Combined specific electrostatic composite fluorescent array sensor as well as preparation method and application thereof

Abstract

The invention discloses a combined specific electrostatic composite fluorescent array sensor and a preparation method and application thereof, the fluorescent array sensor comprises tetramethyl rhodamine (TAMRA) modified cationic antibacterial peptide (AMP) and fluorescein (FAM) modified anionic single-stranded DNA (ssDNA), the ssDNA sequence is shown as SEQ ID NO.1-4, and the AMP sequence is shown as SEQ ID NO.5-7. The invention further discloses a preparation method and application of the combined specific electrostatic composite fluorescent array sensor. The sensor prepared by the invention is relatively high in sensitivity and good in selectivity and does not need to be operated by professional technicians, and a fingerprint spectrum is formed by combining detection results of different bacteria through the cross reactivity of array sensing so as to realize trace detection of different types of bacteria at the same time. According to the sensor prepared by the invention, the pretreatment time of a bacterial sample can be saved, and rapid detection and real-time detection are realized.

Description

Technical Field

[0001] The invention belongs to the field of fluorescence detection sensors, and in particular relates to a combination specific electrostatic composite fluorescence array sensor and a preparation method and application thereof. Background Art

[0002] Bacteria are widely present in human life and can cause a series of environmental pollution and lead to human infections. At present, the main treatment for bacterial infections is broad-spectrum antibiotics, which mainly act on the surface or nucleus of bacteria, affect the normal state of bacteria, and kill bacteria indiscriminately. However, as bacteria mutate, drug-resistant strains gradually increase, and the selection of antibiotics is gradually developing towards more complex advanced antibiotics, and the dosage is greatly increased compared to before. As bacteria gradually become resistant to drugs, it becomes very important to promptly identify the type of bacteria that the patient is infected with and prescribe the right medicine.

[0003] At present, the main bacterial identification technologies include macroscopic colony morphology observation based on laboratory culture, polymerase chain reaction (PCR) and other technologies. Macroscopic colony morphology observation based on laboratory culture is still the gold standard for bacterial identification, but the bacterial culture cycle is too long and cannot provide correct guidance for patient treatment in a timely manner. Polymerase chain reaction has high sensitivity and fast detection speed. However, the instruments used in this method are relatively expensive, and the operators need to undergo professional training, which is difficult to achieve in some underdeveloped areas where bacterial infections are most likely to occur, limiting the widespread application of polymerase chain reaction.

[0004] Sepsis is a common infectious disease in the intensive care unit (ICU). It can easily cause vascular dysfunction and multiple organ failure, and has a rapid onset and high mortality rate. The blood of sepsis patients has complex components and can contain multiple bacteria at the same time, and the golden time for treatment is short. At present, the detection method in general hospitals is still the time-consuming blood culture method. Although this method has accurate results, it will delay the best time for diagnosis of patients.

[0005] The fluorescence analysis method is simple in design, highly sensitive, and has intuitive test results. Fluorescence array sensing technology simulates the functions of the animal olfactory system and taste system to achieve multi-channel signal transmission of data, and has a broad-spectrum response and interactive response. At present, the existing fluorescence array sensors for detecting bacteria have certain defects. When used for bacterial species testing, the detection element must be incubated with the bacteria for a period of time. Although the detection cycle has been shortened compared to traditional methods, it still cannot detect and distinguish bacterial species in a timely and rapid manner. In addition, the construction of this type of sensor is generally more complicated and is not suitable for large-scale promotion and use. Therefore, there is still a need to establish a more convenient, rapid, and accurate detection method for the rapid real-time detection of bacterial infectious diseases such as sepsis. Summary of the invention

[0006] Purpose of the invention: In order to solve the above technical problems, the present invention aims to provide a combination specific electrostatic composite fluorescence array sensor, which solves the problems of needing professionals to pre-treat samples during the test process, long time consumption and expensive detection instruments.

[0007] The present invention also provides a method for preparing a combination-specific electrostatic composite fluorescent array sensor and its application. The fluorescent array sensor prepared by the present invention is simple in construction, easy in operation, and quick in response. It can be applied to the detection of serum from actual clinical samples of sepsis patients, and can quickly and accurately distinguish between a variety of bacterial samples.

[0008] Technical solution: In order to achieve the above-mentioned purpose, the present invention provides a combination specific electrostatic composite fluorescence array sensor, which includes a plurality of sensing units composed of cationic AMP modified by TAMRA (TAMRA-AMP) and anionic ssDNA modified by FAM (FAM-ssDNA), wherein the ssDNA includes any one of the sequences of ssDNA1 to 4 as shown in SEQ ID NO.1 to 4, the AMP includes any one of the sequences of AMP1 to 3 as shown in SEQ ID NO.5 to 7, and the molecular structure of the FAM is as follows:

[0009]

[0010] The TAMRA molecular structure is as follows:

[0011]

[0012] Among them, ssDNA1 SEQ ID NO.1, its sequence is: 5'-GGGACAGGGAGTGCGCTGCTCCC C-3',

[0013] ssDNA2 SEQ ID NO.2, the sequence of which is: 5'-GGGGACTAGAGGACTTGTGCGGCC-3',

[0014] ssDNA3 SEQ ID NO.3, the sequence of which is: 5'-CCTTCTAACAGAATGTTGTTAGATAGCTTTGAAGCTGGTC-3',

[0015] ssDNA4 SEQ ID NO.4, its sequence is: 5'-ATACCAGCTTATTCAATTGGGTGAGGGG GGGTTCACAACGTTAAAGATAGACGGGGGAAGATAGTAAGTGCAATCT-3';

[0016] AMP1, whose sequence is shown in SEQ ID NO.5:

[0017] H2N-Arg-Leu-Phe-Lys-Lys-Leu-Leu-Lys-Tyr-Leu-Arg-Lys-Phe-NH2,

[0018] AMP2, whose sequence is shown in SEQ ID NO.6:

[0019] H2N-Trp-Trp-Trp-Leu-Arg-Lys-Ile-Trp-OH,

[0020] AMP3, the sequence of which is shown in SEQ ID NO.7:

[0021] H2N-Val-Lys-Leu-Phe-Pro-Val-Lys-Leu-Phe-Pro-OH,

[0022] Furthermore, the sensor includes 12 sensing units, which are any one of FAM-ssDNA1 / TAMRA-AMP1, FAM-ssDNA1 / TAMRA-AMP2, FAM-ssDNA1 / TAMRA-AMP3, FAM-ssDNA2 / TAMRA-AMP1, FAM-ssDNA2 / TAMRA-AMP2, FAM-ssDNA2 / TAMRA-AMP3, FAM-ssDNA3 / TAMRA-AMP1, FAM-ssDNA3 / TAMRA-AMP2, FAM-ssDNA3 / TAMRA-AMP3 or FAM-s sDNA4 / TAMRA-AMP1, FAM-ssDNA4 / TAMRA-AMP2, FAM-ssDNA4 / TAMRA-AMP3.

[0023] Preferably, the sensing unit of the sensor includes any one of FAM-ssDNA1 / TAMRA-AMP1, FAM-ssDN A1 / TAMRA-AMP2, FAM-ssDNA1 / TAMRA-AMP3, FAM-ssDNA2 / TAMRA-AMP1, FAM-ssDNA2 / TAMRA-AMP3, FAM-ssDNA3 / TAMRA-AMP3 or FAM-ssDNA4 / TAMRA-AMP3.

[0024] The preparation method of the electrostatic composite fluorescence array sensor of the present invention comprises the following steps: diluting FAM-ssDNA with a buffer solution, and then adding TAMRA-AMP of different concentrations to carry out a composite reaction.

[0025] Furthermore, the concentration of the FAM-ssDNA is 1 to 100 nM, and the concentration of the TAMRA-AMP antimicrobial peptide is 1 to 100 nM.

[0026] Preferably, the preparation process of the electrostatic composite fluorescence array sensor unit is as follows: dilute FAM-ssDNA with PBS buffer solution, then divide it into several portions, add different concentrations of TAMRA-AMP respectively, measure the fluorescence intensity under different temperatures and pH conditions respectively, and select the composite ratio when the TAMRA-AMP concentration is 10nM for constructing the sensor.

[0027] Preferably, a 12×6 well plate array sensor is established based on 12 sensing units, where 12 is the number of columns of the 12 sensing unit well plate array, 6 is the number of wells in each column, each column constitutes a sensor, and the 6 wells in the same column are all filled with 100 μL of PBS solution or serum of the above sensing unit.

[0028] Furthermore, the composite reaction is carried out at 25-37°C for 0.5-1h.

[0029] Preferably, the complexing reaction is carried out at 37° C. for 1 h.

[0030] Application of the electrostatic composite fluorescence array sensor of the present invention in bacteria detection.

[0031] Furthermore, the electrostatic composite fluorescence array sensor is used in the preparation of reagents or tools for bacteria detection.

[0032] Preferably, the fluorescent array sensor is used in real-time rapid detection of sepsis bacterial infectious diseases.

[0033] Furthermore, the application process of the electrostatic composite fluorescence array sensor in bacteria detection is as follows:

[0034] (1) Bacterial culture: The bacterial mother liquid was inoculated on a solid culture medium by the plate streak method and cultured overnight in a constant temperature incubator at 37°C. When the bacteria grew to the stable phase, single colonies with good growth conditions were selected and placed in a liquid culture medium and cultured overnight in a constant temperature shaker at 37°C.

[0035] (2) Collection of bacteria: Centrifuge the culture medium obtained above, discard the supernatant, and resuspend the bacteria in buffer;

[0036] (3) Detection of unknown bacteria: Add the unknown bacteria sample to be tested to the well plate containing the sensor unit and incubate, use a multifunctional microplate reader to measure the fluorescence intensity, and obtain the fluorescence spectrum intensity I of the unknown bacteria sample. Repeat the above process N times to obtain different fluorescence spectrum data I, deduct the corresponding blank value I0, and calculate the relative fluorescence change intensity;

[0037] (4) The "fingerprint" of the unknown bacterial sample to be tested is obtained by a machine learning algorithm based on the fluorescence change intensity in step (3), and is compared with the "fingerprints" constructed in advance for different bacterial standard samples. The unknown bacterial sample is identified as belonging to the bacterial category to which the "fingerprint" of the bacterial standard sample is closest, and the unknown bacterial sample is identified as the category of the bacterial standard sample that matches its "fingerprint".

[0038] The present invention collects the fluorescence response data of the sensing unit of the fluorescence array sensor after the interaction with bacteria, uses the statistical analysis software SYSTAT (version 13.0) to process and analyze the data, classifies the data matrix through a machine learning algorithm, and visualizes the data, thereby constructing a model to detect and distinguish a variety of bacteria. The bacteria can also be blindly tested in a random order, and the test is performed according to the same steps, and the fluorescence response data is recorded. The machine learning algorithm is used to verify the model's ability to detect unknown samples.

[0039] Furthermore, in step (3), the incubation time of the bacteria and the sensor unit is 1 to 2 hours, and the temperature is 25 to 37°C.

[0040] Furthermore, in step (3), the excitation wavelength for measuring the fluorescence intensity is 470-548 nm, the emission wavelength is 520-588 nm, and the number of repetitions is 3 to 7 times.

[0041] Furthermore, the fluorescence change intensity in step (4) is analyzed and trained by a machine learning algorithm to obtain multiple discriminant functions, and two discriminant functions with the largest contribution to discrimination are selected from the obtained multiple discriminant functions as discriminant function F1 and discriminant function F2, and the value of discriminant function F1 is used as the X value, and the value of discriminant function F2 is used as the Y value. With these as the horizontal and vertical coordinates and retaining a 95% confidence interval, a discriminant function scatter plot of different types of bacterial standard samples is drawn. The discriminant function scatter plot can be used as a "fingerprint" of T different types of bacterial standard samples. The relative fluorescence change intensity of the unknown bacterial sample to be tested is classified by a machine learning algorithm. The training results are used to obtain multiple discriminant functions, and the two discriminant functions with the largest contribution to discrimination are selected from the multiple discriminant functions as discriminant function F1 and discriminant function F2. The value of discriminant function F1 is taken as the X value, and the value of discriminant function F2 is taken as the Y value. With these as the horizontal and vertical coordinates and a 95% confidence interval is retained, the "fingerprint" of the unknown bacterial sample is drawn, and the "fingerprint" of the unknown bacterial sample is matched with the "fingerprint" established by standard samples of T different types of bacteria. When the "fingerprint" of the unknown bacterial sample is closest to or overlaps with the "fingerprint" of the standard sample of a certain type of bacteria, the unknown sample is determined to be that type of bacteria.

[0042] Furthermore, the bacteria include bacteria causing sepsis infection, including any one or more combinations of Klebsiella pneumoniae, Proteus mirabilis, Enterobacter hallii, Staphylococcus aureus, Enterococcus faecium, Escherichia coli, Serratia marcescens, Klebsiella oxytoca, Pseudomonas aeruginosa, Staphylococcus lugdunensis, Enterococcus faecalis, Enterobacter aerogenes, Acinetobacter baumannii, and Salmonella.

[0043] Furthermore, the method for distinguishing different types of bacteria is as follows: 100 μL of different types of bacteria (bacterial concentration OD 600 =0.001), including Klebsiella pneumoniae (Bact.1), Proteus mirabilis (Bact.2), Enterobacter hallii (Bact.3), Staphylococcus aureus (Bact.4), Enterococcus faecium (Bact.5), Escherichia coli (Bact.6), Serratia marcescens (Bact.7), Klebsiella oxytoca (Bact.8), Pseudomonas aeruginosa (Bact.9), Staphylococcus lugdunensis (Bact.10), Enterococcus faecalis (Bact.11), Enterobacter aerogenes (Bact.12), Acinetobacter baumannii (Bact.13), and Salmonella (Bact.14). Each bacterium was measured 6 times and the fluorescence response signal was recorded.

[0044] The method for real-time rapid detection of septic bacterial infectious diseases of the present invention comprises the following steps:

[0045] (1) Construction of array sensor and data processing: The constructed electrostatic complex fluorescence array sensor has Z sensing units. A Z×N well plate array sensor is established, where Z is the number of well plate array columns of Z sensing units, and N is the number of wells in each column. Each column constitutes a sensor, and each of the N wells in the same column is filled with 100 μL of the PBS solution of the above sensing unit;

[0046] (2) Bacterial culture: Place different bacteria in MH medium and culture at 37°C with shaking for 8 to 12 h;

[0047] (3) Collection of bacteria: Centrifuge at 3000-5000 rpm for 5 min, discard the supernatant, wash the bacteria three times with saline, and finally resuspend the bacteria in 10 mM PBS buffer, pH = 7.4.

[0048] (4) Establish the fluorescence spectrum data of the bacterial standard sample: add 100 μL of different types of bacterial standard samples to the 96-well plate array sensor, and repeat the above process N times (it is known that the sample N is 6 groups, that is, add 100 μL of different types of bacterial standard samples to the 6 wells in each column). Under the excitation wavelength of 470 nm, collect the signals corresponding to the emission wavelengths at 520 nm and 588 nm. Under the excitation wavelength of 548 nm, collect the signals corresponding to the emission wavelength at 588 nm. According to the number of sensor units, a total of M signal channels are collected (M is the signal channel of Z sensor units). Use a multifunctional microplate reader to measure the fluorescence intensity and obtain the fluorescence spectrum intensity I of the bacterial standard sample. Then, the N×M fluorescence spectrum data I are obtained, and the corresponding blank value I0 is deducted. The relative fluorescence change intensity is calculated by the formula (I-I0) / I0.

[0049] (5) Repeat step (4) for T different types of bacterial samples to obtain T×N×M fluorescence spectrum data I, deduct the corresponding blank value I0, and calculate the corresponding fluorescence change intensity using the formula (I-I0) / I0. Then, the corresponding fluorescence change intensity is analyzed using a machine learning algorithm to obtain the "fingerprint" of T different types of bacteria.

[0050] (6) Detection of unknown bacteria: Add 100 μL of unknown bacteria sample to be tested to the 96-well plate array sensor, and repeat the above process N times (N is 4 groups of unknown samples, that is, 100 μL of different types of bacterial standard samples are added to the 4 wells in each column). At an excitation wavelength of 470 nm, collect the signals corresponding to the emission wavelengths of 520 nm and 588 nm. At an excitation wavelength of 548 nm, collect the signals corresponding to the emission wavelength at 588 nm. According to the number of sensor units, a total of M signal channels are collected (M is the signal channel of Z sensor units). Use a multifunctional microplate reader to measure the fluorescence intensity and obtain the fluorescence spectrum intensity I of the unknown bacterial sample, thus obtaining N×M fluorescence spectrum data I. Deduct the corresponding blank value I0, and calculate the relative fluorescence change intensity by the formula (I-I0) / I0.

[0051] (7) The corresponding fluorescence change intensity of the unknown bacterial sample to be tested in step (6) is compared with the "fingerprint spectrum" of T different types of bacterial samples through a machine learning algorithm, and the unknown bacterial sample is identified as the category of the bacterial standard sample that matches its "fingerprint spectrum".

[0052] (8) Screening the sensor units: The relative fluorescence response values ​​obtained in steps (4) and (6) are respectively subjected to a machine learning algorithm to screen the sensor units with the largest factor contribution. The sensor units finally obtained constitute the electrostatic complex fluorescence array sensor.

[0053] The fluorescent array sensor of the present invention has Z sensing units, and the Z sensing units are composed of a combination of specific complexes formed by multiple TAMRA-AMPs and multiple FAM-ssDNAs through electrostatic action. Among them, FAM-ssDNA can specifically identify multiple common components such as peptidoglycan, lipopolysaccharide and outer membrane vesicles on the surface of bacteria. TAMRA-AMP has the function of identifying a wide spectrum of bacteria. By responding to the different fluorescence signals generated by different types of bacteria and analyzing the fluorescence intensity before and after detection, multiple bacteria can be quickly distinguished.

[0054] The sensing unit of the present invention is compounded by TAMRA-AMP and FAM-ssDNA through electrostatic interaction.

[0055] The present invention uses a variety of machine learning algorithms to process the fluorescent response signals of the array sensor unit and the bacterial sample, classifies the data matrix, and obtains the array visualization "fingerprint" of the bacteria to achieve visual identification. The obtained array differentiation "fingerprint" is used as a model to detect unknown samples, and the accuracy of the unknown sample prediction is calculated to achieve the detection and differentiation of unknown bacteria.

[0056] Design principle: The present invention selects a sensor element that has specific recognition for bacteria. The selected 4 ssDNAs can specifically target the peptidoglycan (SEQ ID NO.1-2), lipopolysaccharide (SEQ ID NO.3) and outer membrane vesicle (SEQ ID NO.4) common to all bacterial surfaces; the selected 3 AMPs are broad-spectrum antimicrobial peptides, which can also be specifically adsorbed on the bacterial surface. Secondly, the positively charged AMP and the negatively charged ssDNA modify TAMRA and FAM respectively. The sensor elements with opposite charges bind through electrostatic interactions, shortening the distance between TAMRA and FAM, allowing TAMRA to play a quenching role, and FAM shows a decrease in fluorescence intensity. On the contrary, the G base in ssDNA can limit the rotation of TAMRA, and TAMRA also shows a decrease in fluorescence intensity. Since FAM-ssDNA and TAMRA-AMP have different abilities to target bacteria, when bacteria are added, the two sensing elements compete with each other to bind to the bacteria, the composite system is destroyed, the party with weaker targeting ability is replaced, and fluorescence is restored; if the targeting abilities are equivalent, the two sensing elements combine with the bacteria to form a ternary composite system, which affects the fluorescence recovery. Therefore, the output of the fluorescence signal depends on the force of the composite system on the bacteria. A single sensing element may not be able to distinguish multiple bacteria at the same time. Therefore, the present invention uses 4 FAM-ssDNAs and 3 TAMRA-AMPs to form 12 sensing units to construct an array sensor. The fluorescence signals output by the 12 sensing units to detect all bacteria are classified by discriminant analysis (LDA) to obtain the distribution of different bacteria, i.e., the "fingerprint spectrum".

[0057] The sensor element used in the present invention is a biomacromolecule with good biocompatibility. Compared with other sensor elements, the sensor element used has a stronger ability to recognize bacteria and can eliminate interference with bacterial detection in the environment. The sensor element does not need to be synthesized and can achieve the purpose of rapid and accurate identification of bacteria through simple compounding.

[0058] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages:

[0059] The sensor prepared by the present invention has high selectivity and does not require professional technicians to operate. Through the cross-reactivity of array sensing, the detection results of different bacteria are combined to form a "fingerprint" to achieve simultaneous trace detection of different types of bacteria. The sensor prepared by the present application scheme can save the pretreatment time of bacterial samples and achieve rapid detection and real-time detection. The sensor prepared by the present invention has a strong ability to recognize bacteria and can eliminate interference with bacterial detection in the environment. The sensor element does not need to be synthesized, and only needs to be simply compounded to achieve the purpose of rapid and accurate identification of bacteria.

[0060] The array sensor preparation method of the present invention is simple, low-cost, and highly sensitive, and can distinguish 84 clinical samples. It can detect multiple samples or multiple targets at the same time, and complete the detection of a large number of samples in a short time. It not only reduces fixed costs, such as equipment and reagent costs, reduces the detection cost of each sample, but also can detect in time. In addition, the sensor can accurately distinguish a variety of bacteria, and has a high accuracy rate and good reproducibility for distinguishing unknown samples. In the test of actual clinical samples, the array sensor has an accuracy rate of 85% in distinguishing the serum of sepsis patients infected with different pathogens from normal people, and has a high practical value.

[0061] The sensor element of the present invention uses commercial fluorophore-modified ssDNA and AMP, which are easy to obtain and do not require professional synthesis, which is conducive to the subsequent transformation of achievements in food safety and medical detection technology. In real life, problems such as food safety, environmental pollution and disease infection are often not caused by a single bacterium. The array retains the advantage of array sensing that it can detect multiple bacteria at the same time and quickly obtain accurate results. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 The UV absorption spectrum and fluorescence spectrum of FAM-ssDNA and the UV absorption spectrum and fluorescence spectrum of TAMRA-AMP;

[0063] Figure 2is a bar graph of relative fluorescence changes exploring the FAM-ssDNA / TAMRA-AMP complex conditions (where ssDNA1: FAM-ssDNA1; AMP3: TAMRA-AMP3);

[0064] Figure 3 It is the fluorescence response signal of 14 kinds of bacteria in PBS to the 8 channels of the constructed array sensor;

[0065] Figure 4 It is the fluorescence response signal of 14 kinds of bacteria in serum to the 8 channels of the constructed array sensor;

[0066] Figure 5 It is a visualized LDA graph of rapid bacterial recognition by the fluorescent array sensor in PBS;

[0067] Figure 6 It is a visualized LDA graph of rapid identification of bacteria in serum by the fluorescence array sensor;

[0068] Figure 7 This is a visualized LDA diagram of the rapid identification of serum from patients with sepsis by the fluorescent array sensor. DETAILED DESCRIPTION

[0069] The technical solution of the present invention is further described below in conjunction with the accompanying drawings.

[0070] Unless otherwise specified, the materials and reagents used in the following examples can be obtained from commercial sources. Experimental methods without specific conditions specified in the examples are usually carried out under conventional conditions or under conditions recommended by the manufacturer.

[0071] The sensor elements described in the present invention are of two types, namely FAM-ssDNA and TAMRA-AMP. The sensor units FAM-ssDNA and TAMRA-AMP in the embodiments are synthesized directly by a biological company through the 5' end of the sequence.

[0072] Sensing units: 12, 50nM FAM-ssDNA1~4 / 10nM TAMRA-AMP1~3;

[0073] Sensor: Each sensor unit has 6 sets of repetitions, and the 6 replicates added in each column constitute a sensor;

[0074] Array sensor: refers to all sensors.

[0075] The 14 strains detected in the present invention are all wild-type strains provided by China Pharmaceutical University.

[0076] Example 1

[0077] Exploring FAM-ssDNA / TAMRA-AMP complexation conditions and complexation ratios

[0078] Since nucleic acids and polypeptides are two important types of biomacromolecules, both are polyelectrolytes, which can carry both positive and negative charges. Therefore, we first used an ELISA reader to determine that the UV absorption of the sensor elements FAM-ssDNA and TAMRA-AMP were 480nm and 558nm, respectively, and the fluorescence emission was 520nm and 588nm, respectively. Figure 1 ). The net charge ratio of FAM-ssDNA and TAMRA-AMP was calculated to be Q = 2:1 at pH = 7.4. According to the charge ratio FAM-ssDNA:TAMRA-AMP = 1:2 (1 / Q), the effects of different temperatures, pH values ​​and incubation times on the composite system were explored when the molar concentration ratio of FAM-ssDNA:TAMRA-AMP was 1:30, 1:10, 1:6, 1:2, 1:1, 1:0.2 and 1:0.1 ( Figure 2 ). It was found that under different pH conditions of 25℃ and 37℃, incubation for 1h had the most obvious quenching effect, and FAM-ssDNA:TAMRA-AMP=1:0.2 maintained a quenching effect of 20% at 1h. At pH=7.4, the quenching degree of the composite system at 37℃ was equivalent to that at 25℃, but the composite system at pH=7.4 and 37℃ showed a stable quenching effect from 0 to 1h. Therefore, the optimal composite conditions of FAM-ssDNA / TAMRA-AMP were determined to be pH=7.4, reaction time 1h, composite temperature 37℃; the composite molar ratio was FAM-ssDNA:TAMRA-AMP=1:0.2. In order to simplify the detection steps, a sensor unit was arbitrarily selected for exploration to unify the detection conditions of all sensor units.

[0079] Example 2

[0080] System construction and application of fluorescence array sensor

[0081] FAM-ssDNA was diluted to a final concentration of 50 nM with PBS buffer (pH = 7.4) and serum (human serum diluted to 2% with PBS), and TAMRA-AMP was diluted to a final concentration of 10 nM with PBS buffer and serum, respectively. The prepared FAM-ssDNA and TAMRA-AMP were complexed at a volume ratio of 1:1 (the complex molar ratio was FAM-ssDNA:TAMRA-AMP=1:0.2), the reaction time was 1 h, the complex temperature was 37°C, and the sensor units of the PBS system and the serum system were obtained, which were stored at 4°C for future use.

[0082] Bacterial pretreatment method: Inoculate the bacterial mother liquid on the solid culture medium by plate streaking method, and culture it in a constant temperature incubator at 37°C overnight. When the bacteria grow to the stable stage, select a single colony with good growth condition, place it in MH culture medium, and culture it in a constant temperature shaker at 37°C overnight. Centrifuge at 5000rmp for 5 minutes, discard the supernatant, wash the bacteria with physiological saline 3 times, and dilute to OD with 10mM, pH=7.4 PBS buffer and 2% serum respectively. 600 =0.002 and then resuspend the bacteria for subsequent sensing of the bacteria (OD 600 =0.001) response.

[0083] Screening process of fluorescence array sensor and bacterial detection in PBS system and serum system, wherein PBS system: 100 μL of PBS sensor unit prepared in this example and different types of bacterial samples OD were added to each well of a 96-well plate. 600 =0.002 (as shown in Table 1) 100 μL, at an excitation wavelength of 470 nm, the emission wavelength is located at 520 nm (channel I) and 588 nm (channel II), at an excitation wavelength of 548 nm, the emission wavelength is located at 588 nm (channel III). The fluorescence intensity is measured using a multifunctional microplate reader, and the relative fluorescence intensity change is used as the detection signal. The corresponding fluorescence change intensity is calculated by the formula (I-I0) / I0, where I is the fluorescence spectrum data, and I0 is the corresponding blank value minus the corresponding blank value. Then, the corresponding fluorescence change intensity is analyzed by a machine learning algorithm, as shown in FIG. Figure 3 As shown, the "fingerprints" of the fluorescence response signals of 14 bacteria in PBS to the 8 channels of the constructed array sensor were obtained, indicating that each 95% confidence interval had no overlap and had a good discrimination effect.

[0084] Each type of bacteria was tested 6 times, with a total of 14 types of bacteria (Table 1). Each sensor unit has 3 signal channels, and 12 sensor units have a total of 36 signal channels (as shown in Table 2). 6×14×36=3024 fluorescence response data can be obtained. PCA is used to screen the 3024 fluorescence response data. According to the sum of the contribution values ​​of each sensor unit to factor 1-factor 9, PCA analysis of bacteria is performed to screen the top 8 channels with the highest contribution value in the PBS system, which are FAM-ssDNA1 / TAMRA-AMP3 (channel 1.III), FAM-ssDNA1 / TAMRA-AMP4 (channel 2.I), FAM-ssDNA1 / TAMRA-AMP5 (channel 3.III), FAM-ssDNA2 / TAMRA-AMP3 (channel 4.I, 4.III), FAM-ssDNA2 / TAMRA-AMP5 (channel 6.III), FAM-ssDNA3 / TAMRA-AMP5 (channel 9.III) and FAM-ssDNA4 / TAMRA-AMP5 (channel 12.I) (Table 2).

[0085] The 7 sensor units (8 channels) obtained by screening were used to detect 14 kinds of bacteria in the serum system. The difference from the PBS system is that 100 μL of the serum sensor unit prepared in this example and the OD values ​​of different kinds of bacterial samples need to be added to each well of the 96-well plate. 600 =0.002 (as shown in Table 1) 100 μL, the same experimental operation as the PBS system was used to obtain the fluorescence response signal "fingerprints" of 14 bacteria in serum to the 8 channels of the constructed array sensor, as shown in Figure 4 As shown, each 95% confidence interval of the array sensor in serum has no overlap, and has a good discrimination effect.

[0086] Table 1 shows the basic information of 14 bacteria

[0087]

[0088]

[0089] Table 2 PCA analysis of bacteria by 12 sensing units (36 channels) (PBS system)

[0090]

[0091]

[0092] The method of distinguishing different types of bacteria in different systems (PBS and serum systems) is as follows: 100 μL of the sensor unit of each system is added to 100 μL of the bacterial sample to be tested, and the signals corresponding to the emission wavelengths at 520 nm and 588 nm are collected at an excitation wavelength of 470 nm, and the signals corresponding to the emission wavelength at 588 nm are collected at an excitation wavelength of 548 nm. The fluorescence intensity is measured using a multifunctional microplate reader, and the relative fluorescence intensity change is the detection signal (I-I0)I0 / . Each type of bacteria was tested six times, for a total of 14 bacteria (Table 1). The seven sensor units screened (all FAM-ssDNA1 / TAMRA-AMP3 (channel 1.III), FAM-ssDNA1 / TAMRA-AMP4 (channel 2.I), FAM-ssDNA1 / TAMRA-AMP5 (channel 3.III), FAM-ssDNA2 / TAMRA-AMP3 (channels 4.I, 4.III), FAM-ssDNA2 / TAMRA-AMP5 (channel 6.III), FAM -ssDNA3 / TAMRA-AMP5 (channel 9.III) and FAM-ssDNA4 / TAMRA-AMP5 (channel 12.I), where FAM-ssDNA2 / TAMRA-AMP3 is repeated once) has 8 signal channels, and 6×14×8=672 fluorescence response data can be obtained. Then, these 672 fluorescence response data are used to form the data matrix of PBS fluorescence array sensor and serum fluorescence array sensor (Table 3, Table 4). The 8 sensing channels of different array sensors have different fluorescence responses to different types of bacteria (Appendix Figure 3 , Figure 4 ), indicating that the two system models performed well on the training data. The data points of each category were closely clustered together and clearly separated from the data points of other categories. The classification boundaries were clear, thus obtaining different "fingerprints". Table 3 Fluorescence response of the array sensor to different bacterial samples in PBS buffer and the machine learning calculation results (bacterial concentration was OD 600 =0.001)

[0093]

[0094]

[0095]

[0096] Table 4 Fluorescence response of the sensor to different bacteria in serum and machine learning calculation results

[0097]

[0098]

[0099]

[0100] Example 3

[0101] Analysis and processing of fluorescence array sensor data

[0102] The corresponding fluorescence change intensity was calculated by the formula (I-I0) / I0, and the fluorescence response data were processed and analyzed using the statistical analysis software SYSTAT (version 13.0). The fluorescence response patterns (Tables 3 and 4) were converted into standard patterns using linear discriminant analysis (LDA). All variables were used in the model (full pattern), and the tolerance was set to 0.001. The Mahalanobis distance between each single pattern and each group in the multidimensional space was calculated, and the assignment of all species was based on the shortest Mahalanobis distance. It can be seen from the LDA graph that the fluorescence array sensors in different systems (PBS and serum systems) can still distinguish each bacterium well even if there are many bacterial species and low concentrations, such as Figure 5 , Figure 6 As shown in the figure, the visualized LDA graph of the rapid identification of bacteria by the fluorescent array sensor in PBS and serum shows that each bacterium can be distinguished into a non-overlapping and separate cluster. The Jackknifed Classification Matrix shows that the accuracy of the array sensor in distinguishing bacteria in PBS buffer and serum is 99% and 100% respectively (Table 5 and Table 6), indicating that the array used has a good effect in distinguishing bacteria.

[0103] Table 5 Verification results of classification matrix accuracy of array sensor for different bacterial samples in PBS buffer

[0104]

[0105] Table 6 Verification results of the classification matrix accuracy of array sensors for different bacterial samples in serum

[0106]

[0107] Example 4

[0108] Differentiation of unknown bacterial samples using fluorescence array sensors

[0109] The 14 types of bacteria in Table 1 were blindly tested in random order, and each type of bacteria was tested 4 times, for a total of 56 unknown samples. The test was performed according to the bacterial detection steps in Example 2, and the relative fluorescence change intensity was recorded (Table 7, Table 8). The analysis and processing method of Example 2 was then used to perform linear discriminant analysis to verify the ability of the model to test unknown samples and distinguish the types of bacteria. After testing 56 unknown samples, the Jackknifed Classification Matrix showed that the array sensor had an accuracy of 89% and 91% in distinguishing bacteria in PBS buffer and serum, respectively. (Table 7, Table 8) This shows that the fluorescent array sensor used has a good effect in distinguishing bacteria in unknown samples.

[0110] Table 7 Fluorescence response of the array sensor to unknown bacterial samples in PBS buffer and machine learning calculation results

[0111]

[0112]

[0113] Table 8 Fluorescence response data array of the sensor to unknown bacteria in serum

[0114]

[0115]

[0116] Example 5

[0117] Differentiation of serum from sepsis patients and normal subjects using fluorescence array sensor

[0118] Serum from 64 patients with sepsis (23 cases of Klebsiella pneumoniae, 12 cases of Pseudomonas aeruginosa, 16 cases of Staphylococcus aureus, and 13 cases of Escherichia coli) and 20 normal subjects was collected and diluted with saline to a final concentration of 2%. 100 μL of serum-equipped fluorescence array optical sensor was added to 100 μL of different sepsis patient or normal serum dilutions. According to the method in Example 2, at an excitation wavelength of 470 nm, signals at emission wavelengths of 520 nm and 588 nm were collected. At an excitation wavelength of 548 nm, signals at an emission wavelength of 588 nm were collected. The fluorescence intensity was measured using a multifunctional microplate reader, and the relative fluorescence intensity change was the detection signal (I-I0) / I0. A total of 84 serum dilution samples were tested 3 times for each sample. The average relative fluorescence intensity of each channel was taken. The array sensor has 8 signal channels, and 3×8×84=2016 fluorescence response data can be obtained. Then, these 2016 fluorescence response data are used to form a data matrix (Table 9). The data is processed by LDA. The array sensor has an accuracy rate of 85% in distinguishing the serum of sepsis patients infected with different pathogens from that of normal people. ( Figure 7 ) This indicates that the array sensor combined with FAM-ssDNA and TAMRA-AMP provides high-sensitivity detection, can detect multiple samples or multiple targets at the same time, and complete the detection of a large number of samples in a shorter time. It not only reduces fixed costs, such as equipment and reagent costs, and reduces the detection cost of each sample, but also means that a large amount of data can be obtained in a single experiment, improving detection efficiency and throughput. In short, the array sensor constructed with the complex of FAM-ssDNA and TAMRA-AMP provides an efficient, sensitive and cost-effective method in clinical sample detection, which is particularly suitable for processing a large number of samples, whether in clinical diagnosis or research environments.

[0119] Table 9 Fluorescence response data matrix of array sensor to patient serum and normal human serum

[0120]

[0121]

[0122]

[0123] In summary, the final concentration of FAM-ssDNA prepared with PBS buffer in the present invention is 50nM, the final concentration of TAMRA-AMP is 10nM, the reaction condition is 1h, and the composite temperature is 37°C, which can be used to construct a combination specific fluorescence array sensor. Bacteria infected with sepsis, including Acinetobacter baumannii and Escherichia coli, are selected as analytes, and different types of bacteria are added to the fluorescence array sensor (bacterial concentration OD 600 =0.001), each type of bacteria was tested 6 times, and a data model of 6 (number of repetitions) × 14 (bacterial types) × 8 (channels) was obtained. The data was analyzed and processed by machine learning algorithms. The accuracy of the fluorescence array sensor in distinguishing 14 types of bacteria in PBS buffer and serum was 99% and 100%, respectively. The accuracy of the unknown sample in distinguishing 14 types of bacteria in PBS buffer and serum was 89% and 91%, respectively. A total of 84 samples of sepsis patient serum and normal human serum dilution were selected, each of which was tested 3 times. The mean of the relative fluorescence change intensity of each channel was taken to construct 3×8×84=2016 fluorescence response data matrix models. The data was processed by LDA. The array sensor had an accuracy of 85% in distinguishing sepsis patients infected with different pathogens from normal human serum.

Claims

1. A combination-specific electrostatic composite fluorescence array sensor, characterized in that: The fluorescent array sensor includes a plurality of sensing units composed of cationic AMP modified by TAMRA (TAMRA-AMP) and anionic ssDNA modified by FAM (FAM-ssDNA), wherein the ssDNA includes any sequence of ssDNA1 to 4 as shown in SEQ ID NOs. 1 to 4, the AMP includes any sequence of AMP1 to 3 as shown in SEQ ID NOs. 5 to 7, and the molecular structure of the FAM is as follows: The TAMRA molecular structure is as follows:

2. The combination-specific electrostatic composite fluorescence array sensor according to claim 1, characterized in that: The fluorescence array sensor includes 12 sensing units, which are any one of FAM-ssDNA1 / TAMRA-AMP1, FAM-ssDNA1 / TAMRA-AMP2, FAM-ssDNA1 / TAMRA-AMP3, FAM-ssDNA2 / TAMRA-AMP1, FAM-ssDNA2 / TAMRA-AMP2, FAM-ssDNA2 / TAMRA-AMP3, FAM-ssDNA3 / TAMRA-AMP1, FAM-ssDNA3 / TAMRA-AMP2, FAM-ssDNA3 / TAMRA-AMP3 or FAM-ssDNA4 / TA MRA-AMP1, FAM-ssDNA4 / TAMRA-AMP2, FAM-ssDNA4 / TAMRA-AMP3.

3. A method for preparing the combination-specific electrostatic composite fluorescence array sensor according to claim 1, characterized in that: The method comprises the following steps: diluting FAM-ssDNA with a buffer solution, and then adding TAMRA-AMP with different concentrations to carry out a complex reaction.

4. The method for preparing the electrostatic composite fluorescence array sensor according to claim 3, characterized in that: The concentration of the FAM-ssDNA is 1 to 100 nM, and the concentration of the TAMRA-AMP is 1 to 100 nM.

5. The method for preparing the electrostatic composite fluorescence array sensor according to claim 3, characterized in that: The composite reaction is carried out at 25-37°C for 0.5-1h.

6. Use of the combined specific electrostatic composite fluorescence array sensor according to claim 1 in bacteria detection.

7. The use according to claim 6, characterized in that: The application process is: (1) Adding an unknown bacterial sample to be tested to a well plate containing a sensor unit and incubating the sample, using a multifunctional microplate reader to measure the fluorescence intensity, and obtaining the fluorescence spectrum intensity I of the unknown bacterial sample. Repeating the above process N times, different fluorescence spectrum data I are obtained, and the corresponding blank value I0 is deducted to calculate the relative fluorescence change intensity; (2) The fluorescence change intensity in step (1) is used through a machine learning algorithm to obtain a "fingerprint" of the unknown bacterial sample to be tested, which is then compared with different bacterial standard samples.

8. The use according to claim 7, characterized in that: The incubation time of the bacteria and the sensor unit in step (1) is 1 to 2 hours, and the temperature is 25 to 37°C.

9. The use according to claim 7, characterized in that: The excitation wavelength for measuring the fluorescence intensity in step (1) is 470 to 548 nm, the emission wavelength is 520 to 588 nm, and the number of repetitions is 3 to 7 times.

10. The use according to claim 6, characterized in that: The bacteria include bacteria causing sepsis infection, including any one or more combinations of Klebsiella pneumoniae, Proteus mirabilis, Enterobacter hallii, Staphylococcus aureus, Enterococcus faecium, Escherichia coli, Serratia marcescens, Klebsiella oxytoca, Pseudomonas aeruginosa, Staphylococcus lugdunensis, Enterococcus faecalis, Enterobacter aerogenes, Acinetobacter baumannii, and Salmonella.

Images