Production of Superoxide in Bacteria Is Stress- and Cell State-Dependent: A Gating-Optimized Flow Cytometry Method that Minimizes ROS Measurement Artifacts with Fluorescent Dyes

Megan E. McBee, Yok H. Chionh, Mariam L. Sharaf, Peiying Ho, Maggie W. L. Cai, Peter C. Dedon

The role of reactive oxygen species (ROS) in microbial metabolism and stress response has emerged as a major theme in microbiology and infectious disease. Reactive fluorescent dyes have the potential to advance the study of ROS in the complex intracellular environment, especially for high-content and high-throughput analyses. However, current dye-based approaches to measuring intracellular ROS have the potential for significant artifacts. Here, we describe a robust platform for flow cytometric quantification of ROS in bacteria using fluorescent dyes, with ROS measurements in 10s-of-1000s of individual cells under a variety of conditions. False positives and variability among sample types (e.g., bacterial species, stress conditions) are reduced with a flexible four-step gating scheme that accounts for side- and forward-scattered light (morphological changes), background fluorescence, DNA content, and dye uptake to identify cells producing ROS. Using CellROX Green dye with Escherichia coli, Mycobacterium smegmatis, and Mycobacterium bovis BCG as diverse model bacteria, we show that (1) the generation of a quantifiable CellROX Green signal for superoxide, but not hydrogen peroxide-induced hydroxyl radicals, validates this dye as a superoxide detector; (2) the level of dye-detectable superoxide does not correlate with cytotoxicity or antibiotic sensitivity; (3) the non-replicating, antibiotic tolerant state of nutrient-deprived mycobacteria is associated with high levels of superoxide; and (4) antibiotic-induced production of superoxide is idiosyncratic with regard to both the species and the physiological state of the bacteria. We also show that the gating method is applicable to other fluorescent indicator dyes, such as the 5-carboxyfluorescein diacetate acetoxymethyl ester and 5-cyano-2,3-ditolyl tetrazolium chloride for cellular esterase and reductive respiratory activities, respectively. These results demonstrate that properly controlled flow cytometry coupled with fluorescent probes provides precise and accurate quantitative analysis of ROS generation and metabolic changes in stressed bacteria.