Post-stroke neuronal circuits and mental illnesses

Abstract

Stroke is one of the leading causes of death in the United States. It is also associated with severe mental illnesses, such as depression and anxiety, that hinder the rehabilitation of surviving patients. Thus, a better understanding of how stroke causes mental illnesses is crucial, but little is known about the neurological mechanisms involved. In this review, we summarized the most common mental illnesses developed after stroke, as well as the underlying mechanisms at the neuronal circuit level.

Introduction

Stroke is defined as any central nervous system (CNS) damage resulting from an abnormal blood supply [1] and is one of the leading causes of death in the United States. Stroke classifications include cerebral infarction, cerebral hemorrhage, subarachnoid hemorrhage, cerebral venous thrombosis, and spinal cord stroke. Cerebral infarction itself is subdivided into four categories: lacunar, atherothrombotic, cardioembolic, and undetermined types [2]. The incidence of stroke and the frequency of stroke subtypes can be influenced by several factors, including age, race, and sex [3,4].

According to the recent stroke statistics report from the U.S. Centers for Disease Control and Prevention, over 795,000 Americans are affected by stroke annually, with about 140,000 dying, while survivors are at risk for serious long-term disability [5]. Many stroke survivors suffer physical, cognitive or mental impairment and need continuous support for daily activities. In most cases, this has a direct psychological, social, and economic impact on both the patients and their families. Thus, there is a growing need for research and development to improve prevention of and recovery from stroke and its complications. This review summarizes the mental disorders that are associated with stroke and highlights the mechanisms underlying its pathophysiology in the brain remapping that occurs during recovery.

Changes in neural circuits after stroke

Ischemic damage can affect various components of brain structures-most notably neurons, astrocytes, pericytes, the cerebral microvascular endothelium, and the extracellular matrix of the basal lamina [6,7]. The core region of stroke is not the only part of the brain that undergoes apoptosis and necrosis following stroke. It is often bordered by a region called penumbra, a portion of the ischemic territory that is still potentially salvageable. A penumbra may undergo a process called reperfusion if blood flow is restored with partial recovery of structure, which allows for survival and contribution to synaptic networks. However, neurons can only survive for a limited time and will die if reperfusion does not occur within sevel hours or days [8,9]. In neurons, the deprivation of oxygen and energy during a stroke can lead to signs of structural damage within two minutes of stroke [8]. Multiple short and long-term changes in brain after stroke may result in an imbalance of excitatory and inhibitory neuronal circuits [10]. It is well known that changing in both γ-aminobutyric acid (GABA) and glutamate neurotransmissions, as principle inhibitory and primary excitatory neurotransmitters alter neuronal activities [11]. In cortical areas adjacent to infarcts, long-term potentiation (LTP) induction was thought to be due to a down regulation of GABA receptor activity [12]. Electrophysiological and receptor autoradiography data shows that stroke events can cause a long-lasting decrease in GABA_A receptor GABA_A R binding sites and an increase in NMDA receptor binding sites in the cortex [13,14]. These changes take place both adjacent and contralateral to the infarct, diminishing inhibitory post-synaptic potentials and boosting excitatory post-synaptic potentials [13,14]. Models of recovery from stroke are associated with enhanced glutamate signaling through α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors and downstream induction of brain-derived neurotrophic factor (BDNF); the latter will, in turn, lead to changes in axonal structure [12,15].

Within weeks of a stroke event, inhibition through phasic GABA signaling is reduced and focal brain damage leads to a decreased density of inhibitory interneurons [16,17]. Treatment with GABA_A R agonists at the time of stroke leads to decreased stroke size [18]. However, Clarkson and colleagues indicated that the timing of drug administration is important in treating patients for stroke; tonic inhibition soon after stroke led to an increase in cell death and exacerbated stroke damage, whereas a three-day delay in treatment led to improved functional recovery without altering the size of the area affected by the stroke [19]. In spite of the fact that inhibition of tonic GABA signaling during the repair phase had a beneficial effect on the recovery of neuronal function in mice, an increase in phasic GABA signaling using zolpidem during the repair phase enhanced behavioral recovery [20]. A clear understanding of the interplay between phasic and tonic GABA signaling in modulating stroke recovery requires further investigation.

After stroke, spontaneous recovery occurs through brain remapping. Newly developed brain circuits lead to different behavioral patterns and new response strategies to recover performance [9]. Several animal studies have shown that increases in gene expression are essential for the growth of neurons, synaptogenesis, and growth of dendritic spines following stroke [21-24]. These studies also show that neuroplastic processes observed during increased gene expression are similar to those that occur during development [21-24]. However, the correlation between such intrinsic compensatory structural rearrangement in the mammalian central nervous system (CNS) (spontaneous recovery) and the recovery of the brain function (performance recovery) has not yet been demonstrated [25]. Spontaneous recovery circuit remapping is associated with dramatic plasticity of dendritic spines within the region surrounding the infarct, and with an increase in spine density in some regions [26]. Rewiring of neuronal connections can lead to the activation of several plasticity mechanisms, such as the release of activity-dependent neurotrophins. These neurotrophins, specifically brain-derived neurotrophic factor (BDNF) and nerve growth factor (NGF) have been shown to improve recovery by increasing axonal and dendritic sprouting [27-30]. In 2014, Cheng and colleagues used an optogenetic approach to demonstrate that selectively stimulating neurons in layer V of the cortex increased the expression of activity-dependent neurotrophins, such as BDNF and NGF, in the contralesional cortex. Notably, stimulation of neurons in the ipsilesional primary motor-cortex (iM1) improved the recovery of function [31]. Cyclic adenosine monophosphate (cAMP) response element-binding protein (CREB), a transcription factor for BDNF, also plays a role in recovery after stroke. Caracciolo and colleagues found that virus-mediated over-expression of CREB in the anterior regions of peri-infarct motor neurons enhanced recovery and increased neuronal excitability and plasticity. Interestingly, these effects were reversible, suggesting that motor recovery can be switched off or on, respectively, by blocking or inducing CREB signaling [32].

The process of motor recovery after stroke has also been associated with the recruitment of neighboring areas of the motor and somatosensory cortices [13,14]. In 2009, Takatsuru and colleagues reported that the left somatosensory cortex compensates for loss of function in the right somatosensory cortex by establishing new neuronal circuits and/or remapping the neuronal circuits [33]. In addition, Starkey and colleagues suggested that motor circuits are substantially remapped after focal ischemia, and that the extent of recovery likely correlates with sprouting of neurons from hind limb sensory-motor cortex into the cervical spinal cord sprouting of new connections [34]. A study published in 2010 by van Meer and colleagues found that within a few days of stroke, a significant loss of functional connectivity between the two brain hemispheres was compensated by interhemispheric coupling accompanied by behavioral improvement [35].

Post-stroke mental illnesses

The neuroplasticity observed in the brain after a stroke can either accelerate recovery or lead to unexpected behavioral changes that result in mental illnesses [36]. Such psychiatric disorders in a patient can negatively affect quality of life and lead to further complications including increased risk of stroke recurrence, suicidal tendencies, or interference with stroke recovery [37]. Mental illnesses that are commonly associated with stroke are depression [38], anxiety [39], fatigue [40], sleep disturbances [41], and emotionalism [42].

Depression

Depression has only recently been studied as a significant long-term problem in patients with brain injury, and as a result, post-stroke depression is often under-diagnosed and under-treated. Among stroke survivors, 33% show symptoms of depression, and 40% of patients with depression remain symptomatic for at least one year post-stroke event [44,45]. The major contributing factor to PSD is thought to be mental distress caused by post-stroke physical disabilities [45,46]. Studies comparing the incidence of depression in stroke patients to patients with disabilities not caused by stroke have suggested that post-stroke cerebral lesions may be a significant contributing factor to PSD [45,46]. The dorsolateral prefrontal cortex (DLPFC) is an important node of the cognitive control network (CCN), a neuronal circuit that contributes to the modulation of attention and working memory [47]. Lesions in both the right and left DLPFC are associated with depression, but a recent study showed that the severity of stroke outcomes correlates with the extent of damage to the left side [48]. Egorova and colleagues showed that stroke patients can experience depression, even in the absence of a lesion in the left DLPFC, due to low connectivity between this region and the supramarginal gyrus (SMG) [49]. The SMG is located at the border of the parietal and temporal cortices and is connected to the angular gyrus; together these two structures form the temporoparietal junction. This junction is engaged in memory, social processing, and attention [50] and together with the DLPFC, is termed the ‘rich-club’ [51]. PSD can be associated with the presence of a lesion on either side of the brain, although lesions typically occur in the left hemisphere [52,53]. Several studies have suggested that lesions in the cortico-limbic circuitry and alterations in neural activity and projections between the PFC and the basal ganglia could contribute to depression after stroke [54-56].

The neurobiological mechanisms underlying PSD could potentially be revealed by studying the regulatory molecules of this disease. Currently we know that anti-depressents classified as selective serotonin reuptake inhibitors (SSRIs) have a beneficial effect on the post-ischemic outcome. Stroke patients with a polymorphism in the promoter of the gene encoding the serotonin transporter protein (5-HTTLPR) are at higher risk for developing PSD than others [45,57]. Increases in the expression levels of pro-inflammatory cytokines, such as interleukin 1 beta (IL-1β), IL-6 and tumor-necrosis factor alpha (TNFα) lead to increased leukocyte infiltration and are associated with a reduction of serotonin (5HT) in some brain regions, including the basal ganglia, paralimbic regions, and ventral/lateral frontal cortex [58]. Such increases may result in the onset of depression and breathing disturbances during sleep [59]. Stroke is also associated with inflammation due to tissue damage and induction of the expression and activation of cytokines such as TNFs, ILs, and interferons (IFNs). These molecules affect neural plasticity, neurotransmitter metabolism, and neuroendocrine function and are contribute to PSD [60]. Although stroke is also associated with enhanced glucocorticoid (GC) secretion [61], the relationship between stroke, GC secretion, cytokine activity, and mental illnesses are not well defined.

An adipocyte hormone, leptin, is considered a vascular risk factor for myocardial infarct and stroke [62]. Although its role in the development of depression remains controversial and unclear, some studies have shown an association between high leptin levels and PSD [43,63,64]. Leptin and GC are known to have a reciprocal relationship, acting in a positive feedback loop. In adipose tissue, leptin can activate the HPA axis, leading to GC induction; this in turn causes an increase in the synthesis and secretion of leptin [43]. Leptin activated by inflammatory cytokines also plays a role in immune responses [65]. The well-studied neurotropin BDNF might be another candidate molecule that could explain the complicated relationship between depression, antidepressant and post-stroke recovery, although limited research has specifically studied the role of BDNF in PSD. It has been shown that in stroke survivors with PSD the concentration of serum BDNF decreased within 3-6 months of a stroke event [66]. Moreover, a BDNF Val66Met polymorphism has been suggested to be responsible for the association between stroke and depression [67]. BDNF is likely involved in PSD and other mental illnesses after stroke because it acts as an antidepressant [68] and reduces apoptosis in vitro after glucose deprivation [68].

PSD is related to a range of adverse health consequences such as increased disability and mortality [69], and its symptoms may become worse during the chronic phase [45]. Stroke patients with depression are at high risk for suicide and increased mortality [70-72]. In another systematic review, Wu and colleagues showed that post-stroke fatigue is directly associated with depressive symptoms and directly or indirectly associated with anxiety, poor coping, loss of control, emotional disorders, and behavioral disorders [73].

Anxiety

After a stroke event, patients are at increased risk of developing anxiety. About 25-50% of patients show anxiety during the acute phase of stroke, and young patients and those with a history of anxiety or depression are more likely to develop anxiety after stroke [74,75]. Longitudinal data suggests that post-stroke anxiety (PSA) can last as long as ten years [76]. The following symptoms of anxiety have been reported in patients after stroke: physiological arousal (increased heart rate); avoidance of stress; cognitive disruption; hypersensitivity to possible threatening cues and waiting for adverse events to occur unpredictably; avoidance of crowded places, sexual intercourse, being home alone, going out alone, and traveling on public transport; activities related to fear of having another stroke; and headache [74,77]. In 2018, Chun et al. found that a major contributing factor to anxiety in post-stroke patients is fear of stroke recurrence [74].

At the molecular level several studies have shown that following cerebral ischemia, nuclear factor kappa-light-chain-enhancer of activated B cells (NF-kB) is activated in neurons [78,79], endothelial cells, astrocytes, and microglia [80,81]. However, whether the role of NF-kB is protective or pathogenic remains unclear. One study that supports a detrimental role in cerebral ischemia revealed that in transgenic mice lacking the NF-kB subunit p50, infarct size decreased significantly; this was the case for models of both transient and permanent stroke [79]. However, in other studies NF-kB protected against neuronal death [82]. In addition, another study suggested that the activation of NF-kB in glia may worsen ischemia through NF-kB-dependent activation of microglia, whereas activation in neurons might be important for other processes like memory [83]. A recent study by Zhu et al. showed that hippocampal NF-kB mediates anxiogenic behaviors, likely through enhancing the expression and association of nNOS-CAPON-Dexras [84]. Yeh et al. 2002, showed that activation of NF-B in the amygdala was required for fear conditioning [85], and that in the context of a lack of the NF-kB subunit p50, anxiety-like and fear-like responses were less extreme [86]. Few studies have shed light on the mechanisms underlying post-stroke anxiety, and further research exploring the role of brain networks involved in these mental illnesses is needed.

Fatigue

Fatigue is a common symptom in patients with neurological diseases developed via various biological mechanisms. Examples include systemic lupus erythematous [87], multiple sclerosis [88], Parkinson’s disease [89] and stroke [90]. Post-stroke fatigue has been found to occur in 40-74% of stroke patients [90], yet the pathophysiology remains poorly understood [91]. Fatigue is sometimes evaluated subjectively, based on a patient’s feeling of weariness, early tiredness or unwillingness to exert effort; in other studies it is evaluated objectively, based on a measurable reduction in performance during the repetition of a physical or mental task [92]. Several factors may contribute to post-stroke fatigue, including physical impairment, disuse, sleep disorders, and depression [93,94]. High post-stroke fatigue is associated with low motor cortical excitability in the lesioned hemisphere, suggesting that post-stroke fatigue may be a direct consequence of changes in corticomotor control on the affected side [95,96], although such a correlation remains to be documented. Stroke survivors experience more mental and physical fatigue than the general population, which allows fatigue to be considered a multidimensional phenomenon [97].

Some studies have suggested that pituitary dysfunction (PD) is comorbid with stroke and that the PD contributes to the development of post-stroke fatigue [98]. Recent studies have demonstrated that serum levels of glucose and uric acid (UA) are closely associated with stroke [99,100]. UA is a product of purine metabolism and a neuroprotective antioxidant [101]. Both a low level of serum UA and a high level of serum glucose are associated with increased fatigue severity scale (FSS) scores during the acute stage of stroke [102]. Therefore, a stroke patient may develop fatigue and then disability, which may prevent reestablishment of professional and social activities.

Sleep disorders

The sleep disorders characterized as sleep-disordered breathing (SDB) and sleep-wake disorders (SWDs) can be either risk factors for or symptoms of stroke [103]. Approximately 50%-70% and 10%-50% of stroke patients have SDB and SWD, respectively [41]. SDB refers to habitual snoring, obstructive sleep apnea (OSA), and central sleep apnea (CSA) [41].

SDB is more common in recurrent versus new stroke patients, and CSA is usually linked to injury of central autonomic networks such as those of the insular cortex and the thalamus [104,105]. In a recent study it was observed that wakefulness disorders resulting as a consequence of SDB, including hypersomnia, excessive daytime sleepiness (EDS) and fatigue are common after stroke [106]. In spite of the fact that no link has been discovered between SDB and stroke severity, topography or presumed etiology [107-109], these symptoms have negative effects on rehabilitation and quality of life due to their association with depression, anxiety and cognitive disturbances [110]. Both habitual snoring and OSA are considered independent risk factors for stroke in elderly and middle-aged adults [111-112]. Habitual snoring is strongly associated with stroke and may be a risk factor for ischemic stroke [113]. Acute stroke, especially with involvement of the thalamo-mesencephalic structures, is frequently accompanied by an increase in daytime sleepiness [41]. Therefore, CSA may be a consequence, rather than a cause of stroke and can be a heralding symptom of the vascular disease.

Along with SDB, SWD are also frequently observed in stroke patients and can affect stroke outcomes, yet they have not been investigated in detail [41]. Insomnia, sleep-related movement disorders (e.g. restless legs syndrome, periodic limb movement during sleep), disturbances of wakefulness (e.g. hypersomnia, excessive daytime sleepiness, fatigue), and parasomnias (e.g. REM sleep behavior disorder) are SWDs and are found in 10%-50% of stroke patients [41]. In addition, insomnia was found to be related to brainstem damage and most commonly linked to post-stroke complications [62]. One study showed that 57% of stroke patients suffer insomnia in the first month after the onset of stroke [61]. A population study showed that patients suffering from insomnia have a high mortality rate [63], but the effect of insomnia on mortality after stroke remains unknown. A meta-analysis showed that both short sleep duration and long sleep duration are predictors of ischemic stroke [114]. Pharmacologic studies suggested that the cAMP/protein kinase A (PKA) pathway might be involved in the regulation of wakefulness and rapid eye movement (REM) sleep [115]. Another study showed that the cAMP/MAPK/CREB transcriptional pathway is activated during REM sleep, but not non-rapid eye movement (NREM) sleep [116]. The study also found this pathway may contribute to hippocampus-dependent memory consolidation [116]. Neither the effect of sleep disorders on the occurrence and recurrence of stroke, nor the neurological pathway associated with post-stroke sleep disorders has been fully elucidated. Further evaluation of the treatment of post-stroke sleep disorders to improve stroke rehabilitation and stroke outcomes, including mood and cognitive function, would be beneficial. Findings from such studies are expected to lead to the discovery of novel targets in therapies for this stroke-associated mental illness.

Emotionalism

Emotionalism (emotional lability) refers to an increase in the frequency of crying or laughing in a social situation in response to non-emotive or incongruous stimuli. Emotionalism is another common psychological condition that can hinder stroke recovery. About 20-25% of stroke patients are afflicted with emotionalism during the first 6 months after stroke, and 10-15% of affected patients remain symptomatic after one year [42]. The rate of co-occurrence for depression and emotionalism is 38%, and emotionalism develops more often in patients with PSD than in those without depression [117]. Different terminology has been used in reporting characteristics of emotionalism, for example pathological laughing and crying (PLC), emotional incontinence, and emotional lability. However, in 1989 House and colleagues introduced ’emotionalism’ as a general term for all such emotional disorders, to define the habit of weakly yielding to emotion [118]. This group found that persistent emotionalism is associated with anterior lesions in the left hemisphere, and this association has been reported after unilateral stroke. However, the correlation between lesion location and the emergence of emotionalism remains unclear. Some studies have reported that the basal ganglia, pons, cerebral cortex, and cerebellum are all involved in emotionalism [119-121]. Others have shown that chemical changes in the circuits that interconnect the frontal/temporal lobe, basal ganglia, and ventral brainstem may affect the development of emotionalism [119]. Parvizi and colleagues suggested that emotionalism is caused by dysfunction in circuits that involve the cerebellum and exert an impact on brainstem nuclei and the cerebral cortex itself [120]. Another study suggested that dysfunction of a serotonergic system due to partial destruction of the serotonergic raphe nuclei or projections from these nuclei to the hemispheres might underlie emotionalism [122]. It has been indicated that lesions in the lenticulocapsular region, basis pontis, medulla oblongata or cerebellum are more likely related to emotionalism [120]. Moreover, stroke survivors with paramedian basilar infarct showed pathological laughing [123].

Although the etiology of post-stroke emotionalism is unknown, several hypotheses have been proposed, including lesions and imbalances in serotonergic neurotransmission. Some studies have implicated dysfunction of serotonergic neurotransmission as a cause. Using single-photon emission computerized tomography, it has been shown that serotonin-based neurotransmission is reduced in patients suffering post-stroke emotionalism due to an abnormally low density of serotonintransporter (SERT) in the midbrain and pons [124]. Data from another study in which serotonergic neurotransmission was measured using positron emission tomography (PET) suggested that either neurotransmitter depletion or an increase in affinity of the receptor for the neurotransmitter could cause of an increase in binding potential [125]. The binding potential of the 5-HT1A receptor antagonist in limbic areas and raphe nuclei was high, whereas that in the basal ganglia and cerebellum was negligible, suggesting that serotonergic neurotransmission is abnormally low during the early phase of stroke [125].

Conclusion

Stroke is not only associated with high mortality, but is also a risk factor for multiple mental illnesses that significantly hinder rehabilitation of stroke survivors. Most recent studies have suggested that the mechanisms responsible for post-stroke illness might be centralized, affecting mainly the remodeled emotional neuronal circuits after a stroke. However, many questions remain unanswered, including how the post-stroke neuronal circuits are rewired, as well as to what extent and how the infarction area influences remote areas of damage, and thus how behavior is enhanced or inhibited. These unanswered questions open doors for research, advances, and improved treatment and outcomes for stroke survivors in the future.

Disclosure of conflict of interest

None.