Tau and Alzheimer's disease: Past, present and future

My journey with tau started when in 1974 for the first time I isolated neurofibrillary tangles of paired helical filaments (PHFs) from autopsied Alzheimer's disease (AD) brains and discovered that they were made up of a ~50–70 KDa protein on SDS‐polyacrylamide gels. Subsequently my team discovered that this PHF protein and the microtubule‐associated factor called tau were one and the same protein. However, we found that tau in neurofibrillary tangles/PHFs in AD brain was abnormally hyperphosphorylated, and unlike normal tau, which promoted the assembly of tubulin into microtubules, the AD‐hyperphosphorylated tau inhibited microtubule assembly. These discoveries of tau pathology in AD opened a new and a major area of research on tau and on the molecular pathology of this major cause of dementia in middle‐ and old‐age individuals. Tau pathology, which without fail is made up of the aggregated hyperphosphorylated state of the protein, is also the hallmark lesion of a family of around 20 related neurodegenerative diseases, called tauopathies. Currently, tau pathology is a major drug target for the treatment of AD and related tauopathies. Both active and passive tau immunization human clinical trials at various stages are underway. Initial results range from negative to partially promising. Future studies will reveal whether tau therapy alone or in combination with drugs targeting Aβ and/or neurodegeneration will be required to achieve the most effective treatment for AD and related disorders.


| INTRODUCTION
Alzheimer's disease (AD) was discovered by Alois Alzheimer in 1907 as a neurodegenerative disorder with numerous neurons with neurofibrillary tangles and extracellular neuritic plaques in the brain associated with dementia (Alzheimer, 1907).Clinical-pathological correlation studies revealed that the density of neurofibrillary tangles and not Aβ plaques correlates with dementia (Tomlinson et al., 1970).
In 1974, for the first time, neurofibrillary tangles were bulk-isolated from autopsied AD brains and were found to be constituted of a $50-KDa protein (Iqbal et al., 1974).Subsequent studies led to the discovery that the neurofibrillary tangle protein and the microtubuleassociated protein tau reported from porcine brain in 1975 (Weingarten et al., 1975) were one and the same (Grundke-Iqbal, Iqbal, Quinlan, et al., 1986), but in AD, tau was abnormally hyperphosphorylated (Grundke- Iqbal, Iqbal, Tung, et al., 1986;Iqbal et al., 1986).Furthermore, unlike normal tau, which stimulated microtubule assembly, the AD hyperphosphorylated tau was found to inhibit the assembly by sequestering normal tau in a prion-like fashion (Alonso et al., 1994;Alonso et al., 1996;Iqbal et al., 1986).A detailed history of development of the tau field was reviewed previously (Iqbal et al., 2016).In this article, the status of tau-based drug development for the treatment of AD and related tauopathies is described.et al., 2023;van Dyck et al., 2023).The small beneficial effect of Aβ immunotherapies on cognitive function suggests that inhibition of either both Aβ and tau pathologies or in addition, the inhibition of neurodegeneration, might be required for the most effective therapeutic effect.

| TAU-BASED DRUG DEVELOPMENTS
Tau has been a major drug target for the treatment of AD and related tauopathies for the last several years, resulting in over 30 drugs to inhibit or prevent aggregation and expression of tau, and promotion of its clearance from the brain.So far, none of these studies have shown significant clinical effect (Cummings et al., 2022;Imbimbo et al., 2022).The reasons for the failures of these drugs are uncertain, but one reason could be the selection of the wrong targets in the wrong patients.The two major therapeutic approaches to inhibiting tau pathology are (1) to inhibit hyperphosphorylation of tau and (2) to clear tau pathology by active or passive immunization.

| INHIBITION OF HYPERPHOSPHORYLATION OF TAU
Tau is hyperphosphorylated both by the action of one or a combination of proline-dependent protein kinases (PDPKs), such as glycogen synthase kinase-3 (GSK-3) and cycline-dependent kinase-5 (cdk-5), and non-PDPKs, such as protein kinase A (PKA) and calcium, calmodulin-dependent kinase II (CamK-II) (Sironi et al., 1998;Wang et al., 2007).A Phase II clinical trial in patients with mild to moderate AD to inhibit hyperphosphorylation of tau by inhibiting GSK-3 activity with oral administration of tideglusib was unsuccessful (Del Ser, 2010).The hyperphosphorylation of tau is regulated mainly by protein phosphatase 2A (PP2A) (Bennecib et al., 2000;Gong et al., 2000;F. Liu et al., 2005), and the activity of this enzyme is F I G U R E 1 Alzheimer's disease (AD) pathology involves a vicious cycle of neurodegeneration, Aβ and hyperphosphorylated tau pathologies.AD is a multifactorial disorder that appears to involve several different mechanisms.Independent of the etiology, whether familial or sporadic forms of the disease, it is characterized by brain neurodegeneration associated with Aβ and tau pathologies.Aβ immunotherapies have shown, at best, only a partial rescue of the disease.At present, it is not known whether inhibition of both tau and Aβ pathologies alone or also combined with inhibition of neurodegeneration will be required for the most effective treatment of AD.Our hypothesis is that highly effective inhibition and prevention of AD can be achieved by shifting the balance from neurodegeneration to neural regeneration and inhibition of associated tau and Aβ pathologies.Both physiological and reparative regeneration occurs throughout life in human brain to sustain tissue homeostasis and repair as well as synaptic renewal through synaptic turnover (Martino et al., 2011;Sudhof, 2021).AD brain responds to neurodegeneration by stimulating neural regeneration, but the diseased brain lacks sufficient appropriate neurotrophic activity, and hence this attempt fails (Li et al., 2008).Although studies have implicated both Aβ and tau pathologies in causing cognitive impairment, neurodegeneration is the most proximal pathology to the development of dementia in AD patients.Substantial evidence suggests that the Aβ and hyperphosphorylated tau (ptau) oligomers, rather than Aβ plaques and tau neurofibrillary tangles (NFTs), are the primary cause of neurodegeneration and dementia.Plaques and tangles are defence responses of the diseased brain whereby it forms, respectively polymers of Aβ and hyperphoshorylated tau, which are inert space occupying lesions; in the case of NFTs as intraneuronal lesions, they appear to eventually choke the cells to death.Development of compounds such as P021 that can shift the balance in the treatment of AD from neurodegeneration to neural regeneration and consequently inhibit both tau and Aβ pathologies represents a novel and promising therapeutic approach.A recent study showed the therapeutic benefit of inhibition of neurodegeneration by reduction of long non-coding RNA MEG3 in a human neuron xenograft mouse model of AD (Balusu et al., 2023).compromised in AD brain (Gong et al., 1993(Gong et al., , 1994(Gong et al., , 1995)).However, to date, no AD clinical trial with an effective PP2A activator has been reported, and this potential therapeutic approach deserves further investigation.

| TAU IMMUNOTHERAPY
Tau pathology is understood only as the intracellular accumulation of hyperphosphorylated protein, except (1) that by an unconventional mechanism, tau is released from neurons in the extracellular space, and (2) that after neuronal death, the neurofibrillary tangles remain as tombstones in the extracellular space.The hyperphosphorylated tau from the tombstones, also known as ghost tangles, can become the source of tau seeds for the seeding and spread of tau pathology (Hu et al., 2016).Seeding of tau pathology is believed to spread from neuron to neuron, and thus extracellular tau can serve as a drug target for AD and related tauopathies.Although about 0.1%-0.3% of an antibody can pass through the blood-brain barrier, only a small fraction of tau antibodies to Ser 396/404 were reported to enter the neuron via clathrin-dependent Fcγ receptor endocytosis (Congdon et al., 2013;Gu et al., 2013).Thus, unless tau antibodies can enter neurons in sufficient quantities, primarily the only target of tau immunotherapy is the extracellular tau (Ji & Sigurdsson, 2021).The level of extracellular tau in aged human brain is around 300 pg/mL in the cerebrospinal fluid (CSF), and this level is elevated by about 2-to 3-fold in the brains of AD patients.Thus, only about 200 ng (1 ng/mL Â $ 200 mL CSF = 200 ng) of extracellular tau is targeted by a tau antibody, suggesting that only a few micrograms of the antibody in the brain should be sufficient to inhibit the seeding and spread of tau pathology.Both active and passive immunization studies by several laboratories in transgenic rodent animal models were found to be very encouraging (Asuni et al., 2007;Ayalon et al., 2021;Boimel et al., 2010;Boutajangout et al., 2011;Chai et al., 2011;Collin et al., 2014;Dai et al., 2015;Dai et al., 2017;Dai et al., 2018;Gu et al., 2013;Sankaranarayanan et al., 2015;Theunis et al., 2013;Troquier et al., 2012;Umeda et al., 2015;Yanamandra et al., 2013).Based on these studies, one active and several passive immunotherapy clinical trials in humans are currently underway (Cummings et al., 2022).
Results of a phase II clinical trial of active immunization against tau 294-305 carried out by Axon Neuroscience (Bratislava, Slovakia) in patients with mild to moderate AD were overall negative, but post hoc analysis revealed its effectiveness in young patients with recent onset and a high level of tau in the CSF (Novak et al., 2021).Results of a Phase II clinical trial of passive immunization with an antibody to tau 25-30 using dosages of $250-750 mg in patients with progressive supernuclear palsy, and of a Phase II clinical trial in patients with mild AD, both carried out by AbbVie Pharma, were negative (Florian et al., 2023).Results of a Phase II clinical trial with semorinemab, an antibody to tau 9-23, carried out by Genetech and AC Immune at dosages of 1500-8100 mg in patients with MCI to mild AD, were negative (Teng et al., 2022).All of these tau immunotherapy trials were found to be safe, with no serious side effects, such as the brain edema and microhemorrhages found with Aβ immunotherapy.
However, for reasons currently not clearly understood, these tau immunotherapy trials failed to show any significant therapeutic effect.
In the case of semorinemab, even at a dosage of 8100 mg, no decrease in tau pathology was found with tau positron emission tomography.One possibility is that the tau antibodies used so far in the human clinical trials had low affinity and specificity.Alternatively, because the extracellular tau targeted by the antibodies used in the clinical trials was a non-toxic waste product, its removal had no significant therapeutic effect.
We generated several high-affinity mouse monoclonal antibodies to tau and found that antibody 43D against Tau 6-18 could inhibit both tau and Aβ pathologies and reverse cognitive impairment in the 3ÂTg-AD transgenic mouse model of AD (Dai et al., 2017).Furthermore, we found that antibody 43D could inhibit the seeding and spread of tau pathology induced with hyperphosphorylated tau isolated from AD brain tissue (AD ptau) in 3ÂTg-AD mice (Dai et al., 2018).We have successfully humanized and optimized antibody 43D, and its further development is currently underway.Based on preclinical studies in mice, its calculated therapeutic dose for humans is only $50 μg/Kg, which is a small fraction of the $5 mg-$ 111.5 mg/kg used in Phase II AD clinical trials by some pharmaceutical companies.

| PREVENTION OF NEURODEGENERATION AND TAU AND Aβ PATHOLOGIES WITH A NEUROTROPHIC COMPOUND: A NOVEL THERAPEUTIC APPROACH
Both tau and Aβ pathologies are associated with neurodegeneration in AD.We have discovered that a small peptidergic compound, P021, which promotes neurogenesis and neural plasticity, can inhibit neurodegeneration and both tau and Aβ pathologies in the 3ÂTg-AD transgenic mouse model of AD (Baazaoui & Iqbal, 2017a, 2017b;Kazim et al., 2014;Li et al., 2010).P021 is orally bioavailable, has $90% stability in gastric juice, and $>97% stability in intestinal fluid at 37 C.It has a plasma half-life of over 3 h and is blood-brain barrier-permeable.P021 competitively inhibits leukemia inhibitory factor (LIF) in JAK-STAT3 signaling and increases the CREB activity and the transcription and expression of brain-derived neurotrophic factor (BDNF).
BDNF binds to its receptor, TrkB, which leads to activation of the PI3Kinase pathway, and downstream inhibitory phosphorylation of GSK-3 kinase at Ser 9 and, consequently, prevention of hyperphosphorylation of tau and tau pathology.The inhibition of GSK-3 activity also leads to reduction of Aβ pathology through prevention of phosphorylation of APP at Thr 668 (Kazim et al., 2014).
In preliminary studies, P021 did not show any off-target binding activity or any adverse effects, even at $550-fold the therapeutic dose in rats (unpublished data).Because of the ability of P021 to inhibit neurodegeneration and promote neural regeneration, it showed therapeutic efficacy in animal models of Down syndrome, traumatic brain injury, autism, macular degeneration, and cognitive aging (Bolognin et al., 2014;Chohan et al., 2011;Kazim et al., 2017;Kazim et al., 2014;Kazim & Iqbal, 2016;Khatoon et al., 2015;Y. Liu et al., 2019).Currently, further development of this compound for human clinical trials is underway.

2
| AD: A VICIOUS CYCLE OF NEURODEGENERATION, TAU PATHOLOGY, AND Aβ PATHOLOGY Neurodegeneration, tau pathology, and Aβ pathology are the three histopathological hallmarks of AD, each of which can lead to the other two, forming a vicious cycle (Figure 1).After many failures over a period of almost 30 years and an expense of $$20 billion, the U.S. Food and Drug Administration (FDA) approved two Aβ immunotherapies: aducanumab/Aduhelm ® developed by Biogen Pharma, on June 7, 2021, and lecanemab/Lecambi ® developed by Eisai Pharma, on January 7, 2023.A third Aβ immunotherapeutic drug, donanemab developed by Eli Lilly, is expected to receive approval from the FDA later in 2023.However, for reasons currently not clearly understood, in Phase III clinical trials in mild cognitive impairment (MCI) to mild AD cases, these three Aβ immunotherapies, although very effective in clearing Aβ plaques from the brain in AD patients, only reduced the rate of cognitive decline by $27%-40% as compared with placebotreated controls over a period of 18 months (Soderberg