Touting tau targeting at the time of uncertainty

Last week was tough for the drug development and research communities focusing on Alzheimer's disease. After reaching Phase III clinical trail stage, Biogen Inc and its partner Eisai Co Ltd announced that they were stopping two trials of aducanumab, a human monoclonal antibody that targets amyloid deposits of Aβ peptide formed from amyloid precursor protein (APP). These deposits are an easily recognizable feature in brains of Alzheimer's patients, and believed to be linked to pathology of the disease, which is accompanied by progressive decline in many cognitive functions and behavioral changes.

The disease is unstoppable, and its burden on the patients, their families, friends and caregivers is extreme. This makes the news of another drug trial failure devastating for many on a personal level. The current state of the affairs is also extremely concerning from the public health perspective, given that based on a forecasting study done more than a decade ago, the number of people battling Alzheimer's will rapidly increase as the global population ages, and reach 1 in 85 prevalence by 2050.

Is stopping and/or preventing this avalanche possible? Unfortunately, Biogen/Eisai case is just the last bit of bad news in a long line of disappointments in this area. This has led many to question whether the current hypothesis that targeting amyloid deposits of Aβ peptide will have clinical benefits is correct. Basically, if Aβ amyloid is not causatively linked to the disease, targeting it is not going to make a difference, suggesting that we should look for targets elsewhere.

One such alternative target is tau protein, a microtubule-binding protein. Pathogenic form of tau is hyperphosphorylated, resulting in protein misfolding and aggregation that have been associated with a range of neurodegenerative diseases collectively referred to as tauopathies. These include not only Alzheimer’s disease, but frontotemporal dementia (the most common dementia in those younger than 60), progressive supranuclear palsy and others as well. In 2018, 54% of agents in phase III clinical trials for Alzheimer's disease were targeting Aβ amyloid (aggregation and other aspects), while only 4% were directed at tau (n=1; targeting aggregation). In the same year, 9 phase II clinical trials in this area were going after tau, with 4 (out of 9 total) directed at tau aggregation, mostly indirectly by mechanisms not aimed at dissociating pathogenic tau aggregates.

New work published earlier this week approaches targeting tau in the context of frontotemporal dementia from a unique perspective. Instead of seeking to prevent tau hyperphosphorylation, or tau protein production, as some other strategies have done, the authors have focused on marking early toxic tau species for clearance through the proteasome system. In this way, toxic tau species is degraded and removed from the affected neurons, thus erased without touching the native, normally functioning tau in healthy neurons.

The proteasome system is sometimes called the cellular garbage disposal due to its ability to recognize proteins that are no longer functionally needed or have been deemed detrimental for health of the cell, and chew them up. What signals to the proteasome that a protein is ready for degradation is a special tag made of small molecule ubiquitin chain, which is attached to target protein via the enzymes called E3 ubiquitin ligases. We have recently learned how to use small molecules, called degraders, to hijack this system and degrade proteins of clinical interest at will. The most common type of degrader molecules are PROTACs, bispecific molecules with two handles, one that recruits an E3 ligase and the other that binds to the target protein of interest. This means that to be able to make an effective PROTAC you have to have a ligand that binds your target of interest and a ligand that binds an E3 ligase, which can then be linked together. Plus: the linker needs enable the interactions between the target and the ligase that will result in target's ubiquitination, a prerequisite for subsequent proteasomal degradation.

So, this new work, done in collaboration between Stephen Haggarty and Nathanael Gray laboratories, and led by M. Catarina Silva and Fleur M. Ferguson, the authors had a marvelous idea: why don't we use 18F-T807, one of the most clinically advanced PET tracers for tau that binds the pathological form of this protein, to develop a degrader? And this is what they did - they linked T807 to pomalidomide, a molecule that binds cereblon, an E3 ubiquitin ligase to create QC-01-175, a degrader molecule that resulted in clearance of tau in frontotemporal dementia patient-derived neuronal cell models. At the same time, tau in neurons of healthy controls was not affected.

There is lot to be excited about here as the results suggest that tau-directed degrader molecules that selectively affect toxic forms of tau are feasible. But the authors rightly point out that there is a lot that needs to be done first before we can celebrate. For example, QC-01-175 is quite large and may not be able to reach tau in the brain, and further medicinal chemistry optimization will be essential to address this. Additionally, the compound did lead to degradation of several zinc-finger transcription factors, proteins that are known off-targets of IMiD-based PROTACs, such as pomalidomide used here. Perhaps using other cereblon-recruiting handles might help, or developing T807 PROTACs that recruit different E3 ligases.

To me this study is worth talking about because it shows that you can make a degrader out of any selective binder, like a PET probe. This may open future opportunities for systems where we have FDA-approved selective imaging agents which have been proven to be safe and effective, but lack other types of agents, like inhibitors. It also seems worth highlighting given the current state of the affairs in Alzheimer's disease drug development. Perhaps going after pathogenic tau may prove to be more relevant than going after the Aβ amyloid?