Multiscale Assay of Unlabeled Neurite Dynamics Using Phase Imaging with Computational Specificity

Area of Science:

• Neuroscience
• Cell Biology
• Biophysics

Background:

• Primary neuronal cultures are crucial for studying neuronal function, but their complex morphology poses challenges for analysis.
• Existing methods like fluorescence microscopy have limitations in quantifying dynamic processes and can cause phototoxicity.
• There is a need for advanced imaging techniques to analyze neuronal morphology and dynamics in a label-free, high-throughput manner.

Purpose of the Study:

• To develop a label-free live-cell imaging method for neuronal cultures with antibody-staining specificity.
• To enable quantitative analysis of dynamic activities like intracellular transport and growth in neurons.
• To provide a high-throughput strategy for analyzing neural network arborization without photobleaching or phototoxicity.

Main Methods:

• Quantitative phase imaging combined with deep convolutional neural networks to estimate fluorescence signals from unlabeled live cells.
• Generation of synthetic fluorescence images and semantic segmentation maps for annotating subcellular compartments.
• Application of the method to time-lapse imaging of hippocampal neurons to study development and transport dynamics.

Main Results:

• The method successfully generates label-free, high-specificity images of live neuronal cultures.
• It allows for dynamic, quantitative analysis of neural network arborization and intracellular transport.
• The study highlights relationships between cellular dry mass production and nuclear/neurite transport activity.

Conclusions:

• This computationally inferred fluorescence imaging approach offers a powerful, non-invasive tool for neuroscience research.
• It overcomes limitations of traditional fluorescence microscopy, enabling high-throughput, dynamic studies of neuronal development and function.
• The method facilitates a deeper understanding of synaptic plasticity and neurodegenerative processes.