The rise of antibiotic resistance is a major global health concern, driven partly by the mobilization of antibiotic resistance genes (ARGs). Horizontal gene transfer, including transformation, conjugation, and transduction, plays a crucial role in ARG dissemination. While conjugation has been widely studied, the role of transduction via bacteriophages is increasingly recognized, especially given the abundance of phages in various environments. However, concerns exist regarding overestimation of ARG abundance in phage fractions due to contamination from non-packaged bacterial DNA or OMVs. This study aimed to clarify the extent to which phages, OMVs, or free DNA contribute to ARG carriage in the viromes of retail food sources, focusing on a rigorous protocol designed to minimize contamination.

The literature highlights the importance of understanding ARG mobilization mechanisms. Conjugation, involving plasmid transfer, is a prevalent mechanism, especially in dense microbiomes. While transduction is well-established, its relative contribution to ARG spread has been debated, partly due to challenges in isolating and propagating transducing phage particles. Many studies have shown ARGs in phage DNA fractions but lack confirmation of ARG packaging within phage capsids. The presence of 16S rRNA genes in the phage fraction has also been questioned as a potential indicator of contamination, although 16S rRNA can be packaged within phage particles. Previous research suggests that the frequency of ARGs in phage fractions may be overestimated due to contamination by non-packaged DNA or OMVs. This study addresses this concern by using advanced protocols to minimize such contamination.

The study used samples of chicken, fish, and mussels purchased from local supermarkets. Samples were homogenized, and viral fractions were obtained through filtration (0.22 μm). Chloroform treatment and DNase I digestion were used to remove bacterial cells and free DNA. ARG-carrying phages were propagated in *E. coli* WG5 cultures. CsCl density gradient centrifugation further purified the phage particles. The removal of external DNA was verified by qPCR amplification before capsid lysis. ARGs (*bla*_TEM, *sul1*, and *tetW*) were quantified using qPCR after proteinase K digestion of phage capsids. Plaque blot analysis assessed the infectivity of ARG-carrying phages. SYBR Gold staining and confocal microscopy were used to visualize phage attachment and infection of *E. coli* WG5 cells. Transmission electron microscopy (TEM) characterized phage morphology. Proteomic analysis using LC-MS/MS identified the proteins present in the purified phage fraction, enabling assessment of OMV contamination. Appropriate controls were included throughout the process to rule out the presence of free DNA or plasmid-mediated ARG transfer.

The study found that after the first propagation round, *bla*_TEM and *sul1* were detected in mussels and chicken, and *bla*_TEM and *tetW* in fish. The absence of ARGs in directly analyzed chicken samples was likely due to concentrations below the detection limit. Propagation resulted in a significant increase in ARG copy numbers (3–5 log₁₀ units), suggesting infectivity of some ARG-carrying phages. However, subsequent propagation rounds did not show further increases in ARG copy numbers, indicating that most ARG-carrying phages from the first round were not able to propagate further. Plaque blot hybridization failed to detect plaques positive for *bla_TEM*, suggesting that phages carrying this ARG might be present at low proportions compared to other virulent phages. Confocal microscopy showed phage particles from all samples attaching to *E. coli* WG5 cells, some showing signs of DNA injection. TEM revealed the presence of siphophages and myophages, with no OMVs observed. Proteomic analysis indicated a high percentage of phage proteins (52.9–78.7%) and a low percentage of membrane proteins (3.8–21.4%), confirming effective removal of OMVs. The results show that phages were the primary carriers of ARGs in these food samples.

The findings strongly suggest that phage particles, including both infectious phages and transducing particles, significantly contribute to ARG dissemination in food viromes. The increase in ARG copy numbers after the first propagation round, coupled with the subsequent lack of increase, supports the idea that the initial detection is primarily due to infectious phages, with transducing particles potentially playing a role but unable to generate further propagation. The effective removal of OMVs and free DNA using the employed methodology reinforces the reliability of detecting phage-associated ARGs. The study highlights the importance of considering both infectious and non-infectious phage particles when assessing the role of phages in ARG dissemination. Future research could focus on identifying specific phage types carrying ARGs and exploring the mechanisms of transduction in different food matrices.

This study provides robust evidence that phage particles are major carriers of ARGs in retail food sources. The rigorous methodology employed, including CsCl density gradient centrifugation and DNase treatment, minimized contamination and ensured reliable detection of ARGs within phage capsids. The low abundance of membrane proteins in the proteomic analysis further confirms the minimal contribution of OMVs. The findings highlight the importance of considering phages as key players in ARG mobilization and underscore the need for continuous research to understand the full extent of their role in the spread of antibiotic resistance. Future studies could focus on identifying specific phage-ARG combinations, exploring transduction efficiency in various host strains and food products, and investigating the ecological factors that influence phage-mediated ARG dissemination.

The study focused on a limited number of food types and ARGs. The chosen host strain, *E. coli* WG5, might not represent the full spectrum of potential phage hosts in these environments. The inability to detect plaques positive for *bla_TEM* suggests potential limitations in detecting all ARG-carrying phages. Finally, while the protocol successfully minimized OMV contamination, the possibility of extremely low-level contamination cannot be entirely excluded.