Nonlinear dynamic modeling of synaptically driven single hippocampal neuron intracellular activity

Area of Science:

• Computational Neuroscience
• Systems Neuroscience
• Neuroscience

Background:

• Understanding the input-output dynamics of single neurons is crucial for deciphering neural computation.
• Existing models often struggle to capture both subthreshold and suprathreshold activities within a unified framework.
• Hippocampal CA1 pyramidal neurons play a key role in memory formation and cognitive functions.

Purpose of the Study:

• To develop a high-order nonlinear dynamic model for single hippocampal CA1 pyramidal neurons.
• To create a unified model for both postsynaptic potentials (PSPs) and action potentials.
• To ensure the model is generalizable to other neuron types and computationally efficient.

Main Methods:

• Developed a three-component model: feedforward kernels for PSPs, a spike generation threshold, and a feedback kernel for after-potentials.
• Used whole-cell patch-clamp recordings from CA1 pyramidal neurons stimulated with Poisson random impulse trains.
• Applied the model to predict PSP waveforms and action potential occurrences.

Main Results:

• The model successfully captured the nonlinear dynamics of intracellular activity.
• Achieved an average normalized mean square error of 14.4% for PSP prediction.
• Attained an average spike prediction error rate of 18.8%.

Conclusions:

• The developed model effectively represents the high-order nonlinear dynamics of single-neuron activity.
• The model's general biophysical parameters allow for broad application to spike-input/output neurons.
• This approach offers a computationally efficient tool for analyzing neural dynamics.