Unrecognized risk of perfluorooctane sulfonate in promoting conjugative transfers of bacterial antibiotic resistance genes

Abstract

Antibiotic resistance is a major global health crisis facing humanity, with horizontal gene transfer (HGT) as a principal dissemination mechanism in the natural and clinical environments. Perfluoroalkyl substances (PFASs) are emerging contaminants of global concern due to their high persistence in the environment and adverse effects on humans. However, it is unknown whether PFASs affect the HGT of bacterial antibiotic resistance. Using a genetically engineered Escherichia coli MG1655 as the donor of plasmid-encoded antibiotic resistance genes (ARGs), E. coli J53 and soil bacterial community as two different recipients, this study demonstrated that the conjugation frequency of ARGs between two E. coli strains was (1.45 ± 0.17) × 10^-5 and perfluorooctane sulfonate (PFOS) at environmentally relevant concentrations (2-50 μg L^-1) increased conjugation transfer between E. coli strains by up to 3.25-fold. Increases in reactive oxygen species production, cell membrane permeability, biofilm formation capacity, and cell contact in two E. coli strains were proposed as major promotion mechanisms from PFOS exposure. Weighted gene co-expression network analysis of transcriptome data identified a series of candidate genes whose expression changes could contribute to the increase in conjugation transfer induced by PFOS. Furthermore, PFOS also generally increased the ARG transfer into the studied soil bacterial community, although the uptake ability of different community members of the plasmid either increased or decreased upon PFOS exposure depending on specific bacterial taxa. Overall, this study reveals an unrecognized risk of PFOS in accelerating the dissemination of antibiotic resistance. IMPORTANCE Perfluoroalkyl substances (PFASs) are emerging contaminants of global concern due to their high persistence in the environment and adverse health effects. Although the influence of environmental pollutants on the spread of antibiotic resistance, one of the biggest threats to global health, has attracted increasing attention in recent years, it is unknown whether environmental residues of PFASs affect the dissemination of bacterial antibiotic resistance. Considering PFASs, often called "forever" compounds, have significantly higher environmental persistence than most emerging organic contaminants, exploring the effect of PFASs on the spread of antibiotic resistance is more environmentally relevant and has essential ecological and health significance. By systematically examining the influence of perfluorooctane sulfonate on the antibiotic resistance gene conjugative transfer, not only at the single-strain level but also at the community level, this study has uncovered an unrecognized risk of PFASs in promoting conjugative transfers of bacterial antibiotic resistance genes, which could be incorporated into the risk assessment framework of PFASs.

Keywords: PFASs; antibiotic resistance; bacterial community; emerging pollutant; horizontal gene transfer; risk assessment.

Fig 1

Influence of PFOS on the conjugative transfer of ARGs between two E. coli strains. (a) Change of transfer frequency and bacterial growth in the presence of PFOS at different concentrations. (b) Electrophoresis of two genes of bla and OriV in RP4 plasmid in the donor and transconjugants from different PFOS-exposed groups. (c) Influence of PFOS on transfer frequency of plasmid in three different conjugation systems, in which system I represents the donor that was pre-exposed to different concentrations of PFOS, and then conjugated with the untreated recipient bacteria in PFOS-free buffer. System II represents the recipient bacteria that were pre-exposed to different concentrations of PFOS, and then conjugated with the untreated donor bacteria in PFOS-free buffer. System III represents the conjugation between the donor and recipient bacteria without pre-exposure to PFOS. The asterisk denotes a statistically significant difference from unexposed control at the 0.05 level, as derived from ANOVA followed by post hoc Tukey’s test. Data are means ± SE.

Fig 2

Influence of PFOS exposure on phenotypic characteristics of donor and recipient bacteria. Changing bacterial (a) ROS production, (b) inner membrane permeability, and (c) outer membrane permeability upon PFOS exposure. (d) TEM observation of bacterial aggregation between donor and recipient after 8 h grown in PFOS-free PBS or 10 µg L⁻¹ PFOS-amended PBS, typical flagella of E. coli grown in the above two treatments were highlighted in the black dotted boxes. (e–g) Crystal violet staining experiment examined the biofilm formation capacities of (e) donor alone, (f) recipient alone, and (g) donor/recipient together upon PFOS exposure. The asterisk denotes a statistically significant difference from unexposed control at the 0.05 level, as derived from ANOVA followed by post hoc Tukey’s test. Data are means ± SE. The data associated with this figure are also displayed in supplemental Table S4.

Fig 3

Weighted gene co-expression network analysis of candidate genes that modulate the conjugative transfer between two E. coli strains. Hierarchical cluster tree showing modules of co-expressed genes in (a) the donor and (b) recipient bacteria. Each DEG is represented by a leaf in the tree and each module by a major tree branch. The lower panel shows modules in designated colors. Module-trait associations for (b) the donor and (e) the recipient bacteria were evaluated by correlations between MEs and the traits of transfer frequency (conjugation), ROS generation (ROS), inner membrane permeability (PI), and outer membrane permeability (NPN). Each row corresponds to an ME, column to a trait. Each cell contains the corresponding correlation (first line) and P value (second line). Identified genes that are significantly correlated to the bacterial traits in (c) the donor and (f) the recipient. These genes are displayed with phylogenetic tree, the heatmap of the gene expression level, and function annotation. BP, MF, and CC represent biological process, molecular function, and cellular component groups of GO, respectively.

Fig 4

CLSM images of plasmid transfer from the E. coli MG1655 donor strain to soil bacteria in the absence and presence of PFOS at different concentrations.

Fig 5

Comparison of recipient and transconjugant communities in the absence and presence of different concentrations of PFOS. Bacterial community composition in (a) recipient and (b) transconjugant communities. Alpha diversity of the bacterial community in (c) recipient and (d) transconjugant communities. (e) Comparison of the community composition of transconjugant and recipient pools at different treatments by nonmetric multidimensional scaling (NMDS) analysis. (f) Bary-Curtis distances between different treatments. (g) Phylogenetic composition of transconjugant communities (abundance > 0.05%) at different treatments. The bar charts in the first annulus represent the average abundance of the taxa in the transconjugant communities irrespective of PFOS concentration. Different letters denote a statistically significant difference between different treatments at the 0.05 level, as derived from ANOVA followed by post hoc Tukey’s test. Data are means ± SE.