Study of the sensitization mechanism of immunoassay based on DNA tetrahedron spatial mediation for detecting imidacloprid in vegetables

Graphical abstract

This schematic illustrates an immunoassay based on TDN-mediated to increase the sensitivity of imidacloprid detection.

Highlights

• •The DNA tetrahedron (TDN) was designed first time via Python language.
• •The TDN was labelled on the surface of magnetic beads (MBs) for spatial mediation orientation of antibody.
• •The specific recognition Fab region of antibody was exposed by TDN spatial mediation orientation.
• •The sensitization mechanism of immunoassay was explored based on DNA tetrahedron spatial mediation.
• •A convenient, efficient and sensitive immunoassay was established for detecting imidacloprid in vegetables.

Abstract

Introduction

Enzyme-linked immunoassay (ELISA) is a widely adopted method for detecting pesticide residues such as imidacloprid because of its speed, low cost, and simplicity. However, antibody adsorption immobilization on solid-phase surfaces often leads to spatial crowding, resulting in reduced accessibility of antigen-binding sites and compromised assay sensitivity.

Objective

This study aimed to address the limitations of conventional immunoassays by introducing DNA tetrahedron (TDN) as spatial mediators. Spatial mediation of antibodies using TDN allows for sequential targeting and full exposure of the antibody to the recognition site.

Methods

Firstly, the sequence of TDN was designed using python language. And then, the TDN was labelled on the surface of magnetic beads (MBs) for spatial mediation orientation of antibody to form MB@TDN@Ab immunoprobe, exposing the specific recognition Fab region of the antibody. The sensitization mechanism of immunoassay based on TDN spatial mediation was explored for detecting imidacloprid in vegetables.

Results

The MB@TDN@Ab immunoprobe acquired a higher affinity constant (Dissociation equilibrium constant (K_d) = 3.21 µg/mL) than MB@Ab immunoprobe (K_d = 19.16 µg/mL). The immunoassay based on TDN spatial mediation achieved an IC₅₀ (1.40 ng/mL) and a linear detection range (0.05 ∼ 50 ng/mL). Compared with the immunoassay based on MB@Ab immunoprobe (IC₅₀ = 2.66 ng/mL and linear range = 0.05 ∼ 5 ng/mL), the sensitivity and linear range immunoassay based on MB@TDN@Ab immunoprobe are much higher and broader.

Conclusion

A convenient, efficient and sensitive immunoassay was established for detecting imidacloprid in vegetables in the study. The MB@TDN@Ab immunoprobe improves immunoassay performance while supporting safer food monitoring practices. These findings contribute to sustainable agricultural safety, public health assurance, and innovation in biosensing technologies.

Introduction

Imidacloprid, the first generation of neonicotinoid pesticides, is widely used in agricultural production because of its broad-spectrum insecticidal properties, low dosage, and good systemic properties [1]. Imidacloprid is used for pest control in agricultural production to ensure the yield and quality of vegetables [2]. Although imidacloprid is highly effective in pest control, its misuse poses a substantial threat to the environment and public health, leading to severe ecological damage. A study has shown that excessive imidacloprid can be released into the environment and may persist in the soil for up to 3000 days [3]. The imidacloprid applied to crops can leach into water and enter the food web, which is potentially toxic to aquatic organisms and human health.

ELISA has the advantages of rapidity, convenience, simplicity and economy and has been widely used in the detection of imidacloprid [4]. However, traditional ELISA tend to capture antibody on solid surfaces via adsorption, resulting in disordered orientation and recognition site hiding. TDN, as a DNA nanostructure with three dimensions, is well characterized by programmability, addressability, and high precision [5,6]. Moreover, TDN is comparable in size to the size unit of the recognition element (antibody or aptamer). TDN serve as ideal scaffolds for recognition elements, allowing for modulation of the nanometer spacing of the antibody and full exposure of the recognition site [7,8]. Fan et al. [9], who first realized TDN-modulated antibody detection of PSA in 2011, proved that the rigid structure of TDN could realize the directional modulation of antibodies. The team then further compared the detection performance of the TDN-modulated antibody with that of the dsDNA-modulated antibody. The results revealed that the sensitivity of the TDN-modulated antibody was 20-fold greater than that of the dsDNA-modulated antibody [10]. Moreover, Mi et al. constructed a microfluidic chip to detect PSA via TDN-modulated antibodies. PSA was detected in spiked serum samples with a linear dynamic range of 1 ∼ 100 ng/mL and an LOD of 0.2 ng/mL [11]. Currently, TDNs are mainly used in electrochemical and microfluidic chip methods, which have limitations such as high cost, poor stability and low detection efficiency [12]. Moreover, the antibody is adsorbed or covalently bound to the surface of the two-dimensional electrode or microarray, which has a small specific surface area and low antibody loading efficiency [13].

To address the limitations of conventional immunoassays, this study proposes the use of MBs as carriers because of their high stability, low cost, large specific surface area, and efficient detection capability. By anchoring TDNs onto the surface of MBs, the antibody was spatially mediated on the surface of MBs, which is proposed to improve target accessibility and signal sensitivity (IC₅₀ and linear range). Horseradish peroxidase-Hapten (HRP-Hapten) and an imidacloprid standard solution competitively bind to the antibody on the surface of the MB@TDN@Ab immunoprobe to produce a colorimetric signal. TDN spatial mediation realized the spatially oriented arrangement of antibody on the surface of MBs. Compared with that of immunoassay based on MB@Ab immunoprobe, the sensitivity of immunoassay based on MB@TDN@Ab immunoprobe was improved by 1.9-fold, and the linear range was increased by one order of magnitude, which can achieve trace detection of imidacloprid in agricultural products.

Materials and methods

Reagents and materials

DNA sequences were synthesized and purified by Sangon Co., Ltd. (Shanghai, China). MBs (10 mg/mL, 180 nm) with 150 μmol/g of –COOH were obtained from Allrun Co., Ltd. (Shanghai, China). Imidacloprid monoclonal antibodies were provided by Zhejiang University (Hangzhou, China). Hapten (imidacloprid) was synthesized by the Institute of Chemistry, Chinese Academy of Sciences (Beijing, China). The imidacloprid standard (100 mg/kg) was procured from Beijing Manhage Bio-Technology Co., Ltd. (Beijing, China). TMB (3,3,5,5-tetramethylbenzidine) single-component substrate solution, phosphate-buffered saline (PBS), horseradish peroxidase (HRP) and tris (hydroxymethyl) aminomethane were obtained from Beijing Solarbio Science & Technology Co., Ltd. (Beijing, China). 1-(3-Dimethylaminopropyl)-3-ethylcarbodiimide (EDC) and bovine serum albumin (BSA) were obtained from Shanghai Yuanye Biotechnology Co., Ltd. (Shanghai, China). Tween 20 was acquired from Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China). An immuno-clear flat-bottom 96-well plate (#439454), NHS-PEG4-Azide, SuperBlock (37532) and 10 × TBE Electrophoresis Buffer were supplied by Thermo Fisher Scientific (Pittsburgh, PA, USA). Methanol, N,N-dimethylformamide (DMF), hydrochloric acid (HCl) and sodium hydroxide (NaOH) were obtained from Sinopharm Chemical Reagent Shaanxi Co., Ltd. (Xi’an, China). GelRed dye was purchased from Biotium (San Francisco, USA). A polyacrylamide kit was purchased from CWBIO Co., Ltd. (Jiangsu, China). Polyethylene glycol (PEG) was obtained from Yeasen Biotechnology Co., Ltd. (Shanghai, China).

Characterization and instruments

The morphology and size of TDN and MB were confirmed with a laser particle analyzer (Zetasizer Nano ZS90, UK), an atomic force microscope (AFM, Bruker AXS, USA) and field-emission scanning electron microscopy (TEM, FEl Tecnai G2 F30, USA). The absorbance was measured via a multifunctional microplate reader (Tecan spark, Salzburg, Austria). The microplate was washed with a DEM-3 washing machine (Tuopu, Beijing, China). A polyacrylamide gel electrophoresis instrument (JUNYI 300HE, Beijing, China) was used. A gel imaging analyzer (Sinsage ChampGel, Beijing, China) was used. A PCR instrument (Bio-Rad T100, California, USA) was used. High-performance liquid chromatography (Shimadzu LC-20AD, Japan) and volume exclusion columns (Phenomenex, Yarra TM 3 μm SEC-4000, USA) were used.

Construction of the TDN

Programming of TDN sequence design

The code (Fig. S1) was pasted into a Python program such as Pycharm, the number of bases in the tetrahedral prisms of the DNA needed was entered, and the code was run to obtain four base sequences. The four oligonucleotide chains obtained via the NUPACK website were entered, the generation of secondary structures for each oligonucleotide chain was verified, and the binding free energy between different oligonucleotide chains was analyzed.

Optimizing the TDN synthesis buffer (TM Buffer)

TDNs were synthesized according to the previous method [14]. In brief, four DNA strands (S1, S2, S3, S4) were mixed in TM buffer to a final concentration of 1 µmol/L. An equal amount of DNA was denatured at 95 °C for 10 min, followed by rapid annealing at 4 °C for 30 min to produce TDNs.

The synthesis of TDN was carried out using TM buffers with varying combinations of Tris-HCl and Mg²⁺ concentrations, and the yields were assessed by polyacrylamide gel electrophoresis (PAGE) and ImageJ. The synthesized TDNs were characterized by DLS and TEM. The synthesized TDN was purified via LC and then characterized via AFM.

Construction and characterization of immunoprobes

MB@Ab

MB@Ab was prepared via the same method as previously reported [15]. One hundred microliters of magnetic beads (MBs) were dispersed in MEST buffer and washed twice (0.01 M, 0.05 % Tween 20, pH 6.0), followed by the addition of 200 µL of EDC (5 mg/mL, 0.01 M MES, pH = 6.0) and 200 µL of NHS (5 mg/mL, 0.01 M MES, pH = 6.0) and activated by rotation at 25 °C for 30 min. After magnetic separation, 10 µg of imidacloprid Ab (0.1 M PB = 8.0) was rotationally mixed with the MB and incubated at 25 °C for 3 h. The supernatant was removed and incubated for 2 h at 25 °C. After magnetic separation, the supernatant was pipetted, and 10 % ethanolamine (0.01 M PBS, pH = 7.4) was added to the supernatant by rotary incubation for 1 h at room temperature to block redundant sites. Finally, MB@Ab was synthesized, washed 3 times and resolubilized with PBS (0.01 M, pH = 7.4) containing 0.5 % BSA and 1 % PEG.

MB@TDN@Ab

100 µL of MBs were dispersed in MEST buffer, washed twice (0.01 M, 0.05 % Tween 20, pH = 6.0), followed by the addition of 200 µL of EDC (5 mg/mL, 0.01 M MES, pH = 6.0) and 200 µL of NHS (5 mg/mL, 0.01 M MES, pH = 6.0), and activated by rotation at 25 °C for 30 min. Following magnetic separation, 100 µL of TDN (1 µM, 0.01 M PBS, pH = 7.4) was rotationally mixed with MBs and incubated at 25 °C overnight [16]. After magnetic separation, the supernatant was pipetted and incubated with 10 % ethanolamine (0.01 M PBS, pH = 7.4) for 1 h at room temperature to block redundant sites. The supernatant was subsequently pipetted, and 10 µL of NHS-PEG4-Azide (10 mM, DMSO) was added for 3 h of rotary incubation at room temperature to activate the NHS ester group at the top of the TDN. The supernatant was aspirated and washed 3 times with PBS, and 10 µg of imidacloprid Abs (PB 0.1 M, pH = 8.0) was added to spin mix with the sediment, which was subsequently incubated at 25 °C for 2 h. MB@TDN@Ab was reconstituted with PBS (0.01 M, pH = 7.4) containing 0.5 % BSA and 1 % PEG [17]. To verify the successful construction of these immunoprobes, ultraviolet–zeta potential and confocal laser scanning microscopy were utilized to characterize these three immunoprobes.

Synthesis of the HRP-Hapten conjugate

For the EDC/NHS method, 0.1 mmol of imidacloprid hapten and 0.2 mmol of NHS were weighed and dissolved in 0.5 mL of DMF. First, 0.2 mmol of EDC dissolved in 0.5 mL of DMF was weighed, the reaction mixture was mixed, and the mixture was stirred at room temperature overnight. Hapten-HRP conjugates were synthesized by reacting different molar ratios of hapten (Hapten: HRP = 20:1, 40:1, 60:1) with a fixed molar amount of HRP. Specifically, 2 % HRP solution (0.01 M PBS) was added to Hapten dropwise in an ice bath followed by continuous stirring for 4 h. The mixed solution was loaded into a pretreated dialysis bag (30 kDa) and dialyzed with 0.01 M PBS for 4 d, and the dialysate was changed once every 24 h.

Immunoassay procedures

Many critical factors, including the coupling concentration of surface Ab to MB@Ab and MB@TDN@Ab, the HRP-Hapten concentration, the type of blocker, and the PEG concentration, were optimized to achieve a simpler procedure and higher performance.

First, 350 µL of blocker was added to a 96-well microplate, and the plate was washed to obtain a blocked plate after incubation at 37 °C for 60 min. The immunoprobe was added to 50 µL of HRP-Hapten and 50 µL of 10 % methanol pesticide. The plate was then incubated and shaken at 37 °C for 1 h. HRP-Hapten competes with the imidacloprid pesticide with a limited amount of antibody. To remove unbound HRP-Hapten, the 96-well plate was separated by placing it on a 96-part magnetic separation plate for 1 min, and the plate was washed three times with a pipettor. Then, 50 µL of PBS was added to each well for reconstitution, 100 µL of TMB single-component color development solution was added, and the reaction was carried out at room temperature for 15 min. The absorbance at 652 nm was measured quickly with a microplate reader. Three independent replicates were performed per group to validate experimental reproducibility and rigor [18].

Specificity and stability evaluation of the MB@TDN@Ab

50 µL of acetamiprid, thiamethoxam, chlorpyrifos, triazophos, malathion, fenthion, isocarbophos or imidacloprid (10 % in methanol) at a concentration of 50 ng/mL were added to blocked microtiter plates. In addition, 50 µL of 0.3 mg/mL MB@TDN@Ab immunoprobe and 50 µL of 1 mg/L HRP-Hapten were added. The mixture was shaken at 37 °C for 1 h and then washed 3 times with PBS (Tween-20 0.1 %). Then, 50 µL of PBS was added to each well for reconstitution, 100 µL of TMB single-component color development solution was added, and the reaction was carried out at room temperature for 15 min. The absorbance at 652 nm was measured quickly with a microplate reader.

50 µL of 10 % in methanol at a concentration of 50 ng/mL were added to blocked microtiter plates. Other experimental steps for stability evaluation were performed using the same experimental procedures as for specificity evaluation.

Evaluation of antibody coupling efficiency and affinity

The concentration of unbound antibody in the supernatant following MB coupling was quantified via the Trace Protein Quantification Kit (Thermo #23235). The immunoaffinity of the MB@Ab and MB@TDN@Ab immunoprobes was evaluated via ELISA. Briefly, 96-well microtiter plates were coated with OVA-hapten in PBS buffer (0.5 µg/mL, 100 µL/well) and incubated at 37 °C for 2 h. After washing, 300 µL of 2 % BSA was added to each well for blocking and incubated at 37 °C for 1 h. Serial concentrations (0.1 ∼ 0.8 mg/mL) of each immunoprobe were then added and incubated at 37 °C for 1 h. Subsequently, 100 µL of IgG-HRP (1:1000 dilution) was added and incubated for another hour at 37 °C. Afterward, 100 µL of TMB substrate was added, and the reaction was stopped after 15 min. Absorbance was measured at 652 nm.

To investigate the sensitization mechanism, molecular docking simulations of the antibody variable regions and the TDN structure were performed via AutoDock and PyMOL software. The interaction mechanism was analyzed in terms of spatial site resistance.

Sample preparation

Chinese cabbage, cucumber, and zucchini were used as spiked samples. The standard concentrations of imidacloprid (0.5, 1, and 5 ng/mL) were determined by adding 5 mL of methanol to the homogenized samples (5 g), shaking them manually for 10 s, extracting them for 30 min, and then centrifuging them for 15 min at 8824g (4 °C). The entire supernatant was transferred to a 50 mL plastic centrifuge tube and fixed with methanol to 20 mL. Fivefold dilution with methanol was continued, and a final dilution with PBS buffer solution (10 mM, pH = 7.4) was made to ensure that the proportion of methanol was 10 %.

Results and discussion

Python code for designing the 4 oligonucleotide strands

The Python code is provided in the supporting information. The DNA sequences generated by Python are shown in Table 1.

Table 1.

Sequence of the four oligonucleotide chains.

Name	Sequence (5′-3′)
S1	NH2-TGGAATGCACACCACGCACAAAGACGGAATACACGTTCCGACAAGCGAAAGTGAGAGCCAGACG
S2	NH2– TGTGCGTGGTGTGCATTCCAAAATTGCCTCTAACCGACGTGCAACCGAACTCACGAGGGACGGA
S3	DBCO-TTTTTTTTTTCGGTCGCAATCGGTCACGTAAAGTCGGAACGTGTATTCCGTCAATCCGTCCCTCGTGAGTTCGG
S4	NH2– TACGTGACCGATTGCGACCGAAGCACGTCGGTTAGAGGCAATAACGTCTGGCTCTCACTTTCGC

Optimizing the TM buffer

TM Buffer consists of Tris-HCl and Mg²⁺. Tris-HCl primarily stabilizes the solution pH, whereas Mg²⁺ neutralizes the negative charges on DNA strands, thereby reducing electrostatic repulsion between oligonucleotides. Together, the optimal concentrations of Tris-HCl and Mg²⁺ facilitate efficient assembly and increase the yield of TDNs. Fig. 1 shows the 8 % PAGE characterization of TDNs synthesized in different TM buffers. Recovery calculations via ImageJ were performed on the target bands, and the TDN yield was initially estimated by calculating the area of the bands according to ImageJ. Fig. 1 shows that the TDN synthesis yield was determined by the ionic concentrations of both Tris-HCl and Mg²⁺. The highest yield was found at a Tris-HCl concentration of 20 mM and a Mg²⁺ concentration of 25 mM. The study used a 3 % agarose gel for cut gel recovery experiments and calculated TDN yields of 75 % to 80 %. Therefore, 20 mM Tris-HCl and 25 mM Mg²⁺ are the best choices for TM buffer. The Python-based sequence design used in this study has not yet been fully optimized, and the generated sequences require further validation through molecular simulations or experimental synthesis to confirm their reliability.

Fig. 1.

(a) PAGE characterization from different TM buffers. (b) Yields of TDN synthesized from different TM buffers. Note: TDNs synthesized in different buffers are represented by 1–12 in Fig. 1 (ab). (1): 20 mM Tris-HCl, 50 mM Mg²⁺; (2): 20 mM Tris-HCl, 25 mM Mg²⁺; (3): 20 mM Tris-HCl, 12.5 mM Mg²⁺; (4): 20 mM Tris-HCl, 5 mM Mg²⁺; (5): 10 mM Tris-HCl, 50 mM Mg²⁺; (6): 10 mM Tris-HCl, 25 mM Mg²⁺; (7): 10 mM Tris-HCl, 12.5 mM Mg²⁺; (8): 10 mM Tris-HCl, 5 mM Mg²⁺; (9): 5 mM Tris-HCl, 50 mM Mg²⁺; (10): 5 mM Tris-HCl, 25 mM Mg²⁺; (11): 5 mM Tris-HCl, 12.5 mM Mg²⁺; (12): 5 mM Tris-HCl, 5 mM Mg²⁺.

Characterization of TDNs

The synthesis of TDN was confirmed and characterized. As shown in the TEM images in Fig. 2a, TDN exhibited a distinct tetrahedral morphology. The hydrodynamic diameter of the TDNs determined via DLS analysis was 8.85 ± 0.07 nm (Fig. 2c). To further determine the successful synthesis of TDN. The size of the DNA tetrahedra was characterized via PAGE, and the following mixtures were synthesized under the same annealing conditions: three single-stranded DNAs (S123), two single-stranded DNAs (S12), and a single-stranded DNA (S1). The formation and stability of the structures were verified via PAGE. The migration distance decreased as the amount of added ssDNA increased (Fig. 2b). The TDNs were purified by liquid chromatography, and the target peak fractions were collected, concentrated in MWCO (30 kDa) ultrafiltration tubes and replaced with TM buffer. The P1, P2, and P3 peaks were characterized by gel electrophoresis (Fig. S3). The results showed that P3 was the target TDN. The size of the purified TDN, characterized by AFM, was consistent with the above characterization (Fig. 2d). All the above characterization methods demonstrate the successful synthesis of TDN [19,20]. Furthermore, the structure prediction of the four synthesized DNA sequences was performed in this study using AlphaFold. The results in Fig. 2(e) show that TDN has a good spatial tetrahedral morphology proving the good feasibility of the model designed.

Fig. 2.

(a) TEM image of TDN (scale lines = 200 nm). (b) A PAGE image of TDN (Lane 1: DNA marker; Lane 2: S1S2S3S4; Lane 3: S1S2S3; Lane 4: S1S2; Lane 5: S1). (c) Particle size distribution histogram of TDN. (d) AFM images of TDN. (e) TDN structure prediction via AlphaFold (model_2.cif).

Evaluation of HRP-Hapten

HRP and Hapten were coupled via combinations of different molar ratios of Hapten:HRP = 25:1, 40:1, and 60:1. The imidacloprid antibody (10 mg/L) was encapsulated on electrostatic adsorbent plates, and the same concentration of HRP-Hapten was added after blocking. The absorbance values were detected after the plate was washed and TMB one-component color development solution was added for 15 min. Fig. S8 shows that the OD values increased with increasing molar ratio of Hapten:HRP. The difference between the OD values of 60:1 and 40:1 is small, and 40:1 is chosen as the optimal ratio of Hapten to HRP for cost-saving purposes.

The checkerboard method was used to select the optimal antibody and HRP-hapten ratio. The combination with an OD of 1.0 and the smallest amount of antigen–antibody mixture was selected as the optimal combination. Table S2 shows that the optimal antibody concentration of HRP-Hapten was 10:0.9 (mg/L). The imidacloprid standard curve Y = 0.4677X-0.3552, R² = 0.9792, was established on the plates via direct competition with a linear range of 5 ∼ 200 ng/mL (Fig. S9).

Characterization of the MB@TDN@Ab

The morphology and distribution of the MBs were characterized via SEM, TEM, DLS, and energy dispersive spectroscopy (EDS). Fig. S4-5 show that the MBs were monodisperse and nearly spherical, with an average particle size of 200 nm. The DLS results revealed that the average hydrated particle size of the MBs was 211.2 ± 3.8 nm (Fig. S6). The EDS results revealed that the elemental distribution of the MBs was mainly composed of Fe (Fig. S7).

MBs were prepared via EDC/NHS-mediated and TDN-mediated coupling strategies for MB@Ab and MB@TDN@Ab. The binding of Ab to the MB surface and MB@TDN surface was coupled via amide bonds. Fig. 3A shows that the zeta potentials of the MBs, MB@TDN and MB@TDNA@Ab were −3.58, −24.50 and −4.72 mV, respectively, for the three nanomaterials, with the surface charge increasing and then decreasing. A dramatic decrease in the surface potential after the modification of MB with TDN can be observed, caused by the negative charge of the phosphate backbone of TDN, which is consistent with the results of a previous study [21]. The zeta potential of the MB@TDN@Ab was dramatically greater than that of the MB@TDN, which was caused by the binding of the antibody to TDN.

Fig. 3.

(A) Zeta potentials of MB, the MB@TDN and MB@TDNA@Ab. (B) Confocal laser scanning microscopy images of (a)(b)(c) MB@TDN@Ab, (d)(e)(f) MB@TDN, and (g)(h)(i) MB@Ab. where the excitation wavelengths of (a)(d)(g) are 525 nm and 595 nm, those of (b)(e)(h) are 525 nm, and those of (c)(f)(i) are 595 nm. (C) DLS of MB, the MB@TDN and MB@TDNA@Ab. (The inset shows the size and PDI of the three nanomaterials). Note: Red fluorescent labeling indicates TDNs; green fluorescent labeling indicates antibodies.

Confocal laser scanning microscopy was employed to characterize the MB@TDN@Ab. Fig. 3 (a–c) shows fluorescence confocal images of MB@TDN@Ab when 525 nm and 595 nm, 525 nm, and 595 nm fluorescence excitations were used. At both excitation wavelengths, the red fluorescence of TDN overlapped with the green fluorescence of Ab, confirming the successful synthesis of MB@TDNA@Ab. Fig. 3 (d-f) shows fluorescence confocal images of the MB@TDN. The absence of green fluorescence in Fig. 3e proves that no antibody is coupled to the surface of the MB@TDN and that only red fluorescence proves the successful synthesis of the MB@TDN. Fig. 3 (g-i) shows fluorescence confocal images of MB@Ab with only green fluorescence at both excitation wavelengths, confirming the successful coupling of the antibody on the surface of MB.

Furthermore, DLS analysis revealed that the particle sizes of MB, MB@TDN, and MB@TDN@Ab were 166.4 nm, 216.3 nm, and 235.9 nm, respectively, confirming the successful stepwise assembly of TDNs and antibodies onto the magnetic beads.

Optimization of immune factors

First, the antibody concentrations on the surfaces of MB@Ab and MB@TDN@Ab were optimized by adding 1, 2, 5, 6, 8, or 10 µg of antibody to 1 mg of MB. Similarly, 0.5, 1, 2, 5, 8, 10 mg of antibody was added to 1 mg of MB@TDN, respectively. As shown in Fig. 4(a-b), an antibody mass of 8 µg was chosen as the MB@Ab and MB@TDN@Ab probe coupling concentration in this study. Under the same MB and antibody concentrations, the absorbance value and inhibition rate of MB@TDN@Ab immunoprobe were greater than those of the MB@Ab immunoprobe. MB@TDN@Ab immunoprobe showed 80 % inhibition with an absorbance of 1.20, whereas MB@Ab immunoprobe showed 60 % inhibition with an absorbance of 0.7 (8 µg of antibody). This demonstrated that TDN served to order the antibody spatially, allowing the antibody active site to be fully exposed and achieving full binding to HRP-Hapten.

Fig. 4.

(a) Absorbance values and inhibition of MB@TDN@Ab immunoreactivity at different antibody masses (b) Absorbance values and inhibition of MB@Ab immunoreactivity at different antibody masses. (Inhibition well concentration of 50 ng/mL). Optimization of the experimental parameters (c) different sealer types (d) PEG concentrations, (e) MB@TDN@Ab concentrations, and (f) HRP-Hapten concentrations.

MB@TDN blocker species optimization. The large protein of BSA and small molecule of ethanolamine were chosen to enclose the unreacted COOH on the surface of MB to avoid covalent coupling of antibody to MB. The MB@TDN reaction system was blocked by adding 1 % BSA and 10 % ethanolamine solution respectively for 1 h. The results showed that the absorbance value and inhibition rate of ethanolamine-contained were significantly higher than BSA-contained MB@TDN@Ab immunoprobe (Fig. 4c). The reason that ethanolamine is significantly better than BSA for containment is that BSA is a protein that may cause TDN to interact with the antibody when it is coupled to TDN.

PEG, as a key reagent in immunoreactions, can effectively inhibit the non-specific adsorption between nanomaterials and proteins. The results of Fig. 4d showed that the PEG concentration from 0 ∼ 1 % range was followed by increasing inhibition, and the highest inhibition was observed at PEG = 1 %. Therefore, 1 % PEG concentration was chosen as the subsequent experimental condition.

MB@TDN@Ab immunoprobe concentration and HRP-Hapten concentration optimization. Different concentrations of MB@TDN@Ab immunoprobe (0.1, 0.15, 0.2, 0.3, and 0.5 mg/mL) and 1, 2, 5, 8, and 10 µg/mL HRP-Hapten were added. The results by Fig. 4e-f showed that HRP-Hapten in the range of 1 ∼ 10 µg/mL was not a determining factor for the absorbance. The increasing concentration of MB@TDN@Ab immunoprobe was followed by increasing absorbance values. Absorbance values at greater than 1.0 and the highest inhibition rate were chosen as the optimal concentration combinations. MB@TDN@Ab immunoprobe (0.3 mg/mL) and HRP-Hapten (1 µg/mL) concentrations were used as the optimal concentrations.

Immunoassay based on MB@TDN@Ab immunoprobe

Fig. 5a shows that at an imidacloprid concentration of 50 µg/L, the inhibition of immunoassay based on MB@TDN@Ab immunoprobe was 82 %, whereas the inhibition of MB@Ab was 55 %. These findings indicate that the antibody recognition site of MB@TDN@Ab immunoprobe was more fully exposed, and TDN reduced the spatial site resistance of Hapten-antibody recognition. The results are shown in Fig. 5(b–c). Under the same conditions of MB concentration, antibody concentration, and coupling conditions, the IC₅₀ of immunoassay based on MB@TDN@Ab immunoprobe is as low as 1.40 ng/mL, with a linear range of 0.05 ∼ 50 ng/mL, while that of immunoassay based on MB@Ab immunoprobe was only 2.66 ng/mL, with a linear range of 0.05 ∼ 5 ng/mL. The sensitivity is improved by a factor of 1.9 and the linear range is improved by an order of magnitude. As a result, compared with immunoassay based on MB@Ab immunoprobe, immunoassay based on MB@TDN@Ab immunoprobe has advantages in terms of the linear range and sensitivity.

Fig. 5.

(a) Fitting curves of immunoassays based on MB@Ab and MB@TDN@Ab immunoprobe at different imidacloprid concentrations. (b) Linear regression equation of immunoassay based on MB@Ab immunoprobe. (c) Linear regression equation of immunoassay based on MB@TDN@Ab immunoprobe.

Analysis of the sensitization mechanism

The study was conducted to investigate the efficiency of MB and MB@TDN coupled antibody. This study investigated the efficiency of antibody coupling to MB and MB@TDN. As shown in Fig. 6(a-b), when the amount of antibody exceeded 8 µg, the coupling on the MB@TDN was likely dominated by nonspecific adsorption. In contrast, at antibody concentrations less than 8 µg, covalent coupling between the TDN and the antibody was more prominent. Therefore, 8 µg of antibody was selected as the optimal amount for surface coupling in subsequent affinity studies. To evaluate immunoaffinity, the Hill equation was applied to calculate the K_d, which is inversely correlated with binding affinity [22]. As shown in Fig. 6(d), the K_d values for MB@Ab immunoprobe and MB@TDN@Ab immunoprobe were 19.16 µg/mL and 3.21 µg/mL, respectively, indicating that the MB@TDN@Ab immunoprobe exhibited significantly stronger binding affinity than the conventional MB@Ab immunoprobe.

Fig. 6.

(a) Quality of MB surface-coupled antibody; (b) Quality of MB@TDN (ethanolamine-enclosed) surface-coupled antibody; (c) Quality of MB@TDN (BSA-enclosed) surface-coupled antibody; (d) Comparison of affinity between MB@Ab immunoprobe and MB@TDN@Ab immunoprobe. Note: Background values were deducted during protein supernatant quantification.

The NHS ester was modified at the tip of TDN for antibody coupling by click chemistry, and the NHS ester at the tip of TDN was coupled to the antibody amino group in the same way as the carboxyl group on the MB surface was coupled to the antibody amino group, thus realizing the consistent orientation of the antibody. Based on the consistent orientation of the antibody, the study further elucidated the sensitization mechanism from the perspective of spatial site resistance. The study performed molecular simulations of the variable region of the imidacloprid antibody, and labeled the heavy and light chains of CDR3 in red and yellow, respectively (Fig. 7). CDR3 is located at the recognition Fab region of the antibody, and plays a major role in recognition. The spatial mediation of the TDN achieves the spatial distance between the Fab region and the surface of the MBs in solid phase of 8.95 nm, which allows the full exposure of the Fab of the antibody.

Fig. 7.

Molecular mimicry of the variable region and TDN of the imidacloprid antibody. Note: The heavy and light chains of the CDR3 of the antibody are labeled in red and yellow, respectively.

Selectivity and analysis of imidacloprid in spiked vegetable samples

To further evaluate the applicability of the immunoassay based on MB@TDN@Ab immunoprobe, the researchers chose Chinese cabbage, cucumber and zucchini as the subjects of applied research to confirm the feasibility of the MB@TDN@Ab immunoprobe in the detection of real samples. The absence of imidacloprid in the above samples was confirmed by the LC-MS/MS method. For the MB@TDN@Ab immunoprobe (Fig. 8), the sample matrix was found to have a large impact on the sensitivity of the assay, with the sensitivity of the matrix calibration curve being reduced by one order of magnitude compared to the solvent calibration curve. The possible reason for this is that the matrix in the sample interferes with the recognition of the antibody to HRP-Hapten. First, some components of the matrix may compete with the target antigen for binding to the antibody, reducing the proportion of effective binding sites. In addition, large matrix components (e.g., polysaccharides or proteins) may impede physical contact between the antigen and the antibody, affecting binding efficiency [23]. In Chinese cabbage, cucumber, zucchini matrix, the recoveries were within the acceptable range (Table 2). The recovery experiments verified the accuracy and precision of the assay, and the RSDs of all the results were within the acceptable range according to Chinese standards (NY/T 788–2018, ER: 60 ∼ 120, RSD＜20 %) and European Union standards (Document N°SANTE/12685/2019, ER: 70 ∼ 120, RSD＜20 %) [24,25]. The results indicate that the proposed method can be used to characterize imidacloprid in complex vegetable matrix samples. Recovery experiments showed a high degree of consistency, demonstrating the robustness and reliability of MB@TDN@Ab immunoprobe in quantitative analysis. According the Chinese standards (GB-2763–2021), The LOD of the assay was lower than (maximum residue limits for cabbage, cucumber and zucchini = 0.2 mg/L, 1 mg/L and 1 mg/L) and comply with the maximum residue limits [26].

Fig. 8.

Immunoassay based on MB@TDN@Ab immunoprobe in three vegetables (Chinese cabbage, cucumber and zucchini).

Table 2.

Detection and recovery of imidacloprid in spiked vegetable samples.

Sample	Spiked sample concentration (μg/L)	Found (μg/L)	Recovery (%)	RSD (%)
Chinese cabbage	0.5	0.33	66.5	3.2
	1	0.67	67.4	1.8
	5	4.57	91.4	2.8
Cucumber	0.5	0.35	69.3	6.6
	1	0.69	69.0	4.7
	5	4.62	92.4	3.0
Zucchini	0.5	0.36	72.6	2.8
	1	0.67	67.3	4.4
	5	3.79	75.9	6.6

The results in Fig. 9a show that MB@TDN@Ab has the highest specificity for imidacloprid, indicating the high selectivity of this proposed method. The results of Fig. 9b show that the absorbance value of MB@TDN@Ab probe was stable at 0.75 under 4 °C conditions and the same experimental conditions proving the good stability of the probe over 10 d. Furthermore, the proposed immunoassay based on MB@TDN@Ab immunoprobe was compared with other detections of imidacloprid. As shown in Table 3, the proposed immunoassay based on MB@TDN@Ab immunoprobe has a very low LOD. The limit of detection (LOD) for immune-competitive assays is often defined using an inhibition rate of 10 % (IC₁₀) [27]. In particular, the introduction of TDN improves the sensitivity compared to randomly labeled antibody and the specific recognition of the antibody.

Fig. 9.

(a) Specificity evaluation of MB@TDN@Ab immunoprobe (b) Stability evaluation of MB@TDN@Ab immunoprobe at 4℃ for 10 days.

Table 3.

Comparison of the published methods for imidacloprid detection.

Method	Linear range (μg/L)	LOD (μg/L)	Reference
HPLC	10.0–3 × 103	3.19	[28]
Zr(Ⅳ) Metal organic framework sensor	—	5.57	[29]
Graphene quantum dot/Au chemosensor	10.0 ∼ 102	6.98	[30]
Electrochemistry	7.67 ∼ 3.06 × 103	2.30	[31]
Electrochemistry	51.13 ∼ 452.51	9.56	[32]
Nanofluidic biosensor	10.0 ∼ 105	6.28	[33]
SERS sensors	1.53 ∼ 102.26	0.21	[34]
QDs-SA-FLISA	0.50 ∼ 5.72	0.5	[35]
FPIA	1.7 ∼ 16.3	4.8	[36]
IFE immunoassay	1.39 ∼ 335.81	18.7	[37]
UCNPs-based strips	0.97 ∼ 250	8.37	[38]
CDs-FL	37.0 ∼ 200	1.87	[39]
MB@TDN@Ab	0.05 ∼ 50.0	0.063	This work
MB @Ab	0.05 ∼ 5.0	0.087	This work
dc-ELISA	5.0 ∼ 200	15.13	This work

Conclusion

The MB@TDN@Ab immunoprobe acquired a higher affinity constant (K_d = 3.21 µg/mL) than MB@Ab immunoprobe (K_d = 19.16 µg/mL). A highly sensitive MB@TDN@Ab immunoprobe was constructed. A convenient, efficient and sensitive immunoassay based on MB@TDN@Ab immunoprobe was established for detecting imidacloprid in vegetables in the study. Under the optimal conditions, the sensitivity IC₅₀ of immunoassay based on MB@TDN@Ab immunoprobe was as low as 1.40 ng/mL, with a linear range of 0.05 ∼ 50 ng/mL, whereas that of immunoassay based on MB@ Ab immunoprobe was 2.66 ng/mL, with a linear range of 0.05 ∼ 5 ng/mL. The sensitivity of MB@TDN@Ab was enhanced by 1.9-fold, and the linear range was increased by one order of magnitude. Thence, the immunoassay based on TDN spatial mediation broaden the avenue for pesticides detection. These findings contribute to sustainable agricultural safety, public health assurance, and innovation in biosensing technologies.

Although the proposed immunoassay based on TDN spatial mediation has made a significant progress for pesticide residue detection, there is still a minor limitation with the low yield of TDN in this study. The Python language program and the synthesis conditions for TDN need to be optimized, improving TDN yield in future.

Compliance with ethics requirements

This article does not contain any studies with human or animal subjects.

Funding

This work has been supported by the National Natural Science Foundation of China (32302207, 32402242), the National Key Research and Development Program of China (2024YFD2000603), and the National Center of Technology Innovation for Comprehensive Utilization of Saline-Alkali (GYJ2023004).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Data availability

Data will be made available on request.