Antidepressant drugs promote the spread of broad-host-range plasmid in mouse and human gut microbiota

Abstract

Antibiotic resistance is a global public health challenge. The gut microbiota serves as a reservoir for antibiotic resistance genes (ARGs), facilitating their transfer between bacteria. With the rising incidence of major depressive disorders (MDD), antidepressant prescriptions have surged. Previous pure-culture studies suggest that antidepressants exhibit antibiotic-like side effects, but their impact on gene transfer in microbial communities remains unclear. Here, we report that clinically relevant doses of antidepressants duloxetine and sertraline enhance the transfer of a broad-host range conjugative plasmid across bacterial phyla from mice gut microbiota. Tests in human gut simulators confirmed that duloxetine facilitates plasmid transfer in human gut microbiota. Mechanistic analyses revealed that antidepressants increase reactive oxygen species levels and alter bacterial cell membrane permeability. Using fluorescence-activated cell sorting and 16S rRNA gene sequencing, we discovered that antidepressants alter plasmid transfer patterns at both phylum and genus levels, driving ARG exchange among opportunistic pathogens. Our findings suggest that antidepressant use may promote the spread of antibiotic resistance between commensal and pathogenic bacteria, raising important public health concerns.

Keywords: Antidepressants; antibiotic resistance; conjugation; gut microbiota; horizontal gene transfer; plasmid.

Figure 1.

Conjugative transfer of plasmid pKJK5 from an E. coli donor to mice gut microbiota in the presence of antidepressants at different concentrations. (a) Schematic of the conjugation assay. (b) The Fold change of plasmid pKJK5 conjugation ratio upon the exposure to antidepressants at different concentrations. Significant differences between antidepressant treated groups and the control group were calculated using independent-sample t test, p values were then corrected by the “Benjamini–Hochberg” method as p, *p_adj < 0.05, and **p_adj < 0.01.

Figure 2.

The proportion of live (purple), damaged (green), and dead (pink) bacteria after exposing the gut microbiota to antidepressants at various concentrations. (a) Live and dead staining results of gut microbiota after the exposure antidepressants for 2 h (instant toxicity). (b) Live and dead staining results of gut microbiota after the exposure antidepressants for 24 h (toxicity after the exposure of mating time). Significant differences between antidepressant treated groups and the control group were calculated using independent-sample t test, p values were then corrected by the “Benjamini–Hochberg” method as p, *p_adj < 0.05, and **p_adj < 0.01.

Figure 3.

Effects of antidepressants (D-duloxetine;F-fluoxetine;S-sertraline) on ROS production in the donor (E. coli K-12 MG1655; pink bar) and the recipient (mice faecal bacteria; blue bar). Significant differences between antidepressant treated groups and the control group were calculated using independent-sample t test, p values were then corrected by the “Benjamini–Hochberg” method as p, *p_adj < 0.05, and **p_adj < 0.01.

Figure 4.

Exposure to antidepressants altered the transconjugant pools in mice gut microbiota. (a) Effect of antidepressants on alpha diversity of faecal microbiota, assessed by observed species, Chao1 index, Shannon index, Simpson index, and goods’ coverage index. (b) Effect of antidepressants on beta diversity of faecal microbiota, assessed by principal component analysis (c) venn diagram of recipient (R: extracted mice fecal bacteria) and antidepressants treated transconjugant pools (0: control; D-duloxetine;F-fluoxetine;S-sertraline) on phylum level. (d) Relative abundance of common taxa shared by the recipient and transconjugant pools on phylum level. (e) Phylogenic composition of recipient and transconjugant pools showing the relative abundance of 91 identified operational taxonomic units (OTUs) on genus level. The heatmap-circle at the periphery of the tree represents the bacteria counts in each OTU.

Figure 5.

(a) Percentage of donor E. coli carrying pKJK5 plasmid in different colon sections of the human gut simulator system with and without 24 h duloxetine treatment, donor percentage was divided by total microorganism counts. (b) Conjugative transfer of plasmid pKJK5 in the different colon sections of the human gut simulator system with and without the 24 h exposure to duloxetine. Significant differences between duloxetine treated groups and the control group were calculated using independent-sample t test, p values were then corrected by the “Benjamini–Hochberg” method as p, *p_adj < 0.05, and **p_adj < 0.01.