PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Trends Cogn Sci. Author manuscript; available in PMC 2013 December 1.
Published in final edited form as:
PMCID: PMC3501603
NIHMSID: NIHMS416298

The Role of Default Network Deactivation in Cognition and Disease

Abstract

A considerable body of evidence has accumulated over recent years on the functions of the default-mode network (DMN) – a set of brain regions whose activity is high when the mind is not engaged in specific behavioral tasks and low during focused attention on the external environment. In this review, we focus on DMN suppression and its functional role in health and disease, summarizing evidence that spans several disciplines, including cognitive neuroscience, pharmacological neuroimaging, clinical neuroscience, and theoretical neuroscience. Collectively, this research highlights the functional relevance of DMN suppression for goal-directed cognition, possibly by reducing goal-irrelevant functions supported by the DMN (e.g., mind- wandering), and illustrates the functional significance of DMN suppression deficits in severe mental illness.

Keywords: default-mode network, suppression, cognition, schizophrenia, computational modeling

Brief history of the default-mode network (DMN)

Non-invasive neuroimaging, especially functional MRI (fMRI), has enabled the characterization of large-scale neural systems involved in human cognitive processes. A set of regions, identified via fMRI, includes hubs around the anterior and posterior medial cortex, bilateral temporal lobes as well as superior frontal and parietal cortices – characterized by high activity when the mind is not engaged in specific behavioral tasks and low activity during focused attention on the external environment. These regions are collectively termed the ‘default mode network’ (DMN) [1, 2]. Leading hypotheses regarding DMN function postulate its involvement in self- referential processing [3], which is typically in opposition to externally-oriented goal-directed cognition. Indeed, a distinctive feature of large-scale distributed neural networks, involving the DMN, is functional ‘antagonism’ [4]. This antagonism was initially observed in task-based neuroimaging studies [510] documenting reduced activity across DMN regions during task conditions relative to a resting baseline. Given the nature of the fMRI signal and the potential sources of error, initial concerns arose that these deactivations were spurious [11, 12]. Two decades later, there is compelling evidence that nearly every cognitive task involves the modulation of these specific deactivated regions [1, 2]. Nonetheless, research in the field continues to emphasize the functional relevance of task-positive deflections in blood-oxygenation-level-dependent (BOLD) signals, perhaps due to the long-standing hypothesis that neural computations function mainly in the service of stimulus processing. Recently, the focus has shifted toward the functional importance of task-based deactivation [13]. In this review, we highlight recent findings that allow a deeper understanding of such task-based deactivations.

In parallel with task-based studies, there has been an explosion of research that exploits resting-state functional connectivity fMRI (rs-fcMRI) to characterize the large-scale organization of neural networks via analyses of temporally correlated BOLD signals [14]. Seminal work by Biswal, Fox, Greicius, Raichle and colleagues (among others) delineated an anti-correlated system-level organization of the human brain [4, 1416]: illustrating that large-scale distributed networks exhibit negatively correlated signal over time (see Glossary). These rs-fcMRI investigations identified functional networks with striking correspondence to findings from task-induced deactivation studies [9], which have been extensively replicated across sites and studies [14], within subjects [17, 18], and across species [19]. This approach also delineated the DMN, complementing studies of cortical deactivation [1, 2, 20]. Although some controversies remain related to data analysis and potential sources of artifact [21, 22], this anti-correlated organization appears to be a fundamental property of the mammalian central nervous system [2325].

Building on these insights, there has been intensive study of the DMN’s role in higher cognition [26, 27], with an emphasis on internally-directed thought [3]. Based on the anti-correlated architecture of the human brain, there is also increasing evidence that DMN deactivation is functionally relevant for cognitive performance. Here, we summarize evidence in support of the view that DMN suppression supports certain types of goal-directed cognitive processes [28]. We detail recent evidence that demonstrates the functional relevance of breakdowns in the relationship between task-positive and default networks in pharmacological models of psychoses and two exemplar psychiatric disorders, depression and schizophrenia. In making this case, this review will draw on multiple complementary approaches including cognitive and systems neuroscience [29, 30], electrophysiology [31], clinical neuroscience [3235], computational neuroscience [36], and emerging pharmaco-fMRI studies [37, 38].

Functional relevance of DMN deactivation: Can the DMN signal be thought of as a source of distraction?

Before we discuss DMN deactivation, it is important to consider briefly what functions DMN activation might reflect [13]. As noted, the prevailing hypotheses suggest that it supports computations necessary for self- referential thought (i.e. internal mentation, daydreaming, etc. [1, 2, 3944]; for a review, see [3, 45]). The dominant proposal argues that the DMN may enable internal construction of mental models that support self- referential computations [3, 42]. This set of functions seems to contrast with externally-oriented goal-directed thought. Thus, by its nature, DMN activity could conflict with computations necessary for externally-oriented processes, reflected in the divergent architecture of the DMN and control networks [23]. Although there is a general consensus that the DMN is anti-correlated with an external attention system (EAS) [46], it is important to briefly distinguish between several sub-networks that appear to be antagonistically coupled with the DMN [46, 47]: (i) the dorsal attention network (DAN); (ii) the cingulo-opercular network (CON); (iii) and the fronto-parietal control network (FPCN) (see Glossary). This distinction is important because there is evidence to suggest that the DMN and the FPCN network positively couple during certain self-referential, cognitively demanding tasks [28, 48]. Nevertheless, here we focus on the role of DMN suppression in support of externally-oriented cognition.

Several studies have found that lower DMN activity is associated with more successful performance across a number of stimulus-driven goal-directed cognitive tasks [29, 30, 37, 49, 50]. Building on an earlier meta-analysis [9], Shulman and colleagues showed that a node proximal to the DMN (i.e., the right temporo- parietal junction, TPJ) was suppressed during correct performance of a demanding visual search task [50]. Deactivation of regions considered to be part of the DMN during encoding of novel information has been shown to enhance subsequent retrieval of learned information [30]. Similarly, less DMN suppression is associated with less efficient stimulus processing during attention lapses [51]. In addition, in a set of regions that overlap considerably with the DMN, a weaker signal during working memory (WM) encoding predicts better performance [29] (Figure 1a), which was recently replicated [37]. These findings suggest that lower DMN activity on a trial-by-trial basis is associated with better cognitive performance, indicating that DMN suppression is functionally important. One hypothesis [13] suggests that such suppression during externally focused cognition might be necessary for adaptive disengagement from certain distracting cognitive operations such as mind-wondering [40], possibly imposing a filter.

Figure 1
Functional relevance of DMN suppression

However, deactivations within parts of the DMN may depend on task characteristics [52], whereby the DMN can be subdivided into nodes that deactivate indiscriminately (“the core DMN”) and regions whose deactivation depends on task demands. This suggests a functional heterogeneity [53], which in part challenges the hypothesis that that the entire DMN is suppressed in response to any cognitive demand. Notwithstanding the possibility of functional heterogeneity, the core observation of functionally relevant DMN suppression has been replicated and extended by a number of studies [49, 52, 5456]. Animal electrophysiology adds converging evidence: non-human primates exhibit reliable suppression of neuronal firing rates in the posterior cingulate node of the DMN during performance of demanding attention and WM tasks [31]. Furthermore, higher firing in the posterior cingulate DMN node is associated with higher error rates and slower task performance [31].

Recently, an EEG-fMRI study found that increased θ oscillations were associated with suppression of the DMN and successful WM performance [49]. As discussed later, increased θ power may reflect top-down modulation of the DMN via control signals [57]. More evidence for the top-down vs. bottom-up mechanisms in DMN function was put forth by a recent elegant experiment [58] that selectively manipulated task-relevant vs. task-irrelevant distraction during WM, while measuring functional connectivity between stimulus-selective visual regions (faces vs. scenes) and the DMN as well as the FPCN. The authors reported that, depending on task demands; DMN coupling with visual cortex was predictive of task performance. Specifically, they found that the DMN is functionally correlated with stimulus-selective visual regions only in the presence of task-irrelevant distraction, postulating that: “…suppression of externally generated distracting information (that is, suppression of visual cortical activity) is intimately coupled with the suppression of internally generated distracting information (that is, suppression of default network regions).” Moreover, this connectivity finding highlights how there may be competition for computational resources between the DMN and FPCN with sensory regions (Box 1). This may be one mechanism that in part underlies the antagonistic relationship between the DMN and control regions.

Box 1

Is there competition for computational resources between anti-correlated systems?

What is the function of the antagonistic relationship between control systems and DMN? One possibility is that it may result from competition for control of shared computational resources between distributed networks. For both networks, their associated computations are likely to entail coordination with additional brain regions. For example, visual cortex is differentially coupled with each network according to task demands [58]. Visual WM or selective attention robustly activates the task-positive, fronto-parietal dorsal attention network (DAN), whose activity is often found to be anti-correlated with that of the DMN [28, 48]. Engagement in such attention-demanding tasks is also associated with positive coupling between the DAN and parts of visual cortex [58]. Interestingly, cognitive functions linked to DMN activation similarly entail coordination between visual cortex and the DMN. In some mental processes, parts of visual cortex and elements of the DMN are positively coupled [30, 58, 105], and other processes that activate the DMN cause suppression of visual cortex [106]. Therefore, DMN influence on visual cortex may disrupt its coordination with the DAN, interfering with cognitive processing. Similarly, DAN influence on visual cortex may interfere with the DMN recruiting or suppressing it. Therefore, for optimal cognitive processing, it may be beneficial for the DAN and DMN to have a reciprocally antagonistic relationship. Activation of one network would suppress the other; to limit the other’s interference in the coordination with shared neural resources relevant for supporting ongoing computations.

In addition to competition for use of sensory computational resources, the DMN and DAN appear to compete for positive coupling with brain networks involved in cognitive control, such as the cingulo-opercular network or the fronto-parietal control network (FPCN) [48]. In that sense, while the FPCN is typically conceptualized as anti-correlated with the DMN, it may actually positively couple with the DMN when organizing internal self-relevant thought [48]. Moreover, the antagonistic DAN-DMN architecture provides a substrate for control networks to flexibly perform circuit selection. The FPCN may exert cognitive control by biasing the competition between the DMN and DAN, to guide selection of the active network according to task demands. Thus, there may be a dynamic interplay for computational resources between these distributed systems, with the FPCN positioned to integrate information depending on present demands [47].

Emerging task-relevant effects clearly argue that DMN suppression is functionally important for successful operation of certain cognitive processes, such as focused attention and WM. Group differences in the capacity to suppress the DMN may also be meaningful. Indeed, a recent study identified greater deactivation of key DMN nodes in experienced meditators relative to non-meditators [59], suggesting that DMN suppression may reflect a state that could be refined with practice. Future studies should ascertain whether such meditation practice leads to better performance on externally directed cognitive tasks. In contrast, older adults exhibit deficits in filtering external distraction [6062] and less ability to suppress the DMN during cognitive tasks [3, 6365]. Thus, the capacity to “tune” DMN activity may have clinical significance.

The ability to suppress the DMN may be related to individual differences in cognitive control. To examine this possibility, Cole et al. recently identified a portion of the FPCN, lateral prefrontal cortex (PFC), in a sample of college-age adults that was associated with cognitive control and was highly globally connected with the rest of the brain [66, 67]. A follow-up analysis examined resting-state connectivity of this region with three major cortical networks and its relationship with general fluid intelligence (Figure 1b) [67]. This study found that the PFC-DMN relationship was predictive of intelligence. Strikingly, the DMN-FPCN relationship was negative: individuals with higher intelligence exhibited more negative PFC-DMN coupling. These findings suggest that individual differences in the relationship between control regions and the DMN may be related to cognitive control and the ability to modulate DMN suppression (although it remains unclear whether variability in this relationship is completely explained by activity in the FPCN: Box 2).

Box 2

What causes lack of DMN suppression in mental illness?

The DMN is increasingly implicated in mental illness by functional connectivity studies. In this review we take a different perspective on the DMN – namely the possibility that the lack of DMN suppression represents a mechanism whereby certain cognitive deficits and/or symptoms may be exacerbated in neuropsychiatric illness. If we can establish functional relevance of DMN suppression for goal-directed cognition, what does this suggest about a breakdown in the ability to accomplish such suppression? As noted in the main text, there is strong evidence for a lack of DMN suppression in neuropsychiatric disease, in particular schizophrenia. Such DMN over-activity during goal-directed cognition may in part contribute to cognitive deficits observed in this illness. That is, it is not sufficient to only consider lack of task-positive signal during cognitively demanding tasks as the only neural signature of cognitive deficits in schizophrenia. Nevertheless, while DMN signal, if left unsuppressed, may compromise optimal goal-directed cognitive function it still remains unresolved whether the ability to adequately suppress the DMN (or lack thereof) could be explained by FPCN deficits alone. Therefore, there is an interesting “chicken or egg“ problem: it is not clear whether the lack of DMN suppression adds a unique source of variance or whether we can explain the success of DMN suppression purely by considering FPCN activity. This problem highlights the issue of causality in the interaction of large-scale anti-correlated distributed neural systems. We argue that it is key to conceptualize such anti-correlated networks as a complex dynamical system where the state of each network is critical when considering the impact on each another. For instance, elevated DMN signal may adversely impact the ability to recruit the computations necessary for externally focused goal directed cognitive tasks, as it may suppress activity of such regions [37]. In turn, if there exist primary deficits in the ability to recruit computations supported by FPCN, this may result in inadequate suppression of DMN signals at times when such deactivation is critical. Nevertheless, regardless of the causal direction, if unconstrained during certain goal-directed cognitive processes, DMN signal could interfere with goal-directed cognitive function, which may be exacerbated in neuropsychiatric conditions presenting with cognitive impairment. We discuss a biophysically-realistic computational model that offers a hypothetical framework for this mechanism[37] (Box 3).

Box 3

Synaptic mechanisms of DMN suppression

In the main text, we highlight evidence that there exists an anti-correlation between regions involved in externally-directed cognition (i.e. DAN/FPCN) and regions involved in self-relevant thought (i.e. DMN). We also discussed recently emerging pharmacological neuroimaging studies that highlight how modulating both fast and slow neurotransmission can affect the interplay among these systems. In [88], Anticevic et al. explicitly manipulated the NMDA receptor component of glutamatergic neurotransmission via ketamine administration in the context of WM performance. This resulted in an attenuation of both task-positive and task-negative signals in WM-related regions, similar to prior observations in schizophrenia [32, 35] (Figure 2a–b). This finding was extended by implementing a leading hypothesis for ketamine’s effects at the synaptic level – namely cortical disinhibition – within a well-validated biophysically plausible computational model of WM [37]. The model consists of two modules: a task-activated module that is a recurrent microcircuit capable of WM computations, and a task-deactivated module representing the DMN characterized by high baseline firing rate and deactivation at task onset [31]. To capture the observed anti-correlation during task performance, the modules interact reciprocally through net inhibitory long-range projections. Disruption of NMDA conductance onto GABAergic interneurons (i.e. E–I conductance) closely reproduced observed BOLD effects following ketamine administration, which also matched prior observations from an independent WM study involving patients with schizophrenia [32] (Figure 2). These modeling results suggest that local disinhibition renders the DMN-type microcircuit hyperactive and therefore less sensitive to the long-range suppressive input from the task- activated WM-related microcircuit. Thus, the already high-firing task-deactivated module cannot be shut down at task onset. These modeling simulations demonstrate that it is vital to consider the complex dynamics between these large-scale neural systems because a given neuropathology (or pharmacological manipulation) can impact both systems due to its downstream synaptic effects (namely cortical disinhibition [89]).

Figure 2
DMN deactivation findings across clinical, pharmacological and computational approaches

The evidence reviewed in this section suggests that DMN suppression serves a vital function in the context of goal-directed cognition and cognitive control. Put simply, DMN suppression appears to be a mechanism through which the brain quiets certain internal activity (e.g. mind-wondering) to optimize externally- directed cognitive function. This observation suggests that such cognitive function would be impaired if DMN suppression was compromised, which is indeed the case in patients with neuropsychiatric disorders.

The role of DMN suppression in mental illness

A growing body of work using rs-fcMRI implicates the DMN in cognitive impairments and symptoms associated with neuropsychiatric disorders, such as schizophrenia and depression (see [3, 68]), both of which impact cognitive function. Schizophrenia is a disabling disorder associated with delusions and hallucinations [69], cognitive impairment [70], social isolation, and emotional dysfunction [71]. Lack of DMN suppression in schizophrenia has been observed when affected individuals perform demanding cognitive tasks [3235, 7279]. Studies have reported deficits in DMN suppression during the performance of the “n-back” WM task in schizophrenia patients, even before the DMN literature evolved [75] (subsequently replicated and found to be associated with worse task performance [35]), as well as schizoaffective disorder [74] and to a lesser extent, 1st-degree relatives of patients with schizophrenia. A recent study showed that, even when matched to healthy individuals for WM performance, schizophrenia is associated with reduced suppression of DMN nodes during encoding and maintenance of delayed WM [32] (Figure 2a). This challenges the notion that DMN suppression deficits in schizophrenia are purely a result of suboptimal task engagement [80]. Nonetheless, cognitively relevant DMN suppression in schizophrenia warrants further investigation. Moreover, PFC activation is associated with DMN suppression in healthy individuals, but not in schizophrenia [32, 33]. Together, these findings support the hypothesis that DMN suppression may be compromised in schizophrenia during performance of cognitively demanding tasks and contribute to cognitive impairment observed in this illness. DMN suppression deficits possibly exemplify additional forms of cortical circuit dysfunction associated with schizophrenia adding to task-related cortical activation deficits that underlie cognitive impairment (see Box 2). These findings tempt one to speculate that DMN suppression deficits compromise task-relevant signal processing, given that the weaker DMN signal may reflect the suppression of cognitive operations carried out by the DMN [5, 79, 13].

Lack of DMN suppression has also been reported in depression. Sheline and colleagues found that depressed individuals fail to reduce DMN activity while reappraising negative images [81], also observed by other groups [82]. Unlike schizophrenia, where lack of DMN suppression has been identified during cognitive demand, depression findings have been linked to negative rumination [83]. Therefore, a failure to suppress DMN nodes (in particular medial prefrontal cortex in depression) may be representative of a regime that manifests as negative internal thought, rather than the primary inability to engage executive resources (as hypothesized in schizophrenia), which is perhaps linked to aberrant automatic emotion processing [81, 84]. In the framework of dynamical systems such a regime may constitute a robust ‘attractor state’ in which the neural system is not readily dislodged from, and thus returns to, this stable state [85]. Indeed, as recently hypothesized [86], in depression there may be hyper-activity in medial prefrontal nodes, in turn inducing a hypo-function of control-related regions (due to antagonistic architecture of the two systems) without primary abnormalities in control regions. Nonetheless, a hyper-active DMN in depression may result in detrimental effects on cognitive performance [87]. However, it may be that DMN over-activity persists – possibly due to heightened rumination – even in the absence of cognitive engagement. While impaired DMN suppression may manifest across neuropsychiatric conditions, the emergent property of DMN dysfunction could possibly occur due to dissociable pathophysiological mechanisms across diagnoses, a hypothesis that warrants direct prospective investigation (see Box 4).

Box 4

Outstanding Questions Box

  • What regions/mechanisms mediate the suppression of the DMN? Is it an emergent property of the system dynamics or are there key cortical/subcortical nodes that may mediate the relationship?
  • Is the anti-correlation between DMN and task-positive networks always present? What are the mental states that occur when they anti-correlate vs. correlate and why is the anti-correlation so important functionally?
  • Is it possible that anti-correlation reflects internal competition for shared computational resources?
  • What are the common vs. unique synaptic mechanisms for the breakdown of DMN suppression across different neuropsychiatric conditions (e.g. schizophrenia vs. depression)?
  • Are individual differences in the ability to accomplish DMN suppression completely explained by variance related to cognitive control region activation?
  • Is DMN suppression equally functionally relevant across cognitive domains? Is attention the key variable? Future studies should systematically examine, in the same sample of subjects, which tasks require performance-relevant DMN suppression and to what extent.
  • Do DMN suppression deficits across neuropsychiatric conditions reflect separable processes (e.g. excessive rumination in depression vs. compromise of executive function in schizophrenia)? Prospective studies should directly compare patients diagnosed with schizophrenia vs. depression to examine whether similar effects emerge across relevant cognitive/emotional manipulations.
  • What is the dynamical nature of the DMN network, and of the spontaneous activity in the brain in general? Can it be captured theoretically using the presently available theory of deterministic chaos or stochastic processes? Alternatively, is some new type of mathematics needed for understanding the complex spatiotemporal dynamical patterns of the DMN system?

Next, we review human pharmaco-fMRI studies to identify mechanisms that might contribute to DMN dysfunction. We focus on schizophrenia as one candidate neuropathology where there is emerging understanding of the underlying neural mechanisms that may compromise DMN suppression [37]. We hope that similar evolving mechanistic understanding will help us understand DMN abnormalities across other neuropsychiatric conditions.

Pharmacological neuroimaging studies: Identifying the synaptic mechanisms of DMN suppression

To move forward in research on the clinical significance of DMN abnormalities, it will be important to understand the synaptic mechanisms underlying DMN suppression and suppression failures. Recently, Mayer and colleagues postulated that DMN suppression [52] involves long-range projections onto inhibitory interneurons. Dysfunction in inhibitory mechanisms and optimal functional antagonism across circuits has been implicated in cognitive impairment [88, 89]. However, there has been little direct testing of this hypothesis. One strategy that may enable progress in this area is pharmacological neuroimaging (ph-fMRI) [90], whereby pharmacological agents are employed to explore specific synaptic mechanisms contributing to DMN suppression.

Recent studies have implicated monoaminergic mechanisms in DMN suppression. Minzenberg and colleagues found that the stimulant modafinil enhances DMN suppression [38], and was related to successful WM performance. A subsequent study also found that at rest, increased dopamine was associated with stronger positive FPCN-DMN coupling and stronger negative DAN-DMN coupling [91]. Related findings were observed with the 5HT2A/2C receptor agonist psychedelic hallucinogen, psilocybin [92]; the authors found that at rest psilocybin increased DMN suppression. These investigations implicate possible monoaminergic modulation of cortical inhibition in the drug effects on DMN suppression.

Building on this work, Anticevic et al. [37] addressed the impact of disrupting glutamatergic neurotransmission by blocking N-methyl-D-aspartate (NMDA) receptors on DMN suppression. Anticevic et al. found that the NMDA receptor antagonist, ketamine, disrupted FPCN activation and DMN deactivation during performance of delayed WM [37]. Critically, ketamine effects on BOLD signal during delayed WM were very similar to observations in schizophrenia (Figure 2). This study also showed that less DMN suppression predicted worse WM performance within subjects – replicating prior findings [29]. Interestingly, subjects with the least DMN suppression showed more severe negative symptoms associated with schizophrenia following ketamine. As noted, one mechanism for DMN suppression may involve an active inhibition of regions not engaged in ongoing goal-directed cognition [13]. This hypothesis was formalized in a biophysically-realistic computational model of WM [93] that also implements an explicit pharmacological manipulation [94] of hypothesized ketamine effects [95], shedding light on the possible mechanisms underling DMN suppression via long-range inhibition (see Box 3 & Figure 2c). The modeling simulations were consistent with the proposal that local excitation/inhibition balance (E/I balance) may be one critical property mediating the interaction between task-based activation and suppression. These simulations were in line with the pharmacological manipulation of the NMDA receptor via ketamine and observations in schizophrenia (Figure 2a–c). Thus, the findings of this work support the hypothesis that long-range inhibition may be critical for FPCN/DMN modulation. Together, these early ph-fMRI studies illustrate one path for probing synaptic mechanisms underlying the interrelationship of FPCN and DMN, a strategy that also enables the exploration of the beneficial or maladaptive effects of modulating the functional antagonism of these networks.

Concluding remarks

We have outlined the functional relevance of DMN suppression and argue for its importance in goal- directed externally-oriented cognition. Further, we highlighted two candidate neuropsychiatric conditions where there may be a failure to suppress DMN signals that impacts symptoms and cognition, possibly due to different pathophysiological mechanisms. Based on pharmacological studies we discussed the role of both fast and slow neurotransmission in optimal modulation of DMN suppression. Lastly, we outlined a computational modeling framework for DMN suppression during cognitive operations and articulate the role of optimal inhibitory microcircuit function for coordinated DMN suppression. Further characterization of circuit mechanisms behind these observations will be critical to elucidate the role of DMN signals both in healthy cognition and neuropsychiatric illness (see Box 4). At the theoretical level, it would be interesting to quantitatively characterize the dynamical nature of the DMN and capture its complexity via neural circuit modeling, possibly with the help of chaotic dynamics theory [9698]. Using the complementary approaches of experimentation and computational theory provides a powerful cross-disciplinary approach in this exciting area of basic and clinical neuroscience in the coming years.

Acknowledgments

We thank Megan Ichinose for assistance with this manuscript. We also thank Dr. Deanna M. Barch for insightful discussions and input. We thank three anonymous Reviewers for helpful comments and constructive criticism on how to improve the manuscript.

Glossary

Anti-correlation
Hypothesized as a fundamental property of large-scale neural systems whereby signal in a set of regions is negatively correlated with signal across another set of regions. This property has been identified and extensively replicated via resting-state neuroimaging studies of the BOLD signal [14]. Example anti-correlated systems are the default-mode network and the fronto-parietal control network [4]
Default Mode Network (DMN)
A set of brain regions identified as anti-correlated with the fronto-parietal regions typically associated with cognitive control and also deactivated during demanding cognitive tasks directed towards external stimuli [4]. DMN function has been largely linked to self-relevant, internally directed information processing
Dorsal Attention Network (DAN)
A set of brain regions typically associated with focusing attention on external stimuli [99]
Fronto-parietal Control Network (FPCN)
A set of brain regions most typically implicated in executive, top- down cognitive control processes [47]
Cingulo-opercular Network (CON)
A set of brain regions typically associated with maintenance of a cognitive set over an extended period during performance of externally focused cognitive tasks [100]
Externally-directed Cognition
A set of cognitive processes that involve representing an external source of information in the service of a goal or action. A prototypical example is working memory [101, 102], which involves encoding, manipulating and retrieving information in the service of some goal and is typically associated with activity in a distributed network encompassing fronto-parietal cortical regions. Overall, externally-directed cognitive processes are most typically associated with the FPCN, DAN and the CON [100, 103]
Internally-directed cognition
A set of cognitive processes that involve self-relevant information processing and spontaneous cognition such as autobiographical memories. These processes are most typically associated with the DMN [26]
Biophysically-based Computational Modeling
A branch of mathematical modeling of neural function based on biologically realistic principles of neural function [36]
Schizophrenia
A severe psychiatric syndrome that is most typically associated with co-occurring symptoms such as delusions and hallucinations but is also characterized by severe cognitive and affective disturbances [69]
Ketamine
A non-competitive N-Methyl-D-aspertate (NMDA) receptor antagonist. Ketamine has been successfully used as a pharmacological model of schizophrenia as it transiently and safely invokes the cardinal symptoms associated with this illness. At doses used to produce psychotomimetic effects, ketamine is hypothesized to more selectively block NMDA receptor conductances on inhibitory γ-Aminobutyric acid (GABA) expressing neurons [104]. Therefore, it is hypothesized that ketamine’s effects may induce a state cortical disinhibition, induced by a lack of inhibitory interneuron function [95]

Footnotes

Disclosure statement

J.H.K. has been a paid consultant for AbbVie, Inc. (formerly Abbott Laboratories); Aisling Capital, LLC; AstraZeneca Pharmaceuticals; Bristol-Myers Squibb; Eisai, Inc.; Eli Lilly and Co.; Gilead Sciences, Inc.; Lundbeck Research USA; Medivation, Inc.; Otsuka Pharmaceutical Development & Commercialization, Inc.; Roche F. Hoffmann-La Roche Ltd; Sage Therapeutics, Inc.; Shire Pharmaceuticals; Takeda Industries; and Teva Pharmaceutical Industries, Ltd. J.H.K. serves as a member of the Scientific Advisory Boards for CHDI Foundation, Inc.; Lohocla Research Corporation; Mnemosyne Pharmaceuticals, Inc.; Naurex, Inc.; and Pfizer Pharmaceuticals. J.H.K. is a member of the Board of Directors for the Coalition for Translational Research in Alcohol and Substance Use Disorders, is the President Elect of the American College of Neuropsychopharmacology, and is the Editor of Biological Psychiatry. J.H.K. is listed as an inventor on the patent: Seibyl JP, Krystal JH, Charney DS (1995) US Patent 5,447,948. J.H.K. is listed as an inventor on a patent application by Yale University related to targeting the glutamatergic system for the treatment of neuropsychiatric disorders (application no. PCTWO06108055A1). J.H.K. is listed on a pending patent application related to intranasal administration of ketamine to treat depression. The other authors declare no conflict of interest.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

1. Raichle ME, et al. A default mode of brain function. Proc Natl Acad Sci U S A. 2001;98:676–682. [PubMed]
2. Raichle ME, Snyder AZ. A default mode of brain function: a brief history of an evolving idea. Neuroimage. 2007;37:1083–1090. discussion 1097–1089. [PubMed]
3. Buckner RL, et al. The brain’s default network: anatomy, function, and relevance to disease. Ann N Y Acad Sci. 2008;1124:1–38. [PubMed]
4. Fox MD, et al. The human brain is intrinsically organized into dynamic, anticorrelated functional networks. Proc Natl Acad Sci U S A. 2005;102:9673–9678. [PubMed]
5. Binder JR, et al. Conceptual processing during the conscious resting state: a functional MRI study. J Cogn Neurosci. 1999;11:80–95. [PubMed]
6. Frith CD, et al. A PET study of word finding. Neuropsychologia. 1991;29:1137–1148. [PubMed]
7. Haxby JV, et al. The functional organization of human extrastriate cortex: a PET-rCBF study of selective attention to faces and locations. J Neurosci. 1994;14:6336–6353. [PubMed]
8. Kawashima R, et al. Positron-emission tomography studies of corss-modality inhibition in selective attentional tasks: closing the “mind’s eye. Proc Natl Acad Sci U S A. 1995;92:5969–5972. [PubMed]
9. Shulman GL, et al. Common blood flow changes across visual tasks: I. Increases in subcortical structures and cerebellum but not in nonvisual cortex. J Cogn Neurosci. 1997;9:624–647. [PubMed]
10. Warburton E, et al. Noun and verb retrieval by normal subjects. Studies with PET. Brain. 1996;119:159–179. [PubMed]
11. Buckner R. The serendipitous discovery of the brain’s default network. Neuroimage. 2012;62:1137–1145. [PubMed]
12. Gusnard DA, Raichle ME. Searching for a baseline: functional imaging and the resting human brain. Nat Rev Neurosci. 2001;2:685–694. [PubMed]
13. Binder J. Task-induced deactivation and the “resting” state. Neuroimage. 2012;62:1086–1091. [PMC free article] [PubMed]
14. Biswal BB, et al. Toward discovery science of human brain function. Proc Natl Acad Sci U S A. 2010;107:4734–4739. [PubMed]
15. Biswal B, et al. Functional connectivity in the motor cortex of resting human brain using echo-planar MRI. Magn Reson Med. 1995;34:537–541. [PubMed]
16. Greicius MD, et al. Functional connectivity in the resting brain: a network analysis of the default mode hypothesis. Proc Natl Acad Sci U S A. 2003;100:253–258. [PubMed]
17. Shehzad Z, et al. The resting brain: unconstrained yet reliable. Cereb Cortex. 2009;19:2209–2229. [PMC free article] [PubMed]
18. Van Dijk KRA, et al. Intrinsic functional connectivity as a tool for human connectomics: theory, properties, and optimization. J Neurophysiol. 2010;103:297–321. [PubMed]
19. Vincent JL, et al. Intrinsic functional architecture in the anaesthetized monkey brain. Nature. 2007;447:83–86. [PubMed]
20. Smith SM, et al. Correspondence of the brain’s functional architecture during activation and rest. Proc Natl Acad Sci U S A. 2009;106:13040–13045. [PubMed]
21. Murphy K, et al. The impact of global signal regression on resting state correlations: are anti- correlated networks introduced? Neuroimage. 2009;44:893–905. [PMC free article] [PubMed]
22. Fox M, et al. The Global Signal and Observed Anticorrelated Resting State Brain Networks. J Neurophysiol. 2009;101:3270–3283. [PubMed]
23. Smith SM, et al. Temporally-independent functional modes of spontaneous brain activity. Proc Natl Acad Sci U S A. 2012;109:3131–3136. [PubMed]
24. Lu H, et al. Rat brains also have a default mode network. Proc Natl Acad Sci U S A. 2012;109:3979–3984. [PubMed]
25. Mantini D, et al. Default mode of brain function in monkeys. J Neurosci. 2011;31:12954–12962. [PMC free article] [PubMed]
26. Andrews-Hanna JR, et al. Evidence for the Default Network’s Role in Spontaneous Cognition. J Neurophysiol. 2010;104:322–335. [PubMed]
27. Andrews-Hanna JR. The brain’s default network and its adaptive role in internal mentation. Neuroscientist. 2011;18:251–270. [PMC free article] [PubMed]
28. Spreng RN. The fallacy of a “task-negative” network. Front Psychol. 2012;3 doi: 10.3389/fpsyg.2012.00145. [PMC free article] [PubMed] [Cross Ref]
29. Anticevic A, et al. When less is more: TPJ and default network deactivation during encoding predicts working memory performance. Neuroimage. 2010;49:2638–2648. [PMC free article] [PubMed]
30. Daselaar SM, et al. When less means more: deactivations during encoding that predict subsequent memory. Neuroimage. 2004;23:921–927. [PubMed]
31. Hayden BY, et al. Electrophysiological correlates of default-mode processing in macaque posterior cingulate cortex. Proc Natl Acad Sci U S A. 2009;106:5948–5953. [PubMed]
32. Anticevic A, et al. Working Memory Encoding and Maintenance Deficits in Schizophrenia: Neural Evidence for Activation and Deactivation Abnormalities. Schizophr Bull. 2011 doi: 10.1093/schbul/sbr107. [Epub ahead of print] [PMC free article] [PubMed] [Cross Ref]
33. Metzak PD, et al. Decreased Efficiency of Task-Positive and Task-Negative Networks During Working Memory in Schizophrenia. Schizophr Bull. 2011;38:803–813. [PMC free article] [PubMed]
34. Nejad AB, et al. Impaired temporoparietal deactivation with working memory load in antipsychotic- naïve patients with first-episode schizophrenia. World J Biol Psychiatry. 2011;12:271–281. [PubMed]
35. Whitfield-Gabrieli S, et al. 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. 2009;106:1279–1284. [PubMed]
36. Wang XJ. Neurophysiological and computational principles of cortical rhythms in cognition. Physiol Rev. 2010;90:1195–1268. [PMC free article] [PubMed]
37. Anticevic A, et al. NMDA Receptor Function in Large-Scale Anti-Correlated Neural Systems with Implications for Cognition and Schizophrenia. Proc Natl Acad Sci U S A. 2012;109:16720–16725. [PubMed]
38. Minzenberg MJ, et al. Modafinil modulation of the default mode network. Psychopharmacology (Berl) 2011;215:23–31. [PMC free article] [PubMed]
39. Antrobus J. Dreaming: cognitive processes during cortical activation ad high afferent thresholds. Psychol Rev. 1991;98:96–121. [PubMed]
40. Antrobus JS, et al. Mindwandering and cognitive structure. Trans N Y Acad Sci. 1970;32:242–252. [PubMed]
41. Binder JR, et al. Where is the semantic system? A critical review and meta-analysis of 120 functional neuroimaging studies. Cereb Cortex. 2009;19:2767–2796. [PMC free article] [PubMed]
42. Buckner RL, Caroll DC. Self-projection and the brain. Trends Cogn Sci. 2007;2:49–57. [PubMed]
43. Mason MF, et al. Wandering minds: the default network and stimulus-independent thought. Science. 2007;315:393–395. [PMC free article] [PubMed]
44. Mazoyer B, et al. Cortical networks for working memory and executive functions sustain the conscious resting state in man. Brain Res Bull. 2001;54:287–298. [PubMed]
45. Andrews-Hanna JR. The Brain’s Default Network and Its Adaptive Role in Internal Mentation. Neuroscientist. 2012;18:251–270. [PMC free article] [PubMed]
46. Fornito A, et al. Competitive and cooperative dynamics of large-scale brain functional networks supporting recollection. Proc Natl Acad Sci U S A. 2012;109:12788–12793. [PubMed]
47. Vincent JL, et al. Evidence for a frontoparietal control system revealed by intrinsic functional connectivity. J Neurophysiol. 2008;100:3328–3342. [PubMed]
48. Spreng RN, et al. Default network activity, coupled with the frontoparietal control network, supports goal-directed cognition. Neuroimage. 2010;53:303–317. [PMC free article] [PubMed]
49. White TP, et al. Theta power during encoding predicts subsequent-memory performance and default mode network deactivation. Hum Brain Mapp. 2012 doi: 10.1002/hbm.22114. [Epub ahead of print] [PubMed] [Cross Ref]
50. Shulman GL, et al. Right TPJ deactivation during visual search: functional significance and support for a filter hypothesis. Cereb Cortex. 2007;17:2625–2633. [PubMed]
51. Weissman DH, et al. The neural bases of momentary lapses in attention. Nat Neurosci. 2006;9:971–978. [PubMed]
52. Mayer JS, et al. Specialization in the default mode: Task-induced brain deactivations dissociate between visual working memory and attention. Hum Brain Mapp. 2010;31:126–139. [PubMed]
53. Leech R, et al. Fractionating the default mode network: distinct contributions of the ventral and dorsal posterior cingulate cortex to cognitive control. J Neurosci. 2011;31:3217–3224. [PubMed]
54. Newton AT, et al. Modulation of steady state functional connectivity in the default mode and working memory networks by cognitive load. Hum Brain Mapp. 2011;32:1649–1659. [PMC free article] [PubMed]
55. Pfefferbaum A, et al. Cerebral blood flow in posterior cortical nodes of the default mode network decreases with task engagement but remains higher than in most brain regions. Cereb Cortex. 2011;21:233–244. [PMC free article] [PubMed]
56. Takeuchi H, et al. Failing to deactivate: the association between brain activity during a working memory task and creativity. Neuroimage. 2011;15:681–687. [PubMed]
57. Liebe S, et al. Theta coupling between V4 and prefrontal cortex predicts visual short-term memory performance. Nat Neurosci. 2012;15:456–462. S451–452. [PubMed]
58. Chadick JZ, Gazzaley A. Differential coupling of visual cortex with default or frontal-parietal network based on goals. Nat Neurosci. 2011;14:830–832. [PMC free article] [PubMed]
59. Brewer JA, et al. Meditation experience is associated with differences in default mode network activity and connectivity. Proc Natl Acad Sci U S A. 2011;108:20254–20259. [PubMed]
60. Gazzaley A, et al. Age--related top--down suppression deficit in the early stages of cortical visual memory processing. Proc Natl Acad Sci U S A. 2008;105:13122–13126. [PubMed]
61. Gazzaley A, et al. Top-down suppression deficit underlies working memory impairment in normal aging. Nat Neurosci. 2005;8:1298–1300. [PubMed]
62. Gazzaley A, D’Esposito M. Top--down modulation and normal aging. Ann N Y Acad Sci. 2007;1097:67–83. [PubMed]
63. Grady CL, et al. A multivariate analysis of age-related differences in default mode and task- positive networks across multiple cognitive domains. Cereb Cortex. 2010;20:1432–1447. [PMC free article] [PubMed]
64. Hedden T, et al. Disruption of functional connectivity in clinically normal older adults harboring amyloid burden. J Neurosci. 2009;29:12686–12694. [PMC free article] [PubMed]
65. Lustig C, et al. Functional deactivations: change with age and dementia of the Alzheimer type. Proc Natl Acad Sci U S A. 2003;100:14504–14509. [PubMed]
66. Cole MW, et al. Identifying the brain’s most globally connected regions. Neuroimage. 2010;49:3132–3148. [PubMed]
67. Cole MW, et al. Global connectivity of prefrontal cortex predicts cognitive control and intelligence. J Neurosci. 2012;32:8988–8999. [PMC free article] [PubMed]
68. Fox MD, Greicius M. Clinical applications of resting state functional connectivity. Front Syst Neurosci. 2010;4:19. doi: 10.3389/fnsys.2010.00019. [PMC free article] [PubMed] [Cross Ref]
69. Walker E, et al. Schizophrenia: etiology and course. Annu Rev Psychol. 2004;55:401–430. [PubMed]
70. Barch DM, Ceaser A. Cognition in schizophrenia: core psychological and neural mechanisms. Trends Cogn Sci. 2012;16:27–34. [PMC free article] [PubMed]
71. Kring AM, Moran EK. Emotional response deficits in schizophrenia: insights from affective science. Schizophr Bull. 2008;34:819–834. [PMC free article] [PubMed]
72. Dreher JC, et al. Common and differential pathophysiological features accompany comparable cognitive impairments in medication-free patients with schizophrenia and in healthy aging subjects. Biol Psychiatry. 2012;71:890–897. [PubMed]
73. Fatjó-Vilas M, et al. Effect of the Interleukin-1β Gene on Dorsolateral Prefrontal Cortex Function in Schizophrenia: A Genetic Neuroimaging Study. Biol Psychiatry. 2012;72:758–765. [PubMed]
74. Madre M, et al. Brain functional abnormality in schizo-affective disorder: an fMRI study. Psychol Med. 2012:1–11. doi: 10.1017/S0033291712000943. [Epub ahead of print] [PubMed] [Cross Ref]
75. Meyer-Lindenberg A, et al. Evidence for abnormal cortical functional connectivity during working memory in schizophrenia. Am J Psychiatry. 2001;158:1809–1817. [PubMed]
76. Nygård M, et al. Patients with Schizophrenia Fail to Up-Regulate Task-Positive and Down-Regulate Task-Negative Brain Networks: An fMRI Study Using an ICA Analysis Approach. Front Hum Neurosci. 2012;6 doi: 10.3389/fnhum.2012.00149. [PMC free article] [PubMed] [Cross Ref]
77. Pomarol-Clotet E, et al. Failure to deactivate in the prefrontal cortex in schizophrenia: dysfunction of the default mode network? Psychol Med. 2008;38:1185–1193. [PubMed]
78. Salgado-Pineda P, et al. Correlated structural and functional brain abnormalities in the default mode network in schizophrenia patients. Schizophr Res. 2011;125:101–109. [PubMed]
79. Schneider FC, et al. Modulation of the default mode network is task-dependant in chronic schizophrenia patients. Schizophr Res. 2011;125:110–117. [PubMed]
80. Whitfield-Gabrieli S, Ford JM. Default Mode Network Activity and Connectivity in Psychopathology. Annu Rev Clin Psychol. 2012;8:49–76. [PubMed]
81. Sheline YI, et al. The default mode network and self-referential processes in depression. Proc Natl Acad Sci U S A. 2009;106:1942–1947. [PubMed]
82. Hamilton JP, et al. Default-mode and task-positive network activity in major depressive disorder: implications for adaptive and maladaptive rumination. Biol Psychiatry. 2011;70:327–333. [PMC free article] [PubMed]
83. Lemogne C, et al. Medial prefrontal cortex and the self in major depression. J Affect Disord. 2012;136:e1–e11. [PubMed]
84. Perrin JS, et al. Electroconvulsive therapy reduces frontal cortical connectivity in severe depressive disorder. Proc Natl Acad Sci U S A. 2012;109:5464–5468. [PubMed]
85. Rolls ET. Attractor networks. Wiley Interdisciplinary Reviews: Cognitive Science. 2009;1:119–134.
86. Fox MD, et al. Efficacy of Transcranial Magnetic Stimulation Targets for Depression Is Related to Intrinsic Functional Connectivity with the Subgenual Cingulate. Biol Psychiatry. 2012;72:595–603. [PubMed]
87. Rose EJ, et al. Limbic over-activity in depression during preserved performance on the n-back task. Neuroimage. 2006;29:203–215. [PubMed]
88. Rao SG, et al. Destruction and creation of spatial tuning by disinhibition: GABA(A) blockade of prefrontal cortical neurons engaged by working memory. J Neurosci. 2000;20:485–494. [PubMed]
89. Yizhar O, et al. Neocortical excitation/inhibition balance in information processing and social dysfunction. Nature. 2011;477:171–178. [PubMed]
90. Honey G, Bullmore E. Human pharmacological MRI. Trends Pharmacol Sci. 2004;25:366–374. [PubMed]
91. Dang LC, et al. Dopamine Supports Coupling of Attention-Related Networks. The Journal of Neuroscience. 2012;32:9582–9587. [PMC free article] [PubMed]
92. Carhart-Harris RL, et al. Neural correlates of the psychedelic state as determined by fMRI studies with psilocybin. Proc Natl Acad Sci U S A. 2012;109:2138–2143. [PubMed]
93. Compte A, et al. Synaptic mechanisms and network dynamics underlying spatial working memory in a cortical network model. Cereb Cortex. 2000;10:910–923. [PubMed]
94. Brunel N, Wang XJ. Effects of neuromodulation in a cortical network model of object working memory dominated by recurrent inhibition. J Comput Neurosci. 2001;11:63–85. [PubMed]
95. Krystal JH, et al. NMDA receptor antagonist effects, cortical glutamatergic function, and schizophrenia: toward a paradigm shift in medication development. Psychopharmacology (Berl) 2003;169:215–233. [PubMed]
96. Sompolinsky H, et al. Chaos in random neural networks. Phys Rev Lett. 1988;61:259–262. [PubMed]
97. Sussillo D, Abbott LF. Generating coherent patterns of activity from chaotic neural networks. Neuron. 2009;63:544–457. [PMC free article] [PubMed]
98. Wang X-J. Spontaneous activity in a large neural net: between chaos and noise. In: Babloyantz A, editor. Self-Organization, Emergent Properties and Learning. Plenum Press; 1991. pp. 155–160.
99. Fox MD, et al. Spontaneous neuronal activity distinguishes human dorsal and ventral attention systems. Proc Natl Acad Sci U S A. 2006;103:10046–10051. [PubMed]
100. Dosenbach NU, et al. Distinct brain networks for adaptive and stable task control in humans. Proc Natl Acad Sci U S A. 2007;104:11073–11078. [PubMed]
101. Goldman-Rakic. Cellular basis of working memory. Neuron. 1995;14:477–485. [PubMed]
102. Jonides J, et al. The mind and brain of short-term memory. Annu Rev Psychol. 2008;59:193–224. [PubMed]
103. Dosenbach NU, et al. A dual-networks architecture of top-down control. Trends Cogn Sci. 2008;12:99–105. [PMC free article] [PubMed]
104. Kotermanski SE, Johnson JW. Mg2+ Imparts NMDA Receptor Subtype Selectivity to the Alzheimer’s Drug Memantine. J Neurosci. 2009;29:2774–2779. [PMC free article] [PubMed]
105. Wang K, et al. Spontaneous activity associated with primary visual cortex: a resting-state FMRI study. Cereb Cortex. 2008;18:697–704. [PubMed]
106. Smallwood J, et al. Cooperation between the default mode network and the frontal-parietal network in the production of an internal train of thought. Brain Res. 2012;1428:60–70. [PubMed]