Method and system for microenvironment cell heterogeneity analysis of animal tissue

Abstract

The invention provides a method and a system for microenvironment cell heterogeneity analysis of animal tissues. The method comprises the following steps: obtaining a slice of animal tissue; carrying out multiple fluorescence immunohistochemical staining on the section, and carrying out quantitative image analysis to obtain a cell contour; based on the cell profile, using an automatic laser microdissection technology to obtain a first cell sample including a plurality of cell types on the tissue slice; performing flow cytometry sorting on the single-cell suspension of the animal tissues to obtain a second cell sample which comprises various cell types in a single tumor block; respectively carrying out micro proteome analysis on the first cell sample and the second cell sample to obtain proteome information of the first cell sample and the second cell sample; and determining spatial distribution information of different cell types in the animal tissue microenvironment by using the first cell sample proteome information and the second cell sample proteome information. According to the method, the space and cell type resolution for analyzing the cell heterogeneity of the animal tissue microenvironment is improved.

Description

Technical Field

[0001] The present invention relates to the field of biotechnology, and in particular to a method and system for analyzing microenvironment cell heterogeneity in animal tissues. Background Art

[0002] Cellular heterogeneity is ubiquitous in tissue microenvironments, and analyzing tissue microenvironment heterogeneity plays a vital role in studying human diseases. In recent years, transcriptome technology has made significant progress in analyzing tissue microenvironment heterogeneity, but compared with transcriptome research at the RNA level, proteome-related technology research at the protein level can best reflect the true expression level of genes and cell status.

[0003] The techniques for studying the heterogeneity of tissue microenvironment at the protein level mainly include: (1) dissociating the whole tumor tissue into a single cell suspension and analyzing the various cell types in the tissue microenvironment by flow cytometry or mass spectrometry; (2) performing multicolor fluorescence staining or imaging mass spectrometry analysis at the level of tissue sections to analyze the heterogeneity of tissue microenvironment at the spatial level. However, the above techniques can only study the expression of dozens of proteins at most.

[0004] Therefore, existing methods for analyzing cellular heterogeneity in animal tissues still need to be improved. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, one object of the present invention is to provide a method for effectively analyzing the microenvironment cell heterogeneity of animal tissues, especially tumor tissues.

[0006] In a first aspect of the present application, the present application proposes a method for analyzing the heterogeneity of microenvironment cells of an animal tissue. According to an embodiment of the present application, the method comprises: (a) obtaining a tissue section of the animal tissue; (b) performing multiple fluorescent immunohistochemical staining on the tissue section, and obtaining the cell contour of the section by quantitative image analysis; (c) performing automatic laser microdissection (LMD) on the section based on the cell contour to obtain a first cell sample, wherein the first cell sample comprises a plurality of cell types on the tissue section; (d) performing flow cytometric sorting on a single cell suspension of the animal tissue to obtain a second cell sample, wherein the second cell sample comprises a plurality of cell types in a single tumor mass of the animal; (e) performing microproteomic analysis on the first cell sample and the second cell sample independently to obtain proteomic information of the first cell sample and proteomic information of the second cell sample, respectively; and (f) determining the spatial distribution information of different cell types in the microenvironment of the animal tissue based on the proteomic information of the first cell sample and the proteomic information of the second cell sample.

[0007] Therefore, by adopting this method, the cell distribution of animal tissues can be effectively analyzed, thereby effectively improving the spatial and cell type resolution of analyzing the cellular heterogeneity of the animal tissue microenvironment.

[0008] In a second aspect of the present application, the present application proposes a system for analyzing cell heterogeneity of animal tissue. According to an embodiment of the present application, the system includes: a cell contour recognition module, which is used to obtain a tissue slice of the animal tissue, and perform multiple fluorescent immunohistochemical staining on the slice, and obtain the cell contour of a specific cell type on the tissue slice by quantitative image analysis;

[0009] A microdissection module, configured to automatically perform laser microdissection on a specific cell type on the slice based on the cell outline, so as to obtain a first cell sample, wherein the first cell sample includes a plurality of cell types on the slice;

[0010] A flow cytometry sorting module, used for performing flow cytometry sorting on a single cell suspension in a single tissue block of the animal to obtain a second cell sample, wherein the second cell sample includes a plurality of cell types in the animal tissue;

[0011] a microproteome analysis module, used for independently performing microproteome analysis on the first cell sample and the second cell sample, so as to obtain proteome information of the first cell sample and proteome information of the second cell sample, respectively; and

[0012] The cell distribution information analysis module is used to determine the spatial distribution information of different cell types in the microenvironment of the animal tissue based on the proteomic information of the first cell sample and the proteomic information of the second cell sample.

[0013] By adopting this system, the analysis method described in the first aspect can be effectively implemented, so that the cell distribution of animal tissues can be effectively analyzed, thereby effectively improving the spatial and cell type resolution of analyzing the cell heterogeneity of the animal tissue microenvironment.

[0014] It should be noted that the features and advantages described in this article for the analysis method are applicable to this system and will not be repeated here.

[0015] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0017] Figure 1 A schematic flow chart of a method for analyzing cellular heterogeneity in an animal tissue microenvironment according to an embodiment of the present invention is shown.

[0018] Figure 2 The spatial proteomic analysis of SCPro applied to mouse spleen according to one embodiment of the present invention is shown. Among them, a. The entire SCPro workflow is benchmarked on the mouse spleen. Left: High-quality multiplexed fluorescence microscopy images of 4μm thick mouse spleen FFPE (paraffin embedded) tissue sections after coverslip sealing. CD3e (T cells); CD19 (B cells); DAPI (nucleus). Center: Cutting path optimization and cell contour generation. Right: Cell contour import, aligned with the real-time image of the microscope to guide automatic laser microdissection. b. Comparison of microscope image quality with or without coverslips. c. The number of protein groups and precursors identified in the proteomic analysis of the B cell region and the T cell region. d. The coefficient of variation (CV) of the proteomic analysis of the B cell region and the T cell region. e. PCA (principal component analysis) shows the proteomic distribution of the B cell region and the T cell region. f. GO enrichment entries in the B cell region and the T cell region. g. Unsupervised hierarchical clustering shows the B cell region and the T cell region. The Z score indicates whether the protein is upregulated (red) or downregulated (blue).

[0019] Figure 3 The SCPro platform according to one embodiment of the present application is shown in pancreatic cancer KP f / fC mouse model, where a, demonstrating SCPro spatial proteomics in KP f / f C Workflow for mouse model application. b, 4 μm thick KP f / f C. Multiplex fluorescent immunohistochemistry whole-section images of mouse tissue sections. The dots of different colors in the figure indicate the LMD cutting positions of different cell types (n=3). c. Representative tissue cell flow cytometry analysis of PDAC-1 region. d. The proportion of CD45+ immune cells, αSMA+ fibroblasts and EpCAM+ epithelial cells in different regions of PDAC TME (pancreatic cancer tumor microenvironment). e. Distance map showing the spatial distribution of CD45+ immune cells and αSMA+ fibroblasts centered on EpCAM+ epithelial cells at different distances from tumor cells: 0-25μm (red); 25-50μm (yellow); 50-75μm (light blue); 75-150μm (dark blue). f. Line graph showing the proportion of CD45+ immune cells and αSMA+ fibroblasts centered on EpCAM+ epithelial cells at different distances from tumor cells. g. Display of ROI selection, cell classification and cell mask generation. h, Workflow for image alignment based on LMD real-time images and automated LMD with single-cell resolution. EpCAM (epithelial cells); CD45 (immune cells); αSMA (fibroblasts); DAPI (nuclei). PanIN, pancreatic intraepithelial neoplasia; PDAC, pancreatic ductal adenocarcinoma; IT, CD45+ immune cells in the tumor microenvironment; LN, CD45+ immune cells in lymph nodes; ROI, tissue region of interest.

[0020] Figure 4The results of the proteomic analysis of PDAC TME (pancreatic cancer tumor microenvironment) cell types based on FACS (flow cytometry sorting) according to an embodiment of the present application are shown, wherein a. The proteome coverage of 14 cell types, gray indicates the number of identified proteomes, and dark indicates the number of quantified proteomes. b. The PCA diagram of the average protein label-free quantitative intensity shows the principal component distribution of the analyzed cell types. Ellipses of different colors cover 4 cell lines, covering all cell types. c. The heat map shows the significantly differentially expressed proteins (LIMMA, p value <0.05, fold change>2) of each cell type, and known cell line markers are marked on the right. d. The dot plot shows the top three GO biological process entries enriched in the four cell lines, where the clusterProfiler algorithm is used for the analysis of differentially expressed proteins. e. The proportion of different immune cell subtypes in the IT and LN regions after Tangran deconvolution. f, Box plots of the proportions of lymphocytes and myeloid cells in the IT and LN regions after Tangran deconvolution (normalized). g, Multicolor immunofluorescence verification of the distribution and cell proportions of CD11b+ myeloid cells and CD3+ T cells (normalized). PCC, pancreatic cancer tumor cells; CAF, tumor-associated fibroblasts; iCAF, inflammatory tumor-associated fibroblasts; myCAF, myofibroblastic tumor-associated fibroblasts; apCAF, antigen-presenting tumor-associated fibroblasts; Treg, regulatory T cells; MYE, myeloid cells; NEU, neutrophils; MO, monocytes; MAC, macrophages; DC, dendritic cells. DETAILED DESCRIPTION

[0021] The present invention will be further described below in conjunction with specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and are not intended to limit the scope of the present invention. In addition, it should be understood that those skilled in the art can make various changes or modifications to the present invention, and these equivalent forms fall within the scope limited by the appended claims of the application equally.

[0022] definition

[0023] Unless otherwise specified, the term "sectioning" used in this article refers to fixing, soaking and embedding the tissue specimen to be studied to maintain its morphological and structural integrity, and using a microscopy knife or tissue slicer to cut the fixed tissue specimen into very thin slices. Generally speaking, the slices can be a few microns to tens of microns thick, depending on the requirements of the research purpose. Specifically to the present invention, according to the specific embodiments of the present application, slices with a thickness of no more than 10 microns can be used, which is convenient for subsequent analysis of single cells.

[0024] Unless otherwise specified, the term "multiple fluorescent immunohistochemical staining" used in this article refers to the tyramide signal amplification (TSA) technology, which is an enzymatic detection method that uses horseradish peroxidase (HRP) to perform high-density in situ labeling of target proteins or nucleic acids. Different primary antibodies are used to specifically bind to different target molecules, and the secondary antibody containing HRP is combined with a specific primary antibody to catalyze a fluorescent substrate. Each activated fluorescent substrate has a unique excitation and emission spectral characteristic, so that the position of a specific target molecule can be displayed. Therefore, according to the embodiments of the present application, the outline of each cell in a tissue section can be determined by detecting specific target molecules of specific cells, such as cell membranes, cell nuclei, etc.

[0025] Unless otherwise specified, the term "cell outline" used in this article refers to the outline of a specific cell determined by multiple fluorescent immunohistochemical staining and further image analysis. The cell outline may include the position of the cell membrane, the boundary between cells, or the position of a specific organelle (such as the nucleus). Thus, the cell outline can be subsequently used to perform laser microdissection on a specific cell type on a slice to obtain a single cell or a specific organelle.

[0026] Unless otherwise specified, the term "laser microdissection" used in this article refers to a high-precision molecular biology and histology technique that is used to select and separate specific cell groups or cell parts from complex cell mixtures or tissues. Generally speaking, in laser microdissection, a laser microdissection system is used that combines a microscope and laser cutting technology. First, the tissue specimen is cut into very thin slices and placed under a microscope to observe and select the cell area of ​​interest. Once the target cell area is determined, a laser cutting system is used to focus on the target area and gradually cut these cells or tissues. The cut cells or tissues can be directly collected and further used for molecular analysis, such as gene expression analysis, proteomics, DNA sequencing, etc. This is very useful for studying the functions and characteristics of selected cell subpopulations or specific cell types, especially for the analysis of cell heterogeneity in complex tissues, with high accuracy and resolution. Specifically to the present invention, by using "multiple fluorescent immunohistochemical staining" and "further image analysis" to obtain the "cell outline", single cells or specific organelles can be effectively cut, thereby realizing the analysis of the spatial heterogeneity of animal tissues.

[0027] Unless otherwise specified, the term "SISPROT technology" used in this article refers to an easy-to-use and integrated proteomics sample pretreatment technology based on a special gun tip (spintip) for deep proteomics analysis. For a detailed description of the SISPROT technology, please refer to J Chromatogr A. 2017 May 19; 1498: 207-214. doi: 10.1016 / j.chroma.2017.01.033. Epub 2017 Jan 13. In short, the core of this technology is a special gun tip, which is a small device that combines protein samples and digestive enzymes and is used to digest proteins and separate the resulting peptides. By combining enzymatic hydrolysis and separation steps, the loss of protein during sample pretreatment is greatly reduced, allowing the SISPROT technology to produce deep proteomics identification results with extremely small amounts of cell samples. In this technology, the proteins in the sample are first bound to the digestive enzyme on the gun tip. Then, the gun tip is placed in a centrifuge tube and incubated with the digestion solution. During the digestion process, the protein is enzymatically hydrolyzed into peptides. Finally, the separated peptides are collected from the gun tip and subjected to liquid chromatography-mass spectrometry analysis to identify and quantify the proteins in the sample. According to the embodiments of the present application, the method of Bepiao can be used. The kit implements the above technology. According to the embodiments of the present application, SISPROT technology has the characteristics of simplicity, efficiency and integration, and can effectively enhance the analytical ability of proteomics. It is widely used in proteomics research, especially in the research of complex samples and low-abundance proteins, to help discover and quantitatively analyze more proteins. In an example of the present application, thousands of proteins can be identified using mass spectrometry-based proteomics technology.

[0028] The expression "spatial distribution information of different cell types in microenvironment" used in this article refers to how different types of cells are distributed and organized in the microenvironment inside the organism. In the tissues and organs inside the organism, there are many different types of cells. These cells interact with each other and participate in the development and functional maintenance of tissues through mutual connection and interaction. In the microenvironment, the distribution and organization arrangement between these cell types are usually regular, and these laws form the basis of the organizational structure of different types of cells. These spatial distribution information can provide a basis for in-depth research on the development mechanism of the disease, for example, it can reveal the proteomic changes in the process of tumor progression, including prognostic markers and biological processes related to the progression of PDAC (pancreatic ductal adenocarcinoma, pancreatic cancer for short), which provides a reference for the discovery of potential biomarkers and the development of treatment strategies.

[0029] This application is completed based on the following findings of the inventors:

[0030] The application of traditional proteomic research based on fluorescent flow sorting of cell types in solid tumors is limited by the fact that a large number of cells are required during sample pretreatment. However, the tumor masses obtained from the clinic are usually small, which is not enough to obtain sufficient cell quantities for downstream proteomic research on multiple cell types in the tumor microenvironment (TME). Most of the previous applications of spatial proteomics technology based on microscopic imaging are HE and IHC stained sections, which are often based on manual identification of different regions or cell types by the naked eye, and then laser microdissection technology is used to obtain target cells for proteomic research. These require professional pathological experience to manually classify and identify cells, and cell cutting also requires manual operation. More importantly, these methods cannot achieve the study of the heterogeneity of different cell types in the tissue microenvironment at the level of single-cell resolution. What is obtained by cutting is a large piece of tissue, and the averaged results obtained are usually mixed with other cell types.

[0031] The inventors of this application have proposed an integrated platform for multi-color fluorescent staining on a single slice to identify multiple cell types in the microenvironment of segmented tissues, lossless cutting and collection of extremely small amounts of identified cells (no more than 100 cells), high-sensitivity proteomics sample pre-treatment and mass spectrometry analysis. On the other hand, although the cell type proteomics technology based on flow cytometry loses the spatial information of the tissue in situ during tissue dissociation, multiple cell types (usually more than a dozen) in the microenvironment of a single tissue block can be obtained through flow sorting, while the aforementioned spatial proteomics technology based on multi-color fluorescence of tissue slices cannot currently achieve so many cell types, so the inventors of this application combined the two, and studied the cellular heterogeneity of the tissue microenvironment by combining the cell type proteomics technology based on fluorescent flow sorting and the spatial proteomics technology based on multi-color fluorescent staining microscopic imaging.

[0032] In the first aspect of the present application, the present application proposes a method for analyzing the cell heterogeneity of animal tissue. According to the embodiment of the present application, the method as a whole includes two parts. A part of the animal tissue is made into slices, and the cell contour of the animal tissue is obtained by multiple fluorescent immunohistochemical staining and further image analysis. The cell contour is used to guide laser microdissection to obtain a first cell sample, and these cell samples are further analyzed to obtain the proteomic information of the first cell sample including spatial information. In addition, for the second part of the animal tissue, a second cell sample consisting of multiple single cells is obtained by flow cytometry sorting, and these single cells are further subjected to proteomic analysis to obtain the proteomic information of the second cell sample. The number of cell types in the second cell sample is much larger than that in the first cell sample, but the protein information of the second cell sample does not have the spatial information of the cell. Therefore, the protein information of the first cell sample and the protein information of the second cell sample are further processed comprehensively, for example, by deep learning, the protein information of the second cell sample is used as an input feature to predict the spatial location of each cell type of the second cell sample, so as to finally obtain the cell heterogeneity analysis result of the animal tissue. By combining the two analysis methods, the efficiency of studying the cell heterogeneity of the tissue microenvironment can be effectively improved. According to the embodiments of the present application, new cell subpopulations can be analyzed by the methods of the present application. For example, according to the embodiments of the present application, a Klrg1+Treg cell subpopulation that plays an immunosuppressive role in pancreatic cancer tumors was obtained and precisely located.

[0033] Reference below Figure 1 , the specific steps of the method for cellular heterogeneity analysis of animal tissues are described in detail.

[0034] According to an embodiment of the present application, the specific steps of the method for analyzing cell heterogeneity of animal tissues include:

[0035] (a) Obtaining tissue sections of the animal tissue.

[0036] According to an embodiment of the present application, FFPE sections can be used. Generally speaking, in FFPE processing, tissue samples are first fixed by formalin, which helps to protect the morphology and structure of the tissue and prevent cell and tissue degradation. Then, the fixed tissue samples are embedded in wax blocks, usually paraffin. This process involves immersing the tissue in hot wax, solidifying the wax and wrapping the tissue to maintain the stability of its morphology and structure. During the embedding process, equipment such as tissue specimen sets and embedding machines can be used. Finally, the tissue samples in the wax blocks are cut into very thin slices. According to this embodiment, the thickness of the slices does not exceed 10 microns, preferably 4 microns.

[0037] In addition, according to an embodiment of the present application, the animal tissue is a solid tumor tissue. The inventors of the present application have found that the method of the present application can perform cell heterogeneity analysis on a small amount of solid tumor tissue, while the previous cell type proteomic technology based on flow cytometry sorting mainly performs cell type proteomic analysis on large normal human tissues and blood tumors.

[0038] (b) performing multiple fluorescent immunohistochemical staining on the tissue sections, and obtaining the cell outlines of the sections by quantitative image analysis.

[0039] After obtaining the tissue section, multiple fluorescent immunohistochemical staining can be performed to obtain the cell outlines of the cells contained in the section. It can be understood by those skilled in the art that the cell outlines here include not only the boundaries between cells defined by the cell membrane, but also the boundaries of specific organelles, such as cytoplasm, nucleus, mitochondria, etc.

[0040] According to an embodiment of the present application, the multiple fluorescent immunohistochemical staining is performed using antibodies that specifically recognize cell membrane proteins and nuclear dyes that specifically recognize cell nuclei. According to an embodiment of the present application, the nuclear dye includes at least one of DAPI, hemoglobin, PI fluorescein, and Hoechst dye.

[0041] According to an embodiment of the present application, the cell membrane protein includes but is not limited to at least one of CD3e, CD19, CD45, tumor stromal cell protein SMA and tumor cell surface specific antigen EpCAM. Thus, by different fluorescence, the cell types on the slice can be classified and the range of each single cell can be divided. For example, DAPI can be used to divide the range of the cell nucleus, CD3e and CD19 can be used to visualize T cells and B cells respectively, such as CD45 for marking immune cells), SMA for marking stromal cells, and EpCAM for marking tumor cells to distinguish pancreatic cancer KP f / f C mouse model (transgenic animal model) with multiple cell subpopulations in the tumor microenvironment.

[0042] Based on the fluorescence signal, the cell membrane and cell nucleus information of the animal tissue can be determined, and based on the cell membrane and cell nucleus information, Gaussian blur and comparison algorithms are used to determine the cell contour suitable for guiding the automatic laser microdissection.

[0043] Gaussian blur is a commonly used image blurring method that reduces image noise and details by taking a weighted average of each pixel in the image. The core idea of ​​Gaussian blur is to take a weighted average of each pixel value with the values ​​of its surrounding pixels, and the weight is calculated by a Gaussian function. Generally speaking, the farther away from the center pixel, the smaller the weight of the pixel. The advantage of Gaussian blur is that it can smooth the image and reduce noise, and is suitable for removing details and textures in the image. It is often used in applications such as image preprocessing, image denoising, and image blurring. In practical applications, the degree of blur can be controlled by adjusting the radius parameter of Gaussian blur.

[0044] Comparison algorithms refer to methods for calculating and comparing similarities or differences between images or features. Comparison algorithms are widely used in image processing, pattern recognition, computer vision and other fields.

[0045] Specifically, according to an embodiment of the present application, after the cell nucleus and cell membrane can be identified by StrataQuest (SQ) software, the cell boundary is drawn on the corresponding cell type, and a filling mask is generated for subsequent laser microdissection and proteomics analysis. In order to optimize the cutting path of laser microdissection, the "fill hole" algorithm in the image quantitative analysis software can be used to generate a filling mask to fill the small gaps in the outline of large cells. According to a specific embodiment of the present application, the inventors confirmed the feasibility of multi-color fluorescent staining image-guided cell typing and cutting collection by using 4 μm thick mouse spleen FFPE tissue sections. Specifically, the inventors performed multiple immunofluorescence (mIHC) staining on the sections, using CD3e, CD19 and DAPI to visualize T cells, B cells and cell nuclei, respectively. For high-quality subcellular resolution images of T cells and B cells obtained. According to an embodiment of the present application, by adopting a filling hole algorithm, the cell outlines of T cell-enriched areas and B cell-enriched areas can be obtained, which are suitable for subsequent LMD and proteomics analysis.

[0046] (c) performing automated laser microdissection on the slice based on the cell outline to obtain a first cell sample, wherein the first cell sample includes a plurality of cell types on the single tissue slice.

[0047] After obtaining the cell outline, the slices can be cut using an automatic laser microdissection device. It can be understood by those skilled in the art that, in order to better meet the needs of the automatic laser microdissection device, the cell outline obtained by fluorescence analysis can be converted using an appropriate algorithm. According to an embodiment of the present application, in this step, based on each of the slices, the automatic laser microdissection produces 60 to 100 cells.

[0048] (d) performing flow cytometric sorting on the single cell suspension of the animal tissue to obtain a second cell sample, wherein the second cell sample comprises a plurality of cell types in a single tissue block of the animal.

[0049] According to the embodiments of the present application, by performing multicolor fluorescent staining on a single slice, 60-100 cells can be obtained by automatic laser microdissection, and deep proteomic analysis can be performed while retaining the spatial information of the cells (approximately 5,000 proteins can be identified from 100 cells), thereby analyzing the proteomic changes of tumor cells at different stages and immune cells in different regions during the progression of tumors such as pancreatic cancer.

[0050] However, since multicolor fluorescent staining can currently only stain about 10 markers on a single slice, it cannot cover most of the cell types in the tumor microenvironment. In order to better study the main cell types in the tumor microenvironment, the inventors proposed to simultaneously use fluorescent flow cytometry-based cell type proteomics technology to perform single-cell sorting and proteomic analysis on tissue samples.

[0051] According to an embodiment of the present application, the number of cell types in the second cell sample is greater than the number of the first cell sample. In this step, flow cytometry sorting can sort more than 10 types of cells, and 1000 cells are sorted for each type. According to a specific embodiment of the present application, proteomic analysis can be performed on 14 important cell types in the pancreatic cancer tumor microenvironment. Due to the limited initial amount of tumor samples, a maximum of 1000 cells can be sorted for each cell type, but based on the high-sensitivity microproteomics analysis platform for trace cell samples developed by the inventor, more than 7000 proteins can be identified, and a Klrg+Treg subpopulation with immunosuppressive characteristics was discovered.

[0052] (e) performing microproteomic analysis on the first single cell sample and the second single cell sample independently, so as to obtain proteomic information of the first cell sample and proteomic information of the second cell sample, respectively.

[0053] In this step, according to the embodiments of the present application, the following steps can be used to perform microproteome analysis:

[0054] (e-1) lysing the first cell sample or the second cell sample;

[0055] (e-2) using SISPROT technology to continuously separate the lysate obtained in step (e-1) to obtain multiple protein samples; and

[0056] (e-3) performing mass spectrometry analysis on the multiple protein samples to obtain the first cell sample proteome information or the second cell sample proteome information.

[0057] The SISPROT technology has been explained in detail above and will not be described again here. According to an embodiment of the present application, the mass spectrometry analysis uses a zero dead volume chromatographic column with an inner diameter of 50 microns, and the flow rate of the chromatographic column is 100 nl / min.

[0058] (f) determining the spatial distribution and cell ratio information of different cell types in the microenvironment of the animal tissue based on the proteomic information of the first cell sample and the proteomic information of the second cell sample.

[0059] In this step, by comprehensively analyzing the proteomic information of the first cell sample and the proteomic information of the second cell sample, the spatial distribution and cell ratio information of different cell types in the animal tissue can be finally determined.

[0060] According to an embodiment of the present application, the cell information of the animal tissue includes the cell types contained in the animal tissue and the content and location of each cell type in the animal tissue. For example, for a specific cell type, it can be determined whether it is located in cancer tissue or adjacent tissue.

[0061] According to an embodiment of the present application, at least a portion of the first cell sample proteomic information and the second cell sample proteomic information is input into a trained deep learning model to obtain the cell information of the animal tissue. Specifically, the deep learning model uses a deconvolution algorithm.

[0062] According to an embodiment of the present application, deconvolution is a data processing technique for recovering or inferring the original signal or source from the observed signal. Usually in the field of image processing and signal processing, deconvolution is mainly used to remove blur or restore the original signal. In image processing, deconvolution can be used for image restoration and super-resolution reconstruction. When the image is blurred or interfered by noise, deconvolution can restore the original image by inverse operation. The goal of deconvolution is to find a filter (or kernel) to eliminate the effects of blur or noise to reconstruct a high-quality image. In signal processing, deconvolution can be used to extract the original signal or source from the received signal. For example, in audio processing, deconvolution can be used to remove reverberation or echo to restore a clear audio signal. The process of deconvolution usually involves mathematical operations such as convolution and inverse filtering. Common deconvolution methods include Wiener filtering, Tikhonov regularization, and least squares solution. These methods select appropriate deconvolution algorithms to restore the original signal according to specific problems and signal characteristics. In tumor analysis, deconvolution has many applications. One of the applications is to infer the relative abundance of cell types or cell compositions from gene expression data of tumor tissues. Tumor tissues are complex mixtures composed of many different types of cells. By measuring the expression levels of genes, information about the relative abundance of different cell types in tumor tissues can be obtained. However, due to the presence of confounding effects and noise, it is challenging to accurately infer cell composition from gene expression data. Deconvolution can be applied to gene expression data of tumor tissues to infer the relative abundance of cell types by decomposing the mixed expression profile into components of different cell types. Deconvolution-based methods, such as the Tangram algorithm, can infer the relative abundance of cell types based on known gene expression profiles and mixture models. This method can help researchers understand the proportions of individual cell types in tumor tissues, thereby revealing the role of different cell types in tumor development and treatment.

[0063] Through the application of deconvolution, researchers can better understand the composition of tumor tissues and further explore the mechanisms of tumor development and treatment. This is of great significance for the realization of precision medicine and the formulation of personalized treatment plans.

[0064] According to the embodiments of the present application, inspired by the deconvolution algorithm of the spatial transcriptome, since the spatial proteome of multicolor fluorescence staining studies some large subgroups (with spatial distribution information) in the tumor microenvironment, and the cell type proteome based on fluorescence flow selects the subdivided subgroups under these large subgroups (without spatial distribution information), the inventor uses the deconvolution algorithm to study the spatial distribution changes of these subdivided subgroups. With the help of this algorithm, according to the specific embodiments of the present application, the inventor deconvoluted the CD45+ immune cells in the spatial proteome data, and obtained the expression of small subgroups of immune cells in the tumor microenvironment, and it can be determined that the Klrg1+ Treg subgroup is mainly distributed in the pancreatic cancer tumor microenvironment.

[0065] Therefore, by adopting this method, the cell distribution of animal tissues can be effectively analyzed, thereby effectively improving the resolution of analyzing the cell heterogeneity of animal tissues. According to the embodiments of the present application, the inventors combined the above two technical routes and proposed a method for revealing the heterogeneity of tumor microenvironment at the proteomic level with the help of deconvolution algorithms. This method can be widely applied to the study of various tumors and disease tissue microenvironments, thereby discovering some cell subpopulations with specific functions and potential therapeutic targets in the tumor microenvironment.

[0066] According to the embodiments of the present application, multiple cell types can be identified and segmented on multi-color fluorescent stained sections. The high-sensitivity proteomics sample pretreatment integrated platform established by integrating SISPROT technology, a 50-micron inner diameter chromatographic column, and a high-sensitivity mass spectrometer has extremely high sensitivity to trace tissue and cell samples, especially for the protein identification effect of stained tissue samples, which is difficult to achieve with existing sample pretreatment methods.

[0067] In addition, according to the embodiments of the present application, the cell type proteomics technology based on flow sorting can be used to divide into more subdivided cell subpopulations and obtain their proteomic advantages. The large subpopulations obtained by the spatial proteome are deconvoluted to obtain the proportional distribution of these subdivided subpopulations at the spatial level. This analysis method has only been reported in the transcriptome but not in the proteome.

[0068] Thus, the heterogeneity of tissue microenvironment is fully revealed through the integration of antibody staining-guided cell typing technology and deep proteomic analysis shown in the technology of this application (named by the inventor as SCPro platform). SCPro platform combines two complementary technologies: spatial proteomics guided by multicolor fluorescence staining imaging and cell type proteomics based on flow cytometry sorting. The main highlight of this platform is the spatial proteomics aspect of SCPro, which combines multicolor fluorescence staining images with high-sensitivity proteomic analysis, adding a proteomic information dimension with spatial resolution on the basis of traditional "digital pathology".

[0069] The application of SISPROT technology enhances the sensitivity and usability of the SCPro platform, allowing for in-depth proteomic analysis of a small number of cells (<100 cells) in stained tissue samples. The SCPro platform demonstrates high sensitivity (for example, in DDA MS / MS mode, 2,000, 4,000, and 5,000 proteomes can be identified from 10, 50, and 100 HEK 293T cells, respectively), which is comparable to the 10-100 cell identification depth of the NanoPOTS system in the field. Compared with cell samples, the main challenge in proteomics sample preparation for stained tissue samples is the removal of chemical dyes used for staining. Compared with other methods (such as One-POT technology), the solid phase extraction (SPE) principle of SISPROT technology overcomes these challenges and identifies more proteomes from a small number of tissue cells.

[0070] Microenvironment cell heterogeneity is a universal feature of biological systems, especially in the tumor microenvironment of solid tumors. According to the specific embodiments of the present application, through SCPro, the inventors quantitatively analyzed the in situ pancreatic cancer tumor microenvironment, revealed the spatial distribution heterogeneity of different cell types in the tumor microenvironment, and explained the potential mechanism of pancreatic cancer as the tumor progresses. The inventors combined quantitative tissue cytology and single-cell precise tissue cell separation with in-depth quantitative proteomic analysis, and performed in-depth proteomic analysis using 60-100 cells from a single tissue section, revealing proteomic changes during tumor progression, including prognostic markers and pathways associated with PDAC progression, providing a reference for potential biomarkers and treatment strategies.

[0071] Unlike the technical means of the present invention, previous flow sorting-based cell type proteomics technologies require a large number of cells and lack spatial distribution information, which limits its application in the analysis of limited biological samples. Utilizing the ultra-sensitive performance of the SCPro platform, the cell type proteomes of 1,000 cells in 14 cell types can be quantitatively analyzed from a single pancreatic tumor, including extremely low-abundance Treg cell subtypes. The SCPro platform complements the spatial proteomics data by integrating flow cytometry-based cell type proteomics data of multiple cell subtypes, and obtains the spatial distribution and cell proportions of major immune cell subpopulations. The deconvoluted cell proportions are basically consistent with the results obtained from the multi-color fluorescent staining images used for verification, confirming the feasibility of this concept.

[0072] In addition, according to the embodiments of the present application, by further analyzing the cell type proteomics data based on flow cytometry, a membrane protein screening strategy was developed, and the inventors identified highly specific cell surface membrane proteins, thereby discovering an immunosuppressive subtype of Treg cells, which only accounts for a small proportion in the tumor microenvironment. In addition, the location of Klrg1+Treg cells is predicted to be near the tumor rather than in the adjacent lymph nodes. This finding is significant because previous reports have shown that complete elimination of Tregs will accelerate tumor development due to compensatory myeloid infiltration. Such research is of great significance. Finding a more aggressive Treg subtype may achieve better results by targeting more aggressive Treg subtypes for tumor treatment.

[0073] In summary, the present invention provides a complete workflow for revealing the cellular heterogeneity of the tumor microenvironment, such as the heterogeneity of the pancreatic cancer tumor microenvironment, combining antibody-based cell typing technology and downstream microproteomics analysis platform, and finally proposed the concept of spatial proteome deconvolution, which is also applicable to other biological systems. It is hoped that a better understanding of the complex tissue microenvironment through the proteomics perspective will help make significant progress in the development of therapeutic strategies for tumors, especially pancreatic cancer.

[0074] In the second aspect of the present application, the present application proposes a system for analyzing cell heterogeneity of animal tissue. According to an embodiment of the present application, the system includes: a cell contour recognition module for obtaining a slice of the animal tissue, performing multiple fluorescent immunohistochemical staining on the slice, and obtaining the cell contour of the slice by quantitative image analysis; a microdissection module for automatically laser microdissection of the slice based on the cell contour to obtain a first cell sample, the first cell sample including multiple cell types on the slice; a flow cytometric sorting module for performing flow cytometric sorting on a single cell suspension of the animal tissue to obtain a second cell sample, the second cell sample including multiple cell types in the animal tissue; a microproteome analysis module for independently performing microproteome analysis on the first single cell sample and the second single cell sample to obtain the first cell sample proteome information and the second cell sample proteome information respectively; and a cell distribution information analysis module for determining the spatial distribution information of different cell types in the microenvironment of the animal tissue based on the first cell sample proteome information and the second cell sample proteome information.

[0075] By adopting this system, the analysis method described in the first aspect can be effectively implemented, thereby effectively improving the spatial and cell type resolution of analyzing the cell heterogeneity of the animal tissue microenvironment.

[0076] The technology of the present application is described below through specific embodiments.

[0077] General approach

[0078] If otherwise specified, in the following embodiments, the following methods will be used.

[0079] Mouse experiments

[0080] Mouse spleens were obtained from wild-type C57BL / 6J mice aged about 6-8 weeks. Kras mice were bred and raised in the Animal Experiment Center. LSL-G12D / + ;Trp53 flox ; Pdx1-Cre mice (referred to as KP f / f C mice), KP was determined by PCR using mouse tail identification f / f C. Genotype of mice. In order to collect KP f / f C tumors, mice were sacrificed by cervical dislocation, and then tumors were removed, washed twice with pre-cooled PBS, transferred to 4% PFA solution and fixed for 24 to 48 hours. After fixation, tissues were paraffin embedded for further analysis.

[0081] Multiplex immunohistochemical staining

[0082] The slide frame was treated with 0.1% (w / v) poly-D-lysine solution for 10 minutes. Then, 4-micron thick tissue sections were cut using a microtome and adhered to the membrane. The membrane was dried at 55 degrees Celsius for 30 minutes to promote better tissue adhesion. To ensure optimal staining, the slides were dewaxed before staining and heated in an oven at 65 degrees Celsius for one hour. The tissue sections were then dewaxed 3 times by soaking in 100% xylene for 10 minutes and rehydrated with a series of 100%, 100%, 90%, 80% and 70% ethanol for 5 minutes each, and then the tissue sections were washed with water for 5 minutes. The slides were multi-color fluorescently stained using a TSA kit (TissueGnostics, TGFP7100) according to the manufacturer's instructions. Primary antibodies: CD3 (dilution 1:100), CD19 (dilution 1:100) were used for mouse spleen staining. EpCAM (dilution 1:500), CD45 (dilution 1:1000), and SMA (dilution 1:500) were used to stain KP f / f C Mouse tumor section. Finally, the samples were mounted with antifade mounting medium containing DAPI and coverslipped to obtain high-quality images for further analysis.

[0083] Multiplex immunohistochemistry image acquisition and analysis

[0084] Prior to image acquisition, reference shapes for image alignment were marked onto the membrane by LMD. The entire slide image was first acquired at a 5x objective to identify the location of the tissue section, and then the entire image was acquired at a higher magnification to extract high-magnification grayscale images of each dye channel for downstream analysis. StrataQuest software was used for quantitative analysis and cell typing analysis of high-magnification images. For KP f / f C Cell typing of tumor sections. The cell nucleus recognition algorithm was used to identify the cell nucleus. The cell nucleus size parameter was set to 10 pixels, and weakly stained objects with grayscale values ​​less than 1 were removed. Then, based on the intensity of the biomarker and the cell morphology, the cell membrane recognition algorithm was used to identify the cell boundaries of the corresponding cell type. The cell membrane recognition parameters were set to -0.32μm inner diameter, 0.63μm outer diameter, and a maximum growth step of 4μm. After identifying the cell nucleus and cell membrane, a filling mask was generated on the original image of the corresponding cell type. Then, the original mask was optimized and expanded using the "Gaussian Blur" and "Compare" algorithms. The standard deviation of the "Gaussian Filter Blur" algorithm was set to 30, and the threshold of the "Compare" algorithm was set to 20.

[0085] Laser microdissection

[0086] The CellCut laser microdissection system (MMI) was used to collect the cut samples. The cell outline file exported from the image quantitative analysis software was imported into the MMI cutter. The cell outline was aligned with the live image of LMD. The cell outline was cut in bright field mode. The cutting parameters were set as follows: cutting speed 30 μm / s, laser power 15%, Z-axis deepening 1 μm, repeated 3 times. The cut samples were collected using IsolationCap (MMI) and stored at -20 °C for further analysis.

[0087] Tumor dissociation and cell sorting

[0088] The KP that will be removed f / f C mouse tumors were washed with pre-cooled PBS and excess fat and blood vessels were removed, then the tumors were minced into fragments and dissociated using a dissociator. After the program was terminated, the cell suspension was filtered through a 70 μm cell strainer (Corning) and washed twice with pre-cooled RPMI-1640 to obtain a single cell suspension. The single cell suspension was centrifuged at 300 g for 5 minutes and the supernatant was completely aspirated. One milliliter of staining buffer (BD Pharmingen) was used to resuspend the cell pellet. The number of cells was counted and divided into three groups for cell staining to reduce sample loss. The cells were stained according to the product instructions. The cell sorting procedure was performed using a flow cytometer.

[0089] Sample preparation using SISPROT

[0090] Lysis and Sonication: Samples were treated with lysis buffer containing 1% DDM, 10 mM HEPES (pH 7), 1 mM NaCl, 600 mM guanidine-HCl, and a protease inhibitor cocktail, followed by two cycles of non-contact sonication (30 s on, 30 s off each).

[0091] Sample treatment: Centrifuge the sonicated sample and transfer the supernatant to a new tube. Equilibrate with methanol and ammonia, then load the sample into the SISPROT tip containing DDM coating buffer.

[0092] Washing: Wash with 20% acetonitrile in 3M NH4OH, followed by incubation in pure acetonitrile for 10 min to remove impurities.

[0093] Protease digestion: Incubate at 37°C for 3 hours using a digestion buffer containing trypsin, Lys-C, and iodoacetamide.

[0094] Transfer and desalting of peptides: The digested peptides were transferred to the C18 layer of the SISPROT tip by acidic NaCl solution and then desalted twice with 1% formic acid.

[0095] Analysis of peptides: The eluted peptides were eluted into glass inserts with 80% acetonitrile and acetate and then lyophilized to dryness. Finally, the peptides were dissolved in 0.1% formic acid and subjected to liquid chromatography-mass spectrometry (LC-MS / MS) analysis.

[0096] High pH reversed phase chromatography

[0097] The inventors performed high pH reverse phase chromatography to generate KP f / f The deep proteome library of C tissue was used for DIA data analysis. f / f About 100 μg of peptide from mouse tumor sections was analyzed on the XBridge peptide BEH C18 column ( m, 2.1 mm×1 mm) using a 60-min gradient and formed into 24 steps (Agilent 1260). The peptide samples were vacuum dried and then used for LC-MS / MS analysis.

[0098] Liquid chromatography

[0099] The lyophilized peptides were redissolved in FA aqueous solution. Only 2 μL of the redissolved peptides were injected for single LC-MS / MS analysis. A separation column with an inner diameter of 20 μm was filled with 1.9 μm C18 beads (Dr.maisch) and coupled to a nanoElute liquid chromatography system (Bruker Daltonics) for separation of peptides and fractionation of peptide samples. Mobile phases A and B consisted of 0.1% FA in water and acetonitrile solutions, respectively.

[0100] Bioinformatics processing of proteomic data

[0101] Proteomic data analysis and visualization were performed using Perseus software (2.0.7.0) and R Studio (4.1.0). After excluding potential contaminants, the inventors filtered the quantified proteins with at least two valid values ​​in at least one cell type. Missing values ​​were imputed using a minimum valid value of 0.1. Statistical analysis and GO enrichment analysis were performed using the LIMMA package and clusterProfiler package, respectively.

[0102] To construct a membrane protein database, we first downloaded mouse protein sequences (17,119 entities) from Uniprot and retained 87 proteins with the "transmembrane" annotation. Then, we used Phobius and DeepTMHMM to predict transmembrane proteins using the previously downloaded protein sequences. Proteins that were not predicted as "transmembrane" by either method were discarded.

[0103] Example 1: Deep proteomic analysis and differentiation of two major subpopulations in mouse spleen using the SCPro platform

[0104] In this example, the SCPro workflow was benchmarked using archived FFPE mouse spleens to evaluate the ability of SCPro to perform cell typing, locate cell outlines, and cut and capture very small numbers of cells based on multicolor fluorescent stained sections. In addition, the performance of the SISPROT technology was evaluated for comprehensive proteomic analysis of rare cells isolated from antibody-stained FFPE tissues. Schematic results are shown in Figure 2 middle.

[0105] Specifically, after converting the cell outlines generated by the image quantitative analysis software into a file that can be read by the downstream LMD instrument, the mounting medium and cover glass were removed from the membrane before LMD to avoid interfering with the laser energy of LMD. In addition, a clearly identifiable reference shape (such as a triangle or square) was used to ensure that the cell outlines generated by the image quantitative analysis software were aligned with the real-time image of LMD. Subsequently, approximately 1000 cells (approximately 2000,000 μm) were successfully separated from the T cell area and the B cell area by LMD. 3 ) and proteomic analysis was performed, and all cells were successfully collected thanks to the sticky lid capture technology of the MMI LMD system.

[0106] In order to evaluate the performance of SISPROT technology in the SCPro platform, the inventors analyzed 1 / 5 of the peptide samples (about 200 tissue cells) treated with SISPROT by mass spectrometry. The results obtained nearly 90,000 precursor ions and more than 7,400 proteins with a low coefficient of variation (CV). Principal component analysis (PCA) showed that there were significant differences in the proteomic clustering of T cell regions and B cell regions, and gene ontology (GO) analysis showed that T and B cell-related signaling pathway proteins were enriched in their respective regions. It is worth noting that in the spatial proteomic data of T cell regions and B cell regions, known cell type-specific markers (such as Cd4, Cd3d, Cd3g and Cd8a for T cells; Cd19, Ms4a1, Cd79a and Cd22 for B cells) showed higher expression levels, indicating the accuracy of cell type recognition and cutting.

[0107] The results of the mouse spleen experiment verified the feasibility of the SCPro platform integrating multi-color fluorescent staining to identify cell contours, automatic laser fiber micro-dissection, and deep proteomic analysis of very small numbers of cells isolated from antibody-stained FFPE tissues.

[0108] Example 2: SCPro in pancreatic cancer KP f / f Application of C mouse model

[0109] After confirming the feasibility of the SCPro platform in mouse spleen experiments, the inventors applied the platform to the pancreatic cancer classic animal model KP, which has more clinical research significance. f / f C mice, schematic results are shown in Figure 3 middle.

[0110] Specifically, by archiving a single 4μm thick KP f / f C mouse paraffin-embedded pathological sections use CD45 to mark immune cells, SMA to mark stromal cells, and EpCAM to mark tumor cells at different stages of progression to distinguish six cell types in the pancreatic cancer tumor microenvironment, including acinar cells, pancreatic intraepithelial neoplasia (PanIN), advanced pancreatic cancer cells (PDAC), stromal cells (CAF), immune cells in the tumor microenvironment (IT), and immune cells in lymph nodes (LN). The inventors performed quantitative image analysis on the imaging results and found that the ratio of immune cells and stromal cells was centered on tumor cells and showed a downward trend, which well illustrates that in the pancreatic cancer tumor microenvironment, immune cells and stromal cells tightly wrap tumor cells to form a barrier for tumor treatment.

[0111] The inventors then further identified the cell outlines of the multi-color fluorescent staining imaging results and performed laser microdissection on the obtained cell outlines to obtain 60-100 cells for downstream spatial proteomic analysis. The inventors performed proteomic analysis on about 100 cells cut off and identified nearly 5,000 proteins, discovering the spatial proteomic characteristics of tumor cells, stromal cells, and immune cells in different pathological areas during the progression of pancreatic cancer.

[0112] The tumor microenvironment of pancreatic cancer contains a variety of cell types, and currently it is not possible to analyze so many cell types on a single tissue section. Therefore, the inventors conducted a proteomic analysis based on flow cytometry to supplement the spatial proteomic data and comprehensively describe the cellular heterogeneity of the pancreatic cancer tumor microenvironment. The pancreatic cancer tumor microenvironment is mainly composed of extracellular matrix, tumor-associated fibrous tissue, and immunosuppressive immune cells infiltrating it, forming a typical fibrous stroma and immunosuppressive microenvironment. The flow cytometry-based cell type proteome analysis mainly analyzed 14 major cell types in the pancreatic cancer tumor microenvironment, including CAFs and three CAF subtypes (myCAFs, iCAFs, and apCAFs), as well as nine immune cell subsets (B cells, CD4+T cells (T4), Tregs, CD8+T cells (T8), myeloid cells (MYE), dendritic cells (DC), macrophages (MAC), neutrophils (NEU), and monocytes (MO)), including pancreatic cancer cells (PCCs).

[0113] Although among the 14 cell types, CAFs (e.g., iCAFs and apCAFs) and immune cell subtypes such as Tregs were significantly expressed in KP f / f The proportion in C tumor single cell suspension was lower, but it was also consistent with previous research results.

[0114] Since the proportions of various cell types in the pancreatic cancer tumor microenvironment are low, in order to obtain sufficient cell numbers for proteomic analysis and ensure that all 14 cell types can be sorted, the inventors only sorted 1000 cells for proteomic analysis of each cell type. f / f The proportion of single-cell suspensions of C mouse tumors was extremely low, and only a few hundred cells could be successfully collected. Using the high sensitivity of SISPROT technology, 4,000 to 6,000 proteomes were identified for each cell type, and a total of more than 7,000 proteins were identified for all cell types. The schematic results are shown in Figure 4 middle.

[0115] Studying immune cells in the tumor microenvironment helps to understand the progression and prognostic characteristics of tumors. In order to gain a deeper understanding of the distribution of CD45+ immune cells and corresponding immune cell subsets in the pancreatic cancer tumor microenvironment, the inventors used Tangram's deconvolution algorithm. The aforementioned SCPro spatial proteomics provides spatial resolution of CD45+ immune cells, but due to the limitation of the staining channel of a single slice, it is impossible to conduct proteomic research on subdivided immune cell subsets. Therefore, the inventors proposed a method of deconvolving flow sorting cell type proteomics data that can obtain more cell subset protein expression conditions as a reference to overcome the limitations of limited staining channels in spatial proteomics. By deconvolving CD45+ immune cells, the predicted proportions of CD11b+ myeloid cells in the tumor microenvironment and lymph nodes were 77.53% and 23.61%, respectively, and the predicted proportions of CD3+T cells in the tumor microenvironment and lymph nodes were 22.47% and 73.39%, respectively. In order to verify the accuracy of the result, the inventors co-stained Krt19 (cancer cells), CD11b (myeloid cells), CD3 (lymphocytes) and DAPI to show the spatial distribution and true proportion of myeloid cells and lymphocytes. The image analysis results showed that the above predicted proportions were basically consistent with the true proportions of image analysis, proving the feasibility of the spatial proteomics deconvolution method proposed by the inventors. This method has great application potential in revealing the cellular heterogeneity of complex tumor microenvironments.

[0116] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "illustrative embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0117] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.

Claims

1. A method for analyzing the heterogeneity of microenvironment cells in animal tissues, characterized in that: include: (a) obtaining a tissue section of the animal tissue; (b) performing multiple fluorescent immunohistochemical staining on the tissue sections, and obtaining the cell outlines of the sections by quantitative image analysis; (c) performing automated laser microdissection on the slice based on the cell outline to obtain a first cell sample, wherein the first cell sample includes a plurality of cell types on the tissue slice; (d) performing flow cytometric sorting on the single cell suspension of the animal tissue to obtain a second cell sample, wherein the second cell sample comprises a plurality of cell types in a single tissue block of the animal; (e) performing microproteomic analysis on the first cell sample and the second cell sample independently, so as to obtain proteomic information of the first cell sample and proteomic information of the second cell sample respectively; and (f) determining the spatial distribution and cell ratio information of different cell types in the microenvironment of the animal tissue based on the proteomic information of the first cell sample and the proteomic information of the second cell sample.

2. The method according to claim 1, characterized in that The microproteome analysis further comprises: (e-1) lysing the first cell sample or the second cell sample; (e-2) performing sample pretreatment on the lysate obtained in step (e-1) using SISPROT technology to obtain multiple protein samples; and (e-3) performing mass spectrometry analysis on the multiple protein samples to obtain the first cell sample proteome information or the second cell sample proteome information.

3. The method according to claim 2, characterized in that The mass spectrometry analysis uses a zero dead volume chromatographic column with an inner diameter of 50 μm, and the flow rate of the chromatographic column is 100 nl / min.

4. The method according to claim 1, characterized in that: The cell information of the animal tissue includes the cell types contained in the animal tissue and the content and spatial location of each cell type in the animal tissue.

5. The method according to claim 4, characterized in that In step (f), further comprising: At least a portion of the first cell sample proteomic information and the second cell sample proteomic information is input into a trained deep learning model to obtain the cell information of the animal tissue.

6. The method according to claim 5, characterized in that The deep learning model uses a deconvolution algorithm.

7. The method according to claim 1, characterized in that The animal tissue includes solid tumor tissue.

8. The method according to claim 1, characterized in that The multiple fluorescent immunohistochemical staining is performed using antibodies that specifically recognize cell membrane proteins and nuclear dyes that specifically recognize cell nuclei. Optionally, the nuclear dye includes at least one of DAPI, hemoglobin, PI fluorescein, and Hoechst dye.

9. The method according to claim 8, characterized in that The cell membrane protein includes at least one of CD3e, CD19, CD45, tumor stromal cell protein SMA and tumor cell surface specific antigen EpCAM.

10. The method according to claim 1, characterized in that The multiple fluorescent immunohistochemical staining is used to determine the cell membrane and cell nucleus information of the animal tissue, and based on the cell membrane and cell nucleus information, Gaussian blur and comparison algorithms are used to determine the cell contour suitable for guiding the automatic laser microdissection.

11. The method according to claim 1, characterized in that: The number of cell types in the second cell sample is greater than that in the first cell sample.

12. The method according to claim 1, characterized in that The flow cytometry method can sort more than 10 types of cells, and for each type, 1000 cells or less can be sorted.

13. A system for analyzing microenvironmental cell heterogeneity in animal tissues, characterized in that: include: A cell outline recognition module, used for obtaining tissue sections of the animal tissue, performing multiple fluorescent immunohistochemical staining on the sections, and obtaining cell outlines of specific cell types on the tissue sections by quantitative image analysis; A microdissection module, configured to automatically perform laser microdissection on a specific cell type on the slice based on the cell outline, so as to obtain a first cell sample, wherein the first cell sample includes a plurality of cell types on the slice; A flow cytometry sorting module, used for performing flow cytometry sorting on a single cell suspension in a single tissue block of the animal to obtain a second cell sample, wherein the second cell sample includes a plurality of cell types in the animal tissue; A microproteome analysis module, used for independently performing microproteome analysis on the first cell sample and the second cell sample, so as to obtain proteome information of the first cell sample and proteome information of the second cell sample respectively; and The cell distribution information analysis module is used to determine the spatial distribution information of different cell types in the microenvironment of the animal tissue based on the proteomic information of the first cell sample and the proteomic information of the second cell sample.

Images