Variability of Burkholderia pseudomallei strain sensitivities to chlorine disinfection

Abstract

Burkholderia pseudomallei is a select agent and the causative agent of melioidosis. Variations in previously reported chlorine and monochloramine concentration time (Ct) values for disinfection of this organism make decisions regarding the appropriate levels of chlorine in water treatment systems difficult. This study identified the variation in Ct values for 2-, 3-, and 4-log(10) reductions of eight environmental and clinical isolates of B. pseudomallei in phosphate-buffered water. The greatest calculated Ct values for a 4-log(10) inactivation were 7.8 mg.min/liter for free available chlorine (FAC) at pH 8 and 5 degrees C and 550 mg.min/liter for monochloramine at pH 8 and 5 degrees C. Ionic strength of test solutions, culture hold times in water, and cell washing were ruled out as sources of the differences in prior observations. Tolerance to FAC was correlated with the relative amount of extracellular material produced by each isolate. Solid-phase cytometry analysis using an esterase-cleaved fluorochrome assay detected a 2-log(10)-higher level of organisms based upon metabolic activity than did culture, which in some cases increased Ct values by fivefold. Despite strain-to-strain variations in Ct values of 17-fold for FAC and 2.5-fold for monochloramine, standard FAC disinfection practices utilized in the United States should disinfect planktonic populations of these B. pseudomallei strains by 4 orders of magnitude in less than 10 min at the tested temperatures and pH levels.

FIG. 1.

B. pseudomallei isolates exhibit a variety of colony morphologies. Representative isolates for each colony morphology and origin are shown. A, ATCC 11668 (crenellated, clinical); B, CA 652 (mucoid, clinical); C, SC 763 (nonmucoid, clinical); D, AU 631 (nonmucoid, environmental).

FIG. 2.

Groupings of Ct values for FAC disinfection at pH 7 and 5°C, based on relative sensitivity. Very sensitive, n = 4; sensitive, n = 3; tolerant, n = 1. Clusters were generated using analysis of variance. Error bars represent standard deviations.