Fluid Flow Networks in Brain Disease: Linking Water Channel Orientation with Multiple Brain Systems in Neurodegeneration

Anchoring AQP4 requires the presence of the dystrophin-associated protein complex (DAPC). The DAPC anchors the astrocytic cytoskeleton to the vascular basement membrane. The C terminal PDZ binding site of AQP4 binds with α-syntrophin allowing AQP4 to become anchored in the DAPC. Dystrophin Dp71 acts as a bridge between the DAPC and the actin filaments. Agrin, laminin, and perlecan act as bridges between the DAPC and the ECM via β-dystroglycan [28]. All of these relationships are dynamic and respond to both mechanical and biochemical stimuli from the vasculature. Agrin derived from endothelium is important for creating the perivascular ECM niche where AQP4 is clustered. Aging, diabetes, or vascular stress reduce agrin and therefore weaken anchoring and lead to depolarization. Other scaffold components, including α-dystrobrevin, various syntrophin isoforms, spectrins, and ankyrins, contribute to the anchoring of the DAPC to the cytoskeleton and the recruitment of signaling partners and membrane curvature [29].

The function of AQP4 is influenced by isoform composition. Alternative translational start sites produce two primary isoforms, M1 and M23, which vary in their ability to form orthogonal arrays of particles (OAPs): M23 easily forms large OAPs, while M1 forms small oligomers limiting lattice growth [1,30].

The ratio of M1/M23 influences OAP size and stability, and thus influences water permeability and membrane organization. The ratio of M1/M23 is controlled by multiple mechanisms, including upstream open reading frames, RNA binding proteins controlling transcript stability and localization, and post-translational modifications affecting the oligomerization rate. OAPs preferentially form in perivascular microdomains aligned with the microvilli of astrocytes, facilitating the directionality of water flow [31,32]. The organization of OAPs allows for the interaction with Kir4.1 potassium channels and facilitates ion/water homeostasis. Impairment of the balance of isoforms or the formation of OAPs will disperse AQP4 throughout the plasma membrane, reduce the directionality of water transport, and reduce glymphatic clearance [33].

The exact delivery and retention of AQP4 at the endfeet require coordination of vesicle traffic. New synthesis of AQP4 is delivered from ER through the Golgi apparatus and is targeted to the perivascular membrane by Rab GTPase-regulated pathways, with Rab11 directing recycling endosomes to the astrocyte endfeet [34]. Kinesin and dynein motor proteins facilitate the transport along microtubules for longer distances, while actin filaments facilitate the final positioning of the membrane. The targeting of AQP4 depends on PDZ-containing sorting signals and, once in the membrane, AQP4 is confined to laterally segregated compartments. Lipid rafts rich in cholesterol and sphingolipids, caveolin, and flotillin assist the organization of these domains and constrain AQP4 diffusion [35]. Changes in lipid composition common to aging and metabolic disorders reduce the integrity of lipid rafts and increase the diffusibility of AQP4. Pathways for endocytosis also influence polarity: clathrin-dependent endocytosis removes incorrectly located AQP4, while the caveolae-dependent process controls the turnover of AQP4 at the endfeet. Repeated inflammation or oxidative stress will disturb these pathways and cause the widespread mislocation of AQP4 [36,37].

The polarity of AQP4 is determined by interrelated regulatory systems operating at the level of transcription, epigenetics, post-translational modification, and the cytoskeleton. Transcriptional regulation involves hypoxia-inducible factors, inflammatory transcription factors such as NF-κB, and circadian rhythm-related regulators such as BMAL1 and CLOCK [38]. Epigenetic regulation, including DNA methylation, histone modification, and the regulation of non-coding RNA affects the level and stability of transcripts. The post-translational modification of AQP4 modifies the behavior of channels: phosphorylation influences the organization and opening characteristics of OAPs; ubiquitination labels AQP4 for degradation if it is misplaced; S-nitrosylation modifies the interaction of AQP4 with the DAPC [39]. Components of the cytoskeleton integrate these signals. Astrocytic actin filaments serve as tracks for the directed movement of vesicles containing AQP4; microtubules are involved in the long distance transport; intermediate filaments such as GFAP contribute to the maintenance of structure of the endfeet [40]. Activity-dependent pathways involving rho GTPases and Ca²⁺ dependent pathways influence cytoskeletal remodeling in response to cues from neurons and vasculature. Chronic inflammation, oxidative stress, and chronic metabolic stress disrupt these mechanisms and make the system prone to depolarization [41,42].

Astrocytic AQP4 microdomains are continuously being remodeled by neuronal activity, metabolic needs, vascular dynamics, and circadian rhythms. High resolution imaging studies demonstrate that the clusters of OAPs vary in size, density and location over time (minutes to days). Circadian rhythm play an important role: reduced noradrenergic activity and increased interstitial volume during sleep promote the clustering of AQP4 and enhance glymphatic clearance, while wakefulness promotes the ion buffering and reorganization of AQP4 according to requirements [43]. Chronic sleep disturbance disrupts these fluctuations and impairs clearance efficiency, and represents one of the earliest features of neurodegenerative diseases [44].

Rapid, Ca²⁺-dependent phosphorylation cascades induced by neuronal activity modify the anchoring interactions of AQP4, while mechanotransduction from vascular pulsatility aligns water flow with hemodynamic forces. Decreased mechanosensitive signaling due to vascular stiffening results in decreased perivascular stability [45]. Recent studies using spatial transcriptomics have demonstrated astrocytic heterogeneity with cells adjacent to arterioles expressing more genes involved in the anchoring of AQP4 than cells adjacent to venules, indicating local specialization in maintaining polarity [46].

Modules involved in the maintenance of polarity of AQP4—anchoring complexes, isoform ratio, trafficking pathways, compartmentalization of membranes, and cytoskeletal dynamics—form an extremely connected system. The disturbance of any module is likely to affect other modules: changes in the isoform ratio affect the architecture of OAPs; inflammatory phosphorylation of α-syntrophin will weaken the anchoring of the DAPC; changes in lipidome will impair the organization and stability of lipid rafts and channels [47].

Mathematical models describing the system as exhibiting threshold-dependent behavior demonstrate that the compensatory mechanisms will work until the density of anchoring or membrane integrity fall below a critical value, after which depolarization will accelerate rapidly [48]. Simulations of the systems integrating molecular interactions, membrane mechanics, and fluid dynamics confirm that there exists such a tipping point, demonstrating that slight disturbances can induce complete collapse of the system once structural buffers are exceeded [49].

Understanding AQP4 polarity as an emergent systems property highlights both its fragility and potential for recovery. Due to its proximity to critical thresholds, even minor improvements in anchoring, trafficking, and/or membrane stability will restore coordinated water flow and glymphatic functions [50]. Table 1 summarizes the most relevant structural and regulatory modules forming AQP4 polarity, highlighting the dependence of each module on the others. Identifying this integrated blueprint of AQP4 polarity forms the basis for developing therapeutic strategies for both the prevention of depolarization and the restoration of functional polarity once lost.

AQP4 polarization emerges as an extremely precisely adjusted, integrated molecular blueprint which allows for the construction of scaffolding complexes, the composition of isoforms, vesicular trafficking, membrane organization, control and signaling, and dynamic modification. This attitude, however, implies fragility, for which a degree of adjustment in even one element will induce instabilities upon the whole network, and depolarization hence will ensue [30]. Understanding this blueprint is therefore essential to study how this component has been forfeited and what are the leverage points for therapy. That this machinery, this very molecular blueprint, would serve no other purpose in the long term but dynamically activate, does not only misplace the water channel, but would destroy a parent regulatory nexus upon which all clearance will depend, as will vascular communication, balance of immune phenomena, and cellular functions through which all communications must be conducted.

When combined, these molecular and organizational aspects demonstrate that AQP4 polarity is not merely an architectural preference, but rather a highly regulated and dynamic state dependent upon inputs from cytoskeletal, transcriptional, and membrane anchoring mechanisms. As all of these components work together to function as a cohesive system, disruptions within any one component have the potential to cause instability throughout the entire system. Therefore, it logically follows that understanding how this complex architecture is compromised in vivo would be the first step toward explaining why polarity loss occurs so frequently prior to the manifestation of early symptoms of neurological diseases. Consequently, the next section will discuss the stresses (i.e., vascular, mechanical, inflammatory, metabolic, and genetic) which disrupt the regulatory systems discussed previously and initiate the earliest stages of AQP4 depolarization.

Under normal physiologic conditions, the dynamic shear generated by the pulsatile blood flow provides stabilizing forces for dystrophin/syntrophin anchoring, and OAP clustering at the perivascular region. However, if hemodynamic balance is disrupted either due to chronic hypotension or small vessel disease, the stiffness of the vasculature decreases the pulse pressure, the endothelium produces less agrin, and the β-dystroglycan–ECM interaction is weakened. As a result, the Dp71/α-syntrophin complex is weakened and AQP4 diffuses out of the endfoot and into the surrounding membrane, resulting in the disruption of unidirectional water flow [65]. Further disruption of the blood–brain barrier (BBB), through endothelial damage, can lead to rapid depolarization of AQP4. This occurs through the disruption of the basement membrane by endothelial cells allowing plasma proteins and thrombin to interact with astrocytes and stimulate intracellular cascades that degrade dystrophin/syntrophin complexes and reorganize the actin cytoskeleton [66]. Additionally, even short-lived disruptions to the BBB can cause significant mislocalization of AQP4 within hours indicating the close association between vascular integrity and maintaining polarity [67].

Traumatic injury has also been shown to cause AQP4 mislocalization through multiple pathways. Traumatic injury can cause the shearing of astrocyte endfeet disrupting cytoskeletal connections, dispersing AQP4 clusters, and causing microhemorrhage and the subsequent activation of reactive gliosis. All of these injuries create large areas where AQP4 has lost its proper orientation [68]. Altogether, these data demonstrate that vascular stability and mechanical force represent major regulatory pathways for AQP4 localization, while other peripheral influences affect AQP4 through these central pathways [69].

Cytokines released by activated microglia can rapidly disrupt AQP4 polarity. Upon binding to astrocytic receptors, cytokines such as IL-1β, TNF-α, and IFNγ can initiate the activation of NF-κB dependent transcriptional programs that upregulate AQP4; however, cytokines also induce the downregulation of anchoring components and the degradation of agrin and laminin by matrix metalloproteases, which separates the number of channels from their spatial distribution [70].

Activated astrocytes also undergo transitions to GFAP and vimentin enriched phenotypes that alter the morphology of the terminal processes of the astrocyte. Complementary proteins, such as C1q and C3, accumulate in the perivascular space of astrocytes and induce astrocytes to transition to a GFAP/vimentin-rich phenotype that disrupts the geometry of the terminal processes of the astrocytes. The inhibition of complement signaling prevents the disruption of polarity caused by complement signaling [71]. Leukocytes entering through a damaged or compromised BBB can deliver cytokines and reactive oxygen species to astrocytes, suppressing alpha-syntrophin and Dp71 and disrupting the lipid rafts necessary to maintain AQP4 clustering. Chronic exposure to cytokines also disrupts astrocytic calcium signaling, destabilizes actin filaments, and impairs vesicular transport, all of which disrupts the perivascular microdomains and slows glymphatic transport [72,73].

Genetic variability affects the degree of stability of the AQP4 polarity network. Mutations in AQP4 alter the relative proportion of M1 and M23 isoforms, modify the density of OAPs, or disrupt the C-terminal PDZ binding sites necessary for anchoring to α-syntrophin. Certain haplotypes have been identified to be associated with decreased levels of M23 mRNA and/or altered phosphorylation of AQP4, leading to decreased lattice stability and increasing the risk of depolarization after exposure to stress [74,75].

Mutations in SNTA1 or DAG1 can disrupt PDZ-mediated binding or ECM attachment to AQP4, resulting in a widespread and non-perivascular distribution of AQP4 and decreasing perivascular water flow. Although each mutation alone may not significantly impact basal physiology, each can lower the threshold for depolarization by environmental or physiological stressors. Epigenetic mechanisms can also confer additional vulnerability to disease. Hypermethylation of the promoter of AQP4 can decrease transcriptional responsiveness to stress, and histone modification of genes involved in anchoring and trafficking can vary depending on the metabolic and inflammatory state of the cell. Non-coding RNA (such as miR-29 and miR-130) can regulate both the AQP4 and SNATA1 mRNAs and influence the response to metabolic and systemic stressors [76,77].

Aging results in gradual and cumulative effects on the AQP4 polarity network. Decreased mitochondrial efficiency limits the energy available for transporting vesicles and remodeling cytoskeletons. Oxidative stress can also modify dystrophin–syntrophin and cytoskeletal proteins that weaken anchoring and allow the lateral diffusion of channels [78,79].

Age-related alterations in metabolic signaling can also compromise AQP4 polarity. Decreased levels of insulin/IGF-1 can decrease the transcriptional responsiveness of AQP4. Alterations in cholesterol and sphingolipid composition can also disrupt the lipid raft stability and increase the movement of channels. Decreased circadian rhythm amplitude with age decreases clustering during sleep and reduces glymphatic transport. Each of these factors can independently compromise the AQP4 polarity network and collectively create a narrower margin of safety that can lead to rapid failure when challenged by secondary insults [80,81].

Environmental insults can precipitate or exacerbate the loss of polarity. Cumulative shear forces caused by repetitive head injury can induce microhemorrhages, ECM remodeling, and chronic expansion of depolarization [82,83]. Exposure to airborne particulates can activate inflammatory signaling in both the systemic circulation and endothelial cells, alter the composition of the basement membrane, and disrupt the BBB functions. Heavy metals can disrupt calcium homeostasis and actin filament dynamics in astrocytes, disrupting the cytoskeletal structures that support the membrane domains [84].

The early dynamics of the loss of polarity arise from the nonlinear interactions between vascular injury, inflammation, metabolic dysfunction, and genetic predisposition. Vascular stiffness and focal leaks in the BBB enable the entry of leukocytes into the CNS; inflammation can degrade ECM components and anchoring complexes, thereby creating mechanical instability; metabolic disorders can disrupt the lipid domains and trafficking pathways; and genetic background can determine the thresholds at which these failures will occur. The combination of these factors creates positive feedback loops that can greatly accelerate depolarization once initiated [85]. The mislocalization of AQP4 often occurs before the measurable accumulation of amyloid-beta or cognitive impairment. Modern advanced imaging techniques can detect subtle polarity disruptions following vascular injury or systemic inflammation in individuals who are at risk. Mathematical modeling demonstrates that maintaining polarity exhibits threshold properties. Once the density of anchoring or the stability of the membrane fall below a certain level, depolarization can proceed rapidly regardless of the magnitude of prior insults. These predictions are consistent with the observation that restoring polarity is most effective when initiated before the system crosses this tipping point [86]. Figure 1 illustrates the combined molecular and physiological forces that destroy the polarity network long before overt pathology develops and emphasizes the synergistic and mutually reinforcing relationships between them.

The polarity of AQP4 seems to be dependent on a fine balance of forces which can be easily upset, hemodynamically, inflammatorily, genetically, metabolically, and by environmental influences. These factors are synergistic and induce loops of self-reinforcing which intensify the associated depolarization and prepathology to devolve and proceed to be pathology in a detectable form. The insight and targeting of the mechanisms conducive to the loss of polarity may allow a chance of an intervention being performed whilst the history is still as such in a wholly reversible condition [87].

These initial destabilizers collectively show that AQP4 depolarization is not simply an isolated molecular event, but represents the first “tipping-point” for a larger system(s) dysfunction. The tightly-regulated integration of perivascular water flow, astrocyte communication, and vascular elasticity, once disrupted through loss of polarity, begins to fall apart. This transition will serve as the physiological link between the aforementioned stressor(s) which initiated this process and the subsequent large-scale hydraulic dysfunction(s). In order to determine how a localized aberration in channel position ultimately results in the widespread compromise of brain fluid transport and clearance, we examine the effects of AQP4 dysfunction on glymphatic organization and pressure dynamics in the next section.

Under physiologic conditions, the phenomenon of arterial pulsatility connects with the perivascular spaces, driving CSF into periarterial spaces and affecting its exchange with ISF via the AQP4-rich endfoot processes of the astrocytes with a low resistance pathway [90,91]. The amplitude of flow and reach depend on vascular compliance and effective water permeability. However, during depolarization, permeability becomes patchy by the insertion of high resistance segments, defeating the continuity of hydraulic transmission. The pulsatile energy is wasted and dissipated before it becomes effective, the pressure gradients become modified, CSF influx from the blood vessels and perivenous return decrease [92].

Computer models indicate that a slight decrease in perivascular conductance can cause the disproportionate loss of velocity and clearance, corresponding to a nonlinear relationship of flow resistance on the basis of branching geometries. Pulse transmission is dampened and phase-shifted, with corresponding effects on mechanotransduction in astrocytes and pericytes, with disordered nitric oxide signaling and disordered vascular tone. Thus, defects in signaling for fluidic and vascular impulses develop together from the same structural defect [93,94].

The relationship of AQP4 polarity to CSF movement creates a specific axis of vulnerability in the brain’s fluidic anatomy. AQP4 localized to the perivascular endfeet of astrocytes facilitates the conversion of the directionally defined inflow of CSF near arteries into a similar directional inflow into the surrounding parenchyma; in doing so, it allows for the propagation of arterial pulsations through the low resistance pathway connecting the periarterial space to the parenchyma [95]. This organization provides a stable hydraulic link between the inflowing CSF and the outflowing (both venous and lymphatic) CSF/ISF. The depolarization of AQP4 causes a loss of polarity leading to the redistribution of AQP4 from the perivascular endfeet of astrocytes across the rest of the astrocyte membrane; in doing so, the areas of reduced permeability lead to fragmentation of the CSF–ISF exchange pathways [96]. As a result, the inflow becomes disorganized and variable in terms of the magnitude and direction of convective flow to the various parts of the parenchyma; in addition, flow will occur primarily through diffusion rather than convection [97].

Tracer experiments have shown that loss of AQP4 polarity results in a non-uniform geometry of CSF penetration into the parenchyma, resulting in irregular wavefronts of the tracer and decreased penetration of the tracer into deeper structures of the brain. The irregular geometry of CSF penetration results in an increased time for solutes to be cleared from the parenchyma, an increased incidence of stagnant regions, and an increased susceptibility to fluctuations in blood vessel tone [98]. The CSF–ISF pressure gradient, normally maintained by the pulsatile nature of the vasculature and facilitated by the polarized water flux of the astrocytes, becomes less stable due to the loss of AQP4 polarity. As a result, the perivascular space will become enlarged, or collapsed in a segmental manner, due to the lack of coordinated water transport across the astrocyte membrane. The decreased hydraulic coupling will decrease glymphatic inflow, reduce the consistency of perivenous return, and render lymphatic clearance through the meninges less efficient [99].

While the loss of hydraulic coupling between the CSF and ISF will result in slowed fluid movement, it will also affect the mechanotransduction of astrocytes, pericytes, and vascular smooth muscle cells, all of which require coherent CSF–ISF dynamics to maintain normal levels of nitric oxide signaling, regulation of vascular tone, and maintenance of metabolic support to the parenchyma. Therefore, the depolarization of AQP4 will modify not only the flow of CSF, but the functional relationships that exist between the flow of CSF and the signaling and regulatory mechanisms of the vasculature and parenchyma [100]. Understanding the role of AQP4 polarity as a structural determinant of CSF–ISF exchange, therefore, will help clarify how molecular alterations at the level of the astrocyte endfoot can alter fluid dynamics throughout the glymphatic–venous network [62]. This schematic (Figure 2) illustrates how AQP4 polarity organizes directed CSF inflow and maintains coherent CSF–ISF exchange. Loss of polarity disrupts hydraulic coupling, fragmenting glymphatic flow and destabilizing vascular–glial signaling.

Although AQP4 orientation is primarily responsible for directing CSF flow to the periarterial and parenchymal regions, it is clear, however, that AQP4 organization is a different mechanism than that governing ISF removal. The three mechanisms involved in regulating ISF efflux are as follows: (1) polarity-directed endfoot permeability gradients, (2) anisotropic water conductance through the astrocyte membrane, and (3) the presence of a perivenous pressure gradient to drive ISF outflow, which is independent of the CSF inflow pathways [101]. Therefore, disrupting AQP4 polarity will disrupt ISF clearance through changes in hydraulic resistance within the brain tissue, loss of endfoot-driven convective transport, and disruption of coherence of the venous outflow system. In order to understand how nanoscale disruptions at the astrocyte endfoot result in macroscopic disruptions of glymphatic–venous transport, a detailed understanding of these polarity-dependent mechanisms is required [102].

Convection is essential in the clearance of any and all of the macromolecules possessing low diffusion coefficients in a dense neuropil. Since the convective Peclet numbers fall with depolarization of the pulsatile arterial blood flow, the motion affected by transport becomes diffusion-limited, and the local concentrations of amyloid-β, tau, input, TDP-43, etc., accumulates with progressive increase with the rousting of nucleation and fibrillization. These aggregates then obstruct the perivascular waterways, causing increased resistance and pushing further down the flow [103]. They also magnify the activation of microglia and the release of proteases which degrade the basement membranes and anchoring proteins leading to the architectural collapse: feed forward in which the inability of the products of metabolism to be removed exacerbates the barriers to removal of products of metabolism. Beyond the proteinopathy, however, is an accumulation of lactate, protons, and reactive oxygen species acidifying the local pH, altering the functions of channels and receptors, and oxidizing neutral and cytoskeletal products which maintain stabilization of the endfeet. The chronic presence of cytokines results in glial reactivity. The interstitium becoming from the buffering functions which maintained homeostasis to a milieu which is pro-pathogenic, which enhances the degenerative cascades [104].

Due to the reduction in adrenergic tone during sleep, there is an increase in interstitial volume and the oscillatory control of AQP4, an increase in perivascular clustering and consequent convective fluxes, and an increase in the removal of products of metabolism. The oscillatory function is diminished because of the depolarization during the period of night. The nocturnal peaks of removal of products of metabolism are diminished, but are still present in the interval of natural awakening, accumulating across the cycles [105]. The abnormal rhythms of the solutes are deranging to neuromodulatory systems of, for example, the adenosine signaling, causing the fragmentation of sleep and a decrease in synaptic homeostasis. The altered adrenergic and transcriptional programs feedback to maintain depolarization, which results in a bidirectional loop between the altered sleep and the hydraulic failure. This loop is consistent with the observations that disruptions of sleep precede and worsen cognitive decline [106,107].

Fluid stasis alters the anatomy of the perivascular niche. The diminished shear, diminished endothelial nitric oxide, favors deposition of the matrix in the extra-cellular space and a thickening and stiffening of basement membranes with diminished compliance caused by propagated pulses. Pericytes are contractile and decrease perivascular volume and mechanical conditions for endfeet, which also shifts AQP4. The immune dynamics also change to a slower turnover of cytokines, and DAMPs give a prolonged activation of NY macrophages and astrocytes [108]. Chronic low-level inflammation promotes the matrix of metalloproteinases which downregulate astrocytic endothelial signaling and gives rise to disorders of the blood–brain barrier, additional immune infiltration, and ECM remodeling. This is as far as heteroscedastics go, exaggerated closer to the penetrating arterioles where the most consequential flow deficits occur, and gives rise, again plausibly, to alterations in the topology of aggregate depositions seen in disease, including the alteration of excluded solute gradients, ionic content, and altered neurovascular signaling [109,110]. Convection also aids the maintenance of the ionic homeostasis of the extracellular compartment and with distribution of metabolites and signaling molecules. With depolarization, the microdomains of increased potassium and glutamate concentrations occur leading to a delay of repolarization and hyperexcitability permitting NMDA dependent excitotoxicity. The pH is also rendered less stable leading to altered enzyme kinetics and receptor sensitivity. These changes lead to a breakdown in neurovascular coupling. The potassium and nitrous oxide gradients around vessels normally regulate the smooth muscle tone and local perfusion. Their distortion leads normally to a lessened fidelity of blood flow regulations to neuronal activity limiting the access of oxygen and substrates to captured circuits. Diminished pulsatility and diminished convection are added limits the feedback loop in the dysregulation of vascularity—failure of fluid clearance [111]. Disruption of fluids lead to network hydraulic collapse linked to non-normally functioning neural circuits. The network of the extracellular milieu is for regulatory interface circuit dynamics. With the loss of a convective buffer, the fractional volume in the extracellular milieu, the ionic homeostasis and metabolite distributions are wider ranging. The coherence of oscillatory activity is lost: gamma synchrony is reduced, decreasing attentional and working memory power; slow wave activity is fragmented, leading to impairments in memory consolidation [112].

The “fluidic connectome”—the network of perivascular and interstitial routes which modulate neuronal computation by maintaining the extracellular milieu—is disrupted with AQP4 depolarization. Multiscale models coupling the transport of fluids with network electrophysiology predict a transition from stable oscillatory regimes to irregular or hypoactive states with only modest changes in convective capacity. Regional deficiencies of flow give rise to spatially localized network deficits, providing the mechanistic link between hydraulic failure and gradated deterioration of cognition [97].

Hydraulic collapse has immediate implications for therapy and diagnosis. The manner in which drugs are distributed in the brain parenchyma is partly sufficiently dependent on perivascular routes; impaired convection leads to reduced penetration and altered pharmacokinetics of both intrathecal and some systemic agents. Restoration of polarity might enhance drug delivery by reestablishing convective routes [113].

Similarly the interpretation of biomarkers is dependent upon the profile of clearance dynamics. The soluble indicators of pathology—amyloid-β, tau fragments, and inflammatory mediators—have concentration–time profiles determined by glymphatic transport. Taking account of polarity-dependent clearance will improve staging and aid early diagnosis. Advances in imaging, such as dynamic contrast MRI and molecular tracers, are now available, which visualize the influx of CSF and the clearance of ISF in vivo, showing flow fragmentation consistent with depolarization [114]. The application of these modalities to computational models of hydraulic failure may make possible the recognition of preclinical hydraulic failure and the introduction of polarity-restoring interventions [115]. Based upon this, Table 2 attempts to synthesize how AQP4 depolarization reshapes the therapeutic and diagnostic landscape.

AQP4 depolarization produces a rearrangement of brain hydraulic structure. Water transport in its directional modes becomes fragmented; the transmission of pressure energy is decreased; the clearance of macromolecules becomes diffusion limited; ionic and metabolic microdomains become destabilized; peripheral solid architecture is remodeled; coherence of networks collapses. The outcome of this is a transformation from fluidic homeostasis to fluid-driven pathology. Placing neurodegeneration within this context raises polarity from being a detail of membranes to a determinant of systems. Those strategies which restore perivascular AQP4 and reestablish convective exchange may remodulate vascular, immune, and network functions allowing for a tractable means of modifying morbid trajectories [124].

Once the system switches from organized or coordinated convective fluid transport to a disorganized or fragmented, diffusion-limited pattern of fluid movement, AQP4’s depolarizing effect on the cell membrane changes from being a localized hydraulic defect to a mechanism for causing widespread physiological instability. Once the mechanisms for fluid exchange, pressure transmission, and metabolic buffering become non-coherent in their function, they will begin to cause an increase in disturbances that move outward from perivascular spaces into vascular, immune, and neural domains. The transformation between impaired glymphatic flow and multisystem pathology is thus provided by this breakdown in hydraulic integrity. The next section addresses how the hydraulic defects caused by polarity collapse lead to the widespread physiological deficits described above by examining the vascular regulatory processes, blood–brain barrier functions, immune responses, metabolic pathways, and integrative networks affected by the loss of hydraulic coherency.

Astrocytes, endothelium, pericytes, neurons, and the extracellular matrix form a unit called the neurovascular unit (NVU). Each component relies on perivascular AQP4 to coordinate astrocytic mechanotransduction with vascular pulsatility. Loss of polarity will weaken calcium signaling, reduce nitric oxide and prostaglandin release, and ultimately cause the NVU to lose the ability to regulate vascular flow, and thus lead to localized areas of mild tissue hypoxia despite no apparent arterial disease [125]. Tight junctions between endothelial cells are compromised and, as such, the paracellular pathway becomes less selective, allowing plasma proteins to enter the parenchyma where they can activate astrocytes and microglia. The endothelium begins to upregulate vesicular transport and leukocyte adhesion molecules in response to the increased permeability. Detachment of pericytes from the basement membrane reduces the space surrounding the vessels, impairs mechanosensing due to altered shear forces, and disrupts glucose and lipid delivery to the parenchymal cells. As such, the blood–brain barrier function is now a conduit to pathology, and energy metabolism, myelination, and synaptic function are severely impaired [126,127]. As the degree of permeability continues to increase, it allows for additional signals from peripheral immune cells to propagate low-grade inflammation within the CNS that will continue to degrade junctional integrity, forming a positive feedback loop. Thus, breakdown of the NVU, remodeling of the BCB, and hemodynamic stresses each exacerbate the loss of polarity [128].

Studies using AQP4 knockouts have expanded this concept. While the loss of AQP4 clearly affects the glymphatic flow, solute clearance, and the calcium signaling of astrocytes, the disease-specific effects of AQP4 loss are still incompletely understood. For example, AQP4 knockout models exhibit accelerated amyloid-beta and tau accumulation and increased microglial activation in Alzheimer’s models, increased dopaminergic vulnerability in α-synucleinopathy models, decreased potassium buffering and metabolic support in demyelinating and motor neuron disorder models, although the causal relationships are still unclear [129,130,131].

Similarly, AQP4 genetic studies in humans show similar results: restriction fragment length polymorphism (RFLP) and regulatory variants of AQP4 have been shown to affect susceptibility to edema, cognitive resilience during aging, and sleep dependent glymphatic efficiency. Variants have also been associated with an increased risk or severity of Alzheimer’s disease, NMOSD, and TBI. Presumably, variants alter water permeability, anchoring density, and/or stress responses; however, the penetrance and functional impacts of the variants are still unknown [132]. Collectively, both knockout and genetic studies suggest that variations in the expression or localization of AQP4 result in the loss of NVU stability, blood–brain barrier competence, and the chemical environment in which neurodegeneration occurs [133].

The continuous removal of cytokines, complement signaling, and DAMPs by polarity-dependent convective flow will prevent the accumulation of pro-inflammatory and pro-degenerative substances in the parenchyma. However, when the removal of these substances is disrupted, the accumulation of these substances will induce microglial activation, release of IL-1β, TNF-α, ROS, and proteases, which will break down the synapses, the extracellular matrix, and ultimately, neuronal integrity. Astrocytes will also develop neurotoxic programs, including the release of complement signaling, decrease in glutamate uptake, and reduction in metabolic support for neurons [134].

The cytokine environment will suppress the transcription of proteins responsible for anchoring AQP4 to the cell surface, while increasing the production of matrix metalloproteinases, resulting in the degradation of the extracellular matrix, while ROS will destabilize cytoskeletal and membrane components of AQP4 clusters, thereby continuing the inflammatory process [135,136].

At the same time, the ionic and metabolic homeostasis will deteriorate. Decreased convection will lead to an accumulation of potassium ions and delay repolarization and hyperexcitability of neural networks. Increased glutamate will drive NMDA receptor-mediated calcium overloading, excitotoxic cascades, and pH instability, while exhausted astrocytes will fail to buffer calcium ions and regulate blood flow [137]. Energy demands will increase due to elevated levels of ATP, while impaired mitochondrial functions, due to oxidative stress, will worsen the metabolic deficiencies [138]. At the network level, the convergence of these molecular failures will result in fragmentation of gamma synchronization, deterioration of slow wave activity, and loss of coherent communication between distant parts of the brain. Mechanisms of plasticity (LTP/LTD) will fail to operate under conditions of ionic and metabolic instability, and cognitive deficits will arise prior to substantial neuronal loss [139]. These CNS disruptions will also interact with peripheral immune and neuroendocrine systems. The disruption of drainage through the meningeal lymphatics will lead to altered antigen presentation and weakened immune surveillance, whereas the accumulation of CNS cytokines in the systemic circulation will reinforce chronic peripheral inflammation, which will feed back into the brain through endothelial and vagus nerve pathways [140,141,142].

The disrupted hypothalamic and brainstem detection of CSF solute gradients will lead to the disruption of rhythmic processes of appetite, arousal, autonomic tone, blood pressure, and metabolic functions—frequently observed features of many neurodegenerative disorders, but insufficiently explained by localized neurodegeneration [143].

Across various scales, AQP4 depolarization produces common pathophysiological logic. Loss of NVU function and BCB dysfunction create an opportunity for immune invasion; the subsequent inflammation will dismantle the structures that maintain AQP4 anchoring; ionic imbalance and excitotoxic stress will modify network dynamics; and feedback loops through peripheral tissues will extend the scope of dysregulation to include regions outside the CNS [144]. This common structure explains the presence of AQP4 depolarization, glymphatic dysfunction, neuroinflammation, and vascular instability in Alzheimer’s disease, Parkinson’s disease, multiple sclerosis, traumatic brain injury, and numerous other disorders [145]. Looking at the progressive degradation of brain physiology through this integrated perspective, the differences between these disorders appear to be largely a matter of varying degrees of a common degenerative sequence initiated or exacerbated by AQP4 polarity collapse. Therapeutic interventions targeting AQP4 polarity may provide a means to influence disease trajectory across diagnostic categories and serve as a conceptual anchor for multi-system therapeutic strategies [146]. Figure 3 attempts to illustrate this sequence: perivascular depolarization disrupts communication between the NVU and vascular pulsatility → chronic inflammation and metabolic deficiency accumulate → ionic and synaptic instability occur → network coherence is lost → peripheral immune and neuroendocrine systems are entrained. Thus, the hierarchical architecture of the above sequence establishes AQP4 polarity as a key determinant of systemic neurological decline, and possibly as a potential target for reversal of neurological decline.

The loss of polarity of AQP4 is afflicted with stupendous effect from the local perivascular membranes to that of the consideration of the much greater architecture of the function of the brain. It is disassembly of the coordinated functionalizing of the NVU, it remodels the blood–brain barrier, it stimulates chronic inflammatory processes, it destroys ionic and metabolic equilibrium, and it causes loss of synchrony within the genome of the nervous system. It brings about a loss of immune functions, it wipes out wired neuroendocrine feedback, and their resultant coupling between loss of central function and systemic loss of control. The formed results, we imply, should show a remapping of the lost control of polarity, close to that of systems level defect, and suggest that restoration of this would naturally involve the simultaneous readjustment of vascular, immune, metabolic, and neurological factors [133].

In combination, this disarray in systems-level functions illustrates that AQP4 depolarization profoundly alters the overall functioning of the brain; it is not limited to the interface between the blood vessels and the brain. Once, in turn, vascular signaling, immunological equilibrium, metabolic homeostasis, and neuroendocrine feedbacks all deviate from their usual coordinated levels of function, the organism will be in a condition where its compensatory systems are no longer able to maintain the level of disruption to the organism’s systems. This widespread alteration of the organisms’ systems creates an environment in which the numerous diseased pathways will have ample opportunity to develop, combine, and mutually reinforce each other. In the next portion of this review, we examine how the above described large-scale physiological disruptions lead from early impairment to cascading disease processes, and explain why the loss of polarity so frequently represents the transition from reversible dysfunction to chronic pathology.

Impairment of the glymphatic system leads to the accumulation of amyloid-β, tau, α-synuclein, and TDP-43 outside of the cellular environment, thereby altering their kinetic behavior towards nucleation and subsequent fibril formation. Aggregates form within the perivascular space and impede the flow through the space. As aggregates accumulate, they stimulate glial receptor activation resulting in sustained inflammatory signaling [148]. Released proteases generated during this inflammatory response cause the degradation of laminin and agrin, destabilize the perivascular scaffold, and further reduce clearance capabilities generating a self-reinforcing cycle of protein deposition and structural breakdown. A restoration of AQP4 polarity has been shown to decrease the amount of aggregate formed within the perivascular space, indicating the role of flow dynamics in disease progression [149].

Stagnated solute loads and inflammation propagated throughout the vasculature lead to the development of vascular pathology. Oxidative and inflammatory responses to cytokines and hypoxia damage endothelial cells. Aberrant fluid dynamics also alter ionic gradients that are necessary for vascular reactivity, leading to an inability to regulate blood pressure, thus creating a state of hypoperfusion that generates oxygen glucose deprivation and mitochondrial failure. Mitochondrial failure and oxidative damage generate further oxidative damage and metabolic failure. Hypoxia-driven remodeling of the endothelium increases vascular permeability, allowing leukocytes to enter the CNS where they produce cytokines that support the inflammatory response [150]. Furthermore, hypoxia causes the apoptosis of oligodendrocytes disrupting the structure and function of myelin, thus affecting the integrity and function of white matter. Astrocytes become metabolically stressed and begin to rely on glycolysis for energy production, thus decreasing their ability to recycle neurotransmitters and buffer ions [151]. Accumulation of glutamate and potassium in the CNS can damage axons and synapses, and mitochondrial failure disrupts the transport of materials along axons, making white matter vulnerable to damage across all “gray matter” diseases [152].

At the network level, these failures converge to create an unstable network. The loss of gamma synchrony creates a lack of coordination between different parts of the network. Slow wave activity collapses creating a loss of the rhythmic coordination that was previously present. Long range communication becomes unreliable due to the energy dependent plasticity mechanisms failing due to excessive calcium and insufficient ATP and the deterioration of dendritic structure under oxidative stress. Importantly, these failures occur prior to the widespread loss of neurons and indicate that cognitive decline occurs due to the disequilibrium of fluids and ions rather than solely due to neuron death [153].

While polarity loss generates a common cascade, the disease context affects how the cascade is expressed. For example, in Alzheimer’s disease, impaired clearance of amyloid-β promotes tau propagation and vascular stiffening; in Parkinson’s disease, impaired glymphatic clearance fosters α-synuclein accumulation and dopaminergic vulnerability; in multiple sclerosis, perivascular inflammation and metabolic insufficiency worsen demyelination; after traumatic brain injury, mechanical disruption of endfeet creates prolonged polarity loss, edema, and cognitive dysfunction; in gliomas, altered perivascular dynamics transform the tumor microenvironment affect the blood–brain barrier functions [154].

Computational modeling demonstrates that, despite these differences, the commonality among the diseases lies in the nonlinear nature of the disease process. Specifically, once perivascular AQP4 density drops below a certain threshold, convective clearance fails, solute accumulates, and ionic regulation loses stability. Therefore, small changes in membrane conductivity or ion gradient concentrations rapidly lead to the sudden transition of network behavior from ordered, oscillatory behavior to irregular or hypoactive behavior [155]. These threshold effects can explain why many diseases exhibit an initial period of apparent dormancy followed by a rapid progression to severe symptoms, and why restoring polarity is most effective when performed before the critical transition points have been reached [156].

Depolarization of AQP4 begins a multidimensional cascade of events in which clearance failure promotes aggregation, inflammation degrades structural scaffolding, vascular insufficiency induces metabolic collapse, and network coherence disintegrates. At some point, across many diseases, this represents the tipping point beyond which reversible dysfunction becomes irreversible degeneration. Viewing depolarized AQP4 as a malleable point provides new perspectives on neurological diseases as disorders of fluidic failure across systems. Thus, restoring polarity should be viewed not simply as a fortunate side effect of therapy, but as a strategically positioned therapeutic anchor with the capability to influence disease trajectory [157].

Thus, a therapeutic strategy aimed at this point of convergence will focus on restoring perivascular anchoring, maintaining trafficking, and restoring convective flow. By doing so, it may restore clearance, stabilize vascular signaling, and break the cycles of inflammation and metabolic failure that promote degeneration. The therapeutic strategy outlined in the next section is predicated on this rationale.

Repairing the dystrophin-associated complex (DAC) which anchors AQP4 to perivascular membranes is the basis for structural repair. Damaging dystrophin, α-syntrophin, β-dystroglycan, or related molecules disrupts endfoot anchoring and is a major pathway leading to depolarization. Stabilizing AQP4–α-syntrophin interactions with small molecules or preventing destructive phosphorylation increases channel retention and the speed of recovery of localized perivascular AQP4 [159,160]. Gene therapy vectors expressing functional forms of the DAC components delivered to astrocytes restore clustered AQP4 and improve the clearance of solutes in experimental settings. Modulating the cytoskeleton is an additional approach to affect the ability of AQP4 to move and to maintain perivascular location. Cytoskeletal organization greatly affects the ability of AQP4 to move, and it is vulnerable to damage from inflammatory and mechanical stress. Agents that stabilize filamentous actin or inhibit proteins involved in depolymerizing actin filaments, such as cofilin, support the maintenance of the DAC’s function to maintain polarity and endfoot architecture [161].

Gene editing technologies expand these approaches. Platforms based on CRISPR/Cas can edit deleterious mutations of dystrophin, syntrophins, or regulatory elements; synthetic transcription factors can selectively induce the expression of anchoring proteins in reactive astrocytes; and tools that modify the epigenome can make promoters accessible to support the sustained expression of programs that support polarity without changing the genome [162,163,164,165,166]. Together, these therapies provide precise correction of structural and regulatory vulnerabilities responsible for polarity failure.

Although the expression of the DAC contributes to structural polarity, it is controlled by the transcriptional and epigenetic states of the astrocytes. Neuroinflammation typically reduces the expression of AQP4 and the anchoring gene expression, and can alter the ratio of AQP4 isoforms. The use of transcription factors, such as NF-κB, STAT3, and CREB, can restore AQP4 expression, normalize the M1/M23 ratio, and increase the expression of scaffold proteins [167]. RNA-based therapies are complementary to these approaches. Antisense oligonucleotides and siRNAs can modify the splicing of AQP4 to favor polarity-supportive AQP4 isoforms. Synthetic mRNAs can increase the number of perivascular channels during periods when there is increased susceptibility. These tools allow for the state-dependent modulation of gene expression in astrocytes, and therefore allow for early, very specific intervention [168,169,170].

Any therapies that will be able to correct polarity must be able to access the gliovascular interface—a major barrier to delivery. Nanomedicine offers several ways to address this issue by allowing targeted, condition-responsive delivery. Nanoparticles engineered to bind to endothelial or astrocytic ligands are taken across the blood–brain barrier via transcytosis and accumulate in perivascular compartments where they can release their cargo in response to local signals, such as pH, oxidative stress, or cytokine concentrations. This allows for the delivery of small molecules, RNA, or gene editing components to the site of action with high specificity and reduced systemic exposure [170]. Exosomes and engineered extracellular vesicles provide an alternative biocompatible delivery method by utilizing the natural mechanisms of crossing the BBB and minimizing the elicitation of an immune response. Engineered exosomes can be modified to bind to dystrophin or syntrophin, thereby enhancing the repair of anchoring structures, and can also be loaded with anti-inflammatory or trophic factors that modify the microenvironment in a way that favors the restoration of polarity [171]. Biomaterials provide opportunities for regeneration. Bioactive scaffolds containing astrocytic progenitor cells can populate damaged regions and reconstitute polarity-competent networks. Advances in 3D printing and microfluidic organoid systems now enable the rapid assessment of scaffold compositions that optimize anchoring integrity and perivascular organization [172].

Since inflammation and vascular dysfunction both initiate and perpetuate depolarization, modulating the inflammatory–vascular axis globally provides a second approach to therapy. Broad-spectrum inhibitors of NF-κB or JAK/STAT signaling decrease the activation of microglia and create a reparative environment that supports polarity, such as methods to reprogram microglia and astrocytes to promote anchoring stability and reduce matrix degradation [173].

Peripherally-activated inflammation can also affect polarity of the CNS through cytokine spillover and endothelial activation. The peripheral administration of TNF-α and IL-1β antagonists, metabolic anti-inflammatory regimens, and targeted vascular therapies can reduce inflammatory spillover across the BBB and indirectly protect glymphatic functions [174,175].

Non-pharmacologic modulation of the inflammatory–vascular axis is provided by bioelectronic methods. Vagal nerve stimulation decreases peripheral and CNS inflammation and promotes endfoot stability. Focused ultrasound can temporarily modulate perivascular water movement and BBB permeability to facilitate drug delivery and activate mechanotransduction pathways in astrocytes [176]. Chemogenetic and optogenetic modulation of astrocytes and vasculature further provide a means to precisely regulate vascular tone and fluid dynamics. Systems that monitor glymphatic flux in real time and adaptively stimulate to modulate flow represent an emerging field [177].

Precision medicine approaches are required due to the complexity of regulating polarity. Single-cell transcriptomic, spatial proteomic, lipidomic, and metabolomic multi-omics profiling is identifying new polarity-regulatory molecules and guiding combinatorial strategies that target multiple layers of regulation simultaneously [178].

AI is accelerating the translation of the findings to the clinic. Machine learning algorithms integrate molecular, imaging, and physiological data to predict which patients would benefit from therapies targeting polarity, to stratify disease subtypes based on glymphatic impairments, and to predict treatment responses [179]. AI-assisted drug discovery is identifying compounds that modulate polarity pathways, while digital twins of individual patients simulate the effects of interventions on perivascular flow and network coherence before clinical trials [180]. Computational fluid dynamic models further predict how interventions shape hydraulic behavior over all scales [93].

Because polarity collapse demonstrates threshold behavior, the timing of the treatments is critical. Several biomarkers of early polarity disruption—CSF/ISF tracer kinetics, perivascular diffusion metrics, and soluble AQP4 fragments—are being investigated to detect polarity disruption before the establishment of irreversible feedback loops [181,182]. Combinatorial therapies that combine structural stabilizers with immunomodulators, or gene editing approaches with biomaterials, are intended to correct all of the failure points simultaneously. Initial clinical studies are beginning to apply these concepts, including nanoparticle-based AQP4 modulators, focused-ultrasound protocols to enhance glymphatic flow, and systemic immunotherapies to target inflammation disrupting polarity [183].

AQP4 polarity has developed as a tractable therapeutic axis that may be used to reshape disease progression. Until recently, polarity was considered a static structural feature of astrocytes. However, current knowledge indicates that polarity is a dynamic determinant of fluid transport, vascular signaling, immune tone, metabolic stability, and network functions. The diverse array of therapeutic approaches—from molecular engineering and reprogramming of astrocytes to nanomedicine, biomaterials, neuromodulation, and AI-guided precision therapy—converge to the objective of reestablishing directional water flow and glymphatic efficiency. Reestablishment of polarity has the potential to reverse the clearance deficits and vascular instabilities that contribute to disease, as well as to recalculate the systems physiology that sustains neuronal resilience [184].

While this review combines molecular, physiological, and systems level data to support AQP4 polarity as an essential regulatory mechanism of brain homeostasis, there are a number of limitations that should be identified.

Firstly, the majority of the information presently understood about AQP4 polarity failure was derived from rodent models and, therefore, can be subject to significant interspecies variability in terms of vascular geometry, astrocyte diversity, sleep patterns, and perivascular dynamics. This results in the inability to make direct translations from experimental findings on polarity failure, glymphatic flow, and disease progression in rodents to humans.

Secondly, the available data are highly segmented by scale. While the molecular data regarding dystrophin–syntrophin interaction, isoform regulation, and trafficking pathways are robust, the relationship between these mechanisms and macroscopic changes in hydraulic functions, vascular integrity, or network coherence is rarely directly demonstrated. Still in their infancy, high resolution imaging techniques are being developed, which would allow researchers to investigate the early loss of polarity, changes in perivascular resistance, or alterations in microdomain organization in humans.

Thirdly, the model presented in this review integrates a large number of different biochemical pathways, including inflammatory signaling, vascular remodeling, metabolic failure, and network instability; however, the hierarchical nature of these interactions and how they relate to each other in a temporal sense (i.e., cause and effect) remains poorly defined. Many of the associations reported in the literature are correlative; thus, while causality may be inferred based upon the timing of events in animal models or through indirect biomarkers, it cannot be established with confidence using longitudinal human data.

Fourthly, while depolarization is presented here as an early event in many neurological diseases, due to its convergence on common molecular mechanisms, this conceptually unifying framework remains purely theoretical. There are numerous disease-specific factors that influence the occurrence and impact of polarity loss, including genetic variation, comorbidity of vascular risk, immune profile, and environmental exposure(s), all of which have not yet been fully characterized.

Lastly, while emerging therapies, including gene editing, nanotechnology, biomaterial scaffolds, and bioelectronics, hold promise for restoring polarity in complex human disease states, most of these therapies are still in preclinical or early-stage translational development. Thus, the ability of these therapies to restore polarity in the complex environment of human disease and the safety of sustained modulation of fluid dynamic pathways will require thorough and long-term testing.

Overall, these limitations emphasize the need for multidisciplinary and multiscale research, the continued improvement of non-invasive imaging of perivascular behavior, and the determination of the causal relationships between polarity and pathology in humans, as well as the design and execution of early phase clinical trials. By recognizing these boundaries, the current synthesis can be framed as a working model which will serve to provide guidance for future study rather than a definitive resolution of the underlying mechanistic processes.