Identification of a parasitic symbiosis between respiratory metabolisms in the biogeochemical chlorine cycle

Abstract

A key step in the chlorine cycle is the reduction of perchlorate (ClO₄^-) and chlorate (ClO₃^-) to chloride by microbial respiratory pathways. Perchlorate-reducing bacteria and chlorate-reducing bacteria differ in that the latter cannot use perchlorate, the most oxidized chlorine compound. However, a recent study identified a bacterium with the chlorate reduction pathway dominating a community provided only perchlorate. Here we confirm a metabolic interaction between perchlorate- and chlorate-reducing bacteria and define its mechanism. Perchlorate-reducing bacteria supported the growth of chlorate-reducing bacteria to up to 90% of total cells in communities and co-cultures. Chlorate-reducing bacteria required the gene for chlorate reductase to grow in co-culture with perchlorate-reducing bacteria, demonstrating that chlorate is responsible for the interaction, not the subsequent intermediates chlorite and oxygen. Modeling of the interaction suggested that cells specialized for chlorate reduction have a competitive advantage for consuming chlorate produced from perchlorate, especially at high concentrations of perchlorate, because perchlorate and chlorate compete for a single enzyme in perchlorate-reducing cells. We conclude that perchlorate-reducing bacteria inadvertently support large populations of chlorate-reducing bacteria in a parasitic relationship through the release of the intermediate chlorate. An implication of these findings is that undetected chlorate-reducing bacteria have likely negatively impacted efforts to bioremediate perchlorate pollution for decades.

Fig. 1. Isolation of chlorate-reducing bacteria from perchlorate-reducing cultures.

a Pathways for the respiration of perchlorate (red) and chlorate (orange) involve the enzymes perchlorate reductase (Pcr) or chlorate reductase (Clr), chlorite dismutase (Cld), and a terminal oxidase reducing oxygen to water (TO). b Binning and key genes of genomes from perchlorate-reducing cultures. A previously sequenced perchlorate-reducing enrichment is included for comparison (“1% NaCl Enrichment”). Filled squares indicate gene presence. Relative abundance (%) was calculated as normalized coverage divided by total coverage for all genomes. Compl. (%) refers to percent completeness (single copy genes); dashed lines indicate medium quality (M) and high quality (H) completeness. All genomes had negligible contamination (<3%). c Magnified image of colonies that developed in agar media supplied perchlorate or chlorate from a co-culture of Denitromonas halophilus SFB-1 and Pseudomonas stutzeri CAL. The inoculum for perchlorate agar media was ten-times more concentrated. d Dissimilatory reduction of chlorate and not perchlorate by isolated chlorate-reducing bacteria.

Fig. 2. Co-cultivation of perchlorate-reducing bacteria (red, PRB) and chlorate-reducing bacteria (orange, CRB) in defined and undefined communities.

a Fold change of clrA in defined co-cultures between lag phase and late exponential phase batch growth. For the co-culture consisting of P. stutzeri PDA and A. suillum PS, primers for 16S rRNA genes were used. Arrows indicate the upper and lower bounds of fold change estimated from the initial and final OD600 of the co-culture. Boxplots indicate quartiles in the sample. b Relative abundance of 16S rRNA gene amplicon sequence variants grouped by similarity to the 16S rRNA genes of perchlorate- and chlorate-reducing taxa. Stacked bars are in the same order as the legend. *, strains most closely related to perchlorate-reducing MAGs for which 16S rRNA genes were not available.

Fig. 3. Determination of the perchlorate reduction intermediate that supports growth chlorate-reducing bacteria in defined co-cultures.

a Genotype and phenotype (in pure culture) of chlorate reduction pathway mutants constructed in Pseudomonas stutzeri PDA (PDA). A chlorate reduction mutant would be unable to grow unless it can use the intermediate produced by perchlorate-reducing bacteria as a respiratory electron acceptor. b Fitness of chlorate reduction mutants in co-culture with Azospira suillum PS (PS) provided 10 mM perchlorate and 40 mM lactate, which PDA does not ferment. Relative abundance was calculated from qPCR measurements of both the PS and PDA 16S rRNA genes. *, significance of p < 0.05 (two-sided T-test); n.s., p > 0.05. Boxplots indicate quartiles in the sample with outliers as circles. c Maximum concentration of chlorate during dissimilatory perchlorate reduction by different strains of perchlorate-reducing bacteria (PRB) supplied 10 mM perchlorate. d Concentrations of perchlorate and chlorate during dissimilatory perchlorate reduction by PS or PS and PDA. Errors bars represent standard deviation of at least three replicates.

Fig. 4. Modeling of perchlorate reduction and chlorate reduction.

Simulated growth curves for perchlorate-reducing bacteria (a–c) alone and (d–f) with chlorate-reducing bacteria. [ClO₃⁻] / [ClO₄⁻], the ratio between chlorate concentration and perchlorate concentration; fraction ClO₃⁻ to CRB indicates the relative amount of chlorate consumed by the chlorate-reducing population at each time step; growth rate, the change in cell concentration between each time step.

Fig. 5. Model for the production and degradation of chlorine oxyanions.

The perchlorate reduction pathway (red) accumulates chlorate, which can react with reductants and generate reactive chlorine species (gray) or be consumed by the chlorate reduction pathway (orange). We did not find evidence for the release of chlorite and oxygen by the perchlorate and chlorate reduction pathways, though both chemicals can react with any reductants in the periplasm. Perchlorate and chlorate reduction remove the products of atmospheric oxidation of chlorine (dashed yellow). Co-metabolic or inadvertent enzyme activities are not shown.