Deep learning-enabled analysis reveals distinct neuronal phenotypes induced by aging and cold-shock.

Area of Science:

• Neuroscience
• Cell Biology
• Computational Biology

Background:

• Quantitative analysis is vital for understanding biological systems.
• Traditional image analysis of complex phenotypes, like PVD neuron degeneration in C. elegans, is often qualitative and labor-intensive.
• Aging and environmental stressors induce neurodegenerative changes, such as dendritic protrusions in the PVD neuron.

Purpose of the Study:

• To apply deep learning for quantitative image-based analysis of neurodegeneration in the C. elegans PVD neuron.
• To identify and characterize neurodegenerative patterns induced by aging and cold shock.
• To enable unbiased tracking of subtle morphological changes in neuronal degeneration.

Main Methods:

• Utilized a convolutional neural network algorithm (Mask R-CNN) for image segmentation.
• Applied deep learning to identify neurodegenerative subcellular protrusions in the PVD neuron.
• Developed a multiparametric phenotypic profile to capture morphological changes.

Main Results:

• Successfully identified neurodegenerative protrusions induced by cold shock and aging.
• Demonstrated that cold-shock-induced neurodegeneration is reversible and temperature-dependent.
• Revealed distinct neuronal beading patterns caused by aging versus cold shock.

Conclusions:

• Deep learning enables quantitative and unbiased tracking of subtle morphological changes in PVD neurodegeneration.
• Aging and cold shock induce distinct neurodegenerative patterns, suggesting different underlying mechanisms.
• This approach can identify molecular components involved in neurodegeneration and assess stressor effects.