Air pollution, Glymphatic Impairment, and Alzheimer's Disease

Abstract

Epidemiological evidence demonstrates a link between air pollution exposure and the onset and progression of cognitive impairment and Alzheimer's disease (AD). However, current understanding of the underlying pathophysiological mechanisms is limited. This opinion article examines the hypothesis that air pollution-induced impairment of glymphatic clearance represents a crucial etiological event in the development of AD. Exposure to airborne particulate matter (PM) leads to systemic inflammation and neuroinflammation, increased metal load, respiratory and cardiovascular dysfunction, and sleep abnormalities. All of these factors are known to negatively impact glymphatic clearance. Rescuing glymphatic function by restricting the impact of causative agents, and improving sleep and cardiovascular system health, may increase the efficiency of waste metabolite clearance and subsequently slow down the progression of AD. In sum, we introduce air pollution-mediated glymphatic impairment as an important mechanistic factor to be considered while interpreting the etiology and progression of AD as well as its responsiveness to therapeutic interventions.

Main text:

Glymphatic Impairment as a Potential link Between air Pollution Exposure and Neurodegeneration in Alzheimer’s Disease and Related dementias

Alzheimer's disease (AD) is one of the most common progressive neurodegenerative diseases. It is characterized primarily by dementia and cognitive impairment, with anatomical signatures of excessive Aβ and phosphorylated tau (pTau) deposition in the brain that worsens with age [1]. The aging population is rapidly expanding globally, and so is the prevalence of AD. Concurrently, in many parts of the world, air quality is worsening. This opinion article builds on the many observations that have established an association between exposure to air pollution, AD and dementia/cognitive decline in humans [2–6]. We will critically examine a novel hypothesis that inhaled PM, particularly the ultrafine particles (UFP) in ambient air, can cause glymphatic impairment, linking air pollution exposure with an increased risk of developing AD. While other components of air pollution (including ozone, nitrogen oxides, and coarse particles) should also be considered in terms of addressing the burden of AD and related dementias, space constraints preclude their discussion in this present article.

We will first briefly review how airborne particles enter the body and their fate after deposition in the lungs. We will then discuss the physicochemical and biological mechanisms by which PM can cause adverse outcomes in the brain, the role of the glymphatic system in AD, and how PM accumulation could putatively perturb the glymphatic system to initiate or accelerate AD-related pathology in the brain.

Air pollution: Types, Sources, and Epidemiology

Air pollution is a global threat to human health and the environment, with an estimated ~7 million deaths per year due to combined indoor and outdoor exposures [7,8]. It is a temporally and spatially inhomogeneous mixture of airborne particles (e.g., metals, metal oxides, carbons), gases, and semi-volatile constituents. The sources of these contaminants include motor vehicle exhaust, industrial emissions, volcanic eruptions, wind and water erosion, and wildfires [9]. Particulate matter (PM) can be separated into three subgroups based on airborne size distribution: coarse particles, fine particles (PM_2.5, <2.5 μm), and ultrafine particles (UFP, <0.1 μm) [10]. PM_2.5 and UFP are transported in air over long distances and are significant contributors to the adverse effects of air pollution on human health [11]. PM_2.5 and UFP are largely combustion-derived, and have carbonaceous cores with other organic species, salts, and metals/metal oxides (e.g., Fe, Cu, Pb, Zn) adhered to their surfaces or as part of the aerosol mixture [12–14].

The epidemiological literature supports consistent and significant associations between air pollution exposures, particularly to PM, and cardiopulmonary diseases in humans (including atherosclerosis, stroke, ischemic heart disease, heart failure, asthma, chronic obstructive pulmonary disease, and lung cancer) [15–18]. Climate change could exacerbate air pollution-related health effects by increasing concentrations of airborne pollutants and inducing heat-related physiological dysregulation, which could act synergistically on the same adverse response pathways [19,20]. PM exposure is especially relevant in CNS disorders because of its composition-dependent redox reactivity (stemming for instance from metals and organics), which can lead to particulate-induced inflammation. Indeed, epidemiological evidence suggests an increased prevalence and incidence of AD, dementias, and cognitive impairment in association with exposure to PM_2.5 and traffic-related aerosols, a significant source of UFP [2–6,21].

PM Inhalation, Systemic Effects and Translocation to the Brain

Airborne PM enters the body primarily via the respiratory tract, where it deposits in the nose, conducting airways, and alveolar (gas exchange) region, depending on aerosol diameter and other properties (Fig. 1). Studies in rodents reveal that UFP, upon deposition in the alveoli, rapidly cross the alveolar-capillary barrier to reach the systemic circulation via which they are distributed to other tissues [22,23] (Fig. 1). Particles can also move out of lungs via lymphatic vessels to accumulate in local lymph nodes. Interestingly, a recent study showed an age-related decline in lung-associated lymph node immune function as a result of the accumulation of inhaled particulate matter [24]. Inhaled poorly-soluble particles also make their way to the gastrointestinal tract (GI) via mucociliary clearance [25] (Fig. 1), where the smallest particles can pass through the gut barrier and be transported to other organs [26].

Fig. 1: Airborne particulate matter exposure and entry into the brain.

Left: Small airborne particles (UFP) generated for instance via combustion processes can translocate to the brain directly after depositing in the nose and uptake by olfactory mucosal nerves. Right, middle: UFP also deposit efficiently in the gas-exchange region of the lung, where they can translocate to local lymph nodes. UFP that can make their way from the lung to the bloodstream are distributed to other distant organs in the body including the brain. Right, top: Inhaled UFP can induce systemic inflammation and may interact directly with the endothelial cells of the BBB, disturbing tight junctions and gaining access to the brain parenchyma. Right, bottom: A fraction of UFP in the gastrointestinal tract is also able to make its way to the bloodstream or mesenteric lymphatics via villi and/or breaching the Peyer’s patches.

Systemic inflammation is a well-documented consequence of exposure to airborne PM, including UFP [27–31]. In addition to direct interactions with target cells or tissues, systemic effects can also result from the spill-over of lung inflammatory mediators, autonomic nervous system activation, or the production of small signaling peptides [32]. Systemic inflammation impacts the blood-brain barrier (BBB) [33,34] and, thus, may play a role in the accumulation of blood-borne UFP in the brain. The BBB is a dynamic physical and metabolic interface that allows the selective transfer of material to and from the brain. It consists of tightly packed endothelial cells with a continuous interface composed of tight junctions, which include proteins like occludin, claudin, and other junctional adhesion molecules. The level of tight junction proteins is a measure of BBB integrity. Of note, in mice, vehicle exhaust aerosol exposure for 30 days has been found to reduce the expression of tight junction proteins [35].

UFP in the circulatory system may also reach the brain via the choroid plexus, where the integrity of the blood-cerebrospinal fluid (CSF) barrier is believed to be leakier as compared to the BBB [36]. Although the size of the choroid plexus is much smaller as compared to cerebral vasculature of the brain, increased blood passage (3–5 times faster) [37] and relatively large apical surface area of the choroidal epithelium [38] make it a competitive route for UFP transfer from blood to the brain [39]. Intriguingly, relatively larger amounts of metals have been found in the choroid plexus as compared to the rest of the brain [40].

Evidence also shows that UFP, including metal particles, have the ability to pass through the nose and translocate to the olfactory bulb (OB) (Fig. 1, Top Left Panel). The olfactory mucosa is rich in nerve endings of olfactory sensory neurons, which communicate smell information to the brain. Studies in rodents and non-human primates demonstrate that UFP translocate along the axons of olfactory nerves into the CNS [41]. In humans, evaluations of postmortem olfactory tissue, including from children, have shown the accumulation of UFP [42]. Inhalation exposure studies using poorly-soluble UFP in rodents have been particularly useful for demonstrating olfactory transport of solid particles to the brain even though only small fractions of the deposited dose get transported via this pathway following short-term exposures [41,43,44]. In the human brain, UFPs have been found in olfactory bulb (OB) periglomerular neurons, intraluminal erythrocytes of the frontal lobe, and trigeminal ganglia capillaries [45–47]. Moreover, autopsy tissue from subjects with documented AD/ADRD also confirms the presence of UFP metals and metal oxides in the OB and other brain regions [48].

While UFP can make their way to brain, either directly or indirectly, little information is available about their fate within the brain, including dissolution and bioprocessing, how long they stay there, and their impact on the neuronal and glial populations. In principle, UFP, once in the brain parenchyma, can be taken up by resident immune cells (microglia) and by other cells like oligodendrocytes, astrocytes, and neurons. Autopsy studies in AD patients suggest that UFP damage the endothelial lining of cerebral vasculature and are linked to a breakdown of the perivascular matrix and enlargement of the Virchow-Robin/perivascular spaces [49]. Ultrastructural analyses reveal abnormalities in mitochondria and endoplasmic reticulum of nearby cells, specifically astrocytes, pericytes, and endothelial cells, as well as thickened basement membrane [50,51]. UFP-associated C-centered radicals and transition metals can initiate the formation of reactive oxygen species (ROS) in the parenchyma, leading to oxidative stress and inflammation [52,53]. Indeed, autopsy studies have revealed associations between pollution exposures and increased infiltration of macrophages into the brain [45].

UFP Distribution in the Brain and the Glymphatic System

Inside the brain, UFP are unlikely to be confined to a particular region, but, rather, transported to other parts of the brain by CSF flow. Unlike other organ systems, the brain lacks a fully functional lymphatic system, which is otherwise primarily responsible for clearing toxins, waste, and harmful metabolites. This function in the brain is performed by an analogous system that operates via perivascular CSF circulation named the glymphatic system [54] (See Box 1). This flow directionally (peri-arterial to peri-venous) drives wastes, metabolites, and non-desired products out of the brain (See illustration: Figure I) [54–56]. However, glymphatic transport may also serve as an instrument to spread UFP and their adverse effects to distant brain regions (Fig. 2).

Box 1: The Basic Principles of the Glymphatic Pathway.

The glymphatic system is a glia-dependent cerebrospinal fluid (CSF) transport system, which consists of 1) periarterial influx of CSF, 2) the exchange of CSF and interstitial fluid (ISF) in the brain parenchyma, 3) recollection of fluid, carrying waste products and metabolites, at the perivenous spaces and along white matter tracts including cranial/spinal nerves [117]. The system is facilitated by water channels (AQP4) expressed at high density in astrocyte end feet (Figure I). On a macroscopic level, CSF is produced in the brain by the choroid plexus located in the four ventricles [118–120] as well as at the blood-brain barrier interface [121]. Directional flow of CSF occurs from the ventricles to the cerebral aqueduct and basal cisterna, where it ascent along the periarterial spaces and enters into the brain parenchyma and mixes with ISF. Mixed CSF and ISF are recollected at perivenous spaces as well as by other outflow pathways, including arachnoid granulation [122,123], meningeal lymphatic vessels [124–126], cranial and spinal nerves [117,127], and olfactory nerves traversing the cribriform plate with drainage through the nasal mucosa [110,128–130]. A large fraction is finally collected by the cervical lymphatic vessels and returned to the venous circulation [55,71,78]. One of the hallmarks of brain injury and pathology is the loss of polarized expression of AQP4 in astrocytic vascular end feet resulting in a suppression of glymphatic flow [77]. The role and altered nature of the glymphatic system have been described in several neurological disorders in humans as well as animal models, including traumatic brain injury [77,131,132], meningitis [77,133,134], multiple sclerosis [135], amyotrophic lateral sclerosis [136], Parkinson's disease [137], and ADRD [78,111,138–141]. Moreover, aging is accompanied by reduced glymphatic function [142,143]. One of the most characteristic features of the glymphatic system is that it is highly active during sleep, i.e., its clearance efficiency is at a minimum in the awake state [54].

Figure I: Diagram outlining glymphatic clearance of waste and Aβ.

Aβ40/42 is produced within the brain parenchyma, and is cleared via peri-venous efflux as well as cervical lymphatic vessels. The periarterial inflow of cerebrospinal fluid that exchanges with interstitial fluid is facilitated by astrocytic AQP4 water channels. Similarly, outflow/recycling of CSF containing both soluble and insoluble waste metabolites occurs via meningeal and cervical lymphatic vessels to the lymph nodes.

Fig. 2: A fluid flow impairment as a result of PM exposure may exacerbate AD.

Left: Under normal physiological conditions, CSF moves within the periarterial spaces, exchanges with the interstitial fluid and is recollected by peri-venous spaces. In this path, CSF carries metabolic waste, Aβ, excessive pTau, and UFP and clears them from the brain. This process of waste clearance is facilitated by AQP4 at the astrocyte end feet. Macrophages within the peri-vascular spaces add an additional layer of surveillance; non-desired particles are either engulfed by macrophages or flushed out by CSF flow. Right: Long-term exposure to UFP putatively, results in increased amounts of solid particles within perivascular spaces, the polarized expression of astrocyte AQP4 decreases, and peri-vascular spaces narrow, impairing the clearance of metabolites and exogenous particles from the brain. Long-term fluid retention or malabsorption likely contributes to the enlargement of the periarterial spaces [113]. Direct binding of UFP with the protofibrils may facilitate the formation of Aβ plaques along the vasculature [114–116]. After breaching epithelial barriers, UFP may enter the brain parenchyma, which could initiate neuroinflammation, and further decrease the process of efflux of Aβ.

While in the interstitial spaces and perivascular network, UFP can bind with metabolic waste, proteins, and neurotransmitters. The binding of biomolecules to UFP surfaces may not only result in conformational changes, decreasing the solubility of waste as well as UFP, but also promote the growth of larger complexes [57], which are hard to eliminate. Similarly, this binding may also affect the neurotransmitter signaling, leading to excessive release. Increased levels of excitatory neurotransmitters have been observed in the brain following UFP exposure, which may be further amplified by the release of inflammatory mediators and cytokines [58].

UFP retention and detoxification depend on the local cellular environment, which can promote particle dissolution and reformation (bioprocessing). These processes are particularly complex in the case of metal UFP, as normal physiological functions depend on the homeostasis of many metal ions in the brain. Intriguingly, erroneous uptake, imbalance, and accumulation of Fe, Al, Si, Ti, Cu, Co, Ni, Zn, etc., in brain tissue has been observed and linked to amyloid beta (Aβ) overproduction, tau hyperphosphorylation, as well as to aggregation and buildup of Aβ and pTau [59]. Metals (e.g. Fe, Cu) are essential components of several enzymes, and their transfer across cell membranes is facilitated by special transporters like the Zrt-Irt-like protein (ZIP) metal transporters [60]. However, their excessive levels and accumulation as in AD is a source of redox-generated free radicals [61] and increases oxidative stress [62]. It should be noted that Aβ is a copper/zinc-metalloprotein that aggregates and further becomes redox-active in the presence of excessive amounts of Cu and Fe [63]. In addition, Cu is present at both the core and rim of AD senile plaques [64,65].

Studies in lungs and liver tissue suggest that UFP induce activation of the unfolded protein response and oxidative stress [66]. Moreover, studies in the fetal brain suggest that UPF possibly interact directly with newly synthesized proteins, including Aβ, and promote their misfolding. High induction of ER stress has been particularly noted in perivascular macrophages and reactive astrocytes around brain blood vessels [67].

While metal chelators seem to improve the clinical outcome of several neurodegenerative diseases, their mechanisms of action remain largely unclear and the effects of long-term use remain to be assessed [68]. Overall, further work is needed to clarify the system- and cellular-level processes that promote the clearance of UFP and metal ions from the brain, and how these processes are impacted by physiological comorbidities and genetic disposition (see Outstanding Questions). Studies in animal models and humans suggest that clearance efficiency is not only dependent on the production of waste and metabolites, but also on factors such as sleep, aging, and cardiovascular health [54,69–71].

Outstanding questions.

• Does the degree of glymphatic suppression following PM exposures exhibit individual variability? Can the heterogeneity in glymphatic suppression explain, in part, the variability in the prevalence and progression of AD? Which exposure metrics are best correlated with outcomes?

• Does the interrelationship between PM exposure and glymphatic impairment systematically differ between young and older populations?

• Is there a basis to expect more sudden or dramatic changes in glymphatic efficiency in genetically prone AD populations as a result of PM exposure?

• Sleep disturbances are early signs of AD; how accurately can glymphatic insufficiency be diagnosed by sleep pattern abnormalities in response to PM exposure?

• What physicochemical properties of PM are most strongly associated with glymphatic impairment?

• What perturbations in glymphatic function occur in animal models following inhalation exposures to PM – UFP, in particular – and what mechanistic insights can be gained from such studies?

• Is it possible to establish a relationship between PM exposure and the duration and degree of glymphatic impairment in animal models? Does the impact vary with age and other factors?

• What is the relationship between particulate metal accumulation and bioprocessing in brain tissue and glymphatic insufficiency?

• Can PM act as a seed for beta-amyloid plaques within the brain, and can this contribute to obstructing CSF circulation?

• How can the detrimental effects of PM exposure be limited? Exercise and sleep are among the enhancers of glymphatic function. Can these lifestyle factors, via their effects on glymphatic function, improve AD outcomes?

Aβ/Tau Aggregation and Glymphatic Impairment Hypothesis

AD is anatomically characterized by the deposition of extracellular Aβ plaques and intracellular neurofibrillary tangles [72–76]. Our work, as well as that from several other groups, shows that glymphatic function has a profound role in the clearance of the brain’s protein waste [54,71]. Its efficiency is reduced with aging and by Aβ and pTau deposition [71], as illustrated for instance by reduced glymphatic performance following intracerebral injections of exogenous Aβ [77]. Further, both the glymphatic system and Aβ and pTau levels are influenced by sleep: clearance of these metabolites is at the highest rate during sleep [56,78]. Brain injury and neuroinflammation similarly reduce glymphatic flow and can increase the burden of Aβ and pTau [77]. Reduced glymphatic flow in animal models of brain injury or AD exacerbates cognitive impairment [56,77]. Intriguingly, clinical studies suggest worsening of AD and other neurological disorders upon glymphatic impairment and sleep disruption [79–81].

At the physiological level, glymphatic flow is operated by pulsation of the arterial wall, and is therefore related to cardiac contractions [70], respiration and slow vasomotion [69,82,83]. Cardiovascular abnormalities such as hypertension can result in reduced arterial wall pulsations and suppression of CSF flow in perivascular spaces, and in the long-term, may lead to enlargement of perivascular spaces [70]. An important aspect of glymphatic flow is the recollection of CSF at perivenous spaces and lymphatic vessels within the meninges as well as along cranial and spinal nerves [84]. This intricate web joins together and the resulting larger lymphatic vessels drain into lymph nodes in the neck region [85]. CSF-containing waste also makes its way out of the brain via exiting nerves as well as the olfactory bulb and down to the nasal cavity [86].

It is plausible that UFP act as a seed that binds excess Aβ, resulting in structural conformational changes that lead to plaque formation. Similarly, attachment of Aβ fibrils to metal UFP could be a defensive mechanism that eventually leads to insoluble deposits and plaques. Senile plaque analysis in AD patients confirms Cu, Zn, and Fe as core elements [64,65]. Plaque growth is expected over time via interaction with microglia and astrocytes and, together with reduced glymphatic efficiency following air pollution exposures, may accelerate pathology. At present, there is little direct mechanistic evidence linking PM exposures to the changes in the brain that lead to Aβ plaque formation, neurodegeneration, or behavioral and cognitive changes. Cu and Fe, based on their oxidative reactivity and presence in airborne PM [87], could be good candidates for studies of the ability of UFP to initiate or accelerate the deposition of extracellular Aβ plaques or intracellular neurofibrillary tangles in vivo (see Outstanding Questions).

A far-reaching effect of air pollution exposure is sleep disturbance, as evidenced in humans via obstructive sleep apnea [88–92]. Studies have shown that PM may disturb sleep duration and quality by impacting brain homeostasis, the autonomic nervous system, and sleep centers [93,94]. AD patients also suffer from sleep and circadian rhythm disturbances: approximately 25–40% of them manifest a decrease in total sleep time, increased sleep fragmentation, and a reduction in time spent in rapid eye movement (REM) and slow-wave sleep (SWS) [95–98]. Increasing evidence suggests that sleep disturbances precede AD neuropathology, but are augmented by the onset of Aβ and neurofibrillary tau tangles [99]. Moreover, decreased sleep quality increases the risk of cognitive impairment, AD, and dementia in healthy older adults [100–103]. Intriguingly, glymphatic flow and clearance of metabolite is optimal during sleep [54] and sleep disturbance could facilitate the buildup of metabolites in the brain.

It is worth noting that optimal glymphatic function is largely dependent upon the vascular architecture and perivascular spaces (PVS). Enlarged PVS are a hallmark of many neurodegenerative diseases, most notable small vessel disease [104,105]. With improved resolution of neuroimaging techniques, PVS enlargement has been revealed in other neurodegenerative diseases as well, including AD [106,107], Parkinson’s disease, Huntington’s disease, and amyotrophic lateral sclerosis (ALS) [108,109]. Enlargement of PVS has also been noted in pollution-exposed AD patients [49], and is an important indication of glymphatic impairment [70]. However, the absence of high-resolution scans across large populations of pollution-exposed individuals renders the link between air pollution and enlargement of the perivascular inconclusive [110,111]. Intriguingly, many neurological disorders involve neuroinflammation, as well as the build-up of waste products and metabolites in the brain. Increased inflammation in Aβ has been shown to sometimes coincide with a buildup of metals in the brain, and one may speculate a causative interplay between UFP exposure, metal accumulation in the brain, and Aβ formation and pathology [112].

To summarize, we outlined three important downstream effects of PM exposure and uptake: disrupted sleep, neuroinflammation, and enlargement of perivascular spaces. These three elements are all known to negatively impact the glymphatic system. Thus, it seems plausible that glymphatic impairment is a common event triggered by the entry of particulates into the brain (Fig. 3).

Fig. 3: Multi-level effects of UFP in the brain and exacerbation of AD.

We propose that UFP, upon inhalation, trigger multiple reactions: olfactory bulb inflammation and neuroinflammation, BBB impairment, and perivascular changes including increased thickness of basement membrane and fragmentation [etc.]. These changes result in impaired glymphatic function, which is further augmented by peripheral inflammation, endothelial cell damage as well as impaired breathing related to sleep or circadian disturbances. Impairment of glymphatic function not only increases the amounts of waste and Aβ in the brain, but may facilitate protein aggregation which worsens over time and contributes to neurodegeneration and dementia in AD and related disorders.

Concluding Remarks and Future Perspectives

Acute and chronic exposure to particulate air pollution has devastating consequences in terms of human morbidity and mortality. There is growing evidence that UFP exposure is associated with neuroinflammation, disordered sleep, and neurodegenerative disease-related cognitive impairment. All these factors are closely linked to both AD and glymphatic impairment. Building on these connections, we proposed in this article that dysregulation of glymphatic flow represents a mechanistic link between particulate air pollution exposure – especially in the case of UFP – and the onset or progression of AD pathology. A deeper understanding of the pathophysiological basis of negative health outcomes following air pollution exposure, including neurodegeneration and AD, may lead to strategies for mitigating the health burdens and societal impacts of air pollution.

Highlights:

• Air pollution is a prime target for the prevention of neurological diseases. Epidemiological, neuropathological, and clinical evidence suggests increased prevalence and exacerbation of Alzheimer’s disease (AD) and related dementias in association with airborne particulate matter (PM) exposures.

• The smallest of airborne PM can enter the brain directly via the nasal mucosa and olfactory bulb or indirectly via the blood-brain barrier following their translocation across the lung and gastrointestinal epithelial barriers.

• Inside the brain, CSF circulation has a dual role: it supports, on the one hand, glymphatic dissemination and exchange with interstitial fluid, and on the other hand, removal of waste products. These processes depend upon multiple factors, particularly sleep and cardiovascular health.

• We propose that PM-mediated cardiovascular system impairment, neuroinflammation, and sleep disturbances suppress glymphatic waste clearance and, thus, may represent a critical intermediary step in PM-mediated initiation and progression of Alzheimer’s disease.