Detection method of Aristolochic acid I based on magnetic carrier Fe₃O₄ and gold nanoclusters

2.1 Preparation of AA-I magnetic MIPs

To achieve precise detection of AA-I, it was necessary to first separate and enrich the AA-I molecules in the sample to be tested, and then proceed with the detection. Supercritical fluid extraction (SFE) and solid-phase extraction (SPE) were currently common separation and enrichment methods. SPE included MIPs, macroporous adsorption resin, MOF, and carbon nanotubes. The principle was to enrich and separate the samples by adding adsorbents, achieving the goal of enriching the required substances [12,13,14,15]. The working principle of SFE was to control the polarity of supercritical fluid by controlling the temperature and pressure of the fluid, thereby selectively separating and extracting the required substances from the sample. SFE was often used in research on the removal of AA substances [16]. Figure 1 shows the advantages and disadvantages of the above method.

Figure 1

Characteristics of separation and enrichment methods for AA-I.

In Figure 1, MIPs were selected as adsorbents, which were the best choice for separating and enriching AA-I. Therefore, this study combined magnetic material Fe₃O₄ nanoparticles with molecular surface imprinting technology to synthesize AA-I magnetic MIPs based on Fe₃O₄. The working principle was to polymerize functional monomers, crosslinking agents, and template molecules to form specific polymers, MIPs. Then, by eluting the template molecules in the polymer, a cavity with the same volume size and contour structure as the desired enriched substance could be obtained, allowing MIPs to have a specific recognition and adsorption effect on the substance to be enriched in the experiment [17,18,19]. The specific preparation process is shown in Figure 2.

Figure 2

Schematic diagram of MIPs preparation process.

The experiment aimed to create magnetic MIP microspheres that combine the advantages of good physical and chemical properties, strong recognition ability, large adsorption capacity, and easy detachment to improve the extraction efficiency of AA-I molecules and facilitate subsequent detection, saving detection time. In the selection of experimental drugs, the purity of AA-I reagent was 98%, while the purity of other reagents was analytical grade. The manufacturers were from Shanghai and Chengdu, respectively. Figure 3 shows the detailed information.

Figure 3

Main reagent information for the experiment.

By analyzing existing literature, the process of preparing AA-I magnetic MIPs was divided into three steps [20,21]. In the preparation of AA-I magnetic MIPs, it was necessary to synthesize Fe₃O₄@SiO₂ particles. The materials to be prepared included electronic balance, three-neck flask, isopropyl alcohol, methanol, ultra-pure water, ammonia water, and ethyl orthosilicate. 40 mL of isopropyl alcohol, 500 mg of dried Fe₃O₄ particles, and 8 mL of ultra-pure water were successively added to the three-necked flask. Ultrasonic mixing was used for 25 min to ensure that the components were well mixed. Then, 15 mL of 25% ammonia and 6 mL of ethyl orthosilicate were added, and the solution was fully stirred by a mechanical agitator and reacted at room temperature for 16 h. After the reaction was complete, the black precipitate produced by the reaction was collected using an external magnet and washed several times with ultra-pure water and methanol to remove impurities. Finally, the black precipitate after cleaning was dried in a vacuum environment of 60°C, and finally Fe₃O₄@SiO₂ particles were obtained.

For the preparation of Fe₃O₄@SiO₂@MPS particles, 50 μL of 3-MPS and 500 mg of Fe₃O₄@SiO₂ particles were added to 80 mL of ultra-pure water containing 15% acetic acid using a pipette. The solution was heated to 70°C in a water bath while stirring for 4 h to ensure adequate reaction. After the reaction was over, the reaction products were collected using an external magnet and washed with ultra-pure water until the pH became neutral. After washing, the product was dried in a vacuum at 55°C to obtain Fe₃O₄@SiO₂@MPS particles.

To prepare Fe₃O₄@SiO₂@MPS@MIP particles, as shown in Figure 4, a functional monomer prepolymer was first produced by adding 0.040 mmoL AA-I reagent to 50 mL of nitrile-methanol solution. It was necessary to ensure sufficient dissolution after cooling. Then, 0.160 mmoL of methacrylic acid was added and allowed to react at room temperature for 6 h. 0.52 mmol of ethylene glycol dimethacrylate, 50 mg Fe₃O₄@SiO₂@MPS particles, and 0.06 mmol azodiisobutyronitrile were added. To remove oxygen from the solution, it was treated with ultrasound for 20 min and then aerated with nitrogen for 12 min. In a nitrogen environment, the solution was heated in a water bath to 75°C and stirred for 30 h to promote the completion of the reaction. After the reaction, the products were collected again with external magnets and washed by 8:2 methanol-acetic acid solution for many times until no residue of Fe₃O₄@SiO₂@MPS@MIPs particles could be detected by HPLC. The preparation of Fe₃O₄@SiO₂@MPS@non-imprinted polymers (NIPs) particles was similar, but the AA-I reagent was not required, thus simplifying the process.

Figure 4

Preparation process of Fe₃O₄@SiO₂@MPS@MIPs particles.

The adsorption capacity and final morphology of MIPs were related to the amount of reagents used in the synthesis process, the proportion of each reagent, and the type of drug. Therefore, to select the Fe₃O₄@SiO₂@MPS@MIPs particles with the strongest adsorption capacity, it was necessary to compare polymers made with different ratios and types of reagents. The comparison method was to conduct adsorption experiments on all polymers using standard AA-I reagents, and then high-performance liquid chromatography was used to detect the concentration of AA-I in the remaining solution, to reflect the adsorption capacity of the polymers. The lower the concentration of AA-I in the remaining solution, the stronger the adsorption capacity of the polymer. Moreover, the specific recognition ability of polymers determined their specific adsorption ability, which was influenced by the binding force between functional monomers and template molecules. It was crucial to determine the proportion of template molecules, crosslinking agents, and functional monomers used. Based on existing literature, the dosage ratios of the three were set to 1:40:4 [22]. The calculation formula for imprinting efficiency is given in equation (1).

where IF is the imprinting efficiency factor. Q Fe 3 O 4 @SiO 2 @MPS@MIPs and Q Fe 3 O 4 @SiO 2 @MPS@NIPs represent the adsorption capacities of Fe₃O₄@SiO₂@MPS@MIPs particles and Fe₃O₄@SiO₂@MPS@NIPs particles, respectively. The formula for adsorption capacity is given in equation (2).

where C 0 is the concentration of AA-I in the solution before adsorption. Q is the adsorption capacity. C is the concentration of AA-I in the adsorbed solution. V is the volume of AA-I standard solution added. m is the mass of added Fe₃O₄@SiO₂@MPS@MIPs or Fe₃O₄@SiO₂@MPS@NIPs particles. In summary, the preparation of AA-I magnetic MIPs has been completed.

2.2 Construction of AA-I identification method

The enrichment of AA-I has been achieved, and the next step was to design the detection of AA-I. At present, the analysis and detection methods for AA mainly included thin-layer chromatography, UV visible spectrophotometry, high-performance liquid chromatography, and mass spectrometry. When the above methods had different drawbacks, such as UV visible spectrophotometry being only suitable for trace analysis, thin layer chromatography had poor accuracy, sensitivity, and precision. The equipment for using high-performance liquid chromatography was expensive and requires professional personnel to operate. The combination of high-performance liquid chromatography and mass spectrometry was costly and time-consuming. The specific advantages and disadvantages of the above method are shown in Figure 5.

Figure 5

Characteristics of classical detection methods for AA-I.

Figure 5 shows that existing methods have certain limitations. Currently, the excellent electrical, optical, and catalytic properties of fluorescent nanomaterials were gradually emerging in the scientific community [23]. The most widely studied fluorescent nanomaterials currently included fluorescent metal nanoclusters, quantum dots, etc. Since quantum dots were mostly prepared from heavy metals, they were highly toxic and prone to environmental damage. Therefore, although they had many excellent properties such as narrow emission spectra and strong resistance to photobleaching, their medical applications were also limited [24,25]. Metal nanoclusters had the advantages of high compatibility with biomolecules, simple preparation, and good stability. They were widely used in many fields such as catalysis, environmental monitoring, optics, and electronics, especially in the research of silver nanoclusters and AuNCs. However, the chemical properties of silver nanoclusters were more active than those of AuNCs, and they were prone to form larger nanoparticles in water [26]. Therefore, this study used AuNCs with more stable chemical properties, lower toxicity, and better biocompatibility as fluorescent probes. Its working principle was that the substance under test quenched its fluorescence by changing the system in which AuNCs were located, which could be seen as a sensor that converts chemical signals into optical signals. The specific working principle is displayed in Figure 6.

Figure 6

The working principle of fluorescent probes.

The experimental instruments mainly selected included an electronic balance with an accuracy of one hundred thousandth, a circulating water multi-purpose vacuum pump, a collector type constant temperature heating stirrer, a fluorescence spectrometer, a UV visible absorption spectrometer, a transmission electron microscope, an ultrasonic cleaner, an electric hot air drying oven, a centrifuge, a high-performance liquid chromatography, a Fourier transform spectrometer, and a dynamic light scattering instrument. The manufacturers of the main instruments for the above experiment included companies in Shanghai, Beijing, the United States, Japan, and the United Kingdom. The detailed information of their models and manufacturers is exhibited in Figure 7.

Figure 7

Main instruments and equipment for the experiment.

In the selection of templates for synthesizing AuNCs, this study selected PSA as the template for synthesis based on the characteristic that functional groups of biomolecules could have a good affinity with gold [27,28,29]. In terms of preparation, the first step was to prepare a solution of tetrachloroauric acid. The method was to dissolve 1 g of tetrachloroauric acid in 50 mL of pure water. 5 mL of the obtained solution and 50 mg/mL of PSA solution separately were taken and placed in a three necked flask. The mixture was stirred evenly with magnetic force under water bath heating conditions, and the temperature was controlled at 37°C. The pH value of the solution was adjusted to 11.0, and 0.5 mL of sodium hydroxide solution was added with a concentration of 1 mol/L. The reaction in the water bath was continued to heat for 12 h, and the obtained solution at 4°C was finally stored. To better detect AA-I, the experimental conditions were set as follows: the pH value of the solution system was 7.4, the detection temperature was 37°C. The machine could be immediately tested.

3.1 Performance verification of AA-I magnetic MIPs

A series of experiments were conducted to verify the performance of the synthesized AA-I magnetic MIPs. First, the AA-I standard solution lay the foundation for the dynamic and static adsorption experiments of Fe₃O₄@SiO₂@MPS@MIPs and Fe₃O₄@SiO₂@MPS@NIPs particles. An AA-I standard stock solution with a concentration of 300 μg/mL was prepared. Method: 15 mg of AA-I standard was dissolved in 50 mL of methanol solution, and then, this stock solution was used to prepare a standard working solution of 50–400 μg/mL.

Figure 8 showed the characterization results of Fe₃O₄@SiO₂@MPS@MIPs particles and Fe₃O₄@SiO₂@MPS@NIPs particles. In Figure 8(a), the surface of Fe₃O₄@SiO₂@MPS@MIPs particles was uneven and had a 3D hole structure. In Figure 8(b), the individual Fe₃O₄@SiO₂@MPS@NIPs particles were relatively smooth and did not have a 3D hole structure. Overall, there was not much difference in particle size between the two types of particles, both of which were clustered together. In further transmission electron microscopy and atomic force microscopy analysis, the dark Fe₃O₄ magnetic core was surrounded by a light-colored SiO₂ intermediate layer, and the outermost layer was an MIP-imprinted polymer shell layer with a thickness of about 20–30 nm. The surface roughness of Fe₃O₄@SiO₂@MPS@MIPs particles was significantly higher than that of non-imprinted polymer Fe₃O₄@SiO₂@MPS@NIPs particles, with a root mean square roughness value of about 8.5 nm. The root mean square roughness value of NIPs particles was about 3.2 nm. The significantly increased irregular surface morphology and higher roughness directly confirmed the successful formation of rich imprinting cavity structures on the surface of the MIP layer.

Figure 8

Characterization results of two synthesized particles. (a) Scanning electron micrographs of MIPs. (b) Scanning electron micrographs of NIPs.

The results of isothermal static and dynamic adsorption experiments are shown in Figure 9. In Figure 9(a), in the static adsorption experiment, the adsorption capacity of both materials gradually increased as the concentration of AA-I increased from 50 to 400 mg/mL. 400 mg/mL was significantly higher than that of Fe₃O₄@SiO₂@MPS@NIPs (1.8 mg/g), indicating that MIPs had strong molecular recognition ability and efficient adsorption capacity. The adsorption capacity of Fe₃O₄@SiO₂@MPS@MIPs was relatively low, with a maximum of only about 0.5 mg/g, indicating that materials without specific molecular recognition sites had limited adsorption capacity. In Figure 9(b), in the dynamic experiment, as the time increased from 0 to 20 h, the adsorption capacity of both materials tended to equilibrium. At 20 h, the equilibrium of Fe₃O₄@SiO₂@MPS@MIPs was reached with an adsorption capacity of 0.8 mg/g, and the initial adsorption rate was relatively fast (after about 5 h, the adsorption capacity approached 0.7 mg/g). The adsorption capacity of Fe₃O₄@SiO₂@MPS@NIPs increased slowly, and the final equilibrium adsorption capacity was only about 0.3 mg/g. The specific recognition performance of MIPs enabled them to quickly achieve high adsorption capacity under high concentration and dynamic conditions, demonstrating great potential in adsorption separation technology.

Figure 9

Results of isothermal static and dynamic adsorption experiments. (a) Static experiments. (b) Dynamic experiments.

Next the superiority of the synthesis conditions for AA-I magnetic MIPs was verified. The effects of different synthesis parameters were analyzed, as shown in Figure 10. In Figure 10(a), different ratios of “template molecules: crosslinking agent: functional monomer” (1:20:4, 1:20:6, 1:15:3, 1:25:5, 1:10:4, 1:40:4) had a significant impact on the IF value. Among them, the IF values of 1:10:4 and 1:40:4 were relatively high, close to 3 and 2.5, respectively, while the IF value of 1:15:3 was the lowest, close to 1. In Figure 10(b), there were significant differences in the effect of using different functional monomers on imprinting efficiency. The IF value of methacrylic acid (MAA) was the highest, close to 3.5, followed by methacrylamide (MAM) at around 2. The IF values of AA and 2-vinylpyridine (VP) were relatively low, approaching 1.5 and 1, respectively. MAA exhibited the best imprinting effect as a functional monomer. Figure 10(c) showed that the IF values of different solvents exhibited a certain variation pattern. The IF value of acetonitrile methanol mixed solvent was the highest, close to 4; Methanol ethanol mixed solvent was second, with a value of about 2.5; The IF value of chloroform was the lowest, only 1. This indicated that mixed solvents could provide a more suitable dissolution environment during the molecular imprinting polymerization, thereby improving imprinting efficiency. Choosing MAA as the functional monomer with a reasonable ratio and using acetonitrile methanol mixed solvent could significantly improve the imprinting efficiency.

Figure 10

Analysis of the effects of different synthesis parameters. (a) Imprinting efficiency of different ratios of template molecules, crosslinkers, and functional monomers. (b) Imprinting efficiency of different functional monomers. (c) Stamping efficiency of different solvents.

3.2 Performance verification of AuNCs

The characterization results of PSA-AuNCs are shown in Figure 11. From Figure 11(a), the peak positions of the excitation spectrum and emission spectrum of PSA-AuNCs were approximately 300 and 700 nm, respectively, indicating that the nanoclusters had typical photoluminescence characteristics. The larger stokes shift (∼400 nm) indicated a significant energy loss between photon absorption and emission, which may be related to the surface states of metal clusters and the ligand interactions of proteins. From Figure 11(b), the spectral lines of PSA-AuNCs showed significant changes compared to PSA, indicating that there was an interaction between protein molecules and AuNCs, which may involve deformation of amide bonds or changes in bond energy in the protein. After introducing AA-I, the spectrum of PSA-AuNCs of AA further changed, especially in the absorption peaks at 3,000–3,500 cm⁻¹ (corresponding to the stretching vibration of –OH or −NH groups) and 1,500–1,700 cm⁻¹ (amide I and amide II regions), indicating that AA binds to the nanoclusters through hydrogen bonding or other weak interactions. In Figure 11(c), the nanoparticles were uniformly dispersed with a diameter of approximately 7 nm, exhibiting good monodispersity. This small-sized characteristic provided advantages for its applications in optics, catalysis, and biomedical fields.

Figure 11

Characterization results of PSA-AuNCs. (a) Excitation wavelength and emission wavelength. (b) Comparison of infrared spectra. (c) Particle size image.

The experimental results of the detection conditions were analyzed, as shown in Figure 12. From Figure 12(a), there was not much difference in fluorescence intensity within the pH range of 5.5–9.0, indicating that the solution of PSA-AuNCs can remain stable in this environment. From Figure 12(b), the fluorescence intensity decreased within this pH range when AA-I solution was added. When the pH was 7.4, the fluorescence of PSA-AuNCs was most quenched by AA-I. This confirmed that the detection of AA-I was most sensitive when the pH was 7.4, and this pH matched the pH of the human body environment, making it a potential fluorescent probe for detecting AA-I in biological samples. In Figure 12(c), as the temperature increased from 30 to 80°C, the fluorescence quenching intensity first briefly increased and then gradually decreased, from about 0.4 to nearly 0.1. This indicated that the increase in temperature led to a decrease in fluorescence quenching efficiency, and the most suitable temperature was about 38°C. From the analysis of Figure 12(d), there was not much difference in the degree of quenching with the increase in time, indicating that the detection of AA-I using PSA-AuNCs could be carried out immediately, saving time.

Figure 12

Experimental results of testing conditions. (a) Fluorescence intensities of porcine serum albumin-gold nanoclusters at different pH. (b) Fluorescence intensities of porcine serum albumin-gold nanoclusters at different pH values after the addition of AA-I. (c) Effect of different temperatures on the degree of quenching. (d) Effect of different times on the degree of quenching.

Figure 13 shows the response performance of PSA-AuNCs to AA-I. In Figure 13(a), when a concentration of 0.1 μg/mL of AA-I solution was added, PSA-AuNCs could undergo fluorescence quenching reaction. This indicated that the detection sensitivity of PSA-AuNCs fluorescent probes was good. As the concentration of AA-I solution increased, the fluorescence intensity gradually weakened, indicating that the quenching intensity is also gradually increasing. In Figure 13(b), the concentration of AA-I solution was positively correlated with the fluorescence quenching degree of PSA-AuNCs, with a correlation coefficient of 0.9923. This indicated that the PSA-AuNCs fluorescence probe detection method was suitable for analyzing actual samples.

Figure 13

Detection of the responsiveness of PSA-AuNCs to AA-I. (a) Emission wavelength spectra at different concentrations. (b) The relationship between the concentration of AA-I and the degree of fluorescence quenching of pig serum albumin gold nanoclusters.

The anti-interference experiment is shown in Figure 14. As shown in Figure 14, regardless of the presence of any interfering substances in the sample, PSA-AuNCs only exhibited a significant quenching reaction toward AA-I, with a quenching degree of over 0.40. The results indicated that the system had good anti-interference ability, especially when targeting most metal ions and organic molecules, the signal was not significantly affected.

Figure 14

Anti interference experiment. (a) Quenching strength. (b) Anti-interference degree.

Figure 15 shows the detection results of AA-I in rat urine. In Figure 15(a), the sample recovery rates of rats using the PSA-AuNCs detection method were 116.43, 105.1, and 93.2%, respectively. The recovery rates using high-performance liquid chromatography were higher, at 128.6, 114.5, and 105.7%, respectively. The recovery rate results of the PSA-AuNCs detection method were better. In Figure 15(b), for the three concentrations of AA-I, the detection time using PSA-AuNCs was significantly shorter than that using high-performance liquid chromatography. As the concentration increased, the increase in time required for the research method was not significant, indicating that the method had good detection performance and saved detection time, making it suitable for detecting AA-I components in actual biological samples.

Figure 15

Detection results of AA-I in rat urine. (a) Sample recovery rates. (b) Detection time.