Esophageal squamous carcinoma early diagnosis marker based on spatiotemporal expression and application thereof
Through the early diagnosis markers of esophageal squamous cell carcinoma based on spatiotemporal expression, combined with spatial transcriptome sequencing and multiple immunofluorescence staining, the CAF-Epi niche was detected, and the difficulties in early diagnosis and risk assessment of esophageal squamous cell carcinoma were solved, and early identification and individualized treatment guidance was achieved.
Patent Information
- Application Number
- CN202510604030.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-12
- Publication Date
- 2025-08-12
AI Technical Summary
In the prior art, the precancerous lesions of esophageal squamous cell carcinoma are insufficient, and there is a lack of effective microenvironmental markers and early risk assessment methods. The sensitivity of traditional screening methods is limited, making it difficult to intervene early.
The presence and activity of the CAF-Epi niche was detected by spatial and temporal expression-based markers of early diagnosis of esophageal squamous cell carcinoma, including CAF markers and epithelial cell markers in CAF-Epi interactions, and early risk index (ESI) was calculated for early diagnosis.
It improves the recognition ability of early esophageal squamous cell carcinoma, integrates molecular and spatial positioning evaluation, and has rich information. The detection method can be used with the help of conventional immunostaining platforms, which is suitable for large-scale applications and can provide a decision-making basis for individualized treatment.
Smart Images

Figure CN120468426A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of medical technology, and more specifically, to an early diagnosis marker for esophageal squamous cell carcinoma based on spatiotemporal expression and its application. Background Art
[0002] Esophageal squamous cell carcinoma (ESCC), a common digestive tract cancer, has a high morbidity and mortality rate in my country. Among esophageal tumors, common subtypes include esophageal adenocarcinoma (EAC) and esophageal squamous cell carcinoma (ESCC), with significant differences in their prevalence. EAC is a subtype of esophageal cancer that is prevalent in Europe and the United States. ESCC, on the other hand, is commonly found in the upper and middle two-thirds of the esophagus and primarily originates from squamous cells in the epithelium. The vast majority of cases of this subtype originate in East Asia and Africa, and approximately 50% of deaths occur in my country, posing a serious health threat to the Chinese people. Before developing into typical invasive cancer, ESCC often undergoes atypical proliferation of squamous cells within the epithelial mucosa. Depending on the proportion of atypically hyperplastic epithelial cells, these pathological stages are designated low-grade intraepithelial neoplasia (LGIN) and high-grade intraepithelial neoplasia (HGIN), respectively. These two pathological stages are currently recognized as precancerous lesions of esophageal squamous cell carcinoma.
[0003] Patients with esophageal squamous cell carcinoma usually experience a series of symptoms such as dysphagia, swallowing pain, and weight loss, which bring great pain to patients. Esophageal cancer goes through five long stages, including precancerous lesions, subclinical stage, carcinoma in situ, invasive cancer (invasive cancer), and metastatic cancer, which takes about 2-20 years. The collagen fiber encirclement becomes increasingly weak, and the number of capillaries increases from nothing to many, and finally becomes a capillary mass. In the existing technology, endoscopic examination, pathological examination and other auxiliary examinations are usually used for the diagnosis of esophageal squamous cell carcinoma, specifically ordinary white light endoscopy, ultrasonic endoscopy, staining / magnifying endoscopy, barium meal angiography, CT, MRI, PET-CT, cytological screening or tumor marker auxiliary examination.
[0004] The above method and technology have the following disadvantages:
[0005] 1. Limited sensitivity of traditional screening: Existing screening methods, which mainly rely on tissue pathology, are inadequate in identifying the transformation process of precancerous lesions (such as high-grade / low-grade intraepithelial neoplasia) to invasive cancer, making it difficult to conduct timely and accurate early intervention.
[0006] 2. Lack of clear microenvironmental indicators: The progression of esophageal squamous cell carcinoma involves not only mutations in the epithelial cells themselves but also the synergistic effects of multiple factors within the microenvironment, including fibroblasts and immune cells. Previous studies have been limited to single molecules or cell types, and systematic microenvironmental indicators for early diagnosis and risk assessment have yet to be established.
[0007] 3. Biomarkers are not very applicable: Although some molecular markers are related to tumors, they lack the ability to predict whether a disease will progress rapidly to invasive cancer, making it difficult to widely carry out early risk population screening or individualized treatment decisions in clinical practice. Summary of the Invention
[0008] The purpose of this invention is to provide an early diagnosis marker for esophageal squamous cell carcinoma based on spatiotemporal expression and its application, aiming to address the problems existing in the prior art, such as insufficient ability to identify precancerous lesions, lack of effective microenvironmental markers, and lack of early risk assessment methods.
[0009] The above technical objectives of the present invention are achieved through the following technical solutions: an early diagnosis marker for esophageal squamous cell carcinoma based on spatiotemporal expression, characterized in that: the marker includes CAF markers and epithelial cell markers, which are typical molecules in CAF-Epi interaction; the CAF markers include two or more of α-SMA, FAP, MMP1, MMP11, POSTN, COL1A1, COL3A1 and COL6A1, and the epithelial cell markers include two or more of JAG1, SPP1, EPCAM, NFE2L2, ANXA1, SOX2 and TP63.
[0010] The present invention further provides a probe combination of early diagnosis markers of esophageal squamous cell carcinoma based on spatiotemporal expression, which is used for marker detection for early diagnosis.
[0011] Furthermore, the probe combination comprises the following components:
[0012] (1) Probes that specifically bind to at least two of the markers listed above: α-SMA, FAP, MMP1, MMP11, POSTN, COL1A1, COL3A1, and COL6A1, including but not limited to monoclonal antibodies, peptides, or other reagents that can specifically recognize the target;
[0013] (2) Probes that specifically bind to at least two of the markers JAG1, SPP1, EPCAM, NFE2L2, ANXA1, SOX2, and TP63, including but not limited to monoclonal antibodies, peptides, or other reagents that can specifically recognize the target.
[0014] Furthermore, the probe combination is used to perform immunohistochemistry (IHC) or multiplex immunofluorescence (mIF) staining.
[0015] The present invention further provides an early risk assessment system for early diagnosis markers of esophageal squamous cell carcinoma based on spatiotemporal expression: the assessment system is based on one of the following technical means:
[0016] (1) Spatial transcriptome sequencing data:
[0017] Divide the cell neighborhood into a predetermined radius (10-120 microns) with each cell as the center, count the cell types in the neighborhood, and cluster them to obtain multiple neighborhood categories;
[0018] a. To detect the presence of CAF-Epi niche-enriched neighborhoods characterized by the enrichment of both CAFs and invasive epithelial cells, primarily located at the epithelial-mesenchymal interface;
[0019] b. We continuously quantified the CAF-Epi niche status by weighted integration of multiple spatial metrics. Based on spatial transcriptome data, we calculated the following: the mean expression values of CAF markers at the cellular level (normalized to 0–1) and the mean expression values of invasive epithelial cell markers;
[0020] (2) Multiple immunofluorescence staining data: Tissue sections were immunofluorescently stained using a combination of fluorescent probes that label CAFs and invasive epithelial cells. The CAF infiltration index, epithelial cell invasion score, and collagen deposition index were calculated by spatial colocalization of protein expression or cell distance analysis to comprehensively evaluate the spatial status of the CAF-Epi niche.
[0021] Furthermore, when the determination system is based on spatial transcriptome sequencing data:
[0022] The system comprises:
[0023] Sample acquisition and spatial transcriptome sequencing module: Esophageal tissue sections were collected from the subjects, and the spatial transcriptome sequencing platform was used to detect the spatial expression data of α-SMA, FAP, MMP1, MMP11, POSTN, COL1A1, COL3A1, COL6A1, JAG1, SPP1, EPCAM, NFE2L2, ANXA1, SOX2, and TP63;
[0024] Cellular Neighborhood (CN) Partitioning and CN Classification Module: This module performs neighborhood partitioning on the spatial expression data detected by the Sample Acquisition and Spatial Transcriptome Sequencing modules. For each cell in all tissues, the cell neighborhood is defined as the set of all cells within a given radius, ranging from 10 to 120 microns. The distance between cells is described using the Euclidean distance of two-dimensional spatial coordinates. After summarizing the neighborhoods of all cells, the cell type characteristics of all neighborhoods are calculated. This characteristic is defined as a vector consisting of the proportion of each type of cell in the neighborhood under a specified set of cell types, and the neighborhood is classified according to the CN classification standard.
[0025] After classification, the risk of early diagnosis can be determined in the following two ways:
[0026] The CAF-Epi niche neighborhood is defined as an ecological niche where CAFs and epithelial cells co-enrich (CAF-Epi niche). It is characterized by a band-like distribution at the epithelial-stromal junction and is rich in invasive epithelial cells and cancer-associated fibroblasts. The CAF-Epi niche neighborhood is commonly found at the invasive front of high-grade intraepithelial neoplasia and around early-stage esophageal squamous cell carcinoma nests, representing a typical early spatial pattern. This method uses the presence and proportion of the CAF-Epi niche neighborhood in the sample as an indicator of early diagnosis: the presence of the CAF-Epi niche neighborhood and its proportion above a certain threshold indicates the presence of an early, high-risk lesion.
[0027] b. Continuously quantify the CAF-Epi niche status by weighted integration of multiple spatial metrics. Specifically, based on spatial transcriptome data, we calculated the average cellular expression of CAF markers (normalized to 0–1) and invasive epithelial cell markers. We also determined the area proportions of key cellular neighborhood categories—invasive epithelial-enriched neighborhoods, CAF-Epi niche neighborhoods, and cancer-associated fibroblast-enriched neighborhoods—in the entire tissue. We assigned weights to each metric based on its correlation with tumor invasive potential, and summed these weighted metrics to generate the ESI value, which continuously represents the activity of the CAF-Epi niche.
[0028] Scoring is performed using domain weights to obtain the Early-ESCC Spatial Index (ESI). The ESI calculation module is as follows:
[0029]
[0030] Among them, ω1∈0.2–0.4, ω2∈0.2–0.3, ω3∈0.1–0.3, ω4∈0.2–0.4, ω5∈0.1–0.3, each weight parameter is obtained based on the training sample fitting, and can be optimized and adjusted within the above range according to different populations or sample types; is the 0-1 normalized mean expression of CAF markers in all cells; is the 0-1 normalized average expression of invasive epithelial markers in all cells; P CNa P is the ratio of the area of invasive epithelial-rich neighborhood to the area of the entire tissue; CNb is the proportion of the CAF-Epi niche neighborhood area; P CNc is the area ratio of cancer-associated fibroblast-enriched neighborhood;
[0031] Risk determination module:
[0032] If ESI ≥ 0.50, the CAF-Epi niche is considered highly active, indicating an early high-risk state. It is recommended to consider neoadjuvant treatment interventions based on multiple factors including pathology.
[0033] If 0.3≤ESI<0.50 or P CNb >3%, considered suspicious and closely observed and followed up;
[0034] If 0.3<ESI and P CNa +P CNb +P CNc ≤1%, it is believed that the subsequent early chemoradiotherapy is effective and non-surgical treatment is preferred.
[0035] The present invention is further configured as follows: the following substances are used as markers in the spatial expression data of the sample acquisition and the spatial transcriptome sequencing module detection:
[0036] CAF markers: any combination of two or more of α-SMA, FAP, MMP1, MMP11, POSTN, COL1A1, COL3A1, and COL6A1;
[0037] Epithelial cell markers: any combination of two or more of JAG1, SPP1, EPCAM, NFE2L2, ANXA1, SOX2, and TP63.
[0038] The present invention is further configured as follows: the neighborhood division rules in the determination system are as follows:
[0039] Taking 50 μm as a representative value (which can be set to any radius within the range of 10–120 μm), with each cell as the center, the initial division of the cell neighborhood is performed based on its spatial coordinates obtained in the spatial transcriptome technology. First, Voronoi polygons are used to perform an initial spatial partitioning of the cell position to generate a non-overlapping basic neighborhood; then, based on the 10 nearest neighbors of each cell, an expansion is performed to form a locally overlapping cell neighborhood with a spatially overlapping relationship (Voronoi-knn algorithm). The neighborhood division strategy can be combined with the nearest neighbor expansion mechanism to obtain a more biologically representative microenvironment cell structure for subsequent cell composition vector construction and CN classification determination.
[0040] The present invention is further configured as follows: In a specific implementation, unsupervised clustering can be performed on the neighborhood to construct multiple niche neighborhood types, such as CAF-Epi niche neighborhood, invasive epithelial enriched neighborhood, CAF enriched neighborhood, etc., but the present invention is not limited to the specific classification type and number. The neighborhood includes:
[0041] Invasive epithelial-enriched neighborhood: Invasive epithelial cells are the main constituent cells, forming a tumor cell cluster area, which is a spatial area with malignant invasion characteristics;
[0042] CAF-Epi niche neighborhood: CAF cells are co-enriched with invasive epithelial cells and are mainly distributed in stripes at the interface between the epithelium and the stroma. This is a typical early invasion niche and is commonly found at the invasion front of high-grade intraepithelial neoplasia and around esophageal squamous cell carcinoma nests.
[0043] Cancer-associated fibroblast-enriched neighborhood: CAFs are the main enriched cell type, reflecting the aggregation status of tumor-associated fibroblasts in the matrix area and can serve as an important indicator of tumor interstitial activity.
[0044] When the determination system is based on multiple immunofluorescence staining data:
[0045] The system comprises:
[0046] (a) Multiple Labeling and Image Acquisition Module: Select at least two of each of CAF markers (including but not limited to α-SMA, FAP, MMP1, MMP11, POSTN, COL1A1, COL3A1, and COL6A1) and epithelial cell markers (including but not limited to JAG1, SPP1, EPCAM, NFE2L2, ANXA1, SOX2, and TP63) to form a fluorescent probe panel and construct a multiplex immunofluorescence (mIF) staining protocol to stain tumor tissue paraffin sections and acquire multichannel images. For example, four fluorescence channels can be selected: FAP, EPCAM, JAG1, and COL3A1. Channel selection can be flexibly adjusted based on the experimental purpose.
[0047] (b) Image preprocessing module: Perform image preprocessing such as background subtraction, signal enhancement, channel registration, and noise filtering on multi-channel images to ensure the accuracy of subsequent analysis. For example, import multi-channel images into software such as ImageJ, QuPath, or CellProfiler and perform basic preprocessing such as background subtraction, noise filtering, and inter-channel correction (to avoid false colocalization);
[0048] (c) Regional segmentation and cell recognition module: The image is segmented into tumor and stromal regions. The tumor epithelial region is identified based on the expression intensity of epithelial cell markers, and its periphery is defined as the stromal region. Subsequently, cell recognition and phenotype determination are performed in the tissue section image. The cell position is determined based on the cell nuclear signal, and the cell type is determined based on the expression results of CAF markers and epithelial markers. CAF cells are defined as cells that are positive for CAF markers and negative for epithelial markers, and epithelial cells are defined as cells that are positive for epithelial markers. For example, tumor parenchymal and stromal regions can be delineated. The EPCAM channel can be used to locate the epithelial region of the tumor, with strongly EPCAM-positive cell clusters representing the tumor region and its periphery as the stromal region. QuPath can automatically segment the tumor and stromal regions by training a classifier. ImageJ can manually delineate ROIs or generate masks based on EPCAM thresholds. Next, cell detection and phenotypic determination are performed across the entire slide: DAPI nuclear signals are used to locate cell nuclei, and FAP and EPCAM are combined to determine cell type—FAP+ / EPCAM- cells are considered CAFs, and EPCAM+ cells are considered tumor epithelial cells. Software such as CellProfiler can construct an analysis pipeline to sequentially complete cell nucleus identification, CAF and epithelial cell intensity measurement, and type determination.
[0049] (d) Quantitative index extraction: After identifying cell types and regions, the following key spatial indices (quantitative or semi-quantitative) are calculated to characterize the CAF-Epi niche:
[0050] CAF infiltration index: reflects the spatial infiltration degree of CAF relative to the tumor;
[0051] Epithelial cell invasion score: used to assess the expression level of epithelial markers (e.g., JAG1) in the CAF-Epi niche and their spatial co-localization relationship;
[0052] Matrix collagen deposition index: used to evaluate the abundance of various collagen fiber components (such as COL3A1, COL1A1, etc.) deposited in the tumor matrix area, reflecting the strength of the physical barrier formed by the CAF-Epi niche.
[0053] The present invention further provides the use of the above-mentioned early diagnosis markers and probe combinations in the following situations:
[0054] (1) Preparation of early diagnosis products for esophageal squamous cell carcinoma;
[0055] (2) Prepare a prognosis prediction product for esophageal squamous cell carcinoma.
[0056] The present invention further provides a spatial scoring method for evaluating esophageal squamous cell carcinoma based on early diagnosis markers of esophageal squamous cell carcinoma, characterized by: obtaining spatial expression data of CAF markers and epithelial cell markers in tissue samples; dividing cell neighborhoods within a predetermined radius with each cell as the center and classifying all neighborhoods according to cell composition characteristics; calculating a CAF-Epi niche scoring index based on the co-enrichment status of CAF cells and invasive epithelial cells in the neighborhood, which is used to quantitatively assess the activity of the CAF-Epi niche in the tissue to assist in determining the risk of malignant progression of the tissue sample.
[0057] In summary, the present invention has the following beneficial effects:
[0058] Early identification capability: Detection based on the CAF-Epi interaction niche can detect lesions with high invasive potential earlier than simple epithelial morphology.
[0059] Combining molecules and space: Fusion of multiple molecular markers and their spatial localization assessment provides richer information and higher reliability;
[0060] Strong scalability: The detection method can be used with conventional immunostaining platforms, and large-scale application can be implemented without significantly changing existing clinical processes. The functional modules involved in the system, such as image analysis and data calculation, can also be gradually implemented through existing digital pathology analysis software, and the establishment of new software modules is also relatively easy.
[0061] Precision treatment guidance: The CAF-Epi score not only assists in diagnosis but also provides a basis for decision-making for subsequent immunotherapy or anti-CAF programs, enabling personalized medication. BRIEF DESCRIPTION OF THE DRAWINGS
[0062] Figure 1 This is a schematic diagram of the calculation results of 11 types of cell areas in the spatial transcriptome and the analysis of related markers in the present invention.
[0063] Figure 2 This invention relates to the spatial transcriptome expression of the CAF-Epi niche in neoadjuvant therapy patients who respond and do not respond.
[0064] Figure 3 The present invention is about the multicolor immunofluorescence staining of actual cases and the survival of related samples based on the high and low expression of CAF-Epi niche score. DETAILED DESCRIPTION
[0065] The following is combined with Figure 1-3 The present invention is described in further detail.
[0066] Example 1: An early diagnosis marker for esophageal squamous cell carcinoma based on spatiotemporal expression
[0067] The early diagnosis markers in this embodiment include CAF markers, typical molecules in CAF-Epi interactions, and epithelial cell markers with invasive potential. The CAF markers include two or more of α-SMA, FAP, MMP1, MMP11, POSTN, COL1A1, COL3A1, and COL6A1, and the epithelial cell markers include two or more of JAG1, SPP1, EPCAM, NFE2L2, ANXA1, SOX2, and TP63.
[0068] Based on the above-mentioned early diagnosis markers, a probe panel for esophageal squamous cell carcinoma can be constructed. The probe panel contains the following components:
[0069] (1) Probes that specifically bind to at least two of the markers listed above: α-SMA, FAP, MMP1, MMP11, POSTN, COL1A1, COL3A1, and COL6A1, including but not limited to monoclonal antibodies, peptides, or other reagents that can specifically recognize the target;
[0070] (2) Probes that specifically bind to at least two of the markers JAG1, SPP1, EPCAM, NFE2L2, ANXA1, SOX2, and TP63, including but not limited to monoclonal antibodies, peptides, or other reagents that can specifically recognize the target.
[0071] The above-mentioned early diagnosis markers and related probe combinations can be used for:
[0072] (1) Preparation of early diagnosis products for esophageal squamous cell carcinoma;
[0073] (2) Prepare a prognosis prediction product for esophageal squamous cell carcinoma.
[0074] Example 2. An early risk assessment system based on early diagnosis markers for esophageal squamous cell carcinoma
[0075] The system, based on spatial transcriptome sequencing data, divides tissue samples into multiple cell neighborhood categories using spatial expression data. For neighborhoods containing CAFs and invasive epithelial cells, the spatial ecological niche index (ESI) is calculated using a neighborhood weighted score, where the weight is positively correlated with the neighborhood's invasive potential or malignancy.
[0076] Sample acquisition and spatial transcriptome sequencing module: Esophageal tissue sections were collected from the subjects, and the spatial expression data of FAP, MMP1, MMP11, POSTN, COL1A1, COL3A1, COL6A1, JAG1, SPP1, EPCAM, NFE2L2, ANXA1, SOX2, and TP63 were detected using the 10x Xenium InSitu platform.
[0077] The spatial expression data of FAP, MMP1, MMP11, POSTN, COL1A1, COL3A1, and COL6A1 were mainly screened as judgment markers. Figure 1 In the spatial transcriptome expression data, the expression levels of the selected FAP, MMP1, MMP11, POSTN, COL1A1, COL3A1, and COL6A1 were significantly higher than those of other CAF-related candidate genes. Combined with the existing published literature on fibroblasts (NF) and cancer-associated fibroblasts (CAF) and our laboratory's previous research on the tumor microenvironment of esophageal cancer, these genes have been proven to accurately represent and effectively distinguish CAF cell subpopulations, and therefore can be used as characteristic markers of CAF cells.
[0078] Neighborhood division and CN classification module: Perform neighborhood division on the spatial expression data detected by the sample acquisition and spatial transcriptome sequencing modules, and classify them according to the CN classification standard;
[0079] The neighborhood division rules are as follows:
[0080] With a radius of 50 μm and each cell as a seed, Voronoi-knn expansion was performed on the 10x Xenium space coordinates to obtain multiple overlapping local cellular neighborhoods (CN).
[0081] The classification of CN is as follows:
[0082] All neighborhoods were unsupervisedly clustered based on the CAF-cell ratio and the Invasive-Epi cell ratio (JAG1^high or SPP1^high), and the optimal number of clusters was 11. According to their cell type enrichment, these 11 CNs can be divided into three types: CN1, CN2, and CN4 are epithelial cell-enriched, CN3 and CN5 are mixed epithelial-mesenchymal cell types, and CN6-CN11 are mesenchymal cell-enriched.
[0083] There are three types of epithelial cell enrichment: CN1 is non-invasive cell enriched, which contains a very high proportion of non-invasive epithelial cells; CN2 is non-invasive cell & invasive cell, which is also dominated by epithelial cells, but the proportion of invasive cells and non-invasive cells is equal, and it is an area where the two epithelia exist together; CN4 is invasive cell enriched, which is dominated by invasive cells and is a malignant cell-enriched area.
[0084] There are two main types of epithelial-mesenchymal mixed types: CN3 is non-invasive cell & NF, the main enriched cell types in this type are non-invasive cells and NF, and this type of CN can be seen distributed in strips at the epithelial-mesenchymal interface; CN5 is CAF-Epi niche, this type of CN is an area enriched in invasive and CAF, similar to CN3, this CN is also distributed in strips at the epithelial-mesenchymal interface, the difference is that CN5 is commonly seen at the invasive front of the HGIN stage and around the cancer nest in the ESCC stage.
[0085] There are 6 types of non-epithelial cell-enriched types: CN6 is CAF-enriched, in which CAF is the main enriched cell type; CN7 is TLS (Tertiary Lymphoid Structures), which is named after the tertiary lymphoid structure, in which the main enriched cells are T cells and B cells; observation of its spatial position shows that this type of CN mainly appears in the lamina propria, which is a lymphocyte-enriched area, consistent with the definition of tertiary lymphoid structure; CN8 is NF-enriched, in which the main enriched cells are NF; CN9 is NF&Endothelial cell, which is a mixed area of NF and endothelial cells; CN10 is Glandular cell-enriched, which is a glandular cell-enriched area, mainly distributed in the submucosal layer, forming oval glands; CN11 is Muscle cell-enriched, which is a muscle cell-enriched area with a large number of muscle cells distributed.
[0086] After the classification is completed, scoring can be performed according to the following two scoring strategies:
[0087] Scoring strategy 1: Cell-based Neighborhood Index. The cell neighborhood is divided with each cell as the center according to a predetermined radius of 50 microns, and the cell type composition in the neighborhood is counted and clustered to obtain multiple neighborhood categories (such as CN1 to CN11). Among them, CN5 is defined as a niche neighborhood where CAFs and epithelial cells are co-enriched (CAF-Epi niche), which is characterized by a band-like distribution at the epithelial-stromal junction in the tissue and is rich in invasive epithelial cells and cancer-associated fibroblasts. The CN5 neighborhood is commonly found at the invasion front of high-grade intraepithelial neoplasia and around the early cancer nests of esophageal squamous cell carcinoma, representing a typical early spatial pattern. When using this method, the presence of CN5 type neighborhoods and their area proportion in the sample are used as early diagnosis judgment indicators: if the CAF-Epi niche neighborhood (CN5) is detected and occupies more than a certain proportion, it indicates the presence of early high-risk lesions.
[0088] Scoring strategy 2: CAF-Epi comprehensive scoring model (Early Spatial Index, ESI). This strategy continuously quantifies the CAF-Epi niche status by weighted integration of multiple spatial indices. Specifically, based on spatial transcriptome data, the mean expression values of CAF markers at the cellular level (normalized to 0–1) and invasive epithelial cell markers are calculated. The tissue-wide area proportions of key cellular neighborhood categories (e.g., CN4 invasive epithelial-enriched neighborhood, CN5 CAF-Epi niche neighborhood, and CN6 cancer-associated fibroblast-enriched neighborhood) are also determined. Each indicator is weighted according to its correlation with tumor invasive potential, and the weighted sum of these indices yields the ESI value, which continuously represents the activity of the CAF-Epi niche. Higher ESI values indicate a more active CAF-Epi niche and a higher risk of invasion. Early risk interpretation criteria can be set based on the ESI value. For example, when ESI ≥ 0.50, the CAF-Epi niche is judged to be positive (early high risk, requiring active intervention); when ESI is between 0.30–0.50 or the proportion of CN5 neighborhood exceeds the established threshold, it is judged to be suspicious (moderate risk, requiring close follow-up observation); when ESI < 0.30 and the CN5 proportion is extremely low, it is judged to be CAF-Epi niche negative (low risk, routine observation can be considered).
[0089] like Figure 1 As shown, the calculation results of 11 types of cell areas in the spatial transcriptome and the analysis of related markers.
[0090] Early-ESCC Spatial Index (ESI) calculation module:
[0091]
[0092] where ω1∈0.2–0.4, ω2∈0.2–0.3, ω3∈0.1–0.3, ω4∈0.2–0.4, ω5∈0.1–0.3; is the 0-1 normalized mean expression of CAF markers in all cells; is the 0-1 normalized average expression of invasive epithelial markers in all cells; P CN4 is the proportion of CN4 (invasive cell enriched) neighborhood area; P CN5 is the proportion of CN5 (CAF-Epi niche) neighborhood area; P CN6 is the proportion of CN6 (CAF enriched) neighborhood area;
[0093] The optimal values of ω1, ω2, ω3, ω4, and ω5 are as follows: ω1 = 0.3, ω2 = 0.25, ω3 = 0.2, ω4 = 0.3, ω5 = 0.2
[0094] Risk determination module:
[0095] If ESI ≥ 0.50, it is considered CAF-Epi niche positive / early high-risk and requires neoadjuvant chemoradiotherapy combined with surgery;
[0096] If 0.3≤ESI<0.50 or P CN5 >3%, considered suspicious and closely observed and followed up;
[0097] If 0.3<ESI and P CN4 +P CN5 +P CN6 ≤1%, it is believed that the subsequent early chemoradiotherapy is effective and non-surgical treatment is preferred.
[0098] Example 3. An early risk assessment system based on early diagnosis markers for esophageal squamous cell carcinoma
[0099] The system is based on multiple immunofluorescence staining data: using a combination of fluorescent probes that label CAFs and invasive epithelial cells, tissue sections are immunofluorescently stained. Through spatial colocalization of protein expression or cell distance analysis, the CAF infiltration index, JAG1 expression index, and collagen deposition index are calculated, respectively, to comprehensively evaluate the spatial status of the CAF-Epi niche.
[0100] The marker combination can be used in existing conventional immunohistochemistry or immunofluorescence staining experiments to detect the interaction between CAFs and epithelial cells in esophageal squamous cell carcinoma tissue and the expression distribution characteristics of related proteins.
[0101] The core innovation of this invention lies in the selection and application of a specific combination of CAF and epithelial cell markers, as well as the criteria for interpreting expression results, rather than the experimental process or staining steps of the immunofluorescence staining technique itself. The spatial expression characteristics of this combination are used to determine the extent of CAF-Epi niche formation, thereby guiding clinical diagnosis and prognosis.
[0102] In terms of specific experimental operations, users can directly refer to commercially available or publicly published standard operating procedures for immunofluorescence or immunohistochemistry staining. No special adjustments are required in the steps of sample preparation, antigen retrieval, primary and secondary antibody incubation, fluorescence development, microscopic analysis, etc., and only the specific marker combination provided by the present invention needs to be used for detection and analysis.
[0103] The operation of this embodiment is as follows:
[0104] Experimental Materials:
[0105] Human tissue sections were stained with primary antibodies against the following proteins: Jagged 1 protein (JAG1, ab7771, Abcam), fibroblast activation protein (FAP, 66562, Cell Signaling Technology), epithelial cell adhesion molecule (EPCAM, 21050-1-AP, Proteintech), and collagen type III (Collagen Type III, 22734-1-AP, Proteintech).
[0106] (1) Experimental steps:
[0107] Multiplex immunofluorescence (mIF) staining was performed on formaldehyde-fixed, paraffin-embedded tissue sections using a 7-color multiplex labeling kit (PANOVUE). The specific steps are as follows:
[0108] (a) Tissue processing: Paraffin-embedded tissue sections were deparaffinized by incubation in a 65°C oven for 1.5–2 hours. The sections were then dehydrated sequentially with Histo-clear II (20 minutes, twice), anhydrous ethanol (5 minutes, twice), and graded ethanol (95%, 85%, and 75%, 2 minutes each). Finally, the sections were equilibrated in ddH2O / TBST for 2 minutes. The sections were fixed in 10% formalin for 20 minutes and then washed in ddH2O / TBST for 3 minutes.
[0109] (b) Antigen retrieval: Select the retrieval method based on the antigen type, using sodium citrate retrieval solution or alkaline retrieval (boil over high heat → boil for 45 seconds → heat over low heat for 15 minutes → cool with ice-water mixture for 8 minutes).
[0110] (c) Blocking: Incubate with peroxidase blocking solution at room temperature for 20 minutes, incubate with goat serum blocking solution at room temperature for 20 minutes, wash twice with ddH2O / TBST, 3 minutes each time, and wipe dry the liquid around the slice.
[0111] (d) Primary Antibody Incubation: Prepare the antibody working solution according to the recommended dosage in the manufacturer's instructions and the pre-experimental ratio. Add 50-80 μL to each section and incubate at room temperature for 120 minutes or at 4°C overnight. Antibodies used include: (a) monoclonal antibodies that bind to at least two of the following target proteins: α-SMA, FAP, MMP1, MMP11, POSTN, COL1A1, COL3A1, and COL6A1; (b) monoclonal antibodies that bind to at least two of the following target proteins: JAG1, SPP1, EPCAM, NFE2L2, ANXA1, SOX2, and TP63. After incubation, wash the sections twice with ddH2O / TBST for 3 minutes each and wipe dry.
[0112] (e) Secondary antibody incubation: add goat anti-rabbit / mouse secondary antibody (50 μL / section), incubate at room temperature for 20 min, wash twice with ddH2O / TBST, 3 min each time, and wipe dry.
[0113] (f) Fluorescent signal incubation: TSA fluorescent reagent was diluted 1:1000 (TSA powder was dissolved in 75 μL DMSO), incubated at room temperature for 10 min, washed twice with ddH2O / TBST for 3 min each time and wiped dry.
[0114] (g) Multiple rounds of staining: Repeat the process of "antigen retrieval → washing → blocking → primary antibody incubation → washing → secondary antibody incubation → washing → fluorescent signal incubation → washing" until all antigen staining is completed.
[0115] (h) Cell nuclear staining and mounting: DAPI staining for 10 min, washed twice with ddH2O / TBST for 3 min each time, and mounted with anti-fluorescence quenching mounting medium after drying. Image acquisition was performed after solidification.
[0116] (2) Image analysis implementation plan:
[0117] (a) Multiplex labeling and image acquisition: A multiplex immunofluorescence (mIF) staining protocol, including four fluorescent channels for FAP, EPCAM, JAG1, and COL3A1, was used to label paraffin sections of tumor tissue. To ensure antibody specificity and minimal crosstalk between fluorescent spectra, spectral unconvolution techniques were employed as necessary. High-resolution fluorescence images of the entire section or representative fields of view were acquired. These images can be acquired using a confocal microscope or multispectral imaging system and saved as multichannel images for subsequent analysis.
[0118] (b) Image preprocessing: Import multichannel images into software such as ImageJ, QuPath, or CellProfiler. Perform basic preprocessing such as background subtraction, noise filtering, and inter-channel correction (to avoid false colocalization). For example, in ImageJ, you can apply a Rolling Ball background correction to each channel; in QuPath, you can use the built-in preprocessing module to reduce autofluorescence background.
[0119] (c) Regional segmentation and cell identification: Divide the tumor parenchyma area and the matrix area. The EPCAM channel can be used to locate the tumor epithelial area, with the EPCAM-positive cell cluster as the tumor area and its periphery as the matrix area. QuPath can automatically segment the tumor area and the matrix area by training the classifier; ImageJ can manually outline the ROI or generate a mask based on the EPCAM threshold. Then, cell detection and phenotypic determination are performed across the entire film: the DAPI nuclear signal is used to locate the cell nucleus, and then the FAP and EPCAM are combined to determine the cell type.
[0120] — FAP+ / EPCAM- cells are considered CAFs, while EPCAM+ cells are considered tumor epithelial cells (there is little overlap between the two). Software such as CellProfiler can establish a pipeline: first identify the cell nucleus, then measure the average intensity of FAP and EPCAM around each cell, and use a threshold to determine the cell type.
[0121] (d) Quantitative index extraction: After identifying cell types and regions, the following key spatial indices (quantitative or semi-quantitative) are calculated to characterize the CAF-Epi niche:
[0122] The CAF Infiltration Index (CAF Infiltration Index reflects the spatial infiltration degree of CAF relative to the tumor) is calculated using the following steps:
[0123] Step 1. Spatial scale calibration:
[0124] In image analysis software (such as ImageJ, QuPath, or CellProfiler), calibrate the actual spatial scale according to the scale bar of the microscope image to ensure that each pixel corresponds to a clear spatial distance (μm).
[0125] Step 2. Cell type identification:
[0126] Definition of tumor epithelial cells: EPCAM-positive signaling cells.
[0127] CAF definition: FAP-positive signal cells.
[0128] Step 3. Calculation method of CAF infiltration index (select one or more):
[0129] Method A (density method): The number of FAP-positive CAF cells in the tumor-stroma interface area (within 20 μm from the tumor boundary to the stroma) was counted and the density (number of CAF cells per square millimeter) was calculated.
[0130] Method B (spatial distance method, recommended): In the analysis software, the edge of the tumor cell cluster (EPCAM-positive) is used as a reference. Distance calculation functions such as "Distance Transform" or "MeasureObjectNeighbors" are used to count the spatial distance between CAF cells and the tumor boundary. The proportion of CAF cells located within 20 μm of the boundary to the total number of CAF cells is calculated.
[0131] The formula is:
[0132] Step 4. CAF Infiltration Index Grading Criteria:
[0133]
[0134] Epithelial Cell Invasion Score: This is used to assess the expression level of epithelial cell markers in the CAF-Epi niche and their spatial co-localization relationship (using JAG1 as an example). The specific steps are as follows:
[0135] Step 1. Express quantitative methods:
[0136] The JAG1 fluorescence signal intensity of tumor cells (EPCAM+) and CAF cells (FAP+) or the percentage of positive cells (such as expressed as the percentage of positive cells to the total number of cells) was measured using image analysis software.
[0137] Step 2. JAG1 expression intensity grading standard (positive rate or relative intensity):
[0138]
[0139]
[0140] Step 3. Spatial colocalization index (optional enhanced analysis):
[0141] A colocalization plug-in (such as the Colocalization plug-in of ImageJ or the QuPath colocalization module) was used to measure the ratio of colocalized pixels of EPCAM and JAG1 fluorescence signals at the interface, expressed as a percentage or Pearson coefficient, to assist in determining the degree of close contact between CAFs and tumor epithelial cells.
[0142] The Collagen Deposition Index (CDI) was used to assess the abundance of COL3A1 collagen fibers deposited in the tumor stroma, reflecting the strength of the physical barrier formed by the CAF-Epi niche. The steps are as follows:
[0143] Step 1. Calculation of collagen-positive area:
[0144] The COL3A1 channel was thresholded and segmented in the image analysis software to generate a positive area mask. The ratio of the positive area to the total area of the tumor surrounding stroma was calculated using the following formula:
[0145]
[0146] Step 2. Collagen deposition index grading standard:
[0147]
[0148]
[0149] The final CAF-Epi niche comprehensive score (0–3 points) is a weighted composite of the three indicators above. For example, the following example is used: CAF-Epi niche comprehensive score = (CAF infiltration score × 40%) + (JAG1 expression score × 30%) + (collagen deposition score × 30%).
[0150] Scoring definition: We graded the degree of CAF-Epi niche from 0 to 3 points, combining indicators such as CAF number distribution, JAG1 expression, and collagen deposition.
[0151] The rating levels and criteria are as follows:
[0152]
[0153]
[0154]
[0155] Suggested grading of comprehensive scores:
[0156]
[0157] This specific embodiment is merely an explanation of the present invention and is not intended to limit the present invention. After reading this specification, those skilled in the art may make non-creative modifications to this embodiment as needed. However, as long as such modifications are within the scope of the claims of the present invention, they are protected by patent law.
Claims
1. An early diagnosis marker for esophageal squamous cell carcinoma based on spatiotemporal expression, characterized by: The early diagnosis markers include typical molecular CAF markers and epithelial cell markers in CAF-Epi interaction; the CAF markers include two or more of α-SMA, FAP, MMP1, MMP11, POSTN, COL1A1, COL3A1 and COL6A1, and the epithelial cell markers include two or more of JAG1, SPP1, EPCAM, NFE2L2, ANXA1, SOX2 and TP63.
2. A probe combination comprising the early diagnosis marker for esophageal squamous cell carcinoma based on spatiotemporal expression according to claim 1.
3. The probe combination according to claim 2, characterized in that: The probe combination comprises the following components: (1) Probes that specifically bind to at least two of the markers listed above: α-SMA, FAP, MMP1, MMP11, POSTN, COL1A1, COL3A1, and COL6A1, including but not limited to monoclonal antibodies, peptides, or other reagents that can specifically recognize the target; (2) Probes that specifically bind to at least two of the markers JAG1, SPP1, EPCAM, NFE2L2, ANXA1, SOX2, and TP63, including but not limited to monoclonal antibodies, peptides, or other reagents that can specifically recognize the target.
4. The probe combination according to claim 2, characterized in that: The probe combination is used to perform immunohistochemical staining or immunofluorescence multiplex staining.
5. An early risk assessment system for esophageal squamous cell carcinoma based on the early diagnosis marker of claim 1, characterized by: The determination system is based on one of the following technical means: (1) Spatial transcriptome sequencing data: divide the cell neighborhood with each cell as the center according to a predetermined radius (10-120 μm), count the cell types in the neighborhood and cluster them to obtain multiple neighborhood categories; a. To detect the presence of CAF-Epi niche-enriched neighborhoods characterized by the enrichment of both CAFs and invasive epithelial cells, primarily located at the epithelial-mesenchymal interface; b. We continuously quantified the CAF-Epi niche status by weighted integration of multiple spatial metrics. Based on spatial transcriptome data, we calculated the following: the mean expression values of CAF markers at the cellular level (normalized to 0–1) and the mean expression values of epithelial markers; (2) Multiple immunofluorescence staining data: Tissue sections were immunofluorescently stained using a combination of fluorescent probes that labeled CAFs and epithelial cells. The spatial colocalization of protein expression or cell distance analysis was used to calculate the CAF infiltration index, epithelial cell invasion score, and collagen deposition index, respectively, to comprehensively evaluate the spatial status of the CAF-Epi niche.
6. The early risk assessment system according to claim 5, characterized in that: When the determination system is based on spatial transcriptome sequencing data: The system comprises: Sample acquisition and spatial transcriptome sequencing module: Esophageal tissue sections were collected from the subjects, and the spatial transcriptome sequencing platform was used to detect the spatial expression data of α-SMA, FAP, MMP1, MMP11, POSTN, COL1A1, COL3A1, COL6A1, JAG1, SPP1, EPCAM, NFE2L2, ANXA1, SOX2, and TP63; The Cell Neighborhood Partitioning and CN Classification Module performs neighborhood partitioning on the spatial expression data detected by the Sample Acquisition and Spatial Transcriptome Sequencing Module. For each cell in all tissues, the neighborhood of that cell is defined as the set of all cells within a given radius, ranging from 10 to 120 microns. The distance between cells is described using the Euclidean distance of two-dimensional spatial coordinates. After summarizing the neighborhoods of all cells, the cell type characteristics of all neighborhoods are calculated. This characteristic is defined as a vector consisting of the proportion of each type of cell in the neighborhood under a specified set of cell types, and then classified according to the CN classification standard. After classification, the risk of early diagnosis can be determined in the following two ways: The CAF-Epi niche neighborhood is defined as an ecological niche where CAFs and epithelial cells co-enrich (CAF-Epi niche). It is characterized by a band-like distribution at the epithelial-stromal junction and is rich in invasive epithelial cells and cancer-associated fibroblasts. The CAF-Epi niche neighborhood is commonly found at the invasive front of high-grade intraepithelial neoplasia and around early-stage esophageal squamous cell carcinoma nests, representing a typical early spatial pattern. This method uses the presence and proportion of the CAF-Epi niche neighborhood in the sample as an indicator of early diagnosis: the presence of the CAF-Epi niche neighborhood and its proportion above a certain threshold indicates the presence of an early, high-risk lesion. b. Continuously quantify the CAF-Epi niche status by weighted integration of multiple spatial metrics. Specifically, based on spatial transcriptome data, we calculated the following: the average expression values of CAF markers at the cellular level (normalized to 0–1) and the average expression values of epithelial cell markers. We also determined the area proportions of key cell neighborhood categories—invasive epithelial-enriched neighborhoods, CAF-Epi niche neighborhoods, and cancer-associated fibroblast-enriched neighborhoods—in the entire tissue. We assigned weights to each metric based on its correlation with tumor invasive potential, and the weighted sum of these metrics yielded the ESI value, which continuously represents the activity of the CAF-Epi niche. The domain weights are used to score the early esophageal cancer detection spatial index (ESI). The ESI calculation module is as follows: The weight parameters ω1, ω2, ω3, ω4, and ω5 are in the range of 0.2–0.4, 0.2–0.3, 0.1–0.3, 0.2–0.4, and 0.1–0.3, respectively. This ratio is the optimal parameter obtained based on statistical fitting of a large number of samples and can be appropriately adjusted according to different populations or data sets (within the above given range); is the average expression of CAF markers (such as α-SMA, FAP, MMP11, etc.) in all cells, normalized to 0-1; is the average expression of epithelial markers (such as JAG1, SPP1, etc.) in all cells, normalized to 0-1; P CNa P is the ratio of the area of invasive epithelial-rich neighborhood to the area of the entire tissue; CNb is the proportion of the CAF-Epi niche neighborhood area; P CNc is the area ratio of cancer-associated fibroblast-enriched neighborhood; Risk determination module: If ESI ≥ 0.50, the patient is considered to have a positive CAF-Epi niche / early high-risk disease and requires active intervention. Neoadjuvant chemoradiotherapy combined with surgery is recommended. If 0.3≤ESI<0.50 or P CNb >3%, considered suspicious and closely observed and followed up; If 0.3<ESI and P CNa +P CNb +P CNc ≤1%, the CAF-Epi niche is considered low-risk, and observation of the patient's condition is recommended. Targeted conservative treatment or non-surgical treatment may be suggested.
7. The early risk assessment system according to claim 5, characterized in that: The following substances are used as markers in the spatial expression data obtained from the sample and detected by the spatial transcriptome sequencing module: CAF markers: any combination of two or more of α-SMA, FAP, MMP1, MMP11, POSTN, COL1A1, COL3A1, and COL6A1; Epithelial cell markers: any combination of two or more of JAG1, SPP1, EPCAM, NFE2L2, ANXA1, SOX2, and TP63.
8. The early risk assessment system according to claim 5, characterized in that: The neighborhood division rules in the determination system are as follows: With a predetermined distance (ranging from 10 microns to 120 microns) as the radius and each cell as the center, the initial neighborhood is divided according to its spatial coordinates obtained in the spatial transcriptome technology; the neighborhood can form multiple local cell neighborhoods with overlapping relationships by combining spatial region division with the neighbor extension mechanism.
9. The early risk assessment system according to claim 5, characterized in that: The invasive epithelial-enriched neighborhood is the invasive epithelial cell-enriched type. This type of CN is mainly composed of invasive cells and is a malignant cell-enriched area; The CAF-Epi niche neighborhood is the cancer-associated fibroblast and epithelial niche. This type of CN is an area enriched with invasive epithelial cells and cancer-associated fibroblasts. This CN is distributed in a strip-like pattern at the epithelial-stromal interface and is commonly seen at the invasive front of high-grade intraepithelial neoplasia and around cancer nests in esophageal squamous cell carcinoma. The cancer-associated fibroblast-enriched neighborhood is the cancer-associated fibroblast-enriched type, in which CAFs are the main enriched cell type.
10. The early risk assessment system according to claim 5, characterized in that: When the determination system is based on multiple immunofluorescence staining data: The system comprises: (a) Multiple labeling and image acquisition module: a fluorescent marker combination consisting of at least two of the CAF markers and epithelial cell markers described in claim 1 is selected to construct a multiple immunofluorescence (mIF) staining protocol for labeling on paraffin sections of tumor tissue; (b) Image preprocessing module: performs image preprocessing such as background subtraction, signal enhancement, channel registration, and noise filtering on multi-channel images to ensure the accuracy of subsequent analysis; (c) Region segmentation and cell recognition module: The images were segmented into tumor and stromal regions. The tumor epithelial region was identified based on the expression intensity of epithelial cell markers, and its periphery was defined as the stromal region. Cell identification and phenotype determination were then performed in the tissue section images. Cell location was determined based on the cell nuclear signal, and cell type was determined based on the expression results of CAF markers and epithelial markers. CAF cells were defined as cells that were positive for CAF markers and negative for epithelial markers, and epithelial cells were defined as cells that were positive for epithelial markers. (d) Quantitative index extraction: After identifying cell types and regions, key spatial indicators for evaluating the CAF-Epi niche status are extracted. These indicators include, but are not limited to, the CAF infiltration index (reflecting the degree of spatial aggregation of CAF cells at the tumor boundary), the epithelial cell invasion score (based on one or more selected epithelial cell markers, such as JAG1, SPP1, and NFE2L2), and the matrix deposition index (preferably based on the spatial abundance of collagen components such as COL3A1 or COL1A1). Multiple indicators can be combined to construct a spatial scoring system for the CAF-Epi niche.
11. Use of the spatiotemporal expression-based early diagnosis marker for esophageal squamous cell carcinoma according to claim 1 or the probe combination according to any one of claims 2 to 4 in any of the following situations, characterized in that: (1) Preparation of early diagnosis products for esophageal squamous cell carcinoma; (2) Prepare a prognosis prediction product for esophageal squamous cell carcinoma.
12. A spatial scoring method for evaluating esophageal squamous cell carcinoma based on the esophageal squamous cell carcinoma early diagnosis marker according to claim 1, characterized in that: Obtain spatial expression data of CAF markers and epithelial cell markers in tissue samples; divide cell neighborhoods within a predetermined radius centered on each cell and classify all neighborhoods according to their cellular composition characteristics; Based on the co-enrichment of CAF cells and invasive epithelial cells in the neighborhood, the CAF-Epi niche scoring index was calculated to quantitatively assess the activity of the CAF-Epi niche in the tissue, thereby assisting in determining the risk of malignant progression in tissue samples.
Citation Information
Patent Citations
Esophageal squamous cell carcinoma microenvironmental cell marker molecular model and application thereof
CN109541209A
Reagent and kit for detecting esophageal squamous carcinoma or precancerous lesion and application
CN116356027A
Novel Biomarker for analysing prognosis of esophageal cancer
KR1020160021970A
Automated picking and packing system using sorter
KR1020250007938A
