Hypothesis

The selective vulnerability of reelin+ stellate (grid) neurons in entorhinal cortex layer II stems from their lack of calbindin-D28k, which leads to chronic intracellular calcium elevation during high-frequency grid firing. This calcium excess activates the phosphatase calcineurin, which dephosphorylates NFAT transcription factors, promoting nuclear translocation and upregulation of GSK3β expression. Elevated GSK3β hyperphosphorylates tau at AD-relevant sites, facilitating its aggregation and seeding. Consequently, grid-cell activity itself fuels a feed-forward loop: more firing → more calcium → more calcineurin/NFAT/GSK3β signaling → more tau pathology → further silencing of grid firing and spatial memory decline.

Testable Predictions

1. Genetic rescue – Overexpressing calbindin-D28k specifically in Wfs1+ or RORB+ excitatory neurons of EC layer II in tau-transgenic mice will reduce phospho-tau levels (AT8, AT100) by >40% compared with controls, without altering overall neuronal activity measured by in vivo calcium imaging.
2. Pharmacological block – Acute administration of an L-type calcium channel antagonist (e.g., isradipine) during active navigation sessions will attenuate calcineurin activity (measured by phosphorylated NFATc1) and prevent tau seeding in the posterior-lateral EC subfield after 4 weeks of treatment.
3. NFAT dependency – Conditional knockout of NFATc1 in reelin+ stellate cells will block the calcium-to-GSK3β signaling cascade, resulting in preserved grid-cell firing patterns (spatial information content >0.8 bits/spike) and resistance to tau accumulation despite high amyloid-β exposure.
4. Regional specificity – The posterior-lateral EC, which shows the earliest tau accumulation, will exhibit the highest baseline calcium transients during grid-cell firing, as revealed by fiber-photometry of GCaMP6f in layer II versus more anterior-medial subfields.

Mechanistic Reasoning Beyond Existing Data

While calbindin absence is noted as a marker of vulnerability (3), the downstream calcium-dependent signaling cascade linking neuronal activity to tau phosphorylation has not been explicitly mapped in EC stellate cells. Calcineurin/NFAT is known to regulate GSK3β transcription in neurons (2), and GSK3β is a principal tau kinase in early AD (4). High firing of grid cells during path integration provides a natural, activity-dependent calcium influx that, in the absence of buffering, preferentially activates this pathway. This explains why tau pathology initiates in the most active subpopulation (grid cells) and why posterior-lateral EC, which encodes spatial navigation with the highest firing rates, shows the earliest lesions (5). Amyloid-β may exacerbate the process by impairing calcium extrusion, but the core driver is activity-linked calcium dysregulation.

Falsifiability

If calbindin overexpression fails to lower phospho-tau or if calcineurin inhibition does not rescue grid-cell firing despite reduced calcium, the hypothesis would be refuted. Conversely, demonstration that manipulating calcium buffering or calcineurin/NFAT/GSK3β signaling alters tau pathology independent of amyloid-β burden would support the proposed mechanism.

Experimental Outline

• Use AAV-mediated calbindin-D28k or NFATc1 Cre-dependent constructs in Wfs1-Cre;R26-LSL-GCaMP mice crossed with tau-P301S line.
• Perform in vivo two-photon imaging of calcium spikes during virtual navigation tasks.
• Quantify phospho-tau by immunohistochemistry and ELISA at 3 and 6 months.
• Assess spatial memory via Morris water maze and grid-cell stability via spike-sorting and spatial information analysis.
• Include amyloid-β-free aged wild-type controls to test Aβ-independent effects.

By directly linking the electrophysiological identity of grid cells to a calcium-dependent kinase cascade, this hypothesis provides a concrete, experimentally tractable framework for why EC layer II is the epicentre of tauopathy and how preserving calcium homeostasis might delay cognitive decline.