Antimicrobial resistance (AMR) is a major global health threat. Pharmaceutical manufacturing and wastewater treatment plants contribute significantly to AMR by releasing effluent containing antibiotics and various other biologically active compounds. While the impact of antibiotics on bacterial evolution is well-understood, the effects of long-term exposure to non-antibiotic pharmaceuticals, both active pharmaceutical ingredients (APIs) and formulation components, are less clear. These compounds are ubiquitous environmental contaminants, detected globally in water bodies. Prior research has shown some species- and compound-specific antimicrobial activity of certain non-antibiotic pharmaceuticals and titanium dioxide (TiO2), a common additive. There's also emerging evidence suggesting some non-antibiotic pharmaceuticals may influence bacterial susceptibility to antibiotics, either by enhancing antibiotic resistance gene uptake or promoting plasmid transfer. However, the short- and long-term effects of exposure to these compounds on bacterial populations, including their intrinsic toxicity and potential for selecting for cross-resistance to antibiotics, require further investigation. This study aimed to investigate the short- and long-term effects of a panel of non-antibiotic pharmaceuticals (acetaminophen, ibuprofen, propranolol, metformin) and TiO2 on *E. coli* K-12 MG1655, focusing on environmentally relevant concentrations to assess the potential for selecting cross-resistance to antibiotics. The selection of these compounds is justified by their presence in wastewater and their common usage. TiO2 was included due to its widespread use and potential impact on microbial communities.

Previous research has explored the antimicrobial effects of various non-antibiotic pharmaceuticals. Some studies have demonstrated antimicrobial activity against several bacterial and fungal species, including *E. coli*, *S. aureus*, *P. aeruginosa*, and *C. albicans*, with some compounds exhibiting species-specific activity. A screen of non-antibiotic pharmaceuticals against gut bacteria revealed numerous drugs negatively impacting the growth of at least one species. TiO2, used in various products, also exhibits antibacterial effects against several species. The impact of non-antibiotic pharmaceuticals on antibiotic resistance is also a focus of recent studies. Some compounds, including ibuprofen, diclofenac, and propranolol, have been linked to increased uptake of antibiotic resistance genes, potentially due to changes in cell permeability. Conversely, other studies suggest that metformin and benzydamine may reverse resistance phenotypes. This body of work highlights the complexity of non-antibiotic pharmaceutical effects and underscores the need for further investigation into their potential to influence bacterial populations and antibiotic resistance.

The study employed several methods to investigate the effects of non-antibiotic pharmaceuticals and TiO2 on *E. coli*. Initially, a toxicity screen was performed using a range of concentrations (Supplementary Table 1) for each compound against *E. coli* K-12 MG1655. Growth kinetics were monitored over 24 hours using a microplate reader. Transcriptomic analysis was conducted on *E. coli* populations grown in the presence and absence of ibuprofen to identify differentially expressed genes. RNA sequencing was performed, and differential gene expression was quantified using Kallisto, with an annotated long-read assembly of the ancestral *E. coli* as a reference. A long-term selection experiment was then conducted. Six independent *E. coli* populations were propagated in the presence of each of the selected compounds (and a control), with daily passaging for 30 days. Growth kinetics of both the ancestral and evolved populations were assessed in both the presence and absence of the selection compound. To rule out short-term, reversible toxicity effects, a seven-day recovery experiment was also performed, where evolved populations were passaged daily in nutrient broth without the selection compound. Whole-genome sequencing (Illumina short-read and MinION long-read) was performed on the ancestral and evolved populations to identify single nucleotide polymorphisms (SNPs) and assess insertion sequence (IS) element movement. Finally, minimum inhibitory concentration (MIC) assays were performed on the ancestral and ibuprofen-evolved *E. coli* populations, using various antimicrobial agents, to determine whether ibuprofen exposure altered antibiotic resistance.

The toxicity screen revealed that all five compounds (acetaminophen, ibuprofen, propranolol, metformin, and TiO2) significantly reduced the maximum optical density (OD) of *E. coli* at various concentrations, including environmentally relevant levels (p < 0.05, one-way ANOVA). Transcriptomic analysis of *E. coli* exposed to ibuprofen identified 16 significantly upregulated genes (FDR < 0.05, absolute log fold change ≥ 1), including genes associated with stress response (*yhcN*, *yhiM*, *bhsA*, *insA*), multidrug efflux (*mdtE*, *emrA*, *emrD*), and nickel transport (*nikA*). The long-term selection experiment showed that evolved populations displayed decreased maximum OD compared to the ancestral lineage when grown in the presence of the selection compound, suggesting that prolonged exposure did not lead to improved growth. Genome sequencing of evolved populations revealed no parallel SNPs between independently evolved populations, indicating no strong selective pressure leading to mutation fixation. Analysis of IS elements showed no evidence of movement between evolved isolates and the ancestor. Furthermore, MIC assays demonstrated no significant changes in antibiotic resistance profiles for the ibuprofen-evolved *E. coli* compared to the ancestral strain. This suggests no cross-resistance to the tested antibiotics was selected for despite the differential gene expression observed after ibuprofen exposure.

This study demonstrates that several non-antibiotic pharmaceuticals and TiO2 exhibit toxicity against *E. coli* at environmentally relevant concentrations. The upregulation of stress response and multidrug efflux genes in response to ibuprofen exposure suggests that bacteria employ mechanisms to mitigate the toxic effects of these compounds. Notably, despite this observed toxicity and the alteration of gene expression, long-term exposure did not lead to significant genetic changes or the evolution of cross-resistance to antibiotics. This finding suggests that, under the tested conditions, the selection pressure exerted by these compounds is not strong enough to drive the fixation of mutations conferring cross-resistance. The results provide valuable insights into the impact of non-antibiotic pharmaceuticals on bacterial populations and the potential for contributing to AMR. The lack of cross-resistance development is reassuring, indicating that these specific compounds, at the tested concentrations and timeframe, may not significantly contribute to the rise of AMR in the environment.

This research demonstrates that while several commonly used non-antibiotic pharmaceuticals and TiO2 exhibit toxicity towards *E. coli*, they do not appear to exert sufficient selective pressure to induce significant genetic adaptation or cross-resistance to antibiotics within the 30-day experimental timeframe. These findings provide initial reassurance regarding the contribution of these specific compounds to the global AMR problem, however, further studies involving environmentally-relevant microbial communities and broader ranges of concentrations and exposure times are needed. Investigating the impacts of other non-antibiotic pharmaceuticals and the combined effects of multiple compounds present in industrial effluent remains a critical area for future research.

The study's use of a laboratory strain of *E. coli* might limit the generalizability of findings to naturally occurring bacterial communities. The specific concentrations and exposure duration used may not fully capture the complexity of environmental exposure scenarios. Testing involved a relatively limited panel of non-antibiotic pharmaceuticals, so additional compounds should be examined. Finally, investigating other mechanisms of antibiotic resistance beyond MIC changes could strengthen future studies.