Rasetti et al. AJP 2011). 2) In another study (Tan et al. Brain 2012), the group explored genetic control over component working memory cortical-subcortical networks in humans, and the pharmacogenetic implications of dopamine-related genes on cognition in schizophrenic patients receiving anti-dopaminergic drugs. Using predictions from basic models of dopaminergic signalling in cortical and cortical-subcortical circuitries implicated in dissociable working memory maintenance and manipulation processes, the group examined pharmacogenetic effects on cognition in the context of anti-dopaminergic drug therapy. Using dynamic causal models of functional magnetic resonance imaging in normal subjects, the group identified differentiated effects of functional polymorphisms in COMT, DRD2 and AKT1 genes on prefrontal-parietal and prefrontal-striatal circuits engaged during maintenance and manipulation, respectively. Cortical synaptic dopamine monitored by the COMT Val158Met polymorphism influenced prefrontal control of both parietal processing in working memory maintenance and striatal processing in working memory manipulation. DRD2 and AKT1 polymorphisms implicated in DRD2 signalling influenced only the prefrontal-striatal network associated with manipulation. In the context of anti-psychotic drugs, the DRD2 and AKT1 polymorphisms altered dose-response effects of anti-psychotic drugs on cognition in schizophrenia. These findings suggest that genetic modulation of DRD2-AKT1-related prefrontal-subcortical circuits could, at least in part, influence cognitive dysfunction in psychosis and its treatment. 3) In another study (Marenco et al. Neuropsychopharmacology 2012), the group studied measures of anatomical connectivity between the thalamus and lateral prefrontal cortex (LPFC) in patients with schizophrenia and controls, and assessed their functional implications. Thalamocortical connectivity was measured with diffusion tensor imaging (DTI) and probabilistic tractography The relationship between thalamocortical connectivity and prefrontal cortical blood-oxygenation-level-dependent (BOLD) functional activity as well as behavioral performance during working memory was also assessed. Compared with controls, patients with schizophrenia showed reduced total connectivity of the thalamus to only one of six cortical regions, namely the lateral prefrontal cortical region (LPFC). The size of the thalamic region with at least 25% of model fibers reaching the LPFC was also reduced in patients compared with controls. The total thalamocortical connectivity to the LPFC predicted working memory task performance and also correlated with LPFC BOLD activation. Notably, the correlation with BOLD activation was accentuated in patients as compared with controls in the ventral LPFC. These results suggest that thalamocortical connectivity to the LPFC is altered in schizophrenia with functional consequences on working memory processing in LPFC. The group has also continued to make strides in the field of cognitive aging. Interesting finding have emerged from these studies as well. 1) Using a novel data driven analysis approach with parallel independent component analysis, the group explored the effect of normal aging on the effect of aging on networks common to multiple brain processes, for e.g. working memory and episodic memory cognitive functions that have been shown to decline with age. The results from this study suggested that cognitive aging is associated with alteration in the connectivity within prefrontoparietal networks that is not memory domain specific. These findings are in line with the dedifferentiation hypothesis of neurocognitive aging, and suggest decreased specialization of the brain networks supporting different memory networks with advancing age (Sambataro et al.Eur J Neurosci 2012). 2)In another study, the group explored the neurobiology underlying age-related changes in working memory updating using fMRI with healthy subjects from across the adult age spectrum. The study results indicate that older age is associated with poorer performance, reduced meso-cortico-striatal activation, and reduced functional coupling between the caudate and the VLPFC during working memory updating. Based on prior evidence that normal aging is associated with decline in executive functions along with a progressive decline of neurotransmitter systems including dopamine (DA), these results are consistent with computational models of executive cognition and DA-mediated age-related decline. Given the known relationship between aging and DA system decline, these findings add to the literature on the neurobiology of the central executive and of age-related cognitive decline (Podell et al. Neuroimage 2012.

The Neuroimaging Core has been responsible for optimizing and updating neuroimaging protocols; maintaining, monitoring analyzing and conducting quality control screening of all structural MRI, functional MRI, MR spectroscopy, and DTI data collected by CBDB neuroimaging investigators. In general, the Neuroimaging Core lab continues to cover the following neuroimaging domains: 1. Structural MRI: acquisition, quality control, segmentation, automated and manual regions-of-interest (ROI) definition and measurements, surface extraction; 2. Functional MRI: acquisition, quality control, preprocessing, standardized analyses; 3. Spectroscopy: acquisition, quality control, preprocessing, ROI-based and voxel-based; 4. DTI: acquisition, quality control, preprocessing; and 5. Implementing new MRI techniques To meet these goals, the Neuroimaging Core continues to provide the following services: 1. Medical coverage and image acquisition, image quality control (for data acquired by the core); 2. Maintenance of stimulation equipment, liaison with other NIH imaging core facilities; 3. Implement and maintain acquisition of physiological measurements and interventions, (e.g. galvanic skin response, pupillometry, pulse pressure, in-bore EEG/TMS); 4. Created and update a manual of MR acquisition and data analysis methods; 5. Train and supervise research assistants and fellows in neuroimaging methods and analysis; 6. Perform standardized analyses and analysis quality control; and 7. Databasing and distribution of raw and processed data to CBDB and investigators; integration with genomic and other scientific data maintained within CBDB To provide these services, the core requires the participation of several staff physicians, neuroscientists, computer specialists, and a fairly large group of research assistants (at a minimum 7, divided into function-structural-spectroscopy/DTI groups) who rotate between imaging and image analyses. All these individuals have been trained to maintain a standard level of knowledge in imaging and analysis. In addition, to facilitate the management of the very high throughput of multimodal imaging datasets necessary for large scale neuroimaging genetic studies, members of the group have created a neuroimaging database (XNATGCAP adapted from the early version of XNAT (The Extensible Neuroimaging Archive Toolkit, a BIRN sponsored project). This database allows management of large amounts of data in an efficient and precise manner and is customized to manage and optimize data archiving, data de-identification, data processing, results inspection, data sharing and data mining of neuroimaging data. It facilitates collaborative research across investigators in a secure manner with minimum management overhead.

Over the past year, we focused on several candidate genes for schizophrenia and affective disorders, and studied expression of multiple transcripts of these genes and their associations with schizophrenia risk-associated genotypes. For instance, for a DARPP-32 gene, we examined the association of expression of two major DARPP-32 transcripts, full-length (FL-DARPP-32) and truncated (t-DARPP-32), with genetic variants of DARPP-32 in three brain regions receiving dopaminergic input and implicated in schizophrenia (the dorsolateral prefrontal cortex DLPFC, hippocampus, and caudate) in a much larger set of postmortem samples from patients with schizophrenia, bipolar disorder, major depression and normal controls (>700 subjects). We found that the expression of t-DARPP-32 was increased in the DLPFC of patients with schizophrenia and bipolar disorder and was strongly associated with genotypes at SNPs (rs879606, rs90974 and rs3764352), as well as the previously identified 7-SNP haplotype related to cognitive functioning. The genetic variants that predicted worse cognitive performance were associated with higher t-DARPP-32 expression. Our results suggest that variation in PPP1R1B affects the abundance of the splice variant t-DARPP-32 mRNA and may reflect potential molecular mechanisms implicated in schizophrenia and affective disorders. We also examined copy number variations (CNVs) associated with diverse neurodevelopmental behavioral disorders. We analyzed 1M SNP genotype arrays (Illumina BeadArrays) for evidence of previously reported recurrent CNVs and enriched genome wide CNV burden in DNA from 600 brains, including 441 individuals with various psychiatric diagnoses. We explored gene expression in the dorsolateral prefrontal cortex in selected cases with CNVs and in other subjects using Illumina BeadArrays (568 subjects in total), and additionally in 66-92 subjects using quantitative real-time PCR. CNVs in previously reported genomic regions were identified in 4/193 patients with the diagnosis of schizophrenia (1q21.1, 11q25, 15q11.2, 22q11), 4/238 patients with mood disorders (11q25, 15q11.2, 22q11), and 1/10 patients with autism (2p16.3). No evidence of increased genome wide CNV burden was observed in cases with schizophrenia or mood disorders although the study is underpowered to observe rare events. mRNA expression patterns suggested incomplete molecular penetrance of observed CNVs, particularly in the duplications. Our data confirm in brain DNA the presence of certain recurrent CNVs in a small percentage of patients with psychiatric diagnoses. Finally, we explored epigenetic changes during development of the human prefrontal cortex (PFC), a mastermind of the brain, which is one of the last brain regions to mature. It is also a region implicated in schizophrenia and other major mental disorders. To investigate the role of epigenetics in the development of PFC we examined DNA methylation in 14,500 genes at 27,000 CpG loci focused on 5 promoter regions in 108 subjects ranging from fetal to old age. DNA methylation in the PFC shows unique temporal patterns across life. The fastest changes occur during the prenatal period, slow down markedly after birth and continue to slow further with aging. At the genome level, the transition from fetal to postnatal life is typified by a reversal of direction, from demethylation prenatally to increased methylation postnatally. DNA methylation is strongly associated with genotypic variants and correlates with expression of a subset of genes, including genes involved in brain development and in de novo DNA methylation. Our results indicate that promoter DNA methylation in the human PFC is a highly dynamic process modified by genetic variance and regulating gene transcription. We have made additional discovery by the scientists possible by using a stand-alone application BrainCloudMeth created by our team. We have conducted several investigations using postmortem human brain specimens focused primarily on understanding the pathophysiology of SZ in addition to other complex neuropsychiatric disorders. In addition to our own studies, the Section continues to provide postmortem human brain tissues to researchers and labs within and outside NIH.

From August 1, 2012 - July 11, 2013, we collected 64 brains through the DC and VA Medical Examiners Offices. We made small refinements to the collection this year by discarding and/or transferring cases out a sum of unusable/depleted cases with a history of primary drug abuse or neurological rule-outs, and increased our number of mood disorder cases. Our collection consists of 1,204 brains, including 157 samples from Stanley Foundation and 150 from agreements. The following counts do NOT include Stanley: SZ 183; Bipolar 93; Major Depression Disorder 182; Controls, Adult 266; Other Mood Disorder 31; Affective Disorders 55; Neonate to Teen 82; Fetal Brains 66; Substance Abuse 16; Other Diagnoses 63; and 65 to be characterized. The major focus of our studies is schizophrenia (SZ), which involves cognitive deficits, negative symptoms, and psychosis, usually starting in late adolescence or early adulthood. The recent discovery of genetic variations associated with SZ not only speaks to a significant component of the etiology, but allows for increased understanding of the neuropsychology, neuroimaging, and neuropathology of the disorder. Postmortem human brain studies guided by genetic advances are a valuable, if not essential component, in elucidating the cellular and molecular pathophysiology of SZ. Improved understanding of a genetic component in SZ may enhance diagnostic abilities and hopefully, lead to new treatments. Postmortem human brain studies are limited by the quality and quantity of the specimens. Brain Procurement: The number of brains collected per year has remained relatively constant for the past ten years and we have become very selective in choosing cases. The entire collection has undergone a case-by-case review leading to a reduction in the number of brains for study. Selected cases all have informed consent from the next-of-kin at time of donation, with extensive medical records available for review. Interviews with the next-of-kin, adds clinical information. The tissue is screened for macro- and microscopic neuropathology, and multiple screens are performed for RNA and protein integrity. Toxicological screens are conducted for substances of abuse, and prescribed psychotropic medications. Clinical Review and Characterization: Informed consent for research on the brain and related tissue samples from the next-of-kin is obtained at or before the time of autopsy includes the brain, blood, and hair specimen (for toxicology) as well as permission to obtain medical records. Next-of-kin are interviewed with a 14-item telephone-screening questionnaire. The next-of-kin are advised that they will receive a follow-up interview in greater depth within several months. This in-depth interview relies upon three screening devices: a SCID (First et al., 1997), a modified psychological autopsy (Kelly and Mann, 1996) and a modified version of the DEAD (Diagnostic Evaluation after Death) (Zalcman et al., 1983) to collect relevant clinical information from the next-of-kin in order to enhance our diagnostic classification. The ME offices provide demographic information (age, gender, race, etc.), a police report, interviews with the next-of-kin, and relevant medical personnel and witnesses, toxicology screens, and a general autopsy report that includes the cause and manner of death. The ME offices perform toxicology screens in the vast majority of cases. If toxicology screens are not performed we have contracted a laboratory to complete the screens. An extensive review of clinical records and all other available information is conducted by trained Section personnel and independently reviewed by two board certified psychiatrists. If the latter agree on diagnosis, the case is formally reviewed at a diagnostic conference attended by the two psychiatrists, our neuropathologist, the clinical interview team, and the database specialist where a DSM-IVR (APA, 2000) final diagnosis is recorded. If the two psychiatrists disagree, a third board certified psychiatrist reviews the case independently prior to the diagnostic conference. No final DSM-IVR diagnosis is assigned at the diagnostic conference unless two psychiatrists concur on diagnoses. Similar rigor is used in designating normal control subjects who must have no history of psychiatric or neurological disorders, no history of significant substance abuse or dependence, a negative toxicology screen at autopsy, and a neuropathological screen free from abnormalities. Nicotine use has been carefully reviewed and is not an exclusionary criterion for normal control subjects. The neuropathological screening involves two components, a macro- and microscopic examination and inspection. A board certified neuropathologist conducts the macroscopic examination at time of tissue collection, weighing and inspecting the whole brain, upper cervical spinal cord, pituitary and pineal glands, dura, and intracranial vasculature. The microscopic examination involves sectioning and staining tissue from the frontal, temporal, and occipital poles, the posterior hippocampus, and the cerebellar vermis, looking for evidence of neurodegeneration or other pathological processes. Sections are routinely stained to reveal neurofibrillary tangles and neuritic plaques that are counted to apply Khachaturian criteria to establish the neuropathological diagnosis of AD (Khachaturian, 1985). Sections are taken from any region that shows signs of macroscopic pathology, such as an apparent infarct or abscess, for additional neuropathological analysis. A formal macro- and microscopic neuropathology report is written from the information obtained from these examinations. Screening for Molecular Biology Studies: The following procedures have been performed on every brain prior to study inclusion: a. Measurement of tissue pH from homogenized cerebellar hemispheric tissue. b. Measurement of mRNA expression of a panel of putatively constitutively expressed genes (such as actin and cyclophilin) using in situ hybridization histochemistry on 14 micron cerebellar hemispheric sections. c. Extraction of total RNA by Qiagen and separation of total RNA using agarose gels to qualitatively determine the integrity and intensity of the 28S and 18S ribosomal RNA bands. d. Quantitative analysis of total RNA integrity and 28S to 18S ratios using capillary electrophoresis on a Bioanalyzer 2100 (Agilent, Inc.). e. Acceptable postmortem intervals (PMI) in human brain for each protein studied with Westerns or ELISAs are inferred from rat studies investigating the PMI when protein levels for that protein significantly decline (Halim et al., 2003). f. To determine effects of medications in the human studies of mRNA and protein expression, we are using brain tissue from rats chronically treated with various doses of antipsychotic drugs (haloperidol and clozapine) (Lipska et al., 2001,2003). g. From these screening measurements, we have developed exclusionary criteria for the use of human brain tissue in our experiments: tissue pH < 6.00, PMI > 60 hours; abnormal cerebellar actin and cyclophilin expression levels in cerebellar in situ hybridization (greater than two standard deviations from the mean); and a 28S/18S ribosomal RNA ratio < 1.2. h. The vast majority of our cases (over 90%) come from ME offices, few, if any, of our brains come from patients who have died with prolonged agonal states. The latter is clearly associated with low pH and increased RNA degradation (Tomita et al., 2004).

Progress this past year in understanding how schizophrenia risk genetics operate at the cognitive and neural systems level has been manifold, with important advances in (1) identifying novel phenotypes that distinguish both people with schizophrenia and unaffected family members from unrelated healthy individuals and in (2) discovering associations between specific risk genes and schizophrenia-linked brain phenotypes. Ongoing studies in these two veins permit both better understanding of heritable, trait-related abnormalities in schizophrenia and the underlying molecular biology responsible for such abnormalities, respectively. With regard to novel phenotypes, we have developed and employed a new fMRI paradigm based on the Digit Symbol Substitution Test, a measure of processing speed that is particularly challenging for individuals with schizophrenia and has been forwarded as one of the most sensitive standard neurocognitive measurements distinguishing people with and without schizophrenia. This has allowed us to identify a pattern of under-activated prefrontal cortex that is seen in schizophrenia and in unaffected siblings. This potential endophenotype was replicated in a second sample, paving the way for future genetic association work aimed at understanding how DNA sequence variation translates to functional brain dynamics affected by illness. Similarly, in line with hypotheses of GABAergic disruption in schizophrenia primarily founded on post-mortem brain tissue investigations, we have shown that at least one measure of in vivo levels of GABA, a critical inhibitory neurotransmitter, measured in the dorsal anterior cingulate with MRS is reduced in both individuals with schizophrenia and their unaffected siblings. Moreover, this measure appears to be moderately heritable within families, possibly constituting an intermediate phenotype of modest effect size. This is in contrast to an important negative finding in which we confirmed an absence of MRS measured glutamate levels in the same brain region. Efforts to elaborate on these observations and, in the case of GABA, determine underlying genetic mechanisms are ongoing. In parallel, genetic work aimed at elucidating links between risk genes and risk-associated cognitive and neural signatures has been accelerating. This work offers the opportunity to provide biological validation to putative risk genes, thereby identifying the most promising causative molecular targets a critical step for advancing research on a polygenic, heterogeneous illness such as schizophrenia. Earlier work by our group was able to establish a coherent, latent structure within our extensive cognitive measurements, identifying not only six domain composite factors, but also a higher order factor reflecting general cognitive ability, also called g. Recently, we identified an exciting and novel association between this general cognitive composite and a genetic variant related to sodium channel biology that helps to explain the cognitive impairment in our sample of people with schizophrenia and in their unaffected siblings. We subsequently demonstrated the molecular functionality of this variant through analyses of gene transcript expression in post-mortem brain tissue. Then, by leveraging our fMRI assays of brain function, we were able to successfully tie this same variation in the SCN2A gene to the heritable intermediate phenotype first described by this group prefrontal inefficiency during the Nback task. SCN2A genotype effects have now been confirmed in 531 healthy individuals from 3 different sites and extended, showing with resting state fMRI that those with the cognitively advantaged allele may have greater interregional correspondence between two brain structures discussed above that support higher cognitive function: the dorsolateral prefrontal cortex and dorsal anterior cingulate. We have similarly executed an incisive investigation of NKCC1, a gene that is important in the establishment of GABA as an inhibitory neurotransmitter early in development, showing not only that variation in this gene can modestly increase risk for schizophrenia, but also that it modulates mRNA transcription and confers differential general cognitive function (g) and prefrontal efficiency during working memory, as measured by fMRI. Our collaborative work has been especially productive this year, including a notable discovery that a genetic locus previously associated with educational attainment, strongly and consistently predicted broad cognitive ability. Additionally, new findings are consistent with gene-gene interactions in particular interactions between genes for the dopamine D2 and serotonin 5-HT2A receptors affecting prefrontal function and biasing response to antipsychotic treatment. Innovative approaches to understanding gene-by-environment interactions are also underway, including studies of urban exposure during childhood, an epidemiological risk factor for schizophrenia. For instance, we have now been able to identify robust interactions between COMT genotype and urbanicity impacting prefrontal fMRI signal in two independent cohorts. These data mark a major step forward in characterizing the coalescence of genetic and environmental factors on systems-level brain function, and studies are underway to better define how these factors operate in the context of schizophrenia illness.

The Clinical and Translational Neuroscience Branch continues to make advances on several fronts in order to delineate the neurochemical, neurogenetic, and neuropsychological contributions to neural systems function and development relevant to mental illness. We have embarked on data collection for two unprecedented scientific resources: first, a unique multimodal neuroimaging dataset in adults that includes neuropsychological testing, extensive dopaminergic PET imaging as well as MR spectroscopy, functional and structural MRI; and, second, a longitudinal, neurodevelopmental dataset that incorporates structural and functional magnetic resonance-based brain imaging, and, in conjunction with the Section on Behavioral Endocrinology, precise, state-of-the-art endocrinological measurements of pubertal status. These comprehensive ongoing data acquisition efforts have resulted in a growing repository of integrated information about the brain, which will permit both novel analyses synthesizing disparate but interrelated indices of neurochemical (e.g., dopamine, GABA, and glutamate) functioning and discovery of critical genetic and endocrinological factors guiding neurodevelopment. This work has already begun to yield important insights into social neurocognitive processes, which inform models of abnormal processing of social cues in neuropsychiatric conditions such as schizophrenia, autism and Williams syndrome. In a series of multimodal experiments, we have taken advantage of magnetoencephalography, positron emission tomography and magnetic resonance imaging methods to first identify heretofore unappreciated links between midbrain dopamine stores and how social neural networks react to viewing emotional faces, then dissect how specific types of such brain activity modulate over time as expressions emerge on a viewed face, and most recently, discover genetic association between the Williams syndrome region gene, GTF2I, and relationships between trait neuroticism and social neural activity. Other recent efforts have focused on defining the impact of gene variants associated with schizophrenia and general cognitive functioning and has begun to evaluate new hypotheses about how sequelae of common genetic variation intersects with the biology of mental illness. We have successfully tied the heritable intermediate phenotype first described by this group prefrontal inefficiency during the working memory N-back task to a promising schizophrenia risk variant in the SCN2A gene. Collaborative studies have now confirmed genetic associations between variation in the SCN2A gene and both cognitive ability and interregional brain connectivity in a large cohort of healthy individuals. In parallel, work this year has uncovered links between a genetic locus previously associated with educational attainment and broad cognitive ability. We also remain engaged in studies characterizing the genetic determinants of key neurochemical phenotypes (i.e., MRS- and PET-based signals) that have been linked with neuropsychiatric illness. Finally, ongoing work is aimed at understanding the underlying structure of our neural systems and cognitive-behavioral measurements. For instance, in view of increasingly prominent views that dimensions of psychopathology are extremes of more typical dispositional variation, we have now undertaken analyses of characterological and personality disorder assays of a large cohort of healthy individuals unfettered by confounds inherent in clinical populations. We have preliminarily identified several prominent dimensions Negative Affect, Detached, and Egocentric that complement leading personality theories and may lend insight into latent personality trait structure regardless of diagnostic thresholds.