A Half Century of Tau

From its original trademark as an esoteric factor that stimulates tubulin polymerization into microtubules (MTs) (Weingarten et al., 1975), a nevertheless classic finding in the annals of cytoskeleton history, tau eventually gained notoriety as a protein whose malfunctioning contributes to a multitude of neurodegenerative diseases. Just a few examples of such disorders include Alzheimer’s disease (AD), Parkinson’s disease, chronic traumatic encephalopathy, Pick’s disease, progressive supranuclear palsy, corticobasal degeneration and frontotemporal dementias. All of these brain diseases share something in common: the presence of intracellular filaments made from abnormally phosphorylated tau, which most commonly form in neurons and are not found in corresponding healthy cells. So pervasive is the involvement of tau in these diseases that they came to be known as “tauopathies.”

Tau’s roots as an object of scientific and medical interest can be traced to the dawn of modern cell biology. One of the holy grails of cell biology from the mid 1960s to the early 1970s was the in vitro assembly from tissue and cell extracts of MTs, which were first reported in a 1956 electron microscopic study of centrioles (De Harven et al., 1956), and eventually were provided with their now familiar name in 1963 by David Slautterback (Slautterback, 1963). A principal rationale for this challenging goal was to accelerate biochemical and biophysical understanding of MT-dependent processes, most notably mitosis (Inoue et al., 1967). A crucial step in this journey was the finding by Gary Borisy and Ed Taylor that brain is among the richest tissue sources of a protein, later dubbed “tubulin”, that binds the mitotic spindle poison, colchicine, which therefore presumably represented the principal subunit protein of MTs (Borisy et al., 1967). Building on that key discovery, among others, the first reports of MT polymerization from mammalian brain cytosol were published in late 1972 and early 1973 by Dick Weisenberg (Weisenberg, 1972), Borisy and Joanna Olmsted (Borisy et al., 1972), and Mike Shelanski, Felicia Gaskin and Charles Cantor (Shelanski et al., 1973).

The ability to polymerize MTs from brain extracts quickly led to interest in proteins that co-purify with tubulin, and might therefore regulate MT dynamics and functions. High molecular weight MT-associated proteins, or HMW MAPs (Bloom et al., 1985; Bloom et al., 1984; Murphy et al., 1975; Sloboda et al., 1975), quickly attracted the lion’s share of attention, but Marc Kirschner’s lab focused instead on the “tau factor” that they discovered (Weingarten et al., 1975) and was ignored by virtually all other MT labs at the time, at least in terms of their own experiments. An eloquent and personal tale of tau’s discovery can be found in the article by Kirschner in this volume (add reference when it becomes available).

Fascination with the HMW MAPs and tau eventually gave way to an explosion of interest in MT motor proteins, the kinesins (Hirokawa, 1998; Lasek et al., 1985; Vale et al., 1985) and cytoplasmic dynein (Lye et al., 1987; Paschal et al., 1987), but a resurrection of interest in tau began in 1986, when Inge Grundke-Iqbal, Khalid Iqbal, Skip Binder and colleagues published the first evidence that the neurofibrillary tangles (NFTs) in AD brain contain abnormally phosphorylated tau (Grundke-Iqbal et al., 1986). Two years later, a pair of back-to-back papers in Neuron by Ken Kosik (Kosik et al., 1988) and Jun Kondo (Kondo et al., 1988), and their colleagues showed that epitopes and peptides collectively spanning the entire tau molecule were abundant in isolated paired helical filaments (PHFs), bundles of which correspond to NFTs. While these late 1980s papers established the presence of tau in PHFs/NFTs, it was not until the following decade that Binder (Wilson et al., 1995), Michel Goedert (Goedert et al., 1996), and their colleagues showed that filaments resembling PHFs can be assembled in vitro from purified recombinant human tau. The proof that tau is the principal subunit of PHFs and otherwise similar straight filaments in the brains of AD and other tauopathy patients was thus finally in hand.

Despite this compelling set of evidence for the involvement of tau in neurodegeneration, tau languished in obscurity for many years compared to the amyloid-β (Aβ) peptides, which form the amyloid plaques in AD brain, and unlike tau, were linked by copious genetic evidence to familial early onset AD (Selkoe, 2001). Although no tau mutations have yet been shown to cause AD, since 1998 more than 40 tau mutations have been found to be highly penetrant for non-Alzheimer’s tauopathies (Hutton et al., 1998; Poorkaj et al., 1998; Spillantini et al., 1998; Strang et al., 2019). Tau’s involvement in the pathogenesis of neurodegeneration thus became indisputable, and a flood of studies showing pathogenic pathways leading from Aβ through tau in AD has proceeded steadily since the early 2000’s (Bloom, 2014; Götz et al., 2001; King et al., 2006; Lewis et al., 2001; Rapoport et al., 2002; Seward et al., 2013).

This special issue of Cytoskeleton was inspired by tau’s discovery about 50 years ago. Although the first paper describing tau was published in 1975 (Weingarten et al., 1975), evidence of its existence was obtained beforehand, so celebrating its golden anniversary now does not seem premature. The overriding theme of the issue is the basic biology of tau, including its biomedical impact, but excluding purely clinical topics. We thank all of the authors who contributed their valuable time to bring to fruition this special issue, which we hope will serve as a stark reminder of the importance of basic scientific discovery to the understanding of human disease.