Ultrasensitive Lateral Flow Immunoassay for Aflatoxin B1 Detection via Magnetic Enrichment-Catalytic Signal Amplification

Abstract

Aflatoxin B1 (AFB₁) is one of the most toxic fungal secondary metabolites. High-sensitivity and rapid detection of AFB₁ is crucial for safeguarding consumer health, reducing post-harvest food losses, and promoting agricultural trade. Here, we developed a magnetic enrichment–catalytic lateral flow immunochromatographic assay (E-C-LFIA) for quantitative AFB₁ detection. The approach couples immunomagnetic capture and enrichment with carboxylated magnetite (Fe₃O₄) nanozyme probes and post-assay peroxidase-like catalysis of the H₂O₂–TMB system to enhance colorimetric readout. Compared with conventional LFIA performed without magnetic enrichment or catalytic amplification, E-C-LFIA achieved a visual detection limit of 0.05 μg/L for AFB₁, corresponding to a 20-fold improvement in sensitivity. The quantitative limit of detection (LOD, 3σ) was 0.023 μg/L, representing a 14.8-fold improvement in sensitivity. The method was demonstrated for AFB₁ screening in representative cereal- and nut-based matrices (rice, corn and peanut). Overall, E-C-LFIA provides a sensitive, rapid, and equipment-light option for on-site AFB1 screening and offers a transferrable strategy for other small-molecule contaminants.

1. Introduction

Aflatoxin contamination is a critical factor affecting food safety, particularly the safety of cereal-based foods, and has become a central issue in global food safety management. Aflatoxin B1 (AFB₁) is recognized as the most toxic member of the aflatoxin family, with significant mutagenic, teratogenic, and carcinogenic risks [1,2,3]. AFB₁ contamination occurs throughout the production, processing, and storage chain of agricultural commodities, and because of its heat and acid resistance, conventional processing methods are often ineffective in eliminating it, posing a significant challenge to food safety. To control AFB₁ residue in food, many countries have established strict regulations [4,5,6]. For example, the European Union has set the maximum allowable concentration of AFB1 in grains at 2 μg/kg, while China has established a maximum residue limit of 5 μg/kg for AFB₁ in milk. However, since AFB₁ typically occurs at very low concentrations, the development of sensitive and reliable detection methods is of paramount importance for ensuring the safety and integrity of the food supply chain [7,8].

Currently, the primary methods for detecting aflatoxins include instrumental techniques, such as high-performance liquid chromatography (HPLC) and liquid chromatography-mass spectrometry (HPLC-MS) [9,10], as well as sensor technologies based on bio-recognition elements, including antibodies, aptamers, and molecularly imprinted polymers. While instrumental methods offer high accuracy, they typically rely on sophisticated equipment and complex sample preparation, which not only increases detection costs but also limits their applicability for rapid on-site testing [11,12,13]. In contrast, although novel biosensors show significant potential, they still face technical challenges, such as complex nucleic acid sequence design and template molecule leakage, with issues related to stability and reproducibility [14,15]. Therefore, a key challenge in rapid aflatoxin detection is to improve sensitivity and robustness while maintaining operational simplicity and low cost.

Immunoassay-based rapid detection technology, known for its high specificity, has become a key tool in mycotoxin screening [12,16,17]. Lateral Flow Immunoassay (LFIA), a representative of immunoassay technologies, is widely used in food safety testing due to its advantages, such as simple operation, rapid response, and the absence of complex equipment [18,19]. It holds promise as an effective method for the rapid detection of aflatoxins. However, traditional LFIA using gold nanoparticles (AuNPs) as labels faces challenges, including insufficient colorimetric signal intensity, low antibody loading capacity, and limited detection accuracy due to reliance on a single optical signal. To address the interference of complex food matrices in detection and to improve sensitivity, immunomagnetic nanoparticles (IMNPs) have been used to enhance detection sensitivity due to their large specific surface area [20]. Meanwhile, magnetite (Fe₃O₄) magnetic nanoparticles (MNPs) exhibit intrinsic peroxidase-like activity, enabling catalytic signal amplification via chromogenic substrate oxidation. Therefore, Fe₃O₄ MNPs can serve as dual-function components that integrate magnetic enrichment/separation with nanozyme-catalyzed colorimetric amplification in LFIA, thereby facilitating an “enrichment–catalysis–detection” workflow.

Based on this rationale, we propose a magnetic E-C-LFIA for highly sensitive detection of AFB1. This method combines immunomagnetic separation and enrichment with nanozyme-based catalytic amplification, in which carboxylated Fe₃O₄ MNPs with peroxidase-like activity are constructed and used as the core signal component to achieve efficient enrichment and amplified colorimetric readout for AFB₁.

2. Materials and Methods

2.1. Materials

The sample pads, conjugate pads, absorbent pads, and PVC backing plates were provided by Shanghai Jinbiao Biological Co., Ltd. (Shanghai, China); Nitrocellulose membranes (CN 95) were supplied by Sartorius Stedim Biotech GmbH (Göttingen, Germany); AFB₁ monoclonal antibodies and their working solutions were prepared in the laboratory; AFB₁, Aflatoxin B2 (AFB₂), Aflatoxin G1 (AFG₁), Aflatoxin G2 (AFG₂), Aflatoxin M1 (AFM₁), Aflatoxin M2 (AFM₂), Ochratoxin A (OTA), Zearalenone (ZEN), Deoxynivalenol (DON), ovalbumin (OVA), bovine serum albumin (BSA), and N,N′-Dicyclohexylcarbodiimide (DCC) were sourced from Sigma-Aldrich (St. Louis, MO, USA); Carboxymethoxylamine hemihydrochloride (CMO), N-Hydroxysuccinimide (NHS), FeCl₃·H₂O, and FeCl₂·H₂O were provided by Shanghai Macklin Biochemical Technology Co., Ltd. (Shanghai, China); 3,3′,5,5′-Tetramethylbenzidine (TMB) and dimethyl sulfoxide were prepared in the laboratory; rice, corn, and peanuts were purchased from a supermarket in Tianjin.

For instrumentation, the HM3035 three-dimensional plane dot-spraying gold sputter coater and ZQ2000 microcomputer automatic slicer were provided by Shanghai Jinbiao Biological Co., Ltd. (Shanghai, China); The ELISA reader was supplied by Thermo Fisher Scientific (Waltham, MA, USA); The HHB11 electric thermostatic incubator was provided by Tianjin Sanshui Co., Ltd. (Tianjin, China).

2.2. Synthesis of AFB1-CMO-OVA

(1) Synthesis of AFB₁-CMO [21]: The synthesis of AFB₁-CMO was performed according to the method described by Zhang et al. Specifically, 4 mg of AFB1 was placed in a 10 mL round-bottom flask and dissolved in a mixture of 4 mL methanol, pyridine, and water (4:1:1 volume ratio). Then, 4 mg of carboxymethylhydroxylamine hydrochloride was added, and the flask was heated to 70 °C in a water bath for reflux for 3 h, followed by overnight reaction at room temperature under light protection. After the reaction, the mixture was vacuum-dried at 50 °C to yield the semi-antigen AFB₁-CMO.

(2) Activation of AFB₁-CMO: The dried product of AFB₁-CMO (2 mg) was dissolved in 1 mL N,N-Dimethylformamide, and 2.7 mg of NHS, 4.8 mg of DCC, and 2.5 mg of DMAP were added. The solution was stirred overnight at room temperature to form an activated ester solution of the semi-antigen.

(3) Preparation of AFB₁-CMO-OVA: 8 mg of OVA was accurately weighed and dissolved in 4 mL PBS, and the activated ester solution was added to the protein solution at a rate of 5 μL/min, and the mixture was stirred overnight at 4 °C. Finally, the reaction product was dialyzed against PBS for 72 h, resulting in the coating antigen AFB1-CMO-OVA.

2.3. Preparation and Characterization of Carboxylated MNPs

Carboxylated Fe₃O₄ magnetic MNPs were synthesized by the co-precipitation method [22]. The specific procedure is as follows: 2.2 g of hexahydrate FeCl₃ and 0.8 g of tetrahydrate FeCl₂ were dissolved in 24 mL of deionized water and stirred under a nitrogen atmosphere, then slowly heated to 70 °C. After 30 min, 6 mL of ammonia solution was added, and the mixture was stirred at 70 °C for another 30 min. Next, 1.2 mL of a 0.5 mg/mL citric acid solution was added, and the reaction temperature was gradually increased to 90 °C, followed by reflux for 1 h. Upon completion of the reaction, the crude product was collected by magnetic separation, washed with deionized water, and transferred to a 100 mL centrifuge tube. The product was then washed alternately with deionized water and anhydrous ethanol four times. Finally, the material was dissolved in a small amount of anhydrous ethanol and dried under nitrogen to obtain a powder. The dried powder was weighed and dissolved in a 10 mM PBS solution.

The morphology and particle size of the MNPs were observed using transmission electron microscopy (TEM). The catalytic peroxidase-like activity was confirmed by UV absorption spectroscopy, which revealed a characteristic absorption peak at 652 nm.

2.4. Preparation of MNP@mAb Immunoprobes

An appropriate amount of MNPs was dissolved in PBS to achieve a final concentration of 1 mg/mL. A 200 μL aliquot of the MNPs was subjected to magnetic separation to remove the supernatant, followed by three washing steps, discarding the remaining supernatant. To the washed MNPs, 10 μL of EDC and NHS solutions (both at 10 mg/mL) were added, and the mixture was dissolved in 500 μL of MES buffer (pH 5.0), followed by uniform mixing using a vortex shaker. The activation process was carried out for 30 min at room temperature and 1200 rpm.

The carboxyl groups on the surface of the activated MNPs covalently coupled with the amine groups of the AFB₁ monoclonal antibody (mAb). After the activation, the supernatant was removed by magnetic separation, and the MNPs were washed three times with PBS. Then, 1.5–3 μg of AFB₁-mAb was added to the MNPs in 500 μL of MES buffer for the coupling reaction, which was performed for 2 h at 25 °C and 1200 rpm. After completion of the coupling reaction, the product was washed three times with PBS. To block any unreacted sites, 5 μL of 20% BSA and 2.5 μL of 10% PEG-20000 were added, and the mixture was incubated at 25 °C and 1200 rpm for blocking. After the blocking step, the immunoprobe was resuspended in 100 μL of working solution.

2.5. Assembly of Immunochromatographic Strip

The sample pad, conjugate pad, nitrocellulose membrane, and absorption pad have adhered to the PVC adhesive backing pad with a 2 mm overlap. The coating antigen AFB₁-CMO-OVA (test line, T line), and goat anti-mouse IgG (control line, C line) were sprayed on the nitrocellulose membrane at 0.5 μL/cm and separated at 0.5 cm. After drying at 37 °C for 6 h, the membrane was cut into a single test strip by an automatic cutting machine for subsequent use.

2.6. MNPs Enrichment and Catalytic Dual Signal Amplification Immunochromatography Detection Procedure

Twenty micrograms of MNPs@mAb were mixed with the sample and incubated for 10 min. The mixture was then enriched using a magnet, collecting the sample at the bottom of the EP tube. After discarding the supernatant, the enriched sample was resuspended in 100 μL PBS buffer. The solution was thoroughly mixed and then applied to the sample pad. After 10 min of reaction, 20 μL of H₂O₂-TMB solution was added to both the C and T lines. Three minutes later, the detection results were captured using an iPhone and analyzed using ImageJ v1.8.0 software to determine the grayscale values of the bands.

2.7. Real Sample Processing and Method Validation

Freshly purchased rice, corn, and peanuts were selected as negative samples. Each sample was ground using a grinder, and 2 g of the sample was added to 10 mL of 70% methanol for extraction. The mixture was shaken for 10 min and then centrifuged at 4 °C and 10,000× g for 10 min. The supernatant was collected and diluted at various ratios for testing using the immunochromatographic strips.

2.8. HPLC Validation of Method Effectiveness

To validate the accuracy of the magnetic bead enrichment catalytic immunochromatography method, HPLC testing was conducted. The specific sample processing procedure is as follows:

Four grams of rice, corn, and peanut samples were each spiked with AFB₁ standard solutions at concentrations of 10 μg/L, 5 μg/L, and 1 μg/L. After mixing overnight, 20 mL of extraction solvent (acetonitrile: water: formic acid = 16.8:3:0.2) was added. The mixture was shaken for 2 min, followed by ultrasonic treatment for 20 min. The samples were then centrifuged at 10,000× g for 10 min at 4 °C. The supernatant was dissolved in 5 mL of acetonitrile-saturated n-hexane, mixed to remove fat, centrifuged, and the lower phase was collected. Three milliliters of the lower phase were processed using a solid-phase extraction column (pre-activated with methanol and water). After two washes, the elution was carried out with 3 mL of methanol, and the eluate was collected and dried under nitrogen. Next, 200 μL of n-hexane and 100 μL of trifluoroacetic acid were added, followed by vortex mixing for 30 s. The mixture was derivatized in a 40 °C water bath for 15 min. After drying under nitrogen, the residue was re-dissolved in the mobile phase, filtered through a 0.22 μm organic membrane, and analyzed by liquid chromatography.

The chromatographic conditions were as follows: The chromatographic conditions were as follows: The separation was carried out using a Waters Symmetry C18 column (Milford, MA, USA) (4.6 mm × 150 mm, 3.5 μm). The injection volume was 5 μL, and the column temperature was maintained at 40 °C. The flow rate was set at 0.3 mL/min. The mobile phase consisted of three components: A phase (water), B phase (methanol), and C phase (acetonitrile), with both methanol and acetonitrile being HPLC grade. A linear gradient elution was applied, with the following volume ratio: water:methanol:acetonitrile = 65:27:8. Detection was performed using a fluorescence detector, with an excitation wavelength of 360 nm and an emission wavelength of 440 nm.

2.9. Statistical Analysis

Unless otherwise stated, experiments were performed using three independent test strips per condition (n = 3). Grayscale intensities of the T and C lines were extracted from strip images using ImageJ v1.8.0, and results are presented as mean ± SD. Calibration curves were obtained by linear regression within the specified working range. The quantitative LOD (3σ) was calculated based on the standard deviation (σ) of replicate blank (or low-level) measurements and the corresponding calibration slope.

3. Results

3.1. Principle of E-C-LFIA for AFB₁ Detection

Figure 1 illustrates the principle and workflow of E-C-LFIA for AFB1 detection. Briefly, carboxylated Fe₃O₄ MNPs conjugated with anti-AFB1 monoclonal antibody (MNP–mAb) are first incubated with the sample for 10 min to form immunocomplexes. The probes are then magnetically collected at the bottom of a microcentrifuge tube, the supernatant is discarded, and the enriched probes are resuspended in 100 μL PBS (pH 7.4). The suspension is applied to the sample pad, and the liquid migrates along the strip by capillary action. On the test line (T line), immobilized coating antigen (AFB₁-CMO-OVA) competitively binds the MNP–mAb probes; therefore, a higher AFB1 concentration results in fewer probes captured on the T line and a weaker T-line signal, whereas a lower AFB1 concentration yields a stronger T-line signal. Excess probes are captured by goat anti-mouse IgG on the control line (C line) to confirm proper strip operation. After 10 min of chromatographic reaction, an H₂O₂–TMB substrate solution is added to the C and T lines. Owing to the intrinsic peroxidase-like activity of Fe₃O₄ MNPs, TMB is catalytically oxidized to blue oxTMB, thereby enhancing band contrast and improving the signal-to-background ratio. The strip is imaged after 3 min, and band intensities are quantified by grayscale analysis.

3.2. Preparation and Characterization of MNPs

Carboxylated Fe₃O₄ MNPs were synthesized via the co-precipitation method (Figure 2A). The morphology and size distribution were characterized by TEM and particle-size analysis (Figure 2B). The results revealed that the MNPs were spherical with a uniform size distribution, and the average particle size was 9.56 ± 2.39 nm (Figure 2B). The UV–vis absorbance spectrum showed a gradual decline from 400 nm to 800 nm (Figure 2C), consistent with the findings reported by Li et al. [22]. To assess the dispersion of the MNPs, they were serially diluted, and absorbance at 450 nm was measured. As shown in Figure 2D, the absorbance at 450 nm increased linearly with increasing MNP concentration, indicating good dispersion in water and suitability as a signal probe component.

3.3. Valuation of Peroxidase-like Activity of MNPs

The MNPs exhibited significant peroxidase-like activity toward the H₂O₂–TMB system. When MNPs were added to a TMB solution containing H₂O₂, the solution rapidly changed from colorless to blue, with a distinct absorption peak at ~650 nm (Figure 3A). In contrast, the control without MNPs showed negligible absorbance in this region. As the amount of MNPs increased, the catalytic formation of oxTMB was progressively enhanced, and the absorbance at 650 nm showed a linear increase (Figure 3B and inset; R² = 0.995). To further analyze the steady-state kinetics of the MNPs, the concentration of TMB was adjusted (Figure 3C), and the kinetic parameters were obtained from Lineweaver–Burk linearization (Figure 3D). The calculated results revealed that the Michaelis constant (Km) of the MNPs was 0.30 mM, with a maximum reaction rate (Vmax) of 8.59 × 10⁻⁸ mol·L⁻¹·s⁻¹. Compared to horseradish peroxidase (HRP), which has a Km of 0.434 mM and a Vmax of 1.0 × 10⁻⁷ mol·L⁻¹·s⁻¹, MNPs demonstrated a lower Km (higher apparent substrate affinity) and a comparable Vmax, supporting their use as nanozyme labels for catalytic signal amplification in LFIA.

3.4. Establishment of E-C-LFIA

3.4.1. Optimization of LFIA Conditions and Sensitivity Analysis

MNPs were coupled with mAb using the activated ester method to prepare the immunosensor probe. The coupling efficiency of MNPs, AFB₁-mAb, and MNPs@mAb immunoprobes was evaluated by Zeta potential measurements (Figure S1). The results showed that the Zeta potential of MNPs was −19.88 ± 1.76 mV, primarily due to the presence of surface carboxyl groups, which imparted a negative charge. After dilution with pure water, mAb exhibited a small negative charge, with a Zeta potential of −12.77 ± 0.51 mV. The covalent coupling of the carboxyl groups on the MNPs with the amine groups of mAb resulted in a shift in the Zeta potential of the MNPs@mAb immunoprobes to −28.88 ± 2.04 mV, confirming the successful preparation of the MNPs@mAb immunoprobes.

The concentration of AFB₁-OVA on the T line significantly impacts the sensitivity of the method. In this experiment, the dilution factor of goat anti-mouse secondary antibody (10 mg/mL) was fixed at 40, and the initial concentration of AFB₁-OVA coating solution was 0.98 mg/mL, which was then diluted to 2, 3, 4, 5, and 6 times. A 100 μL sample buffer was mixed with the immunoprobes and applied to the sample pad. The optimal dilution factor was determined by observing the uniformity of the C and T line bands. The test results are shown in Figure S2A. When AFB₁-OVA was diluted to 2 or 3 times, the T line displayed significantly darker coloration than the C line. At dilutions of 5 or 6 times, the T line coloration gradually weakened and eventually became lighter than the C line. When AFB₁-OVA was diluted to 4 times (i.e., 0.25 mg/mL), the coloration of the C and T lines was consistent. During the experiment, it was essential to ensure that the C and T lines exhibited consistent coloration. Significant differences in coloration could lead to invalid results. Therefore, the optimal conditions for subsequent experiments were chosen as a 4-fold dilution of AFB₁-OVA coating solution and a 40-fold dilution of goat anti-mouse secondary antibody.

To enhance the uniformity of the coloration on the C and T lines and improve the visual quality of the test strips, the sample buffer was optimized. Nine different sample buffers were tested, including PBS (pH 5.7, 7.4, 8.5), PB (pH 5.7, 7.4, 8.5), MES (pH 7.4), Hepes (pH 7.2), and BB (pH 7.4), all at a concentration of 0.01 mol/L. This concentration was selected because, in LFIA systems, buffer ionic strength strongly influences antibody–antigen interactions, nanoparticle dispersion, and capillary flow on the nitrocellulose membrane. A moderate buffer strength is commonly adopted to provide sufficient buffering capacity while minimizing salt-induced changes in flow rate and nonspecific background; therefore, to isolate the effects of buffer type and pH, the buffer concentration was fixed at 10 mM throughout the screening. These buffers were mixed with the MNPs@mAb immunoprobes to a final volume of 100 μL and applied to the sample pad. The coloration was observed, and the buffer that produced clear C and T line bands with minimal background interference was selected as the optimal condition. The results, shown in Figure S2B, indicated that PBS buffers at pH 5.7 and pH 8.5 resulted in uneven coloration on the C and T lines. PB buffers at pH 5.7, 7.4, and 8.5 caused the C line to be overly dark, which not only affected the visual effect but also potentially reduced the sensitivity of the test strip. For BB and Hepes buffers at pH 7.4, no T line coloration was observed, likely due to the environmental conditions of these buffers affecting antibody activity. In contrast, PBS at pH 7.4 and MES at pH 6.0 both produced consistent and uniform coloration on the C and T lines, indicating better antibody stability and activity under neutral conditions. Based on this analysis, PBS at pH 7.4 was determined to be the optimal buffer.

The concentration of materials significantly affects the coloration, uniformity, background interference, and detection sensitivity of the C and T lines on the test strips. To optimize the performance of the test strips, different concentrations of MNPs (0.1, 0.5, 1, 1.5, 2 mg/mL) were tested. Using the optimal sample buffer, the MNP@mAb immunoprobe was mixed with the solution and applied to the sample pad. The coloration was observed, and the concentration that provided clear C and T line bands with minimal background interference was selected as the optimal material concentration. The results, shown in Figure S2C, indicated that as the MNP concentration increased, the intensity of the T line color deepened. At concentrations of 0.1 and 0.5 mg/mL, the T line was lighter than the C line, whereas at concentrations of 1.5 and 2 mg/mL, the T line was darker than the C line. After considering all factors, a concentration of 1 mg/mL of Fe₃O₄ was chosen as the optimal material concentration, as it provided uniform coloration of the C and T lines with a clear background.

During the coupling process, the amount of AFB1 antibody added significantly impacted the coloration of the C and T lines and the sensitivity of the method. Different amounts of AFB₁ antibody (1, 2, 3, 4, 5 μg) were added to prepare MNP@mAb immunoprobes at varying antibody concentrations, which were then mixed with the sample buffer and applied to the sample pad. The antibody concentration that resulted in consistent coloration between the C and T lines was selected as the optimal condition. As shown in Figure S3A, when 1 or 2 μg of antibody was added, the T line coloration was faint. When 4 or 5 μg of antibody was added, the T line was darker than the C line, affecting sensitivity. However, when 3 μg of antibody was added, the C and T lines exhibited consistent coloration. Therefore, the optimal amount of AFB1 mAb to add was determined to be 3 μg.

The volume of MNP@mAb immunoprobe applied significantly affects the coloration of the C and T lines on the test strip. In the experiment, 2, 4, 6, 8, and 10 μL of the signal probe were mixed with the sample buffer and applied to the sample pad. The coloration of the test strip was then observed. The results, shown in Figure S3B, indicated that when the probe volume was 2 μL or 4 μL, the coloration of the C and T lines was faint, affecting visual detection. At a probe volume of 6 μL, the coloration of both the C and T lines was uniform and consistent. When the probe volume exceeded 6 μL, the coloration intensity of the C and T lines was similar to that observed at 6 μL. To improve sensitivity while minimizing costs, 6 μL of probe was selected as the optimal amount.

Under optimized conditions, an LFIA detection method was successfully established. AFB1 standard solutions were diluted with PBS buffer to various concentrations (10, 8, 6, 5, 4, 2, 1, 0.5, 0.1 μg/L) and tested using the test strips. The results, shown in Figure 4, demonstrated that as the concentration of AFB₁ increased, the T line gradually weakened in intensity, and at 1 μg/L, the T line almost disappeared. Thus, the visual detection limit was approximately 1 μg/L. The grayscale values were analyzed using ImageJ v1.8.0 software, with the AFB1 concentration plotted on the x-axis and the T line gray intensity plotted on the y-axis to generate the calibration curve (inset: linear fitting from 1 to 6 μg/L). The quantitative detection limit (LOD, 3σ) of the LFIA was 0.34 μg/L, with a linear range of 1 μg/L to 6 μg/L (R² = 0.9861, n = 3).

To evaluate the specificity of the method, five AFB1 structural analogs (AFB₂, AFG₁, AFG₂, AFM₁, AFM₂) and three other mycotoxins (OTA, ZEN, DON) were tested. AFB₁ was tested at 1 μg/L, while the other eight toxins were tested at 10 μg/L. After mixing each toxin with the immunoprobe, the test strip was analyzed for band coloration. The results, shown in Figure S4, indicated that for AFB₁ detection, the T line completely disappeared. When testing structural analogs (Figure S4A), the method showed some recognition ability for AFG₁, AFB₂, AFG₂, and AFM₁, with weaker T line coloration. However, the method exhibited good specificity for AFM₂, likely due to the similarity in epitopes between the analogs. Further validation demonstrated that the method exhibited cross-reactivity rates of 37.1%, 4.3%, 2.1%, 80.5%, and 10.6% toward AFB₂, AFM₁, AFM₂, AFG₁, and AFG₂, respectively. Although such cross-reactivity may introduce interference in the specific detection of AFB₁, it indicates a relatively broad-spectrum recognition capability of the antibody, which is advantageous for simultaneously screening multiple aflatoxins. Notably, AFM₁ and AFM₂ are primarily present in milk and do not interfere with the detection of aflatoxins in grains. Therefore, this method holds practical value in the detection of aflatoxin analogs. For OTA, ZEN, and DON, the T line still displayed coloration, indicating that the LFIA method has good specificity for other typical mycotoxins.

3.4.2. Optimization of C-LFIA Conditions and Sensitivity Analysis

Although the LFIA was successfully established, considering that AFB1 is a toxic mycotoxin, the sensitivity of the existing method has not yet met the practical application requirements. Therefore, based on the peroxidase-like activity of MNPs, a C-LFIA was further developed to enhance sensitivity. To optimize sensitivity, the amount of mAb was reduced to 1.5 μg to minimize the capture efficiency of the T line. At the same time, the peroxidase-like activity of MNPs was utilized to catalyze the TMB reaction, generating a signal change from a small amount of brown MNPs to a large blue band, thus achieving signal amplification. Under optimized conditions, the TMB concentration was set at 41 mM, and H₂O₂ concentrations were tested at 0.1, 0.5, 1, 1.5, and 2 mol/L, mixed with 0.1 mol/L NaAc-HAc buffer (pH 4). A 20 μL volume of the solution was added to the test strip. As shown in Figure S5A, the coloration intensity was positively correlated with H₂O₂ concentration, and the intensity of the C and T lines gradually deepened as the H₂O₂ concentration increased. Therefore, 2 mol/L H₂O₂ and 41 mM TMB were selected as the optimal conditions.

Building on the optimized H₂O₂ concentration, the ratio of H₂O₂ to TMB was further adjusted. Ratios of 2:8, 3:7, 4:6, 5:5, and 6:4 were tested. As shown in Figure S5B, lower ratios of H₂O₂ to TMB resulted in a high blue background, which interfered with visual detection. However, when the H₂O₂: TMB ratio was 6:4, background interference was minimized, and the contrast of the T line was clearer, with the intensity of the T line steadily increasing over time. Based on these results, a H₂O₂: TMB ratio of 6:4 was selected as the optimal ratio for the catalytic colorimetric reaction.

With the optimized substrate conditions, the MNPs-based C-LFIA method was successfully established. As shown in Figure 5, at 0.1 μg/L, the T line was still visible. At 0.5 μg/L, the T line coloration was faint, and no significant blue change was observed. Therefore, the catalytic detection limit was determined to be 0.5 μg/L, representing a two-fold increase in sensitivity, significantly amplifying the signal and improving sensitivity. Grayscale analysis using ImageJ v1.8.0 software was performed, plotting the logarithm of AFB1 standard concentrations on the x-axis and the T line grayscale values on the y-axis to generate the standard curve. The quantitative detection limit (LOD, 3σ) for this method was 0.17 μg/L, with a linear range from 0.1 μg/L to 2 μg/L (R² = 0.9770, n = 3).

3.4.3. Optimization of E-C-LFIA Conditions and Sensitivity Analysis

To optimize the enrichment system, experiments were conducted in PBS buffer at pH 7.4, using the optimal dilution of the extraction solution, with 0.02 μg/L AFB₁ standard added to the immunoprobe. The volume of the enrichment system was set at 200, 400, 600, 800, and 1000 μL. After enrichment, magnetic separation was performed, and the probe was washed and re-dissolved in the quantification working solution. A 5 μL aliquot of the probe was mixed with 95 μL sample buffer and applied to the sample pad, and the coloration was observed. Based on the optimized extraction solution dilution and enrichment time, the enrichment system volume was gradually increased to 200, 400, 600, and 800 μL. The results, shown in Figure S6, indicated that as the enrichment system volume increased, the coloration of the T line gradually weakened. At 800 μL, the difference in coloration between the T and C lines was significant, and at 1000 μL, the T line coloration was similar to that at 800 μL. Therefore, 800 μL was selected as the optimal enrichment system volume.

Under these optimized conditions, the E-C-LFIA method was successfully developed. Solutions containing AFB1 at concentrations of 0.02, 0.05, 0.1, 0.2, 0.4, 0.8, and 1.6 μg/L were mixed with 20 μg of immunoprobe. After shaking for 10 min, the probe was enriched using a magnet, the supernatant was removed, and the probe was re-dissolved in 95 μL of buffer. The re-dissolved probe was applied to the sample pad, and after 10 min, 4 μL of H₂O₂-TMB buffer was added to the Nitrocellulose membrane, and the results were observed. As shown in Figure 6, as the concentration of the standard decreased, the T line coloration gradually deepened. At a concentration of 0.05 μg/L, the T line was noticeably lighter than the C line. Therefore, after enrichment, the visual detection limit of the E-C-LFIA was 0.05 μg/L, representing a 20-fold increase in sensitivity compared to the non-catalyzed and non-enriched method.

Grayscale analysis was performed using ImageJ v1.8.0 software, with the logarithm of AFB1 standard concentrations plotted on the x-axis and T line grayscale values on the y-axis to generate the standard curve. The quantitative detection limit (LOD, 3σ) of the method was 0.023 μg/L, with a linear range of 0.02 μg/L to 0.4 μg/L (R² = 0.9770, n = 3), showing a 14.8-fold increase in sensitivity compared to the non-catalyzed and non-enriched method.

A comparative evaluation of reported AFB₁ detection methods was performed in terms of detection sensitivity (visual LOD and quantitative LOD, μg/L), nanomaterial type, signal output mode, and applicable sample matrix (

Table 1. Performance comparison of the E-C-LFIA with reported nanomaterials-based detection methods.

Nanomaterial	Signal Type	Visual LOD (μg/L)	Quantitative LOD (μg/L)	Sample	Ref.
Time-resolved fluorescence microspheres	Time-resolved fluorescence	0.3	0.04	corn	[23]
Eu-MOFs	Fluorescence	1	0.149	corn	[24]
Magnetic beads	Colorimetric	--	0.063	corn	[25]
AuNP-TDN13bp-mAb	Colorimetric	0.2	0.13	Rice bran oil	[26]
Fe3O4 nanozyme	Colorimetric	0.05	0.023	Rise, corn peanut	This work

). The E-C-LFIA exhibited the highest visual and quantitative detection sensitivity among all reported methods. Furthermore, most of the reported methods were only applicable to single or a limited number of sample types, whereas the method developed in this work covers three sample types (rice, corn, and peanut), thus demonstrating a broader application scope and higher practical value.

3.5. Sample Analysis

To assess the practical applicability of the E-C-LFIA method, spike recovery experiments were performed by adding 5, 1, and 0.5 μg/L of AFB₁ standard to three negative samples (rice, corn, and peanuts). The recovery rates are shown in

Table 2. Determination of AFB₁ in actual samples.

Samples	Added Conc. (μg/kg)	E-C-LFIA		HPLC
		Measured Content (μg/kg)	Recovery Rate (%)	Measured Content (μg/kg)	Recovery Rate (%)
Rice	5	4.33 ± 0.26	86.55%	4.91 ± 0.41	98.27%
	1	0.70 ± 0.04	70.17%	0.75 ± 0.02	75.17%
	0.5	0.41 ± 0.02	82.18%	0.43 ± 0.02	86.18%
Corn	5	4.14 ± 0.29	82.72%	4.34 ± 0.20	87.72%
	1	0.84 ± 0.04	84.47%	0.88 ± 0.02	88.47%
	0.5	0.67 ± 0.03	134.43%	0.52 ± 0.04	104.29%
Peanut	5	4.10 ± 0.33	82.00%	3.90 ± 0.48	78.00%
	1	0.88 ± 0.33	87.69%	0.78 ± 0.03	78.69%
	0.5	0.47 ± 0.03	93.97%	0.41 ± 0.03	81.44%

Note: Recovery (%) = mean ± SD (RSD%, n = 3).

, ranging from 70% to 134.43%, consistent with the results obtained by HPLC. These results indicate that the MNPs-enriched catalytic immunoassay method demonstrates high accuracy and can rapidly and easily perform quantitative detection of AFB1 in cereal samples. During the enrichment process of this method, complex matrix components in the sample may adsorb nonspecifically onto the surface of magnetic beads or antibodies, affecting the efficiency of antigen-antibody binding and leading to fluctuations in recovery rates. Particularly at low concentrations, the target signal is susceptible to interference from the matrix background, resulting in deviations of the measured values from the true concentration. Additionally, when the target concentration is close to the detection limit of the method, even minor fluctuations in efficiency can significantly impact the recovery results. It is worth mentioning that long-term stability evaluation not only provides critical evidence for determining the storage conditions and shelf life of the method but also helps reveal its potential weaknesses by monitoring performance changes, thereby guiding the optimization of formulations and protective strategies and comprehensively enhancing the method’s robustness and practical utility. To transition this method from “successful in development” to “practically reliable,” systematic long-term stability studies must be conducted in the future. This will establish a reliable operational window for the method in real-world application scenarios and lay the foundation for its standardization and large-scale application.

4. Conclusions

This study successfully developed a novel immunochromatographic detection method based on magnetic E-C-LFIA, achieving highly sensitive and rapid quantitative detection of AFB₁. The method significantly enhanced detection performance through the synergistic effect of magnetic nanoparticle enrichment and catalytic signal amplification, with the visual detection limit (0.05 μg/L) and quantitative detection limit (0.023 μg/L) being 20-fold and 14.8-fold improvements, respectively, compared to traditional methods. This technology provides a reliable tool for on-site detection of AFB₁ in agricultural products such as rice, corn, and peanuts. Additionally, the innovative “enrichment-catalytic dual amplification” design opens new avenues for the development of rapid detection methods for small molecule contaminants. The E-C-LFIA technology holds significant application value in food safety monitoring, offering strong technical support for safeguarding consumer health and mitigating the risks associated with mycotoxins.