Label free high throughput screening for apoptosis inducing chemicals using time-lapse microscopy signal processing

Area of Science:

• Cell biology
• Biophysics
• Computational biology

Background:

• Label-free time-lapse microscopy enables studying dynamic cellular events.
• Automated image analysis is crucial for large-scale spatiotemporal quantification.
• Efficient algorithms for detecting apoptosis in microscopy are lacking.

Purpose of the Study:

• To develop a fast image analysis algorithm for detecting chemically induced apoptosis in phase-contrast microscopy images.
• To enable large-scale, cost-effective screening of compounds for apoptosis-inducing activity.

Main Methods:

• Application of a 2D linear matched filter optimized for apoptotic cell morphology.
• Validation of computational apoptosis detection against fluorescence-based caspase activity assays.
• Screening of 2,866 drug-like compounds using HCT116 human colorectal carcinoma cells.

Main Results:

• Successfully detected chemically induced apoptosis using the developed image analysis method.
• Validated computational detections against established fluorescence assays.
• Identified two novel compounds that induce apoptosis, in addition to known inducers.

Conclusions:

• A fast, label-free image analysis approach can effectively detect apoptosis in cell populations.
• This method facilitates high-throughput screening for apoptosis-inducing compounds.
• The study identified new potential therapeutic agents for inducing apoptosis.