Detection and classification of neurotoxins using a novel short-term plasticity quantification method

Area of Science:

• Neuroscience
• Biosensor Technology
• Computational Biology

Background:

• Screening neuroactive compounds is crucial for drug discovery.
• Existing methods for assessing neural function are often time-consuming.
• Short-term plasticity (STP) in hippocampal slices offers a sensitive measure of neural response.

Purpose of the Study:

• To develop and validate a tissue-based biosensor for rapid screening of neuroactive chemical compounds.
• To establish a novel quantification method for STP using Volterra modeling.
• To utilize machine learning for classifying compounds based on their effects on neural plasticity.

Main Methods:

• Utilized acute hippocampal slices as a biological substrate for the biosensor.
• Employed random electrical impulse sequences as input and population spike amplitudes as output.
• Quantified STP using first and second-order kernels derived from a Volterra modeling approach.
• Decomposed the second-order kernel into Laguerre functions for feature extraction.
• Inputted kernel features into a single-layer artificial neural network (ANN) for classification.

Main Results:

• The biosensor demonstrated distinct feature profiles for various tested chemical compounds, including picrotoxin, valproate, and carbachol.
• The ANN successfully classified each compound into its respective pharmacological class.
• The Volterra modeling approach proved more specific and time-efficient than conventional methods.

Conclusions:

• The developed tissue-based biosensor effectively screens compounds affecting the nervous system.
• The novel STP quantification and ANN classification provide a robust platform for neuropharmacological assessment.
• This approach offers a rapid and specific method for identifying drug effects on neural function.