**4. Discussion**

Plants are an important component of terrestrial ecosystems and are a potential pathway for antibiotic transport because of their absorption capacity [33]. Our previous studies indicated that cephalosporin, tetracyclines, and sulfonamides were the most frequently used antimicrobial agents in livestock farms in China, and high ratios of AREB occurred in the livestock manure fertilized field vegetables [28]. The transformation of these antibiotics and their induction of antibiotic resistance in soil or water environments have been frequently reported [6,41,42]. Further study through pot planting experiments confirmed different compositions of AREB presence in vegetable endophytic systems [43]. Therefore, TC, CPL, and SMX were selected to explore their accumulation and induction in pakchoi endophytic systems in this study. To simplify the conditions, a series of hydroponic cultures of pakchoi were used. Based on the Clinical and Laboratory Standards Institute (CLSI) standards [32] and the bacterial community composition in the pakchoi endophytic system [43], the highest MIC values of TC, CPL, and SMX for different genera of endophytic bacteria were selected and set as the levels of antibiotic exposure in the present study. From our results, we can see that although the exposure doses of different drugs were greater than their practical occurrence in agroecosystems, the pakchoi still showed natural growth throughout the planting process. Moreover, according to our previous investigation [28], sometimes animal manure containing high concentrations of antibiotic residue also will be used for field plants. Therefore, this study provided direct evidence for the effects of antibiotics on plant growth and the development of antibiotic resistance, especially under different treatment doses. The changes in growth parameters indicated the phytotoxic levels of each antibiotic at different doses. Generally, at 50% MIC levels, antibiotics stimulated growth, increasing plant fresh biomass production. However, at MIC levels, the antibiotics acted as inhibitors, reducing yields and inducing metabolic disturbances. Previous research had indicated that low antibiotic concentrations are beneficial for plant growth, whereas high antibiotic concentrations can induce toxicity [44]. In a comparison of the three antibiotics, CPL has the lowest toxicity to pakchoi. This may be due to β-lactam's specific actions on bacterial cell wall components, which are targets that do not exist in plant cells [45]. Therefore, the growth inhibition rate was lowest under CPL exposure when compared with controls.

The detection of the three antibiotics in pakchoi tissues indicates the uptake and transfer of antibiotics from the water environment to the vegetable, which is similar to previous results [14,18]. The bioaccumulation of antibiotics in plants can vary depending on plant species and antibiotic class [15,20]. Usually, ionization, as well as the properties of sorption and water solubility, can directly affect how plants uptake pharmaceuticals [21]. CPL was noted to have accumulated to the highest concentration among the study compounds at the MIC exposure level, which may reflect the greater absorbency of CPL compared with the other compounds. Furthermore, the concentrations of antibiotics in pakchoi did not increase unlimitedly as antibiotic dose increased. The probable reasons include: (1) the saturation level of antibiotic accumulation was reached; (2) the incorporated antibiotics were stored in the plant cells, in which they can be degraded; and (3) the degradation of antibiotics was accelerated by the release of plant enzymes during sample grinding [20]. Nevertheless, vegetables that have accumulated antibiotics from contaminated environments will be consumed by humans, and then might be absorbed by the human body, resulting in increased antibiotic resistance, including the emergence of multidrug-resistant bacteria, which leads to antibiotic treatment failures [13,46].

Few studies have explored the influence of antibiotic pollution on endophytic systems in vegetables. Yet vegetables contaminated by antibiotics may contribute to the development of AREB. The present study showed that the rates of AREB occurrence and the relative abundances of ARGs increased in pakchoi endophytic systems after exposure to the three antibiotics. The trends of these changes were comparable with the results from our previous study, in which a distinct increase in some AREBs was shown in manure- or organic fertilizer-amended pakchoi samples [43]. Two possible factors may contribute to such increases. Firstly, during the pakchoi cultivation, there would be a natural rhizosphere microbial consortium forming in the nutrient solution. However, spiked antibiotics as a selective pressure could influence the microbial community compositions and induce the occurrence of high ratios of resistance. This has been proved in many other environments even at much lower concentrations than their MIC values [47–50]. Thus, a special microbial consortium would be established corresponding to different treatments of antimicrobial agents. As we know, environmental bacteria, especially rhizosphere bacteria, are an important source for plant endophytic bacteria [51], which could enter through the tissues to the plant endophytic systems, thus resulting in high occurrence of AREB in the plant. Secondly, the AREB can be persistent in the plant endophytic systems. In the present study, the accumulated antibiotics in plants, although below their corresponding minimum MIC values for various species of bacteria, might also provide a selection pressure to the endophytic

bacteria, thereby providing the AREB survival advantages in the polluted environments. The qPCR also demonstrated that the abundance of ARGs in the endophytic bacteria corresponding to the three antibiotics continuously increased as the antibiotic uptake in the pakchoi increased.

Among the four detected ARGs, the presence of *tet*X, a special enzymatic modification gene for the degradation of TCs, may be related to the low detected TC concentration in pakchoi. In addition, the absence of the *tet*X gene in control samples and the persistence of it in the treated samples may indicate the potential transfer of bacteria carrying the *tet*X gene from the environment to plant. Notably, the third-generation drug tigecycline has been used in clinical treatments due to its broad spectrum of antibacterial activity (especially inhibiting multiple antibiotic-resistant bacteria and super bacteria) [52–54]. However, the bacterial strains containing the *tet*X gene isolated from patients are still resistant to tigecycline [55]. Therefore, the prevalence of the *tet*X gene in edible pakchoi should be highly concerning. The other ARGs, *sul*1, *sul*2, and *bla*CTX-M, are widely present in various environmental media [56–58], and act as the most prevalent mechanisms of sulfonamide and β-lactam resistance, respectively. In particular, the *sul*1 gene is normally found linked to other resistance genes in the Tn21 type integron, while *sul*2 is usually located on small plasmids of the IncQ family [59]. qPCR showed that the abundance levels of these ARGs increased in antibiotic contaminated environments, indicating their enrichment and transmission under antibiotic selection pressure.

Previous studies have demonstrated the antibiotic uptake [22] and the presence of resistant human pathogens or opportunistic pathogens in vegetables planted in manure-amended soil [43]. Thus, accumulated antibiotics in vegetables and the prevalence of antibiotic resistance in endophytic systems might be disseminated to humans when these vegetables are consumed. Consequently, evaluating the biological responses of terrestrial crops to antibiotics, especially frequently consumed vegetables, is important. However, compared with the soil environment, this hydroponic cultivation system is just a simple model to evaluate the influences of antibiotics on plant growth and plant endophytic bacteria. Further research is required to study the community compositions of AREB corresponding to different types of antibiotic exposure under soil cultivation systems. The results will provide basic information for an integrative risk assessment of antibiotic application and food security.

## **5. Conclusions**

The present study investigated the growth of pakchoi and the antibiotic resistance in its endophytic system under TC, CPL, and SMX exposure. Pakchoi was shown to absorb antibiotics from the hydroponic culture environment. The absorption was selective toward different antibiotics, and the absorption amount was related to the antibiotic concentration. The accumulated antibiotics in the plant influenced the growth of the plant and increased the levels of AREB and ARGs, even at sub-inhibitory doses, which should be noted due to considerations surrounding the possible transfer of ARGs through the food chain.

**Supplementary Materials:** The following are available online at www.mdpi.com/1660-4601/14/11/1336/s1, Table S1: PCR primers, annealing temperatures, and resistance mechanisms, Table S2: qPCR primers and annealing temperatures used in the present study, Table S3: qPCR standard curves for 16S rRNA gene and antibiotic resistance genes, Table S4: Root length, plant height, and fresh biomass values of hydroponic pakchoi under different dosages of antibiotic treatment.

**Acknowledgments:** The authors would like to acknowledge the financial support from the National Natural Science Foundation of China (NSFC 21477035 and U1504219), the Key Science and Technology Project of Henan Province (142102210447), and the Specialized Research Fund for the Doctoral Program of Higher Education (20134104110006).

**Author Contributions:** Hao Zhang conducted the experiments, performed the data analysis, and drafted and edited the manuscript. Xunan Li performed the data analysis. Qingxiang Yang designed the study, provided the academic guidance for the work, and contributed to the critical revision for the paper. Linlin Sun, Xinxin Yang, Mingming Zhou, Rongzhen Deng, and Linqian Bi conducted the experiments.

**Conflicts of Interest:** The authors declare no conflict of interest.
