The lymphatic drainage systems in the brain: a novel target for ischemic stroke?

Abstract

Recent studies have proposed three lymphatic drainage systems in the brain, that is, the glymphatic system, the intramural periarterial drainage pathway, and meningeal lymphatic vessels, whose roles in various neurological diseases have been widely explored. The glymphatic system is a fluid drainage and waste clearance pathway that utilizes perivascular space and aquaporin-4 protein located in the astrocyte endfeet to provide a space for exchange of cerebrospinal fluid and interstitial fluid. The intramural periarterial drainage pathway drives the flow of interstitial fluid through the capillary basement membrane and the arterial tunica media. Meningeal lymphatic vessels within the dura mater are involved in the removal of cerebral macromolecules and immune responses. After ischemic stroke, impairment of these systems could lead to cerebral edema, accumulation of toxic factors, and activation of neuroinflammation, while restoration of their normal functions can improve neurological outcomes. In this review, we summarize the basic concepts of these drainage systems, including drainage routes, physiological functions, regulatory mechanisms, and detection technologies. We also focus on the roles of lymphatic drainage systems in brain injury after ischemic stroke, as well as recent advances in therapeutic strategies targeting these drainage systems. These findings provide information for potential novel strategies for treatment of stroke.

Introduction

Stroke is the leading cause of death and acquired disability in adults and can be broadly classified as ischemic stroke and hemorrhagic stroke (Campbell et al., 2019). Ischemic stroke is responsible for 71% of stroke events worldwide (Campbell et al., 2019) and is a disorder of cerebral blood flow caused by vascular obstruction. The pathophysiology includes excitotoxicity, free radical release, protein misfolding, mitochondrial response, and inflammatory changes, leading to neuronal death and neurological deficits (George and Steinberg, 2015; Zhu et al., 2021; Li et al., 2022; Jiang et al., 2022). Ischemic stroke poses a great threat to human health; however, currently effective treatment methods are limited.

Currently, the theory of intracerebral lymphatic drainage systems includes three parts: the glymphatic system, the intramural periarterial drainage (IPAD) pathway, and meningeal lymphatic vessels (Figure 1). The glymphatic system, first proposed by Iliff et al. (2012), is a new concept. This pathway is highly dependent on aquaporin-4 (AQP4) expressed on the endfeet of astrocytes (Jessen et al., 2015). The glymphatic system provides a fast pathway for the inflow of cerebrospinal fluid (CSF) and the outflow of interstitial fluid (ISF), which facilitates the clearance of metabolites from the brain and maintains the homeostasis of cerebral fluid (Jessen et al., 2015). The IPAD pathway acts as another ISF drainage system, draining solutes through the capillary basement membrane and tunica media of the arteries to the cervical lymph nodes (Carare et al., 2008). Meningeal lymphatic vessels, located in the dura mater, serve as downstream channels to drain ISF, macromolecules, and immune cells out of the cranial cavity, and play a role in the regulation of the immune response in the brain (Aspelund et al., 2015; Louveau et al., 2015a).

Figure 1.

The overall lymphatic drainage systems.

This figure summarizes the anatomic structure of three lymphatic drainage systems. CSF flows into the cerebral parenchyma through the periarterial space of the glymphatic system. ISF outflow relies on the perivenous space as well as the basement membrane of capillaries and arterial tunica media (IPAD pathway). The meningeal lymphatic vessels act as an important fluid efflux pathway of the glymphatic system, draining CSF and ISF to the deep cervical lymph nodes. AQP4: Aquaporin-4; CSF: cerebrospinal fluid; IPAD: intramural periarterial drainage; ISF: interstitial fluid.

In recent years, some studies have focused on the roles of lymphatic drainage systems in the pathology of neurodegenerative diseases, such as Alzheimer’s disease (Reeves et al., 2020) and Parkinson’s disease (Sundaram et al., 2019). In addition, much attention has also been paid to the relationship between lymphatic drainage systems and ischemic stroke, especially its roles in the formation of cerebral edema, activation of neuroinflammation, and modulation mechanism of AQP4 changes after stroke. In this study, we focus on the glymphatic system and cover the latest advances of the IPAD pathway and meningeal lymphatic vessels. The drainage routes, physiological functions, regulatory mechanisms, and detection technologies of the lymphatic drainage systems in the brain are reviewed. Furthermore, their roles in brain injury after ischemic stroke, as well as recent therapeutic advances targeting these systems, especially AQP4, are also discussed.

Search Strategy and Selection Criteria

Studies cited in this review were published from 2000 to 2022 and obtained following a PubMed search using the following keywords: stroke, ischemia, ischemic, glymphatic, intramural periarterial drainage, meningeal lymphatic vessel, edema, inflammation, astrocyte, AQP4. The literature search was completed by YJW on March 22, 2022.

Current Understanding of the Glymphatic System

Drainage route and physiological functions of the glymphatic system

The glymphatic system consists of three key compartments: a periarterial CSF influx route, a perivenous ISF efflux route, and a parenchymal exchange pathway dependent on astrocytic AQP4 (Iliff and Nedergaard, 2013; Figure 2). It was first reported by Iliff et al. (2012). This team detected the movement of an intracisternally injected fluorescent tracer by two-photon laser scanning microscopy and visualized the CSF influx route in real time. After injection, the tracers rapidly flowed along the perivascular space of the cortical surface arteries and penetrating arterioles, and then entered the brain parenchyma. The glymphatic drainage route throughout the brain was defined by fluorescence imaging of fixed brain slices. Longer-time observation showed that tracers accumulated along capillaries and parenchymal venules, implying that interstitial solutes were cleared along the perivenous space surrounding large caliber draining veins (Iliff et al., 2012). In mice with AQP4 gene deletion, tracer movement from the perivascular space to the surrounding interstitium and clearance from the brain parenchyma were significantly compromised (Iliff et al., 2012), thus supporting the hypothesis that AQP4 provided low resistance pathways for CSF-ISF exchange. The fluid was ultimately discharged into the dural sinuses through the arachnoid granulations and into peripheral lymphatic vessels through the perineural spaces along the cranial nerves (Koh et al., 2005). The meningeal lymphatic vessels are also proven a route for ISF absorption and transport of fluid into deep cervical lymph nodes (Iliff et al., 2015). The most widely accepted theory of glymphatic fluid dynamics is convective bulk flow, which suggests that solutes with different molecular sizes share the same clearance rate (Iliff et al., 2012).

Figure 2.

Detailed information on glymphatic system and IPAD pathway.

CSF from subarachnoid space flows along the periarterial space, then enters the brain parenchyma via AQP4, which is highly-expressed on the perivascular astrocytic endfeet. In the brain, the neurons and other neurocytes produce metabolic waste (Aβ, for example), which is excreted into the ISF and is mixed with CSF to achieve the ISF renewal. The mixed fluid enters the perivenous space via AQP4, then flows into the dural sinus via arachnoid granulations and the peripheral lymphatic system through the perineural space and meningeal lymphatic vessels. ISF can also enter the basement membrane of capillaries and the basement membrane surrounding SMCs of the arteries, which flows out of the brain triggered by vasomotion of arteries. AQP4: Aquaporin-4; Aβ: amyloid β; CSF: cerebrospinal fluid; IPAD: intramural periarterial drainage; ISF: interstitial fluid; SMC: smooth muscle cell.

Originally, the glymphatic system was noted for its ability to clear waste. CSF circulation functions as an alternative route for metabolic waste that cannot be easily eliminated through the blood-brain barrier (BBB) (Iliff et al., 2015). The glymphatic system is one of the main pathways for the clearance of extracellular metabolites, including amyloid β (Aβ) (Iliff et al., 2012), tau protein (Iliff et al., 2014; Harrison et al., 2020), and lactate (Lundgaard et al., 2017). Furthermore, the glymphatic system also contributes to the brain-wide distribution of some important molecules, such as glucose, amino acids, lipids, growth factors, and neuromodulators (Christensen et al., 2021). The role of regulation of fluid homeostasis is also of concern, since the imbalance between the CSF inflow and the ISF outflow exacerbates the formation of cerebral edema and hinders its resolution. Glymphatic CSF influx contributes to early brain swelling after acute ischemic stroke (Mestre et al., 2020) and cardiac arrest (Du et al., 2022). Meanwhile, glymphatic outflow drives excess water removal from the brain to eliminate vasogenic edema (Papadopoulos et al., 2004).

Regulatory mechanisms of the glymphatic system

To date, research on the regulatory mechanisms of the glymphatic system has focused on arterial pulsation and sleep. The arterial pulsation was initially discovered after observing that unilateral internal carotid artery ligation reduced cerebral arterial pulsation and slowed perivascular CSF-ISF exchange, while increased pulsation with dobutamine could accelerate this process (Iliff et al., 2013b). According to ultrafast magnetic resonance encephalography signal detection, it was observed that although periarterial inflow was driven by arterial pulsation, the perivenous drainage system could be regulated by the respiration cycle (Kiviniemi et al., 2016). Inspiration expanded the perivenous space and facilitated glymphatic outflow from the brain, while exhalation reversed these effects (Kiviniemi et al., 2016). However, several fluid dynamical models have questioned the feasibility of arterial pulsation as a driving force (Asgari et al., 2016). To answer this controversial question, Mestre’s group (Mestre et al., 2018) recently conducted a study in which particle tracking velocimetry was used for the quantification of CSF flow, along with simultaneous measurements of artery diameter and heartbeat to reflect vessel movement. In their experiments, the peak in arterial wall motion velocity was consistent with the peak of CSF spatial root mean square velocity, and both showed a delay after electrocardiogram peak. Therefore, the authors proposed that the local movement of the arterial wall acted as a pump to drive the flow of the surrounding CSF (Mestre et al., 2018). In addition, hypertension causes adverse effects on arterial pulsation, increasing fluid backflow by 21% and thus reducing the drainage rate in the perivascular space (Mestre et al., 2018).

Sleep is an evolutionarily conserved biological behavior, promoting the clearance of harmful metabolites and consolidating memory. Using two-photon imaging, Xie et al. (2013) found that glymphatic function was 90% more active during sleep, while it was significantly suppressed in an awake state. In sleep, the size of the interstitial space increased by 60%, leading to a striking increase in the exchange of CSF-ISF and Aβ clearance. Changes in cell volumes may be related to different noradrenergic signaling derived from the locus coeruleus between awake and sleep periods. Administration of adrenergic antagonists increases interstitial volume, the power of slow waves on electrocorticography, and CSF tracer influx in the same manner as during sleep (Xie et al., 2013). In both rodents and humans, sleep deprivation results in a dramatic increase in Aβ and tau burden (Shokri-Kojori et al., 2018; Holth et al., 2019), reflecting impaired glymphatic function. In addition, the lateral sleeping position could optimize waste removal compared to the supine or prone sleeping position, which may be a choice for long-term evolution (Lee et al., 2015). Increased excitatory neurotransmitters caused by neuronal activity during wakefulness has also been confirmed to interfere with ISF drainage capacity (Li et al., 2020). Furthermore, a recent study found that regulation of glymphatic function was not only dependent on the arousal state, but also exhibited a circadian rhythm that peaked during the day when mice were sleeping and decreased during the night when mice were active (Hablitz et al., 2020). These day-night differences persisted under constant light, supporting the existence of endogenous circadian oscillations. Maintenance of this circadian rhythm depended on up-regulated AQP4 localization to the perivascular endfeet at midday, since the elimination of AQP4 eliminated the day-night difference (Hablitz et al., 2020). Therefore, factors that cause disturbances in the circadian clock may further promote glymphatic dysfunction.

Detection technologies for the glymphatic system

Although the fluorescence imaging method in vivo or in vitro has been confirmed to explore the drainage pathway of the glymphatic system (Iliff et al., 2012), there are inevitable drawbacks to these approaches. Two-photon microscopy has a narrow view field (about 3 mm in diameter), requires an invasive operation on the skull and is incapable of probing the deep brain region (less than 250 µm in depth) (Sweeney et al., 2019). For in vitro imaging, the process of perfusion-fixation collapses the perivascular space and significantly alters the distribution of the tracers (Sweeney et al., 2019). Since magnetic resonance imaging (MRI) can provide a real-time CSF-ISF exchange map of the whole brain in a relatively noninvasive manner, it has become a more favorable detection method for exploring the glymphatic system.

In 2013, dynamic contrast-enhanced MRI was applied to reveal the anatomical routes of perivascular CSF influx (Iliff et al., 2013a). After intracisternal administration of the paramagnetic contrast agent, researchers observed an CSF-ISF exchange process in the rat brain using T1-weighted MRI. Two key influx nodes were identified and they further defined simple kinetic parameters to reflect glymphatic functions (Iliff et al., 2013a). To explore the glymphatic system in humans, the effectiveness of lumbar intrathecal infusion has been evaluated. The results showed that intrathecal infusions could also image the glymphatic pathway similar to intracisternal infusion, although the peak fluorescence intensity was reduced and delayed (Yang et al., 2013). Therefore, clinically established techniques of myelography and cisternography may have the potential to assess glymphatic function. In a case report (Eide and Ringstad, 2015), a patient was subjected to intrathecal gadobutrol-based MRI for the diagnosis of spontaneous CSF leakage. On the MRI images, intrathecal gadobutrol was widely distributed in all subregions of the human brain (Eide and Ringstad, 2015). Diffusion tensor imaging (DTI) is a new method of monitoring the activity of the glymphatic system by assessing the movement of water molecules along the perivascular space. Taoka et al. proposed that diffusion tensor image analysis along the perivascular space (DTI-ALPS) index could serve as a potential neuroimaging marker of glymphatic function, showing a significantly positive correlation with cognitive function in Alzheimer’s disease (Taoka et al., 2017). In fact, the DTI-ALPS index has successfully reflected the dysfunction of the glymphatic system in various diseases, such as Parkinson’s disease (Ma et al., 2021), obstructive sleep apnea (Lee et al., 2022), and ischemic stroke (Toh and Siow, 2021). A recent study confirmed the reproducibility of the DTI-ALPS index and suggested that diffusion-weighted imaging values could be used more widely in clinical practice as a substitute for DTI (Taoka et al., 2022).

Function and Regulation of the Intramural Periarterial Drainage Pathway

In 2008, Carare et al. (2008) proposed another pathway for ISF removal, that is, the IPAD pathway (Figure 2). After injection of fluorescent tracers into the striatum of mice, they observed that the tracers entered the basement membrane of capillaries and the basement membrane surrounding smooth muscle cells (SMCs) of the arteries, then the tracers moved towards the leptomeningeal arteries and deep cervical lymph nodes (Carare et al., 2008). This pathway was later verified by electron microscopy (Morris et al., 2016) and two-photon microscopy (Arbel-Ornath et al., 2013). The IPAD pathway has been proven to be an important route for the clearance of cerebral metabolites. The deposition sites of Aβ in cerebral vessels of amyloid angiopathy were consistent with the drainage route of IPAD, suggesting a failure of Aβ clearance (Hawkes et al., 2014). Tau and α-synuclein proteins could also be cleared through the IPAD pathway (Nimmo et al., 2020). The IPAD route and the perivenous ISF clearance pathway of the glymphatic system were believed to contribute cooperatively to the clearance of fluids and toxic substances in the brain, although their individual contribution has not been fully determined.

Initially, arterial pulsation was naturally considered as the driving force of IPAD (Carare et al., 2008). However, later mathematical modeling tests found that this was insufficient to produce ISF drainage in the opposite direction of blood flow (Diem et al., 2017). Therefore, a new model was proposed, suggesting that the vasomotion generated by rhythmic contraction and dilatation of arterial SMC induced deformation of the basement membrane, which squeezed ISF from the pores within the basement membrane along the direction of the contraction wave (Aldea et al., 2019). Subsequent in vivo studies also showed that increased vasomotion amplitude led to an increased ISF clearance rate (van Veluw et al., 2020). Additionally, since vasomotion is regulated by cholinergic neurons, loss of innervation leads to a significant increase in brain Aβ burden (Diem et al., 2018; Nizari et al., 2021). These results indicated that improvement in SMC activity and neurovascular coupling could be a potentially promising therapeutic target to modulate the function of the IPAD drainage system in the brain.

Detection and Function of Meningeal Lymphatic Vessels

In 2015, Louveau and Aspelund’s research groups (Aspelund et al., 2015; Louveau et al., 2015b) independently demonstrated the presence of lymphatic vessels at both molecular and functional levels in the dura mater of the central nervous system. Lymphatic endothelial cell specific markers were found to be positive in vessel-like structures in mice meninges dissection samples (Louveau et al., 2015b). Fluids from meningeal lymphatic vessels travel along arteries, veins, and cranial nerves, exit the cranial cavity _through the foramina at the base of the skull, and eventually drain into deep cervical lymph nodes (Aspelund et al., 2015). In transgenic mice with lymphatic vessel aplasia, CSF and ISF tracer transfer was significantly inhibited (Aspelund et al., 2015). In addition to rodents, MRI provided convincing evidence on the existence of meningeal lymphatic vessels in nonhuman primates, even in humans. Utilizing MRI contrast agents gadobutrol, T2-fluid attenuated inversion recovery and T1-weighted black-blood imaging visualized lymphatic vessels alongside the dural venous sinuses, especially the superior sagittal and straight sinuses, as well as branches of the middle meningeal artery (Absinta et al., 2017). More recently, a non-contrast fluid attenuated inversion recovery MRI study based on internal macromolecular signals within meningeal lymphatic vessels demonstrated the dorsal and ventral lymphatic vessels and their direct connection with cervical lymph nodes (Albayram et al., 2022).

The lymphatic vessels are considered to be an outflow pathway within the glymphatic system, draining ISF from the central nervous system to the periphery tissue (Louveau et al., 2015b). An MRI study of 35 patients with intrathecal contrast injection measured changes in signal intensity in six regions of the brain (Zhou et al., 2020). The results showed that the clearance of the glymphatic system and meningeal lymphatic vessels was associated with aging, and the drainage rate of the glymphatic pathway was significantly faster in patients with early tracer filling in lymphatic vessels (Zhou et al., 2020). Since meningeal lymphatic vessels can directly drain macromolecules and immune cells from the CSF to cervical lymph nodes, they are believed to participate in the regulation of immune responses in the brain. Labeled T cells and dendritic cells injected into the CSF could be detected in meningeal lymphatic vessels, which further drains to the superficial and deep cervical lymph nodes (Louveau et al., 2018). This process relied on the chemokine receptor 7/chemokine C-C motif ligand 21 pathway through which the immune cells and lymphatic endothelial cells interacted. Pharmacochemical ablation of meningeal lymphatic vessels blocked the immune cell migration pathway and reduced the activation of encephalitogenic T cells in lymph nodes, thus attenuating the severity of neuroinflammatory diseases such as experimental autoimmune encephalomyelitis (Louveau et al., 2018). Conversely, the integrity of the meningeal lymphatic vessels is essential for antitumor immunotherapy. Vascular endothelial growth factor C enhanced meningeal lymphangiogenesis in brain tumors, promoted dendritic cell drainage to lymph nodes, and led to the activation of tumor-specific T and B cell populations (Song et al., 2020). With the help of vascular endothelial growth factor-C, immune checkpoint inhibition therapy showed better efficacy in tumor treatment (Hu et al., 2020).

Impairment of the Lymphatic Drainage Systems Following Ischemic Stroke

The dysfunction of glymphatic drainage in ischemic stroke

In ischemic stroke, glucose and oxygen depletion can lead to irreversible neuronal injury, accompanied by the generation of metabolic byproducts (Gaberel et al., 2014). Infarct areas are rich in abundant factors related to neurotoxicity and neurodegeneration, such as pro-inflammatory cytokines and chemokines (Zbesko et al., 2018). Therefore, the glymphatic system, as a waste clearance pathway, could be involved in the pathological process of ischemic stroke. Researchers have found that glymphatic drainage function was impaired in ischemic stroke mice. In a study by Gaberel et al. (2014), MRI and histological examination showed a blockade of ipsilateral glymphatic perfusion three hours after middle cerebral artery occlusion (MCAO). Twenty-four hours later, the middle cerebral artery was recanalized and the glymphatic system function returned to normal. The authors attributed this dysfunction to decreased arterial pulsation and perivascular space occlusion. Therefore, intravenous injection of fibrinolytic drugs could restore arterial patency and improve glymphatic function (Gaberel et al., 2014). A similar phenomenon could be observed in mice with multiple microinfarctions. The experimental results showed that the flow of fluorescent CSF tracers along the periarterial space was significantly reduced overall and did not recover until 2 weeks later (Wang et al., 2017). The intensity of the CSF tracer in microinfarct cores increased significantly and CSF protein trapping persisted for more than 14 days. Compared to young mice, 12-month-old mice showed more severe damage, implying that aging brain was more vulnerable to ischemia (Wang et al., 2017). A recent study reported impaired glymphatic function in humans. The DTI-ALPS index in the ischemic hemisphere of stroke patients was significantly lower than that of the contralateral hemisphere and normal controls, which gradually recovered 14 days after stroke (Toh and Siow, 2021).

The capture of toxic solutes within infarcted lesions can serve as persistent foci of chronic neuroinflammation, leading to cognitive dysfunction and progression of dementia (Zbesko et al., 2018). For example, when tau protein clearance by the glymphatic system is interfered with, this may aggravate post-stroke dementia (Back et al., 2020). Focal cerebral ischemia can also cause neuronal damage in distal areas with fiber connections to the ischemic area, such as the substantia nigra (Nakane et al., 2001). Lin et al. (2020) found through enhanced MRI that in the acute and subacute stages after ischemic stroke, the clearance rate of contrast agent was slower in the ipsilateral substantia nigra, indicating a decrease in glymphatic function in this area. Potential mechanisms may include increased intracranial pressure, AQP4 depolarization, and swelling of the astrocytic endfeet (Lin et al., 2020).

In recent years, greater attention has been paid to glymphatic system dysfunction in cerebral edema after stroke. Cerebral edema is a detrimental adverse effect of ischemic stroke, leading to secondary hypoperfusion, additional tissue loss, increased intracranial pressure, and potential risk of cerebral hernia (Mestre et al., 2020). In Mestre et al.’s study (Mestre et al., 2020), the dynamics of the CSF influx during the first five minutes after stroke were depicted. MCAO caused a three-fold increase in the influx of ipsilateral tracer in comparison to the contralateral hemisphere. There were two peaks of CSF flow: a first peak (11.4 ± 1.8 seconds) and a second peak (5.24 ± 0.48 minutes), and a decrease in intracranial pressure coincided with the first influx peak (Mestre et al., 2020). Spreading depolarization (SD) is a specific type of depolarization wave that propagates widely through gray matter, causing the breakdown of neuronal ion gradients and aggravating cytotoxic edema (Hartings et al., 2017). Further experiments proved that the glymphatic influx at the second peak was triggered by SD (Mestre et al., 2020). After the SD wave, a potent contraction of the pial and penetrating arterioles occurred, resulting in an enlargement of the distal perivascular space and an establishment of a pressure gradient that could effectively facilitate the drainage of perivascular fluid (Mestre et al., 2020). As a result, more fluid spread into the brain parenchyma, enhancing tissue swelling. This study confirmed for the first time that the influx of CSF from the glymphatic system was a source of early edema observed in ischemic stroke. Possible mechanisms of glymphatic dysfunction following cerebral ischemia are summarized in Figure 3.

Figure 3.

Impaired glymphatic system and IPAD in ischemic stroke and therapeutic targets.

Possible mechanisms of glymphatic dysfunction following cerebral ischemia: (1) Decreased arterial pulsation due to vessel occlusion reduces the driving force; (2) Enlarged perivascular space is associated with aggravated cytotoxic edema; (3) Changes in AQP4 expression and distribution are involved in cerebral edema formation and resolution, as well as solute clearance deficiency; (4) Swelling astrocytes restrict the flexibility of fluid flow. The pathological changes of IPAD following cerebral ischemia include damaged smooth muscle cells and basement membrane degradation. As for therapeutic targets, AQP4 is a promising target for glymphatic function improvement and edema amelioration. Potential treatment strategies may include inhibition of AQP4 function, regulation of AQP4 expression and restoration of AQP4 polarization. Drugs that modulate SMC function can facilitate the IPAD clearance pathway and may be beneficial for preventing the cognitive impairment observed after stroke. AQP4: Aquaporin-4; IPAD: intramural periarterial drainage; SMC: smooth muscle cell.

Changes in the IPAD pathway and meningeal lymphatics in ischemic stroke

Following ischemic stroke, the IPAD drainage pathway has been shown to be noticeably impaired. Using a mouse model of focal stroke induced by photothrombosis, Arbel-Ornath et al. (2013) found that the clearance rate of dyes along the IPAD pathway within occluded vessels was markedly impaired. After ischemic insult, the weak contractile activity of damaged SMC resulted in a significantly reduced solute clearance efficiency through the IPAD pathway (Aldea et al., 2019). Breakdown of the blood-brain barrier (BBB) and degradation of the basement membrane also leads to leakage of toxic factors from blood vessels, serving as a pathological barrier to the IPAD route (Arbel-Ornath et al., 2013). Deposition of Aβ in the vascular wall caused by impaired IPAD drainage can act as a feedforward mechanism to further impede Aβ clearance capacity, thus creating a vicious cycle and improving cognitive dysfunction (Arbel-Ornath et al., 2013). The pathological changes of IPAD following cerebral ischemia are summarized in Figure 3.

Yanev et al. (2020) investigated the relationship between meningeal lymphatic vessels and stroke. Meningeal lymphatic vessels dilated in a mouse MCAO model, although there was no statistically significant change. In mice with lymphatic dysplasia, after ischemic stroke, the increased volume of the infarct and the deteriorated neurological deficits suggested that the lymphatic vessels played a pivotal role in brain injury (Yanev et al., 2020). In particular, lymphatic vessel dilation can also enhance the magnitude of neuroinflammation after ischemic stroke. The proliferation of lymphatic endothelium in cervical lymph nodes after stroke was mediated by the signaling of the vascular endothelial growth factor C/vascular endothelial growth factor receptor 3 (Esposito et al., 2019). Activated lymphatic endothelium could stimulate macrophages to secrete a variety of pro-inflammatory factors, such as tumor necrosis factor-α and interleukin-1β, thus improving immune injury and breakdown of the BBB. Additionally, Esposito et al. (2019) found that cervical node lymphadenectomy reduced the volume of infarction and the immune response in the brain. The study of vascular regeneration after ischemic stroke in zebrafish had some unexpected findings (Chen et al., 2019). Time-lapse imaging revealed that the meningeal lymphatics rapidly grew into the injured parenchymal region and became lumenized. Tracers injected into the ventricle could be detected in the new lymphatic vessels, which indicated the normal capacity of ISF drainage and the resolution of cerebral edema. At the same time, new blood vessels grew along the trace of lymphatic vessels, restoring blood supply to the brain parenchyma. After vasculature regeneration, lymphatic cells experienced apoptotic cascades and the brain parenchyma returned to lymphatic-free status (Chen et al., 2019). Further study by this team showed that early angiogenesis depended on the direct transformation of the ingrown lymphatic vessels, which gradually exchanged to express endothelial-specific markers and communicated with blood flows in the surrounding residual blood vessels (Chen et al., 2021).

AQP4 changes in ischemic stroke

AQP4 is the most abundant aquaporin in the brain, including two isoforms M1 and M23 (Salman et al., 2022). AQP4 expression is highly concentrated in astrocyte endfeet surrounding blood vessels, while localization in cell bodies is less common. The polarity of AQP4 is related to dystrophin associated complex, which helps anchor AQP4 to the vascular basement membrane (Salman et al., 2022). AQP4 is crucial to the physiological function of the glymphatic system, providing a pathway for the exchange of CSF and ISF driven by an osmotic pressure gradient, therefore the proper expression and distribution of AQP4 ensure the efficiency of the glymphatic system.

Due to the important role of AQP4, its changes in ischemic stroke have been a constant concern. The spatial and temporal profile of AQP4 expression has been demonstrated in the transient MCAO model. One hour after stroke onset, the expression of AQP4 in astrocyte endfeet is rapidly up-regulated in the area of the lesion (Ribeiro Mde et al., 2006). At 48 hours, AQP4 expression reaches a second peak and its distribution shifts to the astrocyte soma membrane. Two peaks of brain edema are also observed coinciding with increased AQP4 expression, suggesting that AQP4 might serve as the main route of fluid movement after cerebral ischemia (Ribeiro Mde et al., 2006). In other models of ischemic stroke, such as multiple microinfarcts and neonatal stroke, increased expression of AQP4 could also be observed and normalization occurs 28 days later (Badaut et al., 2007; Wang et al., 2012). A recent study revealed the mechanism of altered subcellular localization and increased expression of AQP4 after hypoxia (Kitchen et al., 2020). Calcium influx in astrocytes activates calmodulin and protein kinase A, leading to the phosphorylation of the carboxyl terminus of AQP4, which facilitates the relocalization of AQP4 from intracellular vesicles to the cell membrane. Protein kinase A can also mediate nuclear localization of the transcription factor Foxo3a, which directly activates the AQP4 gene and leads to increased expression of AQP4 (Kitchen et al., 2020).

How do these changes affect cerebral edema after stroke? The exact answer may depend on the different time course of the disease. After intracerebral vascular occlusion, cerebral edema progresses from cytotoxic to vasogenic edema over time. Cytotoxic edema is caused by the imbalance of ion gradients and subsequent accumulation of intracellular fluid, while vasogenic edema occurs when leakage of the BBB allows plasma components to flow into the interstitial space. Early studies have confirmed that AQP4 can facilitate both water flow into the brain in cytotoxic edema and excess water removal from the brain in vasogenic edema (Manley et al., 2000; Papadopoulos et al., 2004). Conflicting outcomes in AQP4 knockout mice after ischemic stroke also showed the complexity of AQP4 functions. Research also supports the neuroprotective effect of AQP4 deficiency. In the study by Mestre et al. (2020), AQP4 knockout did not exhibit edema within the first 15 minutes after MCAO. After 24 hours of permanent ischemia, in AQP4 deletion mice neurological outcomes showed improvements implied by the reduced mortality rate and deficit score (Papadopoulos et al., 2004). However, other studies held the opposite view and suggested that the lack of AQP4 may hinder the clearance of edema fluid and aggravate neurological injury. For example, after 72 hours of reperfusion, the AQP4^–/– mice showed increased mortality and neurological deficits (Zeng et al., 2012). Therefore, early up-regulation of AQP4 after ischemic stroke accounts for the development of cytotoxic edema, while a later increase in AQP4 expression may partly serve as a self-protection mechanism to improve water clearance in the phase of vasogenic edema.

Apart from increased expression, loss of AQP4 polarity has also been observed. After an ischemic stroke, the pattern of AQP4 distribution transfers from the perivascular endfeet to the entire astrocytic membrane, which may compromise solute clearance (Wang et al., 2012). Mislocalization of AQP4 is related to enhanced amyloid pathology and neuroinflammation, resulting in vascular cognitive impairment (Back et al., 2017; Lyu et al., 2021). In the ischemic brain hemisphere, the expression of AQP4-M1 increased (Hirt et al., 2009), which could promote astrocyte migration and glial scar formation, but is not conducive to perivascular polarization of AQP4 (Smith et al., 2014). AQP4 also interacts with various ion channels on astrocytes, such as transient receptor potential vanilloid 4 and sulfonylurea receptor 1-transient receptor potential melastatin 4, which also influence water transport through astrocyte membranes and facilitate the edema development (Salman et al., 2022).

Potential Therapies Aimed at Lymphatic Drainage Systems

Today, classical strategies such as fibrinolysis and endovascular therapies are still the main treatment methods for ischemic stroke. New treatments are being explored, but their efficacy is not known. Restoring the function of lymphatic drainage systems can improve the prognosis of cerebral ischemia, and some representative drugs are summarized in Table 1. Since AQP4 promotes the formation of cytotoxic edema in early stroke (Mestre et al., 2020), early inhibition of AQP4 function can slow the entry of fluid into the brain parenchyma. There have been multiple attempts to identify AQP4 inhibitors that reduce AQP4-mediated water transport, some of which have shown positive effects. For instance, acetazolamide can inhibit human AQP4-mediated water permeability by 80% (Huber et al., 2007). In a rat model of cerebral infarction caused by bilateral carotid artery ligation, acetazolamide administration reduced brain water content and improved pathological changes of infarction (Hao et al., 2022). 2-(Nicotinamide)-1,3,4-thiadiazole, TGN-020, was another potent AQP4 inhibitor (Igarashi et al., 2011). As assessed by MRI, a single treatment with TGN-020 before ischemia significantly reduced infarcted volume and brain edema (Igarashi et al., 2011). The study by Pirici et al. (2017) demonstrated the beneficial role of TGN-020 more thoroughly, including better motor performance, reduced albumin extravasation, reduced gliosis, and reduced apoptotic cells in the treated group. After high-throughput screening for AQP4 inhibitors, AER-270 was selected and its efficacy was confirmed in controlling cerebral edema and improving neurological outcomes in ischemic stroke (Farr et al., 2019).

Table 1.

Potential therapies targeting lymphatic drainage systems

Target	Mechanism	Representative drugs	Effect	Reference
Glymphatic system: AQP4	Inhibits AQP4 mediated water transport	Acetazolamide	Reduces the brain water content and improves pathology caused by infarction changes	Hao et al., 2022
		TGN-020	Reduces brain edema and infarcted volume, improves motor performance, reduces albumin extravasation, gliosis, and apoptotic cells	Igarashi et al., 2011; Pirici et al., 2017
		AER-270	Controls cerebral edema and improves neurological outcome	Farr et al., 2019
	Downregulates AQP4 expression	miR-29b	Reduces infarct volume and edema formation, attenuates blood-brain barrier disruption	Wang et al., 2015
		Methylene blue	Ameliorates brain edema, astrocytic swelling, and pathological changes of ischemia	Shi et al., 2021
		Trifluoperazine	Reduces total brain water content during the acute phase of stroke	Sylvain et al., 2021
	Promotes AQP4 repolarization	Digoxin	Attenuates brain injury and cognitive dysfunction after cerebral hypoperfusion	Cao et al., 2022
IPAD pathway: SMC	Promotes arterial dilatation and modulates SMC function	Cilostazol	Facilitates Aβ drainage along the IPAD pathway, reduces Aβ accumulation in the brain	Maki et al., 2014
		Fasudil hydrochloride	Increases IPAD drainage and the number of Aβ40-containing vessels	Nizari et al., 2021
		Taxifolin	Restores cerebrovascular reactivity and promotes Aβ clearance from brain to blood	Saito et al., 2021

AQP4: Aquaporin-4; Aβ: amyloid β; IPAD: intramural periarterial drainage; SMC: smooth muscle cell.

Drugs may also influence ischemic injury by interacting with AQP4 expression. Gene transfer of miR-29b downregulated AQP4 expression, thereby reducing infarct volume and edema extent in MCAO mice and attenuating BBB damage (Wang et al., 2015). Based on a similar mechanism, trifluoperazine showed the ability to reduce brain water content during the acute phase of stroke (Sylvain et al., 2021). By inhibiting AQP4 expression through the extracellular signal-regulated kinase 1/2 pathway, methylene blue showed a neuroprotective effect on ischemic stroke and improved cerebral edema and ischemic pathology (Shi et al., 2021). Conversely, restoration of AQP4 polarity is also a potential therapeutic approach to facilitate the clearance of metabolites from infarcted areas; however, few drugs have been reported to have this effect. Experimental studies have demonstrated that slit protein homologue 2 (He et al., 2020) and caveolin-1 (Filchenko et al., 2020) play a role in maintaining perivascular AQP4 expression after cerebral ischemia. Digoxin accelerates AQP4 polarization, thus promoting recovery of glymphatic function after hypoperfusion (Cao et al., 2022).

In addition to AQP4, other factors modulate glymphatic function. Basement membrane fibrosis induced by transforming growth factor-β leads to impaired CSF-ISF exchange and poor functional recovery in MCAO mice (Howe et al., 2019). Therefore, inhibition of the transforming growth factor-β receptor may be beneficial for cerebral ischemia. Adrenergic receptor antagonism facilitates fresh CSF influx and normalizes extracellular potassium concentration, thus inhibiting cortical SD wave propagation and improving stroke outcome (Monai et al., 2019).

Improving IPAD drainage is also beneficial in preventing the adverse prognosis of stroke, especially vascular cognitive impairment. The treatment strategies targeting IPAD work by regulating the function of SMC and promoting arterial dilation. Cilostazol is an inhibitor of phosphodiesterase-3 in SMC, which can modulate SMC function by upregulating the signaling pathways of cyclic adenosine monophosphate and cyclic guanosine monophosphate (Maki et al., 2014). Cilostazol treatment has been shown to facilitate Aβ drainage along the IPAD pathway and reduce Aβ accumulation in the brain (Maki et al., 2014). Another drug, fasudil hydrochloride acts by stimulating the endothelial nitric oxide synthase pathway and promoting arterial dilation (Nizari et al., 2021). The increase in the number of blood vessels containing Aβ₄₀ in animals treated with fasudil suggested an improvement in IPAD drainage (Nizari et al., 2021). Taxifolin exerts therapeutic effects on cerebral amyloid angiopathy in many ways. In addition to its antioxidant and anti-inflammatory effects, taxifolin also restores cerebrovascular reactivity and promotes Aβ clearance from the brain (Saito et al., 2021). The application of these drugs after ischemic stroke warrants further exploration in future experiments. The therapeutic targets for lymphatic drainage systems are indicated in Figure 3.

Conclusions and Perspectives

Three anatomically and functionally related lymphatic drainage systems in the brain, including glymphatic system, IPAD pathway and meningeal lymphatic vessels, were discussed in this review. The glymphatic system involves the perivascular CSF influx and ISF efflux pathway, which promotes metabolite clearance and maintains fluid balance. Arterial pulsation and sleep are the most studied regulatory mechanisms. Meanwhile, the role of the circadian rhythm cannot be ignored. Compared to conventional fluorescence imaging in vivo or in vitro, MRI has more advantages in the detection of the human glymphatic system. DTI, a noninvasive real-time monitoring method, is used to explore the function of the glymphatic system and will contribute to a wider application in daily clinical work in the future, which may help identify patients at high risk of a poor prognosis after ischemic stroke.

The IPAD pathway and the glymphatic outflow pathway work together to drain ISF from the deep cerebral parenchyma out of the brain. Meningeal lymphatic vessels are important routes through which CSF and ISF communicate with the extracranial lymphatic system. Drainage of antigen and immune cells also reveals the importance of the lymphatic vessels on inflammation and the immune response. The characteristics of each lymphatic drainage system have been extensively studied; unfortunately, few studies have simultaneously monitored all three systems and determined their relative contributions to intracerebral fluid drainage. This may be limited by current detection technologies; therefore, better exploration tools for comprehensively studying these lymphatic drainage systems and their interactions need to be developed in the future.

The impairment of the glymphatic system has been proven to be a crucial mechanism in the pathological process of ischemic stroke. According to experimental results, perivascular space enlargement and increased glymphatic flow during early cerebral ischemia could be the potential source of brain edema. In contrast, the decrease of glymphatic inflow and outflow in the later stages aggravates the accumulation of toxic solutes and deteriorates cognitive function. Therefore, it is necessary to intensively explore the entire time course of glymphatic changes. IPAD and meningeal lymphatic vessels are also altered in response to ischemic stroke. Enhancing meningeal lymphatic drainage function is beneficial for improving cerebral edema but is harmful for controlling the extent of the inflammatory injury. Therefore, the long-term effects of meningeal lymphatic drainage on ischemic damage need to be further evaluated.

Restoring the normal functions of the lymphatic drainage systems can improve neurological outcomes. Most studies provide evidence that AQP4-knock-out animals exhibit reduced cerebral edema after ischemic stroke compared to wild-type animals. Therefore, AQP4 inhibitors may have neuroprotective effects if administered during the cytotoxic edema phase. Despite many attempts, the development of a safe AQP4 inhibitor for humans is still far from being achieved. Interestingly, the expression and distribution of AQP4 at the subcellular level may be a promising new target for drug development in the future; however, this pathway also requires more exploration of the potential molecular mechanism involved. Today, few studies targeting IPAD and the meningeal lymphatic system have been reported for the treatment after ischemic stroke, and further exploration of key regulatory sites and modulation mechanisms may offer some improvement.

There are some limitations in this review. First, other pathways of CSF clearance from the brain, such as arachnoid granulations and perineural routes through the cribriform plate, were not fully discussed. Second, less attention was paid to the role of intracerebral lymphatic drainage systems in brain injury after hemorrhagic stroke. Third, the glymphatic influx may also serve as a thrilling route for drug delivery to the brain parenchyma, which has not been thoroughly addressed in this review.

In summary, despite the existence of unresolved questions, the discovery of lymphatic drainage systems still opens a new avenue for the treatment of ischemic stroke.