Identifying and tracking simulated synaptic inputs from neuronal firing: insights from in vitro experiments

Area of Science:

• Neuroscience
• Computational Neuroscience
• Systems Neuroscience

Background:

• Accurate description of synaptic interactions and their temporal changes is crucial for systems neuroscience.
• Intracellular electrophysiology is limited by simultaneous recording capabilities and in vivo technical difficulties.
• Large-scale extracellular recording and statistical inference offer potential for mapping neural connectivity.

Purpose of the Study:

• To validate methods for inferring functional connectivity from spike trains.
• To assess the relationship between inferred functional connectivity and simulated synaptic input.
• To determine the limitations of inferring synaptic input from neuronal firing.

Main Methods:

• Utilized in vitro current injection in layer 2/3 pyramidal neurons.
• Employed partially-defined input (single simulated input with noise) and fully-defined input (controlled synaptic weights and timing).
• Analyzed neuronal firing responses to artificial inputs to infer connectivity.

Main Results:

• Individual current-based synaptic inputs are detectable across various amplitudes and conditions.
• Detectability is influenced by input amplitude, output firing rate, and input type (excitatory vs. inhibitory).
• Modeling more presynaptic inputs improved accuracy in estimating connection strengths and speed of detection.

Conclusions:

• Inferred functional connectivity from spikes can be validated against controlled synaptic inputs.
• The study outlines the possibilities and limitations of inferring synaptic input from neuronal activity.
• These findings advance the understanding of large-scale neural network analysis using extracellular recordings.