Intrinsic threshold plasticity: cholinergic activation and role in the neuronal recognition of incomplete input...

Area of Science:

• Neuroscience
• Computational Neuroscience
• Machine Learning

Background:

• Neurons exhibit activity-dependent changes in membrane excitability, known as intrinsic plasticity (IP).
• IP modifies the action potential threshold, affecting neuronal output without altering synaptic strength.
• Non-homeostatic IP is hypothesized to play a role in learning by adjusting the action potential threshold.

Purpose of the Study:

• Investigate the computational consequences of intrinsic plasticity on neural network performance.
• Assess how changes in action potential threshold affect pattern discrimination and recognition.
• Analyze physiological data to understand the mechanisms and ranges of intrinsic plasticity in cortical neurons.

Main Methods:

• Utilized an artificial neural network model to simulate the effects of intrinsic plasticity on pattern recognition tasks.
• Analyzed published whole-cell patch-clamp recording data from L2/3 pyramidal neurons in the primary visual cortex (V1) and somatosensory cortex (S1).
• Examined the impact of altered spike thresholds on the network's ability to discriminate unknown and recognize known patterns.

Main Results:

• Negative shifts in threshold potentials decreased the network's ability to discriminate unknown input patterns.
• Conversely, reduced threshold potentials improved the network's performance in recognizing known, but incompletely presented, patterns.
• Physiological data from S1 neurons showed a significant resting to threshold potential difference and a plasticity range influenced by cholinergic and electrical stimulation.

Conclusions:

• Intrinsic plasticity, specifically threshold reduction, shifts neural coding from faithful representation to interpretative assignment.
• This plasticity mechanism offers a trade-off between precise input discrimination and robust recognition of learned information.
• Findings suggest that intrinsic plasticity plays a crucial role in adapting neural computations for learning and memory.