Schizophrenia Research Forum - A Catalyst for Creative Thinking

Default Mode Network Acts Up in Schizophrenia

27 January 2009. People with schizophrenia show changes in the activity of an interconnected network of brain regions that is involved in memory and self-reflection, according to a study out in this week’s PNAS online edition. The work, from Susan Whitfield-Gabrieli and colleagues at MIT, used functional magnetic resonance imaging (fMRI) to look at the default mode network, a distributed network that is active when the brain is resting and that powers down during focused mental tasks. The results indicate that the network is hyperactive and hyperconnected in people with schizophrenia, and could be contributing to their symptoms. The network is also affected in healthy, first-degree relatives of subjects, suggesting that the changes could be part of the genetic risk for the disease.

The results fit with the idea that there are widespread deficiencies in cortical processing in schizophrenia, and suggest that these deficiencies affect a brain network thought to be responsible for self-awareness and internal reference.

In other network news, three papers in the January 19 edition of Nature online show progress toward the elusive goal of mapping neuronal networks in the cortex at the finer level of single cells and their synaptic partners. The results offer techniques that will be useful to probe the alterations in micro-scale cortical organization that underlie the functional changes seen in schizophrenia.

Cognitive connections
The concept of the default mode network, a distributed neural network that is active when the brain is resting and that powers down during focused mental tasks, was only first published in 2001, but it has become a hot topic in cognitive neuroscience. The network, which includes the medial prefrontal cortex, the posterior cingular cortex/precuneous, and the lateral parietal cortex, activates during daydreaming, self-referential thought, and during some kinds of memory retrieval. The seesaw deactivation of the default network and activation of task-related brain regions, which may reflect the dynamic allocation of cognitive resources between internal and external demands, is critical for peak performance on memory tasks.

Previous studies have shown changes in the default mode function in people with schizophrenia (Garrity et al., 2007; Zhou et al., 2007; Pomarol-Clotet et al., 2008). The new work adds to those studies by looking at activity in relatives and by correlating changes in the network with schizophrenia pathology.

In the study, Whitfield-Gabrieli and coworkers performed fMRI scans of subjects while they were idle, and then when they engaged in a simple working memory test. The data allowed them to assess both resting network activity, and the extent of deactivation that occurred during a task that required concentration. The study compared 13 volunteers with early-phase schizophrenia, 13 unaffected first-degree relatives, and 13 healthy controls. When subjects performed the recall test, activation occurred in the dorsolateral prefrontal cortex. As expected, the activation was higher in patients and relatives compared to controls. At the same time, deactivation of the default mode network was decreased in patients and relatives compared to controls. Overall, the deactivation in the default network and activation of task-related areas strongly correlated in control subjects, but the seesaw effect was much weaker in patients and relatives. The result was a consistent hyperactivity of the default network, which correlated with worse performance in the memory task.

Altered brain connectivity of default brain network in persons with schizophrenia and first-degree relatives. Colored areas represent an interconnected network of brain regions that show synchronized activity (overlapping black and blue traces) when subjects rest and allow their minds to wander. The amount of synchrony, which reflects the strength of functional connections between the different areas, is increased in patients with schizophrenia. First-degree relatives of persons with the illness also show some increase, although less than patients. Black circle: medial prefrontal cortex. Blue circle: posterior cingulate/precuneous. Image credit: Susan Whitfield-Gabrieli, McGovern Institute for Brain Research at MIT

Measurement of the default network connectivity, or the extent to which regions activate together, showed a stronger connectively between the medial prefrontal cortex and the precuneous and the rest of the default network in patients and relatives. This was seen whether connectivity was measured at rest or during the task. Like the deactivation defect, higher connectively also correlated with worse working memory performance.

The strength of connectivity and defect in deactivation correlated with stronger schizophrenia symptoms, suggesting that the default mode network could play an important role the cognitive and clinical symptoms of schizophrenia. Defects in deactivation could explain problems with working memory and attention. In addition, the authors point out that the default mode network is normally activated during internal, self-referential thought. “Hyperactivity of the default network may blur the normal boundary between internal thoughts and external perceptions,” this write. “Indeed, many symptoms of schizophrenia involve an exaggerated sense of self-relevance in the world, such as paranoid ideation that individuals and groups are conspiring against the patient, and a blurring of internal reflection and external perception, such as hallucinations.”

The finding that unaffected relatives show changes in the default network suggests that the activity stems from genetic risk and is causal, rather than just a consequence of the disease, Whitfield-Gabrieli told SRF. “In the future, it may be possible to use these fMRI measures as a way of diagnosing disease, or to figure out how patients are responding to treatment,” she said.

Other diseases where default mode network activity is known to be altered include autism, epilepsy, depression, attention deficit/hyperactivity disorder, and Alzheimer disease (for a comprehensive review of the literature of default mode activity and disease, see Broyd et al., 2008).

Connect the neurons
Networks like the default mode involve widely distributed areas of the brain, but no matter how far-flung their components may be, the basic unit of any network comes down to neuron-to-neuron communication at individual synapses. A detailed understanding of brain circuits, and how they are altered in schizophrenia, will require a map of the wiring between individual neurons. However, it has been nearly impossible to tease out exact contacts in the spaghetti-like tangle of axons, dendrites, and cell bodies that make up brain tissue. In the January 18 online edition of Nature, three different groups report their efforts to unravel this knotty problem by combining single cell recording with sophisticated imaging and other techniques.

First, Solange Brown and Shaul Hestrin of Stanford University looked at the local connections among three classes of cortical neurons that project over long distances to the contralateral cortex, the contralateral striatum, or the superior colliculus. After retrograde labeling of projections in live mice to distinguish the different kinds of neurons, the researchers simultaneously recorded activity in sets of four cells in cortical slices. By stimulating each cell in turn and watching the others’ responses, Brown and Hestrin showed they could identify local cortical connections with high success. The tendency of cells to make local connections was not random, or simply based on proximity, but depended on the identities of both the presynaptic and postsynaptic cells. For example, a neuron that projected to the contralateral cortex (a corticocortical neuron) was four times more likely to make a synapse with a local corticotectal neuron than with another corticocortical neuron. The results suggest a way to unravel the local circuit architecture in the cortex.

In a second paper, researchers from Matthew Larkum’s lab at the University of Bern in Switzerland used fiber optic imaging to measure dendritic calcium changes and uncover a cortical inhibitory microcircuit in living rats. In the circuit, they found, inhibitory interneurons govern the graded calcium response in L5 pyramidal neuron dendrites after sensory stimuli. The results help explain how neurons can respond to sensory input over a large dynamic range by utilizing cortical micronetworks.

Finally, Karel Svoboda and colleagues of the Howard Hughes Medical Institute in Ashburn, Virginia, traced out an excitatory circuit using photoactivation of neurons expressing channelrhodopsin (Wang et al., 2007) to map out points of contact. Taking cortical slices with channelrhodopsin expressed in axons, first author Leopoldo Petreanu and colleagues used a laser beam to stimulate local neurotransmitter release while simultaneously recording from barrel cortex pyramidal neurons. If the axons made a synapse on the recorded cell, a postsynaptic excitatory current would be triggered. By directing channelrhodopsin expression to different layers of the cortex whose axons overlap with the pyramidal cell dendrites of the barrel cortex, the investigators were able to map input from the thalamus, cortical layers 4, 2/3, 4, L2/3, and part of the motor cortex to the dendrites of L3 and L5 pyramidal neurons. They found many of the known connections, as well as some new ones. In addition, the mapping revealed a high degree of spatial specificity, with different inputs mapping onto more distal or apical locations on the dendrites. By allowing the high-resolution look at neuronal circuits, the technique should help achieve a micro-scale, comprehensive picture of the hardware of the brain.—Pat McCaffrey.

Whitfield-Gabrieli S, Thermenos HW, Milanovic S, Tsuang MT, Faraone SV, McCarley RW, Shenton ME, Green AI, Nieto-Castanon A, LaViolette P, Wojcik J, Gabrieli JDE, Seidman LJ. Hyperactivity and hyperconnectivity of the default network in schizophrenia and in first-degree relatives of persons with schizophrenia. PNAS Early Edition. 2009 Jan 19. Abstract

Murayama M, Pérez-Garci E, Nevian T, Bock T, Senn W, Larkum ME. Dendritic encoding of sensory stimuli controlled by deep cortical interneurons. Nature. 2009 Jan 18. Abstract

Brown SP, Hestrin S. Intracortical circuits of pyramidal neurons reflect their long-range axonal targets. Nature. 2009 Jan 18. Abstract

Petreanu L, Mao T, Sternson SM, Svoboda K. The subcellular organization of neocortical excitatory connections. Nature. 2009 Jan 18. Abstract

Comments on News and Primary Papers
Comment by:  Vince Calhoun
Submitted 27 January 2009
Posted 27 January 2009

In this work the authors test for differences in the default mode network between healthy controls, patients with schizophrenia, and first degree relatives of the patients. They look at both the degree to which the default mode is modulated by a working memory task and also examine the strength of the functional connectivity. The controls are found to show the most default mode signal decrease during a task, with relatives and patients showing much less. The controls, relatives, and patients show increasing amounts of functional connectivity within the default mode regions. In addition, signal in some of the regions correlated with positive symptoms. The findings in the chronic patients and controls are consistent with our previous work in Garrity et al., 2007, which also showed significantly more functional connectivity in the default mode of schizophrenia patients and significant correlations in certain regions of the default mode with positive symptoms, and in both cases the regions we identified are similar to those shown in the Whitfield-Gabrieli paper. Our work in Kim et al., 2009, was a large multisite study showing significantly fewer default mode signal decreases for the auditory oddball task in chronic schizophrenia patients, again consistent with the Whitfield-Gabrieli paper, but in a different task.

The most interesting contribution of the Whitfield-Gabrieli paper is their inclusion of a first-degree relative group. They found that the first-degree relatives are “in between” the healthy controls and the chronic patients in terms of both the degree to which they modulate the default mode, as well as in their degree of functional connectivity. This has interesting implications in terms of the genetic aspects of the illness and suggests that the default mode may be a potential schizophrenia endophenotype. It will be interesting in future studies to examine both the heritability of the default mode patterns and their genetic underpinnings.

View all comments by Vince CalhounComment by:  Edith Pomarol-Clotet
Submitted 28 January 2009
Posted 28 January 2009

The Default Mode Network and Schizophrenia
For a long time functional imaging research has focused on brain activations. However, since 2001 it has been appreciated that there is also a network of brain regions—which includes particularly two midline regions, the medial prefrontal cortex and the posterior cingulate cortex/precuneous—which deactivates during performance of a wide range of cognitive tasks. Why some brain regions should be active at rest but deactivate when tasks have to be performed is unclear, but there is intense speculation that this network is involved in functions such as self-reflection, self-monitoring, and the maintenance of one’s sense of self.

Could the default mode network be implicated in neuropsychiatric disease states? There is evidence that this is the case in autism, and a handful of studies have been also carried out in schizophrenia. Now, Whitfield-Gabrieli and colleagues report that 13 schizophrenic patients in the early phase of illness showed a failure to deactivate the anterior medial prefrontal node of the default mode network when they performed a working memory task. They also find that failure to deactivate is seen to a lesser but still significant extent in unaffected first-degree relatives of the schizophrenic patients, and that the degree of failure to deactivate is associated with both the severity of positive and negative symptoms in the patients.

Importantly, the findings of Whitfield-Gabrieli and colleagues are closely similar to those of another recent study by our group (Pomarol-Clotet et al., 2008), which found failure to deactivate in the medial prefrontal cortex node of the default mode network in 32 chronic schizophrenic patients. This is a striking convergence in the field of functional imaging studies of schizophrenia, which has previously been marked by diverse and often conflicting findings. Additionally, in both studies the magnitude of the difference between patients and controls was large and visually striking. These findings suggest that we may be dealing with an important abnormality which could be close to the disease process in schizophrenia.

If so, what does dysfunction in the default mode network mean? On the one hand, failure to deactivate part of a network whose activity normally decreases when attention has to be turned to performance of external tasks might be expected to interfere with normal cognitive operations. Consistent with this, cognitive impairment is nowadays accepted as being an important, or even a “core” feature of schizophrenia. Perhaps more importantly, could it be that default mode network dysfunction can help us understand the symptoms of schizophrenia? As Whitfield-Gabrieli and colleagues note, if the default mode network is involved in self-reflection, self-monitoring, and maintenance of one’s sense of self, then failure of deactivation might lead to an exaggerated focus on one’s own thoughts and feelings, excessive self-reference, and/or a breakdown in the boundary between the inner self and the external world. The default mode network may thus have the potential to account for two major realms of clinical abnormality in schizophrenia—its symptoms and the cognitive impairment that is frequently associated with them.

View all comments by Edith Pomarol-ClotetComment by:  Samantha BroydEdmund Sonuga-Barke
Submitted 4 February 2009
Posted 4 February 2009

The surge in scientific interest in patterns of connectivity and activation of resting-state brain function and the default-mode network has recently extended to default-mode brain dysfunction in mental disorders (for a review, please see Broyd et al., 2008). Whitfield-Gabrieli et al. examine resting-state and (working-memory) task-related brain activity in 13 patients with early-phase schizophrenia, 13 unaffected first-degree relatives, and 13 healthy control participants. These authors report hyperconnectivity in the default-mode network in patients and relatives during rest, and note that this enhanced connectivity was correlated with psychopathology. Further, patients and relatives exhibited reduced task-related suppression (hyperactivation) of the medial prefrontal region of the default-mode network relative to the control group, even after controlling for task performance.

The findings from the Whitfield-Gabrieli paper are in accordance with those from a number of other research groups investigating possible default-mode network dysfunction in schizophrenia. For example, in a similar working memory task Pomarol-Clotet and colleagues (2008) have also shown reduced task-related suppression of medial frontal nodes of the default-mode network in 32 patients with chronic schizophrenia. However, the findings are at odds with research reporting widespread reductions in functional connectivity in the resting brain of this clinical group (e.g., Bluhm et al., 2007; Liang et al., 2006). As noted by Whitfield-Gabrieli et al., increased connectivity and reduced task-related suppression of default-mode activity may redirect attentional focus from task-related events to introspective and self-referential thought processes. The reduced anti-correlation between the task-positive and default-mode network in patients further supports and helps biologically ground suggestions of the possibility of an overzealous focus on internal thought. Perhaps even more interestingly, the study by Whitfield-Gabrieli and colleagues suggests that aberrant patterns of activation and connectivity in the default-mode network, and in particular the medial frontal region of this network, may be associated with genetic risk for schizophrenia. Although there are some inconsistencies in the literature regarding the role of the default-mode network in schizophrenia, the work of Whitfield-Gabrieli and others suggests that this network may well contribute to the pathophysiology of this disorder and is relevant to contemporary models of schizophrenia. Indeed, the recent flurry in empirical research investigating the clinical relevance of this network to mental disorder has highlighted a number of possible putative mechanisms that might link the default-mode network to disorder. Firstly, effective transitioning from the resting-state to task-related activity appears to be particularly vulnerable to dysfunction in mental disorders and may be characterized by deficits in attentional control. Sonuga-Barke and Castellanos (2007) have suggested that interference arising from a reduction in the task-related deactivation of the default-mode network may underlie the disruption of attentional control. The default-mode interference hypothesis proposes that spontaneous low-frequency activity in the default-mode network, normally attenuated during goal-directed tasks, can intrude on task-specific activity and create cyclical lapses in attention resulting in increased variability and a decline in task performance (Sonuga-Barke and Castellanos, 2007). Sonuga-Barke and Castellanos (2007) suggest that the efficacious transition from rest to task and the maintenance of task-specific activity may be moderated by trait factors such as disorder. Secondly, the degree of functional connectivity in the default-mode network may highlight problems of reduced connectivity, or excess functional connectivity (e.g., schizophrenia), which suggests a zealous focus on self-referential processing and introspective thought. Thirdly, the strength of the anti-correlation between the default-mode and task-positive networks may also indicate a clinical susceptibility to introspective or extrospective orienting. Finally, future research should continue to examine the etiology of the default-mode network in schizophrenia.


Bluhm, R.L., Miller, J., Lanius, R.A., Osuch, E.A., Boksman, K., Neufeld, R.W.J., Théberge, J., Schaefer, B., & Williamson, P. (2007). Spontaneous low-frequency fluctuations in the BOLD signal in schizophrenic patients: Anomalies in the default network. Schizophrenia Bulletin, 33, 1004-1012. Abstract

Broyd, S.J., Demanuele, D., Debener, S., Helps, S.K., James, C.J., & Sonuga-Barke, E.J.S. (in press). Default-mode brain dysfunction in mental disorders: a systematic review. Neurosci Biobehav Rev. 2008 Sep 9. Abstract

Liang, M., Zhou, Y., Jiang, T., Liu, Z., Tian, L., Liu, H., and Hao, Y. (2006). Widespread functional disconnectivity in schizophrenia with resting-state functional magnetic resonance imaging. NeuroReport, 17, 209-213. Abstract

Pomarol-Clotet, E., Salvador, R., Sarro, S., Gomar, J., Vila, F., Martinez, A., Guerrero, A.,Ortiz-Gil, J., Sans-Sansa, B., Capdevila, A., Cebemanos, J.M., McKenna, P.J., 2008. Failure to deactivate in the prefrontal cortex in schizophrenia: dysfunction of the default-mode network? Psychological Medicine, 38, 1185–1193. Abstract

Sonuga-Barke, E.J.S., Castellanos, F.X., 2007. Spontaneous attentional fluctuations in impaired states and pathological conditions: a neurobiological hypothesis. Neuroscience Biobehavioural Reviews, 31, 977–986. Abstract

View all comments by Samantha Broyd
View all comments by Edmund Sonuga-BarkeComment by:  Yuan ZhouTianzi JiangZhening Liu
Submitted 18 February 2009
Posted 22 February 2009
  I recommend the Primary Papers

The consistent findings on default-mode network in human brain have attracted the researcher’s attention to the task-independent activity. The component regions of the default-mode network, especially medial prefrontal cortex and posterior cingulate cortex/precuneus, are related to self-reflective activities and attention. Both of these functions are observed to be impaired in schizophrenia. And thus the default-mode network has also attracted more and more attention in the schizophrenia research community. The study of Whitfield-Gabrieli et al. shows a further step along this research streamline.

The authors found hyperactivity (reduced task suppression) and hyperconnectivity of the default network in schizophrenia, and found that hyperactivity and hyperconnectivity of the default network are associated with poor work memory performance and greater psychopathology in schizophrenia. And they found less anticorrelation between the medial prefrontal cortex and the right dorsolateral prefrontal cortex, a region showing increased task-related activity in schizophrenia, whether during rest or task. Furthermore, the hyperactivity in medial prefrontal cortex is negatively related to the hyperconnectivity of the default network in schizophrenia.

There are two main contributions in this work. First, they found significant correlation between the abnormalities in the default mode network and impaired cognitive performance and psychopathology in schizophrenia. Thus they propose a new explanation for the impaired working memory and attention in schizophrenia, and propose a possibility that schizophrenic symptoms, such as delusions and hallucinations, may be due to the blurred boundary between internal thoughts and external perceptions. Secondly, they recruited the first-degree relatives of these patients in this study, and found that these healthy relatives showed abnormalities in the default network similar to that of patients but to a lesser extent. This is the first study investigating the default mode network of relatives of individuals with schizophrenia. This finding indicates that the dysfunction in the default mode network is associated with genetic risk for schizophrenia.

The findings in schizophrenia are consistent with our previous work (Zhou et al., 2007), in which we also found hyperconnectivity of the default mode network during rest. Considering the differences in ethnicity of participants (Chinese in our study) and methodology, the consistency in the hyperconnectivity of the default mode network in schizophrenia is exciting, which supports the possibility that abnormality in the default-mode network may be a potential imaging biomarker to assist diagnosis of schizophrenia. However, this needs to be validated in future studies with a large sample size, due to other contradictory findings, for example, the reduced resting-state functional connectivities associated with the posterior cingulate cortex in chronic, medicated schizophrenic patients (Bluhm et al., 2007). In addition, further studies should focus on default-mode function in different clinical subtypes, as schizophrenia is a complicated disorder. Finally, it should be noticed that the hyperconnectivity of the default-mode network is not exclusively contradictory with hyperconnectivity in other regions, as we previously found (Liang et al., 2006). It is possible that hyperconnectivity and hyperconnectivity coexist in the brains of individuals with schizophrenia and together lead to the complicated symptoms and cognitive deficits.


Bluhm, R. L., Miller, J., Lanius, R. A., Osuch, E. A., Boksman, K., Neufeld, R. W., et al., 2007. Spontaneous low-frequency fluctuations in the BOLD signal in schizophrenic patients: anomalies in the default network. Schizophr Bull 33, 1004-1012. Abstract

Liang, M., Zhou, Y., Jiang, T., Liu, Z., Tian, L., Liu, H., et al., 2006. Widespread functional disconnectivity in schizophrenia with resting-state functional magnetic resonance imaging. Neuroreport 17, 209-213. Abstract

Whitfield-Gabrieli, S., Thermenos, H. W., Milanovic, S., Tsuang, M. T., Faraone, S. V., McCarley, R. W., et al., 2009. Hyperactivity and hyperconnectivity of the default network in schizophrenia and in first-degree relatives of persons with schizophrenia. Proc Natl Acad Sci U S A 106, 1279-1284. Abstract

Zhou, Y., Liang, M., Tian, L., Wang, K., Hao, Y., Liu, H., et al., 2007. Functional disintegration in paranoid schizophrenia using resting-state fMRI. Schizophr Res 97, 194-205. Abstract

View all comments by Yuan Zhou
View all comments by Tianzi Jiang
View all comments by Zhening Liu

Comments on Related News

Related News: Schizophrenia-associated Variant in ZNF804A Gene Affects Brain Connectivity

Comment by:  James WaltersMichael Owen (SRF Advisor)
Submitted 3 June 2009
Posted 3 June 2009

Andreas Meyer-Lindenberg’s group examine the association between a single nucleotide polymorphism (SNP), rs1344706 in gene ZNF804A, recently identified as a risk factor for schizophrenia in a genome-wide association study (GWAS) (O'Donovan et al., 2008) and functional connectivity as measured by fMRI. The attraction of this polymorphism for a study of this kind is twofold. First, statistically speaking it is the most robust SNP association with schizophrenia reported to date. Second, because a single variant shows strong evidence for association, which is not the case for other reported associations, it is possible to specify a priori for the gene in question directional hypotheses in relation to potential neurocognitive correlates. This militates against the generation of false positives through the testing of multiple SNPs and haplotypes which has rendered problematic the interpretation of at least some previous genetic imaging studies (Walters and Owen, 2007). The function of ZNF804A is unknown but the fact that it contains a zinc finger domain suggests that it may be a transcription factor. It is hoped that the characterization of the actions of SNPs identified by GWAS will identify new pathogenic mechanisms of psychosis. One way in which this can be achieved is via approaches such as that taken in this article.

Esslinger et al. report variations in functional connectivity in 115 healthy individuals according to rs1344706 risk variant status. Given the association of ZNF804A with both schizophrenia and bipolar disorder they employed two fMRI tasks thought to be sensitive to altered function in these disorders: the N Back (2back) task is sensitive to deficits of dorsolateral prefrontal cortex (DLPFC) function in schizophrenia and an emotional face-matching task is linked with amygdala function and thought to be relevant to mood disorder. They compared the activity in these regions and functional connectivity (using time-series correlation) between the three rs1344706 genotype groups.

No differences between genotype groups were found for activation, but the authors did identify altered connectivity with the most activated DLPFC locale. Risk-allele carriers were shown to exhibit a lack of uncoupling of activity (increased functional connectivity) between the right DLPFC and left hippocampus during the 2-back task as well as decreased connectivity within right DLPFC and between right and left DLPFC. Risk variant carriers also showed wide ranging increased connectivity between right amygdala and other anatomical regions. The majority of these findings showed a risk allele dose effect.

The increased DLPFC/hippocampus functional connectivity in carriers of the risk allele is potentially the most interesting finding given that Meyer-Lindenberg’s group has previously shown that those with schizophrenia show increased functional connectivity between DLPFC and hippocampus during working memory (Meyer-Lindenberg et al., 2005). Notes of caution in this regard are that 1) the biological, anatomical or functional significance of fMRI determined functional connectivity is yet to be established and 2) other functional connectivity studies in schizophrenia have produced conflicting results Lawrie et al., 2002. Nonetheless, it is interesting that rs1344706 may affect co-ordination of activity between these two brain regions given their seeming importance in psychotic conditions. The significance of these findings to cognitive deficits and other symptom domains needs further investigation particularly as others have postulated dysconnectivity has more relevance to first rank psychotic symptoms (Stephan et al., 2009).

It is likely that genome-wide association approaches will continue to identify genes with unknown neural function and so approaches such as this are likely to be a valuable way of identifying the biological/neural pathways that involve these genes. It is also imperative that as in this study methodology is employed to allow for multiple testing and also that negative findings are reported. We would also suggest caution until these findings are replicated. As well as such approaches in humans, it is also important to investigate the effects of identified variants at other levels of analysis from gene expression to behavioural genetics work. Finally we find it reassuring that GWAS approaches seem to be successful in identifying risk variants whose functions can be investigated using methods such as that taken by Esslinger et al.


O'Donovan MC, Craddock N, Norton N, et al. Identification of loci associated with schizophrenia by genome-wide association and follow-up. Nature Genetics. 2008;40(9):1053-1055. Abstract

Walters JT, Owen MJ. Endophenotypes in psychiatric genetics. Mol Psychiatry. 2007;12(10):886-890. Abstract

Meyer-Lindenberg AS, Olsen RK, Kohn PD, et al. Regionally Specific Disturbance of Dorsolateral Prefrontal-Hippocampal Functional Connectivity in Schizophrenia. Archives of General Psychiatry. 2005;62(4):379-386. Abstract

Lawrie SM, Buechel C, Whalley HC, Frith CD, Friston KJ, Johnstone EC. Reduced frontotemporal functional connectivity in schizophrenia associated with auditory hallucinations. Biological Psychiatry. 2002;51(12):1008-1011. Abstract

Stephan KE, Friston KJ, Frith CD. Dysconnection in Schizophrenia: From Abnormal Synaptic Plasticity to Failures of Self-monitoring. Schizophr Bull. 2009;35(3):509-527. Abstract

View all comments by James Walters
View all comments by Michael Owen

Related News: Ketamine Elicits Brain State Resembling Early Stages of Schizophrenia

Comment by:  Hugo Geerts
Submitted 26 August 2014
Posted 26 August 2014

This is a very interesting contribution to improve the understanding of the progressive nature of schizophrenia pathology. Ketamine-induced effects have been used for a long time in healthy volunteers or in animal models, both rodents and non-human primates to "mimic" schizophrenia pathology. The observation that ketamine mimics more the very early schizophrenia or the at-risk state but not the more chronic pathology helps to explain the dissociation between effects of compounds in ketamine-induced deficits and in chronic schizophrenia, for example, nicotine (D'Souza et al., 2012), glycine transport inhibitor (D'Souza et al., 2012), haloperidol (Oranje et al., 2009), and lamotrigine (Goff et al., 2007).

The impact of these findings, if reproduced in a longitudinal study, is very important. This model of ketamine-induced deficit can be used in animals to test new experimental interventions in very early psychosis or in at-risk subjects. This is a patient group that is currently underserved in terms of therapeutic interventions and for which there is great interest. For the first time, we now have a model that mimics important aspects of the early schizophrenia pathology, and the hope is that when addressing these changes with the right medication early on, one could postpone or delay the onset of overt schizophrenia pathology, which could be the beginning of a preventive strategy.


D'Souza DC, Ahn K, Bhakta S, Elander J, Singh N, Nadim H, Jatlow P, Suckow RF, Pittman B, Ranganathan M. Nicotine fails to attenuate ketamine-induced cognitive deficits and negative and positive symptoms in humans: implications for schizophrenia. Biol Psychiatry . 2012 Nov 1 ; 72(9):785-94. Abstract

D'Souza DC, Singh N, Elander J, Carbuto M, Pittman B, Udo De Haes J, Sjogren M, Peeters P, Ranganathan M, Schipper J. Glycine transporter inhibitor attenuates the psychotomimetic effects of ketamine in healthy males: preliminary evidence. Neuropsychopharmacology . 2012 Mar ; 37(4):1036-46. Abstract

Oranje B, Gispen-de Wied CC, Westenberg HG, Kemner C, Verbaten MN, Kahn RS. Haloperidol counteracts the ketamine-induced disruption of processing negativity, but not that of the P300 amplitude. Int J Neuropsychopharmacol . 2009 Jul ; 12(6):823-32. Abstract

Goff DC, Keefe R, Citrome L, Davy K, Krystal JH, Large C, Thompson TR, Volavka J, Webster EL. Lamotrigine as add-on therapy in schizophrenia: results of 2 placebo-controlled trials. J Clin Psychopharmacol . 2007 Dec ; 27(6):582-9. Abstract

View all comments by Hugo Geerts

Related News: Ketamine Elicits Brain State Resembling Early Stages of Schizophrenia

Comment by:  Alexandre Seillier
Submitted 9 September 2014
Posted 11 September 2014

The "Ketamine Model" Revived?
Despite the seminal review by Jentsch and Roth arguing that long-term, rather than acute administration of NMDA antagonists, such as phencyclidine (PCP) and ketamine, may more isomorphically model schizophrenia (Jentsch and Roth, 1999), acute ketamine remained one of the most widely utilized pharmacological models in both humans and rodents. Recently, Dawson and co-workers added to the accumulating body of evidence that "alterations in brain circuitry that result from chronic, but not from acute, NMDA receptor blockade most accurately reflect the systems level differences in brain network functioning seen in schizophrenia" (Dawson et al., 2014a; Dawson et al., 2014b). Indeed, using brain network connectivity, they reported that, "at a systems level, the mechanisms through which acute ketamine treatment induces schizophrenia-like symptoms may be profoundly divergent from those that contribute to these symptoms in the disorder."

This new study by Anticevic et al. (2014) might have resolved the paradox of acute ketamine-induced hyperfrontality given the hypofrontality observed in schizophrenia. As previously shown in both healthy humans and mice (Dawson et al., 2014a; Driesen et al., 2013), acute ketamine administration was associated with increased prefrontal cortex connectivity. On the other hand, whereas chronic schizophrenics had reduced prefrontal cortex functional connectivity, patients in the early stage of schizophrenia showed increased prefrontal cortex functional connectivity. As stated by the authors, "these data point to a qualitative difference between NMDAR antagonist and chronic schizophrenia effects, suggesting that ketamine's effect on prefrontal cortex connectivity may be more relevant to particular illness stages."

Longitudinal studies will be necessary to confirm that the "phase of illness is an important moderator of the prefrontal cortex functional connectivity in schizophrenia." The authors also addressed some discrepancies that need to be highlighted. First, the reduced and the increased prefrontal cortex functional connectivity in chronic versus early-stage schizophrenia, respectively, were observed in distinct prefrontal areas, namely the right middle frontal gyrus and the left superior frontal gyrus, respectively. Second, the cross-validation of the pharmacological and clinical analysis failed to reach significance; i.e., acute ketamine did not significantly change connectivity in the regions identified in the clinical study. Although it has limitations, this study should inform the application of acute ketamine as a translational model that may better approximate some early stage of schizophrenia.


Dawson N, McDonald M, Higham DJ, Morris BJ, Pratt JA (2014a) Subanesthetic Ketamine Treatment Promotes Abnormal Interactions between Neural Subsystems and Alters the Properties of Functional Brain Networks. Neuropsychopharmacology 39:1786-98. Abstract

Dawson N, Xiao X, McDonald M, Higham DJ, Morris BJ, Pratt JA (2014b) Sustained NMDA receptor hypofunction induces compromised neural systems integration and schizophrenia-like alterations in functional brain networks. Cereb Cortex 24: 452-464. Abstract

Driesen NR, McCarthy G, Bhagwagar Z, Bloch M, Calhoun V, D'Souza DC et al (2013) Relationship of resting brain hyperconnectivity and schizophrenia-like symptoms produced by the NMDA receptor antagonist ketamine. Mol Psychiatry 18:1199-1204. Abstract

Jentsch JD, Roth RH (1999) The neuropsychopharmacology of phencyclidine: from NMDA receptor hypofunction to the dopamine hypothesis of schizophrenia. Neuropsychopharmacology 20: 201-225. Abstract

View all comments by Alexandre Seillier

Related News: Ketamine Elicits Brain State Resembling Early Stages of Schizophrenia

Comment by:  Albert Adell
Submitted 10 September 2014
Posted 16 September 2014

The paper by Anticevic and co-workers, as well as the commentary by Hugo Geerts, points to a very interesting issue on the similitude of the ketamine model with schizophrenia. In a previous mini-review, we anticipated that it would be conceivable that acute administration of NMDA receptor antagonists would lead to a reversible malfunctioning of PV-containing interneurons, whereas in schizophrenia, a damage of these neurons may take place at early stages of neurodevelopment (Adell et al., 2012). In summary, acute NMDA antagonism would resemble the early stage of schizophrenia, whereas chronic exposure better models the more chronic pathology of the illness.


Adell A, Jiminez-Sanchez L, Lopez-Gil X, Roman T. Is the Acute NMDA Receptor Hypofunction a Valid Model of Schizophrenia? Schizophr Bull 38(1): 9-14. Abstract

View all comments by Albert Adell

Related News: Ketamine Elicits Brain State Resembling Early Stages of Schizophrenia

Comment by:  Didier Pinault
Submitted 25 September 2014
Posted 26 September 2014

Acute Ketamine Dramatically Amplifies Network Gamma (30-80 Hz) and Higher-Frequency Oscillations
N-methyl D-aspartate receptors play a key role in synaptic plasticity, memory processes, and in the modulation of field oscillations. The non-competitive NMDAR antagonist ketamine has context-dependent and dose-dependent multiple properties, including positive and negative effects. For instance, a single sub-anesthetic administration can disturb cognitive and perceptual processes and induce schizophreniform psychosis in healthy subjects (Krystal et al., 1994; Adler et al., 1998; Newcomer et al., 1999; Hetem et al., 2000); puzzlingly, but of importance, ketamine can generate a durable antidepressant effect in patients refractory to conventional antidepressant therapies (Zarate et al., 2006; Fond et al., 2014; McGirr et al., 2014).

More specifically, Anticevic and colleagues (2014), using brain scans, recently revealed that a single sub-anesthetic administration of ketamine in healthy subjects at rest produces in the prefrontal cortex a state of hyperconnectivity, which resembles that recorded in people in the early stages of schizophrenia but not in patients with chronic (several years' duration) schizophrenia. Also, using fMRI in healthy human subjects, Driesen and colleagues (Driesen et al., 2013) demonstrated that the NMDA receptor antagonist ketamine increases global brain functional connectivity and reduces negative symptoms, suggesting that the ketamine-induced increase in connectivity corresponds to a state of enhanced cortical function. The acute ketamine effects appear quickly (~1-2 minutes) and are transient. In the following, we will see that these fresh clinical findings are consistent with preclinical studies demonstrating that, in rodents, non-competitive NMDAR antagonists increase the amount of field (or network) gamma frequency oscillations (GFOs; 30-80 Hz) in cortical and subcortical regions. In healthy human subjects, ketamine increases the power of GFOs and decreases that of delta oscillations during auditory-evoked network oscillations (Hong et al., 2010).

What does "hyperconnectivity" mean? From clinical imaging studies, functional connectivity is assessed by the degree of correlation between pairwise functional connections from multiple functional cortical and subcortical regions. What does it mean in terms of neural excitatory and inhibitory activities, network oscillations, and cellular firing?

In rodents a single sub-anesthetic administration of ketamine quickly and transiently induces abnormal behavior (hyperlocomotion, ataxia), memory deficits, and abnormally persistent and generalized hypersynchronized (200-400 percent increased power) ongoing GFOs (Ma and Leung, 2007; Chrobak et al., 2008; Pinault, 2008; Hakami et al., 2009; Ehrlichman et al., 2009; Kocsis, 2012). The gamma frequency at maximal power is significantly increased by approximately 10 Hz on average (Pinault, 2008). The amount of ongoing higher-frequency (>80 Hz) oscillations is also increased following a single sub-anesthetic administration (<10 mg/kg) of ketamine (Hakami et al., 2009; Hunt et al., 2006; Kulikova et al., 2012).

In addition, NMDAR antagonists transiently disrupt the expression, not the induction, of long-term potentiation in the thalamocortical system (Kulikova et al., 2012), disorganize action potential firing in rat prefrontal cortex (Molina et al., 2014), increase the firing in fast-spiking neurons, and decrease that in regularly spiking neurons (Homayoun and Moghaddam, 2007). These results suggest that the amount of ongoing GFOs is inversely related to synaptic potentiation in the thalamocortical system (Kulikova et al., 2012). They also suggest that the ketamine-induced state results in part from dysfunction of cortical GABAergic interneurons that would lead to hyperexcitation of projection glutamatergic neurons and to glutamate release (Homayoun and Moghaddam, 2007).

Following the subcutaneous administration of a low dose (<10 mg/kg) of ketamine, both the duration and the amplitude of spontaneously occurring GFOs significantly increase in the rat frontal, parietal, and occipital cortices (Pinault, 2008; Hakami et al., 2009). From a mathematical viewpoint (Cadonic and Albensi, 2014), it was proposed that natural, physiological ongoing GFOs operate like damped harmonic oscillators, which would leave room for synaptic potentiation, learning, and memory, whereas ketamine-induced persistently amplified GFOs run like forced harmonic oscillators, which would disrupt the network ability in information processing and reduce the expression of synaptic potentiation (Pinault, 2014). This might be a key neurophysiological substrate of the functional hyperconnectivity observed by Anticevic and colleagues (2014) and by Driesen and colleagues (Driesen et al., 2013).

There is a growing body of evidence suggesting that the NMDAR antagonist ketamine modulates not only GFOs and higher-frequency oscillations, as mentioned above, but also lower-frequency oscillations, including alpha, theta, and delta oscillations (Ehrlichman et al., 2009; Hong et al., 2010; Palenicek et al., 2011; Tsuda et al., 2007). This broad-spectrum effect depends on the injected dose, the experimental and recording conditions, and on the anatomo-functional properties of the structures under investigation. For instance, under in vivo conditions, a single low-dose (<10 mg/kg) ketamine administration alters specifically GFOs and higher-frequency oscillations (Pinault, 2008; Hakami et al., 2009; Ma and Leung, 2007), while higher doses in addition affect slower rhythms (Ehrlichman et al., 2009; Hunt et al., 2006; Palenicek et al., 2011; Caixeta et al., 2013; Hiyoshi et al., 2014; Nicolas et al., 2011; Buzsaki, 1991). Therefore, we must be prudent when comparing results and inferring mechanisms from studies using different doses of NMDAR antagonists and various and diverse animal and network models. This is fundamental for basic-clinical translational understanding.

Investigating the pathophysiology of schizophrenia in relation to "noise and signal-to-noise ratio" is an appealing basic-clinical translational way to understand the neural mechanisms underlying the multiple symptoms that characterize this complex and heterogeneous neurobiological disease (Rolls et al., 2008). I recently argued on the notion of network signal-to-noise ratio (Pinault, 2014). In any system, both the amount of the ongoing (background or baseline) activity and the amplitude (or power) of its global response to the activation of its inputs are indicators of its state and functionality. The possible noise-signal interplay(s) might in part explain some disparities between findings (e.g., increases and decreases in GFOs in patients with schizophrenia; see SRF Live Discussion organized by Peter Uhlhaas and Kevin Spencer).

More precisely, in the rat thalamocortical system, ketamine simultaneously increases the power of spontaneously occurring GFOs (signature of a change in the state of the system) and decreases sensory-evoked GFOs (signature of a disturbance of the functionality of the system) (Pinault, 2008; Hakami et al., 2009; Kulikova et al., 2012). Assuming that sensory-evoked GFOs include a "true" sensory-related component, the ketamine-induced gamma noise amplification decreases the ability of the thalamocortical system to discriminate the sensory-evoked gamma signal drowned in the noise. In other words, the NMDAR antagonist ketamine decreases the gamma signal-to-noise ratio during sensory information processing. Such a ratio may be considered as a suitable neurophysiological marker of neural networks to evaluate their function and dysfunction.

This dramatically excessive ongoing gamma noise, which might be involved in the ketamine-induced state of hyperconnectivity in healthy subjects (Anticevic et al., 2014; Driesen et al., 2013), is thought to affect global brain state and operation, and to contribute to psychosis. Moreover, continuous and stereotyped GFOs might be responsible for clinical positive symptoms (Llinas et al., 1999). Furthermore, ongoing abnormally hypersynchronized GFOs have been recorded in patients experiencing sensory hallucinations (Baldeweg et al., 1998; Behrendt, 2003; Spencer et al., 2004; Ffytche, 2008; Becker et al., 2009). Hypersynchronized GFOs in cortico-thalamo-cortical systems are thought to play a key role during the appearance of hallucinations (Baldeweg et al., 1998; Behrendt, 2003), raising the question as to whether persistent amplification of ongoing GFOs could somehow conceal function-related GFOs in the corresponding brain networks.


Adler CM, Goldberg TE, Malhotra AK, Pickar D, Breier A (1998) Effects of Ketamine on Thought Disorder, Working Memory, and Semantic Memory in Healthy Volunteers. Biological Psychiatry 43:811-816. Abstract

Baldeweg T, Spence S, Hirsch SR, Gruzelier J (1998) Gamma-band electroencephalographic oscillations in a patient with somatic hallucinations. Lancet 352:620-621. Abstract

Becker C, Gramann K, Muller HJ, Elliott MA (2009) Electrophysiological correlates of flicker-induced color hallucinations. Conscious Cogn 18:266-276. Abstract

Behrendt RP (2003) Hallucinations: synchronisation of thalamocortical gamma oscillations underconstrained by sensory input. Conscious Cogn 12:413-451. Abstract

Buzsaki G (1991) The thalamic clock: emergent network properties. Neuroscience 41:351-364. Abstract

Cadonic C, Albensi BC (2014) Oscillations and NMDA receptors: Their interplay create memories. AIMS Neuroscience 1:52-64.

Caixeta FV, Cornelio AM, Scheffer-Teixeira R, Ribeiro S, Tort AB (2013) Ketamine alters oscillatory coupling in the hippocampus. Sci Rep 3:2348. Abstract

Chrobak JJ, Hinman JR, Sabolek HR (2008) Revealing past memories: proactive interference and ketamine-induced memory deficits. J Neurosci 28:4512-4520. Abstract

Driesen NR, McCarthy G, Bhagwagar Z, Bloch M, Calhoun V, D'souza DC, Gueorguieva R, He G, Ramachandran R, Suckow RF, Anticevic A, Morgan PT, Krystal JH (2013) Relationship of resting brain hyperconnectivity and schizophrenia-like symptoms produced by the NMDA receptor antagonist ketamine in humans. Mol Psychiatry 18:1199-1204. Abstract

Ehrlichman RS, Gandal MJ, Maxwell CR, Lazarewicz MT, Finkel LH, Contreras D, Turetsky BI, Siegel SJ (2009) N-methyl-d-aspartic acid receptor antagonist-induced frequency oscillations in mice recreate pattern of electrophysiological deficits in schizophrenia. Neuroscience 158:705-712. Abstract

Ffytche DH (2008) The hodology of hallucinations. Cortex 44:1067-1083. Abstract

Fond G, Loundou A, Rabu C, Macgregor A, Lancon C, Brittner M, Micoulaud-Franchi JA, Richieri R, Courtet P, Abbar M, Roger M, Leboyer M, Boyer L (2014) Ketamine administration in depressive disorders: a systematic review and meta-analysis. Psychopharmacology (Berl) 231: 3663-3676. Abstract

Hakami T, Jones NC, Tolmacheva EA, Gaudias J, Chaumont J, Salzberg M, O'Brien TJ, Pinault D (2009) NMDA receptor hypofunction leads to generalized and persistent aberrant gamma oscillations independent of hyperlocomotion and the state of consciousness. PLoS One 4:e6755. Abstract

Hetem LA, Danion JM, Diemunsch P, Brandt C (2000) Effect of a sub-anesthetic dose of ketamine on memory and conscious awareness in healthy volunteers. Psychopharmacology (Berl) 152:283-288. Abstract

Hiyoshi T, Kambe D, Karasawa J, Chaki S (2014) Differential effects of NMDA receptor antagonists at lower and higher doses on basal gamma band oscillation power in rat cortical electroencephalograms. Neuropharmacology 85:384-396. Abstract

Homayoun H, Moghaddam B (2007) NMDA receptor hypofunction produces opposite effects on prefrontal cortex interneurons and pyramidal neurons. J Neurosci 27:11496-11500. Abstract

Hong LE, Summerfelt A, Buchanan RW, O'Donnell P, Thaker GK, Weiler MA, Lahti AC (2010) Gamma and delta neural oscillations and association with clinical symptoms under subanesthetic ketamine. Neuropsychopharmacology 35:632-640. Abstract

Hunt MJ, Raynaud B, Garcia R (2006) Ketamine dose-dependently induces high-frequency oscillations in the nucleus accumbens in freely moving rats. Biol Psychiatry 60:1206-1214. Abstract

Kocsis B (2012) Differential role of NR2A and NR2B subunits in N-methyl-D-aspartate receptor antagonist-induced aberrant cortical gamma oscillations. Biol Psychiatry 71:987-995. Abstract

Krystal JH, Karper LP, Seibyl JP, Freeman GK, Delaney R, Bremner JD, Heninger GR, Bowers MB, Jr., Charney DS (1994) Subanesthetic effects of the noncompetitive NMDA antagonist, ketamine, in humans. Psychotomimetic, perceptual, cognitive, and neuroendocrine responses. Arch Gen Psychiatry 51:199-214. Abstract

Kulikova SP, Tolmacheva EA, Anderson P, Gaudias J, Adams BE, Zheng T, Pinault D (2012) Opposite effects of ketamine and deep brain stimulation on rat thalamocortical information processing. Eur J Neurosci 36:3407-3419. Abstract

Llinas RR, Ribary U, Jeanmonod D, Kronberg E, Mitra PP (1999) Thalamocortical dysrhythmia: A neurological and neuropsychiatric syndrome characterized by magnetoencephalography. Proc Natl Acad Sci U S A 96:15222-15227. Abstract

Ma J, Leung LS (2007) The supramammillo-septal-hippocampal pathway mediates sensorimotor gating impairment and hyperlocomotion induced by MK-801 and ketamine in rats. Psychopharmacology (Berl) 191:961-974. Abstract

McGirr A, Berlim MT, Bond DJ, Fleck MP, Yatham LN, Lam RW (2014) A systematic review and meta-analysis of randomized, double-blind, placebo-controlled trials of ketamine in the rapid treatment of major depressive episodes. Psychol Med1-12. Abstract

Molina LA, Skelin I, Gruber AJ (2014) Acute NMDA receptor antagonism disrupts synchronization of action potential firing in rat prefrontal cortex. PLoS One 9:e85842. Abstract

Newcomer JW, Farber NB, Jevtovic-Todorovic V, Selke G, Melson AK, Hershey T, Craft S, Olney JW (1999) Ketamine-induced NMDA receptor hypofunction as a model of memory impairment and psychosis. Neuropsychopharmacology 20:106-118. Abstract

Nicolas MJ, Lopez-Azcarate J, Valencia M, Alegre M, Perez-Alcazar M, Iriarte J, Artieda J (2011) Ketamine-induced oscillations in the motor circuit of the rat basal ganglia. PLoS One 6:e21814. Abstract

Palenicek T, Fujakova M, Brunovsky M, Balikova M, Horacek J, Gorman I, Tyls F, Tislerova B, Sos P, Bubenikova-Valesova V, Hoschl C, Krajca V (2011) Electroencephalographic spectral and coherence analysis of ketamine in rats: correlation with behavioral effects and pharmacokinetics. Neuropsychobiology 63:202-218. Abstract

Pinault D (2008) N-methyl d-aspartate receptor antagonists ketamine and MK-801 induce wake-related aberrant gamma oscillations in the rat neocortex. Biol Psychiatry 63:730-735. Abstract

Pinault (2014) N-methyl D-aspartate receptor antagonists amplify network baseline gamma frequency (30-80 Hz) oscillations: Noise and signal. AIMS Neuroscience 1:169-182.

Rolls ET, Loh M, Deco G, Winterer G (2008) Computational models of schizophrenia and dopamine modulation in the prefrontal cortex. Nat Rev Neurosci 9:696-709. Abstract

Spencer KM, Nestor PG, Perlmutter R, Niznikiewicz MA, Klump MC, Frumin M, Shenton ME, McCarley RW. Neural synchrony indexes disordered perception and cognition in schizophrenia. Proc Natl Acad Sci U S A . 2004 Dec 7 ; 101(49):17288-93. Abstract

Tsuda N, Hayashi K, Hagihira S, Sawa T (2007) Ketamine, an NMDA-antagonist, increases the oscillatory frequencies of alpha-peaks on the electroencephalographic power spectrum. Acta Anaesthesiol Scand 51:472-481. Abstract

Zarate CA, Jr., Singh JB, Carlson PJ, Brutsche NE, Ameli R, Luckenbaugh DA, Charney DS, Manji HK (2006) A randomized trial of an N-methyl-D-aspartate antagonist in treatment-resistant major depression. Arch Gen Psychiatry 63:856-864. Abstract

View all comments by Didier Pinault

Related News: Ketamine Elicits Brain State Resembling Early Stages of Schizophrenia

Comment by:  Alan AnticevicJohn Krystal (SRF Advisor)
Submitted 8 October 2014
Posted 8 October 2014

Linking Hyperconnectivity Induced via Acute Ketamine to Neuronal Mechanisms and Emerging System-Level Markers in Schizophrenia
We sincerely appreciate the recent engaging commentaries focusing on our paper. The constructive and thoughtful discussion has raised a number of important issues regarding the ketamine model in schizophrenia research. In particular, in the latest commentary Dr. Pinault has noted several hypotheses regarding our findings on which we would like to further comment.

First, what does the observed "hyperconnectivity" represent functionally? Does this "signature" relate to alterations in excitation (E) and inhibition (I) balance in cortical circuits and in turn reflect an elevation in shared signal between prefrontal areas in our analyses? It is important to note that the reported effect builds on basic preclinical studies and studies examining cortical metabolism, which have established that NMDAR antagonism elevates pyramidal cell activity (Moghaddam and Adams, 1998), extracellular glutamate levels (Homayoun and Moghaddam, 2007), cortical metabolism (Breier et al., 1997; Lahti et al., 1995; Lahti et al., 2001; Vollenweider et al., 1997; Vollenweider et al., 2000), and resting-state functional connectivity (Driesen et al., 2013; Pinault, 2008; Driesen et al., 2013). Moreover, a recent influential study by Schobel and colleagues demonstrated that NMDAR antagonism caused elevated "spreading" of hippocampal activation and metabolism (Schobel et al., 2013). Strikingly, they showed that repeated NMDAR antagonist dosing eventually caused chronic hippocampal atrophy, resembling observations after longer periods of illness. A possible corollary of this preclinical finding suggests that aberrant increases in excitation, and perhaps connectivity induced by NMDAR antagonism, are likely pathological. Therefore, the "hyperconnectivity" pattern may reflect increased pyramidal cell excitability, reduced pyramidal input selectivity, or other sources of aberrant cortical hypersynchrony.

One mechanism could involve reductions in the inhibitory component of cortical connectivity, which may alter excitatory connections. It is possible that such deficits not only increase excitability, but also reduce the input selectivity of cortical excitation (i.e., pathologically increase cortical functional connectivity). It will be critical for future studies to establish whether the observed "hyperconnectivity" measured via BOLD fMRI also relates to concurrently abnormal oscillations measured via EEG and in what frequency band. We comment on this specific issue below. Relatedly, follow-up clinical and pharmacological investigations should more carefully decompose the "correlation" measures between pairwise regions by considering both shared and unshared signal components (Friston, 2011). Put differently, it will be critical to show that the alterations in BOLD functional connectivity measures actually reflect elevations in aberrant shared signal (not just alterations in the variance structure), which could also be altered in schizophrenia (Yang et al., 2014). Pinpointing the precise nature of signal alteration that induces "hyperconnectivity" will have key implications for interpreting this effect and, in turn, for treatment studies designed to reverse this effect.

Second, what might be the precise synaptic contributions to such aberrant "hyperconnectivity"? Dr. Pinault elegantly argues for the role of GABAergic interneurons in the reported effect. We are in full agreement here; a number of studies have implicated interneurons as playing a key role in the effects of NMDAR antagonists (Krystal and Moghaddam, 2011; Krystal et al., 1994; Krystal et al., 2003; Anticevic et al., 2012). It remains unknown, however, which interneuron subtype is predominantly involved in their effects. We posit that the combination of preclinical animal work and computational modeling approaches is well positioned to address this question. For instance, studies that carefully experimentally dissect the potential contributions of specific neuronal subtypes to the observed "hyperconnectivity" pattern will be important to constrain our mechanistic understanding of acute NMDAR antagonist effects on cortical microcircuits (Kwan and Dan, 2012). Similarly, computational studies that incorporate multiple interneuron subtypes will help generate predictions for both electrophysiological and functional connectivity experiments that can be tested in animals and humans (Lisman et al., 2010; Lisman, 2012). It is well established that there are two broad classes of interneurons—fast-spiking cells that are critical for gamma oscillations and slower-spiking cells that strongly contribute to theta oscillations (Lisman, 2012). As noted, it remains unknown which GABAergic interneuron contribution "drives" the hyperconnectivity and at which frequency band (i.e., gamma or theta). It is demonstrated that systemic ketamine administration attenuates theta power but concurrently increases gamma power in mice (Ehrlichman et al., 2009; Lazarewicz et al., 2010), rats (Sabolek, H., et al., 2006), and humans (Hong et al., 2010). Computational simulations established that manipulating NMDAR currents on slow-spiking GABAergic cells (as opposed to fast-spiking cells) captured these dissociable gamma/theta experimental observations in silico (Neymotin et al., 2011). Forthcoming computational studies should strive to extend such micro-circuit models to capture neural system-level connectivity alterations in schizophrenia and following pharmacological challenge (Anticevic et al., 2013). In turn, these computational models can generate cell-specific predictions for NMDAR antagonist effects that can be tested experimentally, both preclinically and in human neuroimaging studies (Kwan and Dan, 2012).

Third, is there a dose effect of acute NMDAR antagonists? We believe that there might be a dose effect, which needs to be carefully considered when interpreting present effects—as argued by Dr. Pinault. Perhaps cognitive activation studies can speak to this issue. For instance, in a recent related study, we found that acute NMDAR antagonism in humans reduced cognitive performance—namely delayed spatial working memory (WM) by preferentially increasing error rates to non-target probes at specific spatial locations that were proximal to original WM locations (Murray et al., 2014). Put differently, volunteers administered ketamine were more likely to report a false alarm than report a memory at a location that they had not previously seen (a miss). We qualitatively and quantitatively captured this effect in our computational model by mildly reducing NMDAR drive onto inhibitory cells in the model—effectively inducing disinhibition and a "blurring" of the WM representation (Murray et al., 2014). The model performance mimicked that of healthy human volunteers at low levels of NMDAR antagonism but not at higher levels, where recurrent excitation on pyramidal cells was also reduced (Wang et al., 2013).

In a related animal study, Arnsten and colleagues examined effects of acute ketamine administration in monkeys during performance of a similar delayed spatial WM task (Wang et al., 2013). In their experiment the animals produced random saccades at previously unseen probe locations (i.e., misses), suggesting a complete loss of memory representation, accompanied by reduced delay-related firing of prefrontal neurons. In our computational simulations we were able to capture both behavioral results as reflecting distinct levels of NMDAR antagonism. In that sense, it may be the case that distinct dose-dependent levels of NMDAR antagonism produce different E/I balance regimes. Put differently, at higher doses of ketamine, where both the inhibitory and excitatory NMDAR synaptic components are affected, the "hyperconnectivity" regime may be replaced by functional connectivity reductions—a hypothesis that needs systematic experimental testing in animals. Therefore, we argue for the importance of carefully dissecting the dose-dependent contributions of ketamine on both behavior and functional connectivity in humans.

Fourth, how might the prefrontal "hyperconnectivity" effect relate to emerging "thalamo-cortical" markers of schizophrenia? An excellent observation in Dr. Pinault's commentary raises the importance of linking the observed "hyperconnectivity," which may possibly reflect alterations in gamma-band oscillations, to the thalamo-cortical gating abnormalities that underlie many theoretical models of schizophrenia (Andreasen, 1997). Unifying these effects constitutes an important theoretical and experimental goal in light of several recent empirical reports showing elevated sensory-thalamic functional connectivity in chronic schizophrenia patients (Woodward et al., 2012; Klingner et al., 2014; Anticevic et al., 2013). It may be the case that NMDAR antagonism "elevates" this sensory-thalamic coupling, either by decreasing the gamma signal-to-noise ratio or possibly by affecting the prefrontal "top-down" control aspect of sensory-thalamic information flow. These scenarios remain to be systematically tested.

In addition, it will be important to establish whether the thalamo-cortical disconnectivity in schizophrenia also exhibits functional dissociations along illness stages or appears during initial illness onset, as reported here for prefrontal hyperconnectivity. In that sense, some neural markers may show "dynamic" change along illness stages or perhaps reflect state-dependent functional causes (e.g., transient symptom exacerbation), whereas other functional connectivity markers may reflect a "trait-like" signature that persists across the illness course. Further establishing how the ketamine model maps onto these distinct neural signatures across illness stages and functional states represents a key goal for harnessing the ketamine model for targeted drug development for schizophrenia.

Moghaddam B, Adams BW. Reversal of phencyclidine effects by a group II metabotropic glutamate receptor agonist in rats. Science. 1998 Aug 28;281(5381):1349-52. Abstract

Homayoun H, Moghaddam B. NMDA receptor hypofunction produces opposite effects on prefrontal cortex interneurons and pyramidal neurons. J Neurosci. 2007 Oct 24;27(43):11496-500. Abstract

Breier A, Malhotra AK, Pinals DA, Weisenfeld NI, Pickar D. Association of ketamine-induced psychosis with focal activation of the prefrontal cortex in healthy volunteers. Am J Psychiatry. 1997 Jun;154(6):805-11. Abstract

Lahti AC, Holcomb HH, Medoff DR, Tamminga CA. Ketamine activates psychosis and alters limbic blood flow in schizophrenia. Neuroreport. 1995 Apr 19;6(6):869-72. Abstract

Lahti AC, Holcomb HH, Medoff DR, Weiler MA, Tamminga CA, Carpenter WT. Abnormal patterns of regional cerebral blood flow in schizophrenia with primary negative symptoms during an effortful auditory recognition task. Am J Psychiatry. 2001 Nov;158(11):1797-808. Abstract

Vollenweider FX, Leenders KL, Scharfetter C, Antonini A, Maguire P, Missimer J, Angst J. Metabolic hyperfrontality and psychopathology in the ketamine model of psychosis using positron emission tomography (PET) and [18F]fluorodeoxyglucose (FDG). Eur Neuropsychopharmacol. 1997 Feb;7(1):9-24. Abstract

Vollenweider FX, Vontobel P, Oye I, Hell D, Leenders KL. Effects of (S)-ketamine on striatal dopamine:a [11C]raclopride PET study of a model psychosis in humans. J Psychiatr Res. 2000 Jan-Feb;34(1):35-43. Abstract

Driesen NR, McCarthy G, Bhagwagar Z, Bloch M, Calhoun V, D'Souza DC, Gueorguieva R, He G, Ramachandran R, Suckow RF, Anticevic A, Morgan PT, Krystal JH. Relationship of resting brain hyperconnectivity and schizophrenia-like symptoms produced by the NMDA receptor antagonist ketamine in humans. Mol Psychiatry. 2013 Nov;18(11):1199-204. Abstract

Pinault D. N-methyl d-aspartate receptor antagonists ketamine and MK-801 induce wake-related aberrant gamma oscillations in the rat neocortex. Biol Psychiatry. 2008 Apr 15;63(8):730-5. Abstract

Driesen NR, McCarthy G, Bhagwagar Z, Bloch MH, Calhoun VD, D'Souza DC, Gueorguieva R, He G, Leung HC, Ramani R, Anticevic A, Suckow RF, Morgan PT, Krystal JH. The impact of NMDA receptor blockade on human working memory-related prefrontal function and connectivity. Neuropsychopharmacology. 2013 Dec;38(13):2613-22. Abstract

Schobel SA, Chaudhury NH, Khan UA, Paniagua B, Styner MA, Asllani I, Inbar BP, Corcoran CM, Lieberman JA, Moore H, Small SA. Imaging patients with psychosis and a mouse model establishes a spreading pattern of hippocampal dysfunction and implicates glutamate as a driver. Neuron. 2013 Apr 10;78(1):81-93. Abstract

Friston KJ. Functional and effective connectivity:a review. Brain Connect. 2011;1(1):13-36. Abstract

Yang GJ, Murray JD, Repovs G, Cole MW, Savic A, Glasser MF, Pittenger C, Krystal JH, Wang XJ, Pearlson GD, Glahn DC, Anticevic A. Altered global brain signal in schizophrenia. Proc Natl Acad Sci U S A. 2014 May 20;111(20):7438-43. Abstract

Krystal, J.H. and B. Moghaddam, Contributions of glutamate and GABA systems to the neurobiology and treatment of schizophrenia., in Schizophrenia, 3nd edition, P.J. Weinberger and P.J. Harrison, Editors. 2011, Blackwell Science:Oxford. p. 433-461.

Krystal JH, Karper LP, Seibyl JP, Freeman GK, Delaney R, Bremner JD, Heninger GR, Bowers MB, Charney DS. Subanesthetic effects of the noncompetitive NMDA antagonist, ketamine, in humans. Psychotomimetic, perceptual, cognitive, and neuroendocrine responses. Arch Gen Psychiatry. 1994 Mar;51(3):199-214. Abstract

Krystal JH, D'Souza DC, Mathalon D, Perry E, Belger A, Hoffman R. NMDA receptor antagonist effects, cortical glutamatergic function, and schizophrenia:toward a paradigm shift in medication development. Psychopharmacology (Berl). 2003 Sep;169(3-4):215-33.Abstract

Anticevic A, Gancsos M, Murray JD, Repovs G, Driesen NR, Ennis DJ, Niciu MJ, Morgan PT, Surti TS, Bloch MH, Ramani R, Smith MA, Wang XJ, Krystal JH, Corlett PR. NMDA receptor function in large-scale anticorrelated neural systems with implications for cognition and schizophrenia. Proc Natl Acad Sci U S A. 2012 Oct 9;109(41):16720-5. Abstract

Kwan AC, Dan Y. Dissection of cortical microcircuits by single-neuron stimulation in vivo. Curr Biol. 2012 Aug 21;22(16):1459-67. Abstract

Lisman JE, Pi HJ, Zhang Y, Otmakhova NA. A thalamo-hippocampal-ventral tegmental area loop may produce the positive feedback that underlies the psychotic break in schizophrenia. Biol Psychiatry. 2010 Jul 1;68(1):17-24. Abstract

Lisman J. Excitation, inhibition, local oscillations, or large-scale loops:what causes the symptoms of schizophrenia? Curr Opin Neurobiol. 2012 Jun;22(3):537-44. Abstract

Ehrlichman RS, Gandal MJ, Maxwell CR, Lazarewicz MT, Finkel LH, Contreras D, Turetsky BI, Siegel SJ. N-methyl-d-aspartic acid receptor antagonist-induced frequency oscillations in mice recreate pattern of electrophysiological deficits in schizophrenia. Neuroscience. 2009 Jan 23;158(2):705-12. Abstract

Lazarewicz MT, Ehrlichman RS, Maxwell CR, Gandal MJ, Finkel LH, Siegel SJ. Ketamine modulates theta and gamma oscillations. J Cogn Neurosci. 2010 Jul;22(7):1452-64. Abstract

Sabolek, H., et al., Ketamine alters synchrony throughout the hippocampal formation. Society for Neuroscience, 2006. 32:751.12.

Hong LE, Summerfelt A, Buchanan RW, O'Donnell P, Thaker GK, Weiler MA, Lahti AC. Gamma and delta neural oscillations and association with clinical symptoms under subanesthetic ketamine. Neuropsychopharmacology. 2010 Feb;35(3):632-40. Abstract

Neymotin SA, Lazarewicz MT, Sherif M, Contreras D, Finkel LH, Lytton WW. Ketamine disrupts modulation of in a computer model of hippocampus. J Neurosci. 2011 Aug 10;31(32):11733-43. Abstract

Anticevic A, Cole MW, Repovs G, Savic A, Driesen NR, Yang G, Cho YT, Murray JD, Glahn DC, Wang XJ, Krystal JH. Connectivity, pharmacology, and computation:toward a mechanistic understanding of neural system dysfunction in schizophrenia. Front Psychiatry. 2013;4():169. Abstract

Murray JD, Anticevic A, Gancsos M, Ichinose M, Corlett PR, Krystal JH, Wang XJ. Linking microcircuit dysfunction to cognitive impairment:effects of disinhibition associated with schizophrenia in a cortical working memory model. Cereb Cortex. 2014 Apr;24(4):859-72. Abstract

Wang M, Yang Y, Wang CJ, Gamo NJ, Jin LE, Mazer JA, Morrison JH, Wang XJ, Arnsten AF. NMDA receptors subserve persistent neuronal firing during working memory in dorsolateral prefrontal cortex. Neuron. 2013 Feb 20;77(4):736-49. Abstract

Andreasen NC. The role of the thalamus in schizophrenia. Can J Psychiatry. 1997 Feb;42(1):27-33. Abstract

Woodward ND, Karbasforoushan H, Heckers S. Thalamocortical dysconnectivity in schizophrenia. Am J Psychiatry. 2012 Oct;169(10):1092-9. Abstract

Klingner CM, Langbein K, Dietzek M, Smesny S, Witte OW, Sauer H, Nenadic I. Thalamocortical connectivity during resting state in schizophrenia. Eur Arch Psychiatry Clin Neurosci. 2014 Mar;264(2):111-9. Abstract

Anticevic A, Cole MW, Repovs G, Murray JD, Brumbaugh MS, Winkler AM, Savic A, Krystal JH, Pearlson GD, Glahn DC. Characterizing Thalamo-Cortical Disturbances in Schizophrenia and Bipolar Illness. Cereb Cortex. 2013 Jul 3. Abstract

View all comments by Alan Anticevic
View all comments by John Krystal

Related News: Corticostriatal Connectivity Predicts Antipsychotic Response

Comment by:  Alexander Fornito
Submitted 18 November 2014
Posted 18 November 2014

A Focus on Dorsal Frontostriatal Circuitry as a Potential Neural Basis for Psychotic Symptoms

Since seminal reports that the clinical efficacy of antipsychotic drugs correlates strongly with their capacity for striatal D2 receptor antagonism (Creese et al., 1976; Seeman and Lee, 1975), the striatum has been thought to play a central role in the pathophysiology of psychotic illness. Accordingly, human positron emission tomography (PET) studies have revealed robust elevations of markers of presynaptic dopamine in this region of the brain (Howes et al., 2012), and recent evidence suggests that these changes may be specific to the dorsal, so-called associative, striatum (Howes et al., 2009). Importantly, these elevations are present before psychosis onset (Howes et al., 2009), correlate with the severity of prodromal symptoms (Howes et al., 2009), and predict which high-risk individuals transition to psychosis (Howes et al., 2011).

Building on evidence that striatal dopamine and prefrontal activity are strongly correlated (Fusar-Poli et al., 2010; Meyer-Lindenberg et al., 2002), we recently used resting-state functional MRI to probe the functional connectivity of distinct frontostriatal circuits in people with first-episode psychosis and their unaffected first-degree relatives (Fornito et al., 2013). We found evidence for a common connectivity reduction in both patients and relatives in a dorsal circuit linking the associative striatum with dorsolateral and medial prefrontal cortex (PFC). Moreover, the magnitude of these reductions correlated with the severity of psychotic symptoms. In a separate study, we found evidence for a similar functional connectivity reduction in people experiencing an at-risk mental state for psychosis (Dandash et al., 2014). In this case, the magnitude of the connectivity reduction correlated with the severity of prodromal symptoms related to perceptual abnormalities. Collectively, these findings suggest that reduced functional connectivity between the associative striatum and dorsolateral PFC represent a candidate risk biomarker for psychosis onset that is intimately related to striatal dopamine dysregulation.

The new study of first-episode psychosis patients by Sarpal and colleagues (Sarpal et al., 2014) adds important evidence in support of this hypothesis. Specifically, they find that improvement of psychotic symptoms following 12 weeks of treatment with either risperidone or aripiprazole is associated with increased functional connectivity between the dorsal caudate and dorsolateral PFC; dorsal caudate and medial PFC; ventral caudate and hippocampus; and ventro-rostral putamen and anterior cingulate and insula cortices. Thus, where other studies have demonstrated that lower functional connectivity of the dorsal caudate-dorsolateral PFC correlates with more severe symptoms, this study shows that clinical amelioration is coupled with increased functional connectivity of this system. There is, therefore, a tight association between dorsal frontostriatal functional connectivity and the emergence of psychotic symptoms. Although it is not yet possible to determine whether the neural changes cause symptom exacerbation or are merely an epiphenomenon, the fact that functional connectivity alterations are observed in symptom-free unaffected relatives (Fornito et al., 2013) points to a causal role for frontostriatal disconnectivity in the expression of psychotic illness.

Another important conclusion that can be drawn from the study by Sarpal et al. is that the ability of antipsychotic agents to effect symptom-related changes in corticostriatal functional connectivity clearly implicates dopamine as playing a central role in this circuit-level abnormality. This result is consistent with PET evidence for dopaminergic abnormalities in patients and high-risk populations being most prominent in the associative striatum (Howes et al., 2009). Indeed, our own work has suggested that dorsal circuit changes cannot be mimicked by acute NMDA antagonism in healthy volunteers via administration of ketamine (Dandash et al., 2014), further underscoring the central role of dopamine (rather than any putative upstream candidate pathophysiological mechanisms) in these changes.

Notably, Sarpal et al. did not find any baseline corticostriatal functional connectivity differences between their patient and healthy control groups, contrasting with our own work and others' reporting evidence for patient-related reductions in this circuit (Fornito et al., 2013; Dandash et al., 2014; Anticevic et al., 2014). The reasons for this discrepancy are unclear, but may be related to methodological differences. The schizophrenia research community would benefit by leveraging the recent emergence of online data-sharing initiatives (e.g., 1000 Functional Connectomes Project) that enable the rapid and open sharing of resting-state fMRI data in order to facilitate the replication of results and the identification of the most robust brain changes underlying the disorder.


Creese I, Burt DR, Snyder SH. Dopamine receptor binding predicts clinical and pharmacological potencies of antischizophrenic drugs. Science. 1976 Apr 30; 192(4238):481-3. Abstract

Seeman P, Lee T. Antipsychotic drugs: direct correlation between clinical potency and presynaptic action on dopamine neurons. Science. 1975 Jun 20; 188(4194):1217-9. Abstract

Howes OD, Kambeitz J, Kim E, Stahl D, Slifstein M, Abi-Dargham A, Kapur S. The nature of dopamine dysfunction in schizophrenia and what this means for treatment. Arch Gen Psychiatry. 2012 Aug; 69(8):776-86. Abstract

Howes OD, Montgomery AJ, Asselin MC, Murray RM, Valli I, Tabraham P, Bramon-Bosch E, Valmaggia L, Johns L, Broome M, McGuire PK, Grasby PM. Elevated striatal dopamine function linked to prodromal signs of schizophrenia. Arch Gen Psychiatry. 2009 Jan; 66(1):13-20. Abstract

Howes O, Bose S, Turkheimer F, Valli I, Egerton A, Stahl D, Valmaggia L, Allen P, Murray R, McGuire P. Progressive increase in striatal dopamine synthesis capacity as patients develop psychosis: a PET study. Mol Psychiatry. 2011 Sep; 16(9):885-6. Abstract

Fusar-Poli P, Howes OD, Allen P, Broome M, Valli I, Asselin MC, Grasby PM, McGuire PK. Abnormal frontostriatal interactions in people with prodromal signs of psychosis: a multimodal imaging study. Arch Gen Psychiatry. 2010 Jul; 67(7):683-91. Abstract

Meyer-Lindenberg A, Miletich RS, Kohn PD, Esposito G, Carson RE, Quarantelli M, Weinberger DR, Berman KF. Reduced prefrontal activity predicts exaggerated striatal dopaminergic function in schizophrenia. Nat Neurosci. 2002 Mar; 5(3):267-71. Abstract

Fornito A, Harrison BJ, Goodby E, Dean A, Ooi C, Nathan PJ, Lennox BR, Jones PB, Suckling J, Bullmore ET. Functional dysconnectivity of corticostriatal circuitry as a risk phenotype for psychosis. JAMA Psychiatry. 2013 Nov; 70(11):1143-51. Abstract

Dandash O, Fornito A, Lee J, Keefe RS, Chee MW, Adcock RA, Pantelis C, Wood SJ, Harrison BJ. Altered striatal functional connectivity in subjects with an at-risk mental state for psychosis. Schizophr Bull. 2014 Jul; 40(4):904-13. Abstract

Sarpal DK, Robinson DG, Lencz T, Argyelan M, Ikuta T, Karlsgodt K, Gallego JA, Kane JM, Szeszko PR, Malhotra AK. Antipsychotic Treatment and Functional Connectivity of the Striatum in First-Episode Schizophrenia. JAMA Psychiatry. 2014 Nov 5. Abstract

Dandash O, Harrison BJ, Adapa R, Gaillard R, Giorlando F, Wood SJ, Fletcher PC, Fornito A. Selective Augmentation of Striatal Functional Connectivity Following NMDA Receptor Antagonism: Implications for Psychosis. Neuropsychopharmacology. 2014 Aug 21. Abstract

Anticevic A, Cole MW, Repovs G, Murray JD, Brumbaugh MS, Winkler AM, Savic A, Krystal JH, Pearlson GD, Glahn DC. Characterizing thalamo-cortical disturbances in schizophrenia and bipolar illness. Cereb Cortex. 2014 Dec; 24(12):3116-30. Abstract

View all comments by Alexander Fornito