Non-antibiotic pharmaceuticals promote the transmission of multidrug resistance plasmids through intra- and intergenerational conjugation

Antimicrobial resistance is a major public health threat, causing an estimated 700,000 deaths annually. Bacterial antibiotic resistance arises via mutations or acquisition of antibiotic resistance genes (ARGs) through horizontal gene transfer (HGT), which occurs by conjugation, transformation, or transduction. Conjugation is the primary mechanism disseminating resistance. Sub-inhibitory antibiotic exposures can promote HGT, with prior work showing aminoglycosides and fluoroquinolones induce transformability in Streptococcus pneumoniae. Despite non-antibiotic, human-targeted pharmaceuticals comprising ~95% of drug consumption, their role in resistance emergence and spread has been underexplored. Recent screening of >1,000 marketed drugs against gut bacteria showed >200 non-antibiotic drugs exert antibiotic-like effects and can contribute to resistance via increased efflux pump expression. The authors previously demonstrated several non-antibiotic pharmaceuticals can enhance natural transformation in Acinetobacter baylyi. However, it remains unknown whether such pharmaceuticals promote conjugation between intra- or intergenerational bacterial populations—of particular concern given conjugative multidrug resistance plasmids can rapidly spread resistance. This study investigates whether commonly used non-antibiotic pharmaceuticals promote conjugative transfer of plasmid-borne ARGs and explores shared features or mechanisms underlying this effect.

The paper situates its research within evidence that: (1) HGT is central to the spread of antibiotic resistance, with conjugation as the dominant route; (2) sub-MIC antibiotic exposures can enhance HGT; (3) non-antibiotic pharmaceuticals, which dominate human drug use, have been shown to exert antibiotic-like effects on bacteria and increase efflux pump expression (Maier et al.), potentially contributing to resistance; and (4) prior work from the authors showed non-antibiotic pharmaceuticals can facilitate natural transformation. The gap identified is whether non-antibiotic pharmaceuticals also promote conjugation across intra- and intergenerational populations and whether common properties drive such promotion.

Two conjugation models were used. Model-1 (environmentally relevant): Donor Escherichia coli K-12 LE392 carrying the conjugative RP4 plasmid (resistant to tetracycline, kanamycin, ampicillin); recipient Pseudomonas putida KT2440 with intrinsic chloramphenicol resistance. Donor and recipient at 1×10^6 cfu/mL were mixed 1:1 in phosphate-buffered saline (PBS, pH 7.2) at 25 °C without shaking for 8 h. Non-antibiotic pharmaceuticals were added at clinically/environmentally relevant, sub-MIC concentrations: ibuprofen, naproxen, gemfibrozil, diclofenac, propranolol each at 0.005, 0.05, 0.5, 50 mg/L; iodipamide at 0.01, 0.1, 1, 5, 50 mg/L. Transconjugants were enumerated on selective LB agar (tetracycline, kanamycin, ampicillin, chloramphenicol), and transfer ratio calculated as transconjugants divided by recipients. A ROS scavenger (thiourea, 100 μM) was included in some sets. Reverse transfer experiments used Model-1 transconjugants as donors to a chloramphenicol-resistant E. coli MG1655 recipient, with selection on M Endo agar with appropriate antibiotics. Model-2 (clinically relevant): Donor E. coli MG1655 harboring pSM198A (blaNDM-1 positive IncA/C plasmid from a multidrug-resistant uropathogenic strain) and recipient E. coli J53 (sodium azide resistant) were grown to OD600 ≈ 1.8, mixed 1:2, and incubated for 2 h at 37 °C static in LB with the same pharmaceutical dosing scheme, followed by enumeration on selective media. Transformation was ruled out given large plasmid sizes (RP4 60.09 kb; pSM198A 137.57 kb) and non-competent recipients. Plasmid verification: Transconjugants were screened by PCR for RP4 traF, tetA, blaTEM; for Model-2, presence of the clinical plasmid was confirmed by PCR (e.g., blaTEM) after genomic DNA extraction; plasmid profiles were assessed by agarose gel electrophoresis. Transmission electron microscopy (JEOL JEM-1011, 80 kV) examined cellular effects after 8-h mating with 0.5 mg/L of ibuprofen, naproxen, gemfibrozil, diclofenac, propranolol, or 1.0 mg/L iodipamide. ROS generation and membrane permeability were measured by flow cytometry using DCFDA (20 μM) and propidium iodide (2 mM), respectively, in donors and recipients after pharmaceutical exposure; thiourea was used to assess ROS dependence. Anaerobic conjugation and ROS assays for Model-1 were performed in an anaerobic chamber with oxygen-depleted media to test the role of ROS. Transcriptomics: Model-1 conjugations were conducted for 2 h with the pharmaceuticals (0.5 mg/L; iodipamide 1.0 mg/L), followed by total RNA extraction (mixed donor/recipient), library prep, and Illumina HiSeq 2500 paired-end sequencing. Reads were QC-filtered (NGSeq QC Toolkit v2.3.3), aligned (SeqAlto v0.5) to a combined reference (E. coli K-12 NC_000913; P. putida KT2440 NC_002947; RP4 plasmid L27758), and differential expression analyzed with Cufflinks v2.2.1/CummeRbund. Differentially expressed genes were defined by log2 fold-change ≥ 2.0 or ≤ −1.0 with P and q < 0.05. Proteomics: SWATH-MS on proteins extracted from mixed cultures after exposure to pharmaceuticals; time-course at 2, 4, 6, 8 h with gemfibrozil or propranolol identified 8 h as optimal based on significant protein changes (q < 0.01), then full proteomics at 8 h for all six drugs (0.5 mg/L; iodipamide 1.0 mg/L). Correlation analyses: Linear regressions of log10(concentration) vs phenotypes (transfer ratio, ROS, membrane permeability); concentration dependence inferred at R^2 > 0.9; Pearson correlations with significance at P < 0.05. Statistics: Triplicate experiments; data as mean ± SD; independent t-tests with Benjamini–Hochberg correction; significance at P < 0.05.

- Multiple non-antibiotic, human-targeted pharmaceuticals significantly accelerated conjugative transfer of plasmid-borne ARGs at clinically and environmentally relevant, sub-MIC concentrations. In Model-1 (E. coli LE392 [RP4] → P. putida KT2440), ibuprofen, naproxen, and gemfibrozil increased the absolute number of transconjugants significantly at all tested concentrations (0.005–50 mg/L; P = 1×10^-8 to 6×10^-4). Diclofenac and propranolol increased transconjugants at higher concentrations (5 or 50 mg/L). Iodipamide did not increase transconjugant numbers. Using ibuprofen as an example, fold change in transconjugant counts rose from 2.8 ± 0.2 at 0.005 mg/L to 7.3 ± 0.8 at 50 mg/L compared to control. Transfer ratios were significantly elevated by all non-antibiotic drugs except iodipamide at concentrations as low as 0.05 mg/L (P = 3×10^-3–0.017). - Mechanistic correlates: Drug-enhanced conjugation associated with increased reactive oxygen species (ROS) production and increased cell membrane permeability, as quantified by flow cytometry (DCFDA and PI staining). Addition of the ROS scavenger thiourea reduced ROS levels, mitigated membrane permeability increases, and diminished the enhancement of conjugation, supporting a ROS-mediated mechanism. Under anaerobic conditions, ROS generation and conjugation enhancement were attenuated, further implicating ROS. - Cellular and molecular responses: TEM and flow cytometry indicated altered cell arrangement and increased envelope permeability. Transcriptomics and proteomics showed induction of antibiotic-like stress responses, including SOS response activation and upregulation of efflux pumps, alongside increased whole-genome permeability, consistent with enhanced plasmid transfer. - The promotion of conjugation was observed at concentrations relevant to human exposure and environmental contamination, underscoring real-world relevance.

The study directly addresses whether non-antibiotic, human-targeted pharmaceuticals can promote conjugative transfer of ARGs. Results demonstrate that several widely used drugs (NSAIDs, a lipid-lowering agent, and a β-blocker) significantly increase conjugation frequencies at low, sub-MIC concentrations typical of clinical and environmental contexts. The enhancement correlates with ROS generation and increased membrane permeability, and is accompanied by induction of SOS and efflux systems—responses commonly seen with antibiotics—suggesting that these pharmaceuticals impose stresses that facilitate plasmid transfer. The attenuation of effects by ROS scavenging and under anaerobic conditions supports a ROS-dependent mechanism. Together, the findings broaden the drivers of ARG dissemination beyond antibiotic exposures, indicating that common non-antibiotic pharmaceuticals can act as facilitators of horizontal gene transfer, potentially accelerating the spread of multidrug resistance in microbial communities in clinical settings (e.g., gut, hospitals) and the environment (e.g., wastewater).

Commonly used non-antibiotic, human-targeted pharmaceuticals can significantly accelerate the conjugative transfer of multidrug resistance plasmids between bacteria at clinically and environmentally relevant concentrations. The promotion of transfer is associated with increased ROS, membrane permeability, and induction of antibiotic-like stress responses (SOS, efflux), revealing shared mechanistic pathways by which non-antibiotic drugs can influence ARG dissemination. These findings highlight the need to consider non-antibiotic pharmaceuticals in antimicrobial resistance risk assessments and management. Future research should: (1) validate these effects in complex microbiomes and in vivo (e.g., gut models, animal studies); (2) expand screening across broader drug classes and doses to identify structural features driving conjugation promotion; (3) quantify impacts across diverse donor–recipient pairs and plasmid backbones; and (4) develop mitigation strategies (e.g., targeted wastewater treatment, stewardship policies) to reduce selection and transfer pressures from pharmaceutical exposures.