Abstract

Neurodegeneration is characterized by progressive neuronal loss in the context of chronic neuroinflammation. We argue that the primary pathological event is not the accumulation of protein aggregates — amyloid-β, tau, α-synuclein — but the failure of the microglial activation-resolution cycle to complete. Microglia activate in response to damage but fail to clear the trigger and fail to resolve. The cycle runs but does not finish.

The broken step is clearance. Microglia require TREM2 signaling to transition to a clearance-competent state. The R47H variant of TREM2 — the second strongest genetic risk factor for late-onset Alzheimer's — reduces TREM2 ligand affinity, impairing this transition. We propose correcting TREM2 R47H to wild-type via adenine base editing in brain microglia, delivered by intrathecal lipid nanoparticles.

---

1. Introduction

1.1 The neuroinflammation cycle

1. detect    — DAMP or aggregate detected, microglia activate
2. respond   — pro-inflammatory cytokines released (TNF-α, IL-1β, IL-6)
3. clear     — phagocytose aggregates, apoptotic cells, debris
               TREM2-dependent transition to clearance state
4. resolve   — anti-inflammatory signals (IL-10, TGF-β)
               return to surveillant phenotype

1.2 The broken step

In Alzheimer's, Parkinson's, and ALS, microglia accumulate around sites of protein aggregation and remain in a chronically activated state. They do not resolve. They continue producing inflammatory cytokines and reactive oxygen species, causing bystander neuronal death. The answer to why they do not complete step 3 is, in a significant fraction of cases, TREM2.

1.3 TREM2 and the clearance transition

TREM2 is a membrane receptor expressed exclusively on microglia in the brain. Its signaling drives the transition from the pro-inflammatory state to the Disease-Associated Microglia (DAM) phenotype — upregulating phagocytic machinery and enabling sustained engulfment of aggregates. Without TREM2 signaling, microglia remain in the inflammatory state and cannot complete the clearance step.

---

2. TREM2 R47H

The TREM2 R47H variant (rs75932628; CGT→CAT at codon 47) is the second strongest genetic risk factor for late-onset Alzheimer's disease after APOE4, conferring approximately 2–4x increased risk. It is also associated with increased risk for Parkinson's disease, frontotemporal dementia, and ALS.

R47H is a missense mutation in the extracellular Ig-like domain. The arginine-to-histidine substitution reduces TREM2's affinity for its key ligands — phosphatidylserine, APOE-lipid complexes, amyloid-β — by approximately 3-fold under physiological pH. TREM2 R47H microglia can still surveil; they cannot efficiently engage and clear pathological aggregates.

The genetic data provide unusual support for the parsimony framing. The fact that a single amino acid change at position 47 — reducing but not eliminating ligand affinity — is sufficient to produce a 2–4x increase in Alzheimer's risk demonstrates that the clearance step is operating near threshold. The minimum intervention is to restore wild-type TREM2 affinity past that threshold.

---

3. Proposed Intervention

3.1 The edit

TREM2 R47H is a G→A transition. On the antisense strand, this is a C→T transition — or for reversion, an adenine base editor (ABE) targeting the antisense strand restores A→G, reverting CAT back to CGT (His→Arg).

LNP payload (or AAV-PHP.eB):
  ABE8e mRNA
  sgRNA targeting TREM2 antisense strand at codon 47

Target: endogenous TREM2 locus, codon 47
Edit: A→G reversion (restores Arg at position 47)
Expression: native TREM2 promoter (unchanged)
Route: intrathecal injection

3.2 Delivery to microglia

Microglia are phagocytic and preferentially take up LNPs. Intrathecal LNP administration distributes through the CSF and is taken up by microglia. Recent work demonstrates >40% microglial editing efficiency with intrathecal LNP delivery in rodents. AAV-PHP.eB provides an alternative with long-term expression and microglia-selective promoters (CX3CR1, P2RY12) for cell-type specificity.

3.3 Predicted mechanism

1. detect    — aggregate detected, microglia activate (unchanged)
2. respond   — pro-inflammatory state (unchanged, transient)
3. clear     — TREM2 (restored) binds aggregate ligands at full affinity
               DAM transition proceeds
               phagocytosis of amyloid, tau, debris      ← RESTORED
4. resolve   — trigger cleared; anti-inflammatory signals
               return to surveillant state

---

4. Discussion

4.1 Why not amyloid clearance therapies

Anti-amyloid antibodies work by recruiting the peripheral immune system to clear amyloid from outside — a workaround for the broken step, not a repair. They help without fixing the underlying cycle failure. Restoring TREM2 addresses the broken step from inside the cycle, using the brain's own clearance machinery.

4.2 Generalization across neurodegenerative diseases

TREM2-dependent microglial clearance is involved in Parkinson's (α-synuclein), ALS (TDP-43), and frontotemporal dementia (tau, TDP-43). The broken step is the same; the trigger differs. Restoring the broken step may be broadly applicable.

4.3 Testable predictions

1. Edited microglia will show enhanced DAM gene expression and increased phagocytic capacity against amyloid-β fibrils in vitro.

2. In TREM2 R47H × APP/PS1 mice, intrathecal LNP delivery will produce >30% TREM2 reversion in microglia within 4 weeks, with associated reduction in diffuse plaque burden.

3. Edited mice will show improved spatial memory and preserved neuronal density in hippocampus and entorhinal cortex relative to unedited controls with comparable amyloid burden.

---

5. Conclusion

Neurodegeneration is a cycle stuck at step 3. Microglia activate, respond, but cannot clear — and without clearance, cannot resolve. The chronic inflammation that kills neurons is the steady state of a cycle that never finishes.

One base. One reversion. One step restored. The brain's resolution mechanism does the rest.