## My Perspective: Neuronal Vulnerability as a Circuit Problem
The question presupposes a single root cause, but from where I sit—building taxonomies of cell types and mapping brain-wide connectivity—the evidence increasingly points to Alzheimer's as a **selective vulnerability of specific neuronal populations propagating through circuits**.
Here's why this framing matters:
**1. Not all neurons die equally.** Cell type atlases reveal that certain excitatory pyramidal neurons in layer II/III of the entorhinal cortex—the first to degenerate—are molecularly distinct. They may share features (connectivity patterns, metabolic profiles, ion channel expression) that make them susceptible *before* amyloid or tau pathology becomes widespread.
**2. The spreading follows connectivity.** Tau pathology advances along anatomically connected pathways, not randomly. This suggests the disease exploits trans-synaptic mechanisms. My work on the mesoscale connectome showed that specific circuit architectures are more vulnerable to disruption.
**3. Amyloid may be upstream but not sufficient.** The amyloid hypothesis dominated for decades, but anti-amyloid therapies have shown modest effects. I suspect amyloid deposition triggers a cascade in *predisposed* cell types, and the root cause is actually that molecular vulnerability—what makes certain neurons unable to handle proteostatic stress, mitochondrial dysfunction, or inflammatory challenge.
**4. Glia set the context.** Microglia and astrocytes aren't passive. My colleagues' work on single-cell atlases shows microglial states shift dramatically in Alzheimer's. The "root cause" may ultimately be a failure of glial-neuronal crosstalk that every brain possesses, but some cell type combinations handle poorly.
In short: the root cause isn't amyloid *per se*—it's **why particular cell types in particular circuits lose resilience first**. That's a question that only systematic, brain-wide cell type mapping can answer.