Introduction

In 2007, McEwen (McEwen 2007) suggested the existence of bidirectional communication between the brain and the cardiovascular, immunological, and other systems, in which neurological and endocrine processes are key players. Since the brain is a primary regulator of the neuroendocrine, autonomic, immunological, and behavioral systems (McEwen 2006), peripheral organ stimulation may also have direct and indirect effects on brain function. However, little is known of the bidirectional communication between the brain and other organs. Brain factors are involved in the regulation of other organs throughout the body, and many have yet to be identified (Wang et al., 2022a). Meanwhile, substances produced or released primarily by organs such as the heart, bone marrow, or adipose tissue may affect normal or abnormal brain function and metabolism (Letra & Santana 2017; Rigueira et al. 2021). A typical example of a myokine expressed simultaneously in cardiac muscle and brain is irisin, a protective factor with antioxidant, anti-inflammatory, and anti-apoptotic effects (Li et al. 2021b). It has been suggested that increased expression of soluble epoxide hydrolases in the brain and liver may play a role in the pathogenesis of major psychiatric disorders, further indicating the importance of cross-talk between brain and liver in major psychiatric disorders (Zhang et al. 2020). Redistribution of local perfusion and pulmonary intravascular microthrombus might cause the increase of alveolar dead space and depletion of pulmonary surfactants due to excessive stimulation of sympathetic never, which may completely disrupt the regulatory mechanisms of ventilation and perfusion balance and cause hypoxemia after TBI (Pelosi et al. 2005). Qualitative differences in gut bacterial composition are linked to mood and behavioral impairment, and exposure to particular non-pathogenic bacteria can modulate brain chemistry and behavior in adult animal (Bravo et al. 2011) and change depression and anxiety-like symptoms in human subjects (Nikolova et al. 2019; Bajaj et al. 2019).

To further increase the complexity of this interaction network, studies published in the last decade have identified a large number of intra- and inter-organ signals mediated by EVs. Almost all types of cells, including those found in brain tissue, have the ability to produce EVs (Gabrielli et al. 2022; Karnati et al. 2019; Ruan et al. 2021). EV-mediated intercellular communication in the brain has been reviewed. The neuron-to-glial cell ratio is 1:10; thus, the former must communicate with neighboring glial cells. This continuous communication ensures homeostasis within the brain. EVs have a significant impact on the central nervous system (CNS) of the brain's information flow, which is a complicated, highly regulated process. Moreover, EVs could be produced, released, and absorbed by neurons (Durur et al. 2022; Huang et al. 2018a; Nogueras-Ortiz et al. 2020) and glial cells (Bakhti et al. 2011; Datta et al. 2020; Li et al. 2021c; Lombardi et al. 2019; Pei et al. 2019; Wang et al. 2011) under both normal and pathological conditions. Substantial numbers of proteins and nucleic acids are present in EVs, and when these materials are transferred to other cells, they may have an impact on biological processes. The cargo of EVs is determined during their biogenesis and by the stimuli affecting the parental cells (Jiang et al. 2022). For example, hypoxic stimulation of neurons and glial cells can cause changes in the size, number, and expression of EVs cargo (Jiang et al. 2022).

Recently, there have been a growing amount of studies on the intercommunication between brain and other organs because EVs can function across the BBB, but there is still no complete review article that extensively discusses the body of knowledge on signals between the brain and other organs carried by EVs. Here, we review the role of EVs in communication between the brain and other organs. Their role has been clearly demonstrated in a few cases; however, in most cases, it has been based on indirect evidence from EVs in cell culture or animal models.

Extracellular Vesicles: Biogenesis and Function

A highly heterogeneous membranous structure that originates from different parental cells, Evs can be classified into three subtypes based on biogenesis, size, composition, and load: exosomes, microvesicles (MVs), and apoptotic vesicles (van Niel et al. 2018; Zhang et al. 2021a, b). Despite trying to further classify EVs based on their size, origin or function, the primary overlap of these characteristics does not clearly distinguish them. Thus, the consensus recommendation of the International Society for Extracellular Vesicles (ISEV) is to use “extracellular vesicles” as a “general term for particles naturally released from cells separated by lipid bilayers that cannot be replicated, i.e., do not contain a functional nucleus (Witwer & Thery 2019)” and we will use this term in this review; nevertheless, in some cases, we will cite the same terminology used by the authors in their article.

The size and genesis of EVs vary widely. EVs range in size from 30 to 5000 nm (Karpman et al. 2017), with exosome diameters ranging from 30 to 150 nm (Rai et al. 2021; Yang et al. 2020) and MVs ranging from 100 to 1000 nm (Kailashiya 2018; Laberge et al. 2018). Apoptotic bodies generally have a diameter of up to 5000 nm (Atkin-Smith et al. 2015). Exosomes are intraluminal vesicles that are intermediates in the endosomal system that are released when multivesicular bodies fuse with the cell surface. They are created by the inward budding of endosomal membranes during the maturation of multivesicular bodies (van Niel et al. 2018; Villarroya-Beltri et al. 2016). In order to release exosomes, several cellular steps need to be accomplished: first, a cup-shaped structure composed of cell surface proteins and hydrophilic proteins forms the first invagination. This leads to the generation of early endosomes (Lloyd et al. 2002). Second, the endosomal limiting membrane invaginates inward to produce multivesicular bodies containing intraluminal vesicles, i.e., the future exosomes (Heijnen et al. 1999). The process ends with degradation of the multivesicular contents by lysosomes or autophagosomes or fusion with the plasma membrane and release of intraluminal vesicles as exosomes (van Niel et al. 2018; Villarroya-Beltri et al. 2016). Exosome secretion has now spread to a wider variety of cell types, and its importance in intercellular communication in both healthy and pathological conditions is well recognized (Colombo et al. 2014) (Fig. 1A).

Fig. 1
figure 1

Schematic diagram of the process of EVs production. A To release exosomes, a cup-shaped structure composed of cell surface proteins and hydrophilic proteins forms the first invagination. This leads to the generation of early endosomes. Second, the endosomal limiting membrane invaginates inward to produce multivesicular bodies containing intraluminal vesicles. When the multivesicular bodies fuse with the plasma membrane and empty their contents, intraluminal vesicles are released and are termed exosomes once they are extracellulars. B Microvesicle formation is calcium-dependent and associated with loss of membrane asymmetry and disruption of the cellular cytoskeleton. C Apoptosis undergoes several stages, beginning with condensation of nuclear chromatin, followed by membrane blistering, and progressing to disintegration of cell contents into distinct membrane-encapsulated vesicles called apoptotic vesicles or apoptotic vesicles. EVs extracellular vesicles; MVs microvesicles

Full size image

Different from exosomes, MV formation and release are the result of a complex process of cytoskeletal reorganization and loss of physiological asymmetry in the membrane bilayer (Hugel et al. 2005). Those MVs were produced by the outward budding and fission of the plasma membrane, followed by the release of vesicles into the extracellular space. The process depends on an increase in cytosolic Ca2+, which activates enzymes, leading to depolymerization of the actin cytoskeletal depolymerization and outward bending of the cell membrane (Comfurius et al. 1990; Kelton et al. 1992; Piccin et al. 2007). When cells are activated, the increased cytosolic calcium activates flippase (which allows lipids to move toward the outer membrane) and scramblase (which moves lipids in both directions), and flippase (which allows lipids to move toward the inner membrane) is inactivated, causing negatively charged phosphatidylserine to flip to the outer leaflet of the phospholipid bilayer. This facilitates membrane budding and the formation of microvesicles (Turturici et al. 2014). The budding process is accomplished by contracting the cytoskeletal structure through actin–myosin interactions (McConnell et al. 2009; Muralidharan-Chari et al. 2009) (Fig. 1B).

Cellular apoptosis consists of these major stages, beginning with condensation of nuclear chromatin, next, membrane blistering and disintegration of cell contents into apoptotic bodies (Kerr et al. 1972) (Fig. 1C). Although exosomes and MVs are secreted during regular cellular processes, apoptotic bodies are exclusively produced during planned cell death. By specifically interacting with the recognition receptors on the phagocytes and the particular alterations in the structure of the apoptotic cell membrane, macrophages can remove apoptotic substances by phagocytosis. The most characteristic of these changes is the translocation of phosphatidylserines to the outer leaflets of the lipid layer. These translocated phosphatidylserines bind to membrane associate protein V recognized by phagocytes (Ishibe et al. 1975). Another well-characterized membrane change involves the oxidation of surface molecules. These changes create sites for the binding of thrombospondin (Savill 1997; Friedl et al. 2002) or complement protein C3b (Takizawa et al. 1996). As a result, C3b and coagulation-responsive proteins are successively acknowledged as three well-known indicators of apoptotic bodies. However, little is known of their function in cell–cell interactions (Battistelli and Falcieri 2020; Elmore 2007; Erwig and Henson 2008).

EVs are important for transmitting information in lipids, nucleic acids, proteins (polypeptides, enzymes, cytokines, etc.), and so on (van Niel et al. 2018). Recently, EV proteins and nucleic acids as potential biomarkers or therapeutic targets for diseases have been attracting extensive attention (Beard et al. 2021; Forro et al. 2021; Karnati et al. 2019; Ko et al. 2020), but EV lipids have received little attention (Su et al. 2021). Overall, EV composition depends mainly on the condition of the parental cell and can change according to physiological or pathological conditions (e.g., inflammation and stress) (Guedes et al. 2020; Valadi et al. 2007).

EV-mediated Bidirectional Communication Between the Brain and Other Organs

The bidirectional communication between the brain and other organs is thought to be mediated by EVs. EVs can cross the BBB to transport lipids, peptides, proteins, and nucleic acids and can reach and target almost any organ (Banks et al. 2020a). Determining the source of EVs is challenging. As far as I know, brain-specific EVs have not yet been discovered so far. However, as we have discussed before, the levels and carriers of EVs are different under different physiopathological conditions. Because EV content depends on the state of the parent cell, it may be easier to understand its origin under pathological conditions. Peraza et al. (Delgado-Peraza et al. 2021) found that Astrocytic EVs showed significant amounts of complement, which is linked to Alzheimer's disease (AD), and circulating neuronal EVs had high quantities of tau and β-amyloid protein (Aβ), suggesting their potential as liquid biopsies for neurological illnesses. Nevertheless, it is possible to identify the origin of EVs under physiological conditions based on the characteristics of cell-specific protein cargo, which is also a direction for future development. Of these, the contents elaborated below are summarized in Table 1 and Fig. 2.

Table 1 EVs-mediated communication between brain and other organs
Full size table
Fig. 2
figure 2

Extracellular vesicle-carried substances mediate communication between the brain and other organs. BMSC bone marrow mesenchymal stem cell; TF Tissue factor, ASC apoptosis-associated speckled protein; ApoE4 Apolipoprotein E4, PDEF pigment epithelium-derived factor, LPS Lipopolysaccharide; NEP neprilysin; GABA gamma-Aminobutyric acid; exRNA extracellular RNA; MHC-II major histocompatibility complex II

Full size image

EVs in Brain–Bone Marrow Crosstalk

The effectiveness of bone marrow mesenchymal stem cells (BMSCs) in the treatment of central nervous system diseases has been a closely investigated and controversial issue. Transplanted BMSCs can promote neural repair by assuming a functional neuronal and glial cell phenotype (Li and Zhang et al. 2021a, b; Maltman et al. 2011). However, mesenchymal stem cells may function primarily in a paracrine manner rather than intracellularly (Phinney & Pittenger 2017). BMSC-derived exosomes (BMSC-exos) have therapeutic potential because of their stable biology, low immunogenicity, and ability to cross the BBB and localize to the injured area (Phinney and Pittenger 2017). These findings promote the study of exosomes released from BMSCs and their protective and repairing effects in the brain. Liu et al. (Liu et al. 2021) reported a reduction in brain infarct size and alleviation of short-term neurobehavioral deficits in rat by BMSC-exos-mediated switching of microglia between the M1 and M2 phenotypes. Hypoxic BMSC-exos rescue oxygen glucose deprivation (OGD)-induced neuronal damage by inhibiting nod-like receptor family pyrin domain-containing 3 (NLRP3)-containing inflammasome-mediated pyroptosis in vitro (Kang et al. 2021). Based on their ability to transmit information, the functions of non-coding RNA in exosomes have been investigated (Zhang et al. 2015). BMSC-exos enriched in miR-124 reduce proinflammatory microglia activation, protect BBB integrity, and ameliorate stroke-mediated injury, possibly in a manner involving peroxiredoxin 1 (PRDX1) (Tian et al. 2022b). MiR-124-3p is downregulated in rats with hypoxic–ischemic brain damage (HIBD), but its upregulation by BMSC-exos improves neurological function, attenuates neuronal pathology and structural damage to neurons, inhibits oxidative stress, and lowers neuronal apoptosis in HIBD rats by inhibiting tumor necrosis factor receptor-associated factor 6 (TRAF6) (Min et al. 2022). Lipocalin 2 (LCN2) is expressed at high levels and is also a target gene for miR-138-5p. MiR-138-5p derived from BMSC-exos attenuates neurological impairment by targeting LCN2 to promote astrocyte proliferation and suppress inflammatory responses after ischemic stroke (Deng et al. 2019). In addition, BMSC-exos ameliorate cerebral infarction by transferring miR-23a-3p, thereby inducing microglia inactivation and M2 polarization (Dong et al. 2022). By targeting the histone deacetylase 4 (HDAC4)/B-cell lymphoma 2 (Bcl-2) axis to prevent hippocampus neuronal death, miR-93 delivered by BMSC-derived EVs (BMSC-EVs) improves HIBD (Shi et al. 2022). BMSC-exos carrying the long non-coding RNA ZFAS1 suppress oxidative stress and inflammation associated with ischemic stroke (Yang and Chen 2022). MiR-410 delivered by BMSC-EVs targets HDAC4 and prevents brain damage in hypoxic–ischemic mice (Shen et al. 2022). The downregulation of phosphatase and tensin homologs, which led to the activation of the PI3K/protein kinase B/rapamycin mechanistic target/glycogen synthase kinase-3 beta signaling pathway, may be partially responsible for the improvement in neural plasticity and recovery after stroke after intravenous administration of miR-17-92 cluster-enriched BMSC-exos. (Xin et al. 2017). In addition, another study found that intravenous delivery of miR-17-92 cluster-enriched BMSC exosomes after traumatic brain injury (TBI) dramatically decreased neuroinflammation and neuronal cell death and boosted neurogenesis and angiogenesis (Zhang et al. 2021b). The BMSC-exos hsa-miR-23b-3p inhibits intracranial aneurysm formation by targeting krueppel-like factor 5 (KLF5) via inhibition of the PI3k/AKT/NF-κB signaling pathway (Sun et al. 2020). BMSC-EVs have therapeutic potential for brain dysfunction and exert a neuroprotective effect after subarachnoid hemorrhage (SAH) (Han et al. 2021). The development of SAH is related to endothelial cell damage and BBB breakdown. BMSC-EVs can deliver non-coding RNA KLF3-AS1 to brain microvascular endothelial cells, thereby attenuating neurological deficits after SAH by promoting cell viability and inhibiting apoptosis (Cheng et al. 2022). Xiong et al. (Xiong et al. 2020) also showed that miR-129-5p from BMSC-exos ameliorates early brain injury after SAH by inhibiting the anti-apoptotic and anti-inflammatory effects of the high mobility group box 1 (HMGB1)/toll-like receptor 4 (TLR4) pathway.

BMSC-EVs are also important in chronic degenerative diseases. Many chronic neurodegenerative diseases, including multiple sclerosis and AD, are associated with demyelination. Demyelination is caused by damaged myelin sheaths, and additional damage to demyelinated axons can be caused by inflammatory responses, oxidative stress, mitochondrial dysfunction, ion channel dysfunction, and excessive excitatory neurotransmitter release (like glutamate) (Dendrou et al. 2015; Schattling et al. 2019; You et al. 2019). EVs enhance myelin maintenance by minimizing oligodendrocyte DNA damage and boosting the expression of myelin-related gene expression and raising the proportion of actively myelinating oligodendrocytes in aged rhesus monkeys (Go et al. 2021). Also, BMSC-exos promote remyelination by acting directly on oligodendrocyte progenitor cells and indirectly on microglia in the demyelinated CNS (Go et al. 2021). Lateral ventricular administration of BMSC-exos improves AD-like behavior in mice by modulating hippocampal glial activation and its associated neuroinflammatory and brain-derived neurotrophic factor (BDNF)-related neuropathological changes (Liu et al. 2022). BMSC-EVs have beneficial effects in AD mouse models, possibly via inhibition of Aβ-induced inducible nitric oxide synthase (iNOS) expression (Elia et al. 2019). BMSC-EVs exert anti-inflammatory effects by modulating cytokine release (Cui et al. 2018; de Godoy et al. 2018), inhibiting iNOS, or preventing Aβ-oligomer toxicity (Wang et al. 2018).

Meanwhile, Cui et al. (Cui et al. 2018) reported that exosomes from hypoxia-preconditioned BMSCs increase the miR-21 level in the brain of AD mice. Furthermore, miR-21 supplementation reversed cognitive impairments and avoided degenerative characteristics in APP/PS1 mice. Sha et al. (Sha et al. 2021) further highlighted that the Wnt/β-catenin pathway is crucial, and that BMSCs-EVs deliver miR-29c-3p to hippocampal neurons to reduce Aβ and inflammatory cytokines to improve AD.

Chronic inflammatory activation in the blood and brain is increased after TBI, including mobilization of bone marrow-derived immune cells (Braun et al. 2018, 2017; Ma et al. 2017; Wang et al. 2017). Therefore, TBI may create a chronic proinflammatory environment in bone marrow, leading to progressive bone loss. Interestingly, EVs may mediate signaling from the brain to bone marrow. Bone marrow-derived EVs from TBI mice increase osteoclast differentiation and colony formation (Singleton et al. 2019). MiR-1224 may mediate TBI-dependent NF-κB activation and bone marrow osteoblast differentiation (Singleton et al. 2019). However, further studies are needed to assess whether EVs in bone marrow after TBI are released locally or transported from the site of injury.

EVs in Brain–Adipose Tissue Crosstalk

The largest endocrine organ, adipose tissue produces a variety of biologically active mediators that work with peripheral organs and the CNS (Hajer et al. 2008; Scherer 2019). Adipokines released from adipose tissue modulate neuroinflammation and oxidative stress, two major pathophysiological and physiological processes in the CNS, and are linked to central nervous system diseases. A significant source of circulating non-coding RNAs, many of which are transported by exosomes, is adipose tissue (Thomou et al. 2017). Thomou et al. (Thomou et al. 2017) suggested that adipose tissue is the source of the majority of circulating exosomes. There is proof that exosomes can mediate communication between the brain and adipose tissue. Otero-Ortega et al. claim that axonal sprouting and development, oligodendrocyte creation, tract connection, and remyelination after subcortical ischemic stroke are all mediated by exosomes of adipose origin (Otero-Ortega et al. 2017). A proteomic analysis of EVs identified more than 2000 proteins associated with brain repair, which if used therapeutically, may improve functional recovery (Otero-Ortega et al. 2017). Adipose tissue contains adipose-derived stem cells (ADSCs), an important source of EVs. Geng et al. (Geng et al. 2019) demonstrated that miR-126-rich ADSC-derived exosomes (ADSC-exos) alleviate ischemic stroke by inhibiting ischemic stroke-induced microglia activation and inflammatory responses and promoting neurogenesis and functional recovery. ADSC-exos rich in miR-30d-5p protect against acute ischemic stroke by inhibiting autophagy-mediated microglia/macrophage polarization to suppress inflammatory responses and reduce the area of infarcted brain damage (Jiang et al. 2018). Moreover, exosomes from hypoxia-pretreated ADSCs lessen brain damage brought by acute ischemic stroke and encourage M2 microglia/macrophage polarization by delivering the non-coding RNA circ-Rps5 (Yang et al. 2022). In addition, Lv et al. (Lv et al. 2021) proposed other molecular participants involved in EV-mediated signaling among adipose tissue and neurons. Exosome-carried miR-31 targeted neurons and downregulated TRAF6, thereby upregulating interferon regulatory factor 5 (IRF5) and ameliorating ischemic stroke-induced neuronal damage (Lv et al. 2021). ADSC-exos protect neurons from damage by suppressing the NF-κB and mitogen-activated protein kinase (MAPK) pathways and preventing microglia activation (Feng et al. 2019). ADSC-exos protect neurons from damage by suppressing the NF-κB and mitogen-activated protein kinase (MAPK) pathways and preventing microglia activation (Huang et al. 2018b). Ischemic hypoxia is associated with angiogenesis or the formation of new microvessels in pre-existing vessels. The induction of cerebral angiogenesis is critical for promoting post-stroke functional recovery (Ruan et al. 2015). In vitro, ADSCs-Exos enhance the angiogenesis of brain microvascular endothelial cells after OGD via miR-181b-5p, revealing a novel role for ADSCs-exos in stroke (Yang et al. 2018).

The relationship between cognitive impairment and ADSC-exos has also been studied. Such exosomes promote cognitive impairment via synaptic loss and chronic neuroinflammation, thereby facilitating the progression of cognitive impairment in high-fat diet-fed rats. Also, suppression of miR-9-3p may prevent insulin resistance cognitive impairment associated with obesity (Wang et al. 2022b). AD is characterized by the accumulation of Aβ in the brain due to an imbalance between Aβ production and clearance. ADSC-exos carrying enzymatically active neprilysin are transferred to N2a cells, a neuronal cell line with Aβ overexpression, where they reduce the extracellular and intracellular levels of Aβ40 and Aβ42. This has therapeutic potential for AD (Katsuda et al. 2013). In vitro, ADSC-exos modulate the cellular phenotype of amyotrophic lateral sclerosis (ALS), including superoxide dismutase (SOD-1) aggregation and mitochondrial dysfunction, suggesting therapeutic potential for ALS (Lee et al. 2016). Therefore, ADSC-exos may promote restoration of brain function, and their therapeutic potential warrants further investigation.

EVs in Brain–Liver Crosstalk

Hepatic encephalopathy demonstrates the close relationship between the brain and the liver (Cabrera-Pastor et al. 2019; Felipo 2013). The average incidence of neurological complications after liver transplantation is about 20% (Bronster et al. 2000). Hyperammonemia is the main cause of neurological impairment in patients with cirrhotic hepatic encephalopathy. Individuals with liver cirrhosis present with persistent hyperammonemia, which induces moderate hepatic encephalopathy, which causes motor incoordination and mild cognitive and motor impairment (Balzano et al. 2020; Cabrera-Pastor et al. 2019). Chronic hyperammonemic rats experience these neurological changes, which cause cerebellar and hippocampal neuroinflammation, altered GABAergic and glutamatergic neurotransmission, and motor incoordination and cognitive impairment (Cabrera-Pastor et al. 2019). EVs are thought to be involved in the transfer of peripheral changes to the brain. Plasma EVs in hyperammonemic rats were changed and the TNF-α level was elevated. When injected into normal rats, EVs from plasma of hyperammonemic rats reached the cerebellum and caused changes in Purkinje neurons and microglia in particular particularly Purkinje neurons and microglia, further resulting in neuroinflammation and motor incoordination (Izquierdo-Altarejos et al. 2022). Also, elevated TNF-α levels in EVs from hyperammonemic rats' blood activate the TNF receptor 1 (TNFR1)/C–C motif chemokine ligand 2 (CCL2)/BDNF/tropomyosin receptor kinase B (TrkB)/KCC2 and TNFR1/NF-kB/glutaminase/GABA transporter 3 (GAT3) pathways, thereby enhancing cerebellar GABAergic neurotransmission. As a result, microglia activation, increased GABAergic neurotransmission, and motor incoordination occur (Izquierdo-Altarejos et al. 2022). Apolipoprotein E4 (ApoE4) alleles are strongly associated with age-related cognitive impairment and AD (Kanekiyo et al. 2014; Strittmatter et al. 1993). Age-related thyroid deficiency can enhance exosomal transport of ApoE4 from liver to brain, which causes AD-related dementia and neuronal dysfunction in ApoE4 allele carriers. These changes also coincide with ApoE4-induced activation of the NLRP3 inflammasome and pyroptosis of neurons (Zhang et al. 2022a, b). Exosomes isolated from serum of rats with hepatic ischemia–reperfusion injury may induce neuronal damage in the hippocampus and cortex, which is associated with the NLRP3 inflammasome and caspase-1-dependent pyroptosis (Zhang et al. 2019). Although there have been fewer studies on the function of EVs in communication between brain and liver under normal and pathological conditions, this is an interesting area for future study.

EVs in Brain–Heart Crosstalk

There are numerous bidirectional connections between the cardiovascular system and CNS. Indeed, 11–18% of stroke patients suffer symptomatic heart failure, and the prevalence of left ventricular systolic dysfunction is 24% (Doehner et al. 2018). Stroke patients often present with myocardial injury, ischemic-like electrocardiographic changes, and arrhythmias (Ay et al. 2006; Oppenheimer 1994; Tokgozoglu et al. 1999). In preclinical studies, an elevated plasma catecholamine level after ischemic stroke is positively associated with the incidences of myocardial lesions and cardiac injury (Meloux et al. 2018). Despite knowledge gaps in our understanding of the processes of brain–heart relationships, emerging evidence suggests that EVs and their cargo might act as mediators of both inter-organ and intercellular interactions (Venkat et al. 2018). Under physiological conditions, EVs can cross the BBB (Banks et al. 2020b), and their ability to do so is enhanced by damage to the BBB (Dickens et al. 2017). Indeed, EVs produced from the brain under healthy or diseased settings can interact, and some of the molecular mechanisms and signaling pathways are similar in myocardial infarction (MI) and stroke (Otero-Ortega et al. 2020). Circulating EVs from patients with MI and stroke have comparable protein and miRNA profiles, despite the fact that they are two distinct clinical diseases affecting various organs and cell types (Otero-Ortega et al. 2020). Patients who have suffered an ischemic stroke, a cerebral hemorrhage stroke, or a SAH have higher circulating EV levels; this is linked to a poor prognosis (Huang et al. 2009; Lackner et al. 2010). Blood-derived MVs promote the produce of the endothelial cell cytokine IL-6 (Mesri and Altieri 1999), the expression of which is linked to the development of vasospasm (Schoch et al. 2007). IL-6 can induce vasospasm in the peripheral vasculature, including the cardiac vessels, leading to coronary syndrome. In cerebral hemorrhage, plasma procoagulant microvesicles levels are also significantly elevated and are linked to stroke development (Huang et al. 2009). Platelet MVs also play an important role in the communication between the brain and the heart. Altered platelet activity caused by platelets and/or MVs produced by the injured brain can lead to thromboembolic events and associated cardiac complications after brain injury (Horn et al. 2016). In preclinical studies, EVs produced by the heart entered the brain via blood and induced sympathetic activation in a rat model of chronic heart failure (Tian et al. 2022a). Also, plasma EVs from rats with chronic heart failure enhanced the inflammatory response in the rostral ventrolateral medulla (RVLM), possibly promoting sympathetic hyperactivity (Tian et al. 2022a). Moreover, cardiac-derived exosomes from mice with MI not only attenuated extracellular secretion of hippocampal synaptic vesicles after crossing the BBB but also led to hippocampal microtubule loss, which was improved by the exosomes production inhibitor GW4869 (Sun et al. 2018).

The non-coding RNAs are prevalent in EVs and have been linked to cardiac and brain dysfunction (Zheng et al. 2021). Long et al. (Long et al. 2013) reported that miR-126 carried by exosomes in circulating blood was decreased significantly for at least 24 weeks in patients with ischemic stroke, which may affect other organs including the heart. Endothelial cell-specific MiR-126 is essential for maintaining vascular integrity and controlling angiogenesis (Wang et al. 2008). Heart failure, atrial fibrillation, coronary artery disease, and possibly severe stroke-induced heart problems may be associated with a miR-126 deficiency (Qiang et al. 2013; Wang et al. 2008). Mesenchymal stem cell-derived EV-carried miR-17–92 cluster is a key regulator of cardiomyocyte proliferation in the neonatal and adult heart, as well as of post-stroke neural progenitor cell proliferation (Chen et al. 2013; Zhang et al. 2021b). In addition, the miR-17-92 cluster protects the ischemic heart and brain by targeting phosphoinositide 3-kinase/protein kinase B and mitogen-activated protein kinases/extracellular signal-regulated kinase signaling pathways (Zhou et al. 2013). MiR-1 is specifically expressed in adult cardiac, and an elevated serum miR-1 level after MI suggests cardiomyocyte necrosis as its source (Chistiakov et al. 2016). MI-induced overexpression of miR-1 in blood and brain may be associated with exosomes released from the heart, which enter the brain and are taken up by hippocampal neurons, resulting in hippocampal neuronal microtubule damage and reduced microtubule protein polymerization (Duan et al. 2018). These changes are prevented by inhibitors of exosome production and by selective inhibition of miR-1. Also, overexpression of miR-1 in the heart in a mouse model of MI suppressed the extracellular secretion of synaptic vesicles. This is caused by exosome-mediated transport from the heart to the brain induced by an elevated miR-1 level in the hippocampus, inhibiting synaptosomal-associated protein, 25 kDa (Snap25) expression (Duan et al. 2018). In a chronic HF rat model, cardiac-derived EVs enriched in miRNA-27a, miRNA-28a, and miRNA-34a entered the RVLM through the BBB to inhibit the nuclear factor-erythroid factor 2-related factor 2 (Nrf2)/antioxidant signaling pathway and promote sympathetic hyperactivity (Tian et al. 2022a).

Using high-throughput whole transcriptome sequencing of TBI mice with Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO) analyses, Zhao et al. (Zhao et al. 2018) showed circRNAs were predicted to influence myocardial contraction and calcium signaling pathways (Zhao et al. 2018). Although these findings are preliminary, they may provide an important basis for future research on exosomes carrying circRNAs in brain–heart communication.

EVs in Brain–Lung Crosstalk

The blood level of EV is markedly elevated in the acute phase of TBI in mice (Morel et al. 2008). Moreover, other neurological conditions like ischemic stroke and cerebral hemorrhagic stroke have higher levels of circulating EV (Chen et al. 2017). Tissue factors (TFs) are abundant on the EV surface and have procoagulant and proinflammatory effects following TBI. TFs initiate the coagulation reaction by activating thrombin to form fibrin (Grover and Mackman 2018). After TBI, the disrupted BBB allows EVs from injured brain tissue to enter the peripheral circulation, where they induce coagulation, fibrinolysis, vascular fibrin deposition, and thrombosis (Zhao et al. 2016). Circulating EVs are primarily consumed by the lung (Takahashi et al. 2013). Hideki et al. (Yasui et al. 2016) showed that brain-derived MVs expressed TFs in the alveoli and that thrombin inhibitors reduced lung injury, suggesting a role for brain-derived MVs in TBI-induced acute lung injury (ALI). After craniocerebral injury, injured brain tissue releases large amounts of damage-associated molecular patterns (DAMPs), triggering an innate immune response (Shi et al. 2019). EVs are implicated in the development of ALI as mediators of inflammation. For instance, EVs containing proinflammatory cytokines were released after experimental TBI and taken up by lung cells, triggering inflammatory vesicle activation. Kerr et al. (Kerr et al. 2018) found that EVs containing inflammasome proteins caused ALI in a TBI mouse model, and detected caspase-1, apoptosis-associated speck-like protein (ASC), and IL-18 in pulmonary type II alveolar epithelial cells. Enoxaparin, a blocker of EVs, significantly ameliorated ALI, implicating EVs in brain–lung communication. An inflammasome component in EVs, ASC, which contains a caspase recruitment domain, promotes TBI-induced lung injury. Targeting of human lung microvascular endothelial cells by EVs from TBI subjects induces inflammasome activation, leading to pyroptosis (Kerr et al. 2019). The lungs facilitate gas exchange through the alveolar capillaries at the blood–air barrier (BAB) (Baffour et al. 2019), which is composed of alveolar type I epithelial cells and type II epithelial cells and endothelial cells. BAB disruption is an important pathological feature of ALI (Nova et al. 2019). EVs can affect the disruption of the alveolar–capillary barrier in TBI. Kerr et al. (Kerr et al. 2019) also reported that serum-derived EVs from TBI patients carry inflammasome proteins that cause lung endothelial cell pyroptosis and disrupt the integrity of the BAB, thereby exacerbating vascular permeability and pulmonary edema.

Respiratory diseases can also lead to pathophysiological alterations in the brain, which is also involved with EVs. Chavez et al. (Chavez et al. 2021) reported that circulating blood exosomes containing caspase-1 crossed the BBB and caused neuroinflammation and caspase-1/gasdermin D activation in neonatal rats with ventilation-induced lung injury (VILI); this interaction could lead to brain damage and neurodevelopmental deficits in infants who receive injurious or prolonged mechanical ventilation. Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection promotes immune processes and oxidative stress, potentially triggering neurological disorders including AD and Parkinson’s disease (PD) (Heneka et al. 2020). J. Ahmed et al. (Ahmed et al. 2021) demonstrated that factors transported in exosomes from the lung to the brain influence SARS-CoV-2 infection by analyzing multiple transcriptomic datasets from SARS-CoV-2-infected lungs, RNA profiles of lung exosomes, and human-brain gene expression profiles. Also, 19 transcription factor-encoding mRNAs were overexpressed in pulmonary SARS-CoV-2 infection; among them, six transcription factors activated 469 genes associated with AD and PD, most of which are involved in inflammatory, apoptotic, and other signaling pathways that contribute to neurodegeneration (Ahmed et al. 2021). This suggests the need for further studies on the role of exosomal brain–lung communication in SARS-CoV-2 infection (Ahmed et al. 2021).

EVs in Brain–Gut Crosstalk

Communication between the gut and the brain is important for maintaining homeostasis. The gut–brain axis (GBA) is a complicated bidirectional relationship between the gut microenvironment and the brain (Raimondi et al. 2019). Disturbances in the gut microenvironment are associated with AD (Kesika et al. 2021; Megur et al. 2020), spectrum disorders (Socala et al. 2021; Srikantha & Mohajeri 2019), and PD (Mulak & Bonaz 2015; Sampson et al. 2016). The GBA contains multiple signaling pathways that transmit signals to the CNS via the vagus nerve or circulating blood. However, until recently, the significance of EVs in brain–gut communication has been gradually gaining attention (Haas-Neill & Forsythe 2020; Zhao et al. 2021).

EVs in the intestinal microenvironment can be from microorganisms and intestinal cells. Like other organisms, bacteria can also produce EVs, which are key mediators of microbial–host interactions and are widely acknowledged as an essential intercellular communication mechanism (Macia et al. 2019). The majority of EVs of gram-negative bacteria are produced by budding of the outer membrane and are termed outer membrane vesicles (OMVs); EVs formed by the outer and inner membranes of gram-negative bacteria have also been documented (Perez-Cruz et al. 2013, 2015). By incorporating cell wall modifications, gram-positive bacteria form cytoplasmic membrane vesicles that are thereafter released (Briaud and Carroll 2020; Toyofuku et al. 2019), however, the mechanism of their release is unclear (Ma and Cao 2021; Toyofuku et al. 2019). Besides bacteria, the human gut can also host a number of other microorganisms, like fungi, whose EVs can also interact with host cells and function (Rizzo et al. 2020). Bacterial EVs, like eukaryotic EVs, carry and stably store polysaccharides, proteins, DNA, RNA, metabolites, and enzymes. Their outer membranes harbor many of the same proteins as the surface of the originating bacterium (Kaparakis-Liaskos and Ferrero 2015; Liao et al. 2014). Bacterial EVs can cross the BBB and deliver cargo (Pirolli et al. 2021). Lipopolysaccharide (LPS) is present in bacterial EVs and may be linked to neuroinflammation (Pirolli et al. 2021). In addition, it is thought that OMVs can disrupt the integrity of the BBB, which may contribute to other harmful substrates and other OMVs in the BBB and thus have an impact on brain pathology (Pirolli et al. 2021). For instance, OMVs increase BBB permeability and promote astrocyte and microglia activation, and cause an inflammatory reaction and tau hyperphosphorylation by activating the glycogen synthase kinase-3 beta (GSK-3β) pathway; the end result is cognitive impairment (Wei et al. 2020). These neural effects of OMVs are similar to those seen in human AD pathology, including GSK-3β-mediated tau phosphorylation, the signature pathological marker of AD (Wei et al. 2020). The vagus nerve and blood can be used by the bacteria EVs to facilitate gut–brain communication, according to a new research by Lee et al. (Lee et al. 2020). Also, Paenalcaligenes hominis EVs increase the amount of activated microglia in the hippocampus and contribute to cognitive impairments in the brain. In animals severed by the vagus nerve, cognitive impairments were considerably decreased and hippocampus cell populations reverted to normal, suggesting that EV–vagus nerve interactions are at least partially to blame for the alterations in the brain (Lee et al. 2020). A recent report also reported that OMVs are transported from the circulatory system to brain monocyte/microglial cells and suggests that OMVs transport to the brain may be associated with infection anywhere in the body, which can lead to neuroinflammation (Ha et al. 2020).

Many neurodegenerative diseases are associated with microglia dysfunction and altered microbiome composition (Abdel-Haq et al. 2019), in this case, OMVs can deliver LPS, DNA, and RNA to the brain to activate neuroinflammation and affect gene expression in the long term. For example, Actinomyces OMVs cross the BBB into the brain in mice, and extracellular RNA (exRNA) transfer of OMVs to the brain causes neuroinflammatory diseases such as AD by activating the TLR8 and NF-κB inflammatory signaling pathways (Han et al. 2019). Similarly, EVs secreted by the periodontopathogen Aggregatibacter actinomycetemcomitans can deliver exRNAs to brain monocytes and microglia to activate IL-6 and NF-κB, thereby inducing neuroinflammation (Ha et al. 2020). Paenalcaligenes hominis proliferation in the gut of aged individuals accelerates LPS absorption into the blood and EV translocation to the brain, leading to cognitive impairment induced by brain inflammation (Lee et al. 2020). Bacterial EVs also carry psychoactive cargo. The inhibitory neurotransmitter gamma-aminobutyric acid (GABA) and its biosynthetic intermediates α-ketoglutarate and glutamate were present in EVs produced by Bacillus fragilis (Zakharzhevskaya et al. 2017). Enteric bacterial EVs contain neurotransmitters that affect the brain, and their effects on neurodegenerative diseases warrant further investigation.

In addition, EVs from gut microorganisms can also exert beneficial effects on the brain. In vitro, BDNF and proBDNF levels in HT22 hippocampus cells were increased by EVs from the gut gram-positive probiotic Lactobacillus plantarum (Choi et al. 2019). Intraperitoneal injection of Lactobacillus plantarum-derived EVs increased BDNF and neurotrophin 4/5 (Nt4/5) expression and blocked stress-induced depression-like behaviors (Choi et al. 2019). Also, EVs from the gram-positive probiotic Bacillus subtilis and the gram-negative probiotic Akkermansia muciniphila exert antidepressant effects by restoring stress-induced produced of methyl-CpG binding protein 2 (MeCP2), sirtuin 1 (Sirt1), and neurotrophic factors in the hippocampus (Choi et al. 2022).

EVs released from intestinal epithelial cells (IECs) affect the brain. The environment of EVs affects their characteristics in addition to the kind of parental cell type. IECs are exposed to food, intestinal pathogens, and microbes. The synthesis and characteristics of EVs produced from the gut can change in this environment, which may have an impact on the pathophysiology of the brain. Katakura et al. (Inotsuka et al. 2020) showed that exosomes derived from GABA-treated IECs activated SH-SY5Y cells and facilitated neurite growth. Also, exosomes found in the serum of mice orally given GABA stimulated neuronal development, suggesting that EVs released from the gut may mediate the positive effects of dietary GABA on the brain (Inotsuka et al. 2021). Intestinal microbial infections, toxins, and other factors can potentially have an impact on EVs produced from the gut (Ayyar and Moss 2021; Deng et al. 2015). Moreover, affected by this microbial–host interaction may be remote organs (Chen et al. 2019). Gut cells infected with Escherichia coli LF82 generate large exosomes rich in miRNAs, including miR-130a (Larabi et al. 2020). MiR-130a is associated with increased BBB permeability (Wang et al. 2016). The intestinal immune system is activated by exosomes released by IECs that contain exogenous peptides that are complexed with the major histocompatibility complex II (MHC-II) (Kojima et al. 2018). Neuroinflammation, brain damage, and alterations in neurogenesis and plasticity all cause the gut immune system to become active (Roubalova et al. 2020). In rats with sepsis-associated encephalopathy, Xi et al. (Xi et al. 2021) found that intestinal flora disturbance promoted the release of intestinal epithelium-derived exosomes, inducing hippocampal neuronal damage and apoptosis in a manner dependent on autophagy and excess circulating IL-1β; the effect was suppressed by the exosome inhibitor GW4869.

Conclusions and Future Directions

To date, studies on EV-mediated bidirectional communication between the brain and other organs are still at early stages. This is an emerging area of research and there are currently many key challenges that may hinder progress in the study of EV-mediated bidirectional communication between the brain and other organs. First, how to precisely identify organs, especially the cell types that create or receive particular EVs, so we need to find organ-specific EVs. Second, how should we find out the actual effect of the specific signals carried by EVs on the function of organs in the body? Finally, whether EVs generated in vitro could be used as next-generation therapeutic approaches, such as alternative viruses or synthetic nanovectors, for organ-targeted biotherapies. Addressing the first challenge requires to find organ-specific EVs or craft strategies to use novel technologies to label organ-specific EVs in appropriate animal models (Chand et al. 2020; Linxweiler et al. 2021). The next challenge may require the generation of genetic animal models with organ-specific defects in the production of EVs. Another thing to think about is whether EVs generated in vitro could replace viruses or synthetic nanovectors as a new vector for organ-targeted biotherapies (Zhang et al. 2022a; Fan et al. 2022; Riazifar et al. 2019; Xu et al. 2021). Also, isolation and purification of EVs is a major challenge, which requires more accurate and reliable techniques and methods to measure the number and distribution of EVs and to isolate functionally and morphologically distinct subpopulations.

This review opens up new horizons: new bidirectional communication networks between the brain and other organs are found and mapped through the use of EVs. The understanding of EVs as a transboundary signaling system provides valuable insights and may lead to more effective biological therapies. At present, more research is needed to better understand how EVs help the brain and other organs communicate with one another. With our understanding of this type of cross-border signaling expanding, there is enormous potential for answering the unanswered problems in this field.