Brain–body communication in stroke

Cardiac dysfunction leads to stroke

Atrial fibrillation (AF) is a common arrhythmia and one of the main risk factors for embolic stroke. In the atrial endocardium, hypercoagulability, flow abnormalities, and endothelial changes must coexist to induce thrombogenesis (Goette et al., 2016). Blot clots formed in the left atrium may travel further with the flow and occlude a blood vessel in the brain. Due to its high dependence on blood supply, already a short period of supply shortage has detrimental consequences to motor, sensory, and cognitive function. AF can also lead to reduced cardiac output as well as transient cerebral hypoperfusion and hypertension involved in cognitive impairment and brain damage (Anselmino et al., 2016).

Hypertrophic cardiomyopathy causes the heart muscle to enlarge and may occur together with AF. In an observational study, patients with hypertrophic cardiomyopathy had a 1% yearly risk for embolic events including stroke, of which half of the patients were not previously documented to have AF (Haruki et al., 2016). Although the prevalence of AF in this population may vary due to ethnic/racial differences and under-recognition of AF, this indicates that hypertrophic cardiomyopathy is an independent cardiac risk factor for stroke.

Atrial cardiomyopathy has been suggested as another source for embolic stroke. It refers to abnormal atrial substrate and function, such as chamber dilation, impaired myocyte function, and fibrosis. Atrial cardiomyopathy can occur in the presence or absence of AF and may explain some of the embolic strokes without a history of AF (Ning et al., 2021). However, the molecular mechanisms of how cardiomyopathy leads to stroke remain to be elucidated.

Oral anticoagulants are commonly used to treat AF to prevent embolic stroke. However, the anticoagulant use is associated with a 7–10-fold increased risk for hemorrhagic stroke (Morotti and Goldstein, 2020). Compared to the vitamin K antagonist warfarin, direct oral anticoagulants confer a significantly lower risk in patients (Bai et al., 2017). This is in line with a preclinical study demonstrating that warfarin promoted deadly ICH in mice with induced microbleeds, whereas direct oral anticoagulants augmented the number of microbleeds without inducing long-term cognitive impairment (Pétrault et al., 2019).

Stroke leads to cardiac dysfunction

Cardiac complications, including acute coronary syndrome, heart failure, and cardiac arrhythmia, frequently occur after ischemic and hemorrhagic stroke (Chen et al., 2017; Lee et al., 2016; Scheitz et al., 2018).

Acute autonomic imbalance has been proposed as a trigger for newly occurring AF after stroke (Paquet et al., 2018). It may be induced due to damage to specific brain regions including the insular cortex that regulates the autonomic control of the heart rhythm or due to inflammatory mechanisms that result in abnormal autonomic responses. In contrast to ischemic stroke, where sympathetic predominance during the first three days and reduced HRV as a risk factor for short-term mortality and cardiac death are well-established (Yperzeele et al., 2015), a shift to parasympathetic predominance and higher HRV within the first 24 h related to the poor three month outcome has been observed in acute ICH (Rass et al., 2021; Szabo et al., 2018). This may be related to increased intracranial pressure that occurs in critically ill patients with large ICH; patients with traumatic brain injury and increased intracranial pressure were also reported to have higher HRV (Szabo et al., 2018).

Another cardiac dysfunction, commonly associated with SAH, which can also develop after ischemic stroke, is Takotsubo cardiomyopathy or Takotsubo syndrome. This syndrome presents as a transient dysfunction of the left heart ventricle and occurs in the absence of obstructive coronary artery disease. Cardiac dysfunction typically recovers spontaneously within days to weeks (Baker et al., 2021). Sympathetic overactivity due to physical or emotional stress (the latter has also led to the term “broken-heart syndrome”) results in a supraphysiological release of catecholamines that directly damages myocytes by inducing a shift from the positive inotropic to a negative inotropic response (Baker et al., 2021).

In mice, ICH induced progressive cardiac dysfunction associated with systemic and cardiac inflammation as well as oxidative stress (Li et al., 2018; Zhang et al., 2021), which were abrogated by splenectomy that also improved neurological outcome (Li et al., 2020). The authors further showed that deficiency in small ubiquitin-like modifier 1 (SUMO1) exacerbates cardiac and neurological dysfunction (Li et al., 2021).

Further, elucidating the molecular mechanisms underlying poststroke cardiac vulnerability and cardiac dysfunction leading to stroke is of major importance to identify therapeutic targets to successfully protect the heart and the brain.

Local and peripheral immune cells are involved in poststroke neuroinflammation

In the CNS, an abrupt drop in perfusion leads to the depolarization of neuronal cells and glia, glutamate release, and triggering of the ischemic cascade involving excitotoxicity, peri-infarct depolarization, and cell death (Dirnagl et al., 1999). Resident phagocytes, microglia, become activated before neuronal death, upregulating proinflammatory genes. Also, astrocytes, local mast cells, and brain-resident macrophages react to the injury contributing to the neuroinflammatory response and disruption of the blood–brain barrier (BBB) facilitating the infiltration of peripheral immune cells (Iadecola et al., 2020).

Neutrophils are the first peripheral cells arriving to the brain and advancing local inflammation and BBB breakdown by secreting proteases, interleukin (IL)-1β, and neutrophil extracellular traps. Neutrophils are followed by monocyte-derived macrophages, natural killer (NK) cells, T cells activated in antigen-dependent and independent processes, and B cells (Iadecola et al., 2020). The kinetics of the immune response and the infiltration of peripheral immune cells have been relatively well-characterized in rodent models of stroke as well as human patients (Beuker et al., 2021). Notably, there are still many gaps in the detailed understanding of the impact of certain cell subpopulations and their roles for the evolution of the lesion and outcome.

Specific immune cells can have distinct (beneficial/detrimental) roles depending on the subpopulation and time after stroke (acute/chronic phase). For example, B regulatory cells are linked to favorable outcomes in experimental stroke settings due to their anti-inflammatory phenotype and ability to attract T regulatory cells to the lesion site (Seifert et al., 2018). On the other hand, B cells and plasma cells have been proposed to contribute to the development of poststroke cognitive impairment by producing antibodies against CNS antigens in the chronic phase of stroke (Doyle et al., 2015). Also, T cells secreting IL-17 (Th17 cells, ɣδT cells) have been linked to detrimental outcome after ischemic stroke, whereas T regs are generally associated with beneficial prognosis (Cramer et al., 2019).

CNS-resident and peripheral immune cells are also important players in the pathophysiological sequelae of hemorrhagic stroke; the detailed mechanisms are, however, less-characterized than in ischemic stroke (Shao et al., 2019).

Brain–body signaling leads to poststroke immunosuppression

Simultaneously, it is important to highlight the role of the brain–body communication axes and the fact that, after brain lesion, the immune response differs between the injured CNS and the rest of the body. In ischemic stroke, the initial systemic inflammatory boost is followed by a suppression of immune functions in the periphery (Meisel et al., 2005). It has been proposed that systemic dampening of the immune response poses, in fact, a protective mechanism to limit the infiltration of peripheral immune cells to the brain (Dirnagl et al., 2007).

Systemic immunosuppression after the insult leads, however, to increased susceptibility to infections that occur in ∼30% of ischemic stroke patients. Specifically, pneumonia has been linked to increased mortality and worsening of neurological outcome (Westendorp et al., 2011). Immunosuppression mediated by the PNS (over the vagus nerve) is mostly connected to the actions of acetylcholine on macrophages upon binding to the ɑ7 nicotinic receptor, decreasing the production of proinflammatory cytokines such as tumor necrosis factor (TNF) and IL-1 (Tracey, 2002). Stimulating the vagus nerve is known to have anti-inflammatory effects, but interestingly, as shown in experimental models, acetylcholine is not derived directly from the vagal fibers. The vagus nerve provides input to the celiac ganglion, the origin of the catecholaminergic splenic nerve. It has been therefore proposed that T cells expressing β2-adrenergic receptors are the producers of acetylcholine reaching the macrophages (Rosas-Ballina and Tracey, 2009).

Concurrently, the SNS can suppress immune function over direct actions of noradrenaline leading to a diminished production of proinflammatory cytokines in cells of the innate and adaptive immune system (Sharma and Farrar, 2020), including invariant NKT cells in the liver, where a decreased intravascular crawling of these cells and increased production of anti-inflammatory IL-10 and IL-5 were shown as effects of increased noradrenergic signaling (Wong et al., 2011). Newly identified indirect effects of SNS activation on immune cells, not explored in experimental stroke settings yet, comprise the inhibition of leukocyte migration via limiting local tissue blood flow leading to decreased leukocyte movement (Devi et al., 2021).

Finally, glucocorticoids as mediators of the HPA axis have been known for their anti-inflammatory actions, decreasing cytokine production, expression of adhesion molecules and nuclear factor kappa B (NF-κB), and induction of apoptosis in cells of innate and adaptive immunity (Cain and Cidlowski, 2017).

A further detailed characterization of regulatory mechanisms in poststroke peripheral immune responses is of great importance for both systemic complications after stroke and pathophysiological events in the CNS, since immune cells from the periphery will be recruited to the site of the lesion.

Stroke is associated with gut dysbiosis

Gut dysbiosis refers to a detrimental imbalance in bacterial composition, metabolic activity, and distribution in the gut. It can result from (1) a diminished representation of beneficial bacteria, (2) overgrowth of “pathobionts” – potentially pathogenic bacteria, and (3) decreased bacterial diversity (DeGruttola et al., 2016).

Gut dysbiosis has been linked to several risk factors for stroke such as hypertension, diabetes, atherosclerosis, aging, vascular dysfunction, and obesity (Durgan et al., 2019) and was reported after stroke (Holmes et al., 2020; Lee et al., 2021). It has the potential to impact ischemic stroke outcome, e.g., through the fine-tuning of the immune response (Benakis et al., 2016; Singh et al., 2016). Additionally, other mechanisms such as increased intestinal permeability and potential bacterial translocation have been proposed as modulators of stroke prognosis (Durgan et al., 2019).

Clinical cohort studies suggested that ischemic stroke triggers dysbiosis due to the overgrowth of pathogenic bacteria that reciprocally exacerbates brain damage (Xu et al., 2021). In another study in patients exhibiting three or more risk factors for stroke without prior history of stroke, opportunistic pathogenic bacteria were reported to be enriched, whereas beneficial bacteria were depleted (Zeng et al., 2019). Hypertensive patients also demonstrated reduced levels of beneficial bacteria, whereas the levels of pathogenic bacteria were increased (Avery et al., 2021). Rodent models of hypertension generally replicate these findings. However, data are still controversial when it comes to the ideal microbial makeup, especially concerning some phyla (i.e., Bacteroidetes) (Marques et al., 2017).

Despite some insight into gut dysbiosis in ischemic stroke, little is known about hemorrhagic stroke. In a mouse model of ICH, alterations of microbial diversity and gut dysbiosis were related to reduced intestinal motility and increased gut permeability. Interestingly, recolonizing ICH mice with healthy microbiota ameliorated neuroinflammation and functional deficits after ICH (Yu et al., 2021).

Microbial metabolites modulate stroke risk and outcome

GM metabolizes phosphatidylcholine from dietary sources (e.g., eggs and red meat), leading to the formation of trimethylamine and finally trimethylamine N-oxide (TMAO), which was associated with an increased risk of major adverse cardiovascular events, carotid artery stenting, and first stroke (Lee et al., 2021). A recent systematic review proposed that inflammation and aging simultaneously contribute to changes in GM composition, and ischemic stroke predisposition was linked to significantly higher TMAO levels and a reduction of SCFA-producing bacteria (Lee et al., 2021). In addition, other GM-derived metabolites such as phenylacetylglutamine enhanced thrombosis risk through the activation of adrenergic receptors in platelets (Nemet et al., 2020).

SCFAs are one of the main bioactive microbial metabolites that constitute major products from bacterial fermentation of dietary fiber in the intestine (mainly acetate, propionate, and butyrate), favoring usually beneficial actions in the host. Propionate showed cardioprotective effects through attenuating systemic inflammation and abrogating cardiac hypertrophy, fibrosis, vascular dysfunction, and hypertension (Bartolomaeus et al., 2019). GM modulation and acetate supplementation protected against obstructive sleep apnea-induced gut inflammation and hypertension and mitigated blood pressure increases, cardiac fibrosis, and left ventricular hypertrophy (Marques et al., 2017).

In acute ischemic stroke patients, SCFA levels were inversely correlated with stroke severity and prognosis, suggesting SCFAs as potential prognostic markers and therapeutic targets (Tan et al., 2021). In aged mice subjected to ischemic stroke, fecal transplantation of SCFA producers attenuated inflammation and neurological deficits, increasing the concentration of SCFAs in the gut, plasma, and brain (Lee et al., 2020). Moreover, acetate, propionate, and butyrate supplementation in drinking water before induction of stroke in the photothrombotic mouse model facilitated lymphocyte recruitment to the infarcted area, resulting in microglial activation and neuronal plasticity as well as improved limb motor function recovery (Sadler et al., 2020). Among the most abundant SCFAs, levels of butyrate were most strongly negatively correlated with neurological outcome and infarct size in a rat model of ischemic stroke. Its supplementation effectively remodeled GM and repaired the leaky gut, not to mention that transplanting SCFA-enriched fecal microbiota also reduced neurological deficits, brain edema, and infarct sizes (Chen et al., 2019).

The described roles of gut commensals and their connection to the brain function offer a new and exciting area of research potentially providing more options in search for therapeutic targets (Durgan et al., 2019).

Protecting the brain is protecting the heart

Tremendous effort has helped to improve anticoagulants for the treatment of AF and to reduce the risk of brain hemorrhage. When it comes to poststroke cardiac dysfunction, alleviating autonomic imbalance will be key to protect the heart. This may be achieved by decreasing sympathetic overactivation inhibiting beta-adrenergic receptors or the renin–angiotensin–aldosterone system. However, pharmacological treatments remain insufficient. Recently, cardiac neuromodulation of various types to restore sympathovagal balance has emerged. Preclinical studies have shown promising results of bioelectric therapies for cardiac diseases, but more clinical studies are needed to identify optimal stimulation parameters and location (Hanna et al., 2018). Whereas neuromodulation of the brain is currently investigated for poststroke motor recovery, its effect on the heart is an outstanding area of research. An alternative approach is to improve heart function and ameliorate autonomic dysfunction of the heart by exercise (Katz-Leurer and Shochina, 2007). Thus, clinical trials on neurorehabilitation should focus on identifying the most beneficial exercise regime that also targets heart health.

Fine-tuning the immune response to impact the poststroke outcome

Most drugs already tested in clinical trials and directed against the poststroke neuroinflammatory response interfered with the entry of peripheral immune cells to the brain or inhibited proinflammatory signaling (Iadecola et al., 2020). When it comes to limiting poststroke pneumonia, preventive antibiotic treatment has not been established as a successful approach (Meisel and Smith, 2015). Also, more studies would be necessary in the field of immunostimulatory therapies to boost peripheral immune responses or block stroke-induced immunosuppression (Faura et al., 2021). Selected agents (beta blockers) modulating the poststroke brain–body signaling have also been tested in clinical studies; however, current data do not show significant benefits in acute ischemic stroke (Balla et al., 2021). Therefore, more research on neuroimmune interactions after stroke is urgently needed to pinpoint potential therapeutic targets.

You are what you eat!

There are several options to target the gut–brain axis and the GM for therapy in stroke: (1) Fecal microbiota transplantation, whereby the microbiota of a healthy donor is transferred to a patient, can be used to increase microbial diversity and the percentage of beneficial microbiota in the gut. Fecal microbiota transplants are currently investigated in a clinical trial for Alzheimer disease and have been successful in preclinical models of stroke (Holmes et al., 2020). (2) Probiotics (live and beneficial bacteria) and prebiotics (nondigestible dietary substances to foster the growth of favorable microorganisms) have been demonstrated to decrease gut inflammation as well as improve cognitive performance in middle-aged and older adults with cognitive impairment (Sanborn et al., 2020). Data on the beneficial effects of probiotics and prebiotics are currently limited to preclinical studies; however, a clinical trial evaluating probiotics for ischemic stroke patients is under way (Holmes et al., 2020). (3) Dietary interventions aim to foster the growth of beneficial bacteria already present in the gut. Whereas certain diets are known to promote healthy aging including cognitive function and have been shown to have significant effects on microbial diversity (Holmes et al., 2020), data on stroke are limited. Overall, further clinical trials are needed to assess the benefit of targeting the GM and gut–brain signaling for stroke therapy.

An ounce of prevention is worth a pound of cure

Whereas most therapeutic targets are developed to ameliorate poststroke outcomes, we also need to better understand how cardiac dysfunction, immune imbalance, and gut dysbiosis lead to CNS vulnerability to stroke. Hence, model systems addressing body–brain communication imbalances should be developed to study these axes and to develop further preventative treatments.