Neurofibrillary tangles (NFTs) of tau proteins are one of the main sources of AD pathology. Tau protein is a microtubule-associated protein encoded by the MAPT gene on chromosome 17 [7]. It comes in six highly soluble isoforms produced by alternative splicing [8]. The longest isoform contains 85 possible phosphorylation sites, allowing for a high degree of heterogeneity in tau phosphorylation patterns in tauopathies [9]. The normal physiological role of tau is to induce tubulin assembly and stabilize microtubules, which are crucial for transporting molecules from the cell body of the neuron to the axon terminal [10]. Tau can take on two distinct forms while bound to microtubules – as independently diffusing molecules or in the form of cohesive envelopes. These cohesive envelopes can regulate interactions of the microtubule with other microtubule-associated proteins, such as kinesin and dynein, which migrate along the microtubule for anterograde and retrograde transport, respectively [11]. Normal tau acts as an obstacle to these motor proteins on the microtubule and modulates the balance of bidirectional transport by preferentially inhibiting kinesin over dynein (Figure 1A) [12]. Filamentous tau, however, is able to inhibit kinesin at normal physiological concentrations, thereby impairing anterograde transport [13]. Moreover, tau plays a role in synaptic vesicle control by interacting with synaptogyrin-3 [14]. Synaptogyrin-3 mediates the binding of pathological tau to synaptic vesicles, which leads to presynaptic vesicle clustering and inhibits neurotransmitter release (Figure 1B) [14]. In AD, tau undergoes a series of post-translational modifications that impact its function and contribute to disease progression. Phosphorylation is regulated in concert by kinases and phosphatases and any disruption of this balance may lead to tau hyperphosphorylation, favoring NFT formation [15]. Acetylation impairs tau-microtubule interaction and promotes tau aggregation [16]. Truncation contributes to apoptosis and tau aggregation [17]. Non-pathological tau contributes significantly to microtubule stability, whereas its post-translational modifications can destabilize the microtubules [18], thereby impairing axonal transport of molecules and organelles and consequently impairing neuronal function [19].

Accumulation of neurofibrillary tangles is one of the hallmarks of AD. Unlike amyloid plaques, which are present in subjects with no cognitive deficits and absent in some AD patients, hyperphosphorylated tau (p-tau) and NFTs correlate strongly with cognitive decline in AD patients [20]. Tau accumulation begins with the formation of pretangles, a soluble form of tau that exists at a higher density, often in monomers or dimers [21]. Pretangles also exhibit AT8 immunoreactivity but can be differentiated from mature tangles with a negative silver Gallyas staining [21]. Pretangles are classified into five stages (a, b, c, 1a, and 1b) based on the extent of their spread from the locus coeruleus (LC): pretangles in stages a, b, and c are more closely associated with the LC while those in 1a and 1b are marked by spread to the EC [21]. Hyperphosphorylation of tau results in its detachment from microtubules and self-assembly into insoluble aggregates [22]. These aggregates exist in two forms–paired helical filaments (PHFs) which are loosely intertwined and straight filaments (SFs) which are tightly wrapped [23, 24]. These two types of filaments further aggregate to form NFTs, which induces microtubule disassembly. This disruption in microtubule organization impairs axonal transport, resulting in synaptic dysfunction, neuronal loss, and reduced neural plasticity, which are critical pathological processes in AD [25].

Cognitive decline in AD initially surfaces as impairments to semantic memory, episodic memory, and visuospatial processing. These impairments manifest as difficulties in speech production, impaired encoding of recent events, and difficulties with spatial navigation [26]. As memory functions progressively decline, AD patients later present with impaired executive functions and sensorimotor symptoms which can impede their daily activities [27, 28].

Many aspects of these cognitive deficits in AD are mediated by the EC – a key component of the hippocampal network. The EC is located in the medial temporal lobe, partially enclosed by the rhinal sulcus which is involved in olfactory perception [29]. As part of the hippocampal formation, which also includes the hippocampus, the dentate gyrus (DG), and the subiculum, the EC plays a pivotal role in memory encoding and retrieval. Structurally, the EC is organized into six distinct layers, where layers I and IV are free of neurons (Figure 2) [29]. A number of subdivisions have been proposed for the EC, but it is most commonly divided into two anatomical regions, the medial entorhinal cortex (MEC) and the lateral entorhinal cortex (LEC) [30]. Despite their similar nomenclature, the MEC and the LEC have drastically different cytoarchitecture and connectivity. The MEC innervates the middle third of the DG whereas the LEC innervates the outer third. Their projections to the hippocampus also differ in that the MEC innervates proximal CA1 which has higher spatial specificity while the LEC innervates the distal CA1 [31]. Consequently, the MEC and the LEC are considered to be in two parallel processing streams into the hippocampus. Their connectivity dictates their function. The MEC is involved in processing spatial information and path integration while the LEC contributes to the processing of olfactory information [32]. These cognitive functions closely align with the initial presentations of deficits observed in AD, specifically impairments in spatial navigation and olfactory perception. Cells in the MEC are responsible for sensing the environment and the animal’s relative position in space. Subsequently, this spatial information is consolidated and transmitted via the perforant path, with layer II neurons of the MEC projecting to the DG and CA3 regions, while layer III neurons extend projections to the CA1 and CA2 regions of the hippocampus (Figure 2) [33]. The microcircuitry underlying olfactory processing initiates within the olfactory bulb, where odors are detected, and sensory information is transmitted to layer I of the LEC. Subsequently, this information is relayed to neurons residing in layers II and III of the LEC [26]. Here, the LEC integrates olfactory signals with contextual information before further relaying them to the CA1 and DG regions of the hippocampus. In both cases, CA1 and the subiculum send feedback to layers V and VI of the EC to refine the spatial and olfactory information. Given the crucial role of the EC in spatial and olfactory processing, the EC may be the key to detecting AD prior to the onset of cognitive deficits.

The EC is composed of chemically, morphologically, and functionally distinct cell populations. These cells can first be chemically distinguished as reelin- or calbindin-expressing based on the expression of reelin, a glycoprotein involved in cell-cell interactions and neuronal migration, or calbindin, a calcium-binding protein that protects neurons from Ca²⁺-induced damage [33]. These cells can be further divided into subgroups based on their morphology. In the MEC, neuronal populations consist of stellate cells and pyramidal cells, while in the LEC, they consist of fan cells, pyramidal cells, and multipolar cells. Stellate cells and fan cells within their respective regions express reelin, whereas pyramidal cells express calbindin [33]. Functionally, the MEC contains a variety of neurons involved in spatial coding, among which grid cells are most prominent in layer II and can exist as either stellate or pyramidal grid cells [34, 35]. In the following section, we will discuss the varying types of susceptibility of these cell types to tau pathology.

According to the prion model of tau seeding, tau aggregates originating from one brain region have the capacity to escape the cells and serve as “seeds,” triggering the abnormal folding of tau proteins in other brain regions, prior to any observable histopathological changes [36]. Although the locus coeruleus (LC) is the site where abnormally phosphorylated pretangle tau is first detected, the prion-like seeding activity of abnormal tau first appears in the EC. This is demonstrated using a fluorescence resonance energy transfer (FRET) biosensor targeting the aggregation-prone repeat domain of tau [4]. The extensive connections linking the EC and other components of the hippocampal formation facilitate the transport and spread of NFTs. Using an epidemic spreading model derived from position emission tomography (PET) scans collected from AD patients across different disease stages, Vogel et al. [6] conclude that the network of anatomical connections can account for 70.2% of tau spatial localization. Furthermore, their model suggests that the EC is most likely the epicenter of tau propagation. Additionally, a rhesus monkey model of tau seeding demonstrates the spread of tau to many hippocampal subfields, encompassing CA3, CA1, the subiculum and dentate gyrus [37]. These insights collectively emphasize the pivotal role of the EC in tau seeding and propagation, prompting further exploration into the precise molecular mechanisms driving the EC’s vulnerability to tau pathology.

In addition to the abnormal aggregation of tau and β-amyloid proteins, chronic neuroinflammation is also a hallmark of AD. Deposition of insoluble NFTs is a classic stimulant for inflammation in the AD brain through the activation of reactive astrocytes and microglia, which subsequently release a wide range of proinflammatory molecules: cytokines such as IL-6, IL-1β, TNF-⍺, chemokines, reactive oxygen species (ROS), and nitric oxide (NO). These mediators in turn induce cellular damage and oxidative stress to neurons, leading to neurodegeneration [38]. Additionally, inflammation leads to the activation of various kinases, such as glycogen synthase kinase 3β (GSK3β) and cyclin-dependent kinase 5 (CDK5), which contribute to tau hyperphosphorylation, resulting in the formation of PHFs and NFTs [39]. In murine models, deficiency in the triggering receptor expressed on myeloid cells 2 (TREM2), a receptor vital for the survival and proliferation of microglia, has been demonstrated to diminish neuroinflammation and provide protection against volume loss in the EC [40]. Furthermore, elevated levels of soluble TREM2 in the cerebrospinal fluid (CSF), observed in symptomatic AD patients, are associated with neuroinflammation and p-tau accumulation [41]. Another proinflammatory mediator that contributes to neuroinflammation in the EC is the glia maturation factor (GMF). GMF is upregulated in regions where NFTs are present, leading to an increase in reactive astrocytes and activated microglia [42]. Additionally, evidence indicates that the administration of anti-GMF antibodies attenuates neuroinflammation and reverses behavioral deficits in a murine model. However, the direct impact of GMF blockade on NFTs remains unclear [43]. Collectively, these studies suggest the potential of TREM2 and GMF as biomarkers for monitoring neuroinflammation and neuronal loss in the EC.

EC hypoperfusion is another key pathogenic feature in early-stage AD. Impaired neurovascular coupling has recently been proposed as a possible mechanism of AD pathophysiology. The initial findings from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) have identified an association between reduced cerebral blood flow (CBF) and tau accumulation in the EC [44]. Kapadia et al. [39] investigated the temporal relationship between these phenomena using a 6-year dataset from ADNI-2. They found that hypoperfusion in the EC precedes NFT deposition. Moreover, in the left EC, they observed a significant negative correlation between CBF and the severity of cognitive decline. Furthermore, CBF exhibited an inverse relationship with tau accumulation in the left EC over the 6-year period [45]. However, it is important to note that only one AD patient participated in the follow-up, necessitating further data to validate these findings and establish clinically relevant reference ranges for CBF in the EC. Nevertheless, these results highlight EC hypoperfusion as a potential predictor for tau presence in the EC, enabling early identification of individuals at risk of AD or in pre-MCI AD stages. These findings shed light on the role of cardiovascular diseases as risk factors for AD, emphasizing the importance of addressing these comorbidities to attenuate disease progression.

The early involvement of the EC in AD tau pathology underscores the clinical potential to combine tau clearance therapies with the use of early EC biomarkers. This approach aims to contain pathologic tau within the EC during the early stages of AD while closely monitoring the region to evaluate therapeutic efficacy. Currently, there is no cure available for tauopathy, but several classes of therapies aiming to halt its progression are in development. These therapies encompass traditional immunotherapies such as monoclonal antibodies against tau, tau vaccines, antisense oligonucleotides (ASO) targeting the MAPT gene, and inhibition of tau aggregation. Additionally, newer immunotherapies leverage mechanisms involving ubiquitin, autophagy, and lysosome for tau degradation [77]. For an extensive review of current tau-targeting therapies for AD, Congdon et al. [78] have meticulously outlined the mechanisms of action for each category of tau-targeting therapy, along with the current clinical trial status of these interventions. In this section, our focus lies primarily on tau clearance strategies in early AD and the monitoring of disease progression through biomarkers associated with EC tau pathology.

Although the cost-effectiveness and feasibility of large-scale screening for AD biomarkers in the EC may be limited, these markers may be invaluable in assessing treatment benefits. A halt in disease progression may be signified by a decrease in CSF levels of TREM2, a marker of neurodegeneration and EC volume loss, or preservation of RORB⁺ excitatory neurons. CSF reelin levels have been highlighted as a proxy for tau levels but further research is needed to evaluate any correlation with tau pathology specifically in the EC [49]. Investigating the relationship between these potential biomarkers and AD tauopathy progression, would allow for more efficient and less resource-intensive methods to assess the benefits of AD therapies.

The EC occupies a central position in AD pathology and cognitive impairment. Its significance extends beyond merely being the site of initial tau aggregation, as it also plays crucial roles in learning and memory processes, particularly in spatial cognition and olfactory learning.

The EC’s intricate connectivity with key memory-related structures like the hippocampus and neocortex underscores its involvement in memory deficits observed in AD. The detrimental effects of tau pathology in the EC manifest in various ways, affecting grid cells' firing patterns, altering electrical gradients in stellate cells, and accelerating the degeneration of specific neuronal populations, such as reelin+ stellate cells. Specific neuronal subtypes within the EC, such as RORB-expressing excitatory neurons and Wsf1-expressing pyramidal cells, are selectively vulnerable to tau pathology. Additionally, reduced levels of neuroprotective factors such as calbindin in EC layer II render it particularly susceptible to tau-induced damage. Furthermore, the presence of NFTs in the EC triggers microglial activation and loss of pyramidal cells, contributing to sustained neuroinflammation.

Understanding the cellular, physiological, and molecular characteristics of the EC not only sheds light on the mechanisms underlying AD progression but also presents opportunities for early detection of tau pathology in this region. Consequently, further research into therapeutic interventions targeting the EC is warranted to develop effective treatments for AD.

Indeed, while various vulnerabilities to tau pathology have been identified in the EC, further research is crucial to elucidate the exact roles of the markers or proteins involved. This deeper understanding is essential for translating these findings into effective therapeutic strategies for AD.

Since slowing disease progression is the primary therapeutic goal for AD, enhancing tools for early detection of NFTs in the EC is critical. Research efforts should focus on evaluating the accuracy and specificity of tau tracers specifically for the EC, as well as exploring blood or CSF AD biomarkers that reflect the histological changes observed in this brain region.

Once more reliable and cost-effective detection methods are established, it will be possible to assess the efficacy and safety of AD therapies in individuals at preclinical stages. Targeting tau pathology in the EC before cognitive deficits manifest could potentially slow disease progression and improve outcomes for individuals at risk of developing AD.