WikiAdultCognitiveDisease.com

The Temporal Architecture Network

Not three diseases but one homeostatic system failing in three successive cell populations. Each node is a phase, a bridge, or a convergence node. Edge thickness and brightness encode connection strength along the causal spine and the node cross-links. Drag nodes to explore; hover for details.

Drag nodes to rearrange · Hover edges for connection details · Node size = number of connections
Phase I — Bioenergetic Ignition
Phase II — Microglial Bridgehead
Phase III — Synaptic Disintegration
Bridge 1 — Locus Coeruleus
Bridge 2 — Proteolytic Turn
Convergence node

The APOE4 Hub

A single polymorphism — the APOE ε4 allele — activating 8+ pathogenic mechanisms, both loss- and gain-of-function, that feed nearly every convergence node across all three phases. Click any step to expand.

The Single Structural Defect
APOE4: Zero Cysteine Residues
APOE4 Hub
ApoE4 carries arginine at positions 112 and 158, where ApoE2/E3 have cysteine, eliminating the disulfide bridge and leaving its polyunsaturated lipid cargo structurally unprotected.
Molecular detail: ApoE2 (Cys112, Cys158) and ApoE3 (Cys112) can form disulfide-linked complexes that conceal PUFA-phospholipid cargo within a protective hydrophobic core. ApoE4 (Arg112, Arg158) remains permanently monomeric, leaving the highly oxidation-prone double bonds of DHA and arachidonic acid exposed to the brain's oxidative environment. This one structural difference is the root from which 8+ downstream mechanisms fan out. As the strongest common risk factor, the ε4 heterozygote carries OR ≈ 3–4 and the ε4/ε4 homozygote OR ≈ 12–15 versus ε3/ε3 (ε4 allele frequency ≈ 14%); ε2 is protective (OR ≈ 0.6).
Eight Mechanisms Fan Out Across the Phases
1. Lipid Peroxidation of Unprotected PUFA Cargo
Phase I
Exposed PUFAs undergo peroxidation → reactive aldehydes (4-HNE, acrolein, MDA) → oxidative load compounding the catecholaminergic burden that ignites the locus coeruleus first.
Where it lands: Phase I — Bioenergetic Ignition. The locus coeruleus and brainstem aminergic nuclei already carry a heavy catecholamine oxidative load with no metabolic reserve; unshielded ApoE4 lipid cargo adds reactive-aldehyde stress to the earliest-failing neurons. Reactive aldehydes prime the NLRP3 effector arm and feed the oxidative-stress / ferroptosis convergence layer.

Key molecules: APOE, PUFA (DHA/arachidonic acid), 4-HNE, ALOX15, ACSL4, GPX4.
2. Meeting the TOMM40 Mitochondrial-Import Gate
Phase IMetabolic-Homeostatic Axis
The TOMM40 locus (rs10524523 poly-T, in linkage disequilibrium with APOE4) sits at the mitochondrial protein-import gate, where Aβ obstructing TOM40 stops import-based mitochondrial repair.
Where it lands: Phase I — Bioenergetic Ignition, at the Metabolic-Homeostatic Axis. PARP-1 hyperactivation and NAD⁺ depletion cripple mitophagy and biogenesis; obstruction of the TOM40 import channel blocks the very repair pathway the neuron needs. The APOE4 hub and the TOMM40 import gate meet at the organelle.

Key molecules: TOMM40, TOM40, TIMM23, PARP1, CD38, NAMPT, PPARGC1A, PINK1.
3. Retromer / Endosomal Trafficking Jam
Phase IEndosomal Nexus
ApoE4 impairs retromer function and prolongs endosomal dwell time → the "endosomal traffic jam," the earliest known cellular phenotype — a quality-control lesion that recurs later in cortical neurons.
Where it lands: the Endosomal Nexus, the compartment where independent pathways jam. ApoE4 reduces VPS35-complex and SORL1/SORLA efficiency so cargo (APP, ApoER2, SORL1) fails to recycle → enlarged endosomes. This retromer/endosomal-trafficking lesion is a Phase-I quality-control failure that recurs in cortical neurons in Phase III. Autophagy-lysosomal failure (BIN1 at the formation end, presenilin at the acidification end) drives the de-acidified lysosome toward PANTHOS.

Key molecules: SORL1, VPS35, VPS26, LRP8, APP, BACE1, Rab5, PSEN1, cathepsins, v-ATPase.
4. Complement Activation & Microglial Priming
Phase IINeuroimmune Interface
Altered lipid-raft organization plus complement activation; APOE is a top disease-associated-microglia marker, amplifying the microglial-complement pruning cascade.
Where it lands: Phase II — the Homeostatic Microglial Bridgehead, at the Neuroimmune Interface. APOE is one of the defining markers of the TREM2-dependent DAM transition; ApoE4 amplifies C1q/C3/C4d deposition on synapses and CR3-mediated engulfment. It compounds the collapse of the TGF-β/SMAD homeostatic microglial signature, in which protective withdrawal and destructive attack become a single inseparable act.

Key molecules: APOE, TREM2, TYROBP, C1QA, C3, C4, LILRB2, CR3, SPP1, CST7.
5. Reduced Excitatory Tone
Phase IIICompensatory Paradigm
ApoE4 lowers excitatory tone; Aβ and tau are reinterpreted as initially adaptive compensation for glutamatergic/NMDA insufficiency that becomes a maladaptive scar.
Where it lands: Phase III — Synaptic Disintegration, at the Compensatory Paradigm Nexus. Reduced excitatory tone drives compensatory APP processing; picomolar Aβ transiently boosts glutamatergic transmission and tau clears microtubule tracks — decades of allostatic compensation that eventually exhaust, leaving plaques and tangles as the maladaptive compensatory scar. This node itself encodes the cause-vs-consequence tension the corpus flags as central.

Key molecules: APOE, GRIN2A, GRIN2B, DLG4, APP, MAPT.
6. Ephexin5–RhoA → Dendritic-Spine Collapse
Phase IIICytoskeletal Collapse
ApoE4 activates the Ephexin5–RhoA–ROCK actin-myosin contraction pathway → dendritic-spine collapse, one of several converging attacks on the synaptic cytoskeleton.
Where it lands: Phase III, at the Cytoskeletal Collapse node, where the actin and microtubule cytoskeleton is attacked from multiple directions. Ephexin5 (ARHGEF15)–RhoA–ROCK contraction converges with LIMK1 hyperactivation / cofilin inactivation (from the GSK3β cascade) and complement-mediated spine destruction on the loss of dendritic spines.

Key molecules: ARHGEF15 (Ephexin5), RhoA, ROCK, GSK3B, LIMK1, cofilin (CFL1), DLG4 (PSD-95).
7. Reelin Competition at ApoER2/VLDLR
Phase III
ApoE competes with reelin at the shared ApoER2/VLDLR receptors, degrading reelin signaling and lifting its Dab1→GSK-3β brake on tau.
Where it lands: Phase III. Reelin is the cortical architect that, through ApoER2/VLDLR → Dab1, brakes tau. ApoE4 occupies the same receptors (apoE-vs-reelin competition) and its reactive lipid aldehydes arrest ApoER2–Dab1 recycling, so the brake is withdrawn just as the perineuronal net that stages the reelin signal is being digested.

Key molecules: APOE, LRP8 (ApoER2), VLDLR, RELN, DAB1, GSK3B, MAPT.
Convergence — One Allele, Every Phase
The Hub That Feeds Nearly Every Node
GroundworkPhase IPhase IIPhase IIIConvergence
One polymorphism reaching into every phase: lipid peroxidation and the TOMM40 gate in Phase I, complement priming in Phase II, and endosomal, cytoskeletal, compensatory and reelin lesions in Phase III.
Why the hub is central:
• It acts both by loss-of-function (no disulfide shield, failed recycling, withdrawn reelin brake) and gain-of-function (reactive-aldehyde toxicity, complement amplification).
• It touches the Endosomal Nexus, the Neuroimmune Interface, the Cytoskeletal Collapse node and the Metabolic-Homeostatic Axis — feeding nearly every other convergence node, and meeting the TOMM40 import gate at the organelle.
• As inherited groundwork it sets the loaded dice, but it does not act alone: COLBOS / Christchurch (APOE3ch) carriers resisted pathology for decades, so APOE4 marks where the system is vulnerable, not the sole author of the outcome.

Key insight: No other common allele hits this many independent mechanisms across all three phases at once — which is why it is the strongest common genetic risk factor, and why it sits at the center of the network as a convergence node rather than a linear cascade.

Anatomical Progression & Disease Timeline

The architecture is anatomically directional — brainstem → limbic → cortex, the Braak rostral order — and substrate-shifting (metabolic → immunological → structural). Hover over regions in the diagram; click cards for detail.

Phase I · Bioenergetic Ignition
Age 20–50 · Brainstem/LC · clinically silent
→
Phase II · Microglial Bridgehead
Age 50–70 · Hippocampus · earliest MCI
→
Bridge 2 · Proteolytic Turn
The threshold · net digestion begins
→
Phase III · Synaptic Disintegration
Age 70+ · Cortical PV/PNN · dementia
LC Raphe EC/input Hipp/CA1 Amyg Cortex PV L III–IV / PNN Brainstem → limbic → cortex (Braak rostral order)
Locus Coeruleus (LC)
Pre-tangle stage a/b · Earliest pathology · Age 20–50 · clinically silent
Phase I Metabolic-Homeostatic Axis
The autonomous-pacemaking, vastly-arborized, catecholamine-loaded LC has no metabolic reserve, so age-dependent quality-control erosion crosses from compensated to decompensated here first. PARP-1 hyperactivation and NAD⁺ depletion cripple mitophagy; Aβ obstructing TOM40 stops import-based repair. Pretangle tau accumulates from early adulthood.
Key genes: PARP1, TOMM40, DBH, SLC6A2 (NET), MAPT (early pTau), CD38
Dorsal Raphe
Pre-tangle · Serotonergic aminergic nucleus
Phase I
A brainstem aminergic nucleus that, like the LC, is metabolically extravagant and among the earliest to accumulate pretangle tau. Early involvement here contributes to the silent noradrenergic/serotonergic prodrome — disturbances of sleep, mood and arousal that precede cognitive decline by years.
Key genes: TPH2, MAOA, MAPT
Entorhinal / Hippocampal Input
Braak I · The Locus Coeruleus Bridge arrives (Age ~50–65)
Bridge 1 Phase II
The single ascending LC projection delivers two cargoes together here, densest where LC innervation is densest: withdrawal of the noradrenergic brake on microglia (β2-AR→cAMP→PKA→NF-κB suppression lost, raising the inflammatory set-point) and trans-synaptic templated tau seeds internalized via LRP1 and heparan-sulfate-proteoglycan endocytosis. A brainstem metabolic disease becomes a limbic immune disease.
Key genes: ADRB2 (β2-AR), LRP1, MAPT, TREM2
Hippocampus — CA1
Braak I–II · The first self-sustaining lesion (the bridgehead)
Phase II Neuroimmune Interface
Under chronic dual-pressure stress the TGF-β/SMAD homeostatic microglial program collapses (SMAD7 rises; P2RY12/TMEM119/CX3CR1/SALL1 markers lost). Resident microglia enter DAM/LDAM/dystrophic post-homeostatic states pivoting through TREM2, where protective withdrawal and destructive attack become one act. PANTHOS neurons appear in CA1; the hippocampus consolidates the disease's first self-sustaining lesion.
Key genes: TREM2, APOE, TYROBP, P2RY12, TMEM119, SMAD7, NLRP3, SPP1
Basolateral Amygdala
Braak II–III · Perineuronal-net-dense · neuropsychiatric manifestations
Phase III Bridge 2
One of the three densest sites of perineuronal-net-ensheathed parvalbumin interneurons (with cortical layers III–IV and hippocampal CA1). As the Proteolytic Turn digests the aggrecan–brevican sheath, PV disinhibition here maps onto the agitation, anxiety and psychiatric symptoms of moderate disease.
Key genes: ACAN (aggrecan), BCAN (brevican), TNR, MMP9, GRIN2B
Neocortex Layers III–IV — PV Interneurons & PNN
Braak III–IV → V–VI · Where the disease finally becomes visible (Age 70+)
Phase III Metabolic-Homeostatic Axis
The densest perineuronal nets sheathe the fast-spiking parvalbumin interneurons of cortical layers III–IV. Three converging arms — NLRP3→IL-1β→MMP-9/ADAMTS proteolysis, ferroptotic-iron Fenton chemistry, and C1q→C4d complement stripping — digest the net. The PV cell loses its coat and inhibitory competence: disinhibition, excitatory–inhibitory collapse and gamma-rhythm degradation ensue, with endosomal/retromer failure recurring in cortical neurons.
Key genes: ACAN, BCAN, TNR, MMP9, ADAMTS4, GPX4, KCNC1 (Kv3), VPS35, GRIN2A
Net-Bearing PV Neurons (Resilience)
Resilience node · Pathology present but not dementing
Convergence
Resilience is the joint preservation of all three layers — homeostatic microglia, an intact matrix, and competent PV synapses; no single layer suffices. Per de Vries, perineuronal-net density predicts cognition independent of amyloid/tau, and neurons that keep their nets carry low tau. The MCI→dementia transition behaves as a threshold (a self-sustaining loop crossing), not a slope — which is why PV silencing may be recoverable rather than simply death.
Key genes: ACAN, RELN, TREM2 (state, not burden), PLCG2 (P522R protective)

Convergence & Shared Substrates

What dissolves the old separate theories into one system: the shared elements that span multiple phases and bridges. Each row is a substrate, signal or pivot; the columns are the phases, bridges and buckets it touches. The longer the bar, the more of the architecture it unifies.

Shared Substrate / Signal / Pivot PⅠ BⅠ PⅡ BⅡ PⅢ Grnd Conv Syn Architecture Spanned
PV interneuron / PNN (shared substrate) ——●●●—●●
5 nodes · the substrate
TGF-β/SMAD signal (shared upstream signal) —●●●——●●
5 nodes · the signal
TREM2 (shared molecular pivot) ——●●—●●●
5 nodes · the pivot
APOE4 hub ●—●—●●●●
6 nodes · feeds all
Mitochondrial / bioenergetic threshold ●—●—●●●●
6 nodes
NLRP3 inflammasome ●—●●——●—
4 nodes
Complement (C1q/C3/C4d) ——●●●—●—
4 nodes
Autophagy-lysosomal / v-ATPase clearance ●———●—●—
3 nodes
Iron / ferroptosis (oxidative) ——●●●———
3 nodes
Tau (LC seeding → cortex) ●●●—●——●
5 nodes
Reelin–ApoER2 brake on tau ———●●—●—
3 nodes
MMP-9 / ADAMTS proteolysis ———●●———
2 nodes

The Unifying Pivots

One shared substrate, one shared upstream signal, one shared molecular pivot — plus the two bridge mechanisms and the hub. These are the joints on which the whole architecture turns. Click to expand.

PV interneuron ↔ Perineuronal net
The shared substrate — one lattice, three offices
9.8
The perisomatic zone of the parvalbumin interneuron and its aggrecan–brevican net is the single substrate on which the disease converges: it shields the neuron, stages the reelin signal, and gates tau uptake. Per de Vries, net density predicts cognition independent of amyloid/tau — the strongest quantitative support for the PNN as the high-value final common bottleneck (recast from "the substrate" to a bottleneck, not the sole terminal substrate). Shared molecules: ACAN, BCAN, TNR, HAPLN1, MMP9, ADAMTS4, RELN, LRP8.
TGF-β/SMAD ↔ TREM2
The shared signal meets the shared pivot — the Metabolic-Homeostatic Axis
9.5
TGF-β/SMAD is the shared upstream signal that maintains microglial homeostatic identity (and, latent in the matrix, restrains complement pruning); TREM2 is the shared molecular pivot (DAP12→SYK→PI3K→AKT→mTOR) gating microglial metabolic and phagocytic fitness. The richest convergence node resolves the "attack vs failure" dichotomy into one upstream event: collapse of the metabolically-maintained homeostatic state. Named upstream levers: ketogenic therapy, GLP-1 agonism, metformin, pioglitazone. Shared molecules: TGFB, SMAD7, TREM2, TYROBP (DAP12), SYK, MTOR, TFEB.
Locus Coeruleus ↔ Microglia
Bridge 1 — the same axons that regulate the forebrain become the axons that poison it
9.2
The single ascending LC projection carries two cargoes to the hippocampus along one substrate: withdrawal of the noradrenergic brake (β2-AR→cAMP→PKA→NF-κB inhibition lost, raising the microglial inflammatory set-point) and release- and seeding-competent pretangle tau internalized via LRP1 and HSPG endocytosis. Their simultaneous arrival converts a brainstem metabolic disease into a limbic immune disease. Shared molecules: DBH, SLC6A2, ADRB2, MAPT, LRP1, HSPG.
NLRP3 ↔ MMP-9
Bridge 2 — the proteolytic switch of the Proteolytic Turn
8.8
Lipid-laden microglia use lipid droplets to assemble NLRP3→caspase-1→mature IL-1β→(NF-κB/AP-1) transcription of MMP-9, MMP-3 and ADAMTS-4/5 — a cytokine-secreting cell becomes a matrix-digesting one. This is one of three converging arms on the net (with ferroptotic-iron Fenton chemistry and C1q→C4d complement stripping); the target is a chemically heterogeneous structure, so it takes three. Therapeutic knife-edge: MMP-9 also serves LTP, neurogenesis and the BBB, so selectivity and pulsed dosing matter. Shared molecules: NLRP3, PYCARD, CASP1, IL1B, MMP9, ADAMTS4, ADAMTS5, C1QA.
APOE4 ↔ TOMM40
The hub meets the mitochondrial import gate
8.5
The APOE4 hub feeds nearly every convergence node and meets the TOMM40 import gate at the organelle: the TOMM40 poly-T locus (rs10524523) sits in linkage disequilibrium with APOE4, and Aβ obstructing the TOM40 channel blocks import-based mitochondrial repair just as PARP-1 hyperactivation and NAD⁺ depletion cripple mitophagy. Groundwork risk (APOE ε4 heterozygote OR ≈ 3–4, homozygote OR ≈ 12–15) that also acts mechanistically at Phase I ignition. Shared molecules: APOE, TOMM40, TOM40, TIMM23, SORL1, VDAC1.