Automated profiling of individual cell-cell interactions from high-throughput time-lapse imaging microscopy in...

Area of Science:

• Cellular and Molecular Biology
• Immunology
• Biotechnology

Background:

• High-throughput time-lapse imaging generates complex data on cell-cell interactions.
• Automated analysis is crucial for profiling immune effector and tumor cell dynamics in adoptive immunotherapy.

Purpose of the Study:

• To develop and validate automated algorithms for single-cell resolution analysis of dynamic cell-cell interactions.
• To enhance the efficiency and accuracy of analyzing immune cell-tumor cell interactions in nanowell-based assays.

Main Methods:

• Co-incubation of fluorescently labeled immune effector cells (T cells, NK cells) and target cells in polydimethylsiloxane nanowells.
• Multi-channel time-lapse microscopy for imaging.
• Development of novel cell segmentation and tracking algorithms exploiting nanowell confinement.

Main Results:

• Achieved 98% yield for correctly analyzed nanowells (effector-target pairs), a significant improvement from 45% with existing algorithms.
• Demonstrated high accuracy: >99% for nanowell delineation, >95% for cell segmentation, and 90% for cell tracking.
• Revealed that NK cells distinguish live from dead targets by modulating conjugation duration and that cytotoxic cells exhibit higher motility.

Conclusions:

• The developed automated methods provide accurate and efficient profiling of cell-cell interactions from time-lapse imaging.
• These algorithms enable robust quantification of cell behavior, including location, morphology, movement, interactions, and death, without manual correction.
• The findings offer insights into immune cell behavior, such as target discrimination and motility patterns, relevant to immunotherapy research.