Hypothesis

Loss of calbindin‑D28k (CB) in basal forebrain cholinergic neurons (BFCN) triggers a self‑reinforcing mitochondrial ROS‑HDAC2 feedback loop that epigenetically suppresses CB transcription, accelerating neurodegeneration in Alzheimer’s disease.

Mechanistic Rationale

Aging‑dependent CB depletion removes a high‑affinity cytosolic Ca²⁺ buffer, causing prolonged microdomains of elevated Ca²⁺ near mitochondria ([3]). This excess matrix Ca²⁺ overloads the mitochondrial calcium uniporter (MCU) complex, leading to oxidative phosphorylation inefficiency and increased superoxide production ([4]). Mitochondrial ROS activate histone deacetylase 2 (HDAC2) via oxidation of its catalytic cysteine, enhancing its deacetylase activity on histone H3 lysine 9 at the Calb1 promoter ([5]). HDAC2‑mediated chromatin compaction reduces CB mRNA synthesis, further diminishing cytosolic buffering capacity and completing a vicious cycle. Notably, HDAC2 activity is already elevated in AD brains and correlates with tangle pathology, providing a mechanistic bridge between calcium dysregulation and epigenetic silencing.

Testable Predictions

1. In aged BFCN, pharmacological inhibition of MCU or ROS scavenging will lower mitochondrial Ca²⁺ overload, decrease HDAC2 activity, and increase CB mRNA and protein levels.
2. Genetic knockdown of HDAC2 in BFCN of aged APP/PS1 mice will rescue CB expression despite ongoing Ca²⁺ dysregulation.
3. Restoring CB expression via AAV‑mediated Calb1 overexpression will blunt mitochondrial ROS production and HDAC2 activation, thereby protecting BFCN from neurofibrillary tangle formation.
4. Conversely, forced HDAC2 overexpression in young BFCN will recapitulate age‑like CB loss and sensitize neurons to Aβ‑induced Ca²⁺ toxicity.

Experimental Design

• Animal models: Young (3 mo) and aged (18‑20 mo) wild‑type mice; aged APP/PS1 transgenic mice; AAV vectors for Calb1 overexpression, shRNA‑HDAC2, or HDAC2‑OE targeted to the basal forebrain using ChAT‑Cre drivers.
• Readouts:
  • Cytosolic and mitochondrial Ca²⁺ measured with GCaMP6f and mito‑GCaMP6f via two‑photon imaging in acute slices.
  • Mitochondrial superoxide assessed with MitoSOX Red.
  • HDAC2 activity quantified by fluorometric deacetylase assay and chromatin immunoprecipitation for H3K9ac at the Calb1 promoter.
  • CB mRNA (qPCR) and protein (immunohistochemistry, Western blot).
  • Neurofibrillary tangle burden (AT8 immunostaining) and neuronal survival (ChAT+ cell counts) in the nucleus basalis of Meynert.
• Pharmacological arms: MCU inhibitor Ru360 (10 µM), ROS scavenger MitoTEMPO (500 nM), and HDAC2 inhibitor (e.g., SB939) administered via osmotic minipumps for 4 weeks.
• Statistical plan: Two‑way ANOVA (age × treatment) with post‑hoc Tukey; n ≥ 6 per group to achieve 80 % power for detecting a 25 % change in CB levels (α = 0.05).

Potential Outcomes and Falsifiability

If the hypothesis is correct, we expect:

• MCU inhibition or ROS scavenging to significantly raise CB levels and reduce HDAC2 activity in aged BFCN (p < 0.01).
• HDAC2 knockdown to elevate CB expression even in the presence of high mitochondrial Ca²⁺, uncoupling the feed‑forward loop.
• CB overexpression to lower mitochondrial ROS, HDAC2 activity, and tangle formation, rescuing cholinergic neuron survival.
• HDAC2 overexpression in young neurons to diminish CB and increase vulnerability to Aβ‑induced Ca²⁺ rises, mimicking the aged phenotype.

Failure to observe any of these directional changes—particularly if manipulating mitochondrial Ca²⁺ or ROS does not affect HDAC2 activity or CB expression, or if HDAC2 modulation fails to alter CB levels—would falsify the proposed mechanism and suggest that CB loss operates independently of mitochondrial ROS‑HDAC2 signaling. Such results would redirect focus toward alternative pathways (e.g., transcriptional repressors like REST or microRNA‑mediated regulation) driving CB decline in aging BFCN.