Parallel, Continuous Monitoring and Quantification of Programmed Cell Death in Plant Tissue.

Area of Science:

• Plant pathology
• Molecular biology
• Biophysics

Background:

• Accurate quantification of hypersensitive response (HR) programmed cell death is crucial for understanding plant immunity.
• Traditional methods for assessing HR are time-consuming and lack sensitivity.
• Developing rapid and sensitive phenotyping tools is essential for plant breeding and disease resistance research.

Purpose of the Study:

• To demonstrate a novel phenotyping platform, PASTEL, for rapid, continuous-time, and quantitative assessment of HR.
• To validate the platform's sensitivity in detecting microscopic levels of cell death.
• To develop machine learning models for classifying HR using PASTEL data.

Main Methods:

• Developed the Parallel Automated Spectroscopy Tool for Electrolyte Leakage (PASTEL) platform for continuous-time impedance data acquisition.
• Induced HR in Nicotiana benthamiana using the AVRblb2 effector protein and Rpi-blb2 resistance protein.
• Applied supervised machine learning models to classify HR based on frequency domain impedance data.

Main Results:

• PASTEL detected cell death at microscopic levels, even when leaves appeared healthy to the naked eye.
• Machine learning models achieved 84.1% accuracy (F1 score = 0.75) at 1 hour and 87.8% accuracy (F1 score = 0.81) at 22 hours for HR classification.
• The platform significantly improved temporal resolution and sensitivity compared to traditional HR assays.

Conclusions:

• PASTEL provides a highly sensitive and rapid method for quantifying plant programmed cell death.
• The integration of PASTEL with machine learning enables high-throughput phenotyping of plant disease resistance.
• This approach can reduce the time required for plant phenotyping from days to weeks to mere hours.