Possible Role of Glymphatic System of the Brain in the Pathogenesis of High-Altitude Cerebral Edema

Introduction

Edema of the brain occurring after an exposure to high altitude, the so-called high-altitude cerebral edema (HACE), is usually considered to be an extension of acute mountain sickness (AMS), although some experts emphasize that there are significant differences between advanced AMS and HACE (Hackett and Roach, 2004; Davis and Hackett, 2017). Nonetheless, although in less severe cases, AMS presents with symptoms comprising headache, sleep disturbances, fatigue, dizziness, and cognitive impairment, HACE is a life-threatening disease manifesting with nausea, vomiting, ataxia, and hallucinations, followed by coma and death.

Although it is well known that this type of cerebral edema is primarily caused by hypobaric hypoxia, precise mechanisms responsible for this pathology remain elusive. Consequently, there are no other effective treatments for HACE than administering oxygen and descending to a lower elevation. Dexamethasone is recommended in HACE patients as an adjunctive treatment, especially if rapid evacuation is not possible (Davis and Hackett, 2017).

In this review, we summarize current knowledge on HACE with focus on the recently discovered glymphatic system of the brain and possible role of this system in the pathogenesis of this type of edema. We also suggest pathways of future research. Still, we emphasize that research in the field of association of an altered function of glymphatic system with HACE is in its infancy; there is little evidence supporting such a link and all conjectures presented in this article should be considered as highly hypothetical.

Glymphatic System of the Brain

Discovery of the glymphatic system of the brain was inspired by an observation that cerebrospinal fluid fluxes into the brain parenchyma along the perivascular spaces (the so-called Virchow–Robin spaces) surrounding the penetrating cerebral arteries. Further research has revealed that water from the periarterial space is actively transported to the interstitial space through aquaporin-4 (AQP-4) water channels. Then, interstitial fluid is cleared through AQP-4 water channels to the perivascular spaces surrounding cerebral veins. Since AQP-4 molecules are expressed on the astrocytes and these glial cells play the main role in this process, researchers coined the term “glymphatic” to emphasize the importance of glial cells serving purpose of the lymphatics.

It seems that most of the water found in the cerebral parenchyma enters this space through AQP-4 channels. Also, water molecules primarily leave intercellular space through AQP-4 channels expressed by astrocytes at the venous side of the perivascular space (Thrane et al., 2014; Jessen et al., 2015; Simka, 2015; Plog and Nedergaard, 2018).

Discovery of astroglial-mediated interstitial fluid bulk flow has resolved the enigma how water enters and leaves cerebral parenchyma. Generally speaking, water cannot freely flow into the extracellular cerebral space as it can in other tissues. Endotheliocytes in blood vessels of the brain that make up the blood–brain barrier, which is responsible for maintaining the homeostasis of the central nervous system, differ greatly from those in the periphery. They are characterized by the presence of tight junctions, low expression of adhesion molecules, lack of fenestration, and minimal pinocytotic activity. Of note, a passive flow of water from the Virchow–Robin space, because of the distance between perivascular space and majority of neurons, cannot be effective. It has been suggested that an altered cerebral interstitial fluid bulk flow plays an important role in several brain pathologies, including neurodegenerative disorders (such as Alzheimer and Parkinson diseases) (Jessen et al., 2015; Plog and Nedergaard, 2018) and also some types of cerebral edema (Thrane et al., 2014; Lawley et al., 2016; Tang and Yang, 2016).

Therefore, it is likely that water congestion in the settings of HACE is associated with impairment of the glymphatic system. Consequently, we propose that the “glymphatic” model of this disease could provide a useful and potentially rewarding framework for future research. A role for the glymphatic system in the pathogenesis of HACE has already been hypothesized by Lawley et al. (2016). In our article, we describe new findings supporting this hypothesis, suggest what could be investigated in the future, and which possible obstacles should be overcome by research in this field.

Findings Supporting the Role for Astrocytic AQP-4 Water Channels in the Pathogenesis of HACE

Acetazolamide, a carbonic anhydrase inhibitor, is the main pharmacological agent used in the prevention of AMS (Kayser et al., 2012; Swenson, 2014, 2016; Wang et al., 2015; Davis and Hackett, 2017). Yet, contrary to milder forms of AMS, efficacy of acetazolamide in the prevention of HACE has not been unequivocally proven, mostly because of the fact that HACE is a rare pathology and is difficult to replicate in laboratory settings. In addition, at least regarding cerebral component of AMS, it seems unlikely that this disease is simply a consequence of hypoxia of the brain. Characteristics of cerebral symptoms of AMS, which are different from those resulting from normal short-term asphyxia and connection of AMS with sleep at high altitude, indicate that more complex mechanisms than neuronal hypoxia play a role here.

Importantly, in addition to carbonic anhydrase, AQP-4 is an alternative target for acetazolamide. Crystallographic experiments have revealed that blocking of AQP-4 by this pharmaceutical agent occurs at the extracellular pore entrance of the AQP-4 molecule (Kamegawa et al., 2016)—thus, in the brain, at the perivascular side of astrocytic end feet.

Recent investigations have shown that water molecules cannot passively move into cerebral parenchyma. Majority of water enters intercellular space of the brain through AQP-4 water channels expressed on glial cells, the astrocytes. These water channels not only enable exchange of water between cerebrospinal fluid and cerebral parenchyma but also facilitate cleansing of the cerebral tissue from waste products (Jessen et al., 2015; Plog and Nedergaard, 2018).

Taking into account these findings, we suggest that it is possible that HACE does not only result from breakdown of the blood–brain barrier, as has been already demonstrated (Lafuente et al., 2016), but an altered function of astrocytic AQP-4 water channels also plays here a significant role.

Although disruption of the blood–brain barrier seems to be a prerequisite of cerebral edema, water molecules—irrespective whether they come from the lumen of capillaries or from the Virchow–Robin space—should enter cerebral parenchyma through these water channels. Therefore, expression and activity of AQP-4 should—at least theoretically—influence the formation of cerebral edema in the settings of hypobaric hypoxia.

Although the glymphatic system has not yet been studied in the context of HACE, there are some findings that make this hypothesis credible. First, AMS is strongly associated with the sleep at high altitude. A common way to avoid AMS is the “climb high, sleep low” rule. It thus seems that an awake brain is less sensitive to hypoxia at high altitude than an asleep brain, which is counterintuitive, since oxygen consumption by this organ is higher in the state of consciousness. Yet, this enigma could be explained by the fact that the flow of water through AQP-4 water channels occurs primarily during sleep (Xie et al., 2013) and, therefore, an asleep brain would be more prone to develop edema.

Second, there are also some recently performed animal experiments that revealed a role of AQP-4 water channels in the pathogenesis of HACE. It has been found that in rodents, hypobaric hypoxia was associated with an increased expression of AQP-4 by astrocytes (Gong et al., 2018; Sun et al., 2018; Wang et al., 2018). A similarly increased AQP-4 astrocytic expression has also been demonstrated in mice exposed to intermittent hypoxia (Wang et al., 2018). However, in another animal experiment, an increased expression of AQP-4 by astrocytes was seen only in rodents that received a combination of hypobaric hypoxia and proinflammatory stimulus (an injection of lipopolysaccharide). In this experiment, researchers have also revealed an increased permeability of the blood–brain barrier and decreased expression of proteins responsible for the integrity of this barrier after combined exposure to hypobaric hypoxia and lipopolysaccharide (Zhou et al., 2017).

Taking into account these findings, we hypothesize that HACE, in addition to the breakdown of the blood–brain barrier, may be associated with an altered astroglial-mediated interstitial fluid bulk flow, known as the glymphatic system.

Suggested Pathways of Future Research

Inhibition of AQP-4 by acetazolamide in the context of anatomical structure of the glymphatic system (Jessen et al., 2015; Plog et al., 2018) may explain low efficacy of this pharmaceutical agent in the prevention of HACE (Swenson, 2014). On the one hand, inhibition of these water channels in the astrocytes surrounding cerebral arteries would decrease influx of water into interstitial space, but at the same time blocking of AQP-4 at the venous side of the glymphatic system would capture water. Thus, a net effect would be close to nil.

Theoretically, a selective inhibition of periarterial AQP-4 molecules should solve this problem; however, periarterial and perivenous AQP-4 molecules probably do not exhibit different conformation and, therefore, such a selective inhibition would not be possible, even by other carbonic anhydrase inhibitors exhibiting blocking capability of the AQP-4 molecule. Therefore, pharmaceutical management of HACE should primarily focus on the prevention of excessive activation of the glymphatic system at high altitude.

It has been suggested that functioning of the glymphatic system may be influenced by venous outflow through the internal jugular veins (Simka, 2013, 2015; Zivadinov and Chung, 2013). It has also been found that some otherwise healthy people present with malformed valves or other obstructing pathologies of these veins, the so-called chronic cerebrospinal venous insufficiency (Zivadinov et al., 2011). Yet, these venous abnormalities are more often seen in patients with neurodegenerative and neuroinflammatory diseases (Parkinson disease and multiple sclerosis) (Zamboni et al., 2009; Zamboni and Galeotti, 2010; Simka et al., 2011; Zivadinov et al., 2011; Simka, 2013). Contrary to intracranial venous pathology (Wilson and Imray, 2016), of as yet a role for impaired flow in the extracranial part of the internal jugular veins in the pathogenesis of HACE has not been studied. However, it is tempting to speculate that venous congestion within cerebral venous microvasculature in the settings of chronic cerebrospinal venous insufficiency, caused by a blockage at the level of jugular valve, can result in a higher risk of developing HACE.

If it were the case, then high-altitude trekkers and mountaineers presenting with this venous abnormality (abnormal jugular valves and impaired flow in the internal jugular veins can be quite easily revealed using standard ultrasound scanner) (Bastianello et al., 2014), to avoid HACE, should be advised to ascend slower than it is usually recommended. Of note, a compression of cerebral sinuses and large cerebral veins, as well as small cerebral veins, has already been suggested to play a role in the pathogenesis of HACE (Lafuente et al., 2016; Sagoo et al., 2016).

Activation of the glymphatic system during sleep is potentially the most promising target in this field. Currently it is not precisely known which mechanism actually activates AQP-4 water channels in the asleep brain. Some researchers suggested that an increased pulsatility of cerebral blood vessels during sleep activates this system (Jessen et al., 2015; Lawley et al., 2016; Plog and Nedergaard, 2018). Even if it actually were the case, what is responsible for this enhanced pulsatility remains elusive. Most likely pulsatility is primarily regulated by the cerebral norepinephrine system through specialized neurons located in the locus coeruleus and could be influenced by targeted pharmacotherapy (Russak et al., 2016; von Holstein-Rathlou et al., 2018).

Angiotensin II inhibitors, such as losartan, represent a group of potentially beneficial pharmaceutical agents that affect norepinephrine axis of the brain. A preliminary animal study has demonstrated an inhibition of glymphatic flow by losartan injected into the cisterna magna (Russak et al., 2016). Of note, it has already been demonstrated that angiotensin-converting enzyme plays a role in the pathogenesis of AMS (Enhao and Huang, 2016). Although inhibition of the renin–angiotensin–aldosterone system by losartan had no observable effect on exercise performance at high altitude (Myers et al., 2017), a role for this drug and other angiotensin-converting enzyme inhibitors in the prevention of HACE cannot be ruled out and should be studied, at least in animal experiments.

Activation of the glymphatic system can also be regulated by exercise. Animal studies demonstrated a decreased activity of the glymphatic system in active awake mice (Xie et al., 2013) probably due to increased activity of the noradrenergic system of the brain. In contrast, basic activity of the glymphatic system was higher in mice that ran regularly, still with normal clearance of interstitial space and no congestion within cerebral parenchyma (von Holstein-Rathlou et al., 2018). It, therefore, remains unclear which type and timing of physical activity in humans, especially in relation to sleep, would be beneficial for the prevention of HACE.

The last area of suggested research is the influence of body posture during sleep on the development of HACE. It has been shown that in rodents, the glymphatic system is differently activated depending on body position during sleep (lateral vs. prone position) (Xie et al., 2013; Lee et al., 2015). If this phenomenon also occurs in humans, it would be possible to regulate activity of the glymphatic system by body position during sleep. This may be of particular importance for mountaineers who are often forced to sleep at high altitude, which—especially after incomplete acclimatization—can result in life-threatening HACE.

Finally, it should be emphasized that a treatment and prevention of HACE could probably be achieved by modification of the functioning of the glymphatic system, but not through its complete blockage. A total obstruction of AQP-4 water channels would probably result in cessation of the cleansing of the neurons from waste products, equally life threatening as HACE, or even in a worsened cerebral edema (Papadopoulos et al., 2004).