Description

The function and expression of tau. Microtubules play a critical role in supporting the structural needs of a neuron as well as facilitating intracellular transport (23). The normal physiological function of tau is to support the assembly and stabilization of microtubules (49; 26). Binding of tau promotes microtubules in their polymerized state and thereby facilitates axonal transport and maintains neuronal integrity (32).

The MAPT gene encoding tau is on chromosome 17q21. In early human development, tau is present throughout the neuron. In developed neurons, however, tau is primarily located in the axons, where it normally contributes to the maintenance of axonal health (16; 49). Grey matter expresses MAPT at higher concentrations than white matter or cerebellum. Tau has also been found to have a physiologic role in dendrites and to be expressed in low levels in glial cells (28).

In the human brain, alternative splicing of the MAPT gene produces six isoforms. Alternative splicing of exons 2 and 3 results in tau protein with either 0, 1, or 2 amino-terminal inserts. The isoforms are labeled 0N, 1N, or 2N based on the number of amino-terminal inserts present. Additionally, alternative splicing of exon 10 creates tau with either three or four microtubule binding regions. These isoforms are respectively labelled 3R and 4R. The fetal brain produces only 0N3R tau, whereas the adult brain produces all six isoforms. Importantly, the healthy adult brain expresses 3R and 4R tau in a 1:1 ratio (16).

Pathologic tau. Tau undergoes a variety of post-translational modifications. The most relevant post-translational modification to the tauopathies is likely phosphorylation, which may play a role in the trafficking and localization of tau in the cell. Notably, phosphorylation reduces tau’s affinity for microtubules. The hyperphosphorylation of tau may allow tau to aggregate and form pathologic oligomers, which, in turn, promote the formation of insoluble paired helical filaments that form the neurofibrillary tangles in Alzheimer disease, suggesting that tau hyperphosphorylation plays a role in disease pathogenesis (07; 16). More research is needed to determine whether hyperphosphorylated tau is a cause or consequence of the tauopathies (41). Other post-translational modifications that may contribute to the development of pathologic tau include acetylation, truncation, ubiquitination, glycation, and SUMOylation (38; 41).

The correlation between pathologic tau burden and disease severity suggests that tau mediates cell death. The Braak rating scale for Alzheimer disease is based on the distribution of tau deposition and tracks well with clinical manifestations of the disease (06). McKee and colleagues proposed a similar rating scale for chronic traumatic encephalopathy (37), and studies have demonstrated the utility of the rating scale with the chronic traumatic encephalopathy stage strongly correlating with the degree of cognitive impairment (01). Similar findings have been reported for variants of progressive supranuclear palsy (42). Sakae and colleagues reported increased frontal, temporal, and white matter tau pathology in patients with progressive supranuclear palsy with frontotemporal dementia compared to patients with progressive supranuclear palsy alone (42).

Studies have demonstrated multiple trajectories of tau spread among Alzheimer patients. One group examined 1143 flortaucipir PET images, of which 443 were positive for tau pathology. The distribution of tau pathology fit into four distinct subtypes. A plurality of the PET scans (S1; 32.7%) fit into a limbic-predominant subtype with a Braak-like pattern. Additional subtypes included a parietal-dominant subtype with relative sparing of the medial temporal lobe (S2; 17.8%), a subtype with early occipital lobe involvement (S3; 30.5%), and a subtype with left-sided temporoparietal involvement (S4; 19.0%). Among all these subtypes, phenotypes were consistent with the pattern of tau deposition. For example, S4 patients had the greatest degree of language impairment, whereas S1 patients had the most amnestic impairment. When tracked longitudinally, patients remained within the same subtype. Overall, this study suggests that only about a third of Alzheimer cases progress through the classic Braak pattern of tau progression (55).

It is not entirely clear how aggregated tau inclusions lead to neurodegeneration. It is possible that pathologic tau disrupts axonal transport and the function of cytoplasmic organelles. In aggregated form, tau is unable to perform its usual physiologic function, which could lead to microtubule instability (49). Gomez-Isla and colleagues quantified the neurons, senile plaques, and neurofibrillary tangles in the superior temporal sulcus of patients who had been diagnosed with Alzheimer disease (14). They found a correlation between neurofibrillary tangle burden and neuronal loss but also found that the degree of neuronal loss greatly exceeded neurofibrillary tangle burden. These findings suggest that mechanisms unrelated to neurofibrillary tangles could contribute to cell death.

In the most common tauopathy, Alzheimer disease, amyloid-beta and tau seem to act synergistically to promote neurodegeneration. The relationship of amyloid-beta and tau with respect to Alzheimer pathogenesis was initially explored in 2001 with the publication of a couple of mouse studies. One study found that injecting amyloid-beta fibrils into the murine brain induced a 5-fold increase in local tau tangle formation (15). Another study crossed mice expressing a mutant strain of tau (tauP301L) with a strain of mice overexpressing amyloid precursor protein (APP) (31). The resultant progeny developed an accelerated rate of neurofibrillary tangle formation relative to the parental tauP301L strain. These studies suggest that amyloid-beta plays an upstream role from tau in Alzheimer disease pathogenesis.

Some evidence suggests that tau and amyloid-beta are mutually influential. One study crossed mice expressing mutant APP and presenilin-1 (PS1) with tau knockout mice. Relative to the APP/PS1/tau(+/+) progeny, the APP/PS1/tau(-/-) mice demonstrated roughly a 50% reduction in cortical amyloid plaque burden. These results suggest that, with respect to Alzheimer pathogenesis, amyloid-beta and tau influence one another in a feedback loop (29; 04).

Advanced age is the strongest risk factor for developing sporadic tauopathies. Models of biological aging have provided some insight into how pathogenic tau aggregation could lead to neurodegeneration (19). For example, accumulation of DNA damage is an important driver of biological aging. A drosophila model with impaired DNA damage checkpoints demonstrated an increase in tau-induced cell death and tau neurotoxicity (24). The authors inferred that DNA damage checkpoints play a protective role against the development of tauopathies.

Categorizing the tauopathies. The human brain normally expresses the 3R and 4R isoforms of tau in a 1:1 ratio (16). Although some tauopathies retain an even proportion of 3R and 4R tau, others exhibit aberrations in this ratio. The table below categorizes some of the known tauopathies by whether the pathologic tau is 3R predominant, 4R predominant, or split between 3R and 4R.

Changes in the ratio of 3R and 4R tau seem to have several adverse effects on neurons. Alterations in 3R and 4R proportions could reduce mitochondrial localization to axons as well as alter axon transport dynamics (48). It is also possible that changes in the isoform proportions alter microtubule binding dynamics; for instance, a study by Lu and Kosik found that 4R tau could displace 3R tau from microtubules (33).

4R tauopathies are more common than 3R tauopathies. The 3R tauopathies include Pick disease, which is characterized by Pick bodies (round, intraneuronal inclusions of tau) that are predominately located in the hippocampus and frontal and temporal cortices.

The 4R tauopathies include progressive supranuclear palsy, in which neurofibrillary tangles, tufted astrocytes, and coiled bodies localize to the pons, subthalamic nucleus, and substantia nigra. In corticobasal degeneration, atrophy occurs in the frontal, parietal, and temporal cortex, along with degeneration of the substantia nigra. Pathology in corticobasal degeneration includes ballooned neurons, astrocytic plaques, coiled bodies, and argyrophilic threads. Another tauopathy, argyrophilic grain disease, is named for one of its pathologic findings; it additionally includes oligodendritic coiled bodies (16).

Tauopathies can additionally be categorized as primary or secondary. In primary tauopathies, tau aggregates are the main driver of neurodegeneration; whereas in secondary tauopathies, other pathologic elements influence tau pathology and neurodegeneration (39). For instance, the most common tauopathy, Alzheimer disease, is a secondary tauopathy in which tau-containing neurofibrillary tangles form in the presence of amyloid plaques (16).

The following table categorizes some of the more commonly encountered tauopathies by whether they are primary or secondary.

Tau propagation. In some tauopathies, specifically Alzheimer disease and chronic traumatic encephalopathy, the tangles formed by aggregated filaments remain in the extracellular space following the death of the affected cells (32). These “ghost tangles” may contribute to disease propagation throughout the brain by making tau accessible to other cells in the extracellular space. This is likely a critical aspect of the pathophysiology of tauopathies. Braak first identified that tau inclusions originated in the transentorhinal cortex in Alzheimer disease and then spread from there (06). Since then, numerous mouse studies have shown that injection of extracts with mutant tau into wild-type animals causes widespread induction of the pathologic protein into neighboring areas that then degenerated (47). Furthermore, injection into animal models of human brain tissue with different pathologic inclusions consistent with confirmed corticobasal degeneration, progressive supranuclear palsy, and argyrophilic grain disease produced lesions in the animal brains demonstrating the same pathologic and phenotypic disease as the respective tauopathy injected (32). This supports the theory of “prion-like” propagation throughout the brain and suggests distinct confirmations of tau for individual diseases (32). The presence of pathologic tau in the extracellular space is also confirmed by the ability to find tau in the cerebrospinal fluid of patients with tauopathies.

Studies have focused on understanding how pathologic tau propagates throughout the brain. Martinez and colleagues focused on aspects of the presynaptic terminal that could promote tau propagation (36). They identified a scaffolding protein of the presynaptic active zone called Bassoon (Bsn), which could help stabilize pathogenic tau seeds. Using mouse models, this group found that downregulating BSN reduced tau propagation and brain atrophy. Additionally, phenotypic effects, such as reduced behavioral impairments, followed the downregulation of Bsn.

Other processes that may be implicated in the pathophysiology of tauopathies include mitopathy and neuroinflammation. Mitopathy is a process that has been implicated in early Alzheimer disease and involves mitochondrial dysfunction and autophagic-lysosomal alterations (22). These processes have been associated in animal studies with tau pathology, but the exact relationship is yet to be understood. Neuroinflammation likely has a contributory role in the neurodegenerative process, with the simultaneous potential to be critically protective and potentially harmful. Supportive of this connection, a study demonstrated colocalization of radiolabeled markers of neuroinflammation and tau pathology in 17 patients with progressive supranuclear palsy (34). Even more compelling was the strong positive correlation between clinical severity and both subcortical neuroinflammation and tau pathology.

Evidence has supported a role for the innate immune system in the development of tauopathies. For example, the “infection hypothesis” of Alzheimer disease posits that immune challenges can provoke neurofibrillary tangle formation (23). One supportive study found evidence that pseudomonas aeruginosa pneumonia precipitates the release of lung endothelial-derived tau, which can ultimately lead to seeding and aggregation of neuronal tau (08). Additional examples of the immune system’s potential involvement in tauopathy pathogenesis include propagation through toll-like receptor signaling, dysregulated complement signaling, and disrupted autophagy of tau (23).

Clinical applications

There is a great deal of interest in developing diagnostic tools and treatment for the tauopathies, particularly Alzheimer disease. Developments in the identification and treatment of tauopathies based on the pathologic understanding of the disease are reviewed below.

Biomarkers. For decades, confirmation of tau was limited to autopsy evaluation. In the past few years, PET imaging has emerged as a way to identify tau pathology in vivo by using radiolabeled ligands against neurofibrillary tangles containing tau (35). In Alzheimer disease, the radiolabeled ligand [18F]-AV-1451 or flortaucipir was shown to successfully identify tau distribution in patients compared to healthy controls, likely due to binding to the paired helical conformation of tau in Alzheimer disease brains. Yet in other tauopathies, the data for flortaucipir to identify tau in the expected areas have been variable, and off-target binding may limit its utility in chronic traumatic encephalopathy and other disorders.

Several other ligands have also been created, and in mild traumatic brain injury and chronic traumatic encephalopathy, promising results suggest that tau can be successfully imaged in vivo. One such ligand, 11C-pyridinyl-butadienyl-benzothiazole 3 (11C-PBB3), was able to identify tau deposits in Alzheimer disease and non-Alzheimer disease tauopathies. This ligand also demonstrated increased MAPT-PET binding in the neocortical grey and white matter in patients with late-onset neuropsychiatric symptoms following traumatic brain injury (50). One small study examined patients at risk for chronic traumatic encephalopathy and was able to demonstrate a tau-PET binding profile consistent with later-stage chronic traumatic encephalopathy in patients who were negative for amyloid-PET binding (30). The authors concluded that MAPT-PET binding may represent a useful biomarker for chronic traumatic encephalopathy, but it is currently only for later-stage disease. Similar results were shown in patients with progressive supranuclear palsy who demonstrated increased tau-PET binding in the globus pallidus and putamen relative to age-matched healthy controls (45). There was, however, a positive correlation between tau-PET binding and age within both groups, which may challenge the application and clinical utility of this technology as a diagnostic tool.

Extracellular tau as measured in cerebrospinal fluid is currently used to help diagnose Alzheimer disease. A standard Alzheimer diagnostic panel includes the biomarkers amyloid-beta42, total tau, and hyperphosphorylated tau (p-tau). When detecting incipient Alzheimer disease in patients with mild cognitive impairment, combinations of these three biomarkers yielded sensitivities of 95% and specificities of 80% to 90% (18). Research identifying fluid biomarkers for tauopathies other than Alzheimer disease is ongoing. One study focused on microtubule-binding region tau (MTBR-tau275 and MTBR-tau282). This tau species showed an increase in the brains of patients with primary tauopathies, especially corticobasal degeneration and FTLD-MAPT. Concurrently, concentrations of these tau species demonstrated a proportional decrease in the CSF of these patients. MTBR-tau275 and MTBR-tau282 show promise as biomarkers for tauopathies other than Alzheimer disease (20).

Skin biopsy has been utilized to detect peripheral alpha-synuclein and help diagnose alpha-synucleinopathies, such as Parkinson disease. A study evaluated whether skin biopsy might have diagnostic utility for tauopathies (53). The researchers examined tau expression in the skin biopsies of patients clinically diagnosed with progressive supranuclear palsy and corticobasal degeneration. This was compared with the tau expression in the skin biopsies of patients with the alpha-synucleinopathies Parkinson disease and multiple system atrophy. The authors found that tau is highly expressed in the nerve fibers surrounding dermal autonomic structures and that there was a significantly increased amount of tau in progressive supranuclear palsy/corticobasal degeneration biopsies relative to Parkinson disease/multiple system atrophy biopsies. These findings suggest that skin biopsy may have utility for diagnosing tauopathies.

Disease-modifying therapies. Given the strong association between tau deposition, cellular toxicity, and neurodegeneration, tau is an obvious target for potential disease-modifying therapies. Active and passive antibody-based clinical trials are underway or have been completed for Alzheimer disease. A humanized monoclonal antibody against tau, ABBV-8E12, was well tolerated without severe adverse effects in a stage 1 trial (57); however, this drug failed to demonstrate efficacy in patients with early Alzheimer disease (11). In a similar vein, a drug targeting progressive supranuclear palsy was well tolerated (05) but failed to prove efficacious in a phase 2 clinical trial (09). Many treatment approaches that have shown promise in animal models of tauopathies have failed to prove to be safe and effective in human clinical trials. The vast majority of these trials are in Alzheimer disease and target microtubule stabilization, prevention of tau hyperphosphorylation, and reduction of tau aggregates. Valproic acid, lithium, methylene blue, and naproxen failed to show cognitive improvement in randomized trials (03; 46). Although tau is an appealing and potentially promising target for disease-modifying therapy, more research is needed to find a safe treatment that slows cognitive decline in tauopathies.

Studies continue to identify potential targets for disease-modifying therapy against tauopathies. A 2023 study identified 82 genetic regulators of tau protein levels. Using mouse tauopathy models, they found that knockdown of three of these genes (USP7, RNF130, and RNF149), which function through the C terminus of Hsc70-interacting protein, reduced pathologic tau levels and improved learning and memory deficits (26). Another study examined tetrandrine, which is thought to correct the lysosomal alkalinization that impairs the autophagy-lysosomal pathway (ALP) (51). The ALP is important for the degradation of hyperphosphorylated tau. The researchers found that, in a mouse model, tetrandrine enhanced tau neurofibrillary tangle clearance and improved memory in a dose-dependent manner. In another study, researchers generated two D-enantiomeric peptides that bind tau with a high affinity (02). In vitro, these ligands inhibit the formation of tau aggregates by stabilizing tau in its nontoxic monomeric form.

In 2023, the U.S. Food and Drug Administration approved one of the first disease-modifying therapies for Alzheimer disease, lecanemab (U.S. Food and Drug Administration 2023). This monoclonal antibody removes amyloid via binding to soluble amyloid-beta protofibrils. In a phase 3 clinical trial (Clarity AD), patients receiving lecanemab declined 1.21 points on the Clinical Dementia Rating-Sum of Boxes (CDR-SB, the primary endpoint), versus a decline of 1.66 points in the placebo group (p< 0.001) (54). Lecanemab also significantly reduced brain amyloid burden. Over the 18 months, CSF T-tau and P-tau181 steadily increased in the placebo group, whereas these biomarkers decreased in the lecanemab group. Another anti-amyloid drug, donanemab, demonstrated promising results in an 18-month phase 3 trial (44). Alzheimer drugs targeting tau remain in development. A phase 1b trial evaluating an anti-tau drug, BIIB080, found that this treatment significantly reduced tau burden (as measured by CSF and PET) among the 46 mild Alzheimer patients over 36 weeks (10). BIIB080’s effect on clinical outcomes will be evaluated in an ongoing phase 2 trial.

Overall, the substantial interest in tauopathy research continues to promote the emergence of promising new therapies.