Machine Learning Establishes Single-Cell Calcium Dynamics as an Early Indicator of Antibiotic Response

Area of Science:

• Microbiology
• Cell Biology
• Biophysics

Background:

• Antibiotic treatment triggers physiological changes before bacterial cell death.
• Disruption of calcium (Ca2+) homeostasis is an early event in antibiotic response.
• Developing rapid markers for antibiotic resistance is crucial for effective treatment.

Purpose of the Study:

• To investigate early disruptions in Ca2+ homeostasis as a marker for antibiotic response.
• To utilize machine learning to quantify temporal information in single-cell Ca2+ dynamics.
• To determine if Ca2+ dynamics can distinguish between antibiotic-sensitive and resistant bacteria.

Main Methods:

• Employing a machine learning framework to analyze single-cell Ca2+ dynamics.
• Utilizing live, single-cell Ca2+ imaging.
• Profiling antibiotic responses including kanamycin and Polymyxin B.

Main Results:

• Ca2+ dynamics differentiate kanamycin-sensitive and resistant cells before observable changes in growth or protein stability.
• Aberrant Ca2+ dynamics are dose- and time-dependent, reflecting pharmacokinetics and pharmacodynamics.
• Ca2+ dynamics precede membrane permeabilization (propidium iodide entry) in Polymyxin B-treated cells.
• Modulating membrane voltage and external Ca2+ affects the timing of membrane destabilization.

Conclusions:

• Single-cell Ca2+ imaging coupled with machine learning can identify antibiotic-resistant bacteria.
• Ca2+ homeostasis disruption is a sensitive early indicator of antibiotic action.
• Calcium plays a significant role in antibiotic-induced membrane destabilization.