Application of deep learning energizing functionalized fluorescent nanoprobe in protein mutation analysis

Abstract

The invention discloses a functionalized targeting fluorescent nanoprobe and application thereof in protein mutation analysis, and relates to the technical field of biology, the fluorescent nanoprobe comprises a targeting molecule-PEG hydrophilic bridging molecule and a matrix; the preparation method comprises the following steps: adding nano-seeds into fluorescent molecules to prepare fluorescent nano-particles, and stirring and mixing the fluorescent nano-particles with targeting molecules-PEG hydrophilic bridging molecules; the method is combined with deep learning to be applied to protein mutation analysis. The multifunctional integrated protein mutation receptor targeted fluorescent nanoprobe provided by the invention can accurately mark the position of a mutant protein at a single cell level, and can realize in-situ, real-time and dynamic living cell imaging of a tumor cell endogenous molecular mutant protein; and qualitative and quantitative analysis of tumor cell protein mutation can be realized through subcellular localization of the targeted fluorescent nanoprobe.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of biotechnology, and particularly to a functionalized targeted fluorescent nanoprobe and its application in protein mutation analysis. BACKGROUND

[0002] In recent years, subcellular mutant proteins play a key role in driving carcinogenesis in various malignant tumors such as lung cancer, gastric stromal tumor, breast cancer, colorectal cancer and melanoma. These mutations (such as EGFR, HER2, c-KIT, BRAF, etc.) lead to abnormal activation of signaling pathways, promoting the proliferation, survival and metastasis of tumor cells. Although traditional treatments such as surgery, chemotherapy and radiotherapy still play an important role in cancer treatment, their efficacy is often limited by tumor heterogeneity and drug resistance, resulting in poor efficacy or recurrence in some patients. With the development of molecular biology and precision medicine, targeted treatment strategies for membrane receptor proteins have become an important breakthrough in cancer treatment. For example, in non-small cell lung cancer, EGFR mutation inhibitors (such as gefitinib) significantly improve patient survival; in gastric stromal tumor, c-KIT mutation inhibitors (such as imatinib) become first-line drugs; in HER2-positive breast cancer, anti-HER2 monoclonal antibodies (such as trastuzumab) have achieved significant efficacy. However, molecular targeted therapy still faces the challenges of tumor cell heterogeneity and drug resistance. The sensitivity of tumor cells to drugs varies significantly in different patients or even in the same patient, making it difficult to predict the treatment effect and leading to recurrence. Therefore, precise identification and quantitative analysis of subcellular protein mutations and exploration of new strategies to overcome drug resistance have become an important direction of current research.

[0003] Existing detection methods such as PCR, sequencing, etc. are widely used in clinical research of tumor mutations, but still have significant limitations. These methods are usually based on the average level of population cells, and cannot reveal the heterogeneity between individual cells, while tumor cell heterogeneity is an important cause of drug resistance and treatment failure. In addition, traditional techniques are difficult to achieve in situ real-time intracellular protein imaging, and cannot spatially locate and dynamically monitor receptor protein mutations at the cellular or tissue level, thereby limiting the in-depth understanding of their functional mechanisms in the tumor microenvironment. In contrast, single-cell detection techniques can accurately identify the expression differences of receptor protein mutations in different cell subpopulations, providing key information for revealing tumor heterogeneity and drug resistance mechanisms. In situ imaging techniques can dynamically monitor receptor protein mutations in real time at the cellular or subcellular level, providing direct evidence for studying their spatial distribution and functional regulation in the tumor microenvironment.

[0004] Fluorescent signal has become one of the core ways of biological probe signal expression and reading due to its unique advantages such as easy reading, fast response, high spatial and temporal resolution, and has been widely used in protein detection and tumor marker diagnosis. However, due to the lack of specific targeting ability to tumor tissues or specific cancer cell subgroups, it is difficult to achieve cancer cell subpopulation and drug resistance diagnosis through precise identification. To achieve this goal, specific targeting molecules must be used to modify and precisely guide the nano probe. Based on the excellent designability of nano fluorescent probes in particle size, surface performance and other aspects, the specific labeling of tumor cell markers and subcellular targets is realized by targeting modification, and combined with fluorescence signal image analysis technology, it provides an important tool for precise diagnosis and quantitative analysis of diseases. This kind of targeted functionalized nano fluorescent probe not only can realize the efficient identification of cancer cells, but also opens up a new way for the study of tumor heterogeneity and drug resistance mechanism.

[0005] In order to intelligently read and quantitatively analyze the fluorescence microscopic image, deep learning has shown significant advantages in the field of biological image analysis. It can extract features from complex data and optimize algorithm performance by simulating human learning behavior, and has been widely used in image intelligent recognition and automatic segmentation tasks. In the field of biological tissue, tumor metastasis and single vessel imaging, deep learning has built an efficient and accurate image segmentation model through data set training and model optimization, realizing multi-dimensional feature extraction and analysis. However, due to the difficulty of single cell labeling and organelle localization through traditional non-targeted probes, there is still a research gap in the field of single cell intelligent image processing based on deep learning. The breakthrough in this field will provide new technical support for precise diagnosis at the single cell level.

[0006] Therefore, the skilled in the art is committed to developing a functionalized targeted fluorescent nano probe, and applying it to protein mutation analysis combined with deep learning. SUMMARY

[0007] In view of the above defects of the prior art, the technical problem to be solved by the present application is how to develop a functionalized targeted fluorescent nano probe, and apply it to protein mutation analysis combined with deep learning.

[0008] To achieve the above purpose, the present application provides a multifunctional integrated protein mutation receptor targeted fluorescent nano probe, which comprises a targeting molecule-PEG hydrophilic bridging molecule and a substrate; the substrate is a fluorescent silica nanoparticle, a polystyrene-co-polyacrylic acid nanoparticle and a metal organic framework nanoparticle; the substrate is mixed in the targeting molecule-PEG hydrophilic bridging molecule.

[0009] The present application also provides a preparation method of a multifunctional integrated protein mutation receptor targeted fluorescent nano probe, comprising the following steps: Step 1, preparing nano-seeds, the nano-seeds are silica nano-seeds, polystyrene-co-polyacrylic acid nano-seeds and metal organic framework nano-seeds; Step 2, adding fluorescent molecules to the nano-seeds obtained in step 1 respectively to prepare fluorescent nanoparticles, the fluorescent nanoparticles are fluorescent silica nanoparticles, polystyrene-co-polyacrylic acid nanoparticles and metal organic framework nanoparticles; Step 3, preparing a targeting molecule-PEG hydrophilic bridging functional polymer, dispersing a mutant protein targeting molecule, a polyethylene glycol crosslinking agent and triethylamine in a solvent and stirring overnight, synthesizing the targeting molecule to the end of the polyethylene glycol through a coupling reaction, and preparing the targeting molecule-PEG hydrophilic bridging functional polymer through dialysis and freeze-drying; Step 4, adding the fluorescent nanoparticles obtained in step 2 to the targeting molecule-PEG hydrophilic bridging molecule prepared in step 3, stirring overnight, and obtaining a targeted fluorescent nanoprobe through centrifugal washing.

[0010] Further, step 1 further comprises: adjusting the particle size of the nano-seeds by adjusting the reaction time, monomer ratio and reaction temperature, so that the particle size of the nano-seeds is finally in the range of 100-500 nm.

[0011] Further, step 2 further comprises: stirring the fluorescent molecules with the nano-seeds, and combining the fluorescent molecules with the nano-seeds through chemical bond connection method, supermolecular force attraction method and physical blending method; the fluorescent molecules are one or more of fluorescein isothiocyanate, cyanine dye, rhodamine dye and coumarin (AF, Alexa Fluor) series.

[0012] Further, step 2 further comprises: Step 2.1, surface modification of the polystyrene-co-polyacrylic acid nanoparticles, surface functionalization using active biological molecules to obtain functionalized polystyrene-co-polyacrylic acid nanoparticles; Step 2.2, surface modification of the metal organic framework nanoparticles, surface functionalization using active biological molecules to obtain functionalized metal organic framework nanoparticles.

[0013] Further, the active biological molecules are: dextran, chitosan, hyaluronic acid, alginic acid, xanthan gum, agarose, pectin.

[0014] Further, the mutant protein targeting molecule is one or more of mutant protein molecules of gefitinib, erlotinib, afatinib, imatinib, sorafenib, dasatinib, bevacizumab, trastuzumab and pertuzumab.

[0015] Further, the light molecules can be selected according to the properties of the dispersion solvent, experimental requirements and application scenarios.

[0016] Further, the polyethylene glycol crosslinking agent can change the end group and molecular weight according to experimental requirements, the end group includes amino, carboxyl, hydroxyl, NHS active ester, and the molecular weight includes 3000, 5000 and 10000. The end group is connected with the target molecule through covalent chemical groups and click chemistry.

[0017] The application further provides an application of the multifunctional integrated protein mutation receptor targeted fluorescent nanoprobes in protein mutation analysis, which comprises the following steps: Step 1) culturing tumor cells using a culture medium containing fluorescent nanoprobes; Step 2) imaging the localization of single cells and target fluorescent nanoprobes of the tumor cells cultured in step 1) by a confocal microscope, collecting a large number of images and performing feature extraction and automatic segmentation on the images by a U-net neural network, then manually calibrating the preliminary model segmentation result; dividing the data set into 70% training set and 30% test set, further dividing the training set into 40% as a validation data set to evaluate the segmentation accuracy of training, and comparing the 30% test set with the manual annotation result to ensure that the accuracy reaches 80% or above, otherwise, the model is further optimized; Step 3) for the confocal images based on the interaction of single cells and target nanoprobes, the image is automatically segmented by a deep learning model by labeling the cell nucleus, cytoplasm, cell membrane and nanoprobes.

[0018] Further, step 3) further comprises taking 90% as a threshold, defining that the pixel area ratio of the localization of the target nanoprobes and the cell membrane is greater than 90% as the membrane target nanoprobes, and the rest as the extracellular / intracellular nanoprobes, and defining the extracellular / intracellular nanoprobes as negative results; performing quantitative analysis, and counting the proportion of the fluorescent nanoprobes targeted on the membrane of a single tumor cell; diagnosing mutation occurrence by the average value of the proportion of the nanoprobes on the membrane of a heterogeneous tumor cell group, and analyzing the mutation degree by the division of the proportion on the membrane of a single cell.

[0019] Further, step 3) further comprises: For highly heterogeneous clinical samples, collect pleural effusion, tissue-derived clinical samples from patients who have not received targeted drug treatment, extract tumor cells by methods including tissue lysates, lysed red, filtration and magnetic bead sorting, resuspend in culture medium, inoculate in 6-well plates, add mutant target fluorescent nanoprobes after the cells are stable, continue to culture in the incubator, centrifuge with phosphate buffer; then fix the tumor cells with paraformaldehyde, stain the cells with cell membrane dye and cell nucleus dye; use confocal microscopy to image and observe the targeting of single cells; for clinical sample imaging, analyze and quantify the localization results of single cells and targeted nanoprobes, analyze the mutation of patients by the average value of the overall membrane targeting ratio of patient-derived single cell images, and quantify the targeting ratio of single positive mutant tumor cells to analyze the mutation degree.

[0020] Further, the tumor cell mutant protein includes tumor-related endogenous active molecule protein mutation and cytoplasmic protein, including transmembrane EGFR mutant protein, C-KIT mutant protein, and cytoplasmic distribution BCR-ABL fusion protein, and the deep learning model is further used to further expand analysis on heterogeneous single cell clinical samples.

[0021] Further, the tumor cells include non-small cell lung cancer HCC827 cells (with EGFR protein mutation), gastric stromal tumor GIST-T1 cells (with C-KIT membrane mutant protein), chronic myeloid leukemia K562 cells (cytoplasmic BCR-ABL fusion protein), and control group MCF-7 breast cancer cells, A549 lung cancer cells, Hela human cervical cancer cells, gastric mucosa cells GES-1 and endothelial cell HUVEC cell lines.

[0022] Further, the tumor cells derived from clinical samples include ex vivo tissue, pleural effusion, peripheral blood and surgically resected tumor, and are selected from patients who have not received targeted drug treatment. The tumor cells have protein mutations related to tumor types and different mutation degrees.

[0023] Further, the deep learning model realizes single cell diagnosis and quantitative analysis of membrane proteins and cytoplasmic proteins by accurately segmenting subcellular structures and capturing the positioning of targeted nanoprobes. The deep learning model includes supervised U-net neural network, CNN convolutional neural network and unsupervised large model.

[0024] In a preferred embodiment 1 of the present application, the process for preparing a multifunctional integrated gefitinib modified protein mutation targeting fluorescent nanoprobes is described in detail; In another preferred embodiment 2 of the present application, the process for preparing a multifunctional integrated dasatinib modified protein mutation targeting fluorescent nanoprobes is described in detail; In another preferred embodiment 3 of the present application, the process for preparing the multifunctionally integrated imatinib-modified protein mutant targeting fluorescent nanoprobes is described in detail; In another preferred embodiment 4 of the present application, the process for preparing the multifunctionally integrated hyaluronic acid-modified FITC-labeled ZIF-8 nanoprobes targeting the marker molecule gefitinib is described in detail; In another preferred embodiment 5 of the present application, the process for preparing the multifunctionally integrated rhodamine-labeled ZIF-70 nanoprobes targeting the marker molecule imatinib is described in detail; In another preferred embodiment 6 of the present application, the process for preparing the multifunctionally integrated FITC-labeled PS-co-PAA nanoprobes targeting the marker molecule dasatinib is described in detail; In another preferred embodiment 7 of the present application, the process for preparing the multifunctionally integrated rhodamine-labeled PS-co-PAA nanoprobes targeting the marker molecule trastuzumab is described in detail; In another preferred embodiment 8 of the present application, the process for locating human lung squamous carcinoma cells and human non-small cell lung cancer EGFR mutant cells with EGFR-targeting nanoprobes is described in detail; In another preferred embodiment 9 of the present application, the process for locating human gastric stromal tumor cells and human gastric mucosa cells with targeting nanoprobes is described in detail; In another preferred embodiment 10 of the present application, the process for locating human leukemia cells and human breast cancer cells with BCR-ABL-targeting nanoprobes is described in detail; In another preferred embodiment 11 of the present application, the process for detecting and analyzing tumor cell mutant proteins by combining deep learning to establish a model is described in detail; In another preferred embodiment 12 of the present application, the process for detecting and analyzing tumor cell mutant proteins by combining deep learning to establish a pixel area ratio threshold for the localization of targeting nanoprobes and cell membranes is described in detail; In another preferred embodiment 13 of the present application, the process for detecting and analyzing tumor cell mutant proteins by combining deep learning to establish a pixel area ratio threshold for the co-localization of targeting nanoprobes and immunofluorescence-labeled proteins is described in detail; In another preferred embodiment 14 of the present application, the process for detecting and analyzing tumor cell mutant proteins by combining deep learning is described in detail; In another preferred embodiment 15 of the present application, the process for detecting and analyzing tumor cell mutant proteins by combining deep learning for highly heterogeneous clinical samples is described in detail; In another preferred embodiment 16 of the present application, the detection and analysis of tumor cell mutant proteins in combination with deep learning are described in detail for clinical samples with subcellular mutations such as highly heterogeneous leukemia.

[0025] The present application has the following beneficial technical effects: Endogenous mutant proteins are directly related to various high-incidence malignant tumors. Therefore, accurate analysis of protein mutations in single living cells and quantitative statistics have important significance for precise diagnosis and treatment of cancer. However, the two bottlenecks of dynamic endocytosis and morphological heterogeneity of living cells make related quantitative research still a blank point in the field. The multifunctional integrated protein mutation receptor targeted fluorescent nanoprobes platform designed in the present application can accurately label the position of mutant proteins at the single cell level, and realize the diagnosis of mutant proteins in single cells according to the localization characteristics of subcells. In addition, the deep learning image processing model constructed in the present application realizes efficient segmentation and intelligent reading of the scale fluorescent image of single cells, and further realizes quantitative analysis according to the quantitative distribution of nanoprobes at the mutant protein target, providing a new scheme for single cell diagnosis and treatment of cancer.

[0026] It is very important to deeply analyze the heterogeneity of tumor cells at the single cell level. Single cell detection technology can accurately identify the expression difference of receptor protein mutations in different cell subgroups, providing key information for revealing the tumor heterogeneity and drug resistance mechanism. Based on this, the receptor mutant protein targeted molecular modified fluorescent nanoprobes disclosed in the present application can realize real-time and in-situ protein mutation targeted imaging. The in-situ imaging technology can dynamically monitor the receptor protein mutation in real time at the cell or subcellular level, providing intuitive evidence for studying the spatial distribution and functional regulation of the tumor microenvironment. In addition, the nanoprobes overcome the defects of poor water solubility of small molecule probes, and increase the efficiency of entering cells through endocytosis of tumor cells, and accurately capture the mutant protein site to realize in-situ mutation diagnosis and drug resistance analysis.

[0027] The covalent modification of the mutant protein targeting molecule disclosed in the present application makes the nanoprobes have in-situ targeting specificity, and the modification of polyethylene glycol and the functionalization of fluorescent molecules increase the stability and fluorescence efficiency of the nanoprobes. Unlike the passive diffusion of small molecule probes into cells due to hydrophobicity, the nanoprobes can enter cells through endocytosis such as pinocytosis, and realize the targeting of mutant protein sites. In addition, the modification of polyethylene glycol hydrophilic polymer chains on the surface of the nanoprobes increases its stability, and it can maintain uniform particle size and dispersibility for a long time, further increasing its efficiency of being internalized by cells and the specificity of targeting mutant proteins.

[0028] The application can realize in-situ, real-time and dynamic live cell imaging of tumor cell endogenous molecular mutant proteins, realize diagnosis and quantitative analysis of tumor cell protein mutations through subcellular localization of targeted fluorescent nanoprobes; and can realize imaging analysis of mutations from the single cell dimension in the tumor cell population of heterogeneous clinical samples, thereby providing intuitive evidence for studying the spatial distribution and functional regulation of tumor microenvironment.

[0029] The application is also based on the excellent designability of the nanofluorescent probe in terms of particle size, surface performance and the like, and the targeted modification thereof enables specific labeling of tumor cell markers and subcellular targets, and the combination of deep learning assisted fluorescence signal image accurate segmentation and quantitative analysis provides an important tool for accurate diagnosis and quantitative analysis of diseases. In addition, in the application, deep learning captures the accurate positioning of the nanoprobes at the subcellular level, and assists in the diagnosis and imaging analysis of mutant proteins through targeted segmentation and quantification.

[0030] Through tumor cell markers and subcellular targets, the substrate of the traditional nanomaterial is further modified (such as targeted antibodies, targeted small molecules) to construct a targeted nanoprobe; a deep learning framework (such as unet, nnunet and MedSAM, etc.) for image segmentation is used to train the data set and optimize the model, thereby constructing an efficient and accurate image segmentation model, realizing multi-dimensional feature extraction and analysis, and then accurately capturing and quantitatively segmenting the subcellular localization of the nanoprobes in tumor cells, and diagnosing and quantifying mutations through data calculation.

[0031] The targeted functionalized nanofluorescent probe disclosed in the application can realize efficient recognition of cancer cells, and further analyze tumor cell heterogeneity and drug resistance mechanism; the new image-based artificial intelligence method can quickly and accurately locate the nanomaterial to assist in diagnosis.

[0032] The concept, specific structure and technical effects of the application will be further described in combination with the accompanying drawings, so as to fully understand the purpose, features and effects of the application. BRIEF DESCRIPTION OF DRAWINGS

[0033] Figure 1 is a schematic diagram of the targeting process of the nanoprobes-cell interaction of a preferred embodiment 8 of the application; Figure 2 is an experimental flowchart for realizing single cell analysis in combination with deep learning of a preferred embodiment 12 of the application. DETAILED DESCRIPTION

[0034] The application can be embodied in many different forms, and the protection scope of the application is not limited to the embodiments mentioned herein.

[0035] Example 1 Preparation of multifunctional integrated gefitinib modified protein mutation targeting fluorescent nanoprobe

[0036] 1. Gefitinib and NHS-PEG-silane coupling agent (Mw=5000) were added in a molar ratio of 1.2:1, and then 20 mL of dimethyl sulfoxide (DMSO) was added, and stirred at room temperature for 24 h to obtain the product gefitinib-PEG-silane coupling agent. After dialysis with a dialysis bag with a molecular weight cutoff of 1000, the gefitinib-PEG-silane coupling agent was lyophilized to obtain the dry product of gefitinib-PEG-silane coupling agent.

[0037] 2. 9.8 mg of fluorescein isothiocyanate (FITC) was added to 1.4 mL of 3- aminopropyltrimethoxysilane (APTMS), and then 20 mL of anhydrous ethanol was added, and stirred in the dark for 24 h to obtain the product FITC-APTMS. 1 mL of tetraethyl orthosilicate (TEOS), 5 mL of 28 wt% ammonia water, and 50 mL of anhydrous ethanol were added to the reaction bottle, and stirred at room temperature for 3 h. Then, FITC-APTMS was added in a mass ratio of nanoparticles to fluorescein of 500:1, and stirred at room temperature in the dark for 5 h. Then, 10 mL of 1 mg / mL gefitinib-PEG-silane coupling agent ethanol solution was added, and stirred at room temperature in the dark for 24 h. After centrifugal washing with anhydrous ethanol and deionized water for 3-4 times, the reaction was completed to obtain the gefitinib modified fluorescent nanoprobe.

[0038] Example 2 Preparation of multifunctional integrated dasatinib modified protein mutation targeting fluorescent nanoprobe

[0039] 1. Dasatinib and NHS-PEG-silane coupling agent (Mw=5000) were added in a molar ratio of 1.15:1, and then 14 mL of N,N-dimethylformamide (DMF) was added, and stirred at room temperature for 24 h to obtain the product dasatinib-PEG-silane coupling agent. After dialysis with a dialysis bag with a molecular weight cutoff of 1000, the dasatinib-PEG-silane coupling agent was lyophilized to obtain the dry product of dasatinib-PEG-silane coupling agent.

[0040] 2. 10.7 mg of fluorescein isothiocyanate (FITC) was added to 1.5 mL of 3- aminopropyltrimethoxysilane (APTMS), and 25 mL of anhydrous ethanol was added, and the mixture was stirred in the dark for 24 h to obtain the product FITC-APTMS. 4 mL of tetraethyl orthosilicate (TEOS), 15 mL of 28 wt% ammonia water, and 150 mL of anhydrous ethanol were added to a reaction bottle, and the mixture was stirred at room temperature for 3 h. Then, FITC-APTMS was added in a mass ratio of nanoparticles to fluorescein of 700:1, and the mixture was stirred in the dark at room temperature for 6 h. Then, 5 mL of a 2 mg / mL ethanol solution containing dasatinib-PEG-silane coupling agent was added, and the mixture was stirred in the dark at room temperature for 24 h. The reaction was completed, and the product was washed with anhydrous ethanol and deionized water by centrifugation for 3-4 times to obtain a dasatinib-modified fluorescent nanoprobe.

[0041] Example 3 Preparation of multifunctionally integrated imatinib-modified protein mutation targeting fluorescent nanoprobe

[0042] 1. Imatinib and COOH-PEG-silane coupling agent (Mw=5000) were added in a molar ratio of 1.12:1, and 20 mL of dimethyl sulfoxide (DMSO) was added, and the mixture was stirred at room temperature for 24 h to obtain the product imatinib-PEG-silane coupling agent. The product was dialyzed with a dialysis bag with a molecular weight cut-off of 1000, and then freeze-dried to obtain the imatinib-PEG-silane coupling agent dry product.

[0043] 2. 15 mg of rhodamine was added to 2 mL of 3-aminopropyltrimethoxysilane (APTMS), and 50 mL of anhydrous ethanol was added, and the mixture was stirred in the dark for 24 h to obtain the product Rhodamine-APTMS. 1 mL of tetraethyl orthosilicate (TEOS), 4 mL of 28 wt% ammonia water, and 50 mL of anhydrous ethanol were added to a reaction bottle, and the mixture was stirred at room temperature for 5 h. Then, Rhodamine-APTMS was added in a mass ratio of nanoparticles to fluorescein of 1000:1, and the mixture was stirred in the dark at room temperature for 4 h. Then, 5 mL of a 2 mg / mL ethanol solution containing imatinib-PEG-silane coupling agent was added, and the mixture was stirred in the dark at room temperature for 24 h. The reaction was completed, and the product was washed with anhydrous ethanol and deionized water by centrifugation for 3-4 times to obtain an imatinib-modified fluorescent nanoprobe.

[0044] Example 4 Preparation of multifunctionally integrated FITC-labeled ZIF-8 nanoprobe for hyaluronic acid-modified marker targeting molecule gefitinib

[0045] Zinc nitrate hexahydrate (Zn(NO3)2·6H2O) was dissolved in 1 mL of water, and 2.85 g of dimethylimidazole was dissolved in 10 mL of water to obtain a zinc nitrate hexahydrate solution and a dimethylimidazole solution, respectively. Fluorescein isothiocyanate (FITC) was dissolved in dimethyl sulfoxide (DMSO) to form a fluorescein isothiocyanate (FITC) dimethyl sulfoxide solution (FITC DMSO solution) with a concentration of 0.5 mg / mL. 1 mL of the zinc nitrate hexahydrate solution, 5.7 mL of the FITC DMSO solution, and 5 mL of deionized water were added to a reaction vessel, and the reaction was stirred at room temperature for 30 min. The concentration of the FITC DMSO solution was adjusted according to a mass ratio of nanoparticles to fluorescein of 20-2000:1. Then, 10 mL of the dimethylimidazole solution was added, and the reaction was stirred at room temperature for 1 h. The resulting product was collected by centrifugation at 14000 rpm for 20 min, and then washed with deionized water by centrifugation for 3 times to obtain FITC-labeled ZIF-8 nanoparticles

[0046] Hyaluronic acid was dissolved in water to form a hyaluronic acid aqueous solution with a concentration of 2 mg / mL. Gefitinib was dissolved in ethanol to form a gefitinib ethanol solution with a concentration of 2 mg / mL, and the hyaluronic acid aqueous solution and the gefitinib ethanol solution were mixed in equal volumes.

[0047] FITC-labeled ZIF-8 nanoparticles were dissolved in water to form a FITC-labeled ZIF-8 nanoparticle aqueous solution with a concentration of 2 mg / mL. The mixture of the FITC-labeled ZIF-8 nanoparticle aqueous solution and the hyaluronic acid aqueous solution and the gefitinib aqueous solution was mixed in a volume ratio of 5:1. Preferably, 5 mL of the FITC-labeled ZIF-8 nanoparticle aqueous solution was mixed with 2 mL of the hyaluronic acid aqueous solution modified with gefitinib, and the mixture was shaken at room temperature for 12 h to obtain a FITC-labeled ZIF-8 nanoprobe for hyaluronic acid-modified labeled targeting molecule gefitinib.

[0048] Example 5 Preparation of a multifunctionally integrated rhodamine fluorescein-labeled ZIF-70 nanoprobe for hyaluronic acid-modified labeled targeting molecule imatinib

[0049] Zinc nitrate hexahydrate (Zn(NO3)2·6H2O) was dissolved in 10 mL of methanol to obtain a zinc nitrate hexahydrate methanol solution; 3.25 g of 2-bromomethylimidazole was dissolved in 12 mL of methanol to obtain a 2-bromomethylimidazole methanol solution; and rhodamine fluorescein was dissolved in water to form a rhodamine fluorescein aqueous solution. First, 12 mL of the 2-bromomethylimidazole methanol solution was added to a reaction bottle, and the zinc nitrate hexahydrate methanol solution was added while stirring, followed by the addition of 1.5 mL of the rhodamine fluorescein aqueous solution, and the reaction was stirred for 60 min.

[0050] The concentration of the rhodamine fluorescein aqueous solution was adjusted according to the mass ratio of the metal organic framework nanoparticles and the fluorescein of 300:1. The obtained product was collected by centrifugation at 12000 rpm for 15 min, and then washed with deionized water by centrifugation for 3 times to obtain the rhodamine fluorescein labeled ZIF-70 nanoparticles.

[0051] Hyaluronic acid was dissolved in water to form a hyaluronic acid aqueous solution with a concentration of 2 mg / mL. Imatinib was dissolved in water to form an imatinib aqueous solution with a concentration of 2 mg / mL.

[0052] The rhodamine fluorescein labeled ZIF-70 nanoparticles were dissolved in water to form a rhodamine fluorescein labeled ZIF-70 nanoparticle aqueous solution with a concentration of 2 mg / mL. The rhodamine fluorescein labeled ZIF-70 nanoparticle aqueous solution and the hyaluronic acid aqueous solution were mixed in a volume ratio of 80:1, shaken at room temperature for 12 h, and then an equal volume of the imatinib aqueous solution was added, shaken at room temperature for 12 h, to finally obtain the rhodamine fluorescein labeled ZIF-70 nanoprobe of the hyaluronic acid modified labeled targeting molecule imatinib.

[0053] Example 6 Preparation of a multifunctional integrated FITC labeled PS-co-PAA nanoprobe of a hyaluronic acid modified labeled targeting molecule imatinib

[0054] 200 μL of acrylic acid and 45 mL of deionized water were added to a reactor, dissolved by stirring at room temperature, then 920 μL of styrene was added, and the reaction was carried out under nitrogen for 30 min, then 0.03 g of potassium persulfate (dissolved in 10 mL of deionized water) was added and the temperature was raised to 70°C, and the reaction was continued under nitrogen for 10 h. After the reaction was completed, the product was collected by centrifugation, and washed with deionized water for 3 times to obtain PS-co-PAA nanoparticles.

[0055] The obtained PS-co-PAA nanoparticles were dissolved in MES buffer with pH 5.5 to obtain a PS-co-PAA solution with a concentration of 4 mg / mL. 6.4 mg of 1-ethyl-(3-dimethylaminopropyl) carbonyl diimide hydrochloride (EDC·HCl) was added to 50 mg of the PS-co-PAA solution, stirred at room temperature for 10 min, then 18.2 mg of sulfo-NHS was added, and the reaction was continued for 3 h, and then FITC was added, and the reaction was continued in the dark for 24 h. The amount of FITC was adjusted according to the mass ratio of the nanoparticles and the FITC of 20-2000:1. After the reaction was completed, the product was collected by centrifugation, and washed with deionized water for 3 times to obtain the FITC labeled PS-co-PAA nanoparticles.

[0056] The FITC-labeled PS-co-PAA nanoparticles were immersed in 5 mL of 5 mg / mL aqueous solution of hyaluronic acid, vortex mixed at 1000 rpm for 1 h, and the hyaluronic acid-modified composite nanoparticles were collected by centrifugation. The nanoparticles were then immersed in 2 mL of 5 mg / mL bis-NH2-PEG, 2 mL of 1 mg / mL 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC-HCl), and 2 mL of 0.5 mg / mL N-hydroxysuccinimide were added. After the reaction overnight, the nanospheres were collected by centrifugation. The composite nanoparticles were then dispersed in 10 mL of an aqueous solution containing 0.5 mg / mL dasatinib, 10 mg / mL 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC-HCl), and 1 mg / mL N-hydroxysuccinimide, reacted overnight, and the nanocomposite particles were collected by centrifugation to obtain the FITC-labeled PS-co-PAA nanoprobe of hyaluronic acid-modified labeled targeting molecule dasatinib.

[0057] Example 7 Preparation of multifunctionally integrated rhodamine-labeled PS-co-PAA nanoprobe of chitosan-modified labeled targeting molecule trastuzumab

[0058] The reactor was charged with 250 μL of acrylic acid, 50 mL of deionized water, and stirred to dissolve at room temperature. Then 1050 μL of styrene was added, and the reaction was continued under nitrogen atmosphere for 30 min. Then 0.04 g of potassium persulfate (dissolved in 10 mL of deionized water) was added, and the temperature was raised to 70 °C. The reaction was continued under nitrogen atmosphere for 12 h with stirring. After the reaction was completed, the product was collected by centrifugation, and washed with deionized water for 3 times to obtain the PS-co-PAA nanoparticles.

[0059] The obtained PS-co-PAA nanoparticles were dissolved in MES buffer at pH 5.5 to obtain a PS-co-PAA solution at a concentration of 5 mg / mL. To 50 mg of the PS-co-PAA solution, 6.8 mg of 1-ethyl-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC-HCl) was added, and stirred at room temperature for 10 min. Then 19.4 mg of N-hydroxysuccinimide sulfo-NHS was added, and the reaction was continued for 3 h. Then rhodamine fluorescein was added, and the reaction was continued in the dark for 24 h. The amount of FITC was adjusted according to the mass ratio of nanoparticles to rhodamine fluorescein of 20-2000:1. After the reaction was completed, the product was collected by centrifugation, and washed with deionized water for 3 times to obtain the rhodamine fluorescein-labeled PS-co-PAA nanoparticles.

[0060] Rhodamine fluorescently labeled PS-co-PAA nanoparticles were immersed in 5 mL of 5 mg / mL aqueous solution of chitosan, vortex mixed at 1000 rpm for 3 h, and centrifugally collected chitosan-modified composite nanoparticles. The nanoparticles were then immersed in 2 mL of 5 mg / mL bis-NH2-PEG, and 4 mL of 1 mg / mL 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride) and 4 mL of 0.5 mg / mL N-hydroxysuccinimide were added. After overnight reaction, the nanospheres were centrifugally collected. The composite nanoparticles were then dispersed in 10 mL of aqueous solution containing 0.5 mg / mL trastuzumab, 10 mg / mL 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride), and 1 mg / mL N-hydroxysuccinimide, reacted overnight, centrifugally collected the nanocomposite particles, and obtained chitosan-modified rhodamine fluorescently labeled PS-co-PAA nanoprobe labeled with the targeting molecule trastuzumab.

[0061] Example 8 Localization of EGFR-targeting nanoprobe in human lung squamous carcinoma cells and human non-small cell lung cancer EGFR mutant cells

[0062] The process of nanoprobe-cell interaction targeting is shown in Figure 1 In this example, human lung squamous carcinoma cells (NCI-H520, EGFR negative) and human non-small cell lung cancer EGFR mutant cells (HCC827) were mixed and seeded in a six-well plate. When the cells reached 70-80% confluence, the old culture medium was removed, and 1.5 mL of culture medium containing 30 ug / mL of fluorescently labeled EGFR mutant targeting nanoprobe was added to each well. The cells were further incubated at 37°C for 6 h. The cells were collected by digestion and centrifugally washed with phosphate buffer at 500 g for 5 min for three times. The tumor cells were then fixed with 4% paraformaldehyde and stained with cell membrane dye and cell nucleus dye. Imaging was performed using a confocal microscope. The localization of EGFR-targeting nanoprobe in human lung squamous carcinoma cells (NCI-H520) and human non-small cell lung cancer EGFR mutant cells (HCC827) was compared. The EGFR fluorescently targeted nanoprobe was enriched in the interior of human lung squamous carcinoma cells (NCI-H520), while the fluorescently EGFR-targeting nanoprobe was aggregated on the cell membrane of non-small cell lung cancer EGFR mutant cells (HCC827).

[0063] Example 9 Localization of targeting nanoprobe in human gastric stromal tumor cells and human gastric mucosa cells

[0064] Human gastric stromal tumor cells (GIST-T1, C-KIT mutant) and human gastric mucosa cells (GES-1, C-KIT negative) were mixed and seeded in a six-well plate. When the cell adhesion confluence reached 70-80%, the old culture medium was aspirated, 1.5 mL of culture medium containing 30 ug / mL of fluorescently labeled C-KIT mutant targeting nanoprobes was added to each well, and the cells were incubated at 37°C for 6 hours. The cells were collected and washed three times with phosphate buffer at 500g for 5 minutes. Then the tumor cells were fixed with 4% paraformaldehyde, and the cells were stained with cell membrane dye and nuclear dye. Imaging was performed using a confocal microscope, and the localization of the C-KIT targeting nanoprobes in human gastric stromal tumor cells (GIST-T1, C-KIT mutant) and human gastric mucosa cells (GES-1, C-KIT negative) was compared. The fluorescent C-KIT targeting nanoprobes of human gastric stromal tumor cells (GIST-T1) were concentrated on the cell membrane, while the fluorescent targeting nanoprobes of human gastric mucosa cells (GES-1) were enriched in the cell interior.

[0065] Example 10 Localization of BCR-ABL targeting nanoprobes in human leukemia cells and human breast cancer cells

[0066] Human leukemia cells (K562, BCR-ABL positive) and human breast cancer cells (MCF-7, BCR-ABL negative) were mixed and seeded in a six-well plate. When the cell adhesion confluence reached 70-80%, the old culture medium was aspirated, 1.5 mL of culture medium containing 30 ug / mL of fluorescently labeled BCR-ABL mutant targeting nanoprobes was added to each well, and the cells were incubated at 37°C for 6 hours. The cells were collected and washed three times with phosphate buffer at 500g for 5 minutes. Then the tumor cells were fixed with 4% paraformaldehyde, and then the cells were permeabilized and incubated with BCR-ABL antibody overnight. The next day, BCR-ABL immunofluorescence staining was performed with secondary antibody, and finally the cells were stained with nuclear dye. Imaging was performed using a confocal microscope, and the localization of the BCR-ABL targeting nanoprobes in human leukemia cells (K562, BCR-ABL positive) and human breast cancer cells (MCF-7, BCR-ABL negative) was compared. The fluorescent BCR-ABL mutant targeting nanoprobes of human leukemia cells (K562, BCR-ABL positive) co-localized with BCR-ABL protein at a higher level, while the fluorescent BCR-ABL targeting nanoprobes of human breast cancer cells (MCF-7, BCR-ABL negative) were randomly distributed in the cytoplasm.

[0067] Example 11 Detection and analysis of tumor cell mutant proteins by establishing a model combined with deep learning

[0068] The localization of single cells and targeted fluorescent nanoprobes was imaged by confocal microscopy, a large number of images were collected and the images were feature extracted and automatically segmented by U-net neural network, and then the preliminary model segmentation results were manually calibrated. The data set was divided into 70% training set and 30% test set, and the training set was further divided into 40% as validation data set to evaluate the segmentation accuracy of the training; 30% of the test set was compared with the manual annotation results to ensure that the accuracy reached 80% or more, otherwise, the model was further optimized.

[0069] Example 12. Establishing a pixel area ratio threshold of targeted nanoprobes and cell membrane localization for detecting and analyzing tumor cell mutant proteins in combination with deep learning

[0070] For confocal images based on the interaction of single cells and targeted nanoprobes, the cell nucleus, cytoplasm, cell membrane, and nanoprobes were labeled, and the deep learning model was used for automatic segmentation of the images. With 90% as the threshold, the pixel area ratio of targeted nanoprobes and cell membrane localization greater than 90% was defined as membrane targeted nanoprobes, and the rest was defined as extracellular / intracellular nanoprobes (as negative results), and further quantitative analysis was performed to count the proportion of single tumor cell membrane targeted fluorescent nanoprobes. The average value of the proportion of nanoprobes on the membranes of heterogeneous tumor cell populations was used to diagnose mutations, and the division of the proportion of single cell membrane was used to analyze the degree of mutation. The experimental process of single cell analysis in combination with deep learning is shown in Figure 2

[0071] Example 13. Establishing a pixel area ratio threshold of targeted nanoprobes and immunofluorescence labeled proteins co-localization for detecting and analyzing tumor cell mutant proteins in combination with deep learning

[0072] For confocal images based on the interaction of single cells and targeted nanoprobes, the cell nucleus, subcellular immunofluorescence labeled protein and nanoprobes were labeled, and the deep learning model was used for automatic segmentation of the images. With 90% as the threshold, the pixel area ratio of targeted nanoprobes and immunofluorescence labeled protein co-localization greater than 90% was defined as protein localization targeted nanoprobes, and the rest was defined as nanoprobes outside protein localization (as negative results), and further quantitative analysis was performed to count the proportion of single tumor mutant protein binding fluorescent nanoprobes. The average value of the proportion of nanoprobes bound to the mutant protein of heterogeneous tumor cells was used to diagnose mutations, and the division of the proportion of single cell mutant protein binding was used to analyze the degree of mutation.

[0073] Example 14. Results of detecting and analyzing tumor cell mutant proteins in combination with deep learning

[0074] ​According to the method of Example 8 and Example 9, for imaging of commercial cell lines, the results of the localization of the cell lines and the targeted nanoprobes are analyzed and quantified by the method of Example 11 and 12, and the proportion of nanoprobes localized on the membrane of tumor cells with membrane mutation target points is significantly higher than that of negative cell lines, and there is a high difference in the proportion of single positive mutant tumor cells.

[0075] According to the method of Example 10, for imaging of commercial cell lines of subcellular mutation sites, the results of the localization of the cell lines and the targeted nanoprobes are analyzed and quantified by the method of Example 13, and the proportion of nanoprobes localized on the immunofluorescence labeled mutant protein binding sites of tumor cells with subcellular mutant protein target points is significantly higher than that of negative cell lines, and there is a high difference in the proportion of single positive mutant tumor cells.

[0076] Example 15 Detection and analysis of tumor cell mutant proteins in highly heterogeneous clinical samples combined with deep learning

[0077] For highly heterogeneous clinical samples, collect clinical samples from patients with chest fluid, tissue and other sources who have not received targeted drug treatment, extract tumor cells by methods including tissue lysates, split red, filtration and magnetic bead sorting, and then resuspend them in culture medium, inoculate them in 6-well plates, and after the cells are stable, add 1.5 mL of mutant targeted fluorescent nanoprobes containing 50 ug / mL, continue to culture in an incubator for 6 h, centrifuge at 500g, 5 min for 3 times of phosphate buffer washing; then use 4% paraformaldehyde to fix the tumor cells, and use cell membrane dye and cell nucleus dye to stain the cells. Use confocal microscope to image and observe the targeting of nanoprobes in single cells. For clinical sample imaging, the localization results of single cells and targeted nanoprobes in clinical samples are analyzed and quantified by the method of Example 11 and 12, the mutation of the patient is analyzed by the average value of the overall membrane targeting proportion of the single cell image of the patient, and the targeting proportion of single positive mutant tumor cells is quantified to explore the mutation degree.

[0078] Example 16 Detection and analysis of tumor cell mutant proteins in highly heterogeneous leukemia and other clinical samples with subcellular mutations combined with deep learning

[0079] For clinical samples with subcellular mutations, such as highly heterogeneous leukemia, pleural effusion and tissue samples were collected from patients who had not received targeted therapy. Tumor cells were extracted using methods including tissue lyase, erythromycin extraction, filtration, and magnetic bead sorting, and then resuspended in culture medium and seeded in 6-well plates. After the cells stabilized, 1.5 mL of fluorescent nanoprobes containing 50 μg / mL of mutation-targeting nanoprobes were added, and the cells were cultured for 6 hours. The cells were then washed three times by centrifugation at 500 g for 5 min with phosphate-buffered saline. The tumor cells were then fixed with 4% paraformaldehyde, permeabilized, and incubated overnight with BCR-ABL antibody. The next day, BCR-ABL immunofluorescence staining was performed with secondary antibody, and finally, the cells were stained with nuclear dyes. Confocal microscopy was used to observe the targeting of individual cells with the nanoprobes. For clinical sample imaging, the localization results of single cells and targeted nanoprobes in clinical samples were analyzed and quantified using the method of Example 13. The patient's mutation status was analyzed by the average proportion of nanoprobes targeting mutant proteins bound to the overall immunofluorescence labeling of single cell images from the patient, and the degree of mutation was explored by quantifying the targeting proportion of a single positive mutant tumor cell.

[0080] The preferred embodiments of the present invention have been described in detail above. It should be understood that those skilled in the art can make numerous modifications and variations based on the concept of the present invention without creative effort. Therefore, all technical solutions that can be obtained by those skilled in the art based on the concept of the present invention through logical analysis, reasoning, or limited experimentation on the basis of existing technology should be within the scope of protection defined by the claims.

Claims

1. A multifunctional integrated fluorescent nanoprobe targeting protein mutant receptors, characterized in that, The fluorescent nanoprobe comprises a targeting molecule-PEG hydrophilic bridging molecule and a matrix; the matrix is ​​fluorescent silica nanoparticles, polystyrene-co-polyacrylic acid nanoparticles, and metal-organic framework nanoparticles; the matrix is ​​mixed in the targeting molecule-PEG hydrophilic bridging molecule.

2. A method for preparing a fluorescent nanoprobe as described in claim 1, characterized in that, The method includes the following steps: Step 1: Prepare nanoseeds, wherein the nanoseeds are silica nanoseeds, polystyrene-co-polyacrylic acid nanoseeds, and metal-organic framework nanoseeds; Step 2: Fluorescent nanoparticles are prepared by adding fluorescent molecules to the nanoseeds obtained in Step 1. The fluorescent nanoparticles are fluorescent silica nanoparticles, polystyrene-co-polyacrylic acid nanoparticles, and metal-organic framework nanoparticles. Step 3: Prepare the target molecule-PEG hydrophilic bridging functional polymer. Disperse the mutant protein target molecule, polyethylene glycol crosslinking agent and triethylamine in a solvent and stir overnight. Synthesize the target molecule to the polyethylene glycol end through a coupling reaction. The target molecule-PEG hydrophilic bridging functional polymer is obtained by dialysis and freeze drying. Step 4: The fluorescent nanoparticles obtained in Step 2 are added to the targeting molecule-PEG hydrophilic bridging molecule prepared in Step 3, stirred overnight, and then centrifuged and washed to obtain the targeted fluorescent nanoprobe.

3. The method as described in claim 2, characterized in that, Step 1 further includes: adjusting the particle size of the nanoseeds by adjusting the reaction time, monomer ratio and reaction temperature, so that the particle size of the nanoseeds is in the range of 100-500 nm.

4. The method as described in claim 2, characterized in that, Step 2 also includes: The fluorescent molecules are stirred with the nanoseeds and then combined with the nanoseeds through chemical bonding, supramolecular attraction, and physical blending. The fluorescent molecules are one or more of the following: fluorescein isothiocyanate, cyanine dye, rhodamine dye, and coumarin series.

5. The method as described in claim 4, characterized in that, Step 2 also includes: Step 2.1: The polystyrene-co-polyacrylic acid nanoparticles are surface modified by using active biomolecules to perform surface functionalization, thereby obtaining functionalized polystyrene-co-polyacrylic acid nanoparticles. Step 2.2: The metal-organic framework nanoparticles are surface modified by using active biomolecules for surface functionalization to obtain functionalized metal-organic framework nanoparticles.

6. The method as described in claim 5, characterized in that, The active biomolecules are: dextran, chitosan, hyaluronic acid, alginate, xanthan gum, agarose, and pectin.

7. The method as described in claim 2, characterized in that, The mutant protein targeting molecule is one or more mutant protein molecules selected from gefitinib, erlotinib, afatinib, imatinib, sorafenib, dasatinib, bevacizumab, trastuzumab, and pertuzumab.

8. An application of the fluorescent nanoprobe as described in claim 1 in protein mutation analysis, characterized in that, The application includes the following steps: Step 1) Culture tumor cells in a culture medium containing the fluorescent nanoprobes; Step 2) Imaging the localization of single tumor cells and targeted fluorescent nanoprobes after culture in Step 1) using confocal microscopy, collecting a large number of images, and extracting features and automatically segmenting the images using the U-net neural network. Then, manually calibrating the preliminary model segmentation results; dividing the dataset into a 70% training set and a 30% test set, with the training set further divided into a 40% validation set for evaluating the training segmentation accuracy; comparing the 30% test set with the manually labeled results to ensure an accuracy of 80% or higher, otherwise, further optimizing the model; Step 3) For confocal images based on the interaction between single cells and targeted nanoprobes, the images are automatically segmented using a deep learning model by labeling the cell nucleus, cytoplasm, cell membrane, and nanoprobes.

9. The application as described in claim 8, characterized in that, Step 3) further includes using 90% as a threshold to define that if the pixel area ratio of the targeted nanoprobe to the cell membrane is greater than 90%, it is considered as a membrane-targeted nanoprobe, and the rest are considered as extracellular / intracellular nanoprobes. The extracellular / intracellular nanoprobes are considered as negative results. Quantitative analysis is performed to count the proportion of fluorescent nanoprobes targeting the membrane of a single tumor cell. Mutations are diagnosed by the average proportion of nanoprobes on the membranes of heterogeneous tumor cell populations, and the degree of mutation is analyzed by the division of the proportions on individual cell membranes.

10. The application as described in claim 8, characterized in that, Step 3) further includes: For highly heterogeneous clinical samples, pleural effusion and tissue-derived clinical samples were collected from patients who had not received targeted therapy. Tumor cells were extracted using methods including tissue lyase, erythromycin, filtration, and magnetic bead sorting, and then resuspended in culture medium and seeded in 6-well plates. After the cells stabilized, fluorescent nanoprobes containing the mutation-targeting nanoprobes were added, and the cells were cultured in an incubator and washed by centrifugation with phosphate buffer. The tumor cells were then fixed with paraformaldehyde and stained with cell membrane dyes and nuclear dyes. Confocal microscopy was used to image and observe the targeting of nanoprobes on individual cells. For clinical sample imaging, the localization results of single cells and targeted nanoprobes were analyzed and quantified. The patient's mutation status was analyzed by the average proportion of the overall membrane targeting in single-cell images from patients, and the degree of mutation was analyzed by quantifying the targeting proportion of individual positive mutation tumor cells.

Images