Time-resolved functional genomics using deep learning reveals global hierarchical control of autophagy

Area of Science:

• Cell Biology
• Systems Biology
• Genetics

Background:

• Autophagy is crucial for cellular homeostasis, particularly in response to nutrient availability.
• While autophagy's activation is well-studied, its inactivation mechanisms upon nutrient restoration are less understood.
• Dynamic regulation of autophagy is essential for adapting to changing cellular environments.

Purpose of the Study:

• To map the genetic network controlling autophagy activation and inactivation during nutrient fluctuations.
• To understand the systems-level dynamics of autophagy regulation.
• To identify novel regulatory pathways involved in autophagy inactivation.

Main Methods:

• Utilized time-resolved high-content imaging and deep learning for quantitative analysis.
• Developed a dataset (AutoDRY) categorizing 5,919 mutants based on nutrient response kinetics.
• Integrated imaging data with functional genomics and genetic network analysis.

Main Results:

• Uncovered hierarchical and multilayered control mechanisms governing autophagy.
• Identified multiple novel regulatory pathways contributing to autophagy dynamics.
• Revealed the retrograde pathway as a critical modulator of autophagy inactivation by tuning core gene expression.

Conclusions:

• The study provides a systems-level understanding of autophagy regulation, particularly its inactivation.
• The AutoDRY dataset serves as a valuable resource for future autophagy research.
• The retrograde pathway is identified as a key player in timely autophagy shutdown.