Activity-specificity balanced Au@Pt nanozyme immunoassay via Dual-Au deposition for enhanced mycotoxin detection

Abstract

Mycotoxins persistently threaten global food and herbal medicine safety, demanding the urgent development of detection methods that integrate high sensitivity with strong specificity. Conventional nanozyme-based immunoassays, however, are constrained by the inherent trade-off between sensitivity and specificity, limiting their performance. In this study, we propose a dual-gold enhanced Au@Pt nanozyme-linked immunosorbent assay (Au@Pt/Au-ELISA), which leverages a rationally engineered sequential optimization paradigm to fundamentally overcome the intrinsic trade-off between sensitivity and specificity. The method innovatively integrates mild gold deposition for background suppression with targeted catalytic site amplification, thereby maximizing the catalytic potential of Au@Pt nanozymes while preserving antibody recognition. The dual-stage design significantly enhances nanozyme activity, reducing the usage of antibody probes to 14.3% and 33.3% of that in the original scheme for zearalenone (ZEN) and aflatoxin B₁ (AFB₁) detection, respectively. As a secondary outcome, the assay maintains excellent analytical sensitivity, achieving half-maximal inhibitory concentrations of 42 pg/mL for ZEN and 28 pg/mL for AFB₁, corresponding to 7.2-fold and 2.0-fold enhancements over conventional ELISA, respectively. Importantly, the method exhibits high quantitative accuracy in complex matrices, with ZEN recoveries of 109.6-119.5% in coix seeds and AFB₁ recoveries of 100.2-113.2% in lotus seeds. Limited cross-reactivity toward related mycotoxins confirms the high immunospecificity of the assay, which shows strong agreement with LC-MS/MS results (R > 0.99). By harmonizing catalytic activity with immunospecificity, this cost-effective platform not only provides a refined design strategy for nanozyme-based immunoassays but also establishes a solid technical foundation for high-performance mycotoxin screening in complex food and herbal medicine matrices.

Graphical abstract

TOC: This study presents a dual-step gold deposition strategy that effectively overcomes the trade-off between probe activity and specificity in Au@Pt nanozyme-based immunoassays, enabling ultrasensitive, highly specific, and broad-spectrum detection of multiple mycotoxins.

Highlights

• •Dual-Au enhanced Au@Pt nanozyme immunoassay breaks sensitivity-specificity trade-off.
• •Gold deposition reduces probe use to 14.3% of that in the unenhanced system.
• •Boosts sensitivity for ZEN/AFB₁ up to 7.2-fold via dual catalytic enhancement.
• •Robust matrix tolerance demonstrated in complex food samples.
• •Excellent consistency with LC-MS/MS confirms accuracy in real sample detection.

1. Introduction

Mycotoxins are secondary metabolites produced by fungi during the metabolism of grains and their derived products, and are widely present in cereals, feed, and related agricultural commodities (Nešić et al., 2021). Due to their strong carcinogenicity, teratogenicity, and immunotoxicity, long-term intake poses a serious threat to human and animal health, potentially leading to acute poisoning or chronic diseases (Yu and Pedroso, 2023). As a result, mycotoxins have become a pressing public health concern in the field of global food safety. To mitigate the harm caused by mycotoxins, regulatory agencies in various countries have progressively tightened their maximum residue limits, which has driven growing demand for rapid, sensitive, and reliable detection methods (Shekhar et al., 2025; Singh and Mehta, 2020) .

The Enzyme-Linked Immunosorbent Assay (ELISA), pioneered by Engvall and Perlmann in 1971 (Eva and Perlmann, 1971), remains a cornerstone technology in immunoassay applications due to its unparalleled technical merits. Three fundamental advantages underpin its enduring value: (1) The heterogeneous detection system integrating solid-phase carriers with enzyme-labeled antibodies demonstrates matrix tolerance for complex biological samples, achieved through synergistic antigen-antibody specificity and physical separation (Waart et al., 2005; Wolters et al., 1976); (2) The enzymatic signal amplification mechanism enables sensitive detection while maintaining procedural simplicity (Gosling, 1990; Saud et al., 2019); (3) Standardized microplate formats facilitate industrial-scale high-throughput screening (Kingsmore, 2006). Current technological innovation primarily focuses on optimizing enzymatic signal amplification systems - the pivotal determinant of ELISA sensitivity. In this context, nanozymes (enzyme-mimicking nanomaterials) present transformative potential by addressing intrinsic limitations of natural enzymes (e.g., thermal lability and production costs) through precisely engineered catalytic sites, remarkable environmental stability, and scalable fabrication (Wu et al., 2019).

Among various nanozymes, gold-platinum core-shell nanoparticles (Au@Pt) have been widely applied for signal amplification in immunoassays owing to their outstanding peroxidase-mimicking catalytic activity (Hendrickson et al., 2022; Zhou et al., 2022). However, a key technical challenge persists in ELISA applications: the essential antibody conjugation for target recognition, followed by blocking steps to reduce nonspecific binding, unavoidably obstructs some catalytic active sites on Au@Pt nanoparticles (Long et al., 2018; Panferov et al., 2021; Que et al., 2014). Despite its significant impact on assay performance, this phenomenon has received inadequate attention in the literature. While current approaches can achieve acceptable sensitivity, the compromised signal amplification efficiency due to enzymatic inhibition presents non-negligible consequences, particularly the requirement for excessive antibody consumption that substantially increases detection costs (Huang et al., 2023). Therefore, maximizing the utilization efficiency of nanozyme catalytic activity constitutes a paramount research priority in labeled immunoassays (Luo et al., 2024; Zhang et al., 2024). To address this issue, we have attempted to adopt a gold enhancement strategy (Wang et al., 2016), where gold ions are deposited onto the surface of Au@Pt nanoparticles to restore and amplify their catalytic activity. However, under conventional gold enhancement conditions, the rapid deposition of gold not only leads to the formation of orange-red precipitates on the well walls but also results in minimal color difference between groups with different analyte concentrations, indicating a problem of nonspecific signal amplification that compromises detection accuracy. Upon adjusting the concentration of chloroauric acid, it was found that when the concentration was reduced to 0.25 mM, nonspecific deposition was significantly suppressed and detection specificity improved. However, the colorimetric signal was notably weakened, suggesting that under low gold ion concentration, the deposition rate decreased, leading to insufficient exposure of catalytic sites and ultimately restricting the signal amplification effect.

To balance signal intensity and detection specificity, this study proposed and optimized a “two-step induced gold enhancement” strategy (see Scheme 1). In the first step, a low-concentration gold ion solution is used for mild initial deposition to reduce nonspecific interference; in the second step, a freshly prepared enhancement solution of the same concentration is applied to promote further exposure of catalytic sites and enhance the signal. This strategy effectively overcomes the nonspecific deposition issue encountered in traditional gold enhancement processes and significantly improves detection sensitivity. In this study, the method was specifically applied to detect zearalenone (ZEN) in coix seeds and aflatoxin B₁ (AFB₁) in lotus seeds, representing typical mycotoxin contamination scenarios in food and herbal medicine matrices. By precisely regulating the gold deposition process, the two-step gold enhancement strategy successfully balances catalytic activity recovery and detection specificity, addressing the long-standing “activity-specificity” contradiction of Au@Pt in immunoassays, and expanding its application potential in highly sensitive immunodetection.

Scheme 1.

Schematic illustration of the two-step induced gold enhancement strategy and its application in Au@Pt/Au-ELISA for mycotoxin detection. (a) Design and construction of the Au@Pt/Au-ELISA based on the two-step gold enhancement process. (b) Detection workflow illustrating the specific recognition and sensitive quantification of mycotoxins using the two-step gold enhancement induced Au@Pt/Au-ELISA.

2. Results and discussion

2.1. Design and validation of a two-step gold enhancement strategy in Au@Pt/Au-ELISA

In this study, Au@Pt nanoparticles were successfully synthesized and systematically characterized in terms of their morphology and elemental composition. Transmission electron microscopy (TEM) revealed that the nanoparticles were uniformly distributed, with an average diameter of 12.12 nm (Fig. 1a). High-angle annular dark-field scanning transmission electron microscopy coupled with energy-dispersive X-ray spectroscopy (HAADF-STEM-EDS) further confirmed the bimetallic structure of Au@Pt, with element mapping indicating a homogeneous distribution of gold and platinum (Fig. 1b and c). To explore its potential application in immunoassays, Au@Pt was conjugated with ZEN monoclonal antibodies to construct an immunoprobe, and its peroxidase-mimicking activity was evaluated (Fig. 1d). However, ELISA results demonstrated a significant decrease in catalytic activity after antibody conjugation and blocking, which may result from steric hindrance, surface coverage of active sites, and alterations in the nanozyme microenvironment that collectively reduce substrate accessibility.

Fig. 1.

Structural analysis of Au@Pt nanozymes and performance optimization via two-step induced gold enhancement. (a) TEM image and size distribution of Au@Pt nanoparticles (scale bar: 20 nm). (b) HAADF-STEM-EDS mapping confirming uniform distribution of Au (red) and Pt (green) in Au@Pt (scale bar: 10 nm). (c) Quantitative elemental composition of Au and Pt obtained from EDS mapping in (b). (d) UV-vis spectra comparing peroxidase-like activity of different Au@Pt conjugates. (e) Microplate images showing specific gold deposition at various HAuCl₄ concentrations. (f) One-step induced gold enhancement at 0.25 mM HAuCl₄ with increasing ZEN concentrations (0.25-1 ng/mL). (g) Two-step induced gold enhancement at 0.25 mM HAuCl₄ with increasing ZEN concentrations (0.25-1 ng/mL). (h) Quantitative analysis shown in figures f and g. (i) Overall signal amplification comparison across no enhancement, one-step, and two-step methods. Data were presented as means ± SD for n = 3. ns (nonsignificant) = p > 0.05, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

To restore the catalytic activity of the Au@Pt probe, a gold enhancement strategy was employed. However, following the method reported in the literature (Wang et al., 2016), after competitive binding between different concentrations of the control analyte and the immobilized antigen with the Au@Pt probe, gold enhancement resulted in the formation of orange-red deposits on the walls of the microplate wells. Furthermore, all groups exhibited similar color intensities, indicating that the assay failed to differentiate between different concentrations of the control analyte. These findings suggest that excessive gold deposition led to significant nonspecific enhancement, thereby compromising assay specificity. Further investigation revealed that as the concentration of chloroauric acid decreased, the formation of orange-red deposits gradually diminished, and no such deposition was observed when the concentration was reduced to 0.25 mM (Fig. 1e). Under this condition, the color intensity decreased with increasing concentrations of the control analyte, demonstrating that assay specificity was restored (Fig. 1f). However, the color intensity was significantly reduced under low chloroauric acid conditions, possibly due to the slower gold deposition rate, which limited the exposure of catalytic sites and impaired both catalytic efficiency and signal amplification.

To enhance color intensity while maintaining specificity, we proposed a two-step induced gold enhancement strategy, in which a second round of gold enhancement was performed using freshly prepared 0.25 mM enhancement solution after the initial enhancement step. Experimental results showed that compared to the one-step enhancement, the two-step induced gold enhancement not only successfully prevented nonspecific deposition but also significantly improved color intensity and enhanced assay sensitivity (Fig. 1g). Further quantitative analysis (Fig. 1h) revealed that the color signal generated by the two-step enhancement was significantly higher than that of the single-step induced gold enhancement and exhibited greater differentiation across varying control analyte concentrations. Additionally, in the colorimetric comparison experiment (Fig. 1i), the color intensity obtained with the two-step induced gold enhancement strategy was approximately sevenfold higher than that without enhancement and threefold higher than that of the one-step enhancement, demonstrating a remarkable improvement in signal amplification. These findings indicate that the optimized two-step induced gold enhancement strategy effectively prevents nonspecific deposition while significantly enhancing the catalytic performance of the Au@Pt probe, thereby achieving stronger colorimetric signals.

Mechanistically, the advantage of the two-step induced gold enhancement strategy likely stems from the precise regulation of the gold deposition process. In high concentrations one-step induced gold enhancement, high concentrations of chloroauric acid induce rapid gold deposition, potentially leading to the formation of an uneven or excessively thick gold layer on the Au@Pt surface. This, in turn, reduces accessibility to the catalytic sites and contributes to nonspecific color development. Such overgrowth may also passivate the nanozyme surface, leading to decreased peroxidase-mimicking activity, as previously reported (Long et al., 2018). In contrast, under low-concentration conditions, gold deposition occurs at a slower and more uniform rate, effectively preventing nonspecific color development, albeit at the cost of reduced catalytic activity. The two-step induced gold enhancement strategy builds upon the initial mild deposition and provides an additional gold source in the second step, thereby improving the exposure of catalytic sites while maintaining uniform deposition. Specifically, the first step stabilizes the core-shell structure without excessive coverage, while the second step reinforces signal generation via controlled growth, this design mimics seed-mediated growth strategies used in nanomaterial synthesis for precise morphological control (Xiong and Xia, 2007). This synergistic effect leads to enhanced catalytic activity and improved signal amplification. Similar dual-phase enhancement strategies have been shown to improve sensitivity in plasmonic and enzymatic biosensing systems (He et al., 2023), validating the efficacy of this approach. The proposed strategy not only enhances the applicability of Au@Pt probes in immunoassays but also provides new insights into optimizing immunoanalytical methods.

2.2. Effect of gold enhancement on the structure and catalytic performance of Au@Pt

TEM characterization revealed that the average particle size of Au@Pt nanostructures increased from 12.12 nm to 12.72 nm after the first-step induced gold enhancement and further to 14.19 nm after the second-step induced gold enhancement (Fig. 2a and b). Because the size difference before and after enhancement is relatively small and close to the resolution and edge-determination limits of TEM analysis, the exact magnitude of the diameter increase is difficult to measure with high precision and should be interpreted mainly as a growth trend. Therefore, compositional analysis was further used to verify the occurrence of gold deposition. HAADF-STEM-EDS mapping (Fig. 2c) confirmed the well-maintained core-shell structure and the homogeneous distribution of gold and platinum. Quantitative EDS analysis further demonstrated an increase in gold content from 15.22% (Fig. 1c) before enhancement to 17.36% (Fig. S1a and b) and 20.37% (Fig. 2d) after the first and second enhancement steps, accompanied by a relative decrease in platinum content. These compositional results provide additional support for successful gold deposition and highlight the ability of the two-step induced gold enhancement strategy to effectively modulate the gold coverage.

Fig. 2.

Impact of two-step induced gold enhancement on Au@Pt structure and catalytic activity. (a) TEM image and size distribution of Au@Pt nanoparticles after first-step induced gold enhancement (scale bar: 20 nm). (b) TEM image and size distribution of Au@Pt nanoparticles after second-step induced gold enhancement (scale bar: 20 nm). (c) HAADF-STEM-EDS mapping confirming showing core-shell structure and uniform distribution of Au (red) and Pt (green) in two-step induced gold enhancement Au@Pt (scale bar: 10 nm). (d) Quantitative atomic composition of Au and Pt from EDS analysis in (c), confirming increased Au content after enhancement. (e) XPS survey spectra of Au@Pt ((conjugated with ZEN antibody)) before and after two-step induced gold enhancement. (f) High-resolution XPS spectra of Au 4f region, showing binding energy shifts after two-step gold enhancement. (g) High-resolution XPS spectra of Pt 4f region, showing slight peak shifts indicating electronic structure modulation. (h) Michaelis-Menten curve of one-step induced gold enhancement Au@Pt for H₂O₂ decomposition (K_m = 21.36 mM). (i) Michaelis-Menten curve of two-step induced gold enhancement Au@Pt for H₂O₂ decomposition (K_m = 20.48 mM). Data were presented as means ± SD for n = 3.

To further investigate the structural and electronic changes induced by gold enhancement, X-ray photoelectron spectroscopy (XPS) analyses were performed. The XPS survey spectra (Fig. 2e) confirmed that the main elemental components of the material (C, O, Au, and Pt) remained unchanged after enhancement, indicating the absence of impurities. The XPS narrow-scan spectra provided further insights into the electronic structure modifications. Prior to enhancement, the binding energies of Pt 4f_7/2 and Pt 4f_5/2 were 71.3 eV and 74.7 eV, respectively, corresponding to Pt (0) (Fig. 2f). After enhancement, these binding energies slightly decreased to 71.1 eV and 74.4 eV, respectively. Similarly, in the Au XPS spectra (Fig. 2g), the binding energies of Au 4f_7/2 and Au 4f_5/2 decreased from 84.2 eV to 87.8 eV to 84.0 eV and 87.5 eV, respectively. Although the binding energy shifts are small (0.2-0.3 eV), they fall beyond instrumental uncertainty. Such a downshift reflects a slight electron transfer from Pt to Au, which optimizes the electronic coupling at the Au/Pt/Au interface. This charge redistribution is beneficial for catalytic performance, as it promotes charge delocalization and accelerates electron transport during the redox process, thereby lowering the catalytic energy barrier and improving the turnover efficiency.

To assess the impact of gold enhancement on the catalytic performance of Au@Pt, we determined the Michaelis-Menten constant (K_m) for the decomposition of H₂O₂ and the oxidation of TMB. In our previous study, the K_m value for H₂O₂ of the protein-coated Au@Pt was determined to be 30.7 mM (Zhang et al., 2024). After the first-step induced gold enhancement, the K_m value decreased to 21.36 mM (Fig. 2h and Fig. S1f), and following the second-step induced gold enhancement, it further declined to 20.48 mM (Fig. 2i and Fig. S1c). This trend indicates an improvement in substrate affinity and catalytic efficiency due to the gold-induced electronic structure modulation. Similarly, for TMB oxidation, the K_m value before enhancement was previously reported as 1.09 mM. After the first-step induced gold enhancement, it decreased to 0.99 mM, and no further change was observed after the second-step induced gold enhancement (Fig. S1d-e and g-h). The consistency of the K_m value for TMB after the first induced gold enhancement suggests that the structural and electronic modifications predominantly influence H₂O₂ decomposition rather than TMB oxidation, likely due to differences in catalytic mechanisms. This is consistent with previous studies reporting that Au@Pt nanozymes primarily facilitate the breakdown of H₂O₂ through Fenton-like or surface-mediated redox pathways, while TMB oxidation often depends on the availability of reactive oxygen species generated during H₂O₂ catalysis (Wu et al., 2019). The observed reduction in K_m values highlight the effectiveness of the gold enhancement strategy in optimizing the catalytic properties of Au@Pt, particularly in H₂O₂ decomposition. This decrease in K_m indicates an improved affinity of the nanozyme surface toward H₂O₂, possibly due to increased surface area or more exposed active sites following gold deposition (Wu et al., 2019). Similar findings were reported by Li et al., where controlled Au overgrowth enhanced peroxidase-like activity of CuCo₂S₄ nanozymes by modulating their surface electron density and catalytic interfaces (Li et al., 2020).

Overall, these results demonstrate that the gold enhancement strategy effectively increases gold deposition, modulates the electronic structure of Pt through gold-induced charge transfer from Pt to Au, and enhances catalytic activity, particularly in H₂O₂ oxidation. This approach provides a promising strategy for optimizing Au@Pt-based nanocatalysts, offering potential applications in biocatalysis and sensing.

2.3. Construction and optimization of the two-step induced gold-enhanced Au@Pt/Au-ELISA

To further improve the catalytic performance and detection sensitivity of Au@Pt nanozymes in immunoassays, a two-step induced gold enhancement Au@Pt-ELISA strategy was developed (Fig. 3a), and its key reaction conditions were systematically optimized. This approach expands the active surface area of Au@Pt through secondary gold deposition, significantly boosting its catalytic activity and enabling highly sensitive detection of target analytes.

Fig. 3.

Construction and optimization of the two-step induced gold enhancement Au@Pt/Au-based ELISA. (a) Schematic illustration showing how different gold enhancement conditions affect signal intensity and detection specificity. (b) Effect of gold enhancement temperature on absorbance and inhibition rate. (c) Optimization of enhancement reaction time. (d) Effect of HAuCl₄ concentration on signal intensity and inhibition rate. (e) Effect of TMB concentration on catalytic signal intensity. (f) Effect of H₂O₂ concentration on catalytic signal intensity. (g) Comparison of standard curves obtained from traditional ELISA, Au@Pt-ELISA, and induced gold enhancement Au@Pt/Au-ELISA. (B/B₀ denotes the inhibition rate, where B₀ is the OD value at 0 ng/mL ZEN and B is the OD value at 0.063 ng/mL ZEN) Data were presented as means ± SD for n = 3.

The effect of the enhancement temperature on both the catalytic signal and inhibition rate was first investigated (Fig. 3b). The results showed that the signal gradually increased with temperature, peaking at 30 °C, and then declined at higher temperatures. The inhibition rate exhibited an opposite trend. These findings suggest that moderate temperatures promote uniform gold deposition, which improves catalytic efficiency, whereas excessive temperatures may lead to nanoparticle aggregation or morphological distortion, ultimately reducing catalytic activity and detection stability. The enhancement reaction time was then optimized (Fig. 3c). When the reaction time reached 5 min, the catalytic signal peaked, indicating that gold deposition was essentially complete and the catalytic sites were optimally exposed. Prolonging the reaction beyond this point resulted in a signal decrease, likely due to excessive growth of the gold shell, which may cover active catalytic sites and hinder substrate accessibility.

Next, the effect of gold ion concentration in the enhancement solution was examined (Fig. 3d). Increasing the concentration led to a progressive rise in the catalytic signal, but the inhibition rate began to increase when the concentration exceeded 0.25 mM. This phenomenon can be attributed to the fact that higher gold ion concentrations not only enlarge the catalytic surface but may also trigger nonspecific signal amplification, which compromises the accuracy of detection. Taking both catalytic activity and specificity into account, 0.25 mM was selected as the optimal gold ion concentration for subsequent experiments. After determining the gold enhancement conditions, the substrate concentrations were further optimized to fully leverage the catalytic potential of Au@Pt nanozymes. As shown in Fig. 3e, the catalytic signal plateaued when the TMB concentration reached 0.6 mM, indicating substrate saturation. Similarly, the H₂O₂ optimization (Fig. 3f) showed that the signal stabilized when the concentration reached 0.9 M. These results are consistent with the previously determined K_m values, which reflect the substrate affinity of the system, confirming that the optimized substrate concentrations meet the catalytic requirements under the enhanced conditions.

Under these optimized conditions, standard curves were constructed to compare the sensitivity of traditional ELISA, Au@Pt-ELISA, and the two-step induced gold enhancement Au@Pt-ELISA (Fig. 3g). The introduction of Au@Pt nanozymes improved the detection sensitivity (half-maximal inhibitory concentration, IC₅₀) from 304 pg/mL to 57 pg/mL, representing a 5.3-fold enhancement over conventional ELISA. This performance was further elevated by a two-step gold enhancement, which pushed the sensitivity to 42 pg/mL, achieving an approximately 7.2-fold improvement. Notably, the final limit of detection (LOD) reached 0.009 ng/mL, representing a 225-fold improvement in sensitivity compared to a prior reported method for detecting ZEN in coix seeds with an LOD of 2.03 ng/mL (Liu et al., 2021). Beyond sensitivity, the gold enhancement strategy substantially boosted probe utilization efficiency. The antibody probes tolerated a 25-fold dilution, enabling a reduction in antibody probe usage to 14.3% of that in the non-enhanced system. Compared to our prior nanoprobe construction method (Huang et al., 2023) using gold nanoparticles as dual carriers for ZEN antibody and HRP co-labeling, the current approach is significantly simpler by requiring only antibody labeling while eliminating HRP conjugation. This streamlined process not only achieves comparable detection sensitivity but also increases the probe's dilution capacity by 11-fold, substantially enhancing overall practicality. Collectively, these results demonstrate that the two-step gold enhancement effectively amplifies the catalytic signal while ensuring detection specificity, offering a reliable and cost-effective platform for high-sensitivity immunoassays.

2.4. Application of the two-step induced gold enhancement Au@Pt/Au-ELISA for ZEN detection in coix seed samples

To evaluate the practical applicability of the two-step induced gold enhancement Au@Pt/Au-ELISA, this method was employed to detect ZEN in coix seed samples, with systematic optimization of the sample pretreatment process. As illustrated in Fig. 4a, the samples were pretreated using a liquid extraction procedure, and the extraction efficiency of different solvents was compared (Fig. 4b). The results showed that 70% methanol-water exhibited significantly higher extraction efficiency than other methanol concentrations and acetonitrile-based solvents, indicating that this composition offers an optimal balance between solubility and extraction efficiency for ZEN.

Fig. 4.

Application of the two-step induced gold enhancement Au@Pt/Au-based ELISA for ZEN detection in coix seed. (a) Schematic workflow for sample pretreatment and two-step induced gold enhancement Au@Pt/Au-ELISA detection. (b) Extraction efficiency of different solvents for ZEN in coix seed. (c) Effect of different dilution solvents on matrix interference. (d) Solvent-based and matrix-matched standard curves for ZEN detection. (e) Linear fitting of the curves in (d) to evaluate matrix effect. (f) Cross-reactivity evaluation with ZEN analogs. (g) Correlation of detection results between two-step induced gold enhancement Au@Pt/Au-ELISA and LC-MS/MS. (B/B₀ denotes the inhibition rate, where B₀ is the OD value at 0 ng/mL ZEN and B is the OD value at 0.063 ng/mL ZEN) Data were presented as means ± SD for n = 3. ns (nonsignificant) = p > 0.05, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

To further reduce matrix interference, the dilution solvent was optimized (Fig. 4c). It was found that 10% methanol-water containing 0.01% Tween-20 could effectively minimize matrix effects and improve the reliability and stability of the assay. Subsequently, a solvent-based standard curve and a matrix-matched standard curve (Fig. 4d and e) were constructed. Both curves exhibited excellent linearity, with correlation coefficients exceeding 0.99, indicating the method's strong accuracy even in complex sample matrices. The matrix effect was further evaluated by comparing the slopes of the two standard curves, which showed a ratio of 98.3%, falling within the acceptable range of 85%-115%. Additionally, the absorbance ratio of the negative sample solution to the blank solvent was 0.953, confirming that matrix interference was well controlled. The method's accuracy was further validated through recovery experiments. Spiked recovery rates ranged from 109.6% to 119.5%, with relative standard deviations (RSDs) below 11.9% (Table S3), demonstrating good precision and reliability for quantitative analysis in coix seed matrices. The specificity of the assay was also investigated by evaluating potential cross-reactivity with ZEN analogs and other common mycotoxins. As shown in Fig. 4f, the cross-reactivity rates for α-zearalenol and β-zearalenol were 36.2% and 19.7%, respectively. Moreover, no significant cross-reactivity was observed with AFB₁, fumonisin B₁ (FB₁), or ochratoxin A (OTA), confirming the assay's strong specificity for ZEN.

The ZEN concentrations in coix seed samples determined by the two-step induced gold enhancement Au@Pt/Au-ELISA were compared with those measured by LC-MS/MS (Fig. 4g). The results demonstrated excellent agreement between the two methods, with a correlation coefficient of R = 0.993, highlighting the method's accuracy and its great potential for the reliable detection of ZEN in coix seed samples. Representative LC-MS/MS chromatograms are provided in Fig. S2.

2.5. Application of two-step induced gold enhancement Au@Pt/Au-ELISA for the detection of AFB₁ in lotus seed samples

Based on the validation of the aforementioned strategy, this study further explored the applicability of the two-step induced gold enhancement Au@Pt/Au-ELISA for detecting AFB₁ in lotus seed samples. First, Au@Pt nanozymes were conjugated with AFB₁ monoclonal antibodies to prepare immunoprobes, and their peroxidase-like catalytic activity was evaluated. The results (Fig. S3a) demonstrated that the nanozymes retained good catalytic performance after conjugation. Additionally, a colorimetric comparison experiment (Fig. S3b) further confirmed the advantages of the two-step induced gold enhancement strategy, which exhibited a more pronounced signal amplification effect compared to single-step induced gold enhancement and no enhancement.

During the optimization of conjugation conditions, the coupling efficiency between the antibody and Au@Pt was mainly affected by the pH of the system and the antibody concentration (Ruiz et al., 2019). The pH optimization results (Fig. 5a) showed that as the pH increased, the surface charge state of the antibody changed, affecting its adsorption orientation on the Au@Pt surface. The colorimetric value peaked at pH 8.0, indicating the highest coupling efficiency under this condition, which is beneficial for improving detection signal. This phenomenon is consistent with findings from AuNP-related studies in the literature (Zeng et al., 2019), where pH modulates the antibody's surface charge state and spatial conformation. The optimization of antibody dosage (Fig. 5b) showed that as the antibody amount increased, the colorimetric value increased, while the inhibition rate first decreased and then increased. Finally, the optimal amount was determined to be 0.5 μL of antibody at a concentration of 1 mg/mL.

Fig. 5.

Two-step induced gold enhancement Au@Pt/Au-ELISA for AFB₁ detection in lotus seed samples. (a) Optimization of pH for efficient antibody-Au@Pt conjugation. (b) Effect of antibody dosage on colorimetric signal and inhibition rate. (c) Influence of enhancement temperature on catalytic activity and specificity. (d) Optimization of HAuCl₄ concentration for maximum enhancement. (e) Effect of enhancement time on signal intensity and assay performance. (f) Comparison of assay sensitivity between gold enhancement Au@Pt/Au-ELISA and conventional ELISA. (g) Solvent-based and matrix-matched standard curves for AFB₁ detection. (h) Linear fitting of the curves in (g) to evaluate matrix effect. (i) Correlation analysis between two-step induced gold enhancement Au@Pt/Au-ELISA and LC-MS/MS results for AFB₁ detection in real samples. (B/B₀ denotes the inhibition rate, where B₀ is the OD value at 0 ng/mL AFB₁ and B is the OD value at 0.063 ng/mL AFB₁) Data were presented as means ± SD for n = 3.

In the enhancement step, temperature significantly influenced the gold enhancement effect (Fig. 5c). The colorimetric value decreased with increasing temperature, while the inhibition rate was highest at −20 °C, suggesting that low temperatures favor the specificity of the enhancement reaction. Notably, the optimal enhancement temperature for AFB₁ detection (−20 °C, Fig. 5c) was much lower than that for ZEN detection (30 °C, Fig. 3b), despite both systems using identical enhancement chemistries and showing similar XPS profiles. It should be emphasized that, for the AFB₁ system, the analytical sensitivity obtained at 30 °C is essentially comparable to that at −20 °C, with only a slight decrease in color intensity, which does not result in a meaningful loss of quantitative performance. This divergence highlights the critical role of conjugated protein properties (such as hydrophilicity, surface coverage, steric hindrance, and interaction with gold precursors) in shaping the nanozyme microenvironment. Protein coronas and structural variations can modulate surface charge, steric effects, and local environments, thereby affecting nanoparticle nucleation and metal ion reduction kinetics (Monopoli et al., 2012; Tenzer et al., 2013). Distinct tertiary structures or charge distributions of AFB₁ versus ZEN antibodies may alter Au³⁺ accessibility or activation energy for gold deposition, consistent with reports that antibody conformation and orientation influence nanoparticle morphology and activity (Yong et al., 2020), making the enhancement more temperature-sensitive in AFB₁ systems. Temperature further modulates protein-gold interactions and nucleation thermodynamics, so even minor antibody structural differences can cause distinct thermal sensitivities (Li et al., 2021). To further evaluate the general applicability of the dual gold-enhancement strategy, we additionally examined a third mycotoxin (sterigmatocystin, ST) detection system. The ST antibody system showed a trend similar to AFB₁, with lower temperatures generally enhancing color development, but the highest inhibition rate was achieved at 30 °C (Fig. S5), consistent with ZEN. These results indicate that although the temperature–color response profiles differ among antibody systems, 30 °C can provide stable and reliable detection across multiple targets. Based on these experimental observations, it remains challenging to precisely predict the optimal enhancement temperature for a new antibody system using a single model, and limited condition screening is therefore still recommended. However, from a practical perspective, when the absolute maximum color intensity is not the sole objective, 30 °C serves as a generally applicable enhancement temperature, ensuring robust detection across ZEN, AFB₁, and ST systems.

Following optimization of detection parameters, the enhancer concentration and reaction time were set at 0.25 mM and 10 min, respectively, to maximize signal amplification (Fig. 5d and e). Under these conditions, gold-enhanced Au@Pt/Au-ELISA outperformed conventional ELISA (Fig. 5f), delivering an IC₅₀ of 28 pg/mL versus 57 pg/mL, thereby underscoring the superior sensitivity conferred by the two-step gold enhancement strategy. Furthermore, antibody probes usage was reduced to 33.3% of that in the non-enhanced system. Based on our previous research on AFB₁ detection using Au@Pt nanozymes (Zhang et al., 2024), the introduction of nano-platinum was found to enhance the activity of antibody-modified Au@Pt nanozyme probes. However, its synthesis and application processes were separate, increasing operational complexity. In contrast, the in-situ gold enhancement strategy proposed in this study features a more streamlined procedure. While maintaining the comparable high detection sensitivity to the prior research (IC₅₀ of 24 pg/mL), this approach reduces antibody consumption by 58%, demonstrating improvements in both efficiency and cost-effectiveness.

The method also demonstrated good matrix tolerance and high recoveries (100.2%-113.2%) with RSDs below 10.4% in spiked lotus seed samples (Fig. 5g and h), supporting its robustness for quantitative analysis in complex matrices. In addition, the excellent batch-to-batch reproducibility as well as short- and long-term storage stability of the Au@Pt probes (Fig. S6) further support the practical applicability of the assay. Finally, the two-step induced gold enhancement Au@Pt/Au-ELISA was compared with LC-MS/MS for AFB₁ detection in lotus seed samples. Correlation analysis showed a high consistency between the two methods, with a correlation coefficient (R) of 0.992 (Fig. 5i), further confirming the accuracy and reliability of this method. Representative LC-MS/MS chromatograms are shown in Fig. S7, further supporting the potential application of this method in real sample detection.

Details on structural characterization (e.g., TEM/XPS), enhancer dosage and time optimization, substrate concentration optimization for the colorimetric reaction, extraction and dilution conditions for lotus seed samples, and methodological validation are provided in the Supporting Information.

3. Conclusion

In this study, a two-step induced gold enhancement Au@Pt/Au-ELISA was developed and optimized. By precisely controlling the gold deposition process, this strategy effectively addressed the catalytic activity loss caused by blocking treatments in conventional nanozyme-based ELISAs and overcame the specificity decline associated with high-concentration one-step induced gold enhancement. The approach not only enhanced catalytic performance and detection accuracy but also markedly reduced probe consumption, decreasing antibody probe usage to 14.3% and 33.3% of that in the original scheme for ZEN and AFB₁ detection, respectively, thereby improving cost-efficiency and practical applicability. In mycotoxin detection, the method exhibited excellent sensitivity and stability, with detection sensitivities improved by 7.2-fold for ZEN and 2.0-fold for AFB₁ compared to conventional ELISA, and showed strong agreement with LC-MS/MS, confirming its accuracy and reliability. In addition, the assay demonstrated good specificity, matrix tolerance, and reproducibility, making it suitable for rapid screening of complex samples. Notably, the enhancement temperature had a significant impact on assay performance, indicating the need for target-specific thermal optimization during gold deposition. With its outstanding detection performance, improved probe efficiency, and operational simplicity, the two-step induced gold enhancement Au@Pt/Au-ELISA offers a sensitive, stable, and cost-effective tool for food and herbal medicine safety monitoring and holds great potential for supporting mycotoxin risk assessment and regulatory applications.

Experimental section

A detailed description of experimental section is provided in the Supporting Information.

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

CRediT authorship contribution statement

Rentang Huang: Conceptualization, Methodology, Investigation, Formal analysis, Data curation, Visualization, Writing-original draft.

Yuxin Huang: Investigation, Data curation.

Jinrong Yao: Investigation, Formal analysis.

Yi Lin: Methodology, Validation.

Xiangsheng Zhao: Resources, Supervision, Funding acquisition.

Qian Li: Funding acquisition, Writing-review & editing.

Shumei Wang: Supervision, Project administration.

Lei Zhang: Conceptualization, Supervision, Writing-review & editing, Funding acquisition.

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.

Handling Editor: Professor Aiqian Ye

Appendix A. Supplementary data

The following is the Supplementary data to this article: