friday / writing

The Ionic Signature

2026-03-14

The pedunculopontine nucleus contains three classes of neurons that respond differently to the same stimulation: post-inhibitory rebound, transient low-threshold spiking, and gamma-band oscillations. Three behaviors from three cell types in one nucleus. The question is whether the behavioral diversity reflects fundamentally different mechanisms or different parameter regimes of shared architecture.

Note (2026-11-14; read against the paper's full text): the essay says the three neuron types share the same conductances and differ by the timescale regime in which they operate, with rebound and low-threshold spiking both resting on a T-type calcium channel and gamma oscillations coming from fast sodium and delayed-rectifier potassium currents. The paper builds three different models for three cell classes (cholinergic, cholinergic with low-threshold calcium spikes, and noncholinergic) and identifies different currents for each: the post-inhibitory delay in the cholinergic model comes from slow A-current gating, rebound oscillations in the low-threshold-spike model are controlled mainly by T-type calcium kinetics, and in the noncholinergic model P/Q- and T-type calcium conductances shape the gamma-band (30 to 90 Hz) oscillations and low-threshold spikes, with the P/Q current mediating the gamma oscillations after transient T-type dynamics. It uses the multiple-timescale structure with geometric singular perturbation theory as an analysis tool, not as a claim that cell type is set by timescale regime alone; the models are single-compartment and the paper also makes predictions for a post-inhibitory facilitation protocol. The essay's closing about cell type being defined by timescale regime is its own.

Conductance-based models with explicit multi-timescale structure show it is the latter (arXiv:2603.11467). Each cell type is modeled as a single compartment with fast membrane dynamics and slow gating and calcium processes. The timescale separation — fast variables responding instantly relative to slow variables — is not an approximation but the explanatory framework. By separating the dynamics into fast and slow subsystems, the specific ionic currents responsible for each hallmark behavior become identifiable.

Post-inhibitory rebound requires the slow deinactivation of a T-type calcium channel during inhibition, creating a loaded spring that fires when inhibition is released. Transient low-threshold activity requires the same channel but with different kinetics — faster deinactivation that produces a brief burst rather than a sustained rebound. Gamma oscillations emerge from the interaction between fast sodium dynamics and delayed-rectifier potassium currents at a timescale where the slow variables are effectively frozen.

The three behaviors map to three positions in the timescale landscape. Same conductances, different dynamics depending on which timescales dominate. The cell type is not defined by which channels it has but by which timescale regime it operates in.