Brian N. Mathur

The claustrum is a subcortical nucleus with widespread, reciprocal connections to the cortex and beyond, inspiring many models of its function. Decades of work from rodents provide a foundational understanding of the anatomy and functional organization of the claustrum, which is more recently aided by human functional neuroimaging studies. However, important questions remain with broad functional implications. These center on claustrum anatomical boundaries, functional organization, and the putative expansion of human claustrum connectivity over that predicted by rodent studies alone. Providing new insights by spanning the large rodent-to-human void, a landmark study by Lei and colleagues recently provided thorough molecular and connectivity profiling of the macaque claustrum. Surprising findings of lateralized and expanded claustrum connectivity with cortical and subcortical partners help to explain human functional connectivity data and the discovery of a well-developed organization of the monkey claustrum into overlapping functional sub-domains provides deep insight into claustrum function when considering rodent synaptic connectivity data. Contextualizing these findings, we provide an elaborated model of the function of the claustrum in cognition.

Cognitive control is believed to arise from task-dependent interactions among networks of brain regions. Although several debilitating neuropsychiatric disorders are characterized by cognitive network dysfunction, the neural circuit mechanisms supporting task-dependent network activity are largely unknown. External and internal task demands elicit opposing responses from key cognitive networks, and claustrum projections target regions associated with both network states. We tested if claustrum supports task-dependent network activity in humans using fMRI during tasks with externally and internally driven demands: working memory ( n = 420) and autobiographical memory ( n = 35). Claustrum activity increased in both tasks. Claustrum exhibited anatomical connectivity with regions representing all implicated networks, and claustrum effective connectivity suggested an excitatory influence on regions in multiple task-associated networks. Task response and connectivity measures differed between the claustrum and regions prominently implicated in directing network states—the anterior insula and pulvinar. These findings establish a role for the claustrum in supporting task-dependent network states subserving cognitive control.

The claustrum is among the most highly interconnected structures in the mammalian brain, with reciprocal projections spanning the cortical mantle and a growing roster of functions that range from sleep regulation to attentional control. Here we review current understanding of claustral circuit organization and function, synthesizing findings from anatomical, physiological, computational, and behavioral studies. We argue that a unifying logic underlies the apparent diversity of claustral functions: the claustrum operates through projection-defined subnetworks whose input selectivity, output targets, and local circuit interactions collectively determine their functional contribution, and whose impact on cortical targets varies with arousal and ongoing brain state. Corticoclaustral connectivity is organized into matched input-output modules, local inhibitory circuits confer spatial filtering properties that amplify strong cortical signals while suppressing weaker ones, and claustral output populations differ in their laminar targets, postsynaptic cell-type preferences, and molecular properties, producing diverse and context-dependent effects on cortical dynamics. These circuit properties translate into claustral regulation of cortical population dynamics across multiple timescales, from the moment-to-moment coordination of brain states to the consolidation of frontal control strategies. In pain and addiction, specific claustral projection populations are selectively dysregulated, and distinct subnetworks exert opposing influences on nociceptive sensitivity and drug-related behavior according to their cortical targets. Together, these findings position the claustrum as a modular subcortical hub for the selective regulation of distributed cortical networks and identify it as a candidate therapeutic target for disorders characterized by dysregulated cortical network activity, including chronic pain, addiction, and impulsivity.

Cognitive impairment is a major component of Alcohol Use Disorder. Optimal cognitive performance requires anterior cingulate input activation of the claustrum, a subcortical nucleus that orchestrates cortical activity. Yet the impact of chronic alcohol exposure on the ability for the anterior cingulate cortex to drive activity of claustrum projection neuron subtypes is unknown. In adult male and female mice, we found that the majority of non-burst firing Type 1 claustrum projection neurons did not express the vesicular glutamate transporter 2 (VGLUT2), while the majority of burst firing Type 2 projection neurons were VGLUT2-expressing. Following chronic intermittent vaporized ethanol exposure (CIE), we found that both Type 1 and VGLUT2-non-expressing neurons exhibited increased responsivity to anterior cingulate cortex input activation that was mediated by increased postsynaptic membrane excitability. In contrast, Type 2 and VGLUT2-expressing projection neurons exhibited increased responsivity to anterior cingulate cortex input due to strengthened pre- and post-synaptic transmission mechanisms. Altogether, we uncovered a hyper-excitatory drive of the claustrum by the anterior cingulate cortex following chronic alcohol exposure. The data provide a foundational resource for the complex effects of chronic alcohol exposure on the claustrum, a critical cognitive control nucleus.

Through its widespread reciprocal connections with the cerebral cortex, the claustrum is implicated in sleep and waking cortical network states. Yet, basic knowledge of neuromodulation in this structure is lacking. The claustrum is richly innervated by serotonergic fibers, expresses serotonin receptors, and is suggested to play a role in the ability of psilocybin, which is metabolized to the non-specific serotonin receptor agonist psilocin, to disrupt cortex-wide network states. We therefore addressed the possible role of serotonin, and the classic psychedelic psilocybin, in modulating cortical signaling through the claustrum. We show that serotonin activates 5-HT_1B receptors on anterior cingulate cortex inputs – a primary driver of claustrum activity – to suppress signaling to parietal association cortex-projecting claustrum neurons. Additionally, we demonstrate that psilocybin injection also activates anterior cingulate cortex presynaptic 5-HT_1B receptors to suppress cortical signaling through the claustrum. Thus, serotonin, via 5-HT_1B, may provide gain-control of cortical input to the claustrum, a mechanism that may be directly targeted by psilocybin to modulate downstream cortical network states.

Alcohol use disorder is characterized by persistent drinking in the face of negative consequences. Such inflexible drinking requires dorsolateral striatum fast-spiking interneurons, which comprise roughly 1% of all striatal neurons. How chronic ethanol exposure affects fast-spiking interneuron physiology is poorly understood. We discover in mice that chronic ethanol exposure induced a dramatic loss of GABAergic, but not glutamatergic, synapses onto dorsolateral striatum fast-spiking interneuron somata and proximal dendrites where perineuronal nets, a subdivision of the extracellular matrix, are enriched. We found that chronic ethanol exposure degraded these perineuronal nets and that enzymatically degrading perineuronal nets similarly reduced GABAergic transmission onto dorsolateral striatum fast-spiking interneurons. Modeling the effect of alcohol, we find that silencing extrinsic GABAergic projections to the dorsolateral striatum increased voluntary ethanol consumption. Taken together, these data suggest chronic alcohol exposure remodels perineuronal nets and inhibitory synapses on fast-spiking interneurons to facilitate alcohol drinking.

Cognitive control is believed to arise from interactions among multiple brain networks depending on task demands. Although several debilitating neuropsychiatric disorders are characterized by cognitive network dysfunction, the neural circuit mechanisms supporting task-dependent network activation are largely unknown. Because the claustrum possesses widespread connections with cortex and can synchronize distant cortical regions, we tested whether the claustrum activates task-dependent network states using fMRI during working memory ( n = 420) and autobiographical memory ( n = 35), tasks which elicit opposing responses from key cognitive control networks. In both tasks, the claustrum exhibited increased activity and excitatory influence on task-associated cognitive control network nodes, with corroborating underlying structural connectivity. The claustrum also displayed stronger excitatory effective connectivity during task performance and greater structural connectivity with task-related network nodes than regions prominently implicated in directing network states—the anterior insula and pulvinar. These findings establish a role for the claustrum in initiating network states for cognitive control. Significance: Cognitive functioning is supported by large-scale networks across the brain. Yet, the neural circuit mechanisms supporting task-dependent network activation are largely unknown. Circuit analyses using human functional and structural neuroimaging in this study found that the claustrum, a subcortical nucleus, activates during multiple cognitive tasks eliciting a wide range of network states, possesses strong anatomical connections with cognitive control network nodes, and exerts excitatory influence on task-associated network regions. These results establish the claustrum as a network activator subserving cognitive control.

Dysregulation of normal reward processing via psychological stress contributes to the development of psychiatric disorders. Estrogen is involved in reward processing in females, but this effect has not been well studied in males despite the abundant conversion of androgens to estrogens in the male brain. Here, we used a combination of genetic deletions, behavioral assays, pharmacology, circuit dissection, electrophysiology, in vivo fiber photometry, and optogenetics/chemogenetics to determine the role of the most prevalent and potent estrogen, 17β-estradiol, in male stress-induced reward processing dysfunction. We found that absence of estrogen receptor (ER) β renders male but not female mice susceptible to stress-induced maladaptive reward-processing behaviors. We demonstrated that activation of ERβ-projecting neurons from the basolateral amygdala to nucleus accumbens induced rewarding effects in male, but not female mice. Moreover, we show that the activity of ERβ-expressing neurons projecting from the basolateral amygdala to nucleus accumbens is reduced in hypogonadal male mice subjected to stress, while activation of this circuit reverses stress-induced maladaptive reward processing behaviors and inhibition induces stress susceptibility. We identified that absence of estradiol, but not testosterone per se, underlies susceptibility to stress-mediated dysfunction of rewarding behaviors and that brain-selective delivery of estradiol and intra-basolateral amygdala administration of an ERβ-specific agonist prevent maladaptive reward-processing behaviors in hypogonadal male mice. These findings delineate an estrogen-based mechanism underlying stress susceptibility and provide a novel therapeutic strategy for the treatment of reward-related disorders associated with hypogonadal conditions.