An AuNPs-Based Fluorescent Sensor with Truncated Aptamer for Detection of Sulfaquinoxaline in Water

Herein, we developed a novel truncation technique for aptamer sequences to fabricate highly sensitive aptasensors based on molecular docking and molecular dynamics simulations. The binding mechanism and energy composition of the aptamer/sulfaquinoxaline (SQX) complexes were investigated. We successfully obtained a new SQX-specific aptamer (SBA28-1: CCCTAGGGG) with high affinity (Kd = 27.36 nM) and high specificity determined using graphene oxide. This aptamer has a unique stem-loop structure that can bind to SQX. Then, we fabricated a fluorescence aptasensor based on SBA28-1, gold nanoparticles (AuNPs), and rhodamine B (RhoB) that presented a good linear range of 1.25–160 ng/mL and a limit of detection of 1.04 ng/mL. When used to analyze water samples, the aptasensor presented acceptable recovery rates of 93.1–100.1% and coefficients of variation (CVs) of 2.2–10.2%. In conclusion, the fluorescence aptasensor can accurately and sensitively detect SQX in water samples and has good application prospects.


Introduction
Sulfaquinoxaline (SQX) is a sulfonamide widely used to prevent infections, treat animals, and promote animal growth [1,2]. However, extra SQX can be released into the environment via animal feces and urine, causing significant water contamination [3]. The SQX in water can be directly drunk or enter the food chain through aquatic organisms and crops, eventually causing human allergies, affecting intestinal homeostasis, threatening thyroid and kidney health, and even triggering carcinogenicity [4][5][6]. For human health, various analytical methods are currently available for the detection of SQX, including liquid chromatography-tandem mass spectrometry [7], high-performance liquid chromatography (HLPC) [8], capillary zone electrophoresis [9], and enzyme-linked immunosorbent assay [10]. Although these methods are highly accurate, they have the disadvantages of requiring expensive equipment and complicated sample handling methods or being unstable. Thus, a reliable and low-cost analytical approach for detecting SQX in the environment is required.
Aptamers are single-stranded DNAs or RNAs that can fold into specific structures and stably bind to their targets via hydrogen bonds, van der Waals interactions, electrostatic forces, and so on [11,12]. Aptamers have some advantages over antibodies, such as low cost, easy modification, and high stability [13]. However, in aptamers, not every sequence region is involved in the binding reaction [14]. These unnecessary regions not involved in the binding can interfere with the reaction [15], and their removal can increase the affinity of the aptamer [16]. Attempts to shorten the sequences of aptamers based on their secondary structure have been previously proposed. For example, Shi et al. developed an SQXspecific aptamer after two truncations [4]. Although a new aptamer with a higher affinity was obtained, the method was cumbersome and required multiple cuts [4,17]. Therefore, an efficient and straightforward aptamers truncation approach is required. Retaining the conserved sequence of aptamers is considered an efficient truncation method [15]. Additionally, molecular docking and molecular dynamics (MD) simulations can be used to evaluate truncation results. Molecular docking can evaluate the interactions between ligands and receptors and predict their binding modes and affinities [18,19]. Meanwhile, MD simulations are used to evaluate the stability of the system [20]. Based on MD results, the binding energy can be decomposed and analyzed by the Molecular Mechanics-Poisson Boltzmann Surface Area (MM-PBSA) approach [21]. Herein, we used molecular docking and MD simulations to develop a novel method to truncate and characterize aptamers.
Gold nanoparticles (AuNPs) have been considered promising nanomaterials for biosensing [22], biological imaging [23], and catalysis [24]. Due to their high chemical stability, large specific surface area, and simple synthesis, AuNPs have been widely used in bioassays [25]. Particularly, the use of AuNPs in aptasensors to detect targets by colorimetry has been widely reported, but its sensitivity remains low [26]. Moreover, the absorption spectrum of AuNPs overlaps with the emission spectrum of Rhodamine B (RhoB). Therefore, AuNPs have the clear ability to quench the RhoB fluorescence [26], which can be used to develop highly sensitive fluorescent aptasensors. However, the use of a fluorescent aptamer based on AuNPs and RhoB for detecting SQX has not been previously reported.
In the present study, we obtained an aptamer against SQX by analyzing its conserved sequence and the results of molecular docking and MD simulations. The affinity of our aptamer was much higher than the aptamer developed by Shi et al. [4]. Then, we analyzed the interactions between SQX and the aptamer and demonstrated the high affinity and specificity of the aptamer. Finally, we developed a fluorescent aptasensor based on this aptamer and AuNPs to accurately detect SQX in water samples, which presented great potential for practical applications.

Molecular Docking
To analyze the homology of the aptamers, a phylogenetic tree was constructed using DNAMAN 8 (Lynnon Biosoft, San Ramon, CA, USA). The chemical structure of SQX was drawn by ChemDraw 19.0 (CambridgeSoft, Cambridge, MA, USA) and transformed into a three-dimensional structure by Chem3D 19.0. For the 3D structure of aptamers, their secondary structures were first predicted and converted to a 3D structure by RNA-composer (http://rnacomposer.cs.put.poznan.pl/, accessed on 8 July 2021). Then the target molecule and aptamers were docked using Autodock Tools 1.5.6 (National Biomedical Computation Resource). Results were analyzed using PYMOL 2.3.2 (Schrödinger) and Discovery Studio (Accelrys Inc., San Diego, CA, USA).

Molecular Dynamic (MD) Simulations
The accuracy of molecular docking results was further verified by MD simulations, and the interaction between the aptamers and SQX was explored. The topology and parameters of the ligand (SQX) were provided by SwissParam. The 60 ns production procedure was performed by GROMACS to generate trajectories for analysis. The Root Mean Square Deviation (RMSD) was used to measure the stability of the trajectory. When the RMSD curve has an upward sloping trend, it indicates that the conformation of the system might undergo significant movement; when the RMSD curve smoothly oscillates near a certain height, it indicates that the system has reached equilibrium [27]. A system with a low RMSD value indicates small structural deviations and high stability [20]. Finally, the binding energy components were calculated using g_mmpbsa software.

Fluorescence Polarization-Based Binding Affinity and Specificity Assays
To evaluate the affinity of candidate aptamers, 1 µg/mL SQX was mixed with different concentrations of aptamers (0, 25, 50, 100, 200, and 400 nM), then incubated in binding buffer (20 mM Tris-HCl, 100 mM NaCl, 5 mM KCl, 2 mM MgCl 2 , 1 mM CaCl 2 , pH 7.5) for 30 min at room temperature in the dark. Next, graphene oxide was added to the mixed solution and incubated for 10 min. The solution was centrifuged at 13,000 rpm for 10 min. Finally, the fluorescence of the supernatant was measured with exciting/emission at 492/518 nm on a Varioskan™ LUX Multimode Microplate Reader. All experiments were carried out three times. A negative control without SQX was also included to assess nonspecific binding. The dissociation constant (K d ) value was obtained by the nonlinear fitting of Origin 9.0 software (Northampton, MA, USA) and calculated according to the nonlinear regression equation Y = B max × X/(K d + X) [28]; where X represents the concentration of aptamer, Y represents the relative fluorescence intensity, and B max represents the most binding sites. The smaller the K d value, the higher the affinity. On the other hand, 200 nM aptamers were incubated with 1 µg/mL of SQX or other antibiotics (SME, SMZ, SDM, SMM, CTC, OTC, OFL, and CAP) to determine their specificity. The fluorescence was measured after adding graphene oxide, and the difference between them and the negative control was calculated.

Synthesis and Characterization of AuNPs
The AuNPs were prepared according to Grabar et al. with some modifications [29]. First, 0.5 g trisodium citrate was dissolved in 50 mL dd·H 2 O, and a 0.22 µm syringe filter was used to filter the solution to obtain a 1% trisodium citrate solution. Then, 1 mL 1% trisodium citrate was mixed with 50 mL 0.01% HAuCl 4 ·4H 2 O at a boiling state. The mixture was heated until it became wine-red, then chilled to 25 • C. The final solution was stored at 4 • C in dim light. The absorption peak at 520 nm of the resultant AuNPs was measured using a VarioskanTM LUX Multimode Microplate Reader (Thermo Fisher Scientific Inc., Waltham, MA, USA).

Optimization of Detection Conditions
Next, we optimized the detection conditions to achieve high sensitivity with the established aptasensor. Briefly, 20 mM NaCl was added to 180 µL AuNPs and incubated for different times (0, 2, 4, 6, 8, and 10 min). The absorbance of the samples at 520 and 695 nm was detected using a Varioskan™ LUX Multimode Microplate Reader (Thermo Fisher Scientific Inc., Waltham, MA, USA). The highest A 695 /A 520 ratio corresponded to the optimal NaCl time. Different doses of NaCl (0, 10, 20, 30, and 40 mM) were also evaluated. The ideal concentration was determined by the greatest A 695 /A 520 ratio. Additionally, different concentrations of SBA28-1 (0, 20, 40, 60, 80, and 100 nM) were mixed with AuNPs and incubated for 30 min, and incubated with the optimal concentration of NaCl (determined above) for 6 min. Finally, SBA28-1 was incubated with or without SQX for 30 min, then AuNPs and NaCl were sequentially added to the solution for incubation, and different concentrations of RhoB (0, 20, 40, 60, 80, and 100 nM) were added to the above solution for fluorescence detection. The excitation and emission wavelengths were 520 nm and 577 nm, respectively. The fluorescence intensities of the SQX (F) and blank (F 0 ) samples were measured, and the value of ∆F was calculated according to the formula; ∆F = F − F 0 .

Fluorescence Assay for SQX Detection
To verify the sensitivity of the aptasensor, different concentrations of SQX (0, 1.25, 2.5, 5, 10, 20, 40, 80, 160, 320, 640, 1000 ng/mL) were incubated with 60 nM SBA28-1 for 30 min, then 180 µL AuNPs were added for 30 min. After the reaction was completed, 20 mM NaCl was added to this mixture and reacted for 6 min, and 3 µM RhoB was added for 10 min. The fluorescence intensity (F, F 0 ) was measured by the Varioskan™ LUX Multimode Microplate Reader (Thermo Fisher Scientific Inc., Waltham, MA, USA), and the ∆F (∆F = F − F 0 ) was calculated. To verify the specificity of the aptasensor, SQX-like antibiotics were tested, including SME, SMZ, SDM, SMM, CTC, OTC, OFL, and CAP. The concentration of total SQX-like antibiotics was 1 µg/mL, and the detection method was the same as above.

Analysis of Real Samples
Finally, the aptasensor was applied to detect lake and tap water samples from a farm to demonstrate its practicability and accuracy. First, the water sample was previously confirmed to be free of SQX compounds using HPLC. SQX was spiked in the samples to the final concentrations of 50, 100, and 150 ng/mL. To prepare the spiked water samples, the supernatants were filtered using a 0.22 µM syringe filter after centrifugation at 10,000 rpm for 10 min, then spiked with the different concentrations of SQX. Finally, the samples were analyzed with the above aptasensor. The recovery was calculated as follows: recovery rates = (concentration/spiked concentration) × 100% [28].

MD Simulations of Aptamer-Target Interactions
Next, we performed MD simulations to verify the SQX and aptamer molecular docking results. After a 60 ns production simulation of the SQX/aptamers complex, the deviation of the structure from the initial configuration was monitored by the RMSD to measure the system's stability. The RMSD curve of the SQX/SBA28-1 and SQX/SBA40-1 complexes rose slowly from 0 to 15 ns (Figure 2a,c), which suggested that the skeleton of the crucial nucleic acid was moving at this time. During the 15 ns-60 ns interval, the SQX/SBA28-1 complex was balanced since the RMSD curve only fluctuated around 6 Å, and the amplitude stayed within 2 Å. The SQX/SBA40-1 complex was also balanced (the RMSD curve only fluctuated around 4.5 Å, and the amplitude stayed within 2 Å). However, the SQX/SBA24-1 complex was not balanced until 35 ns (the RMSD curve fluctuated around 7.5 Å) (Figure 2b). These results indicated that SBA28-1 and SBA40-1 have higher stability in binding to SQX than to SBA24-1.  [4]. These results demonstrated the effectiveness of molecular docking and MD simulations for aptamer truncation. The specificity of the aptamers was also determined using graphene oxide (Figure 3b). Considering the affinity and specificity of aptamers, SBA28-1 was used for the following experiments. The binding energy of the SQX/SBA28-1 complex was decomposed using g-mmpbsa. Based on the Molecular Mechanics/Poisson Boltzmann (Generalized Born) Surface Area method, the binding energy of the SQX/SBA28-1 complex (−17.995 kJ/mol) was composed of Van der Waals (−20.594 kJ/mol), electrostatic (−3.432 kJ/mol), polar solvation (8.209 kJ/mol), and nonpolar solvation (−2.176 kJ/mol) energies. The Van der Waals energy presented the major contribution to the total binding free energy. The polar solvation energy negatively affected the total binding free energy, in contrast to the Van der Waals, electrostatic, and nonpolar solvation energies.

Principle of the Aptasensor
Next, we developed an aptasensor using the SQX28-1 aptamer to detect SQX quantitatively. The sensing mechanism is illustrated in Figure 4. Without SQX, the AuNPs are expected to adsorb SBA28-1 and form AuNPs-SBA28-1 complexes, which allows the good dispersion of AuNPs in high concentrations of NaCl. In this case, the high fluorescence of RhoB is effectively quenched by the dispersed AuNPs [30,31]. In the presence of SQX, the aptamer can attach to SQX rather than the AuNPs, and this well-folded binding complex is more stable than the AuNPs-SBA28-1. Thus, the AuNPs form aggregates in the presence of NaCl and fail to quench RhoB fluorescence [32,33]. Therefore, the content of SQX in the sample can be detected by the fluorescence of RhoB.

Interactions between the Components of the Aptasensor
Further, we evaluated the absorption spectral responses of AuNPs under different conditions to confirm the feasibility of the proposed aptasensor (Figure 5a). The AuNPs have an absorption peak at 520 nm, and the addition of SQX or SBA28-1 did not change this absorption. Since NaCl can significantly aggregate AuNPs, the absorption peak of AuNPs consequently shifts from 520 to 695 nm with the addition of NaCl. Moreover, the AuNPs can bind to SBA28-1 and stabilize without SQX, implying that AuNPs would disperse again in the presence of NaCl, and the absorption spectrum would still present an absorption peak at 520 nm. However, in the presence of SQX, the aptamer reacted with SQX instead of binding to the AuNPs, so the AuNPs aggregated when NaCl was added. The aggregated AuNPs form a new UV absorption peak at 695 nm.
The diameter of AuNPs was approximately 0.60 nm (Figure S8a), and the addition of SQX did not significantly affect the diameter ( Figure S8b). When SBA28-1 was introduced and adsorbed onto the surface of AuNPs without NaCl, the diameter of AuNPs grew to 1.70 nm ( Figure S8c). In the presence of NaCl, the diameter of AuNPs increased to 64.72 nm ( Figure S8d). However, in the presence of SBA28-1, the AuNPs were no longer aggregated by NaCl, and their diameter was reduced ( Figure S8e). After SQX was added to the solution, the AuNPs aggregated again and the diameter enlarged ( Figure S8f). Then, the fluorescence intensity was determined to confirm the sensing feasibility. RhoB alone had the highest fluorescence intensity, which was lowered by adding AuNPs (Figure 5b). After adding NaCl and SBA28-1 to the solution, the fluorescence intensity of RhoB was almost constant and improved again when SQX was added. The fluorescence intensity of RhoB was not affected by SQX, NaCl, and SBA28-1, nor by the binging of SQX and SBA28-1 ( Figure S9a). Similarly, when AuNPs were in the system, the binding of SQX and SBA28-1 did not affect the fluorescence intensity of RhoB without NaCl ( Figure S9b).

Optimization of Detection Conditions
To optimize the aptasensor, the incubation time with NaCl and the concentrations of NaCl, SBA28-1, and RhoB was evaluated. First, the influence of incubation time with NaCl on detection performance was investigated ( Figure S10a). As the incubation time with NaCl increased, the A 695 /A 520 ratio enhanced until 6 min, then stabilized. Thus, the final incubation time with NaCl was determined as 6 min. The A 695 /A 520 ratio was maximum at 20 mM NaCl ( Figure S10b), which was considered the optimal NaCl concentration. The A 695 /A 520 ratio was continuously decreased until the concentration of SBA28-1 was 60 nM ( Figure S10c). Thus, 60 nM SBA28-1 was selected for subsequent experiments. Finally, the ∆F continued to increase until the concentration of RhoB reached 3 µM, considered the optimal RhoB concentration ( Figure S10d).

Properties of the Aptasensor
Furthermore, the ∆F increased in the linear range of 1.25-160 ng/mL (Figure 6a). The linear regression equation was ∆F = 0.1022 C SQX + 1.1479 (R 2 = 0.9942). The limit of detection (LOD) for SQX was 1.04 ng/mL (LOD = 3 × SD/S; where S represents the slope of the calibration curve, and SD is the standard deviation for the blank). Therefore, the aptasensor presented excellent sensitivity for SQX.
Next, eight structurally related or unrelated antibiotics (SME, SMZ, SDM, SMM, CTC, OTC, OFL, and CAP) were used to study the specificity of the aptasensor for SQX. The ∆F of the aptasensor for other antibiotics, such as SME, was significantly smaller than that of SQX (Figure 6b), showing that the aptasensor has good specificity against SQX.

Validation of the Aptasensor
Based on the optimal reaction conditions determined above, different concentrations of SQX (50, 100, and 150 ng/mL) were added to SQX-free lake water and tap water to validate the aptasensor. Each concentration of SQX was analyzed in five replicates (n = 5). The detection data showed a good recovery rate in the range of 96.1 to 100.1%, and the CVs ranged from 2.2 to 10.2% (Table 2). Altogether, these results demonstrated the accuracy of this detection platform for SQX in water samples.

Conclusions
In summary, we successfully truncated the aptamers and selected SBA28-1 (CCC-TAGGGG) due to its higher affinity (K d = 27.36 nM). The truncated approach was validated by molecular docking and MD simulations. The effects of Van der Waals, electrostatic, and solvation energies on target recognition by the SBA28-1 aptamer were also explored. We found that the Van der Waals energy had a major contribution to the recognition of SQX by SBA28-1. Finally, based on the FRET between AuNPs and RhoB, we established a sensitive and accurate platform to detect SQX in water samples. The linear range of detection for the aptasensor was 1.25-160 ng/mL, and the detection limit was 1.04 ng/mL. The recovery range was 93.1-100.1%, and the CVs were 2.2-10.2%. Therefore, the aptasensor has excellent application prospects in water testing.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/bios12070513/s1, Figure S1 Phylogenetic tree of aptamers binding to SQX. These affinity sequences belong to 6 major families. Figure S2 Secondary structure of aptamers in family I. Figure S3 Secondary structure of aptamers in family II. Figure S4 Secondary structure of aptamers in family III. Figure S5 Secondary structure of aptamers in family IV. Figure S6 Secondary structure of aptamers in family V. Figure S7 Secondary structure of aptamers in family VI. Figure  Author Contributions: X.C., conceptualization, methodology, writing. L.Y., methodology, validation, investigation, and writing. J.T., methodology, investigation, and reviewed the manuscript. X.W., methodology, validation, and investigation. X.Z., methodology, validation, and investigation. L.C., methodology, validation, and investigation. J.L., methodology, validation, and investigation. Y.X. and T.L., supervision and project administration. All authors have read and agreed to the published version of the manuscript. Institutional Review Board Statement: Not applicable.