RPA-CRISPR / Cas12a-based magnetic control ratio fluorescent biosensor as well as preparation method and application thereof

Abstract

The invention provides a magnetic control ratio fluorescent biosensor based on RPA-CRISPR / Cas12a as well as a preparation method and application of the magnetic control ratio fluorescent biosensor, and belongs to the technical field of fluorescent biosensors and pathogenic bacterium detection. The magnetic control ratio fluorescence biosensor comprises a fluorescence reference signal element, a fluorescence response signal element and a specific recognition element, according to the magnetic control ratio fluorescent biosensor, high-sensitivity and specific detection of staphylococcus aureus and salmonella is realized through the magnetic and fluorescent characteristics of the fluorescent reference signal element, the fluorescent characteristic of the fluorescent response signal element and the nucleic acid cleavage activity of Cas12a; the magnetic control ratio fluorescent biosensor has a built-in calibration function, can reduce interference brought by the environment and instruments, has good selectivity to staphylococcus aureus and salmonella, has good anti-interference performance to other bacteria, and can be used for rapid detection in the field of food safety.

Description

Technical Field

[0001] This invention belongs to the field of fluorescent biosensors and pathogen detection technology, specifically relating to a magnetically controlled ratiometric fluorescent biosensor based on RPA-CRISPR / Cas12a, its preparation method, and its application. Background Technology

[0002] Eggs are a vital source of nutrition globally, but they are susceptible to microbial contamination, leading to food safety issues. Staphylococcus aureus and Salmonella are common pathogens found in eggs, causing a range of illnesses from minor skin infections to serious, life-threatening conditions such as sepsis and pneumonia. Therefore, rapid and accurate detection of Staphylococcus aureus and Salmonella is of significant clinical and public health importance.

[0003] Currently, the main methods for detecting Staphylococcus aureus and Salmonella include culture methods, molecular biological methods, and immunological methods. While culture methods offer high specificity, they require long incubation times and lack sensitivity in samples with low initial bacterial counts. Molecular biological methods, such as polymerase chain reaction (PCR) and nucleic acid hybridization, are faster but involve expensive equipment, complex procedures, and are susceptible to contamination in the sample. Immunological methods rely on specific antibodies, which are complex and costly to prepare, and cross-reactivity can affect the accuracy of results. These methods are difficult to widely apply in practical testing; therefore, there is a need to develop a new method for detecting foodborne pathogens that is highly sensitive, specific, and stable. Summary of the Invention

[0004] To address some shortcomings in existing technologies, this invention provides a magnetically controlled ratiometric fluorescence biosensor based on RPA-CRISPR / Cas12a, its fabrication method, and its applications. The magnetically controlled ratiometric fluorescence biosensor comprises a fluorescent reference signal element Eu-Fe3O4@SiO2-COOH, a fluorescent response signal element BQD@ZIF-8-ssDNA2, and a specific recognition element RPA-CRISPR / Cas12a. This biosensor achieves high sensitivity and specificity for the detection of Staphylococcus aureus and Salmonella through the magnetic and fluorescent properties of the fluorescent reference signal element, the fluorescent properties of the fluorescent response signal element, and the nucleic acid cleavage activity of Cas12a. Furthermore, this invention utilizes a smartphone to photograph the detection results and introduces deep learning methods for system modeling and analysis of the photographed images. The magnetically controlled ratiometric fluorescence biosensor has a built-in calibration function, reducing interference from the environment and instruments. It exhibits good selectivity for Staphylococcus aureus and Salmonella, and good anti-interference performance against other bacteria, enabling rapid detection in the field of food safety.

[0005] To achieve the above-mentioned technical objectives, the present invention employs the following technical means:

[0006] This invention first provides a method for fabricating a magnetron-controlled ratiometric fluorescence biosensor based on RPA-CRISPR / Cas12a, the method comprising:

[0007] (1) Preparation of the fluorescent reference signal element Eu-Fe3O4@SiO2-COOH:

[0008] Fe3O4 was dispersed in distilled water, and then europium chloride hexahydrate, benzoyltrifluoroacetone, 1,10-phenanthroline, ammonia, ethanol and tetraethyl orthosilicate were added to it. After mixing evenly, the first stirring reaction was carried out. After the reaction was completed, Eu-Fe3O4@SiO2 was obtained.

[0009] Eu-Fe3O4@SiO2 was dispersed in distilled water, citric acid was added and a second stirring reaction was carried out. After the reaction was completed, the mixture was washed and dried to obtain Eu-Fe3O4@SiO2-COOH.

[0010] Eu-Fe3O4@SiO2-COOH was dispersed in distilled water, and then 1-ethyl-(3-dimethylaminopropyl)carbodiimide hydrochloride and N-hydroxysuccinimide were added. The reaction was then carried out for the third time with stirring. After the reaction was completed, the product was washed and dispersed in distilled water to obtain carboxyl-activated Eu-Fe3O4@SiO2-COOH.

[0011] Preferably, in step (1), the mass ratio of Fe3O4, europium chloride hexahydrate, benzoyltrifluoroacetone, 1,10-phenanthroline, ammonia, ethanol, and tetraethyl orthosilicate is 40~60mg:150~250mg:300~400mg:90~110mg:0.1~0.6mL:5~20mL:0.5~2mL;

[0012] The ratio of Eu-Fe3O4@SiO2 to citric acid is 30~100mg:50~150mg;

[0013] The ratio of Eu-Fe3O4@SiO2-COOH, 1-ethyl-(3-dimethylaminopropyl)carbodiimide hydrochloride, and N-hydroxysuccinimide is 0.05~0.5mg:0.1~5mg:0.1~5mg;

[0014] Both the first and second stirring reactions were carried out at room temperature for 10-20 hours.

[0015] The third stirring reaction was carried out at room temperature for 0.5 to 2 hours.

[0016] (2) Preparation of the fluorescent responsive signal element BQD@ZIF-8-ssDNA2:

[0017] BQD@ZIF-8, 1-ethyl-(3-dimethylaminopropyl)carbodiimide hydrochloride, N-hydroxysuccinimide and ssDNA2 were dispersed in distilled water, and then the mixture was reacted at room temperature. After the reaction was completed, the mixture was centrifuged and washed to obtain BQD@ZIF-8-ssDNA2.

[0018] Preferably, in step (2), the preparation method of BQD@ZIF-8 includes:

[0019] Citric acid and polyethyleneimine were dissolved in distilled water and then transferred to a reaction vessel for hydrothermal reaction. After the reaction was completed, blue quantum dots (BQD) were obtained.

[0020] 2-Methylimidazole, nitric acid, BQD and 3-aminopropyltriethoxysilane were dissolved in methanol and stirred to react. After the reaction was completed, the mixture was freeze-dried to obtain BQD@ZIF-8.

[0021] Preferably, the ratio of citric acid to polyethyleneimine is 0.5~5g:0.1~5g;

[0022] The ratio of 2-methylimidazole, nitric acid, BQD and 3-aminopropyltriethoxysilane is 100~200mg: 50~150mg: 5~20mg: 0.1~5mg;

[0023] The hydrothermal reaction conditions are 150~250℃ for 5~8 hours;

[0024] The stirring reaction was carried out at room temperature for 12-36 hours.

[0025] Preferably, in step (2), the amount of BQD@ZIF-8 is 0.5~1.5 mg, and the ratio of 1-ethyl-(3-dimethylaminopropyl)carbodiimide hydrochloride, N-hydroxysuccinimide and ssDNA2 is 0.3~0.5 mg:0.1~0.3 mg:1~10 μL; the initial concentration of ssDNA2 is 1~10 μM;

[0026] The reaction conditions were: stirring at room temperature for 0.5 to 2 hours;

[0027] The sequence of the ssDNA2 is: 5'-COOH-AAAAAAAAAAGAGAACCTGGG-3' (SEQ ID No:6).

[0028] (3) Fabrication of a magnetron-controlled ratiometric fluorescent biosensor:

[0029] Cas12a protein and crRNA were pre-assembled in NEBuffer 2.1 and enzyme-free water. After assembly, RPA product and ssDNA1 were added, mixed and incubated. After incubation, lysed ssDNA1 was obtained.

[0030] The lysed ssDNA1 was mixed with carboxyl-activated Eu-Fe3O4@SiO2-COOH and BQD@ZIF-8-ssDNA2 and reacted. After the reaction was completed, the precipitate was collected using a magnet and dispersed in distilled water to obtain the magnetically controlled ratio fluorescence biosensor.

[0031] Preferably, in step (3), the volume ratio of Cas12a protein, crRNA, RPA product and ssDNA1 is 0.5~2:0.5~2:1~4:2~6;

[0032] The volume ratio of NEBuffer2.1 to enzyme-free water is 1~5:10~30;

[0033] The initial concentration of Cas12a was 0.5~2 µM;

[0034] The initial concentration of the crRNA was 0.5–2 µM;

[0035] The initial concentration of ssDNA1 is 1~10 µM.

[0036] Preferably, in step (3), the pre-assembly conditions are: incubation at 20~50℃ for 1~20 min;

[0037] The conditions for mixed incubation are: incubation at 20~50℃ for 1~50 min.

[0038] Preferably, in step (3), the volume ratio of the lysed ssDNA1, the carboxyl-activated Eu-Fe3O4@SiO2-COOH and BQD@ZIF-8-ssDNA2 is 10~40:100~200:150~250;

[0039] The concentration of the carboxyl-activated Eu-Fe3O4@SiO2-COOH is 400~600 μg / mL;

[0040] The concentration of the BQD@ZIF-8-ssDNA2 solution is 3~7 mg / mL;

[0041] The mixing reaction is carried out at 20-50°C for 1-30 minutes.

[0042] Preferably, in step (3), the crRNA includes Staphylococcus aureus crRNA or Salmonella crRNA;

[0043] The sequence of the Staphylococcus aureus crRNA is: 5'-UAAUUUCUACUAAGUGUAGAUGUUGAAGUUGCACUAUAUAC-3' (SEQ ID No:3);

[0044] The sequence of the Salmonella crRNA is: 5'-UAAUUUCUACUAAGUGUAGAUAAATAGAAGAGTACGCTTAAAAC-3' (SEQ ID No:4);

[0045] The RPA product includes Staphylococcus aureus RPA product or Salmonella RPA product;

[0046] The sequence of the Staphylococcus aureus RPA product is: 5'-GCATCACAAACAGATAATGGCGTAAATAGAAGTGGTTCTGAAGATCCAACAGTATATAGTGCAACTTCAACTAAAAAATTACATAAAGAACCTGCGACATTAATTAAAGCGATTGATGGTGATACGGTTAAATTAATGT-3' (SEQ ID No:1);

[0047] The sequence of the Salmonella RPA product is: 5'-GCGGCTGCTCGCCTTTGCTGGTTTTAGGTTTGGCGGCGCTACGTTTTGCTTCACGGAATTTAAAATAGAAGAGTACGCTTAAAACCACCGATAAAATAACAAAAACCGGCAGTGGGAATCCCGGCAGAGTTCCCATTGAAATGGTCAAAATAGCCGTAACAACCAATACAAATGGGT-3' (SEQ ID No: 2);

[0048] The sequence of the ssDNA1 is: 5'-HN2-AAAAAAAAACCCAGGTTCTCT-3' (SEQ ID No:5).

[0049] The present invention also provides a magnetically controlled ratio fluorescence biosensor prepared by the above method, wherein the magnetically controlled ratio fluorescence biosensor includes a fluorescence reference signal element Eu-Fe3O4@SiO2-COOH, a fluorescence response signal element BQD@ZIF-8-ssDNA2, and a specific recognition element RPA-CRISPR / Cas12a;

[0050] The Eu-Fe3O4@SiO2-COOH consists of a Fe3O4 core and a SiO2 shell covering the core, wherein the SiO2 shell is doped with Eu.

[0051] The BQD@ZIF-8 exhibits a quasi-cluster structure;

[0052] The specific recognition element is obtained by incubating the nucleic acid isothermal amplification (RPA) product, Cas12a protein, crRNA and ssDNA1 in a base solution NEBuffer 2.1 and enzyme-free water.

[0053] The present invention also provides the application of the above-mentioned magnetically controlled ratio fluorescence biosensor in the detection of Staphylococcus aureus and / or Salmonella.

[0054] Preferably, the application includes Staphylococcus aureus and / or Salmonella in food.

[0055] Preferably, the food includes eggs.

[0056] The present invention also provides a method for detecting Staphylococcus aureus and / or Salmonella in food, the method comprising:

[0057] (1) The sample was added to the magnetron ratio fluorescence biosensor for reaction, and then placed in a sealed dark chamber to capture fluorescence images;

[0058] (2) Construct a deep learning model, input the collected fluorescence images into the deep learning model, perform modeling and analysis, and output quantitative detection results.

[0059] Preferably, in step (1), the image capture is performed under a 365nm light source.

[0060] Step (2), the deep learning model includes: a dataset construction and annotation module, an image preprocessing module, a feature extraction and regression prediction module, a training optimization and model saving module, and an inference and result output module;

[0061] The dataset construction and labeling module organizes training data by naming subfolders by concentration, and reads image paths and corresponding concentration labels through a custom dataset class, outputting sample pairs of "image tensor-true concentration-image path";

[0062] The image preprocessing module unifies the image size to 224×224 and performs normalization processing. During the training phase, random flipping and small-angle rotation can be introduced to enhance the robustness of the model to differences in shooting posture.

[0063] The feature extraction and regression prediction module uses a ResNet18 residual convolutional neural network as the backbone feature extractor and replaces its last fully connected layer with a linear regression layer with an output dimension of 1, thereby achieving end-to-end regression prediction from fluorescence image to concentration value.

[0064] The training optimization and model saving module uses the mean squared error loss function as the optimization objective and employs the Adam optimizer (learning rate 1×10⁻⁶). -4 The network parameters are backpropagated and iteratively updated. The training rounds are 100, and the optimal weights obtained during the training process are saved as a model file.

[0065] The reasoning and result output module performs the same preprocessing process as training on the image to be tested during testing or actual detection and inputs it into the model to obtain the predicted concentration value. Furthermore, it can map continuous predicted values ​​to the most recent concentration level to generate a confusion matrix, thereby simultaneously outputting quantitative prediction results and statistical evaluation from a classification perspective.

[0066] The model performance was calculated using scatter plot fitting of the actual concentration and the predicted concentration. 2 The confusion matrix is ​​obtained by statistically analyzing the discretized predicted level and the actual level.

[0067] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0068] This invention combines Eu-Fe3O4@SiO2-COOH fluorescent material, BQD@ZIF-8-ssDNA2 fluorescent material, and RPA-CRISPR / Cas12a to fabricate a magnetically controlled ratiometric fluorescence biosensor. This biosensor utilizes the magnetic and fluorescence properties of the reference signal Eu-Fe3O4@SiO2-COOH, the fluorescence properties of the response signal BQD@ZIF-8-ssDNA2, and the high sensitivity and specificity of the RPA-CRISPR / Cas12a system to rapidly and accurately detect Staphylococcus aureus and Salmonella. Furthermore, this invention incorporates deep learning to achieve stable and high-precision quantitative prediction of ratiometric fluorescence detection data, and is simple to operate and low in cost. In addition, the sensor exhibits good anti-interference capabilities and maintains high detection accuracy in complex sample environments.

[0069] Compared with existing technologies, the magnetically controlled ratio fluorescence biosensor of the present invention has the following advantages:

[0070] (1) High sensitivity and high specificity: This invention combines the magnetic and fluorescent properties of Eu-Fe3O4@SiO2-COOH, the fluorescent properties of BQD@ZIF-8-ssDNA2, and the single-base resolution capability of the CRISPR / Cas12a system. By specifically recognizing the nuc gene of Staphylococcus aureus or the invA gene of Salmonella and cleaving ssDNA1, accurate detection of extremely low concentrations of pathogenic bacteria can be achieved through magnetic separation. When the magnetically controlled ratio fluorescence biosensor detects Staphylococcus aureus and Salmonella, the detection time is reduced compared to the culture method, and compared to the PCR method, no complex thermal cycling instrument is required.

[0071] When the magnetically controlled ratiometric fluorescence biosensor of this invention detects Staphylococcus aureus and Salmonella, Staphylococcus aureus is detected at 3.8 × 10⁻⁶. 0 Up to 3.8×10 5 It exhibited a broad linear response within the CFU / mL range, with a detection limit as low as 3.8 CFU / mL, and Salmonella detection limits at 4.7 × 10⁻⁶ CFU / mL. 0 Up to 4.7×10 5 It exhibits a wide linear response within the CFU / mL range, with a detection limit as low as 6.2 CFU / mL.

[0072] (2) Rapid detection capability: Traditional culture methods take several days to obtain results, while the RPA technology used in this invention can complete DNA amplification in just tens of minutes, and the CRISPR / Cas12a reaction time is 30 minutes, which can complete the detection process in 105 minutes, greatly shortening the detection time.

[0073] (3) Easy to operate and low cost: Compared with PCR technology, which requires expensive equipment and professional operation, the RPA-CRISPR / Cas12a system used in this invention is easy to operate and does not require high-end experimental equipment, which reduces the detection cost and is suitable for environments with limited resources.

[0074] (4) Built-in calibration function: The present invention uses the fluorescence of Eu-Fe3O4@SiO2-COOH as a reference signal and the fluorescence of BQD@ZIF-8-ssDNA2 as a response signal, which can effectively resist the influence of environmental changes and instrument differences, and improve the accuracy and reliability of experimental results.

[0075] (5) Good anti-interference performance: The sensor of the present invention can specifically identify target DNA and maintain high detection specificity even in the presence of other bacteria, avoiding cross-reaction and interference.

[0076] (6) Good quantitative prediction accuracy: The sensor of the present invention introduces a deep learning model to automatically extract and model the features of ratio fluorescence images, so as to realize the quantitative prediction of the detection results of samples with different concentrations. The regression analysis results show that the determination coefficients of Staphylococcus aureus and Salmonella are 0.9843 and 0.9777, respectively, reflecting that the model has a high goodness of fit.

[0077] (7) Wide range of applications: This invention can be used not only for the detection of pathogens in the medical and health field, but also extended to multiple fields such as food safety and environmental monitoring, and has a very broad market prospect and application value. Attached Figure Description

[0078] Figure 1 Characterization diagrams of Fe3O4 and Eu-Fe3O4@SiO2-COOH materials are shown below; where A is the SEM image of Fe3O4; B is the SEM image of Eu-Fe3O4@SiO2-COOH; C is the TEM image of Fe3O4; D is the TEM image of Eu-Fe3O4@SiO2-COOH; E is the elemental distribution diagram of the Eu-Fe3O4@SiO2-COOH composite material; and F is the individual distribution diagram of carbon (C), oxygen (O), silicon (Si), iron (Fe), and europium (Eu).

[0079] Figure 2 The images show the characterization of the BQD@ZIF-8 material; where A is the XPS image of BQD@ZIF-8; B is the TEM image of BQD@ZIF-8; C is the SEM image of BQD@ZIF-8; D is the EDX mapping image of BQD@ZIF-8; E is the Zn elemental mapping image of BQD@ZIF-8; and F is the N elemental mapping image of BQD@ZIF-8.

[0080] Figure 3 The RPA amplification effect of Staphylococcus aureus (A) and Salmonella (B) was verified by agarose gel electrophoresis (2%).

[0081] Figure 4 To verify the cleavage activity of Staphylococcus aureus (A) and Salmonella (B) after addition to the CRISPR / Cas12a system using PAGE.

[0082] Figure 5 Different concentrations of Staphylococcus aureus (3.8 × 10⁻⁶) were used to detect the presence of Staphylococcus aureus. 0 -3.8×10 5 Fluorescence spectrum image (A) and fluorescence intensity ratio F in the sensor at CUF / mL 435 / F 617 Linear curves (B) showing the logarithm of Staphylococcus aureus concentration and different concentrations of Salmonella (4.7 × 10⁻⁶). 0 -4.7×10 5 Fluorescence spectrum image (C) and fluorescence intensity ratio F in the sensor at CUF / mL 435 / F 617 Linear curve (D) of Salmonella concentration versus logarithm.

[0083] Figure 6 Figure 1 shows the results of the specificity (A), repeatability (B), and stability (C) of the magnetron ratio fluorescence biosensor for detecting Staphylococcus aureus, and the specificity (D), repeatability (E), and stability (F) of the magnetron ratio fluorescence biosensor for detecting Salmonella.

[0084] Figure 7 Performance evaluation of the detection prediction model for Staphylococcus aureus and Salmonella: Confusion matrix (A) and concentration prediction (C) results for Staphylococcus aureus, and confusion matrix (B) and concentration prediction (D) results for Salmonella. Detailed Implementation

[0085] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but the scope of protection of the present invention is not limited thereto. Those skilled in the art will understand that the following embodiments are for illustrative purposes only and should not be considered as limiting the scope of the invention. Where specific techniques or conditions are not specified in the embodiments, they are performed according to the techniques or conditions described in the literature in the field or according to the product instructions. Reagents or instruments used, unless otherwise specified, are all conventional products that can be obtained through market purchases.

[0086] The sequences involved in the following embodiments include:

[0087] The sequence of the Staphylococcus aureus RPA product is: 5'-GCATCACAAACAGATAATGGCGTAAATAGAAGTGGTTCTGAAGATCCAACAGTATATAGTGCAACTTCAACTAAAAAATTACATAAAGAACCTGCGACATTAATTAAAGCGATTGATGGTGATACGGTTAAATTAATGT-3' (SEQ ID No:1); the sequence of the Salmonella RPA product is: 5'-GCGGCTGCTCGCCTTTGCTGGTTTTAGGTTTGGCGGCGCTACGTTTTGCTTCACGGAATTTAAAATAGAAGAGTACGCTTAAAACCACCGATAAAATAACAAAAACCGGCAGTGGGAATCCCGGCAGAGTTCCCATTGAAATGGTCAAAATAGCCGTAACAACCAATACAAATGGGT-3' (SEQ ID No:2), obtained by amplifying Staphylococcus aureus or Salmonella using the DNA isothermal rapid amplification kit from Anpu Future (Changzhou) Biotechnology Co., Ltd.

[0088] The sequence of the Staphylococcus aureus crRNA is: 5'-UAAUUUCUACUAAGUGUAGAUGUUGAAGUUGCACUAUAUAC-3' (SEQ ID No:3), synthesized by Qingke Biotechnology.

[0089] The Salmonella crRNA sequence is: 5'-UAAUUUCUACUAAGUGUAGAUAAATAGAAGAGTACGCTTAAAAC-3' (SEQ ID No:4), synthesized by Qingke Biotechnology.

[0090] The sequence of the ssDNA1 is: 5'-HN2-AAAAAAAAACCCAGGTTCTCT-3' (SEQ ID No:5), synthesized by Sangon Biotech.

[0091] The sequence of the ssDNA2 is: 5'-COOH-AAAAAAAAAAGAGAACCTGGG-3' (SEQ ID No: 6), synthesized by Sangon Biotech.

[0092] Example 1: Fabrication of a Magnetically Controlled Ratiofluorescent Biosensor

[0093] (1) Preparation of Eu-Fe3O4@SiO2-COOH:

[0094] 50 mg of Fe3O4 was ultrasonically dispersed in 50 mL of distilled water. 183.21 mg of Eu(NO3)3·6H2O and 333.21 mg of benzoyltrifluoroacetone (btfa) were added to the homogeneous dispersion while stirring. Then, 0.4 mL of ammonia, 99.11 mg of 1,10-phenanthroline (phen), 10 mL of ethanol, and 1 mL of tetraethyl orthosilicate (TEOS) were added. The mixture was stirred at room temperature for 12 h. After the reaction, the product was washed three times alternately with distilled water and ethanol. Finally, it was vacuum dried at 60 °C for 6 h to obtain Eu-Fe3O4@SiO2.

[0095] 50 mg of the above Eu-Fe3O4@SiO2 product was dispersed in 50 mL of distilled water, and then 100 mg of citric acid was added. The reaction was continued to be stirred at room temperature for 12 h. After the reaction was completed, the product was washed three times alternately with distilled water and ethanol, and then dried under vacuum at 60 °C for 6 h to obtain carboxylated Eu-Fe3O4@SiO2, namely Eu-Fe3O4@SiO2-COOH.

[0096] Next, 20 μL of a mixed solution of 1-ethyl-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC, 0.4 M) and N-hydroxysuccinimide (NHS, 0.2 M) was added to 200 μL of Eu-Fe3O4@SiO2-COOH (0.5 mg / mL) and incubated for 1 h. After incubation, the composite microspheres were collected with a magnet and washed three times with distilled water to obtain carboxyl-activated Eu-Fe3O4@SiO2-COOH.

[0097] Figure 1 Characterization diagrams of Fe3O4 and Eu-Fe3O4@SiO2-COOH materials, from Figure 1 As can be seen from A, Fe3O4 particles are generally spherical or near-spherical, with small particle size, rough surface, and severe agglomeration. Figure 1 In B, Eu-Fe3O4@SiO2 exhibits larger and more rounded, smoother particle features, with relatively blurred particle edge outlines and increased spacing between particles; Figure 1 C shows that the boundaries between Fe3O4 particles are clear, and no obvious coating layer is observed; Figure 1 D clearly shows a typical core-shell structure, with a dark Fe3O4 core surrounded by a ring of light gray, amorphous SiO2 shell, indicating that the coating process was successful and uniform. Figure 1 E is the total elemental distribution diagram of Eu-Fe3O4@SiO2-COOH, where Fe can be clearly observed. Figure 1 F elements are mainly concentrated in the central region of the particles, while Si elements are concentrated in the central region. Figure 1 F) and O ( Figure 1The uniform distribution of F elements on the periphery indicates that a typical core-shell structure is formed through the coating of SiO2. Eu ( Figure 1 F) element is sparsely distributed in the shell region, indicating that it has been successfully introduced into the material, suggesting that Eu is incorporated into the shell structure through coordination with SiO2 surface groups or by being embedded during the coating process. C ( Figure 1 The F element mainly originates from residual organic ligands in the carbon film or sample. The overall results further verify that the Fe3O4 nucleation, SiO2 coating, and Eu doping steps in the material synthesis process were all successfully achieved.

[0098] (2) Preparation of BQD@ZIF-8-ssDNA2:

[0099] 1g of citric acid and 0.5g of polyethyleneimine were dissolved in distilled water and then transferred to a reaction vessel for hydrothermal reaction at 180℃ for 5h. After the reaction was completed, blue quantum dots (BQD) were obtained.

[0100] 100 mg of 2-methylimidazole, 50 mg of nitric acid, 5 mg of BQD and 0.1 mg of 3-aminopropyltriethoxysilane were dissolved in methanol and stirred at room temperature for 24 h. After the reaction was completed, the mixture was freeze-dried to obtain BQD@ZIF-8.

[0101] 20 μL of a mixture of EDC (0.4 M) and NHS (0.2 M), 3 μL of ssDNA2 (5 μM), and 200 μL of BQD@ZIF-8 (5 mg / mL) were mixed and incubated for 1 h. The supernatant was then removed by centrifugation, and the remaining precipitate was redispersed in 200 μL of distilled water to obtain the BQD@ZIF-8-ssDNA2 solution.

[0102] Figure 2 The figure shows the characterization of BQD@ZIF-8 material, which mainly contains Zn, C, N and O elements, consistent with the expected composition of BQD and ZIF-8 composite materials. Figure 2 A). Figure 2 B shows that there is a distinct clump-like structure on the periphery, indicating that BQD was effectively encapsulated within or attached to the surface during the growth of the ZIF-8 framework. Figure 2 C shows that the surface exhibits a stacked structure of polyhedral particles. Figure 2 As can be seen from D, E, and F, Zn and N elements are evenly distributed in the composite particles, indicating that the constituent elements of ZIF-8 are uniformly coated on the material surface.

[0103] (3) RPA amplification of Staphylococcus aureus and Salmonella:

[0104] Staphylococcus aureus (ATCC 6538) and Salmonella (ATCC 50761) were cultured on LB medium at 37°C in a shaker at 160 rpm. A series of dilutions were then prepared and incubated on agar plates to obtain the desired concentrations of Staphylococcus aureus and Salmonella. The steps included: inoculating 100 μL of the dilution onto Hektoen enteric (HE) agar plates and incubating at 37°C for 24 hours; counting the colonies on three parallel plates; and taking the concentration of 3.8 × 10⁻⁶ colonies. 0 -3.8×10 5 Staphylococcus aureus at CUF / mL and a concentration of 4.7 × 10⁻⁶ 0 -4.7×10 5 Salmonella dilutions with CUF / mL were used for extraction and amplification.

[0105] The whole genome DNA of Staphylococcus aureus and Salmonella was extracted from the cultured bacteria using a bacterial whole genome DNA extraction kit (Tiangen Biotech (Beijing) Co., Ltd.). The extracted whole genome DNA of Staphylococcus aureus and Salmonella was amplified using a DNA isothermal rapid amplification kit to obtain RPA products. The nucleotide sequences of the RPA products are shown in SEQ ID No:1 and SEQ ID No:2.

[0106] Figure 3 The RPA amplification effect was validated from... Figure 3 As shown in Figure A, lane 1 is the marker, and lanes 2-4 are 3.8 × 10⁻⁶. 2 3.8×10 3 and 3.8×10 4 The RPA product of DNA extracted from Staphylococcus aureus at CFU / mL showed a distinct band between 100 bp and 150 bp in lanes 2-5 compared to the marker, consistent with the theoretical amplification size of 139 bp for the designed primers. Figure 3 As shown in B, lane 1 is the marker, and lanes 2-4 are 4.7 × 10⁻⁶. 2 4.7×10 3 and 4.7×10 4 The RPA product of CFU / mL Salmonella DNA extraction showed a distinct band between 100 bp and 150 bp in lanes 2-5 compared to the marker, consistent with the theoretical amplification size of 177 bp for the designed primers.

[0107] (4) Fabrication of a magnetron-controlled ratiometric fluorescent biosensor:

[0108] 1 μL of 1 μM Cas12a protein, 1 μL of 1 μM crRNA, 3 μL of 10× NEBuffer r2.1 and 20 μL of nuclease-free water were pre-assembled at 25°C for 10 minutes. After assembly, 3 μL of 5 μM ssDNA1 and 2 μL of RPA amplification product were added and incubated at 37°C for 30 minutes. After incubation, the solution was heated to 65°C for 10 minutes to inactivate Cas12a and obtain lysed ssDNA1.

[0109] 150 μL of Eu-Fe3O4@SiO2-COOH (500 μg / mL) and 200 μL of BQD@ZIF-8-ssDNA2 (5 mg / mL) were added to the lysed ssDNA1. The mixture was reacted at 25 °C for 10 min. After the reaction, the precipitate was collected with a magnet and repeatedly dispersed with 400 μL of distilled water to obtain the magneto-controlled ratio fluorescence biosensor.

[0110] The crRNA includes Staphylococcus aureus crRNA or Salmonella crRNA, with amino acid sequences shown in SEQ ID No:3 and SEQ ID No:4, respectively.

[0111] Spectroscopic analysis was performed on the magnetically controlled ratiometric fluorescence biosensor, and the results are as follows: Figure 4 As shown. From Figure 4 It can be seen that the ssDNA band disappears only when Cas12a, crRNA and target DNA are present in the system at the same time, indicating that Cas12a is activated after the target DNA is recognized, triggering its non-specific ssDNA cleavage function, which further proves that the system has good target dependence.

[0112] Example 2: Fabrication of a Magnetically Controlled Ratiofluorescent Biosensor

[0113] (1) Preparation of Eu-Fe3O4@SiO2-COOH:

[0114] 60 mg of Fe3O4 was ultrasonically dispersed in 60 mL of distilled water. 200 mg of Eu(NO3)3·6H2O and 350 mg of btfa were added to the homogeneous dispersion while stirring. Then, 0.5 mL of ammonia, 100 mg of phen, 11 mL of ethanol and 0.5 mL of tetraethyl orthosilicate (TEOS) were added. The mixture was stirred at room temperature for 10 h. After the reaction was completed, the product was washed three times alternately with distilled water and ethanol. Finally, it was vacuum dried at 60 °C for 6 h to obtain Eu-Fe3O4@SiO2.

[0115] 50 mg of the above Eu-Fe3O4@SiO2 product was dispersed in 50 mL of distilled water, and then 120 mg of citric acid was added. The mixture was stirred at room temperature for 10 h. After the reaction was completed, the product was washed three times alternately with distilled water and ethanol, and then dried under vacuum at 60 °C for 6 h to obtain carboxylated Eu-Fe3O4@SiO2, namely Eu-Fe3O4@SiO2-COOH.

[0116] Immediately afterwards, 20 μL of a mixed solution of EDC (0.3 M) and NHS (0.3 M) was added to 200 μL of Eu-Fe3O4@SiO2-COOH (0.5 mg / mL) and incubated for 1 h. The composite microspheres were collected using a magnet and washed three times with distilled water to obtain carboxyl-activated Eu-Fe3O4@SiO2-COOH.

[0117] (2) Preparation of BQD@ZIF-8-ssDNA2:

[0118] 2g of citric acid and 3g of polyethyleneimine were dissolved in distilled water and then transferred to a reaction vessel for hydrothermal reaction at 200℃ for 7h. After the reaction was completed, blue quantum dots (BQD) were obtained.

[0119] 150 mg of 2-methylimidazole, 100 mg of nitric acid, 10 mg of BQD and 1 mg of 3-aminopropyltriethoxysilane were dissolved in methanol and stirred at room temperature for 12 h. After the reaction was completed, the mixture was freeze-dried to obtain BQD@ZIF-8.

[0120] 20 μL of a mixture of EDC (0.3 M) and NHS (0.3 M), 3 μL of ssDNA2 (5 μM), and 300 μL of BQD@ZIF-8 (5 mg / mL) were mixed and incubated for 1 h. The supernatant was then removed by centrifugation, and the remaining precipitate was redispersed in 200 μL of distilled water to obtain the BQD@ZIF-8-ssDNA2 solution.

[0121] (3) RPA amplification of Staphylococcus aureus and Salmonella:

[0122] Staphylococcus aureus (ATCC 6538) and Salmonella (ATCC 50761) were cultured on LB medium at 37°C in a shaker at 160 rpm. A series of dilutions were then prepared. 100 μL of each dilution was inoculated onto Hektoen enteric (HE) agar plates and incubated at 37°C for 24 hours. Colony counts were performed on three parallel plates, and the concentration at 3.8 × 10⁻⁶ was recorded. 0 -3.8×10 5 Staphylococcus aureus at CUF / mL and a concentration of 4.7 × 10⁻⁶ 0-4.7×10 5 Salmonella dilutions with CUF / mL were used for extraction and amplification.

[0123] The whole genome DNA of Staphylococcus aureus and Salmonella was extracted from the cultured bacteria using a bacterial whole genome DNA extraction kit (Tiangen Biotech (Beijing) Co., Ltd.). The extracted whole genome DNA of Staphylococcus aureus and Salmonella was amplified using a DNA isothermal rapid amplification kit to obtain RPA products. The nucleotide sequences of the RPA products are shown in SEQ ID No:1 and SEQ ID No:2.

[0124] (4) Fabrication of a magnetron-controlled ratiometric fluorescent biosensor:

[0125] 1.5 μL of 1.5 μM Cas12a, 1.5 μL of 1.5 μM crRNA, 2 μL of 10× NEBuffer r2.1, and 30 μL of nuclease-free water were pre-assembled at 25°C for 10 minutes. After assembly, 1 μL of 5 μM ssDNA1 and 2 μL of RPA amplification product were added, and the mixture was incubated at 37°C for 30 minutes. After incubation, the solution was heated to 65°C for 10 minutes to inactivate Cas12a, yielding lysed ssDNA1. The crRNA included Staphylococcus aureus crRNA or Salmonella crRNA, with amino acid sequences shown in SEQ ID No:3 and SEQ ID No:4, respectively.

[0126] 100 μL of carboxyl-activated Eu-Fe3O4@SiO2-COOH (500 μg / mL) and 250 μL of BQD@ZIF-8-ssDNA2 (5 mg / mL) were added to the lysed ssDNA1. The mixture was reacted at 25 °C for 10 min. After the reaction, the precipitate was collected with a magnet and repeatedly dispersed with 400 μL of distilled water to obtain the magnetically controlled ratio fluorescence biosensor.

[0127] Example 3: Fabrication of a Magnetically Controlled Ratiofluorescent Biosensor

[0128] (1) Preparation of Eu-Fe3O4@SiO2:

[0129] 40 mg of Fe3O4 was ultrasonically dispersed in 55 mL of distilled water to obtain a Fe3O4 dispersion. 210 mg of Eu(NO3)3·6H2O and 380 mg of btfa were added to the homogeneous dispersion while stirring. After thorough mixing, 0.35 mL of ammonia, 95 mg of phen, 9 mL of ethanol, and 1.5 mL of TEOS were added. The mixture was stirred for 5 h. After the reaction was complete, the product was washed three times alternately with distilled water and ethanol, and finally dried under vacuum at 60 °C for 6 h to obtain Eu-Fe3O4@SiO2.

[0130] 50 mg of the above Eu-Fe3O4@SiO2 was dispersed in 50 mL of distilled water, and then 80 mg of citric acid was added. The mixture was stirred for 5 h. Finally, it was washed three times alternately with distilled water and ethanol, and then dried under vacuum at 60 °C for 6 h to obtain carboxylated Eu-Fe3O4@SiO2, namely Eu-Fe3O4@SiO2-COOH.

[0131] Immediately afterwards, 20 μL of a mixed solution of EDC (0.5 M) and NHS (0.1 M) was added to 200 μL of Eu-Fe3O4@SiO2-COOH (0.5 mg / mL) and incubated for 1 h. The composite microspheres were collected using a magnet and washed three times with distilled water to obtain carboxyl-activated Eu-Fe3O4@SiO2-COOH.

[0132] (2) Preparation of BQD@ZIF-8-ssDNA2:

[0133] 20 μL of a mixture of EDC (0.5 M) and NHS (0.1 M), 3 μL of ssDNA2 (5 μM), and 100 μL of BQD@ZIF-8 (5 mg / mL) were mixed and incubated for 1 h. The supernatant was then removed by centrifugation, and the remaining precipitate was redispersed in 200 μL of distilled water to obtain the BQD@ZIF-8-ssDNA2 solution.

[0134] (3) RPA amplification of Staphylococcus aureus and Salmonella:

[0135] Staphylococcus aureus (ATCC 6538) and Salmonella (ATCC 50761) were cultured on LB medium at 37°C in a shaker at 160 rpm. A series of dilutions were then prepared. 100 μL of each dilution was inoculated onto Hektoen enteric (HE) agar plates and incubated at 37°C for 24 hours. Colony counts were performed on three parallel plates, and the concentration at 3.8 × 10⁻⁶ was recorded. 0 -3.8×10 5 Staphylococcus aureus at CUF / mL and a concentration of 4.7 × 10⁻⁶ 0-4.7×10 5 Salmonella dilutions with CUF / mL were used for extraction and amplification.

[0136] The whole genome DNA of Staphylococcus aureus and Salmonella was extracted from the cultured bacteria using a bacterial whole genome DNA extraction kit (Tiangen Biotech (Beijing) Co., Ltd.). The extracted whole genome DNA of Staphylococcus aureus and Salmonella was amplified using a DNA isothermal rapid amplification kit to obtain RPA products. The nucleotide sequences of the RPA products are shown in SEQ ID No:1 and SEQ ID No:2.

[0137] (4) Fabrication of a magnetron-controlled ratiometric fluorescent biosensor:

[0138] 1 μL of 2 μM Cas12a, 1 μL of 2 μM crRNA, 3 μL of 10× NEBuffer r2.1, and 10 μL of nuclease-free water were pre-assembled at 25°C for 10 minutes. After assembly, 3 μL of 7 μM ssDNA1 and 2 μL of RPA amplification product were added, and the mixture was incubated at 37°C for 30 minutes. After incubation, the solution was heated to 65°C for 10 minutes to inactivate Cas12a, yielding lysed ssDNA1. The crRNA included Staphylococcus aureus crRNA or Salmonella crRNA, with amino acid sequences shown in SEQ ID No:3 and SEQ ID No:4, respectively.

[0139] 160 μL of carboxyl-activated Eu-Fe3O4@SiO2-COOH (400 μg / mL) and 230 μL of BQD@ZIF-8-ssDNA2 (7 mg / mL) were added to the lysed ssDNA1. The mixture was reacted at 25 °C for 10 min. After the reaction, the precipitate was collected with a magnet and repeatedly dispersed with 400 μL of distilled water to obtain the magneto-controlled ratio fluorescence biosensor.

[0140] Example 4: Detection and performance evaluation of Staphylococcus aureus and Salmonella

[0141] This embodiment examines the detection capabilities of the magnetron-controlled ratiometric fluorescence biosensor prepared in Example 1 for Staphylococcus aureus and Salmonella, as detailed below:

[0142] (1) Plotting the standard curve:

[0143] The magnetically controlled ratiometric fluorescence biosensor prepared in Example 1 was used to detect 3.8 × 10⁻⁶ cells. 0 CUF / mL, 3.8×10 1 CUF / mL, 3.8×102 CUF / mL, 3.8×10 3 CUF / mL, 3.8×10 4 CUF / mL, 3.8×10 5 CUF / mL Staphylococcus aureus and 4.7×10 0 CUF / mL, 4.7×10 1 CUF / mL, 4.7×10 2 CUF / mL, 4.7×10 3 CUF / mL, 4.7×10 4 CUF / mL, 4.7×10 5 CUF / mL Salmonella was used to record the fluorescence intensity of Eu-Fe3O4@SiO2-COOH and BQD@ZIF-8-ssDNA2 after adding different concentrations of Staphylococcus aureus or Salmonella. When measuring the fluorescence signal using a fluorescence spectrophotometer, an excitation wavelength of 365 nm was used to measure Eu-Fe3O4@SiO2-COOH and BQD@ZIF-8-ssDNA2.

[0144] A scatter plot was created by comparing the logarithm of Staphylococcus aureus concentration (lg CFU / mL) or Salmonella concentration (lg CFU / mL) with the corresponding fluorescence intensity ratio. Standard curves for the detection of Staphylococcus aureus and Salmonella were established using regression analysis. The results are shown below. Figure 5 As shown.

[0145] from Figure 5 A and Figure 5 As shown in Figure C, with the gradual increase in the concentration of Staphylococcus aureus or Salmonella amplification products, the fluorescence intensity of BQD@ZIF-8-ssDNA2 at 435 nm in the ratiometric fluorescence sensor significantly decreased. Meanwhile, the fluorescence intensity of Eu-Fe3O4@SiO2-COOH at 617 nm remained stable, unaffected by changes in amplification product concentration. This is because the fluorescence emission of Eu-Fe3O4@SiO2-COOH is independent of the presence of Staphylococcus aureus; it serves as an internal reference signal, providing a stable baseline to correct for fluorescence fluctuations caused by external environmental or instrumental differences.

[0146] Figure 5 B and Figure 5 D shows the corresponding fluorescence ratios of the sensor for amplified products of Staphylococcus aureus and Salmonella at different concentrations. A linear equation was constructed by relating the fluorescence ratios to the logarithm of the Staphylococcus aureus or Salmonella concentration. Based on the linear regression curve, the regression equation for detecting Staphylococcus aureus was obtained as y = -0.03468x + 0.3815 (R²). 2=0.9774). Where y and x represent F respectively. 435 / F 617 and lg C S. aureus Furthermore, the logarithm (lg) of the fluorescence ratio to the Staphylococcus aureus concentration was within 3.8 × 10⁻⁶. 0 -3.8×10 5 A good linear relationship was observed between CUF / mL, and the limit of detection (LOD) was 3.8 CFU / mL; the regression equation for the detection of Salmonella was obtained as y = -0.04003x + 0.4067 (R²). 2 =0.9888). Where y and x represent F respectively. 435 / F 617 and lg C Salmonella Furthermore, the logarithm (lg) of the fluorescence ratio to the Salmonella concentration was within 4.7 × 10⁻⁶. 0 -4.7×10 5 A good linear relationship was observed between CUF / mL, and the limit of detection (LOD) was 6.2 CFU / mL.

[0147] (2) Specificity assessment:

[0148] Escherichia coli (ATCC 25922), Listeria monocytogenes (ATCC 54004), Bacillus cereus (CMCC 6303), Vibrio parahaemolyticus (ATCC 17802), and Staphylococcus aureus or Salmonella were respectively subjected to RPA amplification. The RPA amplification products were then used in the magnetically controlled ratio fluorescence biosensor described in Example 1 for detection. The detection results are as follows: Figure 6 A and Figure 6 As shown in D.

[0149] from Figure 6 A and Figure 6 As can be seen from D, the magnetically controlled ratio fluorescence biosensor of the present invention has excellent specificity for Staphylococcus aureus or Salmonella. It is evident that the prepared sensor not only has high specificity for Staphylococcus aureus or Salmonella, but can also effectively eliminate interference from non-target pathogens, greatly improving the accuracy and reliability of detection.

[0150] (3) Repeatability assessment:

[0151] Five parallel assays were performed on bacterial cultures of different concentrations to evaluate the repeatability of the magnetically controlled ratiometric fluorescence biosensor. The results are as follows: Figure 6 B and Figure 6 As shown in E.

[0152] from Figure 6 B and Figure 6As can be seen from E, the relative standard deviations of the obtained ratio fluorescence signals are all below 1.5%, indicating that the sensor has good repeatability.

[0153] (3) Stability assessment:

[0154] This step involved using a magnetically controlled ratiometric fluorescence biosensor to continuously detect the same concentration of Staphylococcus aureus for 20 days to evaluate the stability of the biosensor. The results are as follows: Figure 6 C and Figure 6 As shown in F.

[0155] from Figure 6 C and Figure 6 As can be seen from F, this magnetron-controlled ratiometric fluorescence biosensor exhibits extremely high stability in continuous measurements, with a relative standard deviation of less than 0.51% in the detection of Staphylococcus aureus and less than 0.78% in the detection of Salmonella.

[0156] (4) Evaluation of the predictive capability of deep learning models:

[0157] This step involves constructing a deep learning model and using it to predict the concentration of Staphylococcus aureus or Salmonella. The specific steps are as follows:

[0158] S1. Place the sample in a dark chamber and use a mobile phone to capture fluorescence reaction images under the conditions of a dark chamber and fixed excitation at 365nm. Crop the reaction region ROI of the image, unify the size to 224×224 and normalize it (it can be converted to single channel or grayscale). Then, use a residual convolutional neural network ResNet18 as the feature extraction backbone (Conv1 7×7, stride=2 + max pooling + Layer1–Layer4 residual block + global average pooling GAP) to automatically learn the concentration-related image features. The feature vector obtained after GAP is entered into the output layer to complete the concentration regression prediction.

[0159] S2. Construct training, validation, and test sets using image-concentration labeled data of known concentration gradient samples. Perform backpropagation iterative updates with mean squared error as the optimization objective, and select the weights that perform best on the validation set as the final model.

[0160] The deep learning model includes a dataset construction and labeling module, an image preprocessing module, a feature extraction and regression prediction module, a training optimization and model saving module, and an inference and result output module. The dataset construction and labeling module organizes training data using subfolders named by concentration and reads image paths and corresponding concentration labels through a custom dataset class, outputting sample pairs of "image tensor-true concentration-image path". The image preprocessing module unifies the image size to 224×224 and performs normalization. During training, random flipping and small-angle rotation can be introduced to enhance the model's robustness to differences in shooting posture. The feature extraction and regression prediction module uses a ResNet18 residual convolutional neural network as the backbone feature extractor and replaces its final fully connected layer with a linear regression layer with an output dimension of 1, thereby achieving end-to-end regression prediction from fluorescence images to concentration values. The training optimization and model saving module uses the mean squared error loss function as the optimization objective and employs the Adam optimizer (learning rate 1×10⁻⁶). -4 The network parameters are iteratively updated via backpropagation for 100 training epochs, and the optimal weights obtained during training are saved as a model file. During testing or actual detection, the inference and output module performs the same preprocessing procedure as during training on the image to be tested and inputs it into the model to obtain predicted concentration values. Furthermore, continuous predicted values ​​can be mapped to the most recent concentration level to generate a confusion matrix, thereby simultaneously outputting quantitative prediction results and statistical evaluation from a classification perspective. The model performance is calculated using R² by fitting a scatter plot of the true concentration and the predicted concentration. 2 The confusion matrix is ​​obtained by statistically analyzing the discretized predicted level and the actual level.

[0161] The final model is a regression model based on the residual convolutional neural network ResNet18, which maps the input fluorescence image to the concentration prediction value. The model can be represented as a composite function:

[0162]

[0163] in, The input image tensor (in this embodiment, after preprocessing) , For RGB channels; if converting to single-channel grayscale... ), To predict concentration values, The mapping function is determined by the ResNet18 network structure. This is the set of all trainable parameters of the model (i.e., weights and biases).

[0164] The core of ResNet18 consists of "convolutional / pooling feature extraction + residual block stacking + global average pooling + linear regression output layer". For the l-th residual block, its feature update formula is:

[0165]

[0166] in, and These are the input and output feature tensors of the residual block, respectively. This represents the residual mapping consisting of two convolutional layers (Conv), a normalization process (BN), and an activation function. These are the trainable parameters corresponding to the residual mapping; For non-linear activation functions, ReLU is used in this embodiment.

[0167] After the network completes the extraction of convolutional residual features, the final convolutional feature map is obtained. (in ResNet18) The feature vector is obtained by global average pooling (GAP):

[0168]

[0169] in, Spatial mean for each channel.

[0170] The final regression output layer is a fully linear connected layer (obtained by replacing the fully connected (fc) layer of ResNet18 with a linear layer with an output dimension of 1 in the code), and its model formula is as follows:

[0171]

[0172] in, For the regression output layer weight vector, For the regression output layer bias term, (The full-c input feature dimension of ResNet18 is 512). To predict concentration.

[0173] Figure 7 The performance of the detection prediction model for Staphylococcus aureus and Salmonella was demonstrated, among which Figure 7 A and Figure 7 C represents the confusion matrix and concentration prediction results for Staphylococcus aureus, respectively, indicating that the model has high accuracy in prediction. R 2 The value is 0.9843; Figure 7 B and Figure 7 D represents the predicted results for Salmonella, with predicted values ​​also concentrated near the diagonal. R 2 The value was 0.9777, all showing good prediction accuracy. During the testing phase, the predicted concentration was output for each sample, and a scatter plot of the actual and predicted concentrations was plotted to calculate the goodness of fit. Figure 7 C and Figure 7 R in D 2Value, and Figure 7 A and Figure 7 B's confusion matrix generates a statistical count of the true and predicted levels after discretizing the continuous regression results into levels, thus simultaneously reflecting the accuracy of level discrimination and the ability to quantitatively predict concentration.

[0174] In summary, the magnetically controlled ratiometric fluorescence biosensor of this invention can detect Staphylococcus aureus and Salmonella with high sensitivity and high specificity, wherein Staphylococcus aureus is detected at a concentration of 3.8 × 10⁻⁶. 0 Up to 3.8×10 5 It exhibited a broad linear response within the CFU / mL range, with a detection limit as low as 3.8 CFU / mL, and Salmonella detection limits at 4.7 × 10⁻⁶ CFU / mL. 0 Up to 4.7×10 5 The sensor exhibits a wide linear response within the CFU / mL range, with a detection limit as low as 6.2 CFU / mL. Significant changes in fluorescence signal can be observed after adding RPA products containing DNA extracted from Staphylococcus aureus or Salmonella. The sensor has a built-in calibration function, effectively reducing the impact of environmental variations and instrument errors. It exhibits extremely high selectivity for Staphylococcus aureus and Salmonella, and good anti-interference performance against other bacteria or common contaminants. It can be used for rapid detection in clinical and food safety fields, demonstrating excellent practicality.

[0175] The embodiments described above are preferred embodiments of the present invention, but the present invention is not limited to the above embodiments. Any obvious improvements, substitutions or modifications that can be made by those skilled in the art without departing from the essence of the present invention shall fall within the protection scope of the present invention.

Claims

1. A method for fabricating a magnetron-controlled ratiometric fluorescence biosensor based on RPA-CRISPR / Cas12a, characterized in that, The preparation method includes: (1) Preparation of the fluorescent reference signal element Eu-Fe3O4@SiO2-COOH: Fe3O4 was dispersed in distilled water, and then europium chloride hexahydrate, benzoyltrifluoroacetone, 1,10-phenanthroline, ammonia, ethanol and tetraethyl orthosilicate were added to it. After mixing evenly, the first stirring reaction was carried out. After the reaction was completed, Eu-Fe3O4@SiO2 was obtained. Eu-Fe3O4@SiO2 was dispersed in distilled water, citric acid was added and a second stirring reaction was carried out. After the reaction was completed, the mixture was washed and dried to obtain Eu-Fe3O4@SiO2-COOH. Eu-Fe3O4@SiO2-COOH was dispersed in distilled water, and then 1-ethyl-(3-dimethylaminopropyl)carbodiimide hydrochloride and N-hydroxysuccinimide were added. The reaction was then carried out for the third time with stirring. After the reaction was completed, the product was washed and dispersed in distilled water to obtain carboxyl-activated Eu-Fe3O4@SiO2-COOH. (2) Preparation of the fluorescent responsive signal element BQD@ZIF-8-ssDNA2: BQD@ZIF-8, 1-ethyl-(3-dimethylaminopropyl)carbodiimide hydrochloride, N-hydroxysuccinimide and ssDNA2 were dispersed in distilled water, and then the mixture was reacted at room temperature. After the reaction was completed, the mixture was centrifuged and washed to obtain BQD@ZIF-8-ssDNA2. (3) Fabrication of a magnetron-controlled ratiometric fluorescent biosensor: Cas12a protein and crRNA were pre-assembled in NEBuffer 2.1 and enzyme-free water. After assembly, RPA product and ssDNA1 were added, mixed and incubated. After incubation, lysed ssDNA1 was obtained. The lysed ssDNA1 was mixed with carboxyl-activated Eu-Fe3O4@SiO2-COOH and BQD@ZIF-8-ssDNA2 and reacted. After the reaction was completed, the precipitate was collected using a magnet and dispersed in distilled water to obtain the magnetically controlled ratio fluorescence biosensor.

2. The preparation method according to claim 1, characterized in that, In step (1), the mass ratio of Fe3O4, europium chloride hexahydrate, benzoyltrifluoroacetone, 1,10-phenanthroline, ammonia, ethanol and tetraethyl orthosilicate is 40~60mg:150~250mg:300~400mg:90~110mg:0.1~0.6mL:5~20mL:0.5~2mL; The ratio of Eu-Fe3O4@SiO2 to citric acid is 30~100mg:50~150mg; The ratio of Eu-Fe3O4@SiO2-COOH, 1-ethyl-(3-dimethylaminopropyl)carbodiimide hydrochloride, and N-hydroxysuccinimide is 0.05~0.5mg:0.1~5mg:0.1~5mg; Both the first and second stirring reactions were carried out at room temperature for 10-20 hours. The third stirring reaction was carried out at room temperature for 0.5 to 2 hours.

3. The preparation method according to claim 1, characterized in that, In step (2), the preparation method of BQD@ZIF-8 includes: Citric acid and polyethyleneimine were dissolved in distilled water and then transferred to a reaction vessel for hydrothermal reaction. After the reaction was completed, blue quantum dots (BQD) were obtained. 2-Methylimidazole, nitric acid, BQD and 3-aminopropyltriethoxysilane were dissolved in methanol and stirred to react. After the reaction was completed, the mixture was freeze-dried to obtain BQD@ZIF-8.

4. The preparation method according to claim 1, characterized in that, In step (2), the amount of BQD@ZIF-8 used is 0.5~1.5 mg, and the ratio of 1-ethyl-(3-dimethylaminopropyl)carbodiimide hydrochloride, N-hydroxysuccinimide and ssDNA2 is 0.3~0.5 mg:0.1~0.3 mg:1~10 μL; the initial concentration of ssDNA2 is 1~10 μM; The reaction conditions were: stirring at room temperature for 0.5 to 2 hours; The sequence of the ssDNA2 is: 5'-COOH-AAAAAAAAAAGAGAACCTGGG-3' (SEQ ID No:6).

5. The preparation method according to claim 1, characterized in that, In step (3), the volume ratio of Cas12a protein, crRNA, RPA product and ssDNA1 is 0.5~2:0.5~2:1~4:2~6; The volume ratio of NEBuffer2.1 to enzyme-free water is 1~5:10~30; The initial concentration of Cas12a was 0.5~2 µM; The initial concentration of the crRNA was 0.5–2 µM; The initial concentration of ssDNA1 is 1~10 µM.

6. The preparation method according to claim 1, characterized in that, In step (3), the pre-assembly conditions are: incubation at 20~50℃ for 1~20 min; The conditions for the mixed incubation are: incubation at 20~50℃ for 1~50 min; The volume ratio of the cleaved ssDNA1, carboxyl-activated Eu-Fe3O4@SiO2-COOH, and BQD@ZIF-8-ssDNA2 is 10~40:100~200:150~250. The concentration of the carboxyl-activated Eu-Fe3O4@SiO2-COOH is 400~600 μg / mL; The concentration of the BQD@ZIF-8-ssDNA2 solution is 3~7 mg / mL; The conditions for the mixed reaction are to react at 20~50℃ for 1~30 min.

7. The preparation method according to claim 1, characterized in that, In step (3), the crRNA includes Staphylococcus aureus crRNA or Salmonella crRNA; The sequence of the Staphylococcus aureus crRNA is: 5'-UAAUUUCUACUAAGUGUAGAUGUUGAAGUUGCACUAUAUAC-3' (SEQ ID No:3); The sequence of the Salmonella crRNA is: 5'-UAAUUUCUACUAAGUGUAGAUAAATAGAAGAGTACGCTTAAAAC-3' (SEQ ID No:4); The RPA product includes Staphylococcus aureus RPA product or Salmonella RPA product; The sequence of the Staphylococcus aureus RPA product is: 5'-GCATCACAAACAGATAATGGCGTAAATAGAAGTGGTTCTGAAGATCCAACAGTATATAGTGCAACTTCAACTAAAAAATTACATAAAGAACCTGCGACATTAATTAAAGCGATTGATGGTGATACGGTTAAATTAATGT-3' (SEQ ID No:1); The sequence of the Salmonella RPA product is: 5'-GCGGCTGCTCGCCTTTGCTGGTTTTAGGTTTGGCGGCGCTACGTTTTGCTTCACGGAATTTAAAATAGAAGAGTACGCTTAAAACCACCGATAAAATAACAAAAACCGGCAGTGGGAATCCCGGCAGAGTTCCCATTGAAATGGTCAAAATAGCCGTAACAACCAATACAAATGGGT-3' (SEQ ID No:2); The sequence of the ssDNA1 is: 5'-HN2-AAAAAAAAACCCAGGTTCTCT-3' (SEQ ID No:5).

8. The magnetron-controlled ratiometric fluorescence biosensor prepared by the preparation method according to any one of claims 1 to 7, characterized in that, The magnetized ratio fluorescence biosensor includes a fluorescent reference signal element Eu-Fe3O4@SiO2-COOH, a fluorescent response signal element BQD@ZIF-8-ssDNA2, and a specific recognition element RPA-CRISPR / Cas12a. The Eu-Fe3O4@SiO2-COOH consists of a Fe3O4 core and a SiO2 shell covering the core, wherein the SiO2 shell is doped with Eu. The BQD@ZIF-8 exhibits a quasi-cluster structure; The specific recognition element was obtained by incubating RPA product, Cas12a protein, crRNA and ssDNA1 in a base solution NEBuffer 2.1 and enzyme-free water.

9. The application of the magnetron-controlled ratio fluorescence biosensor of claim 8 in the detection of Staphylococcus aureus and / or Salmonella.

10. A method for detecting Staphylococcus aureus and / or Salmonella in food, characterized in that, The method includes: (1) The sample is added to the magnetically controlled ratio fluorescence biosensor of claim 8 for reaction, and then placed in a closed dark box to take a picture and collect fluorescence images; (2) Construct a deep learning model, input the collected fluorescence images into the deep learning model, perform modeling and analysis, and output quantitative detection results.

Images