Integrons (Stokes and Hall, 1989; Akrami et al., 2019; Ghaly et al., 2022) are natural cloning and expression vectors that are common multifunctional gene capture systems in bacterial genomes that enable bacteria to generate phenotypic diversity and form adaptive responses (Partridge et al., 2018; Ahmadian et al., 2020; Ali et al., 2020; Bhat et al., 2023). This capability plays a vital role in the emergence and spread of antibiotic resistance. Class 1 integrons are most commonly seen in clinical settings, followed by class 2 integrons (Jové et al., 2017; Ahmadian et al., 2020). The majority of class 2 integrons currently isolated from clinical samples are non-functional (Jiang et al., 2023), their integrases cannot translate the complete integrase protein, rendering them unable to integrate and excise antibiotic resistance gene cassettes. Functional class 2 integrons (Márquez et al., 2008; Wei et al., 2014; Lu et al., 2022) are a unique subset characterized by a mutation in the integrase gene that replaces a stop codon (TAA), inserted after the 178th amino acid, with a codon for an active amino acid. This mutation enables the translation and synthesis of a complete integrase protein, facilitating the integration and excision of resistance gene cassettes. Functional class 2 integrons are extremely rare in clinical environments (Lu et al., 2022). Gene cassettes carried by integrons have evolved antimicrobial resistance in key clinical environments (Bhat et al., 2023). Class 2 integrons typically harbor three resistance gene cassettes: dfrA1, sat2, and aadA1 (Jové et al., 2017). The dfrA1 gene encodes dihydrofolate reductase (DHFR), which mediates resistance to trimethoprim. The sat2 gene encodes streptothricin acetyltransferase, which confers resistance by transferring an adenylyl group to the streptothricin molecule, thereby inactivating it. The aadA1 gene encodes aminoglycoside adenylyltransferase, which modifies specific sites on aminoglycoside antibiotics (such as streptomycin and spectinomycin), preventing their binding to bacterial ribosomes.

Studies on class 1 integrons have revealed that various drugs can trigger the SOS response, which collaborates with host regulatory networks to control the expression of integrases and resistance gene cassettes. This mechanism ultimately transforms integrons into a genetically stable and efficient system, endowing bacteria with a low-cost, highly adaptable evolutionary potential (Ahmadian et al., 2020; Mohanraj and Mandal, 2022). However, current research on class 2 integrons mainly focuses on their clinical distribution and the types and structures of gene cassettes in the variable region (Jové et al., 2017; Alonso et al., 2018). Reports on functional class 2 integrons are rare. The mechanisms underlying the TAA to CAA conversion in the integrase gene, the expression status and regulatory mechanisms of functional class 2 integrase, and its relationship with variable region gene cassettes remain poorly understood. Therefore, this study aims to investigate the mutation and expression regulation of class 2 integron integrase genes under the induction of common drugs. Furthermore, the research explores the role of these drugs in modulating the mutation and expression of both class 2 integron and functional class 2 integron integrase genes and examines their correlation with gene cassettes in the variable region.

Bacteria develop resistance under the pressure of antibiotics. The emergence and widespread clinical dissemination of resistant strains have become a significant challenge to public health (Wistrand-Yuen et al., 2018). Class 2 integrons typically harbor three resistance gene cassettes: dfrA1, sat2, and aadA1, which confer resistance to trimethoprim, streptothricin, and aminoglycosides, respectively (Jové et al., 2017). The intI2 integrase of non-functional class 2 integrons contains a stop codon, TAA, inserted after the 178th amino acid, rendering this class of integrases incapable of integrating and excising resistance gene cassettes, resulting in a truncated integrase protein. In contrast, functional class 2 integrons possess a stop codon mutation from TAA to CAA in the integrase gene, enabling the production of an active integrase, which facilitates the acquisition or rearrangement of gene cassettes.

Bacteria exposed to antibiotics can induce the SOS response (Sheng et al., 2021; Van Alstine and Sandler, 2022), which serves to maintain genomic integrity and enhance survival chances (Blázquez et al., 2018). Numerous studies on class 1 integrons have shown that fluoroquinolones, trimethoprim, and β-lactams can trigger the SOS response, activating horizontal gene transfer elements, including integrons (Ahmadian et al., 2020), thereby facilitating the spread of antibiotic resistance genes (Phillips et al., 1987; Goerke et al., 2006; Maiques et al., 2006; Mohanraj and Mandal, 2022). Drug concentrations below the minimum inhibitory concentration (MIC) are referred to as sub-inhibitory concentrations (sub-MIC) (Davies et al., 2006; Gu et al., 2020). The widespread clinical use of antibiotics (Patel et al., 2019; Singh and Bhunia, 2019; Polianciuc et al., 2020) as well as their improper use in agriculture and animal husbandry (Shen et al., 2008; Liu et al., 2023), results in low-level antibiotic exposures (sub-MIC) that are insufficient to kill bacteria but can induce changes in gene expression, leading to phenotypic changes (Sato et al., 2018; Davarzani et al., 2021) and increased mutation rates (Drlica, 2003; Drlica and Zhao, 2007; Gullberg et al., 2011; Barrett et al., 2019), thereby enhancing the potential for adaptive genetic variations. Studies have indicated that sub-MIC levels of antimicrobial agents may be more significant in contributing to bacterial resistance than lethal concentrations (Andersson and Hughes, 2012; Hughes and Andersson, 2012; Laureti et al., 2013). Class 2 integrons, as key mobile elements carrying resistance gene cassettes, are critically linked to the dissemination of resistance genes. This study systematically explores the dynamic relationship between SOS stress response, integron expression, and resistance phenotypes by inducing constructed strains (pAC-FA1, pAC-FA3, pAC-FA6) and clinical isolates (5b, 7b) with functional or non-functional class 2 integrons using two sub-MIC concentrations of antibiotics and mutagens (AMP, CIP, KAN, MMC). The findings provide experimental evidence for understanding the mechanisms of resistance evolution under low-level antibiotic exposure in vitro.

The SOS response plays a critical role in protecting bacterial communities from extreme environmental conditions, mediating survival and adaptation to genotoxic stress (Miura et al., 2020; Niccum et al., 2020), and contributing to the development of antibiotic resistance (Maslowska et al., 2019; Podlesek and Žgur Bertok, 2020; Xu et al., 2024). This response is a widespread regulatory network controlled by various regulatory proteins and enzymes and is highly conserved across different pathogens. Among these, RecA and LexA are essential proteins required universally for DNA damage-mediated SOS repair in bacteria. Even under non-inducing conditions, RecA is expressed at basal levels, but in induced cells, RecA rapidly binds to ssDNA and facilitates the degradation of LexA, ensuring a swift and effective SOS response (Ghodke et al., 2019). The expression level of recA is a key indicator of SOS response activation. In this study, the SOS response was monitored by measuring the relative expression levels of recA using real-time quantitative PCR. Our experiments revealed that the relative expression of recA in experimental strains increased to varying degrees following induction with various drugs, exhibiting both concentration and time dependency. Drug concentrations at 1/2 MIC were more effective in elevating recA expression than 1/4 MIC concentrations. Comparing days 1, 3, and 8 of induction showed that recA expression significantly increased after just 1 day, peaked on the third day, and slightly decreased by the eighth day. This dynamic change suggests that bacteria can quickly mobilize related genes to trigger the SOS response when faced with adverse conditions, aiding in their survival until they adapt to the stressor. There were notable differences in how different strains responded to the drugs: after CIP induction, pAC-FA1 exhibited the greatest increase in recA expression. In pAC-FA3, the effects of the drugs were relatively similar, while in pAC-FA6, KAN had the most notable effect. In strains 5b and 7b, MMC showed the strongest induction effect. These findings demonstrate that all four antibiotics tested in this experiment can enhance recA expression, thereby triggering the bacterial SOS response to combat adverse conditions. However, the tolerance of each strain to different drugs varies, and even at the same concentration, different levels of SOS response may be triggered to counteract these stressors. Additionally, although the relative expression of recA was highest in pAC-FA6, the change in resistance after 8 days of continuous antibiotic induction was not the most significant. The high heterogeneity within bacterial populations (Orman and Brynildsen, 2013), coupled with potential indirect feedback inhibition and differences in regulatory proteins due to different antibiotic mechanisms (Schaenzer and Wright, 2020; Wilson et al., 2020; Huang et al., 2022; Baranova et al., 2023), may collectively contribute to a non-linear relationship between recA levels and resistance changes, underscoring the need for further research in this area. Under identical induction conditions, differences in the relative expression levels of recA were observed between the constructed strains and clinical isolates, with the constructed strains showing a more pronounced increase. The survival mechanisms of cellular stress responses are closely linked to bacterial metabolism (Recacha et al., 2024). We hypothesize that E. coli TOP10, being a genetically engineered strain, has well-developed regulatory systems such as cAMP-CRP (Heyde et al., 2021) and LexA (Cory et al., 2024). Engineered expression vectors often contain strong promoters and terminators, and the strain's faster growth rate can support the higher energy demands of foreign protein expression. In contrast, the natural genome of Proteus mirabilis may have target genes regulated by upstream inhibitory elements, leading to slower growth and a greater allocation of energy to basic metabolism rather than foreign gene expression.

After drug induction, the relative expression levels of class 2 integron integrase and resistance genes in the variable region of experimental strains increased to varying degrees. In the constructed strains pAC-FA1 and pAC-FA3, CIP exhibited the strongest induction effect compared to AMP and KAN, generally showing an initial increase with a peak on day 3, followed by a decline. In pAC-FA6, the induction effects of all four drugs were relatively similar, although KAN specifically led to a noticeable rise in the expression levels of sat2 and aadA1. In clinical isolates, 5b and 7b, MMC strongly induced expression of the class 2 integron integrase gene, whereas the other three drugs did not demonstrate significant differences in their effects. It can be observed that non-functional intI2, although unable to modify its gene cassettes array, can still effectively express the gene cassettes (Jové et al., 2017). The expression of integrases must be regulated, “turning on” and “off” as required to obtain the functional diversity encoded by gene cassettes. From an evolutionary perspective, the ability of intI2 to upregulate transcription in strains with non-functional integrases may serve as a stress reserve (Escudero et al., 2015). Moreover, even among constructed strains, the same drug can exert substantially different effects. It is hypothesized that the presence of non-functional class 2 integrons, functional class 2 integrons, and even the number of resistance genes linked to class 2 integrons might influence the adaptive capabilities of bacteria, leading to variations in growth rates and drug sensitivities. As independent clinical isolates, 5b and 7b exhibit inherent genomic differences beyond the presence of class 2 integrons. The genetic heterogeneity of clinical strains may also lead to differential responses in stress regulatory networks, such as the SOS response system, to the same antibiotic concentration, resulting in inconsistent upregulation trends of resistance genes between the two strains. Further experiments are needed to confirm these hypotheses.

Integrons facilitate the rapid dissemination of bacterial resistance by capturing and expressing resistance gene cassettes, and their activity is regulated by environmental stress. This study found that, following drug-induced culture, the relative expression levels of class 2 integron integrase and resistance genes in the variable region of experimental strains increased to varying degrees, aligning with the expression trend of recA (Figures 2–4). This suggests a potential association between the expression of class 2 integrons and the SOS response. Research on class 1 integrons has identified a potential LexA binding site on the intI1 promoter, suggesting that the SOS response can specifically and directly induce the expression of class 1 integrase (Guérin et al., 2010). In class 2 integrons, a potential LexA binding site overlaps with the PintI2 promoter; however, subsequent experiments have shown that the LexA protein does not bind to this region. The antibiotic-induced SOS response could activate other LexA binding sites within the bacterial genome, such as promoters of DNA repair genes, thereby indirectly promoting the expression of intI2 and resistance genes through shared regulatory networks like RecA-mediated transcription factor activation. The expression of integrons is closely linked to bacterial physiology and environmental conditions, adapting as needed. Additionally, class 2 integrons might possess alternative stress-sensing mechanisms, potentially regulating expression by responding to other stress signaling pathways like oxidative stress or membrane damage response. This hypothesis requires validation through future transcriptomic or protein interaction studies (Erill et al., 2007).

The inhibition zone assay demonstrated that sub-MIC concentrations of antibiotics can significantly alter bacterial resistance phenotypes (Figure 5). After AMP induction, the sensitivity of constructed strains pAC-FA1 and pAC-FA3 to AMP decreased markedly, shifting from sensitive to resistant/intermediate. Clinical isolates 5b and 7b exhibited significantly enhanced resistance to CIP, with a reduction in the inhibition zone size. This finding supports the conclusion that low-level antibiotic exposure can drive resistance evolution (Barrett et al., 2019; Anderson et al., 2023). The induction effect at 1/2 MIC concentration was stronger than at 1/4 MIC, suggesting that higher sub-MIC concentrations exert greater selective pressure on resistance genes, thereby more readily promoting the emergence of resistant phenotypes. The differences in resistance between clinical and constructed strains are also noteworthy. Clinical isolates rapidly acquired resistance to the corresponding drugs after induction (especially with CIP), while the phenotypic changes in constructed strains were more limited. This may be attributed to the accumulation of gene elements, such as multiple mobile genetic elements, in clinical strains through long-term natural evolution, allowing for more efficient mobilization of resistance genes in response to stress. In contrast, constructed strains, with simpler genomes like that of E. coli TOP10, carrying only a single integron element, have limited adaptive evolutionary capacity. Additionally, some strains exhibited increased resistance to unrelated antibiotics after induction with a specific drug, corroborating the idea that antibiotics can promote multidrug resistance through “cross-selection” (Blázquez et al., 2018).

Sequencing of the class 2 integrase genes before and after induction did not reveal mutations in the stop codon TAA, which is inserted after the 178th amino acid in the integrase gene. The evolutionary capacity of bacteria depends not only on the rate at which genetic diversity is generated but also on the population size (Blázquez et al., 2018). In clinical and environmental settings, bacteria are often continuously exposed to multiple antibiotics, with pressure intensity dynamically changing over time. This “chronic mixed pressure” can accelerate the accumulation of adaptive mutations (Good et al., 2017; Lenski, 2017). Short-term antibiotic exposure may not lead to mutations at critical sites of class 2 integrons. Additionally, it is possible that any induced mutations are of low frequency and were not detected by the methods employed in this study. Occasionally, tester strains exhibited a base deletion downstream of the stop codon; however, due to the presence of the stop codon, their integrase gene would not have translated an intact integrase protein, rendering the mutation apparently ineffective on the integrase function. Further investigation is required to elucidate the mutation mechanisms of functional class 2 integrons.

Sub-MIC concentrations of drugs have been shown to induce an increase in the relative expression of recA, associated with the bacterial SOS response. This induction also elevates the expression levels of class 2 integron integrase and resistance genes. Continuous induction results in a sustained upregulation of these genes, which stabilizes or slightly decreases upon reaching a plateau, ultimately driving the evolution of clinical-level resistance in different antibiotic classes with various cellular targets. Moreover, higher sub-MIC levels more effectively trigger bacterial responses to adverse environments, enhancing population survival. The induction capacities of different drugs vary significantly, and the underlying mechanisms require further investigation. Short-term exposure to a single antibiotic did not result in critical mutations that would transform class 2 integrons into a functional form. The widespread use of antimicrobial agents in clinical settings, agriculture, and livestock inevitably leads to low-level antibiotic exposure, providing sufficient evolutionary pressure for bacterial resistance. It is imperative to develop effective, targeted strategies to address this challenge.