Tau Pathology

Tau is a microtubule-associated protein whose normal job is axonal: it binds and stabilises the microtubule tracks along which a neuron moves cargo. Tau pathology is what happens when that protein detaches, becomes abnormally phosphorylated, and aggregates — first as soluble oligomers, later as the paired helical filaments that constitute the neurofibrillary tangle. The distribution of those tangles is what Braak staging measures, and it is the pathology that tracks cognitive decline most closely.

It is the single most heavily used concept in this corpus, referenced across roughly two hundred documents. That is not incidental. Much of the framework assembled here is an argument about where tau pathology starts, what starts it, and what it does on the way — and the corpus disagrees with itself, productively, on all three.

Where the corpus locates the beginning

The three-phase architecture places the earliest tau pathology not in the cortex but in the brainstem, decades before symptoms — in the locus coeruleus, on Braak's own staging. Several theses develop the mechanism by which it is kindled there: the noradrenaline metabolite DOPEGAL activates asparagine endopeptidase, which cleaves tau at N368 and, in the same stroke, cleaves SET and so silences the phosphatase PP2A that would otherwise erase the phosphorylation. One enzyme, two crimes — a writer engaged and an eraser disabled.

This is a claim about sequence, and the corpus treats it as such rather than as settled fact. What is established is the topography: tau appears in the locus coeruleus early and in the isocortex late. What is inferred is that the brainstem event is causally upstream of everything after it.

Where the corpus disagrees with itself

Three positions in the corpus are mutually incompatible on the origin of tau pathology, and the disagreement is worth preserving rather than smoothing over.

Tau as licensed by amyloid. The steelman of the amyloid cascade (research/amyloid-steelman/) argues that limbic, medial-temporal tau is common in normal ageing and relatively inert on its own; what converts it into the neocortically spreading, dementia-producing tauopathy is the presence of amyloid. In amyloid-negative individuals tau tends to remain confined; in amyloid-positive individuals it spreads. On this reading amyloid is not the toxin but the permission.

Tau as developmentally independent. Abbate's account, evaluated in research/abbate-evaluation/, holds that tau pathology is mechanistically independent of the amyloid process at its origin. It arises from the biochemistry of the neurogenic niche, where the three-repeat tau isoform expressed by migrating neuroblasts is naturally hyperphosphorylated at the same epitopes that characterise the tangles of established disease. The pathology is a developmental program running in the wrong place.

Tau as a toxicological endpoint. Roggen's submission (56) maps tau as the convergence point of an adverse outcome pathway: twenty-seven environmental neurotoxicants converge on mitochondrial dysfunction and metabolic dysregulation, cascading through oxidative stress into tau hyperphosphorylation and failed autophagy. Calpain-mediated activation of the tau kinases Cdk5 and GSK-3β correlates with Braak stage II–III and precedes both tau phosphorylation and synaptic loss.

These three cannot all be the whole story. The corpus does not adjudicate between them by assertion; each is argued in its own document and graded there.

What drives it, once started

Several independent lines in the corpus converge on the kinase–phosphatase balance rather than on tau itself:

GSK-3β as the shared writer. The noradrenergic arm reaches it through a dual-armed pincer; the inflammasome arm reaches it through IL-1β and IL-18. Two upstream systems, one kinase.
PP2A as the silenced eraser. The ungoverned microglion silences tau's phosphatase through PME-1 and SET — the mirror image of engaging a kinase, and the reason phosphorylation persists.
NLRP3 as amplifier. Ising et al. (Nature, 2019) showed inflammasome activation drives tau pathology, and that loss of ASC or NLRP3 in THY-Tau22 mice prevents the hyperphosphorylation. The corpus grades this as amplifier rather than initiator.
Cholesterol metabolism as a separate axis. Van der Kant and colleagues showed in patient-derived iPSC neurons that cholesterol metabolism regulates tau and amyloid-β independently, and that the protective effect of statins on tau depended on cholesterol esterification — an ACAT1 reaction, and ACAT1 is a contact-site enzyme.

What it does downstream

Tau oligomers, not tangles, carry most of the toxicity in these accounts. They correlate with HMGB1 and with T-cell recruitment into the brain, linking tau directly to neuroinflammation rather than leaving it as a parallel process. The tangle, on this reading, is closer to a tombstone than to a weapon — a theme the corpus repeats for the amyloid plaque.

Fit within the framework

Tau pathology sits at the Cytoskeletal Collapse Node convergence node, which is where the Roggen submission scores highest (6), alongside Transcriptional-Epigenetic Dysregulation Node (5). It is the through-line that connects Phase I in the brainstem to Phase III in the cortex, and it is the substrate on which the corpus's propagation arguments — receptor hijack, strain templating, prion-like spread — all operate.

Advanced by

Submission 56, Erwin Roggen — a tau-driven adverse outcome pathway for sporadic Alzheimer's disease. The submission's own framing is worth quoting for why it centres tau at all: the decision to focus on tau-mediated effects was triggered by research showing that tau oligomers and aberrant phosphorylation, rather than Aβ, are linked to early tau pathology and sporadic disease.

Tau Propagation · Transsynaptic tau propagation · Tau as conditional prion · Tauopathy primacy over amyloidopathy · Protein Misfolding · Proteostasis · Locus coeruleus as ground zero · Synaptic Plasticity · Cytoskeletal Collapse Node