Establishment of an indirect competitive immunoassay for the detection of dicamba based on a highly specific nanobody

Abstract

Dicamba, a traditional highly effective and low toxicity herbicide, has gained new life with the development of dicamba-tolerant transgenic crops in recent years. However, dicamba is highly volatile and therefore easy to cause drift damage to sensitive crops. The development of efficient and sensitive detection methods is essential for monitoring of trace dicamba in the environment. Nanobody-based immunoassay plays an important role in on-site detection of pesticides. However, now rapid and sensitive immunoassay methods based on nanobody for dicamba detection were lacking. In this study, the nanobodies specifically recognizing dicamba were successfully obtained by immunising camels and phage display library construction, and then an indirect competitive immunoassay based on Nb-242 was constructed with IC₅₀ of 0.93 μg/mL and a linear range of 0.11~8.01 μg/mL. Nb-242 had good specificity with no cross-reactivities against the dicamba analogs other than 2,3,6-trichlorobenzoic acid and the developed immnoassay had a good correlation with the standard HPLC in the spike-recovery studies. Finally, the key amino acid Ala 123, Tyr 55, Tyr 59 and Arg 72 of Nb-242 that specifically recognizing and binding with dicamba were identified by homologous modeling and molecular docking, laying an important foundation for further structural modification of Nb-242. This study has important guiding significance for constructing immunoassay method of dicamba based on nanobody and provides a sensitive, specific, and reliable detection method that is suitable for the detection of dicamba in the environment.

1. Introduction

Dicamba (3,6-dichloro-2-methoxybenzoic acid), a hormonal herbicide belonging to the benzoic acid series of herbicides, is one of the representatives of traditional herbicides in the world. It has broad spectrum, high efficiency, low toxicity and is used to control annual and perennial broadleaf weeds (Behrens et al., 2007). Due to the introduction of dicamba-tolerant transgenic crops developed by Monsanto, the use of dicamba in the world has reached another climax. The environmental issues caused by this herbicide are also receiving increasing attention, especially the drug damage caused by its drift to non target and sensitive crops such as soybean (Jones et al., 2019), cotton (Hamilton and Arle, 1979), and watermelon (Culpepper et al., 2018) is becoming increasingly prominent. Arkansas and Missouri in the United States have already banned the sales and use of dicamba in 2017 because of the sensitive crop damage caused by its drift (Wang Guobin, 2021). In addition, the residue of dicamba can also affect the genetic toxicity of aquatic organisms and cause DNA single strand breaks(Ruiz et al., 2014). Dicamba has teratogenic effects on zebrafish embryos, causing malnutrition and yolk sac edema in zebrafish, leading to a decrease in hatching rate at high concentrations (Felisbino et al., 2023). A study (Attademo et al., 2021) has shown that the addition of dicamba had a significant impact on the liver tissue and cell function of tadpoles, exhibiting high biological toxicity to local species of amphibians. More worrisome is that the widespread use of dicamba can also cause direct harm to the human body and the agricultural health analysis shows that the use of dicamba is associated with colon cancer and lung cancer (Lerro et al., 2020). Given this, it is particularly important to construct a simple and fast detection method for dicamba.

At present, the commonly used approaches for dicamba detection are instrumental methods, such as GC-MS (Zhang et al., 2023), LC-ESI-MS/MS (Guo et al., 2016), LC-MS/MS (Larose et al., 2023) and so on. Although these methods has the advantages of accuracy, sensitivity, and good selectivity, and are widely used in the field of pesticide residue detection, the experimental operation requires high technical requirements for operators, and the pre-treatment is relatively complex, making them unsuitable for on sit detection in the field. Immunoassays have been considered efficient, fast, highly sensitive, and easy to operate detection methods (Gao et al., 2020; He et al., 2021; Watanabe et al., 2013) and have been widely used in detection of pesticides, their metabolites or degradation products, pharmaceuticals, veterinary drugs, food and other environmental pollutants in recent years (Ren et al., 2018; Yu et al., 2022; Yu et al., 2023; Zeng et al., 2022). A polyclonal antibody against dicamba was firstly developed in 2001 (Clegg et al., 2001), and based on this study, we had designed and synthesized three haptens for dicamba, and successfully prepared polyclonal antibodies with higher sensitivity (Huo et al., 2019). However, there is still a certain gap in the stability and specificity of polyclonal antibodies compared to nanobodies due to its ability to bind to multiple antigenic epitopes stimulated by multiple antigens, and polyclonal antibodies face difficulties in structural modification and mass production.

Nanobody, discovered in the 1990s, is a naturally occurring heavy chain antibody that lacks light chains (Hamers-Casterman et al., 1993) and a research hotspot and development trend of ELISA technology. Nanobody has more advantages than traditional antibody, and it has been widely used in many fields. With the advantages of small molecular weight, capacity of penetration into tissues and easy modification, nanobodies playing an important role in prevention, diagnosis and treatment of infectious diseases, especially in immunotherapy against tumor cells (Alderton, 2019; Schütze et al., 2018). Currently, nanobodies have been widely used in the field of small molecule detection (Zhang et al., 2019a). For example, the anti-FB1 idiotypic nanobodies were screened from a phage displayed nanobody library after three rounds of selection, and a competitive ELISA was established based on anti-idiotypic nanobodies to detect FB1, with an IC₅₀ of 0.95±0.12 ng/mL (Shu et al., 2015); Zhang et al (Zhang et al., 2022) designed six haptens and then immunized alpaca, followed by establishing an efficient phage display nanobody library for panning paraquat nanobodies, and constructing TRFICA for quantitative analysis and recognition of paraquat through strip readers; The nanobodies targeting cyanofenzamide and chloramphenicol were obtained through alpaca immunization and phage display library construction, and an ELISA was developed for detecting these two insecticides based on the nanobodies (Xu et al., 2021). Haptens targeting 2,4-D were synthesized (Li et al., 2020) and evaluated using polyclonal antibodies, based of which, a nanobody-alkaline phosphatase fusion protein was prepared and one-step fluorescence enzyme immunoassay was developed with an IC₅₀ of 1.9 ng/mL for 2,4-D. Nanobody-based immunoassays have significant advantages in terms of sensitivity, stability and specificity, and there are currently no reports on the detection and analysis methods of dicamba based on nanobodies, so it is of great significance to construct an enzyme-linked immunosorbent assay method for dicamba based on nanobodies.

Based on the agricultural environment issues caused by dicamba, the prepared immunegen was used for camel immunization, and then a phage display nanobody library was constructed. The specific nanobodies against dicamba were successfully obtained by using dicamba as the target for affinity elution. Based on the nanobody, an indirect competitive immunoassay was constructed, and the immunoassay showed good sensitivity and specificity. According to the spike and recover studies, the immunoassay method had good correlation with the instrumental analysis method. Also, the detection limit was lower than the maximum residue limit of dicamba in environmental samples, indicating that this detection method was suitable for detection of dicamba in the real environmental samples. In addition, the mechanism of nanobodies recognizing and binding to dicamba had been preliminarily clarified using computer-aided technology, laying an important foundation for structural modification of nanobodies in the future study.

2. Materials and methods

2.1. Chemicals and reagents

Dicamba, Toxadine, 3,6-dichloropyridine carboxylic acid, 2,3,5-trichlorobenzoic acid, Chloramben, 2,3,6-trichlorobenzoic acid and 2,5-dichloro-3-hydroxy-6-methoxybenzoic acid were purchase from Sigma-Aldrich (St.Louis, MO, USA) or Thermo Fisher Scientific (Rockford, IL). T4 DNA ligase, restriction enzyme SfiI and helper phage M13KO7 were obtained from New England Biolabs, Inc. (Beverly, MA). LeukoLOCK Total RNA Isolation System, Top 10 F’ competent cell and bacterial protein extraction reagent (B-PER) were supplied by Thermo Fisher Scientific (Rockford, IL). HisPur Ni-NTA resin were purchased from Shanghai Shengong Biotechnology Service Co.,Ltd. (Shanghai, China). 3, 3’, 5, 5’-tetramethylbenzidine (TMB) two-component substrate solution (for ELISA), Isopropyl-β-d-thiogalactopyranoside (IPTG), RnaseZap and DNA marker were purchased from Solarbio Science & Technology Co.,Ltd. (Beijing, China). Goat anti-alpaca IgG-HRP was supplied by Jackson Immunoresearch Laboratories (West Grove, PA, USA). The pComb3x and pET22b vector were gifts from Prof. Bruce Hammock (UC Davis, California, USA). The E. coli ER2738 was purchased from Biosearch Technologies Co. Ltd. (Shanghai, China). Immunogen and coating antigens (Figure 1A) were previously prepared and kept in our laboratory.

Fig. 1.

(A) The best combination of immunogen and coating antigen used in this study; (B) The titer of camelid antiserum, the concentration of different coating antigens was 1 μg/mL; (C) Inhibition of free dicamba (1 μg/mL) to different concentration of camelid antiserum, hapten 24-OVA (with different concentration) was employed for coating antigen; (D) The standard curve of ic-ELISA based on polyclonal antiserum, with the IC₅₀ value of 15.8 μg/mL.

2.2. Immunization and antiserum assessment

Immunogen (JQ-21-Thy) was used to generate nanobodies according to standard camel immunization procedures and the briefly immunization steps were as follows. 50 mL of serum was taken from the camel under the neck as a negative control before the first immunization. At the initial immunization, 100 μg of JQ-21-Thy mixed with an equal amount of complete Freund’s adjuvant was injected into the camel intramuscularly. Then, booster injections was performed every 2 weeks using the same immune methods, but the difference was that the immunogen (JQ-21-Thy, 50 μg) was emulsified with an equal amount of incomplete Freund’s adjuvant. Starting from the 3rd immunization, 10 mL of serum was taken for titer measurement one week after each immunization. The immunization was terminated until there was no significant difference in serum titer compared to the previous immunization. For the titer measurement of camel serum, an enzyme-linked immunosorbent assay procedure was developed and the details of the operation were listed in the Supplementary Material. 150 mL of blood was collected in blood collection tubes with EDTA anticoagulants one week after the last booster for the isolation of peripheral blood lymphocytes.

2.3. Library construction, biopanning and screening the nanobody

The total mRNA was extracted from the isolated peripheral blood lymphocytes and then reverse transcribed to cDNA by RT-PCR. DNA fragments encoding the nanobody genes were amplified twice by PCR using two sets of primers (F1:5’-GTCCTGGCTGCTCTTCTACAAGG-3’,R1:5’-GGTACGTGCTGTTGAACTGTTCC-3’;F2:5’-ACTGGCCCAGGCGGCCGAGGTGCAGCTGSWGSAKTCKG-3’,R2:5’-ACTGGCCGGCCTGGCCTGAGGAGACGGTGACCWGGGTC-3’). The DNA fragments were digested and then ligated into the vector pComb3x, which were electroconverted, and finally a phage display nanobody library was successfully constructed by assisting in the infection of phage M13KO7. Positive nanobody candidates, screened by phage-ELISA after amplification, were expressed in Top 10F’ as vector. In addition, the expression conditions were optimized and the nanobodies were identified by SDS-PAGE. The performance of the selected nanobody candidates were further evaluated by indirect competitive enzyme-linked immunosorbent assay (ic-ELISA).

2.4. Characterization of anti-Dicamba nanobody

Sensitivity of Nb-242

The ELISA based on nanobody plays an important role in the rapid on-site detection and monitoring the contamination of pesticides in the environment. In this study, nanobody Nb-242 showing best performance in ic-ELISA was selected for the subsequent studies. Firstly, the suitable concentration of 24-OVA (coating antigen) and Nb-242 was determined by checkerboard procedure. The detailed procedure for sensitivity evaluation of Nb-242 based on ic-ELISA was listed in the Supplementary Material.

Specificity of Nb-242

Nanobodies often have better specificity when compared to traditional antibodies. In this study, the specificity of the obtained nanobody Nb-242 was evaluated by measuring its cross reactivity against the structural analogs of dicamba, and then the cross reactivity was compared with that of polyclonal antibody obtained in our previous study. The relative cross reactivity was calculated by the following equation: cross reactivity (%) = [IC₅₀ (dicamba)/IC₅₀ (structural analogs)]×100.

2.5. Analysis of spiked samples

In order to evaluate the accuracy of the immunoassays developed in this study, the spike and recovery study was performed using tap water and soil. The blank samples used for spiked recovery assessments were confirmed to be free of dicamba by HPLC. Tap water samples were spiked with dicamba in methanol at final concentrations of 0.4 μg/mL, 4.0 μg/mL, and 32.0 μg/mL, while, the soil samples were spiked with dicamba in methanol at final concentrations of 4.0 μg/mL, 40 μg/mL, and 320 μg/mL. The spiked samples were allowed to stand at room temperature overnight after thorough mixing. The tap water samples were directly diluted and analyzed by the ic-ELISA and HPLC, while the solid samples (10 g) were extracted using 20 mL methanol by a vortex mixer for 15 min. The mixture was then centrifuged at 4000 rpm for 5 min, and the supernatant was collected, diluted appropriately and analyzed by the ic-ELISA and HPLC.

For HPLC, tap water samples (20 mL) were added to a solution of acetonitrile (40 mL) contained NaCl (15 g). Then, anhydrous sodium sulfate (5 g) was added to the above mixture. The organic phase was collected, filtered and evaporated to dryness. The soil samples (20 g) were mixed with acetonitrile (50 mL) and water (5 mL) and extracted by shaking for 1 h. The mixture was filtered and NaCl (15 g) was added. Organic phase was collected and evaporated to dryness. The residues of tap water and soil samples were dissolved with methanol (5 mL) and analyzed by HPLC. The detection of HPLC was performed using methanol-water (1:1, v/v) as the mobile phase at 230 nm with a flow rate of 0.5 mL/min and an injection volume of 20 μL.

2.6. Mechanism of dicamba specifically recognized by Nb-242

Computer-aided technology is one of the most important means to study the process of small molecule recognized by nanobodies. In this study, the homology modeling and molecular docking were used to further clarify the molecular mechanism of specific recognition of dicamba by Nb-242. The key amino acid sites that determine the specificity of nanobody will lay an important foundation for further targeted structural modification of Nb-242. First, the three-dimensional structure of Nb-242 was predicted by YASARA (Version 23.4.25.W.64). The high resolution template required for homology modeling was derived from the protein data bank (PDB ID 7TJC, resolution 1.35 Å)(Swofford et al., 2022). Procheck, errat and verify3D were used to evaluate the protein models. The 3D structure of dicamba (CAS ID: 1918-00-9) was retrieved from the ChemSpider database (http://www.chemspider.com/) and optimized to the minimum energy by YASARA. And then, the 3D structure of dicamba was docked to the binding site of the Nb-242 model by YASARA. Docking calculations was performed using Vina and each molecule was tested for 25 times. The docked structure was scored using the built-in scoring function and was clustered using 1.0 Å of root mean square deviation criterions(Yang et al., 2023). The binding energy and dissociation constants were calculated and the intermolecules interactions between the key amino acid sites of Nb-242 and dicamba were presented in visualization format.

3. Results and discussion

3.1. The response of serum from the immunized camel

Starting from the 3rd immunization, antiserum was collected from the camel one week after each immunization for the performance evaluation. The concentrations of coating antigens (21-BSA, 21-OVA, 24-BSA and 24-OVA) and different batches of antiserum were firstly determined by the checkerboard method (data not shown). As shown in Figure 1B, the titer performed by 24-OVA was higher than that of the other coating antigens. Hence, 24-OVA was chosen for the following study. Also, the titers of antiserum increased from the 3rd to 5th immunizations, and then reduced at the 6th immunization. Thus, blood from the 5th immunization was chosen for the evaluation of inhibition and nanobody isolation.

The optimal concentration of 24-OVA and the 5th antiserum was determined by the checkerboard method. The results (Figure 1C) showed that the highest inhibition rate was observed when the concentration of 24-OVA was 1 μg/mL and the 5th antiserum was diluted 1000-fold. So, the standard curve of ic-ELISA (Figure 1D) for the 5th antiserum was prepared with the IC₅₀ and the linear range of 15.80 μg/mL and 4.52–56.29 μg/mL, respectively.

3.2. Library construction and isolation of nanobodies

Total RNA was extracted from the peripheral blood lymphocytes isolated from 10 mL of blood by centrifugation using LeukoLOCK total RNA isolation system and then cDNA was obtained after reverse transcription. The amplification product of the first round using cDNA as a template and F1/R1 as primers was identified by 1% agarose gel electrophoresis. A 750 bp fragment of the heavy chain antibody was yielded from the the first round of amplification and was used as a template for the second round of amplification. A 450 bp fragment of the nanobody was yielded from the second round of amplification which using F2/R2 as primers. The SfiI sites introduced with the primers were used for subsequent cloning of the nanobody fragments into the phagemid pComb3X vector. The ligated vector was then electroporated in E. coli ER2738 cells and applied to obtain the nanobody display phage by amplification with M13K07 helper phage. The phage-displayed nanobody library with a size of 1.6×10⁸ clones was bio-panned in three rounds against the selected coating antigen 24-OVA. Amplified phage solution was eluted using phage-ELISA, and three positive phage clones were screened out. The specific sequence of the three clones were shown in Figure S1. Based on the screening results of the positive clones, the performance of the three nanobodies was initially evaluated. The results were shown in Figure 2, which indicated that the IC₅₀ of Nb-242 was significantly lower than that of other two nanobodies, and therefore the nanobody Nb-242 was chosen for the followong studies.

Fig. 2.

Standard curve of ic-ELISA based on the isolated three nanobodies, with the IC₅₀ value of 0.89 μg/mL, 1.01 μg/mL, and 5.21 μg/mL, respectively.

3.3. Characterization of anti-Dicamba nanobody

Sensitivity of Nb-242

The assay parameter conditions of ionic strength, pH and orgnic solvent are often optimized to increase the sensitivity and accuracy of the immunoassay methods. The above optimized conditions could be used in extraction, clean up and concentration of samples. Firstly, the optimal concentrations of coating antigen (24-OVA, 1 μg/mL) and Nb-242 (diluted 1000-flod) were determined by checkerboard titration method. Then, the effect of different ionic strength of buffer on the sensitivity of immunoassay methods was evaluated and the results (Figure S2) showed that the higher the ion strength of the buffer, the lower the maximum absorbance. The difference in IC₅₀ was not significant at different ionic strength, but the lowest IC₅₀ value (0.93 μg/mL) and the highest absorbance value were observed under 1×PBS. Therefore, the optimal ionic strength of the immunoassay method constructed in this study was 1×PBS. The effect of buffer pH between 5.0 and 9.0 on the performance of immunoassay method was evaluated, and the results (Figure S3) indicated that the maximum absorbance showed no remarkable change for the immunoassay, and the lowest IC₅₀ value (2.06 μg/mL) was observed at pH 8.0 which was selected for the following experiments. Organic solvents such as methanol and acetonitrile are generally used for sample extraction, dissolution and dilution of analytes in the process of enzyme-linked immunosorbent assay. However, organic solvents can affect the performance of antibodies to a certain extent, thereby affecting the recognition and binding between antibodies-antigens, or antibodies-analytes. Therefore, a reasonable evaluation of the impact of organic solvents on enzyme-linked immunosorbent reaction is of great significance for the accuracy of this method. Methanol was selected as solvent to dissolve dicamba because of its better solubility. The results (Figure S4) showed that methanol would only affect the maximum absorbance when the concentration was 40%, and when the methanol concentration was 10%, its impact on ELISA results could be ignored. After the optimization of the immunoassay for dicamba, the IC₅₀ value was 0.93 μg/mL with a linear range (IC_20–80) of 0.11–8.01 μg/mL(Figure 3).

Fig. 3.

Standard curve for dicamba by ic-ELISA based on Nb-242. Serial dilutions of dicamba standard were mixed with Nb-242 in optimized buffer. Then 100 μL of the mixtures were added to the coating antigen-coated wells. Each point represents the mean value of three replicates.

Specificity of Nb-242

The specificity of the Nb-242 was evaluated using structural analogs of dicamba, and the results showed (Table 1) that negligible cross-reactivity was observed for all compounds except for the 2,3,6-trichlorobenzoic acid. The cross-reactivity of Nb-242 against 2,3,6-trichlorobenzoic acid in this study was 19.23%. While the cross-reactivity of dicamba polyclonal antibody against 2,3,6-trichlorobenzoic acid in the reported study was 33%. Thus the specificity of the nanobody was significantly improved compared to the traditional polyclonal antibody.

Table 1.

Cross-reactivity of the Nb-242 and polyclonal antibody against dicamba structural analogs.

Compound	Structure	Polyclonal antibody CR (%) (Huo et al.. 2019)	Nb-242 CR (%)
Dicamba		100	100
2,3,5-Trichlorobcnzoic acid		<0.1	<0.1
2,3,6-Trichlorobcnzoic acid		33	19.23
Picloram		<0.1	<0.1
3,6-Dichloropyridine carboxylic acid		<0.1	<0.1
Chloramben		<0.1	<0.1
2,5-Dichloro-3-hydroxy-6- methoxybenzoic acid		<0.1	<0.1

Thermo-stability of Nb-242

The thermal stability of the Nb-242 was evaluated by comparison with the antiserum. The purified Nb-242 and polyclonal antiserum were incubated at 20, 50, 65, 75 to 95°C for 5 minutes, and was also evaluated at 85°C for 5, 15, 25, 35, 45, and 60 minutes. The results showed that (Figure S5A and Figure S5B), the polyclonal antiserum gradually lost its binding ability as the temperature increases, whereas Nb-242 was still able to bind to the coating antigen at temperatures as high as 95°C. And the activity of Nb-242 could maintain nearly 100% after being incubated at 95°C for 5 minutes. The binding activity of both Nb-242 and polyclonal antiserum decreased under the condition of incubation at 85°C, but the activity of Nb-242 could maintain about 65% after being incubated for 60 minutes. These results indicated that the nanobody had better thermal stability compared to polyclonal antibody, which played an important role in the wider applicability of nanobodies.

3.4. Validation of ELISA in spiked water and soil samples

Matrix effects can seriously interfere with the performance of immunoassays and have a certain impact on the accuracy of the detection method when using immunoassay in complex samples (Lin et al., 2016). Therefore, matrix effects should be minimized or eliminated during sample analysis to ensure the accuracy of the immunoassay method. The simplest method for matrix elimination or reduction is to dilute the sample with the assay buffer. In this study, tap water and soil were selected for matrix effect evaluation. The tap water samples confirmed to be none of dicamba were diluted directly with 1×PBS to a total dilution ratio of 2-, 4-, 8-, and 16-fold, and then different concentrations of dicamba solution were prepared with these obtained buffer solutions, respectively. The results (Figure 4A) displayed that there was no significant difference in maximum absorbance under different dilution ratios of tap water. But, tap water samples diluted 2-fold had higher IC₅₀ when compared with assay buffer or other diluted samples. When the tap water samples were diluted 4-, 8- and 16-fold, there was no significant difference in maximum absorbance and IC₅₀ when compared with assay buffer. Therefore, the 4-fold dilution was selected for the tap water samples. Soil samples without dicamba were used as blank samples and extracted with methanol, then diluted with 1×PBS resulting in a total dilution ratio of 10-, 20-, 40-, and 80-fold. Dicamba standard solutions were prepared using the above obtained buffer solutions. The results (Figure 4B) showed that the soil samples diluted 10 and 20-fold had the lower maximum absorbance and higher IC₅₀ than those of other dilutions. When the soil matrix was diluted 40 and 80-fold, there was no significant difference between their maximum absorbance and IC₅₀. Therefore, the 40-fold dilution was selected for the next spike and recovery study. The results of matrix effects indicated that higher concentration of matrix could affect the recognition and binding between antibodies and coating antigens/analytes.

Fig. 4.

The effect of tap water (A) and soil matrix (B) on the performance of Nb-242 based ELISA.

In order to evaluate the accuracy and reliability of the ELISA developed in this study, the spike and recovery study was applied using tap water and soil samples. The results showed that (Table 2) the average recoveries ranged from 97.8% to 105% for tap water and 100.3% to 105% for soil samples analyzed by ELISA. Also, the results of the ELISA method were approximately the same as that of HPLC. This results indicated that the ic-ELISA based on Nb-242 developed in this study could be utilized in the detection of dicamba residues in tap water and soil with acceptable accuracy and reproducibility.

Table 2.

Spike-recovery results for tap water and soil samples determined by ELISA and HPLC.

Samples	Spiked concentration (μg/ml)	Concentration (μg/mL)		Average recovery (%)		CV (36)
		ELISA	HPLC	ELISA	HPLC	ELISA	HPLC
Tap water	0.4	0.42±0.04	0.36±0.04	105	90.0	9.52	11.1
	4.0	4.08±0.11	3.76±0.0 5	102	94.0	2.70	1.33
	32.0	31.3±1.35	32.2±2.06	97.8	100.6	4.31	6.40
Soil	4.0	4.16±0.33	3.60±0.3l	104	90.0	7.93	8.61
	40	42.0±3.15	41.2±3.23	105	103	7.50	7.84
	320	320.8±4.26	319.6±5.18	100.3	99.8	1.33	1.62

3.5. The recognition mechanism of nanobody against dicamba

Studying the recognition mechanism between antibody and antigen has important guiding significance for targeted modification of antibody structure and preparation of antibodies with high quality, high specificity and affinity(Kusharyoto et al., 2002; Zhang et al., 2019b). In the study of antibody recognition mechanism based on the structure-activity relationship model, due to the lack of accurate grasp of the three-dimensional spatial structure of the antibody, the development of high-quality antibody preparation was limited (Yuan et al., 2011). Therefore, obtaining the three-dimensional spatial structure of antibody has gradually become an important way to analyze the recognition specificity of antibodies. Computer-aided technology is important in the study of antibody-antigen recognition (Lippow et al., 2007). In this study, in order to clarify the molecular mechanism of the specificity between nanobody and dicamba, the 3D structure of Nb-242 was constructed by homogenous modeling, and then was docked with dicamba by molecular docking. A high resolution template required for homology modeling was derived from PDB database based on the sequences of Nb-242 (M K K T A I A I A V A L A G F A T V A Q A A E V Q L V HSGGGSVQAGGSLRLSCAASGNRNNVAYMGWYRQRPGKERERVARIDTRASGNKWYADSVKGRFTISQDYNAKNTVYLQMNSLKPEDTAVYTCAAALPYDTQPRYWGQGTLVTVSSGQAGQHHHHHHGAYPYDVPDYAS) and a homology model of Nb-242 was then created (Figure 5A). The PROCHECK evaluation results (Figure S6A) indicated that the residues in most favored regions rate were 94.6%. The evaluation results of model ERRAT (Figure S6B) showed that the overall quality factor was 91.064 which was higher than 50. Also, the results of model Verify3D evaluation (Figure S6C) showed that 93.08% of the residues had averaged 3D-1D score≥0.1. All the above evaluation indicated that the constructed protein model was a high-quality model. After docking the Nb-242 with dicamba one hundred times, the most suitable structure for stacking dicamba with Nb-242 was selected based on the binding energy. As shown in Figure 5B, the binding of dicamba to Nb-242 was mainly in the pocket formed by CDR2 and CDR3. The binding pocket of Nb-242 had many interactions with dicamba, including one hydrogen bonds (green, Tyr59) and several hydrophobic force (Ala 123, Tyr 55 and Arg 72). This docking study illuminated the binding pattern of Nb-242 against dicamba. In the future study, molecular dynamics simulation technology can be used to study the entire process of remote attraction, recognition, binding and adjustment of Nb-242 and dicamba, then using molecular biology technology to further identify the recognition mechanism of Nb-242 against dicamba. This results has important guiding significance for the targeted structure modification of the nanobody to obtain the antibodies with better performance.

Fig. 5.

The results of homologous modeling and molecular docking. (A) The three-dimensional structure of Nb-242 obtained by homologous modeling; (B) The recognition mode between Nb-242 and dicamba in 3D docking plot

4. Conclusion

With the widespread adoption of genetically modified crops in the world, the use of dicamba will continue to increase, and the environmental issues caused by this herbicide will also be increasingly valued. A simple and efficient detection method is important in guiding the correct and reasonable use of dicamba. In this study, we obtained a novel nanobody against dicamba by immuning camel, serum assessment, phage-display library construction, and affinity panning. The nanobody Nb-242 showed better specificity against dicamba than the polyclonal antibody obtained in our previous study. According to the optimization of enzyme-linked immunosorbent reaction conditions, an ic-ELISA based on Nb-242 was successfully developed with an IC₅₀ of 0.93 μg/mL. This immonoassay method had good sensitivity and accuracy for the detection of dicamba contamination. In addition, the mechanism of Nb-242 specific recognition of dicamba was preliminarily clarified using computer-aided technology, and four key binding amino acids Ala 123, Tyr 55, Tyr 59 and Arg 72 were obtained, laying the foundation for the targeted structural modification of nanobody in the future study.