Considerations for future tau-targeted therapeutics: can they deliver?

KEYWORDS:

• Tau
• Alzheimer’s disease
• drug discovery
• tauopathy
• neurodegeneration

1. Introduction

Alzheimer’s disease (AD) is the most common cause of dementia in the elderly and a leading cause of death. The progressive accumulation and spread of β-amyloid and tau in AD brain are linked with synaptic loss and neurodegeneration. Recent evidence indicates that while alterations in amyloid precursor protein processing and β-amyloid accumulation likely trigger AD, abnormal tau acts as the executioner.

Tau is normally a highly soluble cytoplasmic protein that dynamically interacts with microtubules to stabilize the cytoskeleton and regulate axonal transport. When modified, including by hyperphosphorylation, cleavage and aggregation, tau accumulates as neurofibrillary pathology in AD and related tauopathies such as progressive supranuclear palsy (PSP) and some forms of frontotemporal dementia (FTD) [1]. Pathological tau species show altered interactions and localization, loss of physiological function, prion-like spread characteristics that promote the progressive appearance of tau pathology through diseased brain, and gain of toxic functions. For example, the mislocalization of pathological tau from the cytoplasm to synapses causes loss of presynaptic function, promotes post-synaptic excitotoxicity, and is closely associated with dementia in AD [1,2]. Here, we discuss the current state of tau-targeting therapies.

2. Targeting tau modifications

Tau phosphorylation is mediated by the activity of protein phosphatases and kinases including glycogen synthase kinase-3 (GSK-3) and fyn [1]. Reducing tau phosphorylation with kinase inhibitors decreases tau aggregation, axonal degeneration, synapse loss, cognitive and behavioral deficits in rodent models of tauopathy [3,4]. Highly selective inhibitors for kinases with improved pharmacokinetic and CNS retention profiles, as well as agents that reduce the activity of multiple tau kinases, have been developed [3]. Several of these molecules, including the GSK-3 inhibitor tideglusib and the fyn kinase inhibitor AZD0530, entered clinical trials for AD and/or PSP. Unfortunately, no kinase inhibitors have met primary endpoints and shown acceptable safety profiles [5,6], so this approach has largely fallen out of favor. A common difficulty is on-target toxicity since most kinases lie within fundamental cell signaling pathways and most have broad substrate specificity in the CNS and periphery. Future efforts may focus on brain-specific kinases such as lemur tyrosine kinase 2 (LMTK2), an upstream modulator of GSK-3 activity. LMTK2 is reduced in AD brain [7] and restoring its activity would inhibit GSK-3 specifically in the CNS, avoiding detrimental effects of peripheral inhibition.

Agents that modify tau aggregation have also been a focus for drug development since tau aggregates are seed-competent and are the primary species enabling pathological tau to spread through the brain [1]. Perhaps the most notable tau disaggregating agent to have reached clinical trials is LMTX, a derivative of methylene blue. While LMTX did not reach its primary endpoints in a phase III trial and was criticized for some aspects of the study design, related agents have shown promising pre-clinical efficacy [5,8]. TauRx and Lilly have recently partnered to investigate the effects of the next generation tau anti-aggregating agent TRx0237, with the first trial results expected in 2020. Strategies that can effectively clear tau aggregates without producing soluble toxic tau species could have enormous benefit in slowing disease progression. Cryo-electron microscopy has revealed the structure of tau aggregates which may allow development of agents that interfere with the core structure of tau filaments [9].

Recently, agents that target other tau modifications such as GlcNAcylation have entered Phase I trials following successful pre-clinical studies. While the exact mechanism of action is not precisely understood, it is believed that tau GlcNAcylation modulates tau phosphorylation and aggregation, and so GlcNAcylation inhibitors could clear abnormal tau to maintain tau function. The first studies using the O-GlcNAcylase inhibitors ASN120290 are underway for PSP. Importantly, target engagement will be aided in these studies by the use of specific PET tracers against O-GlcNAcylase enzymes [10].

3. Regulation of tau levels

Reducing tau expression protects against detrimental effects of β-amyloid in models of AD [1], leading many to believe that tau is necessary to facilitate neurodegeneration. Strategies to reduce overall tau levels have been developed, most notably using antisense oligonucleotides (ASO) to reduce tau expression. The advantage of this approach is that all forms of tau mRNA are reduced, so all types of tau are targeted equally [11]. The first phase I/II clinical trial of a tau-lowering ASO, Ionis-MAPTRx CS1, in mild AD is due to complete in early 2021. Although there are concerns that ablating tau may itself lead to neurodegeneration or interfere with physiological effects of tau [1], substantial pre-clinical evidence from mouse models showing that knocking out tau protects from toxicity at synapses [1] suggests that partial reduction of tau expression may well be beneficial.

4. Antibody-based approaches

Perhaps the most promising tau-targeted approach to date is tau immunotherapy. Asuni et al. [12] first reported that passive and active immunization of mice with antibodies against phosphorylated tau, or phospho-tau peptides, effectively reduced tau pathology and behavioral abnormalities. Early amyloid vaccination therapies were halted due to the development of encephalitis in some trial participants, but this has not been reported in tau immunotherapy studies to date, although some patients have withdrawn due to adverse reactions [4]. A promising monoclonal antibody-based tau immunotherapy using an antibody against the N-terminus of tau recently showed good tolerance in a phase 1b trial for PSP [12]. However, a similar phase II trial was halted after futility analysis [13], with the relatively late disease stage being proposed as a possible reason for the failure. A number of other tau antibodies, including those isolated from blood, show good selectivity for pathological tau and promise in pre-clinical studies [4]. These include antibodies against the mid region of tau protein such as UCB0107 which recently reported good tolerance and an acceptable safety profile in a PSP trial. Whilst it is clearly not possible to resolve existing neurodegeneration in AD, preventing tau spread to unaffected brain regions could slow or even halt disease progression.

One concern is a lack of consensus regarding the domain of tau that is important for tau pathology spread. Extracellular tau is variously reported as being full-length, C-terminally or N-terminally cleaved, phosphorylated and non-phosphorylated. Some extracellular tau species have inter-cellular signaling functions, and so designing tau immunotherapeutics based on antibody- or peptides involves some uncertainty. Another important consideration is that tau phosphorylation, cleavage, and conformation are heterogenous and dynamic, varying between individuals at equivalent disease stages and further changing during disease progression. In this context, the development of accurate biomarkers to enable measurement of specific tau modifications at baseline and during the course of treatment would be beneficial. This technology would greatly inform clinical trials and improve opportunities to administer the most effective immunotherapy for each patient.

5. Other approaches

A range of other approaches are also under investigation, that while not specifically targeting tau, may reduce tau toxicity. These include molecules that promote microtubule stability in response to loss of tau function in disease. Others studies are designed to promote the clearance of damaging tau aggregates by targeting degradation and clearance mechanisms that are known to be defective in tauopathies. These include correction of autophagic flux, modulating components of the ubiquitin proteasome system (including phosphodiesterase E4 inhibitors) or endoplasmic reticulum stress responses (including protein kinase-like endoplasmic reticulum kinase inhibitors) [4,5,8].

6. Expert opinion

Several tau-directed therapies have shown promise in transgenic rodent models of tauopathies. However, several of these treatments have since failed in clinical trials and this raised questions about the usefulness of current animal models, particularly those with supraphysiological levels of tau over-expression. Human neural cell cultures can be used to ensure that potential therapies are effective in human systems. However, these show embryonic tau splicing patterns unless cultured for prolonged periods of time. Human induced pluripotent stem cell (iPSC)-derived neurons can be integrated into adult mouse brain [14], and these chimeric models allow the study of human neural cell responses to treatment in vivo.

Clinical trial design may also be improved. Post-hoc analysis of some clinical trial data shows benefits of treatments in specific patient groups, usually those at the earliest stages of disease, with no serious co-morbidities, and lacking specific genetic risk variants. However, such highly stratified groups are unlikely to represent the general patient population. Improvements in diagnosis are also needed so that patients can be routinely identified at early disease stages when treatments are likely to show the greatest benefit. Recent advances in peripheral fluid tests for AD [15] may improve early diagnosis, and could allow more sensitive tracking of treatment effects. Finally, the availability of target engagement biomarkers still needs to be improved. Several new tau PET tracers have been developed which enable tracking of some forms of tau, and monitoring of disease stage and treatment efficacy, although the specificity of some tracers has been questioned.

Despite these issues, there is optimism that advances in our understanding of tau aggregate structure, tau functions in different cellular compartments, and improved peripheral biomarkers will allow the development of tau-directed therapies with the capacity to modify the course of AD and related tauopathies. With renewed interest in tau biology, more is emerging about the biological functions of tau and this will identify which tau species should be targeted and which are needed to support a healthy CNS.

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer Disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

This manuscript has not been funded.