Apparatus and methods for high-throughput continuous sampling and gene detection of single live cells

Abstract

This invention relates to the field of biomedical engineering, specifically to a device and method for high-throughput continuous sampling and gene detection of single live cells. The device includes a cell capture layer and a gene detection layer stacked sequentially from top to bottom. The cell capture layer includes a fluid inlet and a fluid outlet, and at least one capture array disposed between the fluid inlet and the fluid outlet. The at least one capture array is connected to the fluid inlet and the fluid outlet respectively via a microfluidic network. Each row of the capture array includes multiple capture units spaced apart and connected along the fluid flow direction. Each capture unit includes a main channel extending along the fluid flow direction and branch channels, the branch channels sharing the same fluid inlet and the same fluid outlet with the main channel, thereby forming a θ-shaped capture unit. The gene detection layer has multiple collection chambers spaced apart, each containing an electrode. A nanomembrane is disposed at the top of each collection chamber, and a gene detection probe is disposed at the bottom.

Description

Technical Field

[0001] This invention relates to the field of biomedical engineering, specifically to a device and method for high-throughput continuous sampling and gene detection of single live cells. Background Technology

[0002] This method enables real-time, dynamic monitoring of gene expression changes in live cells under external stimuli (such as immunotherapy and drug treatment) at single-cell resolution. Applications include studying the dynamic immune escape mechanisms of tumor cells under immunotherapy stress. For example, clinicians use immune checkpoint blockade therapies such as PD-1 / PD-L1 inhibitors to treat lung cancer patients. However, a significant proportion of patients develop drug resistance after the initial response, leading to tumor recurrence. The fundamental reason lies in the high heterogeneity of the tumor cell population; some cells can survive treatment by dynamically adjusting their gene expression profiles (such as upregulating other immunosuppressive molecules besides PD-L1, such as CEA and IDO1), activating alternative immune escape pathways. Understanding this dynamic evolution is crucial for predicting efficacy, overcoming drug resistance, and developing combination therapy strategies.

[0003] Given this need, the field mainly relies on the following two technical approaches.

[0004] The first technical approach, single-cell RNA sequencing, involves isolating single cells from tissue or cell suspensions, lysing the cells, extracting mRNA, and obtaining gene expression data from tens of thousands of single cells through reverse transcription, amplification, library construction, and high-throughput sequencing.

[0005] The second technical approach is based on nanotube-based live cell sampling technology, which uses a nanoscale hollow needle to puncture the cell membrane to extract cytoplasmic contents for biochemical analysis.

[0006] However, the above-mentioned existing technologies have the following drawbacks: (1) Endpoint analysis, unable to monitor dynamically: This method requires lysing and killing cells, and can only provide a "static snapshot" at a certain point in time. It cannot continuously observe the same batch of live cells, and therefore cannot capture the dynamic changes in gene expression. Its value for studying the time-varying process of immunotherapy response is limited.

[0007] (2) Unable to associate cell phenotype: Because the cells are lysed, the gene expression data obtained by sequencing cannot be associated with the subsequent behavior of the cells (such as whether they are killed by T cells, whether their morphology changes, etc.).

[0008] (3) Extremely low throughput and lack of statistical significance: This technique can only operate on one cell at a time. Obtaining statistically significant population data requires a lot of time and manpower, which is inefficient and difficult to apply to studies that require the analysis of thousands of cells to capture rare subpopulations.

[0009] (4) Complex technical operation: The technique of accurately locating and puncturing a single cell without causing damage is highly demanding and difficult to standardize and automate. Summary of the Invention

[0010] The purpose of this invention is to provide a device and method for high-throughput continuous sampling and gene detection of single live cells, which partially solves or alleviates the above-mentioned deficiencies in the prior art, and can realize dynamic monitoring of large-scale single-cell populations, multiple genes, and long-term scales while maintaining cell viability.

[0011] To solve the aforementioned technical problems, the present invention specifically adopts the following technical solution: A first aspect of the present invention provides a device for high-throughput continuous sampling and gene detection of single live cells. The device includes a cell capture layer and a gene detection layer stacked sequentially from top to bottom, wherein... The cell trapping layer includes a fluid inlet and a fluid outlet, and at least one trapping array disposed between the fluid inlet and the fluid outlet, wherein the at least one trapping array is connected to the fluid inlet and the fluid outlet respectively through a microchannel network; Each row of the capture array includes multiple capture units spaced apart and connected along the fluid flow direction; each capture unit includes a main channel extending along the fluid flow direction and branch channels located on both sides of the main channel, wherein the branch channels share the same fluid inlet and the same fluid outlet with the main channel, thereby forming a θ-shaped capture unit; wherein the size of the capture area in the main channel is smaller than the size of a single cell, and the size of the branch channels is greater than or equal to the size of a single cell; The gene detection layer is provided with multiple collection chambers with open tops for collecting samples. Electrodes are provided in the collection chambers, and a nanopore array is provided on the top of the collection chambers. Each nanopore array corresponds to a capture array. Furthermore, when there are multiple capture arrays, the size of the main channel of the capture unit in the multiple capture arrays gradually decreases along the fluid flow direction; and / or, The microchannel network includes: a first multi-level branched microchannel disposed between the fluid inlet and the at least one trapping array (9), and a second multi-level branched microchannel disposed between the fluid outlet and the at least one trapping array; Both the first multi-level branched microchannel and the second multi-level branched microchannel include at least two branch networks that are sequentially connected along the fluid flow direction. In the first multi-level branched microchannel, the number of branches in each branch network gradually increases along the fluid flow direction, so that each row in the capture array corresponds to a branched microchannel; In the second multi-level branched microchannel, the number of branches in each branch network gradually decreases along the fluid flow direction, so that all branched microchannels converge at the fluid outlet.

[0012] Further devices include: a filtering module disposed between two adjacent capture arrays; and / or, At least one filter module is disposed in the microchannel network between the fluid inlet and the at least one capture array.

[0013] Furthermore, the filtering module includes: A group of multiple rows of baffles spaced apart along the fluid flow direction, the group of baffles comprising: a plurality of first baffles extending perpendicular to the fluid flow direction, and a plurality of second baffles extending parallel to the fluid flow direction; The first and second baffles in each column are arranged alternately, thereby forming a transverse microchannel extending along the fluid flow direction between two adjacent first and second baffles; and The first and second blocks in each row extending along the fluid flow direction are arranged alternately, so that a longitudinal microchannel extending perpendicular to the fluid flow direction is formed between two adjacent first and second blocks.

[0014] Furthermore, the spacing between two adjacent capture units in the same row of the capture array is 200 μm; and / or, The spacing between adjacent rows of capture units in the plurality of capture arrays along the fluid flow direction gradually decreases and then gradually increases.

[0015] Furthermore, the device also includes: a plurality of microbeads disposed at the bottom of the collection chamber, wherein probe markers are fixed on the microbeads, and each probe marker has a plurality of fluorescent groups and a plurality of fluorescent quenchers.

[0016] Furthermore, the multiple microbeads in the collection chamber have at least two different sizes.

[0017] A second aspect of the present invention provides a method for high-throughput continuous sampling and gene detection of single live cells. This method utilizes the aforementioned apparatus for high-throughput continuous sampling and gene detection of single live cells, and specifically includes the following steps: S101, Single-cell loading and capture: The cell suspension to be tested is injected into the device through the fluid inlet. Under the action of the fluid, the cells are evenly distributed into the capture array through the microchannel network, so that each capture unit in the capture array captures one cell. S102, Cell sampling and biomolecule enrichment: A pulsed electric field is applied between the electrodes of the capture unit and the collection chamber to form nanochannels on the cell membrane, thereby allowing negatively charged RNA in the cell to be directionally enriched into the collection chamber through the nanochannels and surround the probe-labeled microbeads pre-fixed in the collection chamber. S103, In situ hybridization and signal amplification: Biomolecules enriched around the microbeads hybridize with the probe, thereby triggering a chain displacement reaction, which separates the fluorescent group and quenching group in the probe and generates a fluorescent signal; subsequently, the free H2 probe in the solution replaces the RNA through a chain displacement reaction, allowing the RNA to participate in a new round of reaction, thereby achieving signal amplification; S104, Multi-channel fluorescence imaging and intelligent recognition analysis: Multi-channel fluorescence imaging of microbeads on the detection layer is performed using a fluorescence microscope to obtain fluorescence images; and an intelligent recognition algorithm based on computer vision is used to process the images to obtain the type of RNA branch and the fluorescence intensity quantification value characterizing RNA concentration.

[0018] Furthermore, the method also includes the following steps: S105, Dynamic monitoring and continuous sampling: The cells are left to be cultured on the cell capture layer for a specified time. After the specified time is reached, the liquid in the collection chamber is replaced, and steps S102 to S104 are repeated to track the dynamic changes in gene expression of the same batch of single cells at different time points.

[0019] Furthermore, step S104, which involves processing the image using a computer vision-based intelligent recognition algorithm, specifically includes the following steps: S1041, The fluorescence image is scanned based on preset circular templates of various sizes to identify candidate regions corresponding to all microbeads in the fluorescence image; S1042, Remove the overlapping or occluded regions corresponding to the microbeads in the fluorescence image to obtain the effective region corresponding to each microbead; S1043, classify each target RNA signal according to the red / green fluorescence intensity ratio of each effective region and its corresponding template size, and calculate its corresponding fluorescence intensity quantification value. The fluorescence intensity quantification value is proportional to the concentration of the target RNA and is used to characterize the concentration of the target RNA.

[0020] Beneficial effects: While traditional single-cell sequencing technology can provide high-throughput gene expression data, its principle of cell lysis makes it impossible to dynamically track the same living cells, and it can only obtain static "snapshots". On the other hand, live cell analysis technologies such as nanotube sampling are limited by extremely low throughput, making it difficult to conduct statistically significant population analyses.

[0021] The device provided by this invention, through its unique layered structure and functional integration, realizes the integration of the entire process from single-cell capture and live sampling to in-situ detection, providing core device support for single-cell dynamic gene expression analysis.

[0022] This invention integrates several key capabilities—high throughput (parallel processing of over 4000 cells), live cell (cell viability > 98%), dynamic monitoring (long-term repeated sampling of the same batch of cells), and multiplex detection (simultaneous analysis of 6 RNA targets)—by combining a theta-shaped single-cell capture array, nanochannel electric field focusing sampling, dual-coded microbead in-situ detection, and intelligent recognition algorithms.

[0023] This invention not only overcomes the core contradiction in existing technologies—the difficulty of balancing "dynamics" and "throughput," and "activity" and "efficiency"—but also reveals the heterogeneity of cell populations through its high-throughput characteristics and captures the temporal evolution of gene expression in tumor cells under immunotherapy stress through dynamic monitoring capabilities. Its integrated in-situ detection design greatly simplifies the process, avoids sample loss, and enables detection to be completed within 30 minutes, achieving single-cell sensitivity. Therefore, the device provided by this invention serves as a powerful platform tool, offering unprecedented possibilities for studying dynamic biological processes (such as tumor immune escape mechanisms) at single-cell resolution, demonstrating enormous application potential in both basic research and clinical precision medicine. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. In all the drawings, similar elements or parts are generally identified by similar reference numerals. The elements or parts in the drawings are not necessarily drawn to scale. Obviously, the drawings described below are some embodiments of the present invention, and those skilled in the art can obtain other drawings based on these drawings without any creative effort.

[0025] Figure 1A This is a schematic diagram of the structure of the device in this application; Figure 1B This is a side view of the device in this application; Figure 2 This is a detailed schematic diagram of the device in this application; Figure 3A This is a schematic diagram illustrating the structure of the cell capture layer; Figure 3B This is a schematic diagram illustrating the structure of the capture unit; Figure 4 A schematic diagram illustrating the principle of chain displacement reaction; Figure 5 This is a schematic diagram of two overlapping fluorescent regions; Figure 6 The simulation results of the flow velocity distribution of the capture unit obtained from the simulation experiment are as follows: a) the capture rate (solid line) and single cell capture rate (dashed line) of the device of this application under different input cell densities; b) an image of a single cell captured in the capture array (small white circles in the figure represent single cells; scale bar: 50 micrometers); c) a simulation diagram of the flow velocity distribution before and after the capture unit captures cells, where: the dashed line represents the flow velocity distribution simulation when no cells are captured; the solid line represents the flow velocity distribution simulation diagram after cell capture; Figure 7 To obtain the simulation results of the flow velocity distribution in the cell capture layer, the following diagrams are provided: a) shows the simulated flow velocity distribution of the entire cell capture layer; b and c) show the simulated flow velocity distribution of three capture units connected at the same position before and after the capture cell, respectively; d and e) show the vertical flow velocity distribution curves of the capture units shown in b and c at the same position before and after the capture cell, and their adjacent units, indicating that the capture cell does not affect the adjacent structures (the blue curve is the flow velocity curve before capture; the red curve is the flow velocity curve after capture); f) shows the simulated flow velocity distribution of capture units of different sizes in the cell capture layer in a; g) shows the flow velocity distribution curves of the center lines of capture units of different sizes in f (the green, purple, and orange curves are the flow velocity distribution curves of large, medium, and small sizes, respectively), indicating that size does not affect the flow velocity within the channel. Figure 8 To obtain the simulation results of the electric field distribution based on nanochannels in the simulation experiment: a) is the simulation result of the electric field distribution during the sampling process under the electric field manipulation based on nanochannels; b) is the transmembrane potential difference of the cell membrane near the nanochannel; Figure 9 To simulate the electric field distribution without nanochannels, the following results were obtained: a) shows the electric field distribution during sampling under electric field manipulation in a conventional device (i.e., a device without nanopores); b) shows the transmembrane potential difference of the cell membrane at the location indicated by the thick line in a. Figure 10 This is a graph showing the distribution of cell viability (i.e. the proportion of surviving live cells) after safe live cell sampling under different electrical parameters (voltage amplitude, number of pulses, pulse duration). Figure 11ATo reflect the changes in cell viability (i.e. the proportion of surviving cells) after data collection at different intervals under the same electrical parameters (9V, 100 pulses, 4 ms per pulse) (ns indicates no significant difference; *** indicates p<0.001; ** indicates p<0.01); Figure 11B This is a comparison chart of cell viability (i.e. the proportion of surviving cells) obtained from an experiment with five consecutive samplings every 12 hours. Figure 12 To reflect: a) the electric field simulation showing the concentrated electric field lines around the microsphere; b) the directional enrichment of charged molecules under no electric field and with an electric field, respectively; scale bar: 100 micrometers; c) fluorescence intensity diagrams showing the enrichment effect achieved by the device of this application, fluorescence intensity diagrams corresponding to the group without electric field, and fluorescence intensity diagrams corresponding to the group with electric field but no target. Figure 13 To reflect the fluorescence images obtained by scanning after using the device of this application to detect six target RNAs at different concentrations in a capture experiment (scale bar: 10 micrometers); Figure 14 A schematic diagram illustrating the process of analyzing fluorescence images obtained after a capture experiment using the device of this application to detect a mixture of six target RNAs (scale bar in i: 5 micrometers; scale bar in vi: 1 micrometer) using intelligent recognition algorithms based on computer vision. Figure 15 shows the linear relationship between target molecule concentration and fluorescence intensity obtained from data statistics of multiplex detection experiments using the device of this application for each of the six probes targeting RNA molecules: a) is a statistical graph showing the linear detection range and detection limit of target molecule concentration and fluorescence intensity when using probe GAPDH to detect target molecules, based on experimental data statistics; b) is a statistical graph showing the linear detection range and detection limit of target molecule concentration and fluorescence intensity when using probe S100A8 to detect target molecules, based on experimental data statistics; c) is a statistical graph showing the linear detection range and detection limit of target molecule concentration and fluorescence intensity when using probe PD-L1 to detect target molecules, based on experimental data statistics. The data are presented in the following graphs: d, which shows the linear detection range and detection limit of the target molecule concentration versus fluorescence intensity when using probe MAGE-A3 to detect the target molecule, based on experimental data; e, which shows the linear detection range and detection limit of the target molecule concentration versus fluorescence intensity when using probe CEA to detect the target molecule, based on experimental data; and f, which shows the linear detection range and detection limit of the target molecule concentration versus fluorescence intensity when using probe IDO1 to detect the target molecule, based on experimental data. *** indicates p < 0.001; ns indicates no significant difference; the error bars in a to f represent the standard deviations of three independent experiments. Figure 16 To reflect the Figure 13 The relationship between RNA concentration and fluorescence intensity (FI - fluorescence intensity; C - concentration) was obtained by classifying and quantifying the fluorescence images generated by six target RNA substances: a) Relationship between different concentrations of GAPDH and fluorescence intensity and Halo value; b) Relationship between different concentrations of S100A8 and fluorescence intensity and Halo value; c) Relationship between different concentrations of PD-L1 and fluorescence intensity and Halo value; d) Relationship between different concentrations of MAGE-A3 and fluorescence intensity and Halo value; e) Relationship between different concentrations of CEA and fluorescence intensity and Halo value; f) Relationship between different concentrations of IDO1 and fluorescence intensity and Halo value; *** indicates p < 0.001; ns indicates no significant difference; the error bars in a to f are the standard deviations of three independent experiments.

[0026] Summary of attached labeling and identification: 1. Base layer; 2. Detection layer; 3. Nanomembrane; 4. Cell capture layer; 5. Fluid inlet; 6. Filtration module; 7. Fluid outlet; 8. Electrode; 9. Capture array; 10. Microfluidic network; 11. Collection chamber; 12. Gene detection layer; 13. Microbeads; 101. First multi-level branched microchannel; 102. Second multi-level branched microchannel; 901. Main channel; 9010. Diversion region; 9011. Capture region; 9012a. First flow region; 9012b. Second flow region; 9013. Convergence region; 9014. Connecting region; 902. Branch channel; 601. First stop; 602. Second stop. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0028] In this document, suffixes such as "module," "part," or "unit" used to denote elements are used only for the purpose of illustrative purposes and have no specific meaning in themselves. Therefore, "module," "part," or "unit" may be used interchangeably.

[0029] In this document, the terms "upper," "lower," "inner," "outer," "front," "rear," "one end," and "the other end," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the present invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the present invention. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0030] In this document, unless otherwise explicitly specified and limited, the terms "installed," "equipped with," "connected," etc., should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection, a direct connection, or an indirect connection through an intermediate medium; it can be a connection within two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0031] In this document, "and / or" includes any and all combinations of one or more of the listed related items.

[0032] In this article, "multiple" means two or more, that is, it includes two, three, four, five, etc.

[0033] Example 1: like Figure 1A , Figure 1B and Figure 2 As shown, this embodiment provides a device for high-throughput continuous sampling and gene detection of single live cells, comprising: a cell capture layer 4 and a gene detection layer 12 stacked sequentially from top to bottom, wherein... The gene detection layer 12 is provided with a plurality of collection chambers 11 for collecting samples and open at the top. Electrodes 8 are provided in the collection chambers 11, and nanopore arrays are provided at the top openings of the collection chambers 11.

[0034] Preferably, in this embodiment, the gene detection layer 12 is provided with three collection chambers 11 spaced apart. Specifically, the gene detection layer 12 includes a base layer 1 and a detection layer 2 overlapping the base layer 1. The detection layer 2 is provided with three through holes spaced apart. A gold-plated substrate is provided on the base layer 1 corresponding to the position of each through hole. That is, the base layer 1 and the through hole walls on the detection layer 2 enclose and form three collection chambers 11. The gold-plated substrate on the base layer 1 serves as the electrode 8 in the collection chamber 11.

[0035] In some embodiments, the cell trapping layer 4 is used to trap and immobilize single cells, the nanochannel membrane (i.e., nanomembrane 3) is used to achieve efficient electroporation, the detection layer is used to collect samples, and the nanomembrane 3 is used for optical detection. An array of nanopores is disposed on the nanomembrane 3.

[0036] In some embodiments, a nanofilm 3 is provided with multiple nanopore arrays, preferably three nanopore arrays, and each nanopore array corresponds to a capture array.

[0037] In other embodiments, multiple nanofilms 3 are provided, each nanofilm having an array of nanopores.

[0038] In some embodiments, the cell trapping layer 4 includes a fluid inlet 5 and a fluid outlet 7, and at least one trapping array 9 disposed between the fluid inlet 5 and the fluid outlet 7, wherein the at least one trapping array 9 is connected to the fluid inlet 5 and the fluid outlet 7 respectively via a microchannel network 10; and each collection chamber 11 corresponds to one trapping array 9 (i.e., each nanopore array corresponds to one trapping array 9).

[0039] Each row of the capture array 9 includes elements along the fluid flow direction (e.g., ...). Figure 1AMultiple capture units are spaced apart and connected (in the direction indicated by arrow A). Each capture unit includes a main channel 901 extending along the fluid flow direction and branch channels 902 located on both sides of the main channel 901. The branch channels 902 share the same inlet and outlet with the main channel 901, thus forming a θ-shaped capture unit. The capture area 9011 in the main channel 901 is smaller than the size of a single cell, and the size of the branch channels 902 is greater than or equal to the size of a single cell. Furthermore, both ends of the main channel 901 extend along the fluid flow direction and are connected to the main channels 901 of the capture units on its front and rear sides.

[0040] Specifically, see Figure 3A and Figure 3B The main channel 901 includes: a diversion region 9010, a first flow region 9012a, a capture region 9011 (large at both ends and small in the middle), a second flow region 9012b, a confluence region 9013, and a connection region 9014 located between two adjacent capture units and connected at both ends to the diversion region 9010 and the confluence region 9013 in the two adjacent capture units, respectively.

[0041] The sizes (such as width or pipe diameter) of the first and second flow regions 9012a and 9012b gradually increase from the end closer to the capture region 9011 to the end farther away from the capture region 9011. The diversion region 9010 and the confluence region 9013 are respectively connected to the two ends of the branch channel 902, so that in the same capture unit, the fluid delivered from the connection region 9014 by the diversion region 9010 is diverted to the first flow region 9012a and the branch channel 902 respectively, while the confluence region 9013 combines the fluids from the branch channel 902 and the second flow region 9012b (when no cells are captured) and delivers them to the next connection region 9014.

[0042] In some embodiments, the nanomembrane 3 is attached below the cell trapping layer 4 and has a pore size of approximately 800 nanometers. When a pulsed electric field is applied between the cell suspension in the trapping array 9 and the gold-plated substrate (specifically, electrodes are connected between the fluid inlet and the fluid outlet, for example, the metal needle of a syringe inserted at the fluid inlet and the fluid outlet is powered on, and it and the gold-plated substrate at the bottom of the collection chamber 11 serve as the upper and lower electrodes, respectively, thereby forming an electric field), the electric field lines preferentially concentrate in the nanochannels due to the low conductivity of the nanomembrane 3, resulting in an electric field focusing effect. This focused electric field can generate a transmembrane potential sufficient to reach the electroporation threshold in the region of the cell membrane near the nanochannels, thereby forming reversible pores in the cell membrane at low voltages (e.g., 9V).

[0043] After electroporation, negatively charged intracellular substances (such as RNA) are extracted and enriched in the collection chamber below through pores and nanochannels in the cell membrane under the influence of electrophoretic forces. After the electric field is removed, the cell membrane recovers due to the self-healing properties of the phospholipid bilayer, thus achieving efficient and low-damage sampling of living cells.

[0044] This device utilizes an innovative design that focuses the electric field through nanochannel thin films. By leveraging the local enhancement effect of the electric field on the nanochannels, it can achieve highly efficient electroporation of cells attached to the channels at low voltage, greatly reducing the electrical stimulation damage to the cells and thus resolving the contradiction between "high throughput" and "cell viability".

[0045] This device combines electrophoretic enrichment (e.g., driving sampled RNA onto magnetic beads in the collection chamber) with a signal amplification probe based on strand displacement reaction, enabling in-situ, highly sensitive detection on a chip. This avoids sample transfer and directly addresses the challenge of balancing "micro-sampling" and "detection sensitivity."

[0046] This invention, through its unique layered structure and functional integration, achieves a complete workflow integration from single-cell capture and live sampling to in-situ detection, providing core device support for dynamic gene expression analysis of single cells. Compared to the endpoint analysis of traditional single-cell RNA sequencing and the low-throughput limitations of other live-cell sampling technologies, it represents a fundamental paradigm shift. While traditional single-cell sequencing technology can provide high-throughput gene expression data, its principle of cell lysis prevents dynamic tracking of the same live cell, only providing static "snapshots." Furthermore, live-cell analysis technologies such as nanotube sampling are limited by extremely low throughput, making statistically significant population analyses difficult.

[0047] In some embodiments, when there are multiple capture arrays 9, the size of the capture region in the main channel 901 of the capture unit in the multiple capture arrays 9 gradually decreases along the fluid flow direction.

[0048] In one specific embodiment, the capture array 9 is configured as three groups, wherein the sizes of the capture areas in the main channel 901 from left to right (fluid flow direction) are 30μm, 20μm, and 15μm, respectively. Preferably, the first group of capture array 9 includes 18*64 capture units; the second group includes 18*96 capture units; and the third group includes 18*128 capture units.

[0049] In other embodiments, the microchannel network 10 includes: a first multi-level branched microchannel 101 disposed between the fluid inlet 5 and the at least one trapping array 9, and a second multi-level branched microchannel 102 disposed between the fluid outlet 7 and the at least one trapping array 9.

[0050] The first multi-level branched microchannel 101 and the second multi-level branched microchannel 102 each include: at least two branch networks connected sequentially along the fluid flow direction, preferably at least five branch networks; In the first multi-level branched microchannel 101, the number of branches in each branch network gradually increases along the fluid flow direction, so that each row in the capture array corresponds to a branched microchannel. In the second multi-level branched microchannel 102, the number of branches in each branch network gradually decreases along the fluid flow direction, so that all branched microchannels converge at the fluid outlet 7.

[0051] In this embodiment, the first multi-level branched microchannels 101 and the second multi-level branched microchannels 102 of the microchannel network resemble a tree diagram. The first multi-level branched microchannels increase in number along the fluid flow direction, while the second multi-level branched microchannels decrease in number along the fluid flow direction. Specifically, the branch network can have more segments. Each row of the capture array 9 is connected to each branched microchannel.

[0052] In some embodiments, the apparatus further includes a filtering module 6 disposed between two adjacent capture arrays 9.

[0053] In other embodiments, the apparatus further includes at least one filter module 6 disposed in a microchannel network 10 between the fluid inlet 5 and the at least one capture array 9.

[0054] In some embodiments, when there are multiple filtering modules 6, the size of the cell channels through which cells pass in the filtering modules 6 gradually decreases.

[0055] In some embodiments, the filtration module 6 includes: a plurality of columns of baffles spaced apart along the fluid flow direction, the baffles group including: a plurality of first baffles 601 extending perpendicular to the fluid flow direction, and a plurality of second baffles 602 extending parallel to the fluid flow direction; the first baffles 601 and second baffles 602 in each column are alternately arranged, thereby forming a transverse microchannel extending along the fluid flow direction between two adjacent first baffles and second baffles; and the first baffles 601 and second baffles 602 in each row extending along the fluid flow direction are alternately arranged, thereby forming a longitudinal microchannel extending perpendicular to the fluid flow direction between two adjacent first baffles 601 and second baffles 602.

[0056] In some embodiments, the size of the transverse microchannels in the plurality of filter modules 6 gradually decreases along the fluid flow direction.

[0057] In other embodiments, the size of the longitudinal microchannels in the plurality of filter modules 6 gradually decreases along the fluid flow direction.

[0058] In this embodiment, the dimensions of the transverse and longitudinal microchannels are 19 μm to 14 μm along the fluid flow direction.

[0059] In some embodiments, the dimensions of the transverse microchannels and the longitudinal microchannels in the filter module located at the cell inlet end of the capture array 9 are greater than or equal to the dimensions of the capture region in the main channel of the capture array.

[0060] In other embodiments, the dimensions of the transverse microchannels and the longitudinal microchannels in the filter module located at the cell outlet end of the capture array 9 are smaller than the dimensions of the capture region in the main channel of the capture array.

[0061] In some embodiments, the spacing between two adjacent capture units in the same row of the capture array 9 is 200 μm.

[0062] In other embodiments, the spacing between adjacent rows of capture units in the plurality of capture arrays 9 along the fluid flow direction gradually decreases and then gradually increases. For example, in this embodiment, the spacing between adjacent rows of capture units in the three capture arrays is 50 μm, 18 μm, and 23 μm.

[0063] In some embodiments, the device further includes: a plurality of microbeads 13 disposed at the bottom of the collection chamber, wherein probe markers are fixed on the microbeads 13, and each probe marker has a plurality of fluorescent groups and a plurality of fluorescent quenchers, preferably three fluorescent groups and three fluorescent quenchers.

[0064] In some embodiments, the plurality of microbeads 13 within the collection chamber 11 have at least two different sizes. Preferably, each collection chamber 11 contains two different sizes of microbeads 13. In practical applications, the number of microbeads 13 in each collection chamber 11, as well as the number of microbead sizes, can be adjusted according to the actual amount of target RNA to be detected.

[0065] Specifically, probe-labeled microbeads are immobilized on a gold-plated substrate via thiol-biotin bonds. These microbeads come in two different sizes (e.g., 1 μm and 2.8 μm) and their surfaces are modified with molecular beacon probes that specifically recognize target RNA. When sampled RNA molecules are enriched around the microbeads under an electric field, a hybridization chain reaction signal amplification is triggered, causing the microbeads to produce a fluorescent signal. By imaging the microbeads of different sizes and fluorescent colors using a fluorescence microscope and combining this with an intelligent recognition algorithm, simultaneous and quantitative in-situ analysis of multiple RNA targets from single cells can be achieved.

[0066] The device of this invention enables high-throughput single-cell capture through a cell capture layer, which is the foundation for high-throughput single-cell analysis. It provides a single-cell localization and fixation structure capable of spatially separating and immobilizing thousands of cells for independent manipulation. Furthermore, it offers a non-destructive sampling mechanism that can repeatedly extract intracellular substances from living cells without causing cell death (e.g., by applying a pulsed electric field to the device, driving target RNA from living cells to accumulate on microbeads within the collection chamber through a nanopore array on a nanomembrane). It also provides a detection and signal conversion interface that can convert the sampled trace biomolecular information into detectable signals in situ (e.g., fluorescence images; the image of the microbeads in the fluorescence image can be converted into a quantitative value characterizing the concentration of the target RNA, such as a Halo value, using an image processing algorithm in a subsequent embodiment).

[0067] Example 2: This example provides a method for high-throughput continuous sampling and gene detection of single live cells, applied to the aforementioned device for high-throughput continuous sampling and gene detection of single live cells. Accordingly, the method specifically includes the following steps: S101, Single-cell loading and capture: The cell suspension to be tested is injected into the device through the fluid inlet 5. Under the action of the fluid, the cells are evenly distributed into the capture array 9 through the microchannel network 10, thereby enabling the capture array 9 to capture the cells.

[0068] S102, Cell sampling and biomolecule enrichment: A pulsed electric field is applied between the electrode 8 of the capture unit and the collection chamber 11 to form nanochannels on the cell membrane, thereby allowing negatively charged RNA in the cell to be directionally enriched into the collection chamber 11 through the nanochannels and surround the probe-labeled microbeads 13 pre-fixed in the collection chamber 11.

[0069] In some embodiments, the key parameters of the pulsed electric field include: voltage, number of pulses, and pulse width. Preferably, the voltage range is 3V-9V, the number of pulses ranges from 50-100, and the pulse width ranges from 2ms-4ms.

[0070] S103, In situ hybridization and signal amplification: Biomolecules enriched around microbeads 13 hybridize with the probe, thereby triggering a strand displacement reaction, which separates the fluorescent group and quenching group in the probe and generates a fluorescent signal; subsequently, the free H2 probe in the solution replaces the RNA through a strand displacement reaction, allowing the bioRNA to participate in a new round of reaction, thereby achieving signal amplification.

[0071] See Figure 4 In the absence of target biomolecules, the DNA of the probe on microbead 13 (such as...) Figure 4In the H1 group, complementary pairing forms a hairpin structure, with the fluorophore and quencher closely adjacent, quenching the fluorescence; while when the target biomolecule, such as Figure 4 The red fragment shown in the image, where RNA binds strongly to H1, opens the card structure, separating the fluorophore and quencher, and illuminating the fluorophore. However, since there are also free H2 probes in the collection chamber 11, they bind even more strongly to H1, thereby displacing the target biomolecule, completing one strand displacement reaction. The displaced target biomolecule can then continue to undergo the same strand displacement reaction with the next DNA strand until all the DNA on the microbead 13 is illuminated. This process is repeated until all the microbeads 13 in the collection chamber 11 are illuminated.

[0072] The sequence listings of probe H1 and free probe H2 on the microbeads 13 in the collection chamber 11 in this application are shown in Table 1 below (the sequence listings in Table 1 are the sequence listings obtained after inserting fluorescent groups into the sequence listings in Table 2): Table 1. List of probe sequences with modified groups Table 2. Probe sequence listing S104, Multi-channel fluorescence imaging and intelligent recognition analysis: Multi-channel fluorescence imaging of microbeads on the detection layer is performed using a fluorescence microscope to obtain fluorescence images; and an intelligent recognition algorithm based on computer vision is used to process the images to obtain the type of RNA and the fluorescence intensity quantification value characterizing the RNA concentration.

[0073] In some embodiments, since each microbead 13 is equipped with only a probe capable of detecting one type of RNA, the corresponding target RNA can be identified simply by identifying the corresponding probe.

[0074] In some embodiments, the detection method further includes the step of: S105, Dynamic monitoring and continuous sampling: The cells are cultured on the cell capture layer or nanomembrane 3 for a specified time. After the specified time is reached, the liquid in the collection chamber 11 is replaced, and steps S102 to S104 are repeated to track the dynamic changes in gene expression of the same batch of single cells at different time points.

[0075] After a sampling is completed and the electric field is turned off, the cell membrane will heal itself, meaning the cells will not die. Therefore, they can continue to be cultured on the cell trapping layer or nanomembrane 3. Experiments have shown that cell viability remains above 85% even at a specified duration of 15 minutes. Therefore, preferably, the specified duration is greater than or equal to 15 minutes. More preferably, the specified duration is greater than or equal to 60 minutes.

[0076] In some embodiments, the step of processing the image using a computer vision-based intelligent recognition algorithm in step S104 specifically includes the following steps: S1041, the fluorescence image is scanned based on preset circular templates of various sizes to identify candidate regions corresponding to all microbeads in the fluorescence image.

[0077] Since the microbeads are spherical, circular templates corresponding to different sizes of microbeads are prepared in advance based on the different sizes of microbeads placed in the collection chamber. That is, one type of microbead corresponds to one circular template (the diameter of the microbead is the same as the diameter of the circular template).

[0078] In this embodiment, the candidate region refers to the fluorescent region in the fluorescence image whose shape matches the circular template. Specifically, since there may be overlap or occlusion in the fluorescence image, matching here means that only when the overlapping area between the fluorescent region and the circular template is greater than a preset threshold, such as more than 99%, it can be considered a match.

[0079] For example, for two different diameter microbeads, firstly, two different sized circular templates are configured, meaning each microbead corresponds to one circular template. See, for example... Figure 14 Circular templates with diameters of 1 micrometer and 2.8 micrometers are provided for microspheres with a diameter D = 1 micrometer and a diameter D = 2.8 micrometers, respectively. See [link / reference]. Figure 5 When two circular templates are compared with fluorescent regions A and B respectively, the overlap area between fluorescent regions A and B and their respective circular templates is 100%.

[0080] Although fluorescent regions A and B partially overlap (i.e., fluorescent region B is partially obscured by fluorescent region A), the distance between the center of fluorescent region A and the center of fluorescent region B is greater than the preset pixel threshold. Therefore, fluorescent regions A and B are selected as two candidate regions.

[0081] S1042, Remove the overlapping or occluded regions corresponding to the microbeads in the fluorescence image to obtain the effective region corresponding to each microbead.

[0082] See Figure 5Since fluorescent regions A and B overlap, the valid region of fluorescent region B consists of the entire circular fluorescent region A and the non-circular fluorescent region of fluorescent region B minus the overlapping portion. Similarly, since circular regions are extracted based on the circular template in step S1041, even obscured or overlapping regions are considered candidate regions. Therefore, the obscured or overlapping regions are removed, and the remaining regions are the valid regions. For example, since the sum of the absolute differences (SAD) of adjacent pixels at the boundary of fluorescent region A and fluorescent region B exceeds a preset threshold, fluorescent region B is marked as a special region. Then, the Canny edge detection algorithm is used to segment the part of fluorescent region B that is not occluded by fluorescent region A to obtain an irregular macroblock contour, which is taken as the effective region.

[0083] During the localization process, three situations may occur that lead to inaccurate localization of MBs (Microbeads): (1) a smaller mask is contained within a larger mask, (2) duplicate masks are generated from the same microbeads due to pixel blurring, and (3) signal overlap is caused by the microbeads being too close together. Therefore, candidate regions are merged based on the center-to-center distance, and then signal interruption regions are identified and segmented based on SAD to obtain irregular macroblock contours. That is, based on SAD, a region is identified as not a perfect single circle, but as an overlapping, broken, or irregular region, and then an edge detection algorithm is used to depict its actual appearance, such as Figure 5 The effective area obtained by subtracting the irregular shape of the overlapping part of circle B on the right and circle A on the left is the area of ​​circle A and circle B minus the irregular shape of the overlapping part.

[0084] S1043, classify each target RNA signal according to the red / green fluorescence intensity ratio of each effective region and its corresponding template size, and calculate its corresponding fluorescence intensity quantification value. The fluorescence intensity quantification value is proportional to the concentration of the target RNA and is used to characterize the concentration of the target RNA.

[0085] In some embodiments, the R / G ratio of each effective region is obtained (e.g., if the R / G ratio of the effective region meets the condition: 0 < R / G ratio < 0.1, then the effective region is identified as corresponding to the green channel; if the R / G ratio of the effective region meets the condition: 0.1 < R / G ratio < 10, then the effective region is identified as corresponding to the yellow channel; if the R / G ratio of the effective region meets the condition: 10 < R / G ratio < +∞, then the effective region is identified as corresponding to the red channel), and the size of the circular mask corresponding to each effective region is obtained (to distinguish two effective regions with the same color channel but different sizes), thereby distinguishing six targets: target RNA, as shown in Figure 17; then, the concentration of the corresponding target RNA is quantified by calculating the average gray intensity (defined as the Halo value in this application) within each effective region.

[0086] Example 3: To verify the performance of the device of the present invention in single-cell capture, flow rate simulations were performed on the capture unit and the cell capture layer, respectively. Specifically, using... Figure 3A The device shown was constructed, and a simulation model of it was built. Different cell densities (including 50 cells / µL, 100 cells / µL, 150 cells / µL, 200 cells / µL, and 250 cells / µL) were input into its fluid inlet. Figure 6 a) Then, fluorescence images obtained after capture experiments at different cell densities were acquired and statistically analyzed to obtain single-cell capture data. See [link to relevant documentation]. Figure 6 b and Figure 6 c.

[0087] Depend on Figure 6 As can be seen from a, with the increase of cell density, the cell capture rate gradually increases, while the single-cell rate gradually decreases. At an input cell density of 150 cells / µL, the cell capture rate is approximately 90% (this 90% refers only to the rate used within the specified range). Figure 3A The data obtained from the experiment using the display device shown is statistically analyzed. The single-cell rate is also approximately 90% (this 90% refers only to the rate achieved using...). Figure 3A The data obtained after the experiment with the display device shown is statistically analyzed, that is... Figure 3A The device shown captures approximately 4199 cells using three capture arrays: (64*18 + 96*18 + 128*18) * 90% * 90% ≈ 4000. In other words, over 4000 single cells are captured. Generally, a cell capture count exceeding 1000 is considered high-throughput capture sampling; therefore, this is sufficient to demonstrate that the device of this application can perform high-throughput single-cell sampling. Specifically, the capture rate refers to the ratio of the total number of captured cells to the total number of input cells. However, in practical applications, a small number of capture units may capture two or more cells; therefore, the single-cell rate refers to the ratio of the total number of single cells captured by a capture unit to the total number of captured cells.

[0088] Depend on Figure 6 As can be seen from c, when no cells are captured, the flow velocity at the trap structure in the capture unit is significantly higher than that in the side channels (i.e., the flow velocities in the first flow region 9012a, capture region 9011, and second flow region 9012b of the main channel 901 are significantly higher than those in the branch channels 902 on both sides, such as...). Figure 6 (As shown by the dashed line in c), this indicates that cells are more likely to pass through the trap structure (i.e., the first flow region 9012a, the trapping region 9011, and the second flow region 9012b in the main channel 901). However, cells (such as...) Figure 6 (As shown in the small circles in b) Once captured by the capture area 9011, the flow velocity in the main channel 901 decreases significantly, while the flow velocity in the branch channels 902 on both sides increases, as shown in b. Figure 6 As shown by the solid line in c. This change in flow rate guides uncaptured cells through the branch channels 902 on both sides into the next capture unit / capture array, thereby preventing cells from accumulating in the main channel 901 and having no effect on adjacent capture units.

[0089] See Figure 7 a, Figure 7 b and Figure 7 d. Before each capture unit in the cell capture layer captures a cell, the flow rate in the main channel 901 is significantly higher than that in the branch channel 902; see also Figure 7 a, Figure 7 c and Figure 7 e. After the capture unit captures the cell, the flow rate in the main channel 901 decreases significantly, while the flow rate in the branch channels 902 on both sides increases, such as... Figure 7 The red curve in e shows the vertical velocity distribution at the same location before and after the cell, as well as in adjacent units. See [reference needed]. Figure 7 b and Figure 7 The red lines in c indicate that capturing cells does not affect adjacent structures.

[0090] See Figure 7 f and Figure 7 As can be seen from g, even if the size of the capture units is different: larger capture units (such as...) Figure 3A Capture cells in a 64*18 capture matrix), medium-sized capture cells (such as...) Figure 3A The capture units in a 96*18 capture matrix and small-sized capture units (such as...) Figure 3A In the 128*18 capture matrix, the flow velocities of the three elements (capture units, capture units, and capture units) are almost equal at the capture point (or main channel), and the flow velocities of the two side branch channels are also almost equal. That is to say, the flow velocity of the cells in the channel is still uniform, and the size of the capture unit does not greatly affect the flow velocity of the fluid in the channel.

[0091] Example 4: In order to verify the influence of nanochannels on the electric field, simulation was performed to obtain the simulation results of the electric field distribution. The focused electric field lines confirmed the electric field focusing effect generated by the nanochannels.

[0092] Specifically, experimental and control groups were set up. The experimental group simulated applying a voltage to the device in the above embodiments to obtain the simulated electric field distribution results, see [link to relevant documentation]. Figure 8 Control group: Simulated electric field distribution results were obtained by applying voltage to a conventional detection device (device without nanopore array), see [link to relevant documentation]. Figure 9 .

[0093] Depend on Figure 8 and Figure 9 It can be seen that when a 9V voltage is applied, the focusing effect increases the transmembrane potential of the cell near the nanochannel from 0.8V to 2.23V, thus meeting the threshold (approximately 1V) required for effective cell electroporation. That is, in the control group, if a 9V voltage is applied, the resulting transmembrane potential is typically 0.8V (e.g., ...). Figure 9 As shown), in the device of this application, when a voltage of 9V is applied, the transmembrane potential obtained is 2.23V (as shown). Figure 8 As shown in the figure, this means that under the same applied voltage, the transmembrane potential is greatly increased, thereby reducing the requirements for power supply equipment.

[0094] Example 5: To further verify that the device of this application does not affect cell viability, cell viability was measured after the initial sampling and subsequent continuous sampling. Specifically, a preliminary safe electric field region (e.g., ...) was first determined experimentally. Figure 10 The cuboid region shown is such that applying an electric field according to the parameters corresponding to the cuboid will not affect cell viability. Then, under this safe electric field region, a specified electric field is applied to the device to perform initial sampling or continuous sampling, and the cell viability data after the initial sampling or continuous sampling is detected and analyzed statistically.

[0095] Determine the safe electric field area: By combining different voltages, pulse numbers, and pulse widths, and employing a single-variable experiment (i.e., changing only one condition at a time), multiple electric fields were obtained and applied to the aforementioned device of this application for cell sampling (i.e., driving the target RNA in the cell to move through the nanopore into the collection chamber by applying an electric field). Cell viability was then assessed after the initial sampling, and the results are shown in [reference needed]. Figure 10 .

[0096] For example, with a voltage of 9V and a pulse count of 100, different pulse widths were set: 2ms, 4ms, 6ms, 8ms, and 10ms, resulting in multiple electric fields for experiments. Similarly, with a voltage of 9V and a pulse width of 2ms, different pulse counts were set: 50, 100, 150, 200, and 250, resulting in multiple electric fields for experiments. Likewise, with voltages of 3V, 6V, 12V, and 15V, the same single-variable experiment was used to obtain multiple electric fields for experiments. Finally, the data were statistically analyzed, resulting in... Figure 10 The statistical results shown are as follows: The combination of the three indicators corresponding to the cuboid region in the figure (such as voltage of 3V-9V, pulse count of 50-100, and pulse width of 2-4ms) is safe for cells. Specifically, applying 9V, 100 pulses, and 4ms per pulse to the experimental group for sampling (driving the target RNA in the cells into the collection chamber) resulted in a 98% viable cell rate after the first sampling.

[0097] Continuous sampling: Furthermore, to verify the effect of different intervals on cell viability during continuous sampling using the device of this application, under the same conditions (e.g., applying an electric field of 9V, 100 pulses, each pulse lasting 4 milliseconds, and other conditions being the same), different intervals were set: 15 min (experimental group 1), 30 min (experimental group 2), 60 min (experimental group 3), and 90 min (experimental group 4) as experimental groups for continuous cell sampling experiments. The cell viability data obtained after continuous sampling at different intervals (i.e., secondary sampling) were then compared with the control group (i.e., no electric field was applied, but cells were continuously cultured on a membrane using the device of this application). Figure 11A .

[0098] Depend on Figure 11A It was found that after the initial cell sampling with an electric field of 9V, 100 pulses, and 4 milliseconds per pulse, and subsequent secondary sampling at different intervals, cell viability increased with increasing interval time. Furthermore, when the interval was greater than or equal to 60 minutes, the cell viability remained above 97% after the secondary sampling. This is because at least 60 minutes passed after the initial sampling before the secondary sampling, allowing the cells sufficient rest time to recover.

[0099] Furthermore, to verify that continuous cell sampling using the device of this application does not affect cell viability as long as the interval between continuous sampling (i.e., sampling the same batch of cells at least twice according to a preset interval) is long enough (e.g., greater than 60 minutes), experimental and control groups were set up for comparative analysis: Experimental group: Under an electric field of 9V, 100 pulses, and 4 milliseconds per pulse, continuous cell sampling was conducted at different intervals: 12h, 24h, 36h, and 48h after the initial sampling. Corresponding cell viability data were obtained after each sampling. Figure 11B As shown by the black lines in the middle; Control group: Cells were cultured on a membrane without an applied electric field, using the device described in this application, and cell viability data were acquired at 12h, 24h, 36h, and 48h, respectively. Figure 11B As shown by the medium gray line.

[0100] Depend on Figure 11B It can be seen that after five consecutive long-term samplings every 12 hours, the proportion of live cells did not decrease significantly compared with the control group (i.e., sampling without applying an electric field, with other conditions being the same).

[0101] In the aforementioned live cell sampling process, the applied electric field not only facilitated the extraction of target biomolecules (such as RNA) from the cell by mediating cell membrane electroporation, but also drove the enrichment of RNA molecules around the probe-labeled microbeads. The results are shown in [link to results]. Figure 12 .

[0102] Depend on Figure 12 As shown in diagram a, the electric field lines exhibit significant aggregation around the microspheres (also known as an enrichment effect), indicating that charged molecules tend to move towards the microspheres. Furthermore, this enrichment effect was further verified by the distribution of sodium fluorescein under an applied electric field, as shown in diagram a. Figure 12 b. Utilizing this enrichment effect, detection efficiency is significantly improved (e.g., detection time < 30 minutes, while detection time using traditional detection devices is typically > 30 minutes); and compared to enrichment conditions without an electric field, the fluorescence signal intensity is increased by 1.5 times, such as... Figure 12 c.

[0103] In this article, "live cell sampling" refers to the process where, after cells are captured by the capture unit and cultured on a nanomembrane, a specified voltage is applied to induce target molecules in the cells to enter the collection chamber through nanochannels on the cell membrane and nanopore arrays on the nanomembrane. Continuous sampling refers to the process where, after the initial sampling, a specified voltage is applied again after a certain period of time to induce target molecules in the cells to enter the collection chamber through nanochannels on the cell membrane and nanopore arrays on the nanomembrane; or the specified voltage is applied multiple times to induce target molecules in the cells to enter the collection chamber through nanochannels on the cell membrane and nanopore arrays on the nanomembrane, with a certain time interval between adjacent applications of the electric field.

[0104] Furthermore, by employing probe-based signal amplification technology, the device of this application is capable of detecting concentrations as low as 10 fM (i.e., 10). 15For target RNA (mol / L), see [link / reference]. Figure 13 and Figure 15 a- Figure 15f .

[0105] Example 6: To support multiplex detection of sampled target RNA molecules, probes targeting six biomarkers were designed, including a reference RNA, GAPDH, and five potential biomarkers for predicting immunotherapy: S100 calcium-binding protein A8 (S100A8), programmed cell death ligand 1 (PD-L1), melanoma-associated antigen 3 (MAGE-A3), carcinoembryonic antigen (CEA), and indoleamine 2,3-dioxygenase 1 (IDO1). Then, different concentrations of the corresponding target RNA molecules were individually detected for each probe to verify the linear relationship between the Halo value obtained from the fluorescence image using the device of this application, the fluorescence intensity, and the target RNA molecules. Finally, multiplex detection experiments were performed on the target RNA molecules of each of the six probes using the device of this application (see subsequent embodiments for details).

[0106] Regarding the linear relationship between Halo value and fluorescence intensity, and target RNA molecules: Six devices according to this application were set up, each with a large number of microbeads placed in each collection chamber. Each microbead was equipped with a probe for detecting a target RNA, resulting in six sets of fluorescence images of the microbeads. Figure 13 ; Furthermore, for each target molecule, cell fluid containing different concentrations of the target molecule was injected into the corresponding device, thereby obtaining fluorescence images after cell sampling experiments with different concentrations of different target molecules, such as... Figure 13 As shown: Specifically, for the reference probe GAPDH, the concentrations of the target molecules were set to 0M and 10M, respectively. 14 M, 10 10 M, 10 6 M was used in the experiment: Compared to the control group with a target molecule concentration of 0M, the intensity of the fluorescence image increased with increasing target molecule concentration. See [link to experiment]. Figure 13 and Figures 15a-15f ; Specifically, for the probes S100A8, PD-L1, MAGE-A3, CEA, and IDO1 in the experimental group, the concentrations of the target molecules were set to 0M, 10M, and 10M, respectively. 14 M, 10 10 M, 10 6 In experiments conducted with M as the target molecule concentration increased, compared to the control groups (i.e., target molecule concentration of 0M), the fluorescence image intensity increased with increasing target molecule concentration. (See [reference needed]). Figure 13 and Figures 15a-15f .

[0107] Furthermore, using computer vision-based intelligent recognition algorithms for MBs (microspheres) localization and signal analysis (see subsequent examples for the specific analysis process), the target molecule concentration and fluorescence intensity of the fluorescence image corresponding to each probe can be obtained (e.g., Figures 15a to 15f The relationship between the target molecule concentration and the Halo value of the effective region in the fluorescence image (showing a linear relationship) is as follows: when the concentration of the target molecule reaches a certain value, the concentration value of the target molecule and the Halo value of the fluorescence image show a linear relationship. Figure 16 As shown.

[0108] For example, such as Figures 15a-15f It can be seen that for probes GAPDH-S100A8, PD-L1, MAGE-A3, CEA, and IDO1, when the concentration of the target molecule is greater than 1 pM, the concentration of the target molecule is linearly related to the fluorescence intensity of the fluorescence image, and the linear detection range is 1 pM-40 pM.

[0109] Regarding multiplex detection of target RNA molecules: To verify the feasibility of multiplex detection of sampled target RNA molecules, experiments were conducted using the aforementioned six probes: namely, in Figure 3A The device depicted contains numerous microbeads of two different sizes in each collection chamber, with each microbead bearing only one probe for detecting the corresponding target RNA molecule. A specified electric field (e.g., 9V, 100 pulses, 4 milliseconds per pulse) is applied to the device for sampling and scanning, resulting in fluorescence images for each collection chamber. Since the probe types on the microbeads in each collection chamber include the six types mentioned above, a complex fluorescence image generated by the six target RNA molecules corresponding to one collection chamber is shown below. Figure 14 .

[0110] Furthermore, the complex fluorescence images generated by the above six target RNA substances were classified and their signals analyzed. For details, see [link to documentation]. Figure 14 Intelligent recognition algorithms based on computer vision are used for MBs (microspheres) localization and signal analysis. (1) MBs localization. First, two different sizes of circular masks (i.e., circular templates) were pre-defined, corresponding to microspheres of 1 μm and 2.8 μm respectively. The entire fluorescence image was scanned using these two masks. Then, any region with a signal intensity (i.e., grayscale intensity, or Halo value) higher than 10 and where the masked area covered more than 99% was marked as the effective region. See [link to relevant documentation]. Figure 14 i-iv.

[0111] During the localization process, three situations may occur that lead to inaccurate macroblock localization: (I) a smaller mask is contained within a larger mask; (II) duplicate masks are generated from the same microspheres due to pixel blurring; and (III) signal overlap occurs due to the microspheres being too close together. To address situations (I) and (II), if the distance between the center of any mask and the center of another mask is less than a preset threshold, such as two circular candidate regions of two pixels, they will be merged into a single continuous region as candidate regions. Masks with a distance greater than or equal to the preset threshold will each be considered as candidate regions. To address situation (III), firstly, the sum of absolute differences (SAD) of adjacent pixels at the mask boundary is calculated. When the SAD exceeds a preset threshold, such as 1388, signal interruption regions are identified, and these regions are subsequently classified as special regions. Then, the Canny edge detection algorithm is applied to segment these regions and depict irregular macroblock contours.

[0112] (2) Signal analysis. Six different fluorescence signal types were classified and their intensities were quantified. The fluorescence signal was distinguished using a two-parameter method: R / G ratio + mask diameter.

[0113] First, calculate the R / G ratio for each valid region (including special regions marked as valid regions). Then, classify the valid regions by comparing their R / G ratios (if the R / G ratio of a valid region meets the condition: 0 < R / G ratio < 0.1, then the valid region corresponds to the green channel; if the R / G ratio of a valid region meets the condition: 0.1 < R / G ratio < 10, then the valid region corresponds to the yellow channel; if the R / G ratio of a valid region meets the condition: 10 < R / G ratio < +∞, then the valid region corresponds to the red channel). See [link to relevant documentation]. Figure 14 Then, based on the mask size, it further distinguishes two types of macroblocks with the same color channel but different sizes, that is, effective regions of the same color but different sizes, thereby identifying six unique targets, such as... Figure 14 As shown in vi-vii: Two valid areas of the same color can be distinguished based on the diameters of different circular templates, D=2.8 and D=1.0.

[0114] Finally, the average gray intensity of each effective region is calculated to quantify the concentration of the corresponding target RNA molecule, thereby enabling the accurate classification and quantification of complex fluorescence images generated by six target RNA substances.

[0115] Specifically, see Figure 14 As shown in vii: Similarly, in the red channel, since probes GAPDH and S100A8 are set on microbeads of different sizes, they can be distinguished based on the diameter of the corresponding microbead's circular template; similarly, in the yellow channel, since probes PD-L1 and MAGE-A3 are set on microbeads of different sizes, they can be distinguished based on the diameter of the corresponding microbead's circular template; similarly, in the green channel, since probes CEA and IDO1 are set on microbeads of different sizes, they can be distinguished based on the diameter of the corresponding microbead's circular template.

[0116] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0117] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. A device for high-throughput continuous sampling and gene detection of single live cells, characterized in that, include: The cell capture layer (4) and gene detection layer (12) are stacked sequentially from top to bottom, wherein, The cell trapping layer (4) includes a fluid inlet (5) and a fluid outlet (7), and at least one trapping array (9) disposed between the fluid inlet (5) and the fluid outlet (7), wherein the at least one trapping array (9) is connected to the fluid inlet (5) and the fluid outlet (7) respectively through a microchannel network (10); Each row of the capture array (9) includes a plurality of capture units spaced apart and connected along the fluid flow direction; the capture unit includes: a main channel (901) extending along the fluid flow direction, and branch channels (902) located on both sides of the main channel (901), the branch channels (902) and the main channel (901) sharing the same fluid inlet (5) and the same fluid outlet (7), thereby forming a θ-shaped capture unit; wherein, the size of the capture area in the main channel (901) is smaller than the size of a single cell, and the size of the branch channels (902) is greater than or equal to the size of a single cell; The gene detection layer (12) is provided with a plurality of collection chambers (11) for collecting samples and open at the top. Electrodes (8) are provided in the collection chambers (11) and nanopore arrays are provided at the top of the collection chambers (11). Each nanopore array corresponds to a capture array (9).

2. The device for high-throughput continuous sampling and gene detection of single live cells according to claim 1, characterized in that, When there are multiple capture arrays (9), the size of the main channel (901) of the capture unit in the multiple capture arrays (9) gradually decreases along the fluid flow direction; And / or, The microchannel network (10) includes: a first multi-level branched microchannel (101) disposed between the fluid inlet (5) and the at least one capture array (9), and a second multi-level branched microchannel (102) disposed between the fluid outlet (7) and the at least one capture array (9). The first multi-level branched microchannel (101) and the second multi-level branched microchannel (102) each include at least two branch networks that are sequentially connected along the fluid flow direction. The number of branches in each branch network of the first multi-level branched microchannel (101) gradually increases along the fluid flow direction, so that each row of the capture array (9) corresponds to a branched microchannel; In the second multi-level branched microchannel (102), the number of branches in each branch network gradually decreases along the fluid flow direction, so that all branched microchannels converge at the fluid outlet (7).

3. The device for high-throughput continuous sampling and gene detection of single live cells according to claim 2, characterized in that, Also includes: A filter module (6) is disposed between two adjacent capture arrays; And / or, At least one filter module (6) is disposed in the microchannel network (10) between the fluid inlet (5) and the at least one capture array (9).

4. The device for high-throughput continuous sampling and gene detection of single live cells according to claim 3, characterized in that, The filtering module (6) includes: A group of multiple rows of baffles spaced apart along the fluid flow direction, the group of baffles comprising: a plurality of first baffles (601) extending perpendicular to the fluid flow direction, and a plurality of second baffles (602) extending parallel to the fluid flow direction. The first baffle (601) and the second baffle (602) in each column are arranged alternately, thereby forming a transverse microchannel extending along the fluid flow direction between two adjacent first baffles (601) and second baffles (602); and The first baffle (601) and the second baffle (602) in each row extending along the fluid flow direction are arranged alternately, so that a longitudinal microchannel extending perpendicular to the fluid flow direction is formed between two adjacent first baffles (601) and second baffles (602).

5. A device for high-throughput continuous sampling and gene detection of single live cells according to any one of claims 1 to 4, characterized in that, The spacing between two adjacent capture units in the same row of the capture array (9) is 200 μm; and / or, The spacing between adjacent rows of capture units in the plurality of capture arrays (9) along the fluid flow direction gradually decreases and then gradually increases.

6. The apparatus for high-throughput continuous sampling and gene detection of single live cells according to any one of claims 1 to 4, characterized in that, Also includes: Multiple microbeads (13) are disposed at the bottom of the collection chamber (11), and probe markers are fixed on the microbeads (13), and each probe marker has multiple fluorescent groups and multiple fluorescent quenchers.

7. The device for high-throughput continuous sampling and gene detection of single live cells according to claim 6, characterized in that, The collection chamber (11) contains multiple microbeads (13) having at least two different sizes.

8. A method for high-throughput continuous sampling and gene detection of single live cells, characterized in that, The apparatus for high-throughput continuous sampling and gene detection of single live cells as described in any one of claims 1 to 7, correspondingly, the method specifically includes the following steps: S101, Single-cell loading and capture: The cell suspension to be tested is injected into the device through the fluid inlet. Under the action of the fluid, the cells are evenly distributed into the capture array through the microchannel network, so that each capture unit in the capture array captures one cell. S102, Cell sampling and biomolecule enrichment: A pulsed electric field is applied between the electrodes of the capture unit and the collection chamber to form nanochannels on the cell membrane, thereby allowing negatively charged RNA in the cell to be directionally enriched into the collection chamber through the nanochannels and surround the probe-labeled microbeads pre-fixed in the collection chamber. S103, In situ hybridization and signal amplification: Biomolecules enriched around the microbeads hybridize with the probe, thereby triggering a chain substitution reaction, which causes the fluorescent group and quenching group in the probe to separate and generate a fluorescent signal. Subsequently, the free H2 probe in the solution replaces the RNA through a strand displacement reaction, allowing the RNA to participate in a new round of reaction, thereby amplifying the signal. S104, Multi-channel fluorescence imaging and intelligent recognition analysis: Multi-channel fluorescence imaging of microbeads on the detection layer is performed using a fluorescence microscope to obtain fluorescence images; and an intelligent recognition algorithm based on computer vision is used to process the images to obtain the type of RNA branch and the fluorescence intensity quantification value characterizing RNA concentration.

9. The method for high-throughput continuous sampling and gene detection of single live cells according to claim 8, characterized in that, It also includes the following steps: S105, Dynamic monitoring and continuous sampling: The cells are left to be cultured on the cell capture layer for a specified time. After the specified time is reached, the liquid in the collection chamber is replaced, and steps S102 to S104 are repeated to track the dynamic changes in gene expression of the same batch of single cells at different time points.

10. The method for high-throughput continuous sampling and gene detection of single live cells according to claim 9, characterized in that, Step S104, which involves processing the image using a computer vision-based intelligent recognition algorithm, specifically includes the following steps: S1041, The fluorescence image is scanned based on preset circular templates of various sizes to identify candidate regions corresponding to all microbeads in the fluorescence image; S1042, Remove the overlapping or occluded regions corresponding to the microbeads in the fluorescence image to obtain the effective region corresponding to each microbead; S1043, classify each target RNA signal according to the red / green fluorescence intensity ratio of each effective region and its corresponding template size, and calculate its corresponding fluorescence intensity quantification value. The fluorescence intensity quantification value is proportional to the concentration of the target RNA and is used to characterize the concentration of the target RNA.

Images