Deep Learning and Single-Molecule Localization Microscopy Reveal Nanoscopic Dynamics of DNA Entanglement Loci

Area of Science:

• Nanotechnology
• Polymer Physics
• Biophysics

Background:

• Nanoscale molecular dynamics are difficult to study due to imaging limitations.
• Single-Molecule Localization Microscopy (SMLM) offers potential for high-resolution imaging.
• Deep learning integration can enhance SMLM capabilities for dynamic studies.

Purpose of the Study:

• To investigate fast timescale entangled polymer dynamics at the single-molecule level.
• To develop a deep learning approach for analyzing SMLM data of polymer dynamics.
• To understand the impact of entanglement on local molecular environments and dye kinetics.

Main Methods:

• Utilized Lambda DNA as a model system for polymer entanglement.
• Modeled DNA entanglement using the self-avoiding wormlike chain model.
• Trained a deep learning algorithm on simulated SMLM data to predict polymer chain contours.
• Analyzed localization data and dye photoswitching kinetics.

Main Results:

• Demonstrated heterogeneous distribution of localizations along DNA contours.
• Identified chain entanglement as a source of local diffusion barriers affecting dye photoswitching.
• Observed stochastic and subdiffusive migration of entanglement loci through segment tracking.
• Showcased direct visualization of nanoscale polymer dynamics and local environments.

Conclusions:

• Deep learning-SMLM provides unprecedented visualization of nanoscale polymer dynamics.
• Fluorophore switching kinetics can serve as a probe for nanoscopic local environments.
• The study advances understanding of polymer entanglement and its influence on molecular interactions.