High-Sensitive FAM Labeled Aptasensor Based on Fe3O4/Au/g-C3N4 for the Detection of Sulfamethazine in Food Matrix

In this study, we developed a fluorescent aptasensor based on Fe3O4/Au/g-C3N4 and a FAM-labeled aptamer (FAM-SMZ1S) against sulfamethazine (SMZ) for the specific and sensitive detection of SMZ in food matrix. The FAM-SMZ1S was adsorbed by the Fe3O4/Au/g-C3N4 via π–π stacking and electrostatic adsorption, serving as a basis for the ultrasensitive detection of SMZ. Molecular dynamics was used to explain the reasons why SMZ1S and SMZ were combined. This aptasensor presented sensitive recognition performance, with a limit of detection of 0.16 ng/mL and a linear range of 1–100 ng/mL. The recovery rate ranged from 91.6% to 106.8%, and the coefficient of variation (CV) ranged from 2.8% to 13.4%. In addition, we tested the aptasensor for the monitoring of SMZ in various matrix samples, and the results were well-correlated (R2 ≥ 0.9153) with those obtained for HPLC detection. According to these results, the aptasensor was sensitive and accurate, representing a potentially useful tool for the detection of SMZ in food matrix.


Introduction
Sulfamethazine (SMZ) is a member of the sulfonamide family of antibiotics, and it is widely used to treat bacterial infections caused by pathogenic bacteria and parasites in livestock [1]. However, the overuse of SMZ in animal breeding leads to SMZ residues, which then accumulate in the human body through the food chain and cause allergic and toxic reactions [2,3]. The European Union, the United States, and China have established maximum residue limits of 100 µg/kg SMZ [4,5]. In previous studies, researchers have used numerous methods for the detection and quantification of SMZ, including liquid chromatography-mass spectrometry (LC-MS), high-performance liquid chromatography (HPLC), and the enzyme-linked immunosorbent assay (ELISA) [6][7][8]. Although LC-MS and HPLC are sensitive and highly specific for SMZ, they require complicated operations and are not suitable for detection in large numbers of samples [9]. Because ELISA is dependent on the surrounding environment, it may be limited in practice [10]. Therefore, there is an urgent need to construct a rapid and sensitive method for detecting SMZ in food matrix.
Aptamers are 70-100 nt RNA or ssDNA that are generally prepared through systematic evolution of ligands by exponential enrichment (SELEX) and often used as bioprobes, which recognize and bind targets with high affinity and specificity [11]. Aptamers can fold into unique spatial structures and can specifically bind to target molecules, such as cells, proteins, antibiotics, toxins, and other biomarkers, through hydrogen bonds, hydrophobic interactions, and van der Waals forces [12]. Kou screened the shortest aptamer (SMZ1S), which can be used as a capture probe to identify and combine SMZ to detect SMZ in the sample matrix. However, SMZ1S was only used in the matrix of eggs and milk [4], so the performance of SMZ1S also needs to be verified in more sample substrates. Fluorescent aptasensors are widely used in food detection due to their sensitivity, specificity and low cost [13,14]. In our previous research, we constructed an aptasensor -based on Fe 3 O 4 /Au/g-C 3 N 4 for the sensitive and accurate detection of sulfameter. In this study, because of its excellent fluorescence-quenching ability, we constructed a fluorescent aptasensor based on Fe 3 O 4 /Au/g-C 3 N 4 to detect SMZ. Previously constructed aptasensors for detecting SMZ are based on graphene oxide (GO) and gold nanoparticles (AuNPs). However, the structure of GO is rich in oxygen-containing groups, which can lead to false-positive detection results for GO-based aptasensors [15,16]. AuNPs are susceptible to salt-induced aggregation, and the adsorption rate of aptamers increases with higher salt concentrations, which makes it difficult to apply AuNPs-based aptasensors to complex biological mechanistic samples [17]. The aptasensor developed in our study bypasses these issues and, moreover, has a lower limit of detection (LOD) and wider ranges in terms of SMZ detection [4,18,19]. However, reports are lacking on the quantitative measurement of SMZ based on the use of Fe 3 O 4 /Au/g-C 3 N 4 .
In this work, we constructed, for the first time, a Fe 3 O 4 /Au/g-C 3 N 4 fluorescent aptasensor, and the LOD was found to be the lowest value for all aptasensors in SMZ detection. The reason for the high specificity of SMZ1S to SMZ was also investigated by molecular dynamics simulations. We also tested the established aptasensor in the analysis of food matrices, including milk, egg, honey, swine, swine kidney, chicken, chicken liver, beef, crucian, and shrimp, which we spiked with different concentrations of SMZ. Finally, we compared the results of HPLC and this method, and explored their correlation, proving that the aptasensor could detect trace SMZ in multiple complex sample matrix.

Synthesis of Fe 3 O 4 /Au/g-C 3 N 4
As shown in Figure 1, Fe 3 O 4 /Au/g-C 3 N 4 was prepared using a hydrothermal synthesis reaction. Briefly, 15 g urea was gradually heated to 550 • C, with a ramp of 2 • C/min, and the temperature was maintained for 4 h to obtain the yellow solid of g-C 3 N 4 . The g-C 3 N 4 (1 g) was added to 50 mL of 1.5 mM trisodium citrate solution, and the mixture was sonicated for 20 min. Subsequently, 25 mL of 1 mM chloroauric acid solution was added to the reaction mixture, which was heated at 60 • C for 2 h. Then, the reacted mixture was suction-filtered and washed 4 times with DI H 2 O and alcohol with subsequent freeze-drying to obtain Au/g-C 3 N 4 . Later, 0.15 g Au/g-C 3 N 4 , 0.5 mmol FeCl 3 ·6H 2 O, and 0.73 mmol of FeSO 4 ·7H 2 O were added to 50 mL DI H 2 O, followed by sonication for 1 h. The mixture was then transferred to a hydrothermal reactor, 3 mmol NaOH was added, and the mixture was then heated at 120 • C for 24 h. After cooling, the mixture was washed 4 times with DI H 2 O followed by ethanol. Finally, we obtained Fe 3 O 4 /Au/g-C 3 N 4 using magnetic separation and freeze-drying [12,20].

Preparation of the Aaptasensor
To prepare the aptasensor, 1 mL of 1 mg/mL of Fe3O4/Au/g-C3N4 was first sonicate for 30 min, followed by adding 10 μL of 1 mg/mL PEG 20,000 and subsequent incubatio for 12 h to block nonspecific binding sites, and the resulting mixture was then stored at °C before use. Subsequently, FAM-SMZ1S was diluted to 100 nM with binding buffer (10 mM NaCl, 2 mM MgCl2, 20 mM Tris-HCl, 1 mM CaCl2, 5 mM KCl, and 0.02% Tween 20 pH 7.6). Then, 199 μL of 100 nM FAM-SMZ1S and 1 μL of different concentrations of SMZ (0, 1, 5, 25, 50, 100 ng/mL) were incubated for 30 min (i.e., to a final volume of 200 μL) a 25 °C in the dark. Along with 60 μL of 1 mg/mL of Fe3O4/Au/g-C3N4, the samples wer incubated for 5 min. The supernatant was then collected by magnetic separation. W measured the fluorescence-intensity value of the supernatant using Varioskan LUX (Thermo Scientific, Waltham, MA, USA) (λex = 492 nm, λem = 518 nm). The standard curve were constructed based on the fluorescence intensity of SMZ standards at different con centrations of 0, 1, 5, 25, 50, and 100 ng/mL. The LOD was calculated as follows: LOD = SD/slope.

Molecular Dynamics Trajectory Analysis
Equilibrium was achieved under a constant particle number, volume and tempera ture, as well as optimal isothermal and isobaric conditions [21,22]. The SMZ1S-SMZ com plex was then subjected to 60 ns implicit solvent model molecular dynamics (MD) simu lations, and the resulting structurally stable complexes were subjected to explicit solven model molecular dynamics simulations. MD simulations of SMZ1S and other antibiotics such as sulfamethazine sulfanilamide, sulfameter, sulfadiazine, sulfamethoxypyridazine chloramphenicol, kanamycin, and chlortetracycline, were performed in the same manner The root mean square deviation (RMSD) was used as a criterion to evaluate the stabilit of the aptamer target. All simulation processes were implemented using GROMACS (Uni versity of Goettingen, Germany) and G_MMPBSA software [23,24]. Step 1: Preparation of g-C 3 N 4 by heating with urea.

Preparation of the Aaptasensor
To prepare the aptasensor, 1 mL of 1 mg/mL of Fe 3 O 4 /Au/g-C 3 N 4 was first sonicated for 30 min, followed by adding 10 µL of 1 mg/mL PEG 20,000 and subsequent incubation for 12 h to block nonspecific binding sites, and the resulting mixture was then stored at 4 • C before use. Subsequently, FAM-SMZ1S was diluted to 100 nM with binding buffer (100 mM NaCl, 2 mM MgCl 2 , 20 mM Tris-HCl, 1 mM CaCl 2 , 5 mM KCl, and 0.02% Tween 20, pH 7.6). Then, 199 µL of 100 nM FAM-SMZ1S and 1 µL of different concentrations of SMZ (0, 1, 5, 25, 50, 100 ng/mL) were incubated for 30 min (i.e., to a final volume of 200 µL) at 25 • C in the dark. Along with 60 µL of 1 mg/mL of Fe 3 O 4 /Au/g-C 3 N 4 , the samples were incubated for 5 min. The supernatant was then collected by magnetic separation. We measured the fluorescence-intensity value of the supernatant using Varioskan LUX (Thermo Scientific, Waltham, MA, USA) (λ ex = 492 nm, λ em = 518 nm). The standard curves were constructed based on the fluorescence intensity of SMZ standards at different concentrations of 0, 1, 5, 25, 50, and 100 ng/mL. The LOD was calculated as follows: LOD = 3 SD/slope.

Molecular Dynamics Trajectory Analysis
Equilibrium was achieved under a constant particle number, volume and temperature, as well as optimal isothermal and isobaric conditions [21,22]. The SMZ1S-SMZ complex was then subjected to 60 ns implicit solvent model molecular dynamics (MD) simulations, and the resulting structurally stable complexes were subjected to explicit solvent model molecular dynamics simulations. MD simulations of SMZ1S and other antibiotics, such as sulfamethazine sulfanilamide, sulfameter, sulfadiazine, sulfamethoxypyridazine, chloramphenicol, kanamycin, and chlortetracycline, were performed in the same manner. The root mean square deviation (RMSD) was used as a criterion to evaluate the stability of the aptamer target. All simulation processes were implemented using GROMACS (University of Goettingen, Germany) and G_MMPBSA software [23,24].

Aptasensor Selectivity
To study its selectivity, we tested the performance of the aptasensor in detecting other sulfonamides, such as sulfamethazine, sulfanilamide, sulfameter, sulfadiazine, sulfamethoxypyridazine, and other commonly used antibiotics (i.e., chloramphenicol, kanamycin, and chlortetracycline). For this, 199 µL of 100 nM FAM-SMZ1S was incubated with 1 µL of the different antibiotics at 100 ng/mL for 30 min. Subsequently, 60 µL of 1 mg/mL of Fe 3 O 4 /Au/g-C 3 N 4 was added to the mixture, which was incubated for 5 min, after which we measured the fluorescence intensity of the supernatant. After measuring the corresponding fluorescence intensity F of different antibiotics and the fluorescence intensity F 0 of blank sample, and the value of ∆F (∆F = F − F 0 ) was calculated.

Sample Preparation
All samples were purchased from the local supermarket and processed using a previously established procedure with slight modifications [12]. To prepare the milk samples, 2 mL of skimmed milk was centrifuged at 14,000 rpm at 4 • C for 20 min and diluted to 20 mL. The milk was then filtered using a 0.22-µm filter membrane. To prepare the egg sample, 2 g egg and 4 mL ethyl acetate were thoroughly mixed and then centrifuged at 12,000 rpm for 10 min. Subsequently, the ethyl acetate in the mixture was evaporated under a nitrogen flow at 40 • C. To prepare the honey sample, 2 mL of honey was diluted 10 times with PBS (136.89 mM NaCl; 2.67 mM KCl; 8.1 mM Na 2 HPO 4 ; 1.76 mM KH 2 PO 4 ; pH: 7.4). The remaining samples (5 g each of crucian, shrimp, swine, swine kidney, chicken, chicken liver and beef) were individually processed using a homogenizer, after which 25 mL of acetonitrile was added to the mixture followed by agitation for an additional 15 min. The resulting mixture was sonicated for 10 min and finally centrifuged at 12,000 rpm for 15 min. The supernatant was collected and transferred to 30 mL acetonitrile-saturated n-hexane, followed by agitation for 10 min to remove the fat. Finally, the organic solvents were evaporated by heating the mixture in a water bath at 80 • C, and the dried residue was dissolved in 50 mL of binding buffer and diluted 10 times for use [25][26][27].

Validation of Aptasensor
We validated the performance of the aptasensor by applying it in the detection of 10 food matrices. HPLC was first used to confirm that all the samples were free of SMZ. The standard curves for the food matrices samples were constructed as described in the preparation of the aptasensor above. The accuracy and precision of the aptasensor were evaluated by analyzing the above samples spiked with SMZ at three different levels (50, 100, 200 µg/kg), with each concentration level tested 5 times. To verify the reliability of the aptasensor in the food matrices, we conducted a comparison of the results for the aptasensor and HPLC using the same 10 food matrices. All spiked samples with different concentrations of SMZ were subjected to HPLC analysis according to a published procedure [28]. Linear regression was used to calculate the correlations between the results obtained from both the aptasensor and the HPLC [29].

Fluorescence-Quenching Effect between SMZ1S and Fe 3 O 4 /Au/g-C 3 N 4
As shown in Figure 2, in the absence of SMZ, the FAM-SMZ1S was adsorbed by Fe 3 O 4 /Au/g-C 3 N 4 , and the fluorescence was quenched by electron-induced transfer [30], electrostatic adsorption, and π-π stacking. However, in the presence of SMZ, FAM-SMZ1S specifically bound SMZ and would thus not be adsorbed by Fe 3 O 4 /Au/g-C 3 N 4 , which prevented quenching of the fluorescence from FAM-SMZ1S. Therefore, the higher the SMZ concentration, the higher the fluorescence value in the supernatant after magnetic separation. Conversely, when the SMZ concentration decreased, the detected fluorescence value decreased. Thus, the basis for our constructed fluorescence aptasensor for SMZ detection was the functional relationship between the SMZ concentration and the fluorescence value.

Optimization of Detection Conditions
We constructed the aptasensor by determining the optimized quality ratio between the FAM-SMZ1S and Fe3O4/Au/g-C3N4. As shown in Figure 3a, the fluorescence value decreased as the mass ratio of FAM-SMZ1S vs. Fe3O4/Au/g-C3N4 decreased. When the mass ratio was reduced to 1:750, the fluorescence value was reduced to a low level, and when the ratio continued to decrease, the fluorescence value did not substantially change. Therefore, we selected the mass ratio of FAM-SMZ1S vs. Fe3O4/Au/g-C3N4 at 1:750 as the optimal mass ratio for constructing the fluorescence aptasensor. As shown in Figure 3b, the LOD of the aptasensor was 0.16 ng/mL, and the aptasensor maintained a good linear relationship between 1 and 100 ng/mL (Y = 0.02241 X + 0.1406, R 2 = 0.9933).

Stability Analysis of Aptamer Target and Selectivity Analysis
After a stable SMZ1S structure was obtained via 60 ns implicit model MD simulations, the RMSD values of the complex SMZ1S system with eight antibiotics, including SMZ, were calculated. As shown in Figure 4, the RMSD value curves of all of the systems

Optimization of Detection Conditions
We constructed the aptasensor by determining the optimized quality ratio between the FAM-SMZ1S and Fe 3 O 4 /Au/g-C 3 N 4 . As shown in Figure 3a, the fluorescence value decreased as the mass ratio of FAM-SMZ1S vs. Fe 3 O 4 /Au/g-C 3 N 4 decreased. When the mass ratio was reduced to 1:750, the fluorescence value was reduced to a low level, and when the ratio continued to decrease, the fluorescence value did not substantially change. Therefore, we selected the mass ratio of FAM-SMZ1S vs. Fe 3 O 4 /Au/g-C 3 N 4 at 1:750 as the optimal mass ratio for constructing the fluorescence aptasensor. As shown in Figure 3b, the LOD of the aptasensor was 0.16 ng/mL, and the aptasensor maintained a good linear relationship between 1 and 100 ng/mL (Y = 0.02241 X + 0.1406, R 2 = 0.9933).
SMZ detection was the functional relationship between the SMZ concentration and the fluorescence value.

Optimization of Detection Conditions
We constructed the aptasensor by determining the optimized quality ratio between the FAM-SMZ1S and Fe3O4/Au/g-C3N4. As shown in Figure 3a, the fluorescence value decreased as the mass ratio of FAM-SMZ1S vs. Fe3O4/Au/g-C3N4 decreased. When the mass ratio was reduced to 1:750, the fluorescence value was reduced to a low level, and when the ratio continued to decrease, the fluorescence value did not substantially change. Therefore, we selected the mass ratio of FAM-SMZ1S vs. Fe3O4/Au/g-C3N4 at 1:750 as the optimal mass ratio for constructing the fluorescence aptasensor. As shown in Figure 3b, the LOD of the aptasensor was 0.16 ng/mL, and the aptasensor maintained a good linear relationship between 1 and 100 ng/mL (Y = 0.02241 X + 0.1406, R 2 = 0.9933).

Stability Analysis of Aptamer Target and Selectivity Analysis
After a stable SMZ1S structure was obtained via 60 ns implicit model MD simulations, the RMSD values of the complex SMZ1S system with eight antibiotics, including SMZ, were calculated. As shown in Figure 4, the RMSD value curves of all of the systems

Stability Analysis of Aptamer Target and Selectivity Analysis
After a stable SMZ1S structure was obtained via 60 ns implicit model MD simulations, the RMSD values of the complex SMZ1S system with eight antibiotics, including SMZ, were calculated. As shown in Figure 4, the RMSD value curves of all of the systems fluctuate significantly, besides that of chloramphenicol, showing that there was significant movement of the ssDNA skeleton in the range of 0-15 ns. Furthermore, in the SMZ1S-SMZ system, the overall RMSD value curve fluctuated around 0.6 nm ( Figure 4a). As shown in Figure 4e,f, the average of the RMSD values of SMZ1S in sulfamethoxypyridazine and chloramphenicol fluctuated between 0.9 nm and 0.8 nm, but the amplitude was much larger than that of SMZ1S-SMZ. As shown in Figure 4b-h, in the MD simulation of other aptamer target systems, the RMSD values fluctuated greatly. This indicated that the complex formed by SMZ1S and other antibiotics undergoes violent movement during the MD simulation in the range of 0-60 ns, and the complex was unstable. Meanwhile, the selectivity of the aptasensor was validated by the selective testing of different non-targeting antibiotics. As shown in Figure 5, the fluorescence intensity (∆F) of other sulfonamide antibiotics similar in structure to SMZ was much lower than that of SMZ, while the fluorescence intensity values of other antibiotics were even lower. These results indicate that the aptasensor had great selectivity for SMZ because SMZ1S stably bound to SMZ.
Biosensors 2022, 12, x FOR PEER REVIEW 6 of 11 fluctuate significantly, besides that of chloramphenicol, showing that there was significant movement of the ssDNA skeleton in the range of 0-15 ns. Furthermore, in the SMZ1S-SMZ system, the overall RMSD value curve fluctuated around 0.6 nm ( Figure 4a). As shown in Figure 4e,f, the average of the RMSD values of SMZ1S in sulfamethoxypyridazine and chloramphenicol fluctuated between 0.9 nm and 0.8 nm, but the amplitude was much larger than that of SMZ1S-SMZ. As shown in Figure 4b-h, in the MD simulation of other aptamer target systems, the RMSD values fluctuated greatly. This indicated that the complex formed by SMZ1S and other antibiotics undergoes violent movement during the MD simulation in the range of 0-60 ns, and the complex was unstable. Meanwhile, the selectivity of the aptasensor was validated by the selective testing of different non-targeting antibiotics. As shown in Figure 5, the fluorescence intensity (ΔF) of other sulfonamide antibiotics similar in structure to SMZ was much lower than that of SMZ, while the fluorescence intensity values of other antibiotics were even lower. These results indicate that the aptasensor had great selectivity for SMZ because SMZ1S stably bound to SMZ.

Validation of Aptasensor
We demonstrated the performances of the aptasensor by detecting the with different concentrations (i.e., 50, 100, and 200 μg/kg) in food matrices. j, we depicted the standard curves in the various matrix samples, includi honey, crucian, shrimps, swine, swine kidney, chicken, chicken liver, and cat in Table 1, the LODs in the different food matrices ranged from 0.29 to 0.6 recovery rate of the aptasensor ranged from 91.6% to 106.8%, and the CVs 2.8% to 13.4%. For HPLC detection, the recovery rate varied from 95.7% to the CVs varied from 1.5% to 11.9%. The results of the aptasensor positively co those of HPLC (R 2 ≥ 0.9153). These results demonstrated that the aptasensor c to accurately and sensitively detect SMZ in food matrix.

Validation of Aptasensor
We demonstrated the performances of the aptasensor by detecting the spiked SMZ with different concentrations (i.e., 50, 100, and 200 µg/kg) in food matrices. In Figure 6a-j, we depicted the standard curves in the various matrix samples, including milk, egg, honey, crucian, shrimps, swine, swine kidney, chicken, chicken liver, and cattle. As shown in Table 1, the LODs in the different food matrices ranged from 0.29 to 0.69 ng/mL, the recovery rate of the aptasensor ranged from 91.6% to 106.8%, and the CVs ranged from 2.8% to 13.4%. For HPLC detection, the recovery rate varied from 95.7% to 107.9%, and the CVs varied from 1.5% to 11.9%. The results of the aptasensor positively correlated with those of HPLC (R 2 ≥ 0.9153). These results demonstrated that the aptasensor can be applied to accurately and sensitively detect SMZ in food matrix.

Conclusions
We developed, for the first time, a sensitive and accurate fluorescent aptasensor based on FAM-SMZ1S and Fe 3 O 4 /Au/g-C 3 N 4 to detect SMZ in 10 food matrixes. Through molecular dynamics simulation, in the SMZ1S and SMZ system, the RMSD value fluctuated near 0.6 nm from 15 ns to 60 ns, indicating that SMZ1S had high selective. According to the experimental results, the aptasensor had a satisfactory linear detection range (1-100 ng/mL), a low LOD (0.16 ng/mL), and a high selectivity in SMZ detection. The accuracy of the established aptasensor for the detection of SMZ was specifically demonstrated by the satisfactory correlation (R 2 ≥ 0.9153) between the aptasensor data obtained in this work and HPLC. Therefore, the aptasensor was a reliable tool and can be used as an analytical means to monitor SMZ in food matrix.