BRAINCELL modelling platform for stochastic nanoscale organisation and dynamic extracellular signalling among neurons and glia

Key results

1. SuggestedPopulating a reconstructed CA1 pyramidal neuron with stochastically generated dendritic spines matched to STED morphometry widened the range of presynaptic firing rates over which input was converted into output spiking.

   The same stochastic excitatory inputs (Pr = 0.36) were placed on dendritic shafts, on spines (100%), as a 50/50 mix, or on post-LTP spines with shortened necks. Including spines expanded the input-output dynamic range from 0-40 Hz to 5-90 Hz (mean ± SEM, n = 10 runs per condition). The Results give 500 synapses; the Methods give 316 spines with one synapse each. · Dendritic spine populations sharply expand the…; The role of dendritic spine populations and…; Stochastic modelling of dendritic spines and…; Figure legends

2. SuggestedFour Kir4.1-expressing 3D-reconstructed CA1 astrocytes placed among the apical dendrites lowered extracellular K+ around the neuron and reduced its firing, compared with astrocytes lacking Kir4.1.

   The neuron was driven by 1493 stochastic excitatory synapses (10 Hz, Pr = 0.5, 400 pS), with [K+]o raised from 2.5 to 5 mM to mimic network activity. Astrocytes carried Kir4.1 at 0.4 mS/cm² or none. At 10 s, [K+]o was lower in perineuronal territories and the firing rate fell by 15%. Grid-refinement tests used a < 5% relative-change criterion for convergence. · Astrocyte-neuron interaction via extracellular…; Astrocyte–neuron interaction via extracellular…; Figure legends

3. SuggestedEach myelin internode was modelled as a separate oligodendrocyte compartment expressing Kir4.1. In this model, removing myelin Kir4.1 let periaxonal K+ accumulate until high-frequency spike propagation failed.

   The axon had ten 100 µm internodes, 1 µm nodes and five radial K+ shells, with intra-myelin K+ diffusivity reduced about tenfold. With myelin Kir4.1 at 0.4 mS/cm², spikes at 200 Hz propagated consistently. Without it, extracellular K+ built up and propagation failed about 1.2 s after spiking began (about 3 s simulated; no longitudinal diffusion). · Modelling myelinated axons: evaluating the role…; Exploring spike propagation in a multi-cell…; Modelling K⁺ dynamics at myelinated axons during…; Figure legends

4. SuggestedWith mean release probability held constant, how it was distributed along the apical dendrites changed spiking output. The experimentally reported linear proximal-to-distal gradient gave the strongest signal transfer.

   Three distributions shared a mean Pr = 0.36: uniform, bell-shaped (σ ≈ 200 μm) and a linear gradient (Pr ≈ 0.2 proximally to ≈ 0.6 distally). The gradient gave the highest postsynaptic firing across input frequencies (mean ± SEM, n = 10 runs per condition). Effect sizes appear only in plots; the text gives no numbers. · Distribution of synaptic release probability…; The role of dendritic spine populations and…; Modelling release probability pattern; Figure legends

5. SuggestedA mesh-free model of GABA release, diffusion and uptake from a single CA3 basket cell was fitted to earlier iGABASnFR2 kinetics. The fit suggests that individual interneurons can act as additive sources of slow extracellular GABA waves.

   Fit target: time course of earlier recordings (rise 0.2-0.3 s, decay 0.9-1.5 s); those waves peaked an estimated 1-2 µM above baseline during synchronous discharges of multiple cells. Model: 6,000 sources within 0.5 µm of axonal sites, 2,000 activated per spike, 3,000 GABA molecules each, apparent D = 0.3 µm²/ms, best-fit uptake τ = 600 ms. Spatial spread is shown only as snapshots, no distance reported. · GABA release, diffusion and uptake; Introducing extracellular ion fluxes and…; Modelling release, diffusion and uptake of GABA; Figure legends