Neurofibrillary tangle (NFT) propagation represents one of the most compelling and transformative concepts in contemporary Alzheimer's disease (AD) research. This phenomenon describes the spreading of pathological tau protein aggregates throughout the brain in a pattern that correlates directly with clinical disease progression and cognitive decline. Understanding the mechanisms underlying NFT propagation has fundamentally reshaped our conceptualization of AD from a diffuse, multi-focal disease to a orderly, network-based neurodegeneration process. The propagation of tau pathology follows anatomically connected neural networks, moving from entorhinal cortex and hippocampus in early stages to cortical association areas in later disease stages, following a predictable staging scheme originally described by Braak and Braak. This pattern of progression suggests that pathological tau may spread along neuronal connections, similar to prion diseases, though the underlying mechanisms appear distinct from classic prion propagation. The identification of NFT propagation has profound implications for understanding disease progression, developing diagnostic biomarkers, and creating therapeutic interventions that might halt or slow the spread of pathology before extensive neuronal loss occurs.
Tau pathology forms one of the two hallmark proteinaceous lesions in Alzheimer's disease, alongside amyloid-beta plaques. Tau is a microtubule-associated protein primarily expressed in neurons, where it plays essential roles in stabilizing axonal microtubules and regulating intracellular transport. In AD, tau becomes hyperphosphorylated, leading to its aggregation into paired helical filaments (PHFs) that constitute the core structure of neurofibrillary tangles. These intracellular inclusions accumulate within neuronal cell bodies and dendrites, disrupting cellular function and ultimately leading to neuronal death. The distribution of NFTs follows a characteristic pattern that has been extensively documented through postmortem studies, establishing the widely used Braak staging system that ranges from Stage I (transentorhinal region) to Stage VI (primary neocortex) [1].
The significance of tau pathology in AD extends far beyond its role as a pathological marker. Numerous studies have demonstrated that the burden of neurofibrillary tangles correlates more closely with cognitive impairment than amyloid-beta plaque burden, suggesting that tau pathology represents the primary driver of clinical symptoms [2]. Furthermore, the predictable spreading pattern of tau pathology has led researchers to investigate how pathological tau spreads from initially affected brain regions to connected areas. This propagation appears to follow the hierarchical organization of functional brain networks, with hub regions that possess high connectivity serving as early recipients of pathological tau seeding and propagation [3].
The concept of tau propagation has been substantially advanced by the discovery that tau can be released from neurons and taken up by neighboring cells, both in vitro and in vivo. Experimental models have demonstrated that extracellular tau can be internalized by recipient neurons and template the misfolding of endogenous tau, creating a self-propagating cycle of pathology spread [4]. This mechanism shares conceptual similarities with prion diseases but differs in important mechanistic details, as tau propagation in AD appears to involve multiple protein conformers and does not require the same level of infectious capability demonstrated by prions. The propagation of tau pathology thus represents a fundamental biological process that explains the predictable clinical progression of AD and provides multiple potential therapeutic targets.
The aggregation of tau into neurofibrillary tangles represents a complex process involving multiple intermediate states and conformational changes. Normal tau protein exists in a largely unfolded state, with minimal secondary structure, allowing it to interact with microtubules and perform its physiological functions. The transition from normal tau to pathological aggregates involves post-translational modifications, particularly hyperphosphorylation, that alter tau's biophysical properties and promote its aggregation [5]. Over 40 potential phosphorylation sites have been identified on tau, and specific phosphorylation patterns appear to correlate with different stages of pathology and different tau conformers.
The misfolding of tau involves the adoption of beta-sheet rich conformations that enable the protein to aggregate into oligomeric species and ultimately into insoluble fibrils. These oligomeric species have attracted particular attention because they appear to be the most toxic form of tau pathology, capable of disrupting synaptic function and propagating to other neurons before the formation of mature tangles [6]. The identification of distinct tau conformers, sometimes termed "strains," has suggested that the aggregation process may produce multiple distinct pathological species, similar to the strain phenomenon in prion diseases. These different conformers may have varying abilities to spread and cause pathology, potentially explaining the heterogeneity in clinical presentation observed among AD patients.
The mechanism of tau misfolding appears to involve the templated conversion of normal tau proteins into the pathological conformation, a process that can be accelerated by the presence of pre-existing pathological tau seeds. This templating capability underlies the theoretical basis for tau propagation, as extracellular tau aggregates can serve as seeds that convert normal tau in recipient neurons [7]. The efficiency of this templating process appears to depend on the conformation of the seed, with certain aggregated forms demonstrating greater seeding capability than others. Additionally, cellular factors including molecular chaperones, post-translational modification enzymes, and degradation pathways influence whether tau that enters a cell will propagate pathology or be cleared before causing damage.
Structural studies using cryo-electron microscopy have revealed the detailed architecture of tau filaments from AD brains, demonstrating the presence of a double helical structure composed of two protofilaments [8]. The core region of these filaments contains residues 306-378 of the tau protein, forming a fuzzy coat region that extends outward. Importantly, the structure of tau filaments differs between AD and other tauopathies, suggesting that the specific conformation of tau aggregates may determine the clinical phenotype. Understanding these structural details has provided insights into how tau aggregates form and how they might be targeted therapeutically.
The concept of prion-like propagation in AD describes the ability of pathological tau species to spread between neurons and template the misfolding of normal tau in recipient cells. This mechanism was first proposed based on the predictable spatial progression of tau pathology that followed anatomical connectivity rather than proximity or vascular supply. The term "prion-like" refers to the capacity of a protein to act as a template for the conversion of normal proteins into the same pathological conformation, similar to the mechanism described in prion diseases such as Creutzfeldt-Jakob disease [9].
Multiple lines of experimental evidence support the prion-like spread mechanism for tau pathology. Studies using rodent models have demonstrated that inoculation of brain homogenates containing pathological tau into wild-type animals leads to the development of tau pathology in the inoculated region, followed by spreading to anatomically connected areas [10]. Additionally, experiments using cultured neurons have shown that extracellular tau aggregates can be internalized and subsequently induce the aggregation of intracellular tau, with the newly formed aggregates capable of being released and taken up by other cells in a self-propagating cycle [11].
The prion-like mechanism involves several sequential steps: release of pathological tau from donor neurons, uptake by recipient neurons, intracellular templating of normal tau, and subsequent release of newly formed aggregates. Each of these steps represents a potential intervention point for therapeutic development. The release of tau appears to occur through multiple mechanisms, including exosomal release, direct secretion in association with extracellular vesicles, and possibly through synaptic transmission. The uptake mechanism likely involves endocytosis and potentially specific receptors that facilitate the internalization of tau aggregates. Once inside the cell, the pathological tau must overcome cellular clearance mechanisms to template the conversion of normal tau, requiring the seed to be sufficiently stable and present in adequate concentrations.
Despite the conceptual similarities to prion propagation, tau propagation in AD differs in important respects. The spreading of tau pathology in AD occurs over years to decades, in contrast to the rapid propagation seen in prion diseases. Additionally, tau pathology appears to require ongoing neuronal dysfunction and specific cellular contexts to propagate efficiently, while prions can propagate in many cell types. These differences suggest that while the fundamental templating mechanism may be similar, the biological context substantially modulates the propagation kinetics and consequences.
The propagation of tau pathology through the brain follows patterns that reflect the underlying architecture of neural connectivity. Neuroanatomical studies have consistently demonstrated that tau pathology spreads in a predictable manner, beginning in the transentorhinal region and olfactory bulb before advancing to the entorhinal cortex and hippocampus in early stages, then progressing to associative neocortical areas and ultimately affecting primary sensory and motor cortices in late-stage disease [12]. This progression pattern correlates with clinical disease progression, as patients with early-stage pathology typically present with memory dysfunction while later-stage patients develop additional cognitive deficits reflecting the involvement of cortical association areas.
The network-based spread of tau pathology has been extensively characterized using positron emission tomography (PET) imaging with tau-specific radiotracers. These studies have demonstrated that the burden of tau pathology in any given brain region correlates with both the local structural connectivity and the integrated network connectivity of that region with areas of early pathology [13]. Hub regions, which serve as highly connected nodes within brain networks, demonstrate early accumulation of tau pathology and serve as conduits for propagation to more peripheral regions. This pattern is consistent with the concept that pathological tau seeds are transported along neuronal pathways to connected brain regions, where they initiate new centers of pathology.
White matter tracts provide important pathways for the propagation of tau pathology between brain regions. Diffusion tensor imaging studies have demonstrated that the integrity of white matter connections influences the spread of tau pathology, with regions connected by compromised white matter showing correlated patterns of pathology [14]. The choroid plexus and ventricular system may also play roles in the propagation of tau, as CSF represents a potential vehicle for the transport of extracellular tau between brain regions. Additionally, the glymphatic system, which facilitates the clearance of interstitial waste products, may influence the spread of tau by modulating the extracellular environment.
The trans-synaptic spread of tau represents another important pathway for propagation. Experimental studies have demonstrated that pathological tau can be detected in presynaptic terminals and post-synaptic densities, suggesting that synaptic activity may facilitate the intercellular transfer of tau [15]. The high metabolic activity and constant vesicle trafficking at synapses may provide opportunities for tau release and uptake, making synapses particularly relevant for propagation. Furthermore, the involvement of specific neuronal populations, particularly those in layer II of the entorhinal cortex that project to the dentate gyrus, may explain the early involvement of hippocampal circuits in AD.
Extracellular vesicles, particularly exosomes, have emerged as important vehicles for the intercellular transfer of pathological tau. Exosomes are small vesicles released from cells that can contain various cargoes, including proteins, lipids, and nucleic acids. In the context of tau pathology, exosomes released from neurons have been shown to contain hyperphosphorylated tau and may serve as protection mechanisms for the release of pathological protein while also potentially facilitating the spread of pathology to other cells [16]. The loading of tau into exosomes appears to be an active process that involves specific cellular machinery, and the exosomal tau may be more efficient at seeding pathology than free tau aggregates.
The role of microglia in tau propagation represents an area of active investigation, with these resident immune cells playing complex and potentially contradictory roles in the spread of pathology. On one hand, microglia may contribute to propagation by phagocytosing tau and subsequently releasing it in a form that retains seeding capability, effectively acting as vectors for the spread of pathology [17]. Microglial activation may also create an inflammatory environment that promotes the release of tau from neurons and increases the permeability of the blood-brain barrier, facilitating the peripheral spread of tau species. On the other hand, microglia are capable of clearing tau aggregates and may represent an important defense mechanism against the propagation of pathology.
The interaction between exosomes and microglia may be particularly important for understanding tau propagation. Microglia can take up exosomal tau and may either degrade the pathological cargo or, under certain conditions, release it to propagate pathology further. The inflammatory state of microglia appears to influence this outcome, with alternatively activated microglia potentially supporting the spread of pathology while classically activated microglia may be more effective at clearing tau. Additionally, the release of cytokines and other inflammatory mediators by microglia can modulate the neuronal environment, potentially affecting the release and aggregation of tau.
The glymphatic system, a macroscopic clearance pathway that involves perivascular cerebrospinal fluid flow, also interacts with tau propagation. Impairment of glymphatic clearance, which occurs with aging and in AD, may contribute to the accumulation and spread of extracellular tau [18]. Sleep disruption, which has been linked to AD risk, appears to particularly impair glymphatic function, potentially explaining part of the relationship between sleep quality and AD pathogenesis. The interplay between cellular and macroscopic clearance mechanisms thus influences the propagation of tau pathology throughout the brain.
The understanding of tau propagation has opened multiple new therapeutic avenues for AD treatment. The key insight that tau pathology can spread between neurons suggests that interventions targeting any step in the propagation cascade could potentially slow or halt disease progression. These therapeutic strategies include preventing the release of pathological tau, blocking the uptake of extracellular tau, inhibiting the templated conversion of normal tau, and enhancing the clearance of pathological species. The development of these interventions represents a major focus of current drug development efforts for AD.
Immunotherapy approaches targeting tau have advanced substantially in recent years, with multiple antibodies targeting various forms of tau已进入临床试验 [19]. These antibodies are designed to neutralize extracellular tau and prevent its uptake by neurons, potentially interrupting the propagation cycle. Active vaccination strategies aimed at generating endogenous antibodies against tau are also under investigation. Early clinical trials of anti-tau antibodies demonstrated some safety concerns related to off-target effects, but newer generation antibodies with improved specificity and targeting are showing more favorable profiles.
Small molecule inhibitors of tau aggregation represent another therapeutic approach that has shown promise in preclinical models. These compounds aim to prevent the formation of toxic tau oligomers and fibrils by stabilizing the normal protein conformation or directly inhibiting the aggregation process. Several aggregation inhibitors have advanced to clinical testing, though demonstrating efficacy in human trials has proven challenging due to the complexity of tau pathology and the potential need for very early intervention [20].
Other therapeutic strategies include the modulation of tau phosphorylation through kinase inhibitors or phosphatase activators, as hyperphosphorylation represents a key step in the initial formation of pathological tau. Additionally, approaches aimed at enhancing tau clearance through autophagy or proteasomal pathways are being explored, as these mechanisms are responsible for the normal turnover of tau and may be impaired in AD. The optimal therapeutic approach may require combinations of these strategies to effectively halt the propagation of pathology.
Current research on NFT propagation encompasses multiple investigative directions, from basic biology to clinical translation. Advances in imaging technology have enabled the visualization of tau pathology in living patients, providing unprecedented opportunities to study the dynamics of propagation in humans. Tau PET imaging has revealed that the rate of tau accumulation correlates with cognitive decline and can predict future progression, supporting the clinical relevance of propagation mechanisms [21]. Longitudinal imaging studies are now investigating the factors that influence the rate of tau spread and how different interventions might modify propagation.
The development of more sophisticated experimental models has accelerated progress in understanding propagation mechanisms. Induced pluripotent stem cell-derived neurons from AD patients allow the study of human tau biology in relevant cell types, while improved animal models that more closely recapitulate human tau pathology enable better translation of findings to human disease [22]. Additionally, advances in cryo-electron microscopy continue to reveal the structural details of different tau conformers, potentially enabling the development of conformation-specific therapeutic interventions.
Research into the biological factors that influence propagation is ongoing, with studies investigating how genetic variations, environmental factors, and comorbidities affect the spread of tau pathology. The role of sleep, exercise, and other lifestyle factors in modulating propagation is being actively investigated, as these represent potentially modifiable factors that could influence disease trajectory. Additionally, the relationship between amyloid-beta and tau pathology continues to be explored, with evidence suggesting that amyloid pathology may facilitate the spread of tau but is not absolutely required for propagation to occur.
The integration of multi-modal biomarker data is enabling more comprehensive understanding of propagation. Studies combining tau PET, CSF measurements, neuroimaging, and cognitive testing are characterizing the sequence of events that lead from early tau pathology to clinical symptoms. This systems-level approach is identifying critical windows for intervention and biomarkers that can guide patient selection and treatment monitoring for clinical trials. The continued refinement of these approaches promises to accelerate the development of effective therapies targeting tau propagation.
Braak H, Braak E. Neuropathological staging of Alzheimer-related changes. 1991. ↩︎
Nelson PT, et al. 'Correlation of Alzheimer disease neuropathologic changes with cognitive status: a review of the literature'. 2012. ↩︎
Zhou J, et al. Functional network connectivity breaks down in default mode network in Alzheimer's disease. 2012. ↩︎
Frost B, et al. Mammalian neurons can induce tau fibrillation in vitro. 2009. ↩︎
Mandelkow EM, Mandelkow E. Tau in physiology and pathology. 2014. ↩︎
Lasagna-Reeves CA, et al. Identification of oligomers at early stages of tau aggregation in Alzheimer's disease. 2012. ↩︎
Clavaguera F, et al. Brain homogenates from Alzheimer's disease mice induce tau pathology in mouse brains. 2013. ↩︎
Fitzpatrick AWP, et al. Cryo-EM structures of tau filaments from Alzheimer's disease. 2017. ↩︎
Prusiner SB. Cell biology. 2012. ↩︎
Sanders DW, et al. Distinct tau prion strains propagate in cells and mice and define two tauopathies. 2014. ↩︎
Wu JW, et al. Small misfolded tau species are internalized via bulk endocytosis and anterogradely transported. 2013. ↩︎
Braak H, et al. 'Stages of the pathogenic process in Alzheimer''s disease: from the category of lesions to the definition of disease'. 2001. ↩︎
Zhou J, et al. Tau-related white-matter degradation in cognitively normal older adults. 2013. ↩︎
Oxford AE, et al. Connectome of the superior temporal gyrus in Alzheimer's disease. 2020. ↩︎
Tai HC, et al. Synaptic activity-related tau pathology in the human entorhinal cortex. 2012. ↩︎
Wang Y, et al. Neuronal activity-induced tau secretion from cortical neurons is associated with exosome enrichment. 2020. ↩︎
Bolos M, et al. 'The role of microglia in the spread of tau: therapeutic implications'. 2018. ↩︎
Iliff JJ, et al. A paravascular pathway facilitates CSF flow through the brain parenchyma and the clearance of interstitial waste. 2012. ↩︎
Sigurdsson EM. 'Tau immunotherapy and diagnostics: progress and prospects'. 2020. ↩︎
Wischik CM, et al. 'Tau aggregation inhibitor therapy: an exploration of long-term clinical effects'. 2020. ↩︎
Xia C, et al. Association of tau PET binding with memory decline in older adults. 2021. ↩︎
Karch CM, et al. 'Human stem cell models of Alzheimer''s disease: progress and challenges'. 2022. ↩︎