Genetic (in)stability of 2,6-dichlorobenzamide catabolism in Aminobacter sp. strain MSH1 biofilms under carbon starvation conditions

Abstract

Aminobacter sp. strain MSH1 grows on and mineralizes the groundwater micropollutant 2,6-dichlorobenzamide (BAM) and is of interest for BAM removal in drinking water treatment plants (DWTPs). The BAM-catabolic genes in MSH1 are located on plasmid pBAM1, carrying bbdA, which encodes the conversion of BAM to 2,6-dichlorobenzoic acid (2,6-DCBA) (BbdA+ phenotype), and plasmid pBAM2, carrying gene clusters encoding the conversion of 2,6-DCBA to tricarboxylic acid (TCA) cycle intermediates (Dcba+ phenotype). There are indications that MSH1 easily loses its BAM-catabolic phenotype. We obtained evidence that MSH1 rapidly develops a population that lacks the ability to mineralize BAM when grown on nonselective (R2B medium) and semiselective (R2B medium with BAM) media. Lack of mineralization was explained by loss of the Dcba+ phenotype and corresponding genes. The ecological significance of this instability for the use of MSH1 for BAM removal in the oligotrophic environment of DWTPs was explored in lab and pilot systems. A higher incidence of BbdA+ Dcba- MSH1 cells was also observed when MSH1 was grown as a biofilm in flow chambers under C and N starvation conditions due to growth on nonselective residual assimilable organic carbon. Similar observations were made in experiments with a pilot sand filter reactor bioaugmented with MSH1. BAM conversion to 2,6-DCBA was not affected by loss of the DCBA-catabolic genes. Our results show that MSH1 is prone to BAM-catabolic instability under the conditions occurring in a DWTP. While conversion of BAM to 2,6-DCBA remains unaffected, BAM mineralization activity is at risk, and monitoring of metabolites is warranted.