Single-compartment model of a pyramidal neuron, fitted to recordings with current and conductance injection

Anton V. Chizhov, Dmitry V. Amakhin, A. Erdem Sagtekin, Mathieu DesrochesBiol. Cybern 2023 · DOI

TL;DR

This paper introduces a new, simplified five-dimensional model of a pyramidal neuron that accurately reproduces key experimental firing characteristics, including firing rate, spike shape, and depolarization block, under varying current and conductance inputs. This model offers a more biophysically realistic and computationally efficient alternative to existing models for large-scale neural simulations.

What Problem Does This Paper Address?

Accurately modeling the behavior of individual neurons is crucial for understanding complex neural networks. However, existing models often struggle to simultaneously capture several essential characteristics of neuronal activity, such as the maximum firing rate, the conditions leading to a depolarization block (where a neuron stops firing), and the influence of synaptic conductance on firing. Many models either oversimplify these aspects or become computationally too expensive for large-scale simulations. This paper aims to bridge this gap by developing a single-compartment model that is both biophysically detailed enough to reproduce these critical features and simple enough for efficient use in network simulations.

Methods

The authors developed a novel five-dimensional dynamical system model for a pyramidal neuron. This model incorporates a single sodium channel with specific inactivation and activation properties, along with two types of potassium channels (fast repolarizing and slow shunting). A key innovation is the use of a dynamic threshold for sodium channel activation, which depends on slow inactivation. This approach allows the model to capture complex behaviors like the depolarization block and the subtle voltage dynamics between spikes. The model was fitted to experimental data obtained using the dynamic patch-clamp technique, which allows for precise control of both injected current and synaptic conductance.

Datasets

The study utilized experimental data from whole-cell patch-clamp recordings of pyramidal neurons in rat brain slices. These recordings were obtained under controlled conditions using the dynamic patch-clamp technique. The paper does not introduce a new benchmark dataset but rather uses experimental data to validate and refine their proposed model, comparing it against existing models like the Hodgkin-Huxley and Izhikevich models.

Key Insights and Ideas

  • Dynamic Threshold is Crucial: The dependence of the sodium channel activation threshold on slow inactivation is a critical factor in reproducing realistic neuronal firing patterns, including the depolarization block.
  • Shunting Effect Matters: The influence of synaptic conductance (shunting) is not a minor detail; it significantly impacts firing rate, spike amplitude, and the conditions for depolarization block, and models must account for this.
  • Simplicity Can Be Powerful: A relatively simple five-dimensional model, by carefully incorporating key ionic channel dynamics and their interactions, can achieve a level of realism comparable to or exceeding more complex models for specific phenomena.
  • Depolarization Block Mechanism: The paper proposes that the depolarization block arises from a "runaway" threshold, where the threshold potential increases beyond the membrane potential due to the dynamics of slow inactivation.

Key Results

The proposed model successfully reproduced several key experimental characteristics of pyramidal neurons:

  • Firing Rate: The model accurately captured the maximum firing rate (around 28 Hz) and the "tongue-shaped" domain of steady-state firing in the current-conductance plane.
  • Depolarization Block: The model predicted critical values for current and conductance that induce depolarization block, which were in good agreement with experimental findings.
  • Spike Dynamics: The model reproduced the shape of spikes, the voltage evolution between spikes, and the timing of voltage minima, which are often poorly captured by simpler models.
  • Conductance Overestimation: The study highlighted that the standard Hodgkin-Huxley model tends to overestimate the contribution of voltage-gated channels, leading to discrepancies with experimental data, a problem the proposed model mitigates.

Limitations and Future Work

The authors acknowledge that the proposed model, while successful, has some limitations:

  • Spike Width: The model's spike width is almost unchanged with increasing stimulating current, which contrasts with experimental observations where spikes tend to broaden.
  • Model Complexity: While simpler than some, the model still involves five differential equations, and further simplification might be explored for specific applications.

Future work could focus on:

  • Extending to Network Models: The model is well-suited for integration into mean-field and refractory density approaches for simulating larger neural populations.
  • Investigating Dynamic Thresholds Further: Exploring the implications of the dynamic threshold concept in more detail and its role in other neuronal phenomena.
  • Addressing Spike Broadening: Incorporating mechanisms that lead to spike broadening under specific conditions.

Summary updated

BibTeX
@article{chizhov2023single,
  author = {Anton V. Chizhov and Dmitry V. Amakhin and A. Erdem Sagtekin and Mathieu Desroches},
  title = {{Single-compartment model of a pyramidal neuron, fitted to recordings with current and conductance injection}},
  journal = {Biol. Cybern},
  year = {2023},
  doi = {10.1007/s00422-023-00976-7},
}
Figure