Fumonisins B1 fluorescent single particle intelligent detection method based on truncated aptamer-mediated nucleic acid cascade amplification

Abstract

The invention discloses a fumonisins B1 fluorescent single particle intelligent detection method based on truncated aptamer mediated nucleic acid cascade amplification, and belongs to the field of biosensing detection. The invention successfully constructs a deep learning assisted fluorescent single-particle aptamer sensing detection method by taking a truncated aptamer with high affinity as a recognition element and AIE dye fluorescence coded microspheres with excellent luminescence property and good dispersibility as a signal marker, and can be used for high-sensitivity detection of FB1 in food. Compared with an existing reported FB1 detection method, the method has the advantages of being wide in linear range, high in sensitivity, low in sample consumption and the like, automatic, rapid and high-throughput recognition and counting detection of red particles in a fluorescence image can be achieved, and the analysis efficiency is greatly improved.

Description

Technical Field

[0001] The invention relates to a fumonisin B1 fluorescent single particle intelligent detection method based on truncated aptamer-mediated nucleic acid cascade amplification, and belongs to the field of biosensor detection. Background Art

[0002] Fumonisin is a type of mycotoxin with a long-chain amino polyol skeleton, mainly produced by Fusarium. Among them, fumonisin B1 (FB1) has the highest toxicity and mortality rate, and has immunotoxicity, neurotoxicity, hepatotoxicity, and renal toxicity. FB1 often contaminates agricultural products such as corn and wheat and their products, and can enter humans and animals through the food chain, causing a variety of serious diseases such as esophageal cancer and liver cancer. Therefore, timely, accurate, and sensitive detection of FB1 in agricultural products and foods is of great significance for ensuring food safety and public health. At present, the traditional detection methods for FB1 mainly include high-performance liquid chromatography (HPLC) and enzyme-linked immunosorbent assay (ELISA). Chromatography has high accuracy, but the detection time is long and the operation is complicated. ELISA can achieve specific detection, but the stability and sensitivity are low. As a new type of biorecognition molecule, nucleic acid aptamers have the advantages of easy chemical synthesis and modification, programmability, low cost, good stability, high affinity, and strong specificity. In recent years, aptamer sensing technologies based on various output signals such as fluorescence, colorimetry and electrochemistry have been developed for FB1 detection. For example, Zhao et al. developed a fluorescence sensing method for FB1 using aptamers to regulate the catalytic activity of carbon dots, with a detection limit of 17.67 ng / mL. Sun et al. constructed a stimulus-responsive colorimetric aptamer sensor based on DNA hydrogel-coated MOF for the detection of FB1 in food, with a detection limit of 0.024 ng / mL. However, these methods still urgently need to be combined with some signal amplification strategies to further improve the sensitivity to meet the needs of trace detection. Summary of the invention

[0003] In view of the above-mentioned deficiencies in the prior art, the present invention provides a fumonisin B1 fluorescent single particle intelligent detection method based on truncated aptamer-mediated nucleic acid cascade amplification, aiming to solve the problems of low sensitivity of the current FB1 biosensor detection method and low efficiency of manual counting in existing fluorescent single particle detection.

[0004] The first technical solution provided by the present invention is a truncated aptamer for specifically recognizing fumonisin B1, and the nucleotide sequence of the aptamer is shown in SEQ ID NO: 1;

[0005] SEQ ID NO: 1:

[0006] AGCACAGAGGTCAGATGCGATCTGGATACCCCTTTGGGGAGACATCCTATGCGT GCT.

[0007] The present invention also provides a method for obtaining the above aptamer, which comprises: using a molecular docking-guided strategy to gradually truncate and optimize the original aptamer FB-80 of FB1 to improve its binding affinity and reduce the synthesis cost, and finally obtaining an optimal truncated aptamer FB-57. The sequence of FB-80 is: AGCAGCACAGAGGTCAGATGCGATCTGGATATTATTTTTGATACCCCTTTGGGGAGACATCCTATGCGTGCTACCGTGAA (SEQ ID NO: 2).

[0008] The second technical solution provided by the present invention is a recognition probe for fumonisin B1, wherein the probe is an Apt-T double strand formed by fully hybridizing the truncated aptamer described in the first technical solution with its complementary Trigger strand, and the sequence of the Trigger strand is: ATCCAGATCGCATCTGACCTCT (SEQ ID NO: 3).

[0009] The third technical solution provided by the present invention is a fumonisin B1 fluorescent single-particle intelligent detection system based on truncated aptamer-mediated nucleic acid cascade amplification, wherein the system comprises the recognition probe, EDC probe, Fuel chain, TtAgo, signal probe and streptavidin-modified magnetic beads described in the second technical solution;

[0010] The EDC probe comprises a SWO complex formed by fully hybridizing a biotin-modified Substrate chain, a Waste chain and an Output chain, the SWO complex is connected to a streptavidin-modified magnetic bead, the sequence of the Substrate chain is: CGAGGTCGATTGAATTTAGGGATAGAGGTCAGATGCGATCTGGAT (SEQ ID NO: 4), the sequence of the Waste chain is: ATCGCATCTGACCTCTATCCC (SEQ ID NO: 5), and the sequence of the Output chain is: P-TAAATTCAATCGACCTCG (SEQ ID NO: 6);

[0011] The sequence of the Fuel chain is: ATCGCATCTGACCTCTATCCCTAAATTCAATCGACCTCG (SEQ ID NO: 7);

[0012] The signal probe comprises a single-stranded DNA (HOOC-ssDNA-Biotin) with carboxyl and biotin modified at both ends respectively, wherein the HOOC-ssDNA-Biotin is connected to red fluorescent coding microspheres (RFEMs) with amino groups on the surface through an amide coupling method, and the sequence of the single-stranded DNA is: TTTTTTTTTTTTCGAGGTCGATTGAATTTATTTTTTTTTTTTTT (SEQID NO: 8); the red fluorescent coding microspheres with amino groups on the surface are obtained by adopting a swelling-diffusion method to incorporate oil-soluble aggregation-induced emission (AIE) dye molecules (TPETPAFN) into monodisperse polystyrene microspheres.

[0013] In certain embodiments, the recognition probe (Apt-T) is prepared by mixing the aptamer FB-57 (10 μM) and its complementary Trigger chain (10 μM) in equal volumes, denaturing at 95°C for 5-10 min, and then slowly cooling to room temperature to allow the two DNA chains to fully hybridize to form an Apt-T double chain.

[0014] In certain embodiments, the preparation of the EDC probe is as follows: 10 μM biotin-modified Substrate chain, 10 μM Waste chain, and 10 μM Output chain are mixed in equal volumes and denatured at 95°C for 5 to 10 minutes, and then slowly cooled to room temperature so that the three DNA chains are fully hybridized to form a SWO complex. Then, streptavidin-modified magnetic beads (SA-MBs) with a final concentration of 1 to 2 mg / mL are added and incubated at 37°C for 30 to 60 minutes to obtain the EDC probe. Finally, the mixture is washed with PBS and resuspended in binding buffer.

[0015] In certain embodiments, the preparation of the signal probe: (1) Preparation of RFEMs: using a swelling-diffusion method to incorporate an oil-soluble aggregation-induced emission (AIE) dye molecule (TPETPAFN) solution into monodisperse polystyrene microspheres (~4 μm) to obtain red fluorescent encoded microspheres (RFEMs), thereby serving as a fluorescent labeling material for single particle imaging;

[0016] (2) Synthesis of RFEM-ssDNA-Biotin: Single-stranded DNA (HOOC-ssDNA-Biotin) modified with carboxyl groups and biotin at both ends was connected to RFEMs with amino groups on the surface by amide coupling method to obtain a signal probe (RFEM-ssDNA-Biotin); the sequence of the ssDNA is shown in SEQ ID NO:8.

[0017] In certain embodiments, in step (1), the mass concentration of monodisperse polystyrene microspheres and TPETPAFN solution is 125:1, and the concentration of the TPETPAFN solution is 0.05-0.4 mg / mL.

[0018] In certain embodiments, in step (2), the volume ratio of HOOC-ssDNA-Biotin to RFEMs is 1:2, the concentration of the HOOC-ssDNA-Biotin is 10 μM, and the concentration of the RFEMs is 10 mg / mL.

[0019] In certain embodiments, the detection system includes 250-500 nM recognition probe, 0.15-0.3 mg / mL LEDC probe, 500-600 nM Fuel chain, 1×reaction buffer, 0.4-0.6 mM MnCl2, 50-100 nM TtAgo, 0.67-0.83 mg / mL signal probe, and 1-1.5 mg / mL streptavidin-modified magnetic beads.

[0020] In certain embodiments, the detection system is reacted in a binding buffer, which includes the following components: 20 mM Tris-HCl, 100 mM NaCl, 2 mM MgCl2, 5 mM KCl, 1 mM CaCl2, pH 7.4.

[0021] The fourth technical solution provided by the present invention is a method for intelligent detection of fumonisin B1 fluorescent single particles based on truncated aptamer-mediated nucleic acid cascade amplification, comprising the following steps:

[0022] S1, incubate the recognition probe and the sample to be tested at 37°C for 30 min, then add 0.15-0.3 mg / mL EDC probe, 500-600 nM Fuel chain and binding buffer, and continue the reaction for 45-60 min;

[0023] S2, after magnetic separation, take the supernatant, add 0.4-0.6 mM MnCl2 and 50-100 nM TtAgo, mix well and react at 75 °C for 30 min, then add 0.67-0.83 mg / mL signal probe and react at 80 °C for 40-60 min;

[0024] S3, add magnetic beads modified with streptavidin at a final concentration of 1-1.5 mg / mL and react at 37°C for 30-60 min; after magnetic separation, take the supernatant and gently drop it on a clean glass slide, and take a single particle fluorescence image using a laser confocal scanning microscope;

[0025] S4, single particle counting analysis of all fluorescence images was performed using the pre-trained YOLOv9 deep learning model.

[0026] In some embodiments, the steps include:

[0027] S1, incubate 10 μL of recognition probe with the sample to be tested at 37°C for 30 min, then add 15 μL of EDC probe, 5 μL of Fuel chain (10 μM) and 60 μL of binding buffer (20 mM Tris-HCl, 100 mM NaCl, 2 mM MgCl2, 5 mM KCl, 1 mM CaCl2, pH 7.4), and continue the reaction for 45 to 60 min;

[0028] S2, after magnetic separation, take 30 μL of supernatant, add 6 μL of reaction buffer (10×), 4 μL of MnCl2 (6 mM), 3 μL of TtAgo (1 μM) and 7 μL of enzyme-free water, mix well and react at 75 °C for 30 min, then add 10 μL of signal probe (4 mg / mL) and react at 80 °C for 40-60 min;

[0029] S3, add 20 μL SA-MBs (5 mg / mL) and react at 37°C for 30-60 min; after magnetic separation, take 5 μL of the reaction supernatant and gently drop it on a clean glass slide, and take a single particle fluorescence image using a laser confocal scanning microscope;

[0030] S4, each sample was measured three times, and at least three fluorescence images were taken each time. Then, the pre-trained YOLOv9 deep learning model was used to perform single particle counting analysis on all fluorescence images, and finally a linear curve was established based on the relationship between different FB1 concentrations and the corresponding number of red fluorescent particles.

[0031] In certain embodiments, the samples are corn and plant-based meat analogs.

[0032] The fifth technical solution provided by the present invention is a biosensor containing the aptamer described in the first technical solution or the probe described in the second technical solution.

[0033] The sixth technical solution provided by the present invention is the use of the aptamer described in the first technical solution, the probe described in the second technical solution or the biosensor described in the fifth technical solution in detecting fumonisin B1.

[0034] Further, the pretreatment of corn samples: the crushed corn was added to acetonitrile-water solution (84%, v / v) and mixed evenly, after ultrasound and centrifugation, the supernatant was collected and filtered, and the filtrate was collected for later use; the pretreatment of plant-based artificial meat: the plant-based artificial meat was made into a fine and uniform powder, added to formic acid-water (0.1%, v / v) and acetonitrile, shook the vortex vigorously, then added MgSO4 and NaCl, the mixture was vigorously shaken, centrifuged, and finally the supernatant was taken and filtered, and the filtrate was collected for later use.

[0035] The technical effects of the present invention are as follows:

[0036] The present invention uses high-affinity truncated aptamers as recognition elements and AIE dye fluorescent encoded microspheres with excellent luminescence performance and good dispersibility as signal markers to successfully construct a deep learning-assisted fluorescent single-particle aptamer sensing detection method that can be used for high-sensitivity detection of FB1 in food. Compared with the existing reported FB1 detection methods, it has the advantages of wide linear range, high sensitivity, and low sample consumption. In addition, the present invention has the following advantages: (1) The molecular docking-guided truncation strategy improves the affinity of the FB1 aptamer and reduces the synthesis cost. The optimal truncated aptamer FB-57 has an affinity that is about 3 times higher than the original chain and a length that is reduced by 23 bases; (2) The EDC-TtAgo cascade design improves the signal amplification efficiency and detection sensitivity, and the detection limit of FB1 is as low as 0.89 pg / mL; (3) The combination of the aptamer and TtAgo expands the target diversity and application range of Argonaute; (4) The introduction of the YOLOv9-based deep learning model can realize the automatic, rapid, and high-throughput recognition and counting detection of red particles in fluorescent images, greatly improving the analysis efficiency. In short, this method is expected to provide a promising tool for sensitive screening of food safety risk factors. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 Schematic diagram of the present invention; (A) is the aptamer-mediated EDC-TtAgo cascade design, and (B) is a schematic diagram of fluorescence imaging and single particle detection.

[0038] Figure 2 The figure shows the optimization results of FB1 aptamer truncation; (A) shows the molecular docking results of FB-80 and FB1, (B) and (C) show the ITC binding affinity analysis of FB-57, (D) shows the binding saturation curve of FB-57 measured by SGI method, and (E) shows the circular dichroism spectrum of FB-57 before and after incubation with FB1.

[0039] Figure 3Preparation and characterization of red fluorescence encoded microspheres (RFEMs); (A) Schematic diagram of RFEMs synthesis, (B) SEM images of blank PS microspheres at different scales, (C) SEM images of RFEMs at different scales, (D) actual photo of blank PS microsphere solution, (E) actual photo of RFEMs solution, (F) hydrodynamic diameter diagram of PS microspheres and RFEMs, (G) Zeta potential diagram of PS microspheres and RFEMs, (H) FT-IR spectra of PS microspheres and RFEMs, (I) confocal fluorescence image of RFEMs solution, and (J) corresponding three-dimensional intensity diagram of RFEMs solution.

[0040] Figure 4 Optimization of dye concentration during RFEMs preparation.

[0041] Figure 5 Synthesis and characterization diagram of RFEM-ssDNA-Biotin; (A) electrophoresis images of the initial HOOC-ssDNA-Biotin solution and the reaction supernatant (1: marker, 2: HOOC-ssDNA-Biotin, 3: reaction supernatant), (B) hydrodynamic diameter and (C) Zeta potential of RFEM-ssDNA-Biotin.

[0042] Figure 6 The feasibility characterization results are as follows; (A) is the electrophoresis verification of the binding of the aptamer to FB1, (B) is the EDC feasibility result diagram, (C) is the electrophoresis verification of the ability of TtAgo to cleave ssDNA, (D) and (E) are the feasibility result diagrams of single particle detection of FB1.

[0043] Figure 7 (A) is a graph showing the performance of fluorescent single particle counting assisted by deep learning; (B) is a graph showing the correlation analysis between the counting results of YOLOv9 and those of other methods;

[0044] Figure 8 Optimization results of (A) EDC reaction time and (B) TtAgo enzyme cleavage time.

[0045] Fig. 9 The following are the evaluation diagrams of FB1 detection performance; (A) is the fluorescence image of the final reaction supernatant with different concentrations of FB1 added (scale bar: 20 μm), (B) is the relationship between the number of red particles in the fluorescence image and the FB1 concentration, (C) is the linear relationship between the relative number of red particles and the logarithmic concentration of FB1, and (D) is the specificity evaluation (****p<0.0001, ns: no significant difference).

[0046] Fig.10The figure is a heat map comparison of the actual sample detection results of the method of the present invention and HPLC. DETAILED DESCRIPTION

[0047] The preferred embodiments of the present invention are described below. It should be understood that the embodiments are for better explaining the present invention and are not used to limit the present invention.

[0048] Test method:

[0049] 1. ITC analysis of aptamer binding affinity

[0050] The aptamer solution dissolved in binding buffer was denatured at 95°C for 10 min and then cooled at 4°C. During titration, 70 μL of aptamer (200 μM) in the loading needle was titrated against 350 μL of FB1 solution (20 μM) in the cell. The parameters were as follows: stirring speed was 750 rpm, the number of titrations was set to 13 drops, the volume of the first drop was 0.4 μL, and the volume of the remaining 12 drops was 3 μL. The binding isotherm was fitted using the "one set of sites" binding model using MicroCalPEAQ-ITC analysis software to obtain the dissociation constant (Kd) value.

[0051] 2. SGI fluorescence analysis of binding affinity

[0052] 1×SGI dye was mixed with different concentrations of aptamers (0, 10, 25, 50, 75, 100, 150, 200, 300 nM) and incubated at room temperature (200 r / min, 5 min). Then an equal volume of FB1 solution with a final concentration of 1 μM was added and incubated for 60 min. Finally, the mixture was transferred to a 96-well plate, and the fluorescence signal (Ex: 495 nm, Em: 525 nm) was recorded with an ELISA reader. The relative fluorescence intensity (ΔF = F0-F) value was calculated, and the saturation binding curve was fitted using GraphPad Prism 9 software. F and F0 are the fluorescence values ​​of the experimental group and the blank group, respectively.

[0053] 3. Circular dichroism experiment

[0054] 10 μM aptamer and 100 μM FB1 were mixed and incubated for 1 hour. At the same time, the same concentration of aptamer alone was used as a control. Then, a circular dichroism spectrometer was used to scan three times at 25°C, with a spectral range of 340-220 nm and a bandwidth of 1 nm.

[0055] The raw materials used in the embodiment:

[0056] Monodisperse aminopolystyrene microspheres (4.0 μm, 2.5% w / v) were purchased from Tianjin Cell Group Technology Co., Ltd. TPETPAFN dye was purchased from Xuzhou Dayang Biochemical Technology Co., Ltd. TtAgo was purchased from New England Biolabs. Pluronic F127 was purchased from Sigma-Aldrich.

[0057] The specific principles of the present invention are as follows:

[0058] like Figure 1 As shown in the figure, the design of this method mainly includes three parts, namely: aptamer recognition for signal transduction, EDC-TtAgo cascade design for signal amplification, and single particle detection for fluorescence signal reading. First, in the presence of target FB1, the aptamer (Apt) in the recognition probe (Apt-T) will specifically bind to FB1, causing the T chain to be released and initiating the subsequent EDC-TtAgo system ( Figure 1 A). Specifically, the T strand first hybridizes with the free pivot region of the Substrate (S) strand and dissociates the Waste (W) strand through a pivot-mediated strand displacement reaction, thereby generating a new pivot region in the middle of the S strand. The Fuel (F) strand can then bind to the new pivot point to initiate a second strand displacement reaction, resulting in the simultaneous release of the T strand and the Output (O) strand. Subsequently, the released T strand can continuously initiate the next round of EDC cycles, thereby generating a large number of O strands. These O strands can further serve as 5'phosphorylated guide DNA (5'P-gDNA) to activate the enzyme cleavage activity of downstream TtAgo for targeted cleavage of the signal probe (RFEM-ssDNA-Biotin), resulting in a significant increase in the RFEMs content in the reaction supernatant after further addition of SA-MBs and magnetic separation. Next, a certain amount of the supernatant was dropped onto a glass slide for single-particle fluorescence imaging ( Figure 1 B). Finally, in order to improve the processing efficiency of fluorescence images and realize automatic, fast, and high-throughput counting detection, the deep learning model YOLOv9 was used to count the number of red bright spots in the obtained fluorescence images and establish a correlation between them and the content of FB1, thereby realizing ultra-sensitive detection of FB1.

[0059] Example 1 Establishment of a fluorescent single particle intelligent detection method for fumonisin B1 based on truncated aptamer-mediated nucleic acid cascade amplification 1. Optimization of truncated FB1 aptamer

[0060] The molecular docking-guided strategy was used to gradually truncate and optimize the aptamer FB-80. First, the molecular docking results of FB-80 and FB1 showed that C6, G10, A11, G12, A59, C62, A64 and T65 in the aptamer structure are the key binding sites ( Figure 2 A).

[0061] The Kd values ​​measured by isothermal titration calorimetry (ITC) and SYBR Green I (SGI) fluorescence were 18.7±29.4 μM and 20.06±2.99 nM, respectively (Table 1).

[0062] Table 1. Kd values ​​of aptamers obtained by ITC and SGI fluorescence methods

[0063]

[0064] Therefore, FB-69 was first obtained by cutting off 3 bases at the 5' end and 8 bases at the 3' end of FB-80. The molecular docking results showed that the binding sites of FB-69 and FB-80 were basically the same; the binding Kd values ​​of FB-69 obtained by ITC and SGI fluorescence were 20.4±14.3μM and 21.77±2.45nM, respectively. FB-69 was further truncated to obtain FB-57, and the molecular docking results showed that G7, A8, G9, G10, T11, C12, G24, G25 and A26 were the key binding sites.

[0065] Compared with FB-69, FB-57 retains some of the binding sites in the stem of FB-69 and exhibits more binding sites on the left side of the ring, namely T11, C12, G24, G25, and A26. The Kd values ​​measured by ITC and SGI fluorescence were 6.7±4.8μM and 7.92±1.22nM, respectively. Figure 2 B, C, D), indicating that FB-57 has significantly higher affinity than FB-69 and FB-80.

[0066] Finally, we attempted to truncate the loop that does not contain the binding site in FB-57 to obtain FB-38. The molecular docking results showed that FB-38 still interacted with FB1, but the key binding site had changed to a certain extent. The Kd values ​​of FB-38 obtained by ITC and SGI methods were 23.3±7.6μM and 29.73±4.82nM, respectively, indicating that its affinity for FB1 was significantly lower than that of FB-57. This may be because although the truncated loop is not directly involved in target binding, it may play an important role in stabilizing the structure of the aptamer. In summary, FB-57 was selected as the optimal truncated aptamer for the establishment of subsequent detection methods.

[0067] Finally, circular dichroism (CD) spectroscopy was used to study the interaction between FB-57 and FB1. Figure 2E shows that FB-57 has a negative peak and a positive peak at 244nm and 274nm, respectively, indicating that it is a B-form DNA. After incubation with FB1, the CD positive peak intensity of FB-57 decreased while the negative peak intensity increased, indicating that the binding of FB-57 to FB1 induced a weakening of the base stacking effect and an enhancement of the helicity of the aptamer. This conformational change confirms the interaction between FB-57 and FB1.

[0068] 2. Preparation and characterization of red fluorescent encoded microspheres (RFEMs)

[0069] In order to achieve good single-particle fluorescence imaging performance, AIE dye molecules (TPETPAFN) were used as fluorescent units and incorporated into polystyrene (PS) microsphere carriers by the swelling-diffusion method to obtain red fluorescent encoded microspheres (RFEMs) ( Figure 3 A). Take 150 μL of monodisperse aqueous solution of polystyrene (PS) microspheres (2.5wt.%) into a 1.5mL centrifuge tube, then add 300 μL of ultrapure water, and after ultrasonic dispersion, add 150 μL of Pluronic F-127 aqueous solution (2wt.%) and vortex mix for 1 min. Then, add 300 μL of AIE dye (dissolved in THF) of different concentrations (0.05, 0.1, 0.2 and 0.4 mg / mL) dropwise, vortex evenly for 10 min, and place on a shaker (600 rpm) to react for 60 to 80 min to achieve diffusion and coating of dye molecules in PS microspheres. Finally, wash the obtained material twice with ultrapure water (9000 rpm, 10 min), redisperse it in ultrapure water, and store it at 4°C for later use.

[0070] Firstly, the concentration of TPETPAFN dye during the preparation of RFEMs was optimized. Figure 4 It shows that when the dye concentration increases from 0.05 mg / mL to 0.1 mg / mL, the fluorescence intensity of RFEMs gradually increases, but further increasing the dye concentration shows a downward trend. Therefore, the optimal doping concentration of the dye is 0.1 mg / mL. SEM images show that the PS microspheres have a uniform spherical morphology and excellent dispersibility ( Figure 3 B). After dye encapsulation, PS microspheres still maintain uniform size and morphology ( Figure 3 C), indicating that the morphology and structure of the samples will not be destroyed after swelling with THF. Figure 3 D and E show that the blank PS solution is milky white and has almost no fluorescence under UV light; while the RFEMs with the same concentration are pink and emit bright orange-red fluorescence under UV light, indicating that the AIE dye is successfully embedded in the PS microspheres. The hydrodynamic diameters of RFEMs and PS microspheres are similar ( Figure 3F), indicating that the PS microsphere carrier did not aggregate and was not destroyed after dye loading. Both PS microspheres and RFEMs have obvious positive potentials, indicating that the surface amino groups were retained during the synthesis of RFEMs ( Figure 3 G). Figure 3 H showed that RFEMs and PS microspheres have similar FT-IR spectra. -1 The bands represent the stretching and bending vibrations of NH, which proves again that the amino group still exists in RFEMs, which is beneficial to the subsequent ssDNA modification. Finally, the single-particle fluorescence imaging performance of RFEMs was investigated using laser confocal fluorescence microscopy. Figure 3 As shown in Figure 1, the RFEMs solution dripped on the slide can easily be observed with the naked eye under the dark field field as a red fluorescent bright spot with distinct particles, and the merged image shows that the fluorescent bright spot is highly overlapped with the monodispersed RFEM particles under the bright field, showing excellent single-particle fluorescence imaging performance. In addition, the red fluorescence is basically evenly distributed throughout the PS microsphere, indicating that the AIE dye is evenly doped. In addition, the fluorescent points in the three-dimensional intensity map show a high signal-to-noise ratio ( Figure 3 J). The above results show that RFEMs was successfully prepared.

[0071] 3. Synthesis and characterization of RFEM-ssDNA-Biotin

[0072] The signal probe (RFEM-ssDNA-Biotin) was obtained by connecting the single-stranded DNA (HOOC-ssDNA-Biotin) with carboxyl and biotin modified at both ends respectively to RFEMs with amino groups on the surface by amide coupling method. Specifically, 1 mg of RFEMs was washed twice with 200 μL MEST buffer (10 mM, pH 6.0, 0.05% Tween-20) and redispersed in 100 μL MES buffer (10 mM, pH 6.0). Then, 50 μL HOOC-ssDNA-Biotin (10 μM) and 10 μL 10 mg / mL freshly prepared 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDAPC) were added, mixed evenly and incubated at room temperature for 60 min. Subsequently, 2 μL of freshly prepared high-concentration EDAPC (100 mg / mL) was added and the reaction continued for 2 h. Finally, the material was collected by centrifugation at 9000 rpm for 10 min, washed twice with water, and redispersed in 1 mL of enzyme-free water for later use.

[0073] The electrophoresis results showed that compared with the initial solution of HOOC-ssDNA-Biotin at the same concentration, the ssDNA band in the supernatant of the coupling reaction disappeared ( Figure 5A), indicating that ssDNA is linked to the RFEMs surface via amide bonds. In addition, after being linked to HOOC-ssDNA-Biotin, the hydrodynamic diameter of RFEMs increases ( Figure 5 B), the zeta potential changes from positive to negative ( Figure 5 C), further confirming the successful combination of RFEMs and HOOC-ssDNA-Biotin.

[0074] 4. Feasibility verification of the method

[0075] The recognition probe formed by premixing the aptamer FB-57 and the Trigger chain (SEQ ID NO: 3) was mixed with FB1 and incubated at 37°C for 30 min. Different control groups were set up at the same time, in which the final concentrations of FB-57, Trigger chain and FB1 were 1 μM, 1 μM and 1000 ng / mL, respectively. The electrophoresis results are shown in Figure 6 As shown in A, an obvious low-mobility band appeared in lane 3 and T (Trigger chain) disappeared, indicating that T hybridized with the aptamer. When FB1 was added, a band corresponding to T appeared in lane 4, indicating that FB-57 could bind to FB1 and cause the T chain to be released.

[0076] The pre-prepared SWO complex was mixed with the Trigger chain, Trigger chain + Fuel chain and Fuel chain respectively and reacted at 37°C for 60 min. Different control groups were set up at the same time. The concentration of each nucleic acid single chain in the final system was 1 μM. The electrophoresis results are shown in Figure 6 As shown in B, when the Trigger chain was mixed with SWO, a band corresponding to the intermediate product STO appeared in lane 7. When T and F were mixed with SWO at the same time, a clear SF double-stranded band was produced, and bands of T and O were observed at the same time (lane 8). However, in the absence of T, F was basically not consumed (lane 9). These results indicate that the EDC reaction is feasible. The enzyme cleavage ability of TtAgo was then verified by electrophoresis.

[0077] 100nM TtAgo and 500nM Output chain were mixed and reacted at 75℃ for 30min, then 100nM ssDNA was added and mixed and reacted at 80℃ for 60min. Different control groups were set up at the same time. The electrophoresis results are as follows Figure 6 As shown in C, the bands in lanes 1 and 2 belong to Output and ssDNA, respectively. In contrast, a new slower band appeared in lane 3, and the ssDNA band disappeared, indicating hybridization of Output and ssDNA. It can be clearly seen from lanes 4-6 that with the addition of Mn 2+ With the increase of Mn concentration, the Output-ssDNA double-stranded band gradually weakened and 2+It completely disappeared at a concentration of 0.4 mM, accompanied by the gradual generation of cleavage bands, indicating that TtAgo can cleave ssDNA under the guidance of Output.

[0078] Finally, the feasibility of FB1 detection is verified as follows:

[0079] 10 μL recognition probe was incubated with 10 μL FB1 at a final concentration of 1 ng / mL at 37°C for 30 min, and then 15 μL LEDC probe, 5 μL Fuel chain (10 μM) and 60 μL binding buffer (20 mM Tris-HCl, 100 mM NaCl, 2 mM MgCl2, 5 mM KCl, 1 mM CaCl2, pH 7.4) were added, and the reaction continued for 60 min. After magnetic separation, 30 μL supernatant was taken, 6 μL reaction buffer (10×), 4 μL MnCl2 (6 mM), 3 μL TtAgo (1 μM) and 7 μL enzyme-free water were added, mixed well, reacted at 75°C for 30 min, and then 10 μL signal probe (4 mg / mL) was added, and the reaction was continued at 80°C for 60 min. Subsequently, 20 μL SA-MBs (5 mg / mL) was added and reacted at 37°C for 30 min. After magnetic separation, 5 μL of the reaction supernatant was gently dropped on a clean glass slide, and single-particle fluorescence images were taken by laser confocal scanning microscopy, and the red particles in the fluorescence images were counted. At the same time, different control groups were set up, including the blank group in which the recognition probe and FB1 were not added to the reaction system, the Trigger group in which the recognition probe was replaced with the Trigger chain in the reaction system and FB1 was not added, and the Apt-T group in which FB1 was not added to the reaction system.

[0080] like Figure 6 As shown in D and E, compared with the addition of the recognition probe, the further addition of FB1 can cause a significant increase in the number of red single particles in the final reaction supernatant. The above results fully demonstrate the detection feasibility of this method.

[0081] 5. Deep learning-assisted fluorescence single particle counting performance

[0082] In order to achieve fast, accurate and high-throughput counting, the YOLOv9 deep learning model was used to automatically count the microsphere bright spots in the image. In order to verify the superiority of YOLOv9 in counting performance, YOLOv9, traditional computer vision methods (findContour), ImageJ software and manual methods were used to count fluorescence images containing different numbers of RFEMs. Figure 7As shown in A, when the number of fluorescent particles in the image is small, these methods can basically accurately identify and count. However, as the number of particles increases, even when some particles overlap and adhere, the accuracy of ImageJ and findContour decreases, while YOLOv9 always maintains a high accuracy. In general, the manual method has good accuracy, but it is time-consuming (2 to 5 minutes per image on average) and laborious, and batch image processing cannot be achieved. ImageJ is a commercial image processing software that has been widely used for single particle counting in images. Although the processing speed of a single image is faster than the manual method, batch processing cannot be achieved. findContour is an image contour detection function that comes with openCV. It can achieve batch image processing, but when the concentration of microspheres is high, it cannot distinguish between overlapping and adherent particle clusters, resulting in a decrease in counting accuracy. The trained YOLOv9 can not only accurately distinguish and count overlapping and adherent microsphere particles, but also takes only about 20ms to process a single image, which can achieve batch input, batch processing and batch output of images, greatly improving detection efficiency. In addition, from Figure 7 B shows that the results of YOLOv9 and the manual counting method are consistent (R 2 =0.999), further confirming the counting accuracy and reliability of YOLOv9.

[0083] 6. Optimization of EDC reaction time and TtAgo enzyme cleavage time

[0084] 1) Optimize the reaction time of EDC

[0085] 10 μL recognition probe was incubated with 10 μL FB1 at a final concentration of 100 ng / mL at 37°C for 30 min, then 15 μL LEDC probe, 5 μL Fuel chain (10 μM) and 60 μL binding buffer (20 mM Tris-HCl, 100 mM NaCl, 2 mM MgCl2, 5 mM KCl, 1 mM CaCl2, pH 7.4) were added, and the reaction was continued for different times (15, 30, 45 and 60 min). After magnetic separation, 30 μL supernatant was taken, 6 μL reaction buffer (10×), 4 μL MnCl2 (6 mM), 3 μL TtAgo (1 μM) and 7 μL enzyme-free water were added, mixed well and reacted at 75°C for 30 min, and then 10 μL signal probe (4 mg / mL) was added and reacted at 80°C for 60 min. Subsequently, 20 μL SA-MBs (5 mg / mL) were added and reacted at 37°C for 30 min. After magnetic separation, 5 μL of the reaction supernatant was gently dropped on a clean glass slide, and single-particle fluorescence images were taken using a laser confocal scanning microscope, and the red particles in the fluorescence images were counted.

[0086] like Figure 8 As shown in A, as the EDC reaction time increases, the number of red particles in the final supernatant fluorescence image of the system gradually increases and tends to stabilize at 45 minutes. Therefore, the optimal reaction time of EDC is 45 minutes.

[0087] 2) Optimization of TtAgo digestion time

[0088] 10 μL recognition probe was incubated with 10 μL FB1 at a final concentration of 100 ng / mL at 37°C for 30 min, then 15 μL LEDC probe, 5 μL Fuel chain (10 μM) and 60 μL binding buffer (20 mM Tris-HCl, 100 mM NaCl, 2 mM MgCl2, 5 mM KCl, 1 mM CaCl2, pH 7.4) were added, and the reaction continued for 45 min. After magnetic separation, 30 μL supernatant was taken, 6 μL reaction buffer (10×), 4 μL MnCl2 (6 mM), 3 μL TtAgo (1 μM) and 7 μL enzyme-free water were added, mixed well and reacted at 75°C for 30 min, then 10 μL signal probe (4 mg / mL) was added, and the reaction was continued at 80°C for different time (10, 20, 40 and 60 min). Subsequently, 20 μL SA-MBs (5 mg / mL) was added and reacted at 37°C for 30 min. After magnetic separation, 5 μL of the reaction supernatant was gently dropped on a clean glass slide, and a single particle fluorescence image was taken using a laser confocal scanning microscope, and the red particles in the fluorescence image were counted.

[0089] like Figure 8 As shown in Figure B, as the enzyme digestion time increases, the number of red particles in the final supernatant fluorescence image of the system gradually increases and stabilizes at 40 minutes. Therefore, the optimal enzyme digestion time of TtAgo is 40 minutes.

[0090] Example 2 FB1 Detection Performance Evaluation

[0091] 1. Evaluate the detection performance of this method

[0092] The details are as follows:

[0093] Preparation of recognition probe (Apt-T): The aptamer FB-57 (10 μM) and its complementary Trigger chain (10 μM) were mixed in equal volumes and denatured at 95°C for 5 min, then slowly cooled to room temperature to allow the two DNA chains to fully hybridize to form an Apt-T double chain. The sequence of the Trigger is: ATCCAGATCGCATCTGACCTCT.

[0094] Preparation of EDC probe: 10 μM biotin-modified Substrate chain, 10 μM Waste chain and 10 μM Output chain were mixed in equal volumes and denatured at 95°C for 5 min, then slowly cooled to room temperature to allow the three DNA chains to fully hybridize to form a SWO complex. Then, streptavidin-modified magnetic beads (SA-MBs) were added at a final concentration of 2 mg / mL and incubated at 37°C for 30 min to obtain the EDC probe. Finally, the mixture was washed with PBS and resuspended in binding buffer.

[0095] 10 μL recognition probe and 10 μL FB1 of different concentrations were incubated at 37°C for 30 min, then 15 μL EDC probe, 5 μL Fuel chain (10 μM) and 60 μL binding buffer (20 mM Tris-HCl, 100 mM NaCl, 2 mM MgCl2, 5 mM KCl, 1 mM CaCl2, pH 7.4) were added, and the reaction continued for 45 min. After magnetic separation, 30 μL supernatant was taken, 6 μL reaction buffer (10×), 4 μL MnCl2 (6 mM), 3 μL TtAgo (1 μM) and 7 μL enzyme-free water were added, mixed well, reacted at 75°C for 30 min, and then 10 μL signal probe (4 mg / mL) was added, and the reaction was continued at 80°C for 40 min. Subsequently, 20 μL SA-MBs (5 mg / mL) was added and reacted at 37°C for 30 min. After magnetic separation, 5 μL of the reaction supernatant was gently dropped on a clean glass slide, and single-particle fluorescence images were taken using a laser confocal scanning microscope. Each sample was measured three times, and at least three fluorescence images were taken each time. The pre-trained YOLOv9 deep learning model was then used to perform single-particle counting analysis on all fluorescence images, and finally a linear curve was established based on the relationship between different FB1 concentrations and the corresponding number of red fluorescent particles.

[0096] from Fig. 9 It can be seen from the fluorescence image of A that as the concentration of the FB1 standard solution increases, the red fluorescent bright spots in the image increase. This is because the addition of FB1 results in more red fluorescent encoded microspheres being retained in the reaction supernatant after the final magnetic separation. Then, YOLOv9 was used to count and analyze the fluorescence images corresponding to different concentrations of FB1 ( Fig. 9 B). The results showed that as the concentration of FB1 increased from 0.001 ng / mL to 100 ng / mL, there was a good linear relationship between the relative number of red fluorescent bright spots (ΔCount) and the logarithm of the target concentration. The linear regression equation can be fitted as follows: y = 51.77x + 178.91 (x = lg C FB1 ), coefficient of determination R 2 is 0.994( Fig. 9C). According to the formula LOD = 3SD / k (SD is the standard deviation of the blank sample, k is the slope of the standard curve), the detection limit was calculated to be 0.89 pg / mL.

[0097] 2. Evaluation of the specificity of this method

[0098] The details are as follows:

[0099] Several other common mycotoxins were selected as interferences for analysis, including ochratoxin A (OTA), aflatoxin B1 (AFB1), deoxynivalenol (DON) and zearalenone (ZEN). The above toxins were added to the above detection system for detection. The results showed that compared with FB1 (10 ng / mL), the number of fluorescent bright spots caused by other mycotoxins (10 times the concentration of FB1) changed very little ( Fig. 9 D), indicating that the proposed method has excellent specificity for FB1, which is mainly due to the good binding affinity of the aptamer and the specific enzymatic activity of TtAgo.

[0100] Example 3 Actual sample detection

[0101] In order to evaluate the effectiveness of this method in detecting FB1 in actual samples, commercially available corn and plant-based artificial meat were used as actual samples, and spike recovery detection experiments were carried out according to the method of Example 2. Both corn and plant-based artificial meat need to be pre-treated. Pre-treatment of corn samples: add acetonitrile-water solution (84%, v / v) to the crushed corn and mix evenly, after ultrasound and centrifugation, collect the supernatant and filter, and collect the filtrate for later use; pre-treatment of plant-based artificial meat: make the plant-based artificial meat into a fine and uniform powder, add it to formic acid-water (0.1%, v / v) and acetonitrile, shake the vortex vigorously, add MgSO4 and NaCl, shake the mixture vigorously, centrifuge, and finally take the supernatant and filter, collect the filtrate for later use.

[0102] The results showed that the recovery of the method was 87.2-113.5%, and the RSD was 5.6-10.9%. In addition, the accuracy of the sensor was verified by HPLC-MS / MS standard method. Fig.10 The results showed that the detection results of this method were consistent with those of HPLC-MS / MS, indicating that this detection method has good accuracy and application potential and can be used for the analysis of FB1 in actual samples.

[0103] Although the present invention has been disclosed as above in the form of a preferred embodiment, it is not intended to limit the present invention. Anyone familiar with this technology can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention should be based on the definition of the claims.

Claims

1. A truncated aptamer for specific recognition of fumonisin B1, characterized in that: The nucleotide sequence of the aptamer is shown in SEQ ID NO:

1.

2. A recognition probe for fumonisin B1, characterized in that: The probe is an Apt-T double strand formed by fully hybridizing the truncated aptamer according to claim 1 and its complementary Trigger strand, and the nucleotide sequence of the Trigger strand is shown in SEQ ID NO:

3.

3. A detection system for fumonisin B1, characterized in that: The system comprises the recognition probe, EDC probe, Fuel chain, TtAgo, signal probe and streptavidin-modified magnetic beads according to claim 2; The EDC probe comprises a SWO complex formed by fully hybridizing a biotin-modified Substrate chain, a Waste chain, and an Output chain, wherein the SWO complex is connected to a streptavidin-modified magnetic bead, the nucleotide sequence of the Substrate chain is shown in SEQ ID NO:4, the nucleotide sequence of the Waste chain is shown in SEQ ID NO:5, and the nucleotide sequence of the Output chain is shown in SEQ ID NO:6; The nucleotide sequence of the Fuel chain is shown in SEQ ID NO: 7; The signal probe comprises a single-stranded DNA with carboxyl and biotin modified at both ends respectively, and the single-stranded DNA with carboxyl and biotin modified at both ends respectively is connected to a red fluorescent coding microsphere with amino groups on the surface through an amide coupling method, and the nucleotide sequence of the single-stranded DNA is shown in SEQ ID NO:8; the red fluorescent coding microsphere with amino groups on the surface is obtained by adopting a swelling-diffusion method to incorporate oil-soluble aggregation-induced emission dye molecules into monodisperse polystyrene microspheres.

4. The detection system according to claim 3, characterized in that: Preparation of recognition probe: The truncated aptamer of claim 1 and its complementary Trigger chain are mixed in equal proportions, denatured at 95° C. for 5 to 10 minutes, and then slowly cooled to room temperature to allow the two DNA chains to fully hybridize to form an Apt-T double chain.

5. The detection system according to claim 3, characterized in that: Preparation of the EDC probe: The biotin-modified Substrate chain, Waste chain and Output chain are mixed in equal proportions and denatured at 95° C. for 5 to 10 minutes, then cooled to room temperature to allow the three DNA chains to fully hybridize to form a SWO complex, and then, magnetic beads modified with streptavidin at a final concentration of 1 to 2 mg / mL are added and incubated to obtain the EDC probe.

6. The detection system according to claim 3, characterized in that: Preparation of the signal probe: (1) Preparation of RFEMs: A 0.05-0.4 mg / mL oil-soluble aggregation-induced emission dye molecule solution was incorporated into monodisperse polystyrene microspheres at a mass concentration ratio of 1:125 using a swelling-diffusion method to obtain red fluorescent encoded microspheres with amino groups on the surface, namely RFEMs; (2) Synthesis of RFEM-ssDNA-Biotin: Single-stranded DNA modified with carboxyl groups and biotin at both ends at a concentration of 5 to 10 μM was connected with RFEMs at a concentration of 5 to 10 mg / mL at a volume ratio of 1:2 by amide coupling method to obtain a signal probe, namely RFEM-ssDNA-Biotin.

7. The detection system according to claim 3, characterized in that: The detection system includes 250-500 nM recognition probe, 0.15-0.3 mg / mL EDC probe, 500-600 nM Fuel chain, 0.4-0.6 mM MnCl2, 50-100 nM TtAgo, 0.67-0.83 mg / mL signal probe, and 1-1.5 mg / mL streptavidin-modified magnetic beads.

8. A method for detecting fumonisin B1, characterized in that: The steps include: S1, incubate the recognition probe and the sample to be tested at 37°C for 30 min, then add 0.15-0.3 mg / mL EDC probe, 500-600 nM Fuel chain and binding buffer, and continue the reaction for 45-60 min; S2, after magnetic separation, take the supernatant, add 0.4-0.6 mM MnCl2 and 50-100 nM TtAgo, mix well and react at 75 °C for 30 min, then add 0.67-0.83 mg / mL signal probe and react at 80 °C for 40-60 min; S3, add magnetic beads modified with streptavidin at a final concentration of 1-1.5 mg / mL and react at 37°C for 30-60 min; after magnetic separation, take the supernatant and gently drop it on a clean glass slide, and take a single particle fluorescence image using a laser confocal scanning microscope; S4, single particle counting analysis of all fluorescence images was performed using the pre-trained YOLOv9 deep learning model.

9. A biosensor comprising the truncated aptamer according to claim 1 or the probe according to claim 2.

10. Use of the truncated aptamer according to claim 1, the probe according to claim 2 or the biosensor according to claim 9 in detecting fumonisin B1.

Images