ABSTRACT
The presence of trace antibiotics in water can lead to the development of drug-resistant bacterial strains, posing risks to ecosystems and human health. Immunomagnetic separation (IMS) using magnetic nanoparticles (MNPs) is an effective technique for targeted enrichment. This study established and optimized a separation system for the immunomagnetic microsphere enrichment of enrofloxacin (ENR) antibiotics, achieving efficient enrichment and isolation of ENR. To address potential elution degradation, an analysis of ENR degradation pathways and toxicity assessment of degradation products was performed. The study manifested the successful conjugation of antibodies to magnetic microspheres, leading to a 97.68% separation efficiency for ENR in water through IMS. Specifically, 1 mg of MNP@Ab could specifically bind to 1.5 ng of ENR at 37 °C for 30 min, and the elution rate exceeded 83%. No degradation products of ENR were detected during the enrichment and isolation procedures. Nevertheless, extending the elution time to 1 h disclosed three major degradation pathways with higher toxicity risks than ENR based on ecological risk assessment. To strictly control the elution temperature and elution time, the increase in temperature and time will heighten the risk of degradation products. This study presents an efficient strategy for water treatment and environmental protection.
HIGHLIGHTS
Successful preparation of immunomagnetic microspheres.
Immunomagnetic microspheres can specifically enrich traces of enrofloxacin (ENR) in water.
The risk of ENR forming degradation products increases with elution time.
A key factor in preventing byproduct formation is reaction time control.
INTRODUCTION
Antibiotics are essential in combating bacterial infections and saving lives (Zhang et al. 2015; Wang et al. 2021). Enrofloxacin (ENR), a potent third-generation quinolone antibiotic widely used in livestock and aquatic products, raises significant concerns due to its high persistence of resistance genes and potential environmental risks (Li et al. 2020b; Gong et al. 2023). The rapid development of industry, agriculture, and aquaculture has led to antibiotic residues in the aquatic environment (Ma et al. 2022). Antibiotics account for 79.9% of the total number of emerging contaminants in surface waters in China (Delgado et al. 2023). Humans and animals are exposed to antibiotics through drinking water (Wang et al. 2023) and irrigated crops using groundwater reservoirs. Trace antibiotics are characterized by low concentrations, poor degradability, high toxicity, and bioconcentration. The presence of trace antibiotics has been detected in groundwater in many countries (Kivits et al. 2018; Tuts et al. 2024). Prolonged exposure to trace amounts of antibiotics can accelerate the horizontal transfer of antibiotic resistance genes (ARGs). These ARGs can then be transferred to bacteria in the human gut microbiota (Leonard et al. 2018), leading to potential contamination of microorganisms within the water column. This has the potential to trigger cyanobacterial outbreaks, increase their toxin-producing capacity, and disrupt ecological environments (Xia et al. 2023). As a result, it poses a threat to global ecosystems and human health (Tello et al. 2012; Tong et al. 2014; Sun et al. 2017; Hou et al. 2021). The currently recognized metabolite of ENR is ciprofloxacin (CIP). It is generated through the dealkylation of ENR, and CIP is a metabolite that enhances the activity of ENR, which in itself is a powerful antimicrobial agent. The degradation of antibiotics under specific conditions can exert an influence on the ecological environment and food safety. Hence, it is indispensable to carry out the identification and analysis of the unknown metabolites of ENR and investigate their ecological toxicity.
Current methods for the removal of antibiotics from water include biodegradation (Alexandrino et al. 2017), chemical oxidation (Patel et al. 2019), activated carbon adsorption (Delgado et al. 2019), and filtration techniques (Le et al. 2018; Bera et al. 2022). Biodegradation is only applicable to biodegradable antibiotics, and the majority of antibiotics are not effectively removed by either bioreactor or activated carbon processes. This is primarily due to their relatively low concentrations (typically at the ng/L level) and their resistance to biodegradation. It has been shown that the total removal of antibiotics by biodegradation is consistently >75%, but among the fluoroquinolones, ciprofloxacin and ENR are the only two antibiotics with removal rates below 90% (Raghavan et al. 2018). Chemical oxidation can enhance the removal of fluoroquinolones in highly contaminated water through oxidative reactions, and the removal rate can reach 94%. Maximum residue limits (MRLs) for ENR in China, Europe, and Japan have been limited to no more than 100, 100, and 50 μg/kg, respectively (Aslam et al. 2016). Trace amounts of antibiotics may still be present in water after undergoing antibiotic removal techniques. Discharge of industrial, agricultural, and other wastewater through watersheds after some treatment into water bodies allows the accumulation of trace antibiotics in water bodies.
The detection methods of antibiotics mainly include instrumental analysis, immunoassays, sensor methods, and so on. Instrumental analysis methods such as chromatography and mass spectrometry are widely used in antibiotic detection due to their high sensitivity and specificity (Cas et al. 2019; Xu et al. 2019). HPLC/MS–MS is a selective and highly sensitive analytical method for determining fluoroquinolone antibiotics in water, with a detection limit of ng/L. It can also be combined with other methods to mitigate limitations related to high equipment costs, poor repeatability, and complex operation. However, these methods are not easily applicable in resource-limited or field rapid detection scenarios. Immunoassays, such as enzyme-linked immunosorbent assay (ELISA) and fluorescent immunoassay, are easy to operate and fast but prone to instability and susceptibility to external factors leading to errors in detection results. Therefore, materials that amplify signals have been synthesized to improve the sensitivity of antibiotic detection. Additionally, combining antibodies with nanocomposite materials (Yang et al. 2023), quantum dots and fluorophores (Zhou et al. 2023b) can enhance the efficiency of antibiotic detection. Sensors including electrochemical sensors (Kling et al. 2016), optical sensors, and biosensors (Lan et al. 2017) offer rapid analysis capabilities along with portability simplicity field analysis less sample reagent consumption relatively low cost. They are widely used in detecting very low concentrations of antibiotics but suffer from high material costs and poor repeatability stability. While each method has its advantages in specific application scenarios, there are limitations that cannot be ignored; further improvement is needed for their application in actual environmental monitoring.
Traditional methods for separation and purification include solid phase extraction (SPE) (Zheng et al. 2022), solid phase microextraction (SPME), immunoaffinity chromatography (IAC), molecular blotting techniques (Cao et al. 2021), and supercritical fluid extraction. However, these methods still suffer from time-consumption, low selectivity, and adsorption capacity. In recent years, immunomagnetic separation (IMS) technology using immunomagnetic beads has shown significant advantages. These beads are modified with specific antibodies or ligands on their surfaces and can accurately recognize and bind to target antibiotic molecules, allowing for efficient separation and enrichment with high specificity (Cao et al. 2021). They can quickly capture antibiotics from complex samples and can be integrated with various detection platforms such as chromatography, mass spectrometry, and biosensors. The beads have strong magnetic responsiveness, good biocompatibility, ease of modification, strong targeting specificity, fast separation speed, and simplicity in operation, making them ideal for trace analysis in sample pretreatment. In one study (Tian et al. 2017), encapsulation of functional groups in Fe3O4 allows simple, sensitive, and efficient separation of quinolone antibiotics in water. Furthermore, immunomagnetic beads were prepared for the extraction and separation of salinomycin with an antibody coupling rate of 84.32% and an adsorption rate of 92.7%. There are also many studies coupling magnetic microspheres with different functional groups and combining them with downstream detection can be quickly and easily used for the detection of antibiotics in different food matrices (Xu et al. 2012; Lu et al. 2021; Li et al. 2022). At present, they have been widely used in the field of fungal toxins (Huang et al. 2018) and cells (Guo et al. 2023), but less in the field of antibiotic small molecules, and their practical application is still limited by stability, and high selectivity and adsorption capacity are also required.
Streptavidin and biotin are recognized as one of the strongest non-covalent forces, providing excellent stability. This effectively reduces the likelihood of antibody detachment from the surface of magnetic microspheres in subsequent experiments when compared with other immunomagnetic microsphere conjugates (Shan et al. 2014). By modifying the biotin on the antibody, the streptavidin magnetic microspheres can be linked to the antibody through the biotin affinity system to synthesize immunomagnetic microspheres, which can precisely capture the trace target while avoiding matrix interference with the help of the good magnetic separation properties of the complex and the specificity of the antibody–antigen. The capture efficiency of magnetic beads is related to the effectiveness and stability of IMS technology, which is mainly affected by the particle size of magnetic beads, the type of antibody, the capture mode, the capture conditions, and other factors. By optimizing the surface modification of magnetic beads and detection conditions, the application range can be further expanded, and the detection performance can be improved. Therefore, the optimization of parameters and conditions for the combined application of magnetic microspheres with antibody technology, as well as their applicability to small molecule antibiotics, still needs to be analyzed and studied.
This study aims to optimize the synthesis conditions of immunomagnetic microspheres, specifically focusing on the particle size of magnetic beads, incubation time, and binding capacity. This optimization is in response to the current issue of ENR residue. The study also explores surface modification strategies for microspheres to enhance their affinity, stability, and reusability toward the target antibiotic; thereby, achieving rapid enrichment and separation of samples. Liquid chromatography-mass spectrometry (LC-MS) and high-performance liquid chromatography-quadrupole time-of-flight tandem mass spectrometry (HPLC-QTOF) were utilized for the determination and analysis of ENR concentrations and its degradation products obtained during the experiment. An extensive investigation was undertaken into the degradation products formed during the separation process of ENR, including an evaluation of their ecological toxicity. These findings not only deepen our understanding of the behavior of ENR in environmental matrices but also provide valuable insights into more effective strategies for antibiotic residue management. These insights are significant for environmental protection and human health.
MATERIALS AND METHODS
Materials
All other chemicals and solvents used in our study are analytical grade. ENR and ciprofloxacin standards were purchased from Dr Ehrenstorfer GmbH (Augsburg, Germany); methanol (MeOH, HPLC grade) was supplied by Merck (Darmstadt, Germany); sodium dodecyl sulfate (SDS) was purchased from Shanghai McLean Biochemistry Technology Co. Ltd; ENR biotinylated monoclonal antibody (BIO107, purification >95%) was purchased from Beijing Biotai Biotechnology Co.; streptavidin (SA) magnetic microspheres were purchased from Weidu Biological Co.; antibody dilution buffer (Phosphate Buffer Saline (PBS), pH 7.4, 0.1% BSA) and Phosphate Buffer Saline with Tween 20 (PBST) buffer (PBS, pH 7.4, 0.05% Tween 20) were used in the experiment; and ultrapure water was supplied by Wahaha Foods Co. Ltd (Tianjin, China).
Synthesis of MNP@Ab
The antibody is added 0.01 M NaHCO3 to adjust the pH to 8.0, then connected to biotin, protected from light for 24 h, centrifuged to remove denatured product, and desalted until the antibody-biotin complex is obtained. To activate the magnetic beads, 1 mg of magnetic streptavidin microspheres was vortexed for 10 s, mixed on a roller mixer for 15 min, and magnetically separated to eliminate the supernatant. PBST buffer was added and mixed for 5 min at room temperature on a roller mixer, followed by two repetitions of the washing procedure to remove the supernatant after magnetic separation. Subsequently, a monoclonal antibody against ENR was introduced, vortexed, and mixed, and then incubated on a roller mixer. Upon completion of the reaction, the supernatant was collected through magnetic separation. The concentration of the monoclonal antibody in the supernatant was determined using the bicinchoninic acid (BCA) method, and the coupling rate was calculated using Equation (1). The resultant coupling of magnetic microspheres with the antibody underwent three washings with PBS and was set aside.
ENR enrichment and elution
One milligram of MNP@Ab was introduced into the water sample containing the target. To ensure thorough mixing and reaction, an appropriate volume of PBS buffer was added to maintain an optimal reaction environment. The mixture was then incubated in a roller mixer set at 37 °C for 30 min to facilitate effective binding of MNP@Ab to the target analyte. Following the incubation period, the centrifuge tubes were placed on a magnetic rack and left undisturbed for 3 min. During this time, the magnetic properties of MNP@Ab caused it to accumulate on one side of the centrifuge tube, tightly binding to the target analyte. Subsequently, the target-bound MNP@Ab could be easily isolated from the supernatant through magnetic separation. The collected supernatant was utilized for subsequent quantitative analysis. After ENR in the supernatant was filtered using a 0.22 μm filter membrane, the concentration of ENR was determined employing a Waters triple quadrupole liquid mass spectrometer. By comparing the mass concentration of ENR before and after the experiment, the binding rate (%) could be calculated using Equation (2). This value reflects the efficiency and specificity of MNP@Ab binding to the target analyte. Equation (3) is utilized to calculate the maximal binding capacity of ENR and antibody, thereby ascertaining the saturation value of the target substance associated with the immune microspheres.
Degradation product analysis and toxicity assessment
Evaluating the formation and risk of byproducts during degradation is a significant and intricate task. Through acute toxicity, chronic toxicity, and ecotoxicity assessments, a better understanding of the potential hazards posed by these byproducts to organisms is gained, crucial for ensuring proper antibiotic storage and accurate residue determination (Lin et al. 2020). The degradation products generated by ENR were analyzed using HPLC-QTOF. ENR and its metabolites underwent screening via diagnostic fragment ion (DFI) to identify the most probable products. The DFI search process involves a node-based workflow in MS-DIAL to preprocess the raw data for extracting features for suspect screening. Key stages encompass peak acquisition, retention time alignment, group differentiation (including isotopes and adducts), and merging of peaks across all samples to create a distinctive signature. Features with intensities below three times the solvent blank or program blank were recognized as background noise and eliminated. The final distinctive fragment ion was detected as fragment 245.1086 with a tolerance of 0.01 Da and an ion abundance of 30%.
Risk assessment calculations for ENR and its degradation intermediates were meticulously conducted using the ECOSAR program to anticipate the acute and chronic toxicity of these substances to fish, daphnia, and green algae. Acute toxicity to algae is quantified by EC50, while LC50 for fish and daphnia represents the concentration causing 50% mortality after 96 and 48 h of exposure, respectively. Notably, EC50 depicts the concentration leading to 50% mortality of green algae following 96 h of exposure. The specific data needed were acquired directly from the calculation process by plotting pertinent chemical formulas.
Data analysis and statistics
Data processing was conducted with Microsoft Excel 2019, while charting, data analysis, and linear fitting were accomplished utilizing Origin 2023 software.
RESULTS AND DISCUSSION
Characterization of MNP@Ab
Condition optimization
Enrichment and isolation of ENR in water
Magnetic microspheres hold significant promise for biomedical applications due to their uniform reaction process, rapid response to external magnetic fields, and ample surface area. ENR forms strong bonds with magnetic nanomaterials through antibody binding, while other impurities are eliminated during magnetic separation. The specific and reversible interaction between the antibody and antigen ensures the targeted enrichment and retrieval of ENR from the water column (Yu et al. 2023).
To optimize the separation efficiency of MNP@Ab, various parameters potentially affecting its performance were examined. The incubation time emerged as a critical factor influencing adsorption efficiency. Initially, the duration required for 1 mg of MNP@Ab to capture the target was investigated, revealing that adsorption efficiency increased over time, plateauing at 30 min with an adsorption rate peaking at 97.68% (Figure 4(d)). In comparison with previous research on nanobody-based IMS platforms (Bai et al. 2023), the rapid separation of Salmonella in food samples takes only 45 min. Furthermore, the separation efficiency of enoxacin can reach an impressive 97.68%, surpassing the capture efficiency of over 80% for tumor cells (Pan et al. 2022), thus making it even faster and more efficient.
To assess specificity, magnetic beads lacking antibody functionalization were exposed to ENR, showing an inability to adsorb ENR under identical conditions. Given the structural similarity between ciprofloxacin and ENR, specificity was assessed by substituting ENR with CIP – a potentially interfering antibiotic at higher concentrations. As depicted in Supplementary Figure S3, magnetic separation in the presence of CIP failed to isolate CIP, affirming the high specificity of the composite for ENR. Simultaneously, the introduction of immunomagnetic microspheres with attached antibodies and their strong binding to target antibiotics forms stable bonds or interactions, which can minimize the interaction with non-target substances (such as salts and metals) to achieve high selectivity in water. Research findings have indicated that magnetic SPE is capable of effecting the separation and pre-concentration of fluoroquinolones (FQ) from water samples. When coupled with HPLC-UV, the quantification of FQ in tap water, wastewater, and river water samples can be accomplished (Bayatloo et al. 2022). Magnetic functional nanocomposites can also function as adsorbents, which can undergo pre-concentration and separation for the detection of doxycycline in marine sediments without matrix interference (He et al. 2022). The magnetic graphene nanoparticle adsorbent can rapidly enrich sulfonamides (SA) in wastewater, and its application to actual wastewater is highly practicable (Wu et al. 2016).
Within a 1.5 mL centrifuge tube, 1 mg of immunomagnetic microspheres was combined with ENR concentrations ranging from 0.25 to 4 ng to determine the maximum adsorption capacity, as depicted in Figure 4(e). The results revealed that MNP@Ab could adsorb up to approximately 1.5 ng of ENR following a 30-min incubation at 37 °C. To obtain a profound comprehension of the adsorption mechanism, the Freundlich–Langmuir model (Figure 4(e) and 4(f)) was employed to analyze and fit the experimental data (Zhao et al. 2015) (refer to Supplementary Table S1 for detailed information on the model's non-linear equations). The Freundlich model effectively characterized the isothermal experimental data, indicating that the interaction between MNP@Ab and ENR is a complex and heterogeneous process not limited to a single molecular layer (Wang & You 2023).
ENR elution
Optimization of elution conditions was conducted to achieve satisfactory recoveries, as illustrated in Supplementary Figure S4. The eluents tested included methanol, SDS, a weak acid solution, and a formamide-EDTA mixture. Notably, the maximum elution of ENR, close to 80%, was attained with the formamide-EDTA mixture. Comparative analysis indicated superior elution using the formamide-EDTA blend over methanol and the other solutions. It was observed that excessively acidic or alkaline conditions adversely affected elution efficiency. At pH levels below 4.0, chelates formed between Fe and ENR, while at pH above 9.0, negative charges on the magnetic particles due to hydroxide group binding led to electrostatic repulsion with anions (Ibarra et al. 2012).
Further investigations were carried out on the influence of various elution times (2, 5, 10, and 15 min) and temperatures on elution efficiency. Notably, the impact of elution time on the elution rate was found to be insignificant (Yu et al. 2019). Similarly, no substantial difference was noted between room temperature and 65 °C. However, at 95 °C, the target compound was undetectable, possibly due to ENR degradation (Gong et al. 2023). Additionally, the elution effect of eluent volume and EDTA in the mixed solution was explored, with the optimal elution achieved using a mixed solution comprising 500 μL with 95% formamide and 20 mM EDTA, resulting in an elution rate exceeding 83%. The washed magnetic microspheres were characterized by SEM and TEM as shown in Supplementary Figure S5. The basic morphology and structure of the magnetic microspheres remained unchanged after use, indicating good stability. Furthermore, the magnetic microspheres demonstrated stable magnetic properties as they could achieve magnetic separation within 2 min and were reusable.
ENR degradation pathways
To investigate the impact of the elution system on ENR, reaction intermediates and degradation pathways were characterized using HPLC-QTOF analysis. No degradation products of ENR were detected under the experimental temperature and duration. However, upon extending the elution time to 1 h, degradation products were identified through HPLC-QTOF analysis, highlighting the pivotal role of reaction temperature in the degradation of ENR. Notably, the degradation of ENR exhibited a marked acceleration with rising reaction temperatures (Geng et al. 2023).
Compounds . | Retention (min) . | Ion formula [M + H]+ . | [M + H]+ (m/z) . | Proposed molecular formula . |
---|---|---|---|---|
ENR | 6.29 | C19H23FN3O3 | 360 | |
P1 | 5.53 | C19H21FN3O3 | 358 | |
P2 | 5.86 | C17H21FN3O3 | 334 | |
P3 (CIP) | 5.99 | C17H19FN3O3 | 332 | |
P4 | 6.33 | C18H23FN3O | 316 | |
P5 | 11.14 | C14H11FN2O4 | 291 | |
P6 | 9.40 | C13H12FN2O3 | 263 | |
P7 | 6.45 | C13H13N2O3 | 245 |
Compounds . | Retention (min) . | Ion formula [M + H]+ . | [M + H]+ (m/z) . | Proposed molecular formula . |
---|---|---|---|---|
ENR | 6.29 | C19H23FN3O3 | 360 | |
P1 | 5.53 | C19H21FN3O3 | 358 | |
P2 | 5.86 | C17H21FN3O3 | 334 | |
P3 (CIP) | 5.99 | C17H19FN3O3 | 332 | |
P4 | 6.33 | C18H23FN3O | 316 | |
P5 | 11.14 | C14H11FN2O4 | 291 | |
P6 | 9.40 | C13H12FN2O3 | 263 | |
P7 | 6.45 | C13H13N2O3 | 245 |
Toxicity assessment
In terms of acute toxicity, only P4 exhibits hazardous properties, with all other products deemed harmless. Concerning chronic toxicity, both ENR and its byproducts fall within safe thresholds. However, end products P4 and P7 from the three pathways in this study display higher toxicity levels compared with ENR in both acute and chronic toxicity assessments. In other studies regarding the degradation of ENR, the potential environmental risks of intermediate products were evaluated using logKow and logBAF values calculated by EPI-Suite™ software (Wang et al. 2024). The results also indicate that most ENR degradation products have greater bioaccumulation potential and stronger toxicity than the parent compound ENR. Environmental factors, such as temperature, pH, light exposure, and dissolved oxygen, can also influence the selectivity and efficiency of different degradation pathways of ENR (Klementová et al. 2022). It is worth noting that, compared with ENR itself, an increase in temperature or prolonged reaction time can lead to the formation of intermediate products with higher levels of biological toxicity. Therefore, strict control of elution temperature and time is crucial to prevent the formation of adverse degradation products of ENR because elevated temperature and extended reaction time will increase the risk of their formation.
Comparison with other methods
Supplementary Table S4 provides a detailed comparison of the significant advantages of our proposed method in the field of antibiotic extraction and separation. Our method significantly reduces the use of adsorbents while maintaining high efficiency in antibiotic separation, especially when dealing with trace water samples. This advantage is particularly prominent, indicating that our method has higher competitiveness in terms of resource conservation and cost-effectiveness. Compared with other complex and time-consuming pretreatment steps, our method offers a simpler and more efficient pretreatment process, reducing operational complexity and labor costs, thus enhancing its practicality and operability. Rapid and efficient magnetic separation of antibiotics is achieved utilizing MNP@SA as adsorbents. This feature is particularly important for rapid response environmental monitoring and food safety testing. It is worth noting that we have analyzed the degradation products generated during the antibiotic separation process. The results demonstrate that, under controlled temperature and time conditions, no toxic degradation products exceeding the parent antibiotics in toxicity are generated, thereby preserving the integrity of the antibiotics' chemical structure. This is crucial for subsequent analysis and detection to ensure accuracy and sensitivity of test results. In summary, based on MNP@SA, our method demonstrates outstanding performance and significant advantages in the field of antibiotic extraction and separation. It not only improves analytical efficiency but also ensures the accuracy and reliability of analytical results. Therefore, it represents an efficient, sensitive, environmentally friendly enrichment analysis method for residual ENR compounds.
CONCLUSION
In this study, we utilized immunomagnetic microspheres based on the streptavidin-biotin system for the efficient separation of trace antibiotics. We optimized and evaluated the conditions for antibody coupling to achieve optimal performance. This technique proficiently isolates the target antibiotic from the matrix, prominently reducing matrix interference during subsequent detection and concurrently enhancing the target. The implementation of the magnetic separation technique markedly shortens the sample separation process. Furthermore, our toxicity assessment of the degradation products indicated that the risk of formation of these products increases with extended reaction time. This emphasizes the crucial role of precise control over the reaction time. Consequently, it is imperative to rigorously manage the elution temperature and time to prevent the generation of undesired degradation products during the test. Elevated temperatures and prolonged elution times exacerbate the risk of such degradation, potentially impacting the accuracy and reliability of the analytical results. The IMS system can also be applied to the separation of different antibiotics by modulating the capture antibody. The technology can separate and detect trace antibiotics in various water samples and be integrated into water treatment pretreatment, combined with other detection methods, for rapid antibiotic separation and detection. In future research, further exploration of the application of this technology in the simultaneous separation of various trace antibiotics is needed to achieve synchronous separation and detection of different types of antibiotics. At the same time, there is a need for an in-depth study of the generation mechanism and toxicity of antibiotic degradation products under different conditions, as well as the development of effective control strategies.
ACKNOWLEDGEMENTS
This work was supported by the National Key Research and Development Program of China (grant number 2022YFC3202103), the Central Public-Interest Scientific Institution Basal Research Fund, the Freshwater Fisheries Research Center, CAFS (No. 2023JBFM02), Identification of New Antibiotic Contaminants in Wuxi Aquatic Products and Development of Rapid Detection Technology (K20231029), and China Agricultural Research System (CARS-46).
DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.
CONFLICT OF INTEREST
The authors declare there is no conflict.