Cyromazine accelerates the dissemination of antimicrobial resistance genes by promoting the conjugative transfer of the SXT integrative conjugative element

Abstract

Antimicrobial resistance (AMR) is an urgent public health challenge, and the horizontal transfer of antimicrobial resistance genes (ARGs) mediated by mobile genetic elements (MGEs) accelerates its spread worldwide. Despite the recognized role of integrative conjugative elements (ICEs) in conjugative gene transfer, ICE-mediated horizontal gene transfer (HGT) remains underexplored, particularly for SXT ICEs, which are widely disseminated and relevant to animal and human health. Cyromazine, a widely used veterinary drug frequently detected in livestock feces and associated environments, is linked to an increased abundance of ARGs, but the underlying mechanisms remain unclear. Here, we investigate the impact of cyromazine on SXT ICE–mediated ARG transfer and delineate potential mechanisms. We show that cyromazine promotes intra- and inter-species conjugative transfer of SXT ICE and validate this effect in vivo. In environmentally relevant models (feces, soil, and water), cyromazine further enhances SXT ICE transfer and reshapes the community structure of transconjugants. Potential mechanisms include cyromazine-induced ROS accumulation and SOS activation, which may promote SXT ICE excision and induce conjugation-related operons. Enhanced energy production and disrupted membrane homeostasis may further facilitate transfer. Collectively, these findings narrow the gap in ICE-mediated AMR transmission and suggest cyromazine could intensify AMR spread by stimulating SXT ICE conjugative transfer.

In this study, authors demonstrate that cyromazine, a widely used veterinary drug, speeds the transmission of antimicrobial resistance by boosting horizontal transfer of SXT integrative and conjugative elements in bacteria.

Introduction

The discovery of antibiotics revolutionized healthcare, laying the foundation for many groundbreaking medical advancements in the 20th century¹. However, the widespread misuse of antibiotics has exacerbated the issue of bacterial antimicrobial resistance (AMR)². Bacterial AMR has emerged as a stealthy yet relentless global crisis, with its threats reaching unprecedented levels^3–5. This AMR crisis is largely fueled by horizontal gene transfer (HGT) through conjugation, transformation, transduction, and vesiduction⁶. Notably, conjugation stands out as the most efficient and predominant mechanism for the dissemination of antimicrobial resistance genes (ARGs), mainly through mobile genetic elements (MGEs), particularly conjugative plasmids and integrative conjugative elements (ICEs)^7–10. ICEs exhibit the unique ability to stably integrate into host chromosomes while retaining conjugative functionality, and they also mediate the transfer of larger genomic segments^6,11. This potentially offers evolutionary advantages and adaptability in certain ecological niches¹². Furthermore, ICEs outnumber conjugative plasmids¹³. However, plasmid-mediated ARG horizontal transfer has been extensively studied, and various antibiotic and non-antibiotic compounds have been reported to promote plasmid-mediated conjugative transfer, whereas research on ICE-mediated transfer remains disproportionately lacking^13–16. Thus, our study was designed to fill the gap in understanding MGE-mediated AMR transmission dynamics.

Among ICEs, SXT ICEs are of particular concern, as they confer multidrug resistance (MDR) to bacteria and play a pivotal role in exacerbating the global AMR crisis¹³. They possess a highly conserved backbone structure, are numerous and widely distributed, exhibit a broad host range, and are significantly associated with human health^17,18. Although clinically critical antibiotics such as tigecycline and meropenem are prohibited in veterinary medicine, corresponding resistance genes (tmexCD-toprJ, tet(X) and bla_NDM) mediated by SXT ICEs have nonetheless been detected in animal-derived bacteria^19–23. SXT ICEs are extensively distributed across diverse biological niches, including animals (both farmed and wild), humans, and the environment^24–26. Importantly, they are SOS-responsive, meaning that DNA damage, indicative of reduced cell viability, confers a selective advantage by promoting their mobilization and horizontal transfer^26–28. Owing to their broad host range, stable chromosomal integration, and efficient self-transmission, SXT ICEs represent one of the most potent vectors for the persistence and propagation of antibiotic resistance within microbial communities^8,26. Environmental and chemical stressors, including DNA-damaging agents²⁷, phages²⁶, ultraviolet light²⁹, antibiotics that disrupt DNA replication³⁰, and metal ions³¹ have been shown to promote SXT ICE conjugation. Given the serious hazards of SXT ICEs and their capacity for horizontal transfer and stable integration into bacterial chromosomes, SXT ICEs were selected as the model system for investigating conjugation-mediated HGT.

Cyromazine is a potent insect growth regulator that inhibits the growth, development, and maturation of fly larvae^32,33. Widely used as a feed additive in certain countries, cyromazine is primarily eliminated in its intact drug form through feces, where it accumulates and effectively controls maggot populations. The detection rate of cyromazine residues in animal feces, organic fertilizer samples, and animal manure-amended soils surpasses that of commonly used antibiotics in animal husbandry, such as sulfonamides, fluoroquinolones, and tetracyclines^34–36. Furthermore, the number of cyromazine premix products approved by the China Institute of Veterinary Drug Control has also steadily increased since 2018 (Fig. 1A). Notably, cyromazine has been shown to enhance the prevalence and dissemination of ARGs³⁷. Extending prior plasmid-based work, we evaluated whether cyromazine exerts a comparable pro-transfer effect on the SXT ICE³⁸. Framing the question in an SXT ICE context provides a more complete basis for assessing cyromazine-related ARGs dissemination risk.

Fig. 1. Cyromazine facilitates the conjugative transfer of SXT ICE.

A Number of approved cyromazine premix products (veterinary drug products) recorded by the China Institute of Veterinary Drug Control since 2018. B In vitro conjugation model of SXT ICE treated with cyromazine. C Fold change in SXT ICE conjugation frequency at different mating durations (4 h, 14 h and 24 h; donor: E. coli J53; recipient: E. coli C600; pH 7; 37 °C) under cyromazine treatment (0–10 μg/mL). D Fold change in conjugation frequency under different temperatures (4 °C, 24 °C, 37 °C and 42 °C; donor: E. coli J53; recipient: E. coli C600; pH 7; mating time 14 h) under cyromazine treatment (0–10 μg/mL). E Fold change in conjugation frequency under different pH values (pH 5–8; donor: E. coli J53; recipient: E. coli C600; 37 °C; mating time 14 h) under cyromazine treatment (0–10 μg/mL). F Fold change in conjugation frequency from different donors (E. coli J53, Proteus, Shewanella and Klebsiella) to E. coli C600 (mating time 14 h; pH 7; 37 °C) under cyromazine treatment (0–10 μg/mL). For (C–F), fold change is calculated relative to the untreated control (0 μg/mL cyromazine) within each subgroup (time/temperature/pH/donor). Bars show mean ± SD; dots indicate individual biological replicates (n  =  3 independent biological replicates per condition). The unit of study is an independently prepared donor–recipient mating assay initiated from independently prepared donor and recipient cultures. Statistical significance was assessed by two-way ANOVA (with interaction) followed by Dunnett’s multiple comparisons test (two-sided), comparing each cyromazine concentration with the 0 μg/mL control within each subgroup. Multiplicity-adjusted P values are denoted as *P  <  0.05, **P  <  0.01, ***P <  0.001, ****P  <  0.0001; adjusted P values (with 95% CIs) are provided in the Source Data. Source data are provided as a Source Data file.

In this study, we established an intraspecific conjugation model to investigate the influence of cyromazine on the excision and transfer of SXT ICEs. To further elucidate the broad-spectrum effects, we also evaluated the influence of cyromazine on interspecies conjugation. Furthermore, mouse intraperitoneal and intestinal conjugation models were used to examine the effect of cyromazine on SXT ICEs transfer in vivo. Meanwhile, complementary in vitro models simulating key environmental reservoirs (feces, soil, and water) were established to elucidate SXT ICE transfer dynamics in ecologically relevant contexts. Our study elucidated the mechanisms of cyromazine-enhanced conjugation using fluorescent probes and RT-qPCR. Collectively, cyromazine stimulates SXT ICE conjugation under both in vitro and in vivo conditions. It further facilitates conjugative transfer to environmentally relevant microbiota and alters the community structure of transconjugants. These findings refine our understanding of ICE-driven AMR transmission dynamics and identify cyromazine as a potent environmental driver of SXT ICE dissemination, thereby accelerating the spread of AMR.

Results

Cyromazine facilitates the conjugative transfer of SXT ICE

To assess the effects of cyromazine on SXT ICE conjugative transfer, we first evaluated its antibacterial activity and determined that the MICs for both the donor and recipient exceeded 100 μg/mL. Furthermore, sub-MIC concentrations (0.001, 0.01, 0.1, 1, and 10 μg/mL) of cyromazine showed no significant impact on the growth of donor and recipient bacteria (Fig. S1A, B). Thus, these concentrations (0.001–10 μg/mL) were selected for use in the intraspecific conjugation model using E. coli J53 harboring SXT ICE as the donor and E. coli C600 as recipient bacteria (Fig. 1B). Given that the conjugation efficiency of SXT ICEs can be affected by environmental factors, parameters including mating duration, temperature and pH were adjusted to better mimic natural conditions. As shown in Fig. 1C, cyromazine exhibited a time-dependent effect on conjugation frequency, peaking at 14 h, with negligible effects at 4 h and a decline observed at 24 h. The strongest effect of conjugation efficiency reached its zenith at 37 °C, which coincides with the optimal growth temperature of E. coli (Fig. 1D). Moreover, cyromazine promoted transfer under slightly acidic to neutral pH conditions, but inhibited it under alkaline conditions, except at 0.001 μg/mL (Fig. 1E). Based on these observations, the optimized parameters for subsequent experiments were established as 37 °C, a pH of 7.0, and a mating period of 14 h. Conjugation experiments were also performed using LB broth diluted to 30%, 60%, and 100% of the standard concentration. Cyromazine showed the strongest promoting effect on conjugative transfer in 100% LB broth, consistent with the more favorable growth conditions in full-strength LB (Fig. S1C). To investigate whether the promoting effect of cyromazine was applicable to interspecies conjugation and whether it is universal, wild-type isolates (Proteus, Klebsiella, and Shewanella) were utilized as donors. Genetic structure analysis of SXT ICEs from these wild-type isolates revealed analogous genetic environments and conserved backbone structure (Fig. S2A). Cyromazine at concentrations ranging from 0.001 to 10 μg/mL significantly enhanced the conjugative transfer frequency of SXT ICEs in these wild-type isolates (Fig. 1F). Taken together, these results demonstrate that cyromazine significantly promotes SXT ICE conjugative transfer both within and across bacterial species. Additionally, transconjugants were randomly selected from each concentration for confirmation through PCR and gel electrophoresis. The band corresponding to the SXT ICE integrase (int) gene (1035 bp) was detected in the donor bacteria and transconjugants but was absent in the recipient bacteria (Fig. S2B), confirming the reliability of the established conjugative system.

Cyromazine promotes SXT ICE conjugative transfer in vivo

Cyromazine is predominantly utilized as a feed supplement and is primarily eliminated by animals in its intact drug form. To simulate natural conditions, mouse models were established to investigate the effect of cyromazine on the in vivo conjugative transfer of SXT ICEs. Prior research indicates that cyromazine residues in organic fertilizers sourced from livestock manure range from 3.27 to 828.76 μg/kg³⁴. Based on these reported environmental levels and the concentrations used in the in vitro experiments, three concentration gradients (10, 100, and 1000 μg/kg) were selected for the in vivo conjugation experiments. A mouse intraperitoneal conjugation model was employed to investigate the impact of cyromazine on SXT ICE conjugative transfer in vivo (Fig. 2A). The results revealed that the conjugation frequency of SXT ICEs was markedly elevated in the liver, spleen, and kidneys after treatment with 10, 100, and 1000 μg/kg of cyromazine (Fig.2B). Additionally, a mouse intestinal conjugation model was designed to examine the effect of cyromazine on SXT ICE conjugation in the gut environment (Fig. 2C). The findings indicated that cyromazine remarkably promotes SXT ICE conjugation frequency in the gut at concentrations of 10, 100, and 1000 μg/kg (Fig. 2D). Consistent with the in vitro findings, cyromazine also enhanced SXT ICE conjugation under in vivo environmental conditions, suggesting that its environmental presence could contribute to the dissemination of AMR.

Fig. 2. Cyromazine increases the frequency of SXT ICE conjugative transfer in vivo.

A Scheme of the experimental protocol for the mouse intraperitoneal conjugation model. B Fold change in SXT ICE conjugation frequency in mouse liver, kidney and spleen following cyromazine treatment (10, 100 and 1000 μg/kg; control: PBS). C Schematic of the experimental protocol for the mouse intestinal conjugation model. D Fold change in SXT ICE conjugation frequency in mouse intestine following cyromazine treatment (10, 100 and 1000 μg/kg; control: PBS). For (B) and (D), fold change is calculated relative to the control group (0 μg/kg). Data are presented as mean ± SD; dots indicate individual mice (n  =  6 mice per group). The unit of study is an individual mouse. Statistical significance was assessed by one-way ANOVA followed by Dunnett’s multiple comparisons test (two-sided). Multiplicity-adjusted P values are denoted as *P  <  0.05, **P <  0.01, ***P  <  0.001, ****P <  0.0001; adjusted P values are provided in the Source Data. Source data are provided as a Source Data file.

Cyromazine promotes SXT ICE conjugative transfer and alters the transconjugant community structure of feces, soil, and water microbiota

To investigate the impact of cyromazine on SXT ICE conjugative transfer in environmentally relevant microbiota, we established feces, soil, and water conjugation models. A Dap-deficient E. coli X7213 harboring SXT ICE was used as the donor, while fresh rat feces-, lawn soil-, and pond water-derived microbiota served as recipient communities (Fig. 3A). Consistent with our in vitro experiments, cyromazine was applied at concentrations ranging from 0.001 to 10 μg/mL. As shown in Fig. 3B–D, all tested concentrations of cyromazine significantly promoted SXT ICE conjugative transfer across all three models. To further examine whether cyromazine altered the composition of the resulting transconjugant communities, 16S rRNA gene sequencing was performed on transconjugants recovered from the feces, soil, and water models. Principal component analysis (PCA) revealed clear separation between cyromazine-treated (0.1 and 1 μg/mL) and control groups within each model, while feces, soil, and water communities formed distinct, non-overlapping clusters, indicating strong environment-specific effects (Fig. 3E–G and Fig. S3A). These patterns were further supported by ANOSIM based on Bray–Curtis distances (Fig. 3H–J and Fig. S3B). To assess changes at the finest taxonomic resolution, Venn diagrams were used to visualize ASV overlap between cyromazine-treated and control samples within each model (Fig. S3C–E). As shown in Fig. 3K–M, alpha diversity changed in a model- and dose-dependent manner across the feces, soil and water models. Under 0.1 μg/mL cyromazine treatment, observed ASVs increased in the feces model but decreased in the soil and water models. In the soil model, Shannon and Simpson diversity increased under 0.1 and 1 μg/mL cyromazine, whereas Shannon diversity decreased at 1 μg/mL in the feces model. At the genus level, cyromazine consistently reshaped dominant transconjugant taxa in an environment-dependent manner (Fig. 3N). In the fecal model, Acinetobacter was dominant, followed by Providencia and Proteus. After cyromazine treatment, the proportion of Proteus increased while Providencia decreased. In the water model, Citrobacter was the most abundant, followed by Aeromonas and Pseudomonas. Following cyromazine exposure, Citrobacter declined, whereas both Aeromonas and Pseudomonas increased. In the soil model, Aeromonas was the predominant genus, followed by Pseudomonas; cyromazine treatment decreased Aeromonas and increased Pseudomonas. To further identify taxa driving these community-level shifts, cladogram and LEfSe analyses were performed (Fig. 3O–R). These analyses revealed that cyromazine selectively enriched distinct phylogenetic lineages in an environment-dependent manner. In the fecal model, Proteus-associated clades were significantly enriched under cyromazine exposure, whereas Providencia-related lineages were preferentially associated with the control group. In soil, cyromazine promoted Pseudomonas-related taxa, while Aeromonas-dominated clades were depleted. Similarly, in the water model, cyromazine treatment favored Aeromonas- and Pseudomonas-associated lineages over Citrobacter-related taxa. Collectively, these results demonstrate that cyromazine not only significantly enhances SXT ICE conjugative transfer into environmentally relevant microbial communities but also reshapes both the taxonomic composition and phylogenetic structure of the resulting transconjugant populations across distinct ecological contexts.

Fig. 3. Cyromazine enhances SXT ICE conjugative transfer and modifies the composition of the transconjugant community in feces, soil, and water microbiota.

A Scheme of the experimental protocol for the feces, soil, and water conjugation models. B–D Fold change in SXT ICE conjugation frequency in the feces (B), soil (C) and water (D) models following cyromazine exposure (0–10 μg/mL; 0, 0.001, 0.01, 0.1, 1 and 10 μg/mL; control, 0 μg/mL). Bars show mean ± SD; dots indicate independent biological replicates (n  =  6 per group). Statistical significance was assessed by one-way ANOVA followed by Dunnett’s multiple comparisons test (two-sided), comparing each cyromazine concentration with the corresponding control within each model. E–G Principal component analysis (PCA) of transconjugant communities. Plots show separation of control and cyromazine-treated groups (0.1, 1 μg/mL) in feces (E), soil (F), and water (G) models. Ellipses represent 95% confidence intervals for each treatment group. The percentage of total variance explained by each principal component (PC) is indicated on axes. Each point represents an independent model replicate (n = 3 per group). Group codes: FCON, F0.1Cyr, and F1Cyr denote feces samples under control, 0.1 μg/mL cyromazine, and 1 μg/mL cyromazine treatments, respectively; SCON, S0.1Cyr, and S1Cyr denote soil samples under the same treatments; WCON, W0.1Cyr, and W1Cyr denote water samples under the same treatments. (H–J) Bray–Curtis ANOSIM for transconjugant community differences in feces (H), soil (I) and water (J) models. Box plots show the ranked Bray–Curtis dissimilarities for between-group comparisons (“Between”) and within-group comparisons for each group (control and cyromazine-treated groups). ANOSIM R statistics and permutation-derived P values are shown above each panel. K–M Alpha diversity of transconjugant communities. Observed ASVs (K), Shannon (L) and Simpson (M) indices across feces, soil and water models under cyromazine exposure. Statistical significance was assessed by two-way ANOVA (with interaction) followed by Dunnett’s multiple comparisons test (two-sided). N Genus-level composition of transconjugant communities across environmental models. Only the top 10 most abundant genera are shown. O Phylogenetic analysis of dominant transconjugant genera. P–R LEfSe cladogram identifying differentially abundant taxa across feces (P), soil (Q), and water (R) models (LDA > 4.0, p < 0.05). The concentric circles represent phylogenetic levels from phylum (inner) to genus (outer). Group codes (applied to panels E–R): FCON, F0.1Cyr, and F1Cyr denote feces samples under control, 0.1 μg/mL cyromazine, and 1 μg/mL cyromazine treatments, respectively; SCON, S0.1Cyr, and S1Cyr denote soil samples under the same treatments; WCON, W0.1Cyr, and W1Cyr denote water samples under the same treatments. Box plots (H–M) show the median (center line) and interquartile range (25–75th percentiles; box). Whiskers indicate the most extreme values within Q1 − 1.5×IQR and Q3 + 1.5×IQR (Tukey). Data are based on n = 3 independent biological replicates per group (unit of study: an independent model replicate). For (B–D) and (K–M), multiplicity-adjusted P values are denoted as *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001; adjusted P values are provided in the Source Data. Source data are provided as a Source Data file.

Cyromazine exacerbates SXT ICE excision

It is widely recognized that the excision of SXT ICE from the chromosome is a prerequisite step for conjugative transfer, and promoting this excision process can enhance transfer efficiency. To assess whether cyromazine facilitates SXT ICE transfer by promoting excision, the excision frequency was quantified using RT-qPCR. SXT ICEs are integrated at the 5′ end of the chromosomal prfC gene, which encodes peptide chain release factor 3 and serves as a highly conserved chromosomal marker. Upon excision, a free attachment site (attB) is generated on the chromosome, while a new attachment site (attP) is formed on the excised circular SXT ICE. The attB/prfC ratio represents the SXT ICE excision frequency, attP/prfC reflects the relative abundance of the circular SXT ICE intermediates, and attP/attB indicates the extrachromosomal copy number of SXT ICE^39,40. As shown in Fig. 4A, cyromazine significantly increased SXT ICE excision frequency after 14 h of exposure to all tested concentrations of cyromazine, with a progressive rise observed as the concentration increased. A similar dose-dependent pattern was also observed after 4 h of treatment; however, the increase reached statistical significance only at 10 μg/mL (Fig. S4A). Furthermore, cyromazine treatment elevated the relative abundance of circular SXT ICE intermediates at both 4 h and 14 h (Fig. S4B, C), and markedly increased extrachromosomal copy numbers in the 0.01 μg/mL group at both time points (Fig. S4D, E). These results indicate that cyromazine promotes SXT ICE excision, providing a prerequisite for enhanced conjugative transfer.

Fig. 4. Cyromazine accelerates SXT ICE excision by stimulating ROS production and triggering the SOS response.

A Effects of cyromazine on SXT ICE excision in the donor strain (E. coli J53 carrying SXT ICE) after 14 h of treatment at concentrations of 0.001, 0.01, 0.1, 1, and 10 μg/mL. B Fold changes in ROS levels in donor and recipient bacteria after cyromazine exposure. C Heatmap of the expression of genes related to the oxidative stress and SOS response after cyromazine addition. The expression of genes associated with conjugative transfer in SXT ICE under cyromazine treatment is shown on the right. This heatmap depicts log₂ fold change computed as log₂(2^−ΔΔCt); values > 0 indicate upregulation (blue) and values < 0 indicate downregulation (green). D Effects of cyromazine on the conjugative transfer of SXT ICE with the addition of 5 mM NAC. Conjugation frequencies were normalized to the mean value of the 0 µg/mL without NAC (control group). E Effects of cyromazine on SXT ICE excision in the presence of 5 mM NAC (ROS scavenger). F Effects of cyromazine on the conjugative transfer of Proteus-derived SXT ICEs carried by E. coli C600 and E. coli C600ΔrecA. G The expression of int, xis, setC, setD and setR in E. coli C600ΔrecA after cyromazine exposure. Data are presented as mean ± SD (n = 3 independent biological replicates). Statistical analyses were performed as follows: (A,C,F,G) one-way ANOVA followed by Dunnett’s multiple comparisons test; (B) two-way ANOVA with Dunnett’s multiple comparisons test; (D) two-way ANOVA with Dunnett’s and Šídák’s multiple comparisons post hoc tests; (E) one-way ANOVA with Tukey’s multiple comparisons test. All tests were two-sided. Significance is indicated as *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001; adjusted P values are provided in the Source Data. Source data are provided as a Source Data file.

Previous studies have shown that cyromazine promotes conjugative transfer of RP4 plasmid by inducing ROS production and activating the SOS response³⁸. However, this plasmid does not require excision from the chromosome. In contrast, SXT ICEs are SOS responsive, with DNA damage triggering their de-repression, leading to SXT ICE excision and the expression of operons associated with conjugative transfer on SXT ICEs^26,27. The expression of genes involved in SOS response was examined after treatment with cyromazine (0.001−10 μg/mL). As shown in Fig. 4C, the expression of recA, recC, recF, recR, recQ, sulA, yebG, mutL, mutS, radA, dpiA, and dpiB was up-regulated, while that of lexA was down-regulated, demonstrating that cyromazine activates the SOS response. ROS directly react with DNA, causing oxidative damage, which in turn triggers the SOS response to repair the damaged DNA. The intracellular ROS levels in the control and cyromazine-treated groups were measured using the DCFH-DA fluorescent probe, showing a significant rise in fluorescence intensity with cyromazine doses ranging from 0.001 to 10 μg/mL, illustrating cyromazine-induced intracellular ROS accumulation (Fig. 4B). Meanwhile, cyromazine treatment also upregulated the expression of genes encoding specific stress proteins (uspA, uspC, uspD, uspE and uspF) as well as antioxidant-related genes (ahpC, ahpF, alkB, sodA, sodC, rpoS and katG), which may help alleviate the deleterious effects of ROS and oxidative stress (Fig. 4C). Herein, rpoS acts as a global regulator of the stress response and regulates the expression of antioxidant genes, ahpC and ahpF encode hydrogen peroxide reductase; alkB repairs DNA alkylation damage, sodA and sodC encode superoxide dismutase, katG encodes catalase. In the presence of the ROS scavenger NAC, SXT ICE conjugation frequency was markedly decreased (Fig. 4D). Moreover, the addition of NAC reduced SXT ICE excision frequency (Fig. 4E). A similar trend was observed for the abundance of circular intermediates (Fig. S4F). Taken together, these results suggest that ROS may contribute to SXT ICE excision and conjugative transfer. Therefore, cyromazine increases ROS, intensifies oxidative stress and triggers the SOS response, which may shift the SXT ICE element toward an excision- and transfer-permissive state.

Given that cyromazine promotes SXT ICE excision and transfer, we next examined the expression of conjugation-related genes within the SXT ICE element by RT-qPCR. After treatment with cyromazine (0.001−10 μg/mL), the expression of int, xis, setC, setD, bet, exo, umuC, umuD, traI, traL, traK, traB, traV, traH, traU, traC, and trhF were up-regulated, while that of setR was down-regulated (Fig. 4C). Among these genes, int encodes integrase, xis encodes excisionase, with int being responsible for integrating SXT ICE into the host chromosome, while xis enables its excision for conjugation. Furthermore, setR represses the expression of the transcriptional activators setC and setD, which control the recombination and tra modules in SXT ICE conjugation. The tra gene cluster is essential for the formation of the conjugation apparatus and regulation of DNA transfer^17,27,41. The coordinated up-regulation of the tra cluster together with reduced setR is consistent with cyromazine shifting the element toward an excision- and transfer-competent state.

To verify the role of the SOS response in cyromazine-mediated conjugative transfer, we examined the effect of cyromazine in the recA mutant strain (E. coli C600ΔrecA). As shown in Fig. 4F, cyromazine increased conjugative transfer of the Proteus-derived SXT ICE from E. coli C600 to E. coli J53, whereas no transconjugants were detected when the recA-deficient donor (E. coli C600ΔrecA) was used under the same conditions. A similar phenomenon was also observed for Shewanella- and Klebsiella-derived SXT ICEs (Fig. S4G, H). Consistently, cyromazine did not alter the expression of setR, setC, setD, xis and int in E. coli C600ΔrecA (Fig. 4G), indicating that the regulation of setR and subsequently the ensuing activation of SXT ICE transfer by cyromazine is RecA-dependent. We propose a potential mechanism whereby cyromazine induces intracellular ROS accumulation and triggers the SOS response. Activated RecA downregulates the setR, leading to the derepression of SXT ICE regulatory module and activation of conjugation-related genes. This regulatory cascade ultimately accelerates SXT ICE excision and conjugative transfer.

Cyromazine enhances proton motive force (PMF) and ATP synthesis

Bacterial cells rely heavily on their internal energy to perform a variety of physiological activities. Conjugation, an energy-dependent process, can be significantly enhanced by an increase in proton motive force (PMF)⁴². The PMF is partitioned between the transmembrane potential (ΔΨ) and the transmembrane proton gradient (ΔpH), with the dynamic balance and interconversion between these components helping maintain the stability of the PMF⁴³. Exposure to cyromazine at concentrations ranging from 0.0001 to 10 μg/mL resulted in reduced DiSC₃(5) fluorescence intensity and increased BCECF fluorescence intensity, indicating increases in ΔΨ and ΔpH and, consequently, in the PMF (Fig. 5A, B). Figure 5C, D show elevated intracellular levels of K⁺ and Ca²⁺, which may be attributed to the increased ΔΨ induced by cyromazine. Meanwhile, the addition of K⁺ and Ca²⁺ to the conjugation system remarkably increased conjugation frequency (Fig. 5E), but the addition of metal ions had no significant effect on SXT ICE excision (Fig. S5A).

Fig. 5. Cyromazine enhances transmembrane proton gradient, intracellular ATP levels, motility, and biofilm formation of bacteria.

A Effects of ΔΨ in the donor and recipient bacteria under cyromazine exposure for 1 h. B Effects of ΔpH in the donor and recipient bacteria after 1 h of cyromazine treatment. C Effects of cyromazine on intracellular K⁺ concentrations in the donor and recipient bacteria. D Effects of cyromazine on intracellular Ca²⁺ concentrations in the donor and recipient bacteria. E Fold changes in SXT ICE conjugation frequency under the addition of 10 mM K⁺ and Ca²⁺, respectively. F Fold changes in intracellular ATP levels of donor and recipient bacteria under cyromazine exposure. G Fold changes in the expression of genes related to ATP synthesis in donor bacteria under cyromazine treatment. H Effects of cyromazine on conjugative transfer of SXT ICE with the addition of uncoupling agent CCCP to the conjugative system. Conjugation frequencies were normalized to the mean value of the 0 µg/mL cyromazine without CCCP (control group). I Effect of CCCP supplementation on SXT excision. J,K Swimming motility of the donor and recipient bacteria after exposure to cyromazine for 48 h. L Changes in biofilm formation ability of donor and recipient bacteria under different concentrations of cyromazine addition. M Fold changes in the expression of pilus-related genes after cyromazine exposure. Data are presented as mean ± SD (n = 3 independent biological replicates per group). Statistical analyses were performed as follows: (A–D,F,K,L) two-way ANOVA with Dunnett’s multiple-comparisons test; (E,G,M) one-way ANOVA followed by Dunnett’s multiple-comparisons test. H two-way ANOVA with Dunnett’s and Šidák’s multiple-comparisons post hoc tests; I one-way ANOVA with Tukey’s multiple comparisons test. All tests were two-sided. Significance is indicated as *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001; adjusted P values are provided in the Source Data. Source data are provided as a Source Data file.

Bacteria harness the energy stored in the PMF to drive ATP synthesis and energy-dependent processes, thereby facilitating conjugation. As shown in Fig. 5F, G, the intracellular ATP levels increased and the expression of ATP-related genes (atpA, atpB, atpC, atpD, atpE, and atpG) was up-regulated under cyromazine exposure. Carbonyl cyanide 3-chlorophenyl-hydrazone (CCCP), an uncoupler of oxidative phosphorylation, inhibits energy production by causing the dissipation of ΔpH. As shown in Fig. 5H, the addition of CCCP significantly reduced the conjugation frequency. Additionally, CCCP treatment also significantly decreased the SXT ICE excision frequency (Fig. 5I). A similar trend was noted for the abundance of circular SXT ICE (Fig. S5B). These findings suggest that the energy supply may play an important role in the efficiency of SXT ICE excision and transfer.

Energy stored in the PMF powers flagellar rotation for motility and supports biofilm formation, both of which are linked to conjugation⁴³. An expansion of the swimming zone was observed in the swimming assay under cyromazine treatment (Fig. 5J, K). Crystal violet staining revealed that the biofilm formation capacity was enhanced after cyromazine treatment (Fig. 5L). Consistently, the expression of pilus-related genes (fimA, fimC, fimD, fimG, fimI, fimH, and yehB) in donor bacteria was up-regulated (Fig. 5M). As reported previously for plasmid systems, cyromazine-promoted motility and biofilm formation act as host-level facilitators of conjugation³⁸.

Cyromazine interferes with bacterial membrane homeostasis

The cell membrane serves as both a physical and functional barrier during conjugation. Growing evidence suggests that changes in the fluidity, permeability and integrity of the cell membrane significantly influence conjugative transfer. Membrane fluidity was assessed utilizing Laurdan GP probes, revealing that cyromazine decreased membrane fluidity in both donor and recipient bacteria at concentrations ranging from 0.001 to 10 μg/mL (Fig. 6A). Meanwhile, an increase in lipopolysaccharide (LPS) induced by cyromazine was observed (Fig. 6B), which potentially further reduces the membrane fluidity. Similarly, the expression of LPS-related genes (waaG, waaJ, waaH, lptA, lptB, lptC, and lptD) was also up-regulated (Fig. S6). In addition, inner-membrane permeability and outer-membrane permeability were assessed utilizing propidium iodide (PI) and N-phenylnaphthalen-1-amine (NPN), respectively. The results show that both PI and NPN fluorescence intensities were increased, indicating that cyromazine enhanced membrane permeability (Fig. 6C, D). Meanwhile, the expression of omp genes related to outer membrane permeability, including pore-forming ompACFGRW and outer membrane channel protein tolC, was also upregulated after cyromazine treatment (Fig. 6E). Furthermore, the accumulation of intracellular ROS could further disrupt membrane homeostasis by affecting membrane permeability. Scanning electron microscopy (SEM) was employed to examine the morphological changes in donor bacteria under 0.1 μg/mL and 1 μg/mL cyromazine exposure. The findings demonstrated that the cyromazine-treated bacteria exhibited a porous and wrinkled surface structure, in contrast to the smooth surface observed in the control group (Fig. 6F). Taken together, the observed reduction in membrane fluidity, increase in permeability, and membrane damage caused by cyromazine may collectively create a permissive cellular state, facilitating SXT ICE conjugative transfer.

Fig. 6. Cyromazine disrupts bacterial membrane homeostasis.

A Changes in membrane fluidity of donor and recipient bacteria after cyromazine treatment. B Effects of LPS levels in the donor and recipient bacteria after exposure to cyromazine. C,D Fold changes in inner-membrane permeability (C) and outer-membrane permeability (D) of donor and recipient bacteria induced by cyromazine. E Fold changes in the expression of outer membrane-related genes in donor bacteria after cyromazine treatment. F Morphological changes of donor bacteria using scanning electron microscopy (SEM) analysis. The area circled in red indicates obvious membrane damage. The experiment was independently repeated three times with similar observations, and representative images are shown. Data are presented as mean ± SD (n = 3 independent biological replicates per group). Statistical analyses were performed as follows: (A–D) two-way ANOVA with Dunnett’s multiple-comparisons test; (E) one-way ANOVA followed by Dunnett’s multiple-comparisons test. Significance is indicated as *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001; adjusted P values are provided in the Source Data. Source data are provided as a Source Data file.

Discussion

The persistent misuse of antibiotics in clinical and agricultural settings has accelerated the spread of AMR, enabling the selection of multidrug-resistant (MDR) pathogens and posing a major global health and economic challenge^44–48. MGE-mediated conjugation is the primary mechanism underlying the horizontal transfer of ARGs, with both antibiotic and non-antibiotic compounds serving as potential amplifiers of this process^16,49. Previous studies have primarily focused on plasmid-mediated conjugation, demonstrating that antibiotic and non-antibiotic stress accelerates intra- and inter-species transfer of ARGs, thereby exacerbating AMR prevalence^14,38,49. However, research on ICE-mediated mechanisms remains limited, particularly regarding SXT ICEs, which pose significant threats to human health due to their carriage of MDR genes and broad host range. Therefore, the SXT ICE-conjugation model was established to better study MGE-mediated HGT.

The accumulation of cyromazine in animal feces is primarily driven by its widespread application in livestock production, posing dual risks of environmental contamination and farmland pollution through organic fertilizer use. Previous studies have detected cyromazine residues in animal wastewater, surface water surrounding farms, farmed animal organic fertilizers, and manure-amended soils, with high detection rates^34–36. Meanwhile, organic fertilizer application has been shown to significantly elevate the abundance of ARGs in soil through both the enrichment of indigenous ARGs and the introduction of exogenous ARGs, with a strong positive correlation observed between the abundance of MGEs and ARGs⁵⁰. In addition, cyromazine has been associated with increased abundance of ARGs and metal-resistance genes during aerobic composting³⁷. Building upon previous studies that focused on plasmid-mediated conjugation, we extended our investigation to the SXT ICE to examine whether cyromazine exerts a similar promoting effect³⁸. This extension provides a more comprehensive assessment of the risks associated with cyromazine in facilitating ARG dissemination.

Cyromazine facilitates SXT ICE conjugative transfer both intra- and inter-species at the tested sub-inhibitory (sub-MIC) concentrations (0.001–10 μg/mL), with similar trends observed across multiple wild-type strains. Cyromazine had no significant effect on conjugation frequency after 4 h, although it promoted SXT ICE excision at 10 μg/mL. This suggests that conjugative transfer is a complex process, involving multiple steps beyond excision, such as conjugative pilus formation, DNA processing, and recipient cell integration, which may require more time for cyromazine to fully influence. Elevated temperature (42 °C) or alkaline pH (pH of 8) substantially reduced this effect, and high cyromazine concentrations even abolished conjugation under alkaline conditions. These results imply that aerobic composting or alkaline treatment of manure using sodium hydroxide or lime could mitigate cyromazine-driven dissemination of ARGs. Moreover, cyromazine-enhanced SXT ICE transfer was observed in vivo at environmentally relevant doses (10, 100, and 1000 μg/kg), suggesting a potential risk in livestock-associated environments. Nevertheless, the present study did not quantify cyromazine concentrations in mouse feces and gut after treatment, nor did it employ native intestinal microbiota for conjugation assays. These omissions limit our ability to assess the relationship between realistic intestinal exposure levels and conjugation efficiency, and to evaluate the broader impacts of cyromazine on gut microbial ecology. Future studies will quantify cyromazine concentrations in both feces and gut, and integrate native intestinal microbiota into conjugation assays to provide a more nuanced understanding of its effects on microbial community dynamics and SXT ICE-mediated gene transfer.

Cyromazine promotes the dissemination of SXT ICE and selectively reshapes transconjugant communities in an environment-dependent manner. The enrichment of specific Gammaproteobacterial clades suggests selection for lineages with inherent potential for SXT-ICE acquisition. However, two important limitations must be acknowledged. First, the use of enriched, culturable recipient communities may not fully capture the complexity of in situ microbiomes, especially in aquatic environments. Second, by focusing solely on transconjugant diversity, without parallel analysis of recipient community dynamics, we are unable to fully separate the direct selective effects of cyromazine on native microbiota from its impact on conjugation efficiency. Future studies incorporating metagenomic tracking of donor, recipient, and transconjugant populations will be necessary to resolve these coupled ecological and genetic processes. Together, these findings underscore the potential of cyromazine to modulate HGT outcomes and reshape patterns of resistance gene dissemination in environmental microbial reservoirs.

Mechanistically, cyromazine exposure markedly promoted SXT ICE excision, indicating a shift of the element toward a transfer-permissive state. This effect was closely associated with the SOS response, as evidenced by the up-regulation of recA, sulA, and related repair genes, together with the suppression of lexA. The ROS scavenger NAC largely abolished cyromazine-induced increases in excision and conjugation frequencies, underscoring the important role of oxidative stress. The SOS system has been recognized as a key regulator of SXT ICE mobility, where RecA-mediated autocleavage of the SetR repressor derepresses the setC/setD module, thereby activating int and tra gene clusters required for excision and transfer^27,41. Consistent with this regulatory paradigm, cyromazine treatment upregulated recA, int, setC, setD, and tra gene clusters while downregulating setR, thereby shifting the SXT ICE element toward a state more permissive to excision and transfer. Notably, in the recA-deficient mutant, cyromazine failed to influence setR, setC, setD, xis and int expression, and no transconjugants were observed. Previous studies have shown that recA mutant prevents SXT transfer induced by mitomycin C, which is consistent with our results, further confirming the essential role of RecA in mediating SXT ICE conjugative transfer²⁷. Collectively, cyromazine induces ROS accumulation and activates the SOS response, leading to RecA-mediated derepression of the SXT ICE regulatory module and subsequent activation of excision and transfer pathways.

Energy metabolism is a recognized determinant of conjugation efficiency in plasmid systems, and consistent with previous reports, our findings indicate that this principle extends to SXT ICEs^38,51,52. Cyromazine exposure elevated both ΔΨ and ΔpH, leading to an increased PMF and higher ATP levels. Although the addition of K⁺ and Ca²⁺ had no significant effect on SXT ICE excision, the rise in intracellular K⁺ and Ca²⁺ may further stabilize membrane potential and promote transfer efficiency. These energetic changes collectively sustain electrochemical gradients that favor pilus assembly and ion transport. Furthermore, cyromazine-treated bacteria exhibited enhanced motility and biofilm formation, both of which serve as host-level facilitators that may strengthen intercellular contact and improve mating-pair formation during conjugation^53–55. Notably, energy availability plays a functional role in SXT ICE excision and transfer, highlighting the coupling between bacterial energy supply and HGT.

Membrane homeostasis is also considered a crucial factor influencing conjugation efficiency. Indeed, cyromazine increased membrane permeability and led to wrinkled, porous bacterial cell surfaces, as observed experimentally. Consistent with prior findings, membrane damage and increased permeability caused by antibiotic and non-antibiotic stressors (including cyromazine) have been associated with higher conjugation frequencies^38,54,56. In contrast to reports in which reduced LPS content and increased membrane fluidity facilitate conjugation, cyromazine exposure increased LPS abundance and decreased membrane fluidity^57–59. The resulting membrane rigidity may increase susceptibility to local damage and permeability, which, together with LPS remodeling and porin upregulation, could favor donor-recipient contact and facilitate SXT ICE transfer. Together, these data suggest that cyromazine compromises membrane integrity and homeostasis, potentially facilitating SXT ICE conjugative transfer by enhancing intercellular exchange interfaces.

In conclusion, cyromazine accelerates the spread of ARGs by stimulating high-frequency SXT ICE conjugation both intra- and inter-species. This promotive effect was consistently observed not only in vivo but also in environmentally relevant models, including feces, soil, and water. Given the widespread use of cyromazine in food animals and its predominant excretion in an unmetabolized form, its potential to drive ARG dissemination in both host-associated and environmental reservoirs merits careful re-evaluation. A plausible mechanism is that cyromazine induces ROS accumulation and activates the SOS response, which may promote SXT ICE excision and upregulate conjugation-related operons. In parallel, cyromazine-driven alterations in energy metabolism and membrane homeostasis may further accelerate the conjugation process (Fig. 7). To mitigate the ongoing dissemination of AMR, it is essential to implement source control measures, such as the rational use of cyromazine in the breeding industry, alongside terminal interventions like composting or alkaline treatment of manure.

Fig. 7. Proposed model illustrating the potential mechanisms by which cyromazine stimulates high-frequency SXT ICE conjugation.

Cyromazine enhances SXT ICE conjugative transfer both in vivo and in vitro, and also facilitates transfer to environmentally relevant microbiota (A). The potential mechanism involves cyromazine increasing intracellular ROS accumulation, intensifying bacterial oxidative stress, and subsequently triggering the SOS response, which promotes SXT ICE excision and activates conjugation-related operons (B). The coordinated regulation of energy supply, bacterial motility, and biofilm formation could facilitate SXT ICE transfer (C). Additionally, disruption of membrane homeostasis could further facilitate this process.

Methods

Bacteria strains, antibiotic and culture reagents

The sulfamethoxazole (SMX)- and trimethoprim (TMP)-resistant Escherichia coli J53 carrying SXT ICE served as the donor strain, while the rifampicin (RIF)-resistant E. coli C600 was set as the recipient strain. Three wild-type strains (Proteus, Klebsiella, and Shewanella) were also used as donor strains. Proteus and Klebsiella strains were isolated from chicken farm feces collected in Henan, China, whereas the Shewanella strain was isolated from aquatic products (Ruditapes philippinarum) collected in Jiangsu, China. Furthermore, E. coli C600, its recA-deficient mutant (E. coli C600ΔrecA), and E. coli X7213 (all carrying SXT ICE) were included as additional donor strains. All strains were preserved in our laboratory. SMX, TMP and RIF were purchased from the China Institute of Veterinary Drug Control. Cyromazine was obtained from Yuanye (Shanghai, China). The LB broth and nutrient agar were acquired from Qingdao Hope Bio-Technology (Shandong, China).

Minimum inhibitory concentration (MIC) and bacterial growth kinetics determination

Donor and recipient bacteria were grown in LB broth with shaking at 37 °C to an OD₆₀₀ of 0.5, diluted 1:1000 in fresh LB medium, and then supplemented with cyromazine at final concentrations of 0, 0.001, 0.01, 0.1, 1, and 10 μg/mL. The cultures were incubated at 37 °C, and the absorbance at 600 nm was measured hourly for 12 h to monitor their growth curves. Furthermore, the EUCAST clinical breakpoint method was utilized to determine the MIC of donor, recipient, and transconjugants to antibiotics.

Conjugation experiments

Conjugation experiments between E. coli J53 carrying SXT ICE (donor) and E. coli C600 (recipient) were performed with or without cyromazine. Bacteria were cultured in LB broth with shaking at 37 °C to an OD₆₀₀ of 0.5. Following centrifugation, the bacterial precipitates were resuspended in fresh LB broth, and the donor and recipient bacteria were mixed at 1:1. Subsequently, cyromazine was added at concentrations of 0, 0.001, 0.01, 0.1, 1, and 10 μg/mL, and the cultures were incubated at 37 °C for 14 h. Transconjugants were screened on agar plates containing 76 μg/mL SMX, 4 μg/mL TMP, and 300 μg/mL RIF, whereas recipient bacteria were selected on agar plates with 300 μg/mL RIF. Conjugation experiments were also conducted using three wild-type strains (Proteus, Klebsiella strains, and Shewanella), as well as E. coli C600 and its recA-deficient mutant (E. coli C600ΔrecA) following the same procedure. Additionally, SXT ICE in E. coli J53 was originally acquired by conjugative transfer from a wild-type Proteus isolate. The conjugation frequency was calculated as the ratio of transconjugants to recipient bacteria. The fold change in conjugation frequency was obtained by dividing each replicate value by the mean value of the control group (0 µg/mL). This approach was applied consistently across all groups, including the control. For all subsequent calculations of conjugation frequency changes, the same method was applied.

Donors, recipients, and transconjugants were randomly selected from the agar plates, and integrase (int) on SXT ICEs was identified by PCR and agarose gel electrophoresis using the following primers (5′–3′): forward, CTGTGGCCAATCATCAACTC; reverse, CGACCGAGATGGGCTAAGTG.

A one-variable-at-a-time experimental design was adopted to assess the effects of temperature, pH, and mating duration on SXT ICE conjugative transfer under cyromazine exposure. Temperature (4 °C, 24 °C, 37 °C, and 42 °C) was varied at constant pH 7 and a mating time of 14 h; pH (5, 6, and 7) was varied at 37 °C for 14 h; and mating duration (4 h, 14 h, and 24 h) was evaluated at 37 °C and pH 7.

In order to further explore the mechanism, the ROS scavenger N-acetylcysteine (NAC) (5 mM) and the uncoupling inhibitor Carbonyl Cyanide3-ChloroPhenylhydrazone (CCCP) (50 μM) were added to the mating system. Moreover, 10 mM K⁺ and 10 mM Ca²⁺ were added to the conjugative system to assess the effect of K⁺ and Ca²⁺ on conjugative transfer, respectively.

Intracellular ROS determination

Intracellular ROS levels were detected by ROS Detection Assay Kit (Beyotime). Following overnight incubation to an OD₆₀₀ of 0.5, the bacterial cells were collected by centrifugation, washed twice, and resuspended in phosphate-buffered saline (PBS) to an OD₆₀₀ of 0.5. The suspension was then incubated with 10 μM 2′, 7′-dichlorodihydrofluorescein diacetate (DCFH-DA) for 30 min at 37 °C in darkness, after which the unbound probe was remove by washing twice with PBS. Subsequently, 190 μL probed bacterial suspension and 10 μL of cyromazine were introduced into the 96-well black plate (final concentrations: 0, 0.001, 0.01, 0.1, 1, and 10 μg/mL) and incubated at 37 °C for 1 h. PBS and 10 μL ROSup were used as negative and positive controls, respectively. The fluorescence intensity was measured using the Infinite M200 Microplate reader (Tecan) at an excitation wavelength of 488 nm and emission wavelength of 525 nm.

Swimming motility assay

The motility assay was conducted on LB agar plates containing 0.3% agar, supplemented with cyromazine at concentrations of 0, 0.001, 0.01, 0.1, 1 and 10 μg/mL. Bacteria were cultured with shaking at 37 °C to OD₆₀₀ of 0.5, and then resuspended in PBS. 2 μL bacterial suspension was inoculated at the center of the agar plate and incubated for 48 h. The swimming zones were subsequently photographed and measured.

Membrane fluidity, membrane permeability, membrane potential (∆Ψ), and transmembrane proton gradient (∆pH) detection

Membrane fluidity was assessed using Laurdan, membrane permeability was evaluated with propidium iodide (PI) and N-phenyl-1-naphthylamine (NPN), membrane potential (∆Ψ) was evaluated using the 3,3-dipropylthiadicarbocyanine iodide (DiSC₃(5)), and the transmembrane proton gradient (∆pH) was measured using BCECF-AM. Bacteria were cultivated to an OD₆₀₀ of 0.5, and then suspended in PBS with the final concentration of 10 μM Laurdan, 0.5 μM PI,10 μM NPN, 0.5 μM DiSC₃(5) or 10 μM BCECF-AM in the dark for 30 min. Subsequently, 10 μL cyromazine was added to 190 μL treated bacterial solution in 96-well black plates, yielding final concentrations of 0, 0.001, 0.01, 0.1, 1, and 10 µg/mL, and incubated for 1 h under dark conditions. The fluorescence intensity was measured using an Infinite E Plex Microplate reader (Tecan) with excitation/emission wavelengths of 435 nm/490 nm for Laurdan, 535 nm/615 nm for PI, 350 nm/420 nm for NPN, 622 nm/670 nm for DiSC₃(5), and 488 nm/535 nm for BCECF-AM. The Laurdan generalized polarization (GP) value was calculated using the formula: GP = (I₄₃₅ - I₄₉₀) / (I₄₃₅ + I₄₉₀).

Intracellular ions concentration measurement

Intracellular K⁺ and Ca²⁺ levels were determined by PBFI-AM and Fura-2 AM (Maokangbio), respectively. Bacteria were grown in LB broth at 37 °C to an OD₆₀₀ of 0.5, resuspended in PBS, and then incubated with 10 μM fluorescence probe for 30 min. The fluorescence intensity was determined after 1 h of interaction with cyromazine at final concentrations of 0, 0.001, 0.01, 0.1, 1, and 10 μg/mL. The excitation and emission wavelengths were 380/550 nm for PBFI-AM and 340/510 nm for Fura-2 AM.

ATP and LPS level determination

Intracellular adenosine triphosphate (ATP) levels were quantified using the Enhanced ATP Assay Kit (Beyotime). Bacteria were cultured to an OD₆₀₀ of 0.5, resuspended in PBS, and exposed to cyromazine (0, 0.001, 0.01, 0.1, 1, and 10 μg/mL) for 4 h. Subsequent to treatment, the bacterial suspension was subjected to centrifuge, lysed, and mixed with the prepared ATP test reagent. ATP levels were measured using a microplate reader (Tecan) in luminescence mode. Intracellular lipopolysaccharide (LPS) was extracted using the LPS Extraction Kit (iNtRON Biotechnology) and quantified with the Chromogenic Limulus Amebocyte Lysate (LAL) Endotoxin Assay Kit (Beyotime, China) following the manufacturer’s protocol. Absorbance was measured at 545 nm using the microplate reader.

The surface morphology was examined using scanning electron microscopy (SEM)

Bacteria were cultivated until reaching an OD₆₀₀ of 0.5, suspended in PBS, and exposed to cyromazine (0, 0.1, and1 μg/mL) for 4 h. Following treatment, the bacteria were washed twice with PBS and fixed overnight at 4 °C with 2.5% glutaraldehyde. After fixation, the samples underwent sequential dehydration using increasing concentrations of ethanol, followed by drying and gold sputter coating. The prepared specimens were then examined with a Gemini SEM 300 (ZEISS) at 40k magnification.

DNA extraction, RNA extraction, and RT-qPCR analysis

After 4 h and 14 h of cyromazine treatment of bacteria, the genomic DNA and bacterial RNA were extracted using the TIANamp Bacteria DNA Kit (Tiangen) and RNA-easy Isolation Reagent (Vazyme), respectively. DNA and RNA quality were assessed by the 260 nm/280 nm absorbance ratio using the Nanodrop spectrophotometer (Thermo). The genomic DNA was diluted to 50 ng/μL. For bacterial RNA, reverse transcription was performed with the HiScript® III RT SuperMix for qPCR (+gDNA wiper) Kits (Vazyme) to obtain cDNA. RT-qPCR was carried out on a 7500 Fast Real-Time PCR System (Applied Biosystems) with ChamQ™ Universal SYBR® Color qPCR Master Mix Kits (Vazyme) and gene-specific primers. The thermal cycling conditions were: 95 °C for 30 s, followed by 40 cycles of 95 °C for 10 s, 60 °C for 30 s, and 72 °C for 30 s. Primer sequences are listed in Supplementary Data 1.

In vivo conjugation assay

For the mouse intraperitoneal conjugation model, the donor (E. coli J53 carrying SXT ICE) and recipient bacteria (E. coli C600) were cultured overnight, and resuspended to OD₆₀₀ of 0.5 (5×10⁸ CFU/mL). A 200 μL mixture of donor and recipient bacteria (1:1) was then injected intraperitoneally. 15 min post-infection, mice received intraperitoneal injections of cyromazine (10, 100, and 1000 μg/kg) or an equivalent volume of PBS. The liver, spleen, and kidney, collected after 24 h, were homogenized in PBS and subjected to serial dilution. RIF agar plates containing 76 mg/L SMX and 4 mg/L TMP were used to select transconjugants, and RIF-only agar plates were used to determine recipient counts. The conjugative transfer frequency was calculated by comparing the CFU counts of transconjugants to recipient bacteria.

For the mouse intestinal conjugation model, mice were pretreated with drinking water containing streptomycin (1.0 g/L), ampicillin (0.5 g/L) and doxycycline (0.5 g/L) for 7 days, followed by 24 h of antibiotic-free water to deplete gut commensal bacteria and reduce colonization resistance. The overnight cultures of donor (E. coli J53 carrying SXT ICE) and recipient bacteria (E. coli C600) were resuspended to an OD₆₀₀ of 0.5 (5×10^8 CFU/mL), and 200 μL of the 1:1 bacterial mixture was administered orally. Fifteen minutes post-infection, mice were orally administered cyromazine (10, 100, and 1000 μg/kg) or an equivalent volume of PBS. Fresh feces were collected, homogenized, diluted, and spread on agar plates 24 h post-infection. Conjugative transfer frequency was assessed by counting the colonies after 14 h of incubation.

Establishment of feces, soil, and water models and microbial diversity analysis

Soil samples were collected from a lawn at a depth of 10–15 cm, and fresh fecal samples were obtained from healthy adult rats. Soil and fecal samples were suspended in PBS, vortexed, centrifuged at low speed, and filtered through a 200-mesh sieve. The resulting supernatants were used as recipient microbial communities. For the water model, pond water samples were filtered through a 200-mesh sieve, centrifuged, resuspended in LB broth, and incubated at 37 °C for 6 h prior to use as the recipient community. For conjugation assays, donor bacteria (E. coli X7213 harboring SXT ICE) and recipient communities were mixed at a 1:1 ratio and subjected to conjugative transfer experiments. Subsequently, cyromazine was added at concentrations of 0, 0.001, 0.01, 0.1, 1, and 10 μg/mL, and the cultures were incubated at 37 °C for 14 h. Transconjugants were screened on LB agar plates containing 76 μg/mL SMX, 4 μg/mL TMP, whereas recipient bacteria were selected on LB agar plates.

Following conjugation experiments in the feces, soil, and water models, the resulting transconjugant communities were collected for microbial diversity analysis. Total community DNA was extracted using a standard CTAB/SDS method and quantified for quality control. The V4 region of the 16S rRNA gene was amplified using barcoded primers (515 F/806 R), and PCR products were purified, pooled in equimolar amounts, and used for library construction. Sequencing was performed on an Illumina platform to generate paired-end reads. Raw sequences were quality-filtered, merged, and chimera-checked to obtain effective tags, which were subsequently clustered into amplicon sequence variants (ASVs) or operational taxonomic units (OTUs) for taxonomic annotation. Alpha- and beta-diversity analyses were conducted to assess changes in microbial community structure among different treatment groups.

Statistical analyses

Statistical analyses were performed via GraphPad Prism 9. Data are presented as mean ± SD from at least three independent biological replicates (exact n is provided in the corresponding figure legends). For experiments with one factor and ≥3 groups, one-way ANOVA was performed followed by Dunnett’s multiple-comparisons test (each treatment vs control) or Tukey’s test (all pairwise comparisons), as specified in the figure legends. For experiments with two factors, two-way ANOVA was used with appropriate post hoc multiple-comparisons tests (e.g., Dunnett’s and/or Šidák’s), as specified. All tests were two-sided. Adjusted P values are reported in the Source Data. P < 0.05 was deemed significant, with significance indicated as: *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Ethical statement and animal studies

Female ICR mice (6–8 weeks old) were used for all mouse experiments. All animal procedures were performed in accordance with the guidelines of the Jiangsu Administrative Committee of Laboratory Animals (SYXK-2022-0044) and were approved by the Animal Care Committee of Yangzhou University. Mice were housed in individually ventilated cages at 22  °C ±  2 °C and 50  ±  10% relative humidity under a 14 h light/10 h dark cycle.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Author contributions

C.W.: Methodology, Software, Validation, Writing—original draft; P.W.: Methodology, Software, Validation; K.P.: Methodology, Validation, Conceptualization; H.Z.: Methodology, Conceptualization; W.Z.: Data curation, Conceptualization; Y.L.: Methodology, Conceptualization; Z.W.: Supervision, Funding acquisition; R.L.: Conceptualization, Supervision, Writing—review & editing, Funding acquisition.

Peer review

Peer review information

Nature Communications thanks Bruno Gonzalez-Zorn, and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Data availability

The 16S rRNA sequencing data have been deposited in the NCBI Sequence Read Archive (SRA) under BioProject accession code PRJNA1415235. Source Data are provided with this paper, and all other supporting data are available within the Article and Supplementary Information files. Source data are provided with this paper.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-026-71554-1.