Tumor grading detection system based on whole proteome cysteine reactivity analysis

Abstract

The invention relates to the technical field of pathological diagnosis, in particular to a tumor grading detection system based on holoproteome cysteine reactivity analysis, which comprises an in-situ labeling module, an in-situ detection module and an in-situ detection module, the histological separation module is used for carrying out molecular weight layering separation on the labeled tumor protein and counting the fluorescence signal intensity of each layer; the processing and analysis module is used for constructing a proteome-level cysteine reactivity map with position resolution capability; the tumor grading module is used for carrying out normalization processing on the average fluorescence intensity of each layer and the highest signal layer to obtain normalized intensity T, drawing a component specificity risk curve by taking the number of layers as a horizontal coordinate, and recording the area under the curve as the normalized intensity T as TSI; the TSI is in positive correlation with the tumor malignancy degree and serves as the basis of tumor grading. The method can accurately diagnose a tumor positive area, and the specificity and the sensitivity are both 100%.

Description

Technical Field

[0001] This invention relates to the field of pathological diagnostic technology, and in particular to a tumor grading detection system based on whole proteome cysteine ​​reactivity analysis. Background Technology

[0002] Tumor grading and staging are core indicators for assessing the malignancy and extent of spread of malignant tumors. Grading reflects the degree of differentiation and invasiveness of tumor cells, while staging describes the primary site of the tumor, lymph node metastasis, and distant spread. The combination of both can guide the selection of treatment plans and prognostic assessment.

[0003] Tumor grading is mainly based on cell morphology, degree of differentiation, and proliferative activity to determine the malignancy of a tumor.

[0004] Based on the degree of differentiation, they are often divided into:

[0005] 1. Highly differentiated (G1): Tumor cells are close to normal tissue, grow slowly, and have low invasiveness;

[0006] 2. Moderately differentiated (G2): Increased cellular atypia, moderate malignancy;

[0007] 3. Poorly differentiated or undifferentiated (G3 / G4): The cells have obvious abnormal morphology, are actively proliferating, and are prone to metastasis.

[0008] Staging is based on the extent of the primary tumor (T), lymph node metastasis (N), and distant metastasis (M), and commonly uses the TNM system or the I-IV stage classification.

[0009] TNM system:

[0010] 1. T (Tumor): Classified into T1-T4 according to tumor size or depth of invasion;

[0011] 2. N (Node): N0 (no lymph node metastasis) to N3 (extensive lymph node involvement);

[0012] 3. M (Metastasis): M0 (no distant metastasis) or M1 (metastasis).

[0013] Currently, tumor grading needs to be confirmed through biopsy or postoperative pathology, which may lead to issues with accuracy and timeliness. Summary of the Invention

[0014] In view of this, the technical problem to be solved by the present invention is to provide a tumor grading detection system based on whole proteome cysteine ​​reactivity analysis, which can achieve rapid and accurate tumor grading.

[0015] This invention provides a tumor grading detection system based on whole-proteome cysteine ​​reactivity analysis, comprising:

[0016] In situ labeling module: Tumor-selective dyes containing halogen substituents and / or alkynyl groups are used to react in situ with tumor characteristic proteins containing active cysteine ​​in tissue sections to specifically label the tumor characteristic proteins;

[0017] Histological separation module: Separates labeled tumor proteins by molecular weight into layers and counts the fluorescence signal intensity of each layer;

[0018] Processing and Analysis Module: By statistically summing the fluorescence signal intensities of each layer, the reactivity of cysteine ​​is reflected, and a proteomics-level cysteine ​​reactivity map with position resolution is constructed.

[0019] Tumor grading module: The average fluorescence intensity of each layer and the highest signal layer are normalized to obtain the normalized intensity T. The component-specific risk curve is plotted with the layer number as the abscissa. The area under the curve is the normalized intensity T, denoted as TSI. TSI is positively correlated with the degree of tumor malignancy and is used as the basis for tumor grading.

[0020] The tumor-selective dye structure includes halogen substituents and / or alkynyl structures, such as C-Cl bonds, which can be inserted into the hydrophobic pocket of tumor characteristic proteins and undergo nucleophilic substitution reactions with their thiol groups (-SH) to generate stable NIR-FPs.

[0021] In some specific embodiments, the tumor-selective dye is RC-NIRFR or RC-NIRFR-alkyne.

[0022] RC-NIRFR; RC-NIRFR- alkyne.

[0023] In some specific implementations, the tissue slices are slices of cancerous tissue, including but not limited to one or more of breast cancer, cervical cancer, lung cancer, bile duct cancer, and thyroid cancer.

[0024] The tumor-selective dye covalently binds to characteristic proteins with thiol groups in tissue sections in situ through a nucleophilic substitution reaction via halogen substituents and / or alkynyl groups in its structure, forming stable biomimetic near-infrared fluorescent proteins, thereby achieving highly selective fluorescent labeling of tumor characteristic proteins.

[0025] Preferably, the in-situ marking module includes the following steps:

[0026] S1) Frozen embedding of cancerous tissue;

[0027] S2) Cut the frozen-embedded tissue into tissue sections;

[0028] S3) The tissue slices are attached to the surface of the microporous array sample plate;

[0029] S4) Protein lysis and labeling were performed by co-incubating cell tissue lysate with dye.

[0030] First, the cancerous tissue removed during surgery is rapidly frozen and embedded.

[0031] Step S2) can specifically involve preparing tissue sections using a cryostat, wherein the thickness of the tissue sections is preferably 5 to 60 micrometers, and for example, it can be 20 micrometers.

[0032] The micropore array sample plate can be a micropore array sample plate well known to those skilled in the art, preferably made of transparent material, containing 10 to 100 micropores per square centimeter.

[0033] The cell tissue lysis buffer can be any cell tissue lysis buffer known to those skilled in the art, including but not limited to RIPA lysis buffer.

[0034] The histological separation module is used to spatially separate NIR-FPs-labeled proteins while preserving tissue morphology and spatial structure information, so that proteins of different molecular weight ranges can be analyzed in layers while retaining their histological location information.

[0035] Preferably, the histological separation module includes: a cathode plate, a microporous array template, a separation gel medium, an anode plate, an electrophoresis tank, and a constant voltage power supply.

[0036] In some specific implementations, the histological separation module specifically employs a HARP-C three-dimensional tissue electrophoresis device for detection and analysis.

[0037] The preferred separation gel medium is sodium dodecyl sulfate-polyacrylamide gel.

[0038] Preferably, the histological separation module specifically involves performing sodium dodecyl sulfate-polyacrylamide gel electrophoresis on the labeled proteins.

[0039] Preferably, the processing and analysis module specifically comprises:

[0040] After electrophoresis, the corresponding gel bands were cut and digested with trypsin and chymotrypsin. The processed samples were then analyzed by liquid chromatography-tandem mass spectrometry.

[0041] In some specific implementations, the histological separation module includes:

[0042] The labeled tumor proteins were separated into multiple layers by gel electrophoresis according to their molecular weight, and the fluorescence signal of each layer was counted.

[0043] In some specific implementation schemes, the labeled tumor proteins are separated into layers by gel electrophoresis according to their molecular weight. The key molecular weight components are selected into eight layers, denoted as F1-F8, and the fluorescence signals of F1-F8 are counted.

[0044] F1–F8 refers to bands 1–8 in the gel containing the target protein signal obtained during electrophoresis.

[0045] In some specific implementations, the processing and analysis module includes: reflecting the reactivity of cysteine ​​based on the sum of the fluorescence intensities of each layer from F1 to F8, and constructing a proteomics-level cysteine ​​reactivity map with position-resolved capability; wherein, the higher the fluorescence intensity, the more active cysteine-containing tumor proteins there are; by corresponding the location of the fluorescence to the tissue section and combining it with the fluorescence intensity, a proteomics-level cysteine ​​reactivity map with position-resolved capability can be constructed.

[0046] Preferably, the tumor grading module specifically includes:

[0047] (i) Quantitative T intensity measurement was performed on the specified tumor-positive electrophoretic components (T refers to the signal intensity of a certain molecular weight tumor protein labeled with dye).

[0048] (ii) Normalize the average T fluorescence intensity of each layer and the highest signal layer to obtain the normalized T intensity for cross-case comparison;

[0049] (iii) Construct component-specific risk curves from the normalized intensity of T;

[0050] (iv) The area under the risk curve is used as the final TSI value for breast cancer risk stratification (T scaledintensity, TSI refers to the normalized brightness curve of different layers obtained by normalizing the average fluorescence intensity of dye-labeled tumor proteins in each layer according to the brightest layer, and the area under the curve is the TSI value). The higher the TSI value, the more tumor proteins with active cysteine ​​are present in the tumor, and thus the more malignant the tumor. The TSI value and the degree of tumor malignancy are linearly positively correlated.

[0051] In some specific implementations, F1–F8 are imaged using a scanner (Azure, Sapphire) at 800 nm channels, with fluorescence intensity and pixel size set to 100 µm. Quantitative analysis of the sample signal and prognostic assessment are performed automatically by a risk assessment procedure; specifically, ImageJ software can be used for quantitative analysis of the fluorescence signal.

[0052] In some specific implementation schemes, the average T fluorescence intensity of F1-F8 is normalized to the highest signal layer to obtain the normalized T intensity. A component-specific risk curve is constructed by plotting the normalized T intensity against the layer number. The area under the curve is the normalized T intensity, denoted as TSI. TSI is directly proportional to the malignancy of the tumor, and the tumor is graded based on TSI.

[0053] Compared with existing technologies, this invention provides a tumor grading detection system based on whole-proteome cysteine ​​reactivity analysis, comprising: an in-situ labeling module: using tumor-selective dyes including halogen substituents and / or alkyne structures to perform in-situ reactions with tumor characteristic proteins containing active cysteine ​​in tissue sections, specifically labeling the tumor characteristic proteins; a histological separation module: separating the labeled tumor proteins by molecular weight stratification and statistically analyzing the fluorescence signal intensity of each layer; a processing and analysis module: reflecting the reactivity of cysteine ​​by statistically analyzing the sum of fluorescence signal intensities of each layer, and constructing a proteome-level cysteine ​​reactivity map with position resolution capability; and a tumor grading module: normalizing the average fluorescence intensity of each layer and the highest signal layer to obtain the normalized intensity T, plotting a component-specific risk curve with the layer number as the abscissa, and the area under the curve is the normalized intensity T, denoted as TSI; TSI is positively correlated with the degree of tumor malignancy, and is used as the basis for tumor grading.

[0054] This invention employs tumor-selective dyes containing halogen substituents (such as C-Cl) and alkynyl structures, which undergo specific nucleophilic substitution reactions with characteristic proteins containing thiol groups (-SH) in tissues, generating stable biomimetic near-infrared fluorescent proteins (NIR-FPs) in situ. This enables rapid and accurate tumor grading and pioneering spatial mapping of the whole proteome cysteine ​​reactivity in tissue sections. It allows for intraoperative diagnosis and prognostic assessment of tumor-positive areas in sections within 45 minutes, accurately diagnosing tumor-positive areas in human surgical resection specimens with 100% specificity and 100% sensitivity.

[0055] This invention establishes Tumor Standardized Intensity (TSI), a quantitative indicator based on the whole proteome cysteine ​​reactivity within a tumor region. It significantly optimizes clinical risk stratification for breast cancer patients, surpassing standard histological grading, T-staging, and N-staging. Compared to currently used independent staging methods (TNM staging system), this invention provides a continuous grading approach; that is, the higher the TSI, the higher the malignancy of the tumor. Attached Figure Description

[0056] Figure 1A proteomic range test map for assessing cysteine ​​reactivity using RC-NIRFR; where, a: RC-NIRFR structure and the use of each functional group; b: RC-NIRFR markers of human breast cancer tissue lysates (untreated, heat-denatured, trypsin-digested); c: Workflow for assessing the difference in cysteine ​​reactivity between tumor (T) and adjacent normal (N) tissues in breast cancer patients (n=13); d: Protein counts of T-specific, N-specific, and non-specific categories before and after RC-NIRFR enrichment; e: Proportion of cysteine-reactive proteins: T-specific (81%), non-specific (27%), N-specific (20%); f: GO terminology enrichment analysis of total protein versus RC-NIRFR-captured proteins; g: Functional annotation of RC-NIRFR-rich proteins: cluster 1 (purple) and cluster 2 (red).

[0057] Figure 2 To differentiate the responsiveness of tumor and normal tissues to differentially expressed cysteine; a, b: RC-NIRFR labeling process and results of isolated lysed breast cancer tumors (T) and adjacent normal tissues (N) from 13 patients; c: Distribution map of RC-NIRFR-labeled reactive cysteine ​​proteome (2,685 proteins) in matched T / N pairs; d: Heatmap of the ratio of tumor to normal protein expression intensity (T / N intensity), with proteins categorized as upregulated (T / N>1.5), non-regulated (0.67≤T / N≤1.5), or downregulated (T / N<0.67); e: Distribution of RC-NIRFR-labeled T-specific proteins; f: Heatmap distribution of RC-NIRFR-labeled upregulated proteins;

[0058] Figure 3 The HARP-C workflow for intraoperative tissue diagnosis and prognostic risk stratification includes: 1. Rapid RC-NIRFR labeling of proteins in tissue sections, followed by in-situ molecular weight-based separation via layer-by-layer 3D tissue electrophoresis (fractions 1 to 8); 2. Quantification of the signal matrix for fractions 1 to 8; summation of the quantized fluorescence signals at corresponding pixel coordinates to obtain location-specific signal intensities; 3. Spatial fluorescence distribution reflects local cysteine ​​reactivity within the proteome, thus distinguishing cancerous and normal tissue regions; 4. The average quantized signal for fractions 1 to 8 within the cancerous region defines the T quantization intensity; normalization produces the T normalized intensity; the AUC of the T normalized intensity for each fraction defines the T scaled intensity (TSI) for each fraction.

[0059] Figure 4To improve the stratification of breast cancer patients by TSI and identify prognostic subgroup detection maps; a: molecular weight distribution of RC-NIRFR marker protein in breast cancer tissue by LC-MS / MS (n=13); b: overview of breast cancer cohorts analyzed by HARP-C: BC1 (postoperative collection with >10-year follow-up information), BC2 (postoperative collection without 10-year results), BC3 (intraoperative collection; 2 / 10 diagnosed as positive cases); c: TSI stratification of 191 sections in cohorts grouped by TSI range (150-250, 250-320, >320); d: scores 1 to 8 in TSI groups. The normalized intensity distribution of T; e: stratified OS of TSI obtained by Kaplan-Meier analysis (P=0.0022); fh: prognostic assessment of TSI for metastasis-free survival (****P<0.0001, f), 5-year DFS (****P<0.0001, g), and 5-year OS (P=0.0015, h); i: significant association between TSI and molecular characteristics: clinical T stage (T1 to T4), N stage (N0 to N3 / NX), histological grade (I to III), intrinsic subtype (luminal / HER2+ / TNBC), and histological subtype (catheter / lobule / other);

[0060] Figure 5 Mapping for HARP-C identification of tumor-positive regions in tissue sections of breast cancer patients; a: Comparison between histological features (H&E report confirmed by pathologist) and active cysteine ​​signal (reported by HARP-C) (Patient ID: R0315); b, c: Quantified signal after HARP-C analysis of tissue sections; d: Protein signal atlas (cancer threshold in c); e: Quantification of the ratio of tumor to adjacent normal tissue (three gel layers per section); ROC analysis showed an AUC of 1.000;

[0061] Figure 6 TSI detection graphs for queues BC1, BC2, and BC3;

[0062] Figure 7Prognostic assessments of DFS and OS using TSI stratification and clinical grading / staging; a: Prognostic assessment of DFS in the BC1 cohort: TSI (****P<0.0001, Pearsonr=-0.4686), T stage (****P<0.0001, Spearmanr=-0.3568), N stage (P=0.0513, Spearmanr=-0.1534), and histological grade (P=0.8243, Spearmanr=-0.01763); b: Prognostic assessment of OS in the BC1 cohort: TSI (***P=0.0001, Pearsonr=-0.2980), T stage (**P=0.0014, Spearmanr=-0.2496). N stage (*P=0.0472, Spearmanr=-0.1562) and histological grade (P=0.3641, Spearmanr=-0.07199); c: Correlation analysis between TSI and T stage (*P=0.0102, Spearmanr=0.2013), N stage (P=0.4945, Spearmanr=0.05405), and histological grade (P=0.1434, Spearmanr=0.1155); df: Analysis of differences in TSI grouped by T stage (d), N stage (e), and histological grade (f); unpaired t-test; g: Correlation analysis between patient age and TSI (P=0.0515, Pearsonr=0.1411).

[0063] Figure 8A diagram illustrating how protein-level stratification differences can be used for prognostic assessment of breast cancer via TSI; a: Cysteine ​​response structures (pink: Cluster#Q8NI60; green: Cluster#P63092; purple: Cluster#P42224) and RC-NIRFR binding sites in typical cancer pathway proteins: AKT1 (Cys344), ERK (Cys81 / Cys177 / Cys182), RAS (Cys80), STAT3 (cysteine ​​108); b: Signal quantification of each protein in each TSI stratum (150–250, 250–320, 300–350, >320); to better describe the contribution of different proteins to TSI stratification, samples with TSI >300 were used as the root... Based on their protein expression characteristics, they were divided into different subgroups (n=3 / group); c: RC-NIRFR labeling efficiency verification using the target recombinant protein at different incubation temperatures (37℃ and 50℃); d: Contribution factor determination: the ratio of fluorescence intensity of RC-NIRFR labeled protein standard (37℃) to RC-NIRFR unlabeled vimentin standard (37℃); e: Average fluorescence intensity of each protein in each TSI stratification (150–250, 250–320, 300–350, >320, n=3 / group b); f: Considering the influence of contribution factors, to reproduce the signal contribution of each protein after RC-NIRFR labeling; g: Overview of the contribution of stratified proteins to TSI classification; h: RC-NIRFR labeled proteins can achieve patient stratification.

[0064] Figure 9 Schematic diagram of the application of HARP-C in the diagnosis of thyroid, lung, and cervical cancer tissues; a: Overview of thyroid, lung, and cervical cancer cohorts analyzed by the Tissue Active Cysteine ​​Analysis Platform (HARP-C); the TC1 cohort consists of thyroid cases collected postoperatively; the TC2 cohort includes 19 thyroid cases collected intraoperatively, of which 14 were diagnosed as positive; the LC cohort includes 25 lung cases collected intraoperatively, of which 13 were diagnosed as positive; the CC cohort consists of cervical cases collected postoperatively; b: Calculated ratio of tumor to adjacent normal tissue (three gel layers per slice); c, d: Performance of HARP-C in the intraoperative diagnosis of thyroid cancer;

[0065] Figure 10 Maps showing the prediction of malignant regions in thyroid, lung, and cervical cancer using HARP-C; ad represents the location of malignant regions predicted by the platform in the TC1, TC2, CC, and LC cohorts, which are consistent with the location of tumor cells in tissue sections confirmed by H&E staining.

[0066] Figure 11A schematic diagram of the affinity purification-mass spectrometry workflow for identification of RC-NIRFR labeled proteins; a: RC-NIRFR selectively forms covalent bonds with proteins containing reactive cysteine ​​in tissue lysates; b: RC-NIRFR-alkyne can target and separate labeled proteins, including: (i) incubation with tissue lysates facilitates selective binding to proteins containing reactive cysteine; (ii) click chemoconjugation with an azide-modified biotin tag; (iii) streptavidin column affinity purification enrichment of the RC-NIRFR-alkyne labeled complex; (iv) bead elution prior to MS identification. Detailed Implementation

[0067] To further illustrate the present invention, a detailed description is provided below with reference to embodiments. However, it should be understood that these descriptions are merely for further illustrating the features and advantages of the present invention, and not for limiting the scope of the claims.

[0068] There are no particular restrictions on the source of any raw materials used in this invention; they can be purchased from the market or prepared using conventional methods known to those skilled in the art.

[0069] Example 1: Preparation of the tumor-selective dye RC-NIRFR-alkyne

[0070] The reaction equation for the synthesis of RC-NIRFR-alkyne is shown below:

[0071]

[0072] The specific steps for synthesizing compound 1 in this embodiment are as follows: 2,3,3-trimethyl-3H-indole (1.50 mL, 9.34 mmol) and anhydrous acetonitrile (10 mL) were added to a sealed reaction vessel, followed by 6-iodo-1-hexyne (1.00 mL, 7.70 mmol). The reaction was carried out under heating for a period of time. After the reaction was completed, the mixture was cooled, and ethyl acetate was added to precipitate the product. The product was then filtered to obtain a light gray solid. The yield was in the hundreds of milligrams range, with a yield of approximately 40%. The characterization data of compound 1 are as follows: 1 H NMR (400MHz, CDCl3) δ 7.76 - 7.72 (m, 1H), 7.65 - 7.54 (m, 3H), 4.70 - 4.62 (m, 2H), 3.15 (s, 3H), 2.25 (td, J = 6.6, 2.7 Hz, 2H), 2.15 - 2.03 (m, 2H), 2.00 (t, J= 2.6 Hz, 1H), 1.85 - 1.71 (m, 2H), 1.65 (s, 6H).

[0073] The specific steps for synthesizing compound 2 in this embodiment are as follows: Compound 1 (400 mg, 1.10 mmol) and 2-chloro-3-(hydroxymethylene)-1-cyclohexene-1-carboxaldehyde (100 mg, 0.54 mmol) were added to a reaction flask, followed by sodium acetate (160 mg, 1.80 mmol) and acetic anhydride (20 mL). The mixture was stirred at approximately 35 °C until the reaction was complete (approximately 3-5 h). After the reaction was complete, the solvent was removed by vacuum distillation. The residue was purified by silica gel column chromatography (using dichloromethane / methanol as the eluent, for example, a volume ratio of approximately 100:1) to obtain a green solid with a yield of approximately 70%. The characterization data of compound 2 are as follows: 1 H NMR(400 MHz, CDCl3) δ 8.32 (d, J = 14.0 Hz, 2H), 7.41 - 7.34 (m, 4H), 7.27 -7.16 (m, 4H), 6.28 (d, J = 14.0 Hz, 2H), 4.25 (t, J = 7.4 Hz, HR-MS Calcd for [M + 615.3501, Found 615.3492.

[0074] Example 2: Proteomic range of cysteine ​​responsiveness assessed using RC-NIRFR

[0075] To verify the selectivity of RC-NIRFR labeling of active cysteine, lysates of human breast tumor tissue were first labeled with RC-NIRFR. The fluorescence intensity of these protein bands reflected the degree to which the probe labeled active cysteine ​​in the proteome. As a control, the tissue lysates were pre-treated with either heat denaturation at 90°C or enzymatic digestion at 37°C. These treatments disrupted the overall protein spatial structure while preserving cysteine ​​residues.

[0076] The results showed that disrupting cysteine ​​reactivity within the proteome in the control group led to a significant decrease in the fluorescence intensity of the protein bands, such as... Figure 1As shown in Figure b, the above results indicate that RC-NIRFR possesses sufficient sensitivity and selectivity for active cysteine. The reactivity of cysteine ​​is governed by the local protein structural environment, and can vary by several orders of magnitude depending on the local protein microenvironment. When the three-dimensional structure containing active cysteine ​​is disrupted, the reactivity of cysteine ​​decreases, leading to a loss of its reactivity.

[0077] Surgical resection of tumors and adjacent normal tissues was performed on 13 breast cancer patients. Following tissue lysis and protein extraction, baseline whole-proteomics maps of two tissue types were established using mass spectrometry. Figure 1 (Figure C) Figure 1 The workflow for estimating the difference in cysteine ​​reactivity between tumor (T) and adjacent normal (N) tissues in breast cancer patients (n=13) includes: (1) lysing and extracting proteins from 13T and 13N tissues; (2) liquid chromatography-mass spectrometry (LC-MS / MS) proteomic analysis to generate 26 proteomes (13T+13N); (3) comprehensive analysis to define T-specific proteins (tumor-specific), N-specific proteins (normal tissue-specific), and non-specific proteins (shared by both); (4) RC-NIRFR labeling and affinity purification of probe-binding proteins in all 26 lysates; (5) LC-MS / MS analysis of captured proteins; and (6) comparative quantification of T-specific, N-specific, and non-specific RC-NIRFR binding proteins.

[0078] Subsequently, the lysates were labeled with RC-NIRFR and subjected to mass spectrometry analysis to identify proteins containing RC-NIRFR-modified active cysteine ​​residues. To enhance the affinity enrichment of the labeled proteins for mass spectrometry analysis, RC-NIRFR was functionalized with an alkyne group (prepared in Example 1). Proteomics data from all patients were integrated, and proteins were classified as: T-specific (detected only in tumor tissues), non-specific (detected in both tissue types), and N-specific (detected only in normal tissues). RC-NIRFR exhibited a clear labeling preference in these categories due to its enhanced cysteine ​​reactivity (…). Figure 1 (Figure d) shows that >80% of T-specific proteins were recognized. Figure 1 (See Figure e). Conversely, only 27% of non-specific proteins and 20% of N-specific proteins contain cysteine, exhibiting considerably high reactivity. Figure 1 (Figure e in the middle). These quantitative results indicate that the cysteine ​​reactivity in the tumor proteome is significantly enhanced compared to normal tissue, thus making this characteristic a competitive tissue diagnostic strategy.

[0079] Gene ontology (GO) analysis was performed on the proteome before and after RC-NIRFR labeling, with particular attention to functional enrichment associated with the RC-NIRFR labeling process. Figure 1 (Figures f and g). Analysis showed that RC-NIRFR-targeted active cysteine-containing proteins bind to two functionally distinct clusters, which are highly correlated with cancer progression: Cluster 1 (purple, ... Figure 1 The features of cluster 2 (in red) are overexpression of proteasome complex proteins, glycolytic enzymes, and components of the pentose phosphate pathway, while cluster 2 (in red) is characterized by overexpression of proteasome complex proteins, glycolytic enzymes, and components of the pentose phosphate pathway. Figure 1 The findings (Figure g) reveal features of innate immune activation, including complement system overexpression, involvement of eosinophils / neutrophils / mast cells / macrophages, and hypoxia. These findings establish whole-proteomic cysteine ​​reactivity assessment as a powerful tool for elucidating the intrinsic molecular characterization of malignant tissues.

[0080] Example 3: Structural motif of reactive cysteine ​​in RC-NIRFR-tagged protein

[0081] Proteins were divided into RC-NIRFR labeled and unlabeled groups, and structural clustering was performed using Foldseek. Clusters were defined by an E-value < 0.01 and a structural alignment overlap > 30%, and singletons were excluded to obtain representative clusters containing at least two proteins. This procedure identified 276 structural clusters from RC-NIRFR labeled proteins and 356 from unlabeled proteins. Each cluster was named after the UniProt ID of the protein with the highest structural overlap with other members of that cluster.

[0082] Structural characterization was prioritized for clusters derived from RC-NIRFR-tagged proteins to identify motifs characterized by reactive cysteine ​​(named cysteine ​​reactive motifs). To achieve this, conserved structural regions were extracted from the 30 largest RC-NIRFR-tagged clusters. Cluster #P61769 exemplifies this strategy: 83 proteins in this cluster exhibit local conformations similar to the Ig-like domain (residues 25-113) of β-2-microglobulin (P61769). Within this domain, Cys45 and Cys100 show high reactivity to RC-NIRFR via nucleophilic substitution, indicating that the Ig-like domain constitutes a cysteine ​​reactive motif. Subsequent analysis of structurally complex members within cluster #P61769 confirmed that this domain stably confers cysteine ​​reactivity in various protein environments. This conclusive evidence establishes the Ig-like domain as a recognizable cysteine ​​reactive structural motif.

[0083] The diverse functional properties of proteins containing reactive cysteine ​​residues constitute a key area of ​​research. In cluster #P61769, both P61769 (β-2-microglobulin) and P01023 (α-2-macroglobulin) play crucial immunomodulatory roles, and their overexpression is associated with enhanced tumor progression and metastatic potential. Q9R069 (basal cell adhesion molecule) and P43121 (cell surface glycoprotein MUC18) enhance tumor cell migration, invasion, and epithelial-mesenchymal transition (EMT) by activating the PI3K / AKT and MAPK / ERK pathways; their high expression predicts lymph node metastasis and adverse clinical outcomes. Q5ZPR3 (CD276 antigen) promotes tumor immune escape by transmitting inhibitory signals to T cells; its overexpression is independently associated with shortened disease-free survival and overall survival in breast, lung, and colorectal cancers. Overall, these findings link reactive cysteine ​​to molecular recognition and signal transduction functions, and also explain to some extent the RC-NIRFR's labeling preference for tumor-specific proteins.

[0084] Example 4: Clinical identification based on differences in cysteine ​​activity between tissues

[0085] Construct individualized quantitative reactivity profiles of cysteine-reactive proteins across 13 matched tumor-normal tissues from 13 breast cancer patients. Figure 2 (Figure c in the middle) Subsequently, the tumor to normal tissue expression ratios of all analyzed proteins in each patient cohort were calculated ( Figure 2 (Figure d). Proteins were stratified into: upregulated (ratio > 1.5), unregulated (0.67 ≤ ratio ≤ 1.5), or downregulated (ratio < 0.67). Upregulated and unregulated proteins were consistently more abundant than their corresponding downregulated proteins in patients. Figure 2 (middle d, top). This unique map shows the inter-individual heterogeneity of expression of RC-NIRFR-tagged T-specific proteins ( Figure 2 (See Figure e). Co-expressed reactive cysteine ​​proteins accounted for 42% of the total reactive cysteine ​​proteome (n = 1,118 / 2,685). Figure 2 (See Figure f). Crucially, 357 proteins (representing 13% of the total) were universally upregulated in tumors across all patients, providing broad cohort diagnostic utility. Figure 2 (See Figure f). Another 467 proteins (17%) showed tumor upregulation in at least one patient, of which 317 (12%) were upregulated in more than 50% of patients (n ≥ 7). Figure 2 Consistent upregulation was observed in the middle (Figure f). Overall, 69% of the reactive cysteine ​​proteomes detected by RC-NIRFR generally enabled NIR fluorescence-based tumor identification, while 81% supported identification in ≥50% of patients.

[0086] Example 5

[0087] This example demonstrates the application of HARP-C for intraoperative tissue diagnosis and prognostic assessment.

[0088] The specific method utilizes tumor-selective dyes RC-NIRFR or RC-NIRFR-alkyne, containing halogen substituents (such as C-Cl) and alkynyl groups, to specifically nucleophilically substitute the thiol groups of tumor-characteristic proteins in tissue sections. This generates stable biomimetic near-infrared fluorescent proteins (NIR-FPs) in situ within the tissue. Combined with an NIR-FPs histological separation and detection system, spatial separation and rapid imaging of the proteins are achieved, thereby aiding in the grading of tumor regions. The steps for applying this method intraoperatively are attached. Figure 3 As shown, the details are as follows:

[0089] 1) Tissue preparation and in-situ marking:

[0090] Fresh breast cancer tissue removed during surgery was rapidly frozen at -80°C for approximately 5 minutes and then cryo-embedded. Frozen sections with a thickness of 60 micrometers were prepared using a cryostat (applicable range: 5–100 micrometers). The tissue sections were then transferred to a custom-made 20×20 microwell array mounting plate (prepared with transparent photocrosslinked resin) and smoothly attached.

[0091] Subsequently, RC-NIRFR or RC-NIRFR-alkyne dye solutions were prepared and mixed with RIPA lysis buffer (5 mL per slice). The mixtures were then incubated in culture dishes at room temperature to 60 °C to simultaneously complete protein lysis and dye labeling. The halogen substituents (-Cl) in the dye covalently bind to the protein thiol groups via nucleophilic substitution reactions, forming stable NIR-FPs.

[0092] 2) Histological separation and molecular weight stratification:

[0093] In situ separation was performed using 3D tissue electrophoresis. Labeled tissue sections, along with a sample plate, were placed in an uncrosslinked stacking gel and fused with a crosslinked 10% separation gel (polyacrylamide) below to form a complete three-dimensional gel structure. This structure was then placed in an NIR-FPs histological separation and detection system. The electrophoresis process was performed under constant voltage, and proteins migrated along the z-axis molecular weight direction, generating spatial cysteine ​​reactivity maps between molecular weight fractions (F1–F8).

[0094] 3) Quantitative fluorescence analysis:

[0095] After tumor tissue sections are labeled with dye, the labeled tumor proteins with active cysteine ​​are longitudinally separated according to molecular weight and artificially divided into 1 to 8 layers (F1-F8). Each layer represents a tumor protein group with a certain molecular weight range. The fluorescence intensity represents the cysteine ​​reactivity. Position-resolved fluorescence signals in the same layer can distinguish between benign and malignant tumors.

[0096] 4) Malignant region identification: Tumor-positive regions are delineated based on reactive cysteine ​​characteristic topology maps (fluorescence intensity);

[0097] 5) The following steps are used to derive the prognostic index of TSI: (i) The signal of each layer corresponding to the tumor region determined in the previous step is used as the quantitative intensity signal of T; (ii) The average T fluorescence intensity of the eight layers F1–F8 is normalized with the highest signal layer to obtain the normalized T intensity and the number of layers to construct a component-specific risk curve. The area under the curve (AUC) is the normalized T intensity (T scaled intensity, TSI). The TSI obtained in this way can be compared across cases. The higher the TSI, the more malignant the tumor, which can be used as a new quantitative index for risk stratification of breast cancer.

[0098] A total of 191 slides from 79 breast cancer patients were analyzed (for each patient, n ranged from 1 to 5 depending on the size of the lesion). Figure 4 A BC1 cohort (n = 162) with over 10 years of follow-up data was used to evaluate the prognostic performance of the platform. Figure 4 (Figure b). To expand the patient cohort to analyze the correlation between TSI and clinical grade / stage provided by the platform, the BC2 cohort (n = 27) was supplemented. While lacking 10-year survival outcome data, this cohort had postoperative pathology reports including histological grade, T stage, N stage, and age information. Additionally, the BC3 cohort (n = 2) was included in the operating room application. Intraoperative analysis was performed on 10 breast tumor cases, 8 of which were determined to be benign by HARP-C (consistent with the final postoperative histological diagnosis) and excluded from risk assessment. HARP-C analysis showed a consistency between the distribution of RC-NIRFR marker proteins in breast tumor tissue and the mass spectrometry of reactive cysteine-containing proteins (…). Figure 4 Figures a and c in the middle section verify the effectiveness of HARP-C in cysteine ​​analysis.

[0099] HARP-C successfully identified cancer regions in all cases across the BC1, BC2, and BC3 cohorts, and RC-NIRFR signal maps co-localized with cancer cells in hematoxylin-eosin (H&E) stained sections. Cancer regions showed significantly higher signal intensity compared to glandular or stromal regions (P < 0.0001; cancer threshold > 54000). Figure 5The ad plot (in the middle section) confirmed the tumor specificity of the reactive cysteine ​​proteome signal. Based on this signal, HARP-C achieved excellent diagnostic performance in tumor detection [Receiver operating characteristic (ROC) analysis, AUC = 1.000; Figure 5 [Figure e]. Furthermore, one case of BC3 that was intraoperatively diagnosed as malignant by the pathologist (but ultimately pathologically and HARP-C diagnosed it as benign) was excluded. This discrepancy may stem from artifacts inherent in frozen sections, such as distorted cell morphology, tissue loss due to ice crystal formation, and section folding / tearing artifacts. These artifacts can cause benign cells to present atypically, potentially leading to false-positive diagnoses. These findings support HARP-C as an intraoperative diagnostic aid. By analyzing signals containing reactive cysteine ​​proteins, HARP-C improves diagnostic accuracy beyond current practices that rely solely on histological features.

[0100] Besides accurate tumor boundary detection, HARP-C's most important capability is its intraoperative prognostic assessment. According to Figure 3 The prognostic strategies outlined in the text are based on their TSI (range: 166.5–522.5). Figure 6 and Figure 4 (See the CD chart) All cases were divided into three risk groups (150–250, 250–320, >320). Survival analysis of the BC1 cohort showed significant prognostic differences among the three TSI strata (P = 0.0022). Figure 4 (See figure). Compared to clinical grading / staging systems, TSI provides a more accurate prognostic assessment (…). Figure 7 (See figure ab). Correlation analysis confirmed that in the pooled BC1, BC2, and BC3 cohorts, TSI was not associated with histological grade (P = 0.1434), N stage (P = 0.4945), or patient age (P = 0.0515), while revealing a partial association with T stage (P = 0.0102). Figure 7 (middle CG figure). This T stage correlation may reflect protein upregulation accompanying tumor progression. To further elucidate the clinical relevance of TSI, the BC1 cohort stratified by metastatic status, five-year disease-free survival (DFS), and five-year overall survival (OS) was analyzed. Figure 4 (See figure). TSI was significantly elevated in patients with metastatic disease (P < 0.0001), regardless of the time of metastasis (postoperative > 5 years). Similarly, higher TSI values ​​were associated with reduced five-year disease-free survival (DFS) and overall survival (OS).

[0101] In breast cancer, histological classification (based on microscopic structural features) categorizes tumors into ductal, lobular, or other types as a preliminary diagnostic method. In contrast, molecular subtyping [luminal, HER2-positive (HER2+), and triple-negative breast cancer (TNBC)] integrates multiple molecular features, including gene expression profiles, protein expression levels, and signaling pathway activation. This approach more comprehensively reflects tumor biology, intrinsic molecular mechanisms, and has a stronger prognostic correlation. This invention evaluated the association between TSI and these two classification systems. TSI showed a significant correlation with molecular subtypes (P < 0.0001), but no correlation with histological classification (…). Figure 4 (Figure i in the middle). This finding confirms that the reactive cysteine ​​proteomic profile captured by TSI can effectively reflect the biological basis of tumors, including growth, proliferation, and metastatic potential, thereby enabling precise prediction of clinical behavior and prognosis. Crucially, HARP-C can provide this molecular-level insight intraoperatively (<45 minutes), offering real-time guidance for surgical decisions and making it a promising adjunctive tool.

[0102] Example 6: HARP-C for intratumoral heterogeneity analysis

[0103] Tumor heterogeneity is a recognized factor contributing to adverse clinical outcomes. Therefore, the risk prediction from multiple biopsies / frozen sections in the same patient can vary significantly. To investigate the association between intratumoral heterogeneity and prognosis, all patients with multiple tumor frozen sections were analyzed. Given that TSI strongly predicts potential metastatic events, patients were further stratified by metastatic status. The standard deviation of TSI in the metastatic group was significantly higher than that in the non-metastatic group, indicating a higher degree of intratumoral heterogeneity. This finding is consistent with clinical observations. These results highlight the ability of HARP-C to assess intratumoral heterogeneity. Traditional clinical methods lack the throughput required for broad-spectrum proteomic analysis of tumor markers intraoperatively to obtain diagnostic conclusions. This invention provides HARP-C as a solution for prognostic assessment through intratumoral heterogeneity evaluation.

[0104] Example 7: Differential expression and regulatory role of key cysteine-responsive proteins in breast cancer pathways across TSI stratification

[0105] Breast cancer molecular subtypes are closely related to the activation state of cellular signaling pathways. Each subtype exhibits a characteristic pattern of pathway activation, in which aberrant signaling plays a crucial role in tumor initiation, progression, invasion, metastasis, and treatment response. Given the significant correlation between TSI and breast cancer molecular subtypes ( Figure 4(Figure i in the original text) This invention investigated the differential expression of key cysteine-reactive proteins in core breast cancer pathways across TSI stratification. To avoid over-reliance on structural clustering and pathway deconvolution analysis, this embodiment focuses on illustrating the differences in protein expression corresponding to TSI stratification and their contribution to prognostic assessment. After labeling relevant lysates with RC-NIRFR-alkyne, the labeled proteins were affinity purified and identified by mass spectrometry, thereby obtaining key protein information related to TSI stratification. The results show that different TSI strata correspond to different protein expression patterns, suggesting that TSI not only reflects the overall cysteine ​​reactivity level but can also characterize the biological differences in breast cancer at the protein level.

[0106] To elucidate the combined expression of these proteins across TSI strata, we analyzed patient tumor samples with TSI values ​​ranging from 150–250, 250–320, and >320. Inclusion criteria for each stratum were that the maximum TSI value in the patient slices fell within the specified range. Overall, the target proteins were overexpressed in at least one stratum in each stratum, while being virtually undetectable in adjacent normal tissues. Figure 8 (See Figure b). As TSI increases, the total expression levels of these proteins also increase accordingly. Figure 8 (Figure b). In patients with TSI of 150–250, only weak RAS upregulation was observed, suggesting low-level activation of the RAS / RAF / MAPK pathway, which may indicate relatively slow tumor growth. A unique protein expression pattern emerged in patients with higher TSI (>300). To better elucidate the relationship between protein levels and TSI stratification, patients were divided into three expression-based groups (250–320, 300–350, and >320). Figure 8 (Figure b). In the TSI 250–320 group, mild ERK upregulation was accompanied by elevated RAS, indicating progressive activation of the RAS / RAF / MAPK pathway. Protein kinases, including RAS and ERK, promote cell proliferation, differentiation, and survival. In the TSI 300–350 group, AKT1 became the dominant protein, suggesting high activation of PI3K / AKT signaling. Overactivated PI3K / AKT pathway enhances tumor cell proliferation, survival, and metastasis, and is associated with tumor invasiveness and poor prognosis. In patients with TSI >320, STAT3, AKT1, and RAS were simultaneously overexpressed. Co-activation of PI3K / AKT and JAK / STAT is a molecular marker of TNBC, a subtype associated with poor prognosis. These findings confirm the oncological significance of cysteine ​​reactivity in proteomic prognostic assessment.

[0107] To deepen our understanding of TSI regulation by RC-NIRFR-tagged proteins, a protein contribution factor was defined—the ratio of fluorescence intensity between RC-NIRFR-tagged protein standards and Vimentin (an untagged RC-NIRFR protein) and a mixture of RC-NIRFR proteins (all systems were incubated at 37°C). Figure 8 (See the CD diagram). These contributing factors represent the relative weight of individual protein levels in the total TSI calculation. Proteins with larger contributing factors have stronger TSI regulatory capabilities, and their upregulation causes a significant increase in TSI. It is worth noting that, without considering their respective "weights," neither single predictors nor multivariate combinations can effectively distinguish patient outcomes. Figure 8 (See figure). In summary, comprehensive stratification of breast cancer patients requires integrating the expression levels of predictive factors and their cysteine ​​content quantified by RC-NIRFR markers.

[0108] To evaluate the diagnostic potential of the aforementioned histological protein analysis platform, it was applied to three clinical tasks: identifying thyroid cancer, lung cancer, and cervical cancer in relevant specimens. Figure 9 Thyroid specimens were divided into two subgroups: malignant samples confirmed by postoperative histopathology and archived in the pathology database (TC1 cohort); and specimens collected intraoperatively (TC2 cohort). Lung specimens were all collected intraoperatively (LC cohort), while cervical specimens were diagnosed postoperatively and derived from the pathology database (CC cohort). Consistent with the breast cancer protocol, benign samples were excluded during intraoperative collection of thyroid and lung specimens in the operating room. Notably, one thyroid specimen (932366E-1) was... Figure 9 (Figure c) Initially diagnosed as malignant during intraoperative frozen section, it was subsequently confirmed as benign in the final histopathology. The detection platform provided by this invention also correctly identified it as benign intraoperatively. Furthermore, HARP-C showed 100% concordance with the final histopathological diagnosis in both intraoperative and postoperative assessments (116 malignant cases, 17 benign cases). Figure 9 In all four cohorts, the protein signal was precisely localized to malignant regions within the tissue sections. Figure 10 These results confirm the diagnostic utility of HARP-C in a variety of oncology applications.

[0109] The invention establishes cysteine ​​reactivity as a novel molecular marker, a potential tool for histological diagnosis and prognostic assessment. The RC-NIRFR probe can form covalent bonds with proteins containing reactive cysteine ​​via a nucleophilic substitution reaction, enabling proteomic assessment of the reactive cysteine ​​range. In characterizing RC-NIRFR-bound human breast cancer histones, significantly higher cysteine ​​reactivity was observed in tumor tissues than in normal tissues. Compared to the native proteome, these RC-NIRFR-labeled enriched cysteine-reactive proteins constitute two functional clusters driving cancer progression. Although the role of hyperreactive cysteine ​​in cellular processes is not fully understood, the results demonstrate its involvement in regulating metabolic reprogramming (functional cluster 1) and innate immune / complement activation (functional cluster 2) during breast cancer development. Figure 1 (See Figure g). This further confirms that assessing tissue proteome cysteine ​​reactivity represents a crucial, yet undeveloped, diagnostic biomarker.

[0110] Application Examples

[0111] Protein binding site analysis:

[0112] First, purified protein or protein@dye samples were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Then, the corresponding gel bands were excised and digested with trypsin and chymotrypsin. The processed samples were analyzed by liquid chromatography-tandem mass spectrometry (LC-MS / MS). Further analysis and comparison of the data were performed using software such as Byonic to obtain identification results.

[0113] The specific operation is as described in Example 5.

[0114] Clinical information and sample collection:

[0115] The collection process for breast cancer cases is illustrated below. This section includes postoperative specimens from breast cancer patients who underwent surgery at least ten years prior (n = 189). Clinical information such as histological grade, T stage, N stage, and patient age were also collected and analyzed from these patients' pathology reports. Intraoperative samples were collected during the surgery and tested within 30–60 minutes (n = 10). Surgically removed tissue was frozen sectioned into 20-micron-thick tissue sections and immediately analyzed using a HARP-C three-dimensional tissue electrophoresis apparatus.

[0116] Clinical information analysis:

[0117] Patients in this study were followed up annually. Overall survival (OS) was defined as the period from the start of follow-up to the date of death from any cause and / or the last follow-up date. Disease-free survival (DFS) was defined as the period from the start of follow-up to the date of local recurrence, distant metastasis, death from any cause, and / or the last follow-up date. Recurrence and metastasis were diagnosed by imaging or pathological examination. Survival functions were stratified by histological grade, T stage, and N stage, and calculated using the Kaplan-Meier method. The endpoint of the survival trial was artificially set at 120 months.

[0118] 3D tissue electrophoresis apparatus:

[0119] The 3D tissue electrophoresis apparatus and 20×20 micropore array mold were designed and fabricated using 3D modeling software (AutoCAD 2020). The micropores, with a diameter of 750 μm, were arranged in an equidistant pattern within a 2 cm × 2 cm area. The apparatus and micropore array mold were printed using a 3D printer (Anycubic Photon M3 Premium) with photocrosslinking resin (3D printing photosensitive resin, HD gray-white, Anycubic; photopolymer standard resin, pure gray, SUNLU) for a 3.5-second exposure time. The printed apparatus and mold were then thoroughly ultrasonically cleaned twice in 99% ethanol, for 3 minutes each time.

[0120] HARP-C analysis of clinical tissue sections:

[0121] Tissue samples were embedded in optimal cutting temperature (OCT) embedding medium (Leica, 39475237) and frozen to a thickness of 20 μm. Proteins in the frozen sections were labeled and separated using the aforementioned method. Then, bands 1-8 containing the target protein signal were imaged in the gel at 800 nm using a scanner (Azure, Sapphire), with fluorescence intensity and pixel size set to 100 µm. Quantitative signal analysis and prognostic assessment of intraoperatively acquired cases were automated using a risk assessment program.

[0122] Western blot:

[0123] The antibodies used include: recombinant anti-GAPDH mouse monoclonal antibody (Servicebio, GB15002), anti-AKT1 rabbit polyclonal antibody (Servicebio, GB111114), anti-IL-6 rabbit polyclonal antibody (Servicebio, GB11117), anti-ERK1+ERK2 mouse monoclonal antibody (Servicebio, GB12087), recombinant anti-PDK1 rabbit monoclonal antibody (abcam, ab202465), recombinant anti-RAS rabbit monoclonal antibody (abcam, ab52939), anti-STAT3 rabbit polyclonal antibody (Servicebio, GB11176), anti-VEGFA rabbit polyclonal antibody (Servicebio, GB11034B), anti-Vimentin rabbit polyclonal antibody (Servicebio, GB11192), and HRP-labeled goat anti-mouse IgG (H+L) (Servicebio, GB23301).

[0124] Tissue was cut into sufficiently small fragments and lysed with an appropriate amount of radioimmunoprecipitation assay (RIPA) lysis buffer (Yeasen, 20101ES60). To promote complete lysis, the tissue was sonicated at 80% power density for 15 minutes at 4°C using a non-contact sonicator. The lysed tissue was centrifuged at 15000 g for 20 minutes to completely remove the precipitate. Proteins in the lysis buffer were separated by SDS-PAGE and transferred to a polyvinylidene fluoride (PVDF) membrane (Servicebio, G6015-0.45) under electrophoresis at 300 mA for 30 minutes. Immediately after transfer, the membrane was rinsed once quickly in an incubator containing TBST buffer (Servicebio, G0001-2L), and then blocked with skim milk powder (Servicebio, G5002) on a shaker at room temperature for 30 minutes. After discarding the blocking solution, diluted primary antibody (diluted according to the antibody instructions) was added, and the membrane was incubated overnight on a shaker at 4°C. After washing with TBST, the membrane was incubated overnight on a shaker at 4°C with diluted secondary antibody (1:5000). The membrane was washed again with TBST, and finally prepared by pelleting, development, and fixing with a developing and fixing reagent (Servicebio, G2019). Western blot images were acquired using a chemiluminescence imaging system (CLINX, 6100).

[0125] H&E staining:

[0126] All postoperative tissue sections (thickness: 15 µm) were stored at -20°C before staining. After fixation in 4% paraformaldehyde, sections were stained with H&E according to the manufacturer's instructions for the H&E kit (Beyotime Biotechnology, catalog number C0105). H&E staining images of the tissues were acquired using an upright microscope (Nikon Eclipse 80i). Intraoperative H&E staining images of tissue sections (thickness: 4 µm) were provided by the Department of Pathology, First Hospital of Jilin University.

[0127] Immunofluorescence staining:

[0128] The antibodies used include: anti-AKT1 rabbit polyclonal antibody (Servicebio, GB111114), anti-ERK rabbit polyclonal antibody (Servicebio, GB112238), recombinant anti-RAS rabbit monoclonal antibody (abcam, ab52939), anti-STAT3 rabbit polyclonal antibody (Servicebio, GB11176), anti-Vimentin rabbit polyclonal antibody (Servicebio, GB111308), and Cy3-labeled goat anti-rabbit IgG (H+L) (Servicebio, GB21303).

[0129] Frozen tissue sections were placed in a 37°C oven for 20 minutes. Subsequently, the sections were fixed in fixative (Servicebio, G1101) for 30 minutes and washed three times (5 minutes each time) with PBS (pH = 7.4) on a destaining shaker. The sections were then incubated with 3% BSA (dissolved in PBS) for 30 minutes. The tissue sections were then placed in a humidified chamber and incubated overnight at 4°C with primary antibody solution. The sections were washed three times (5 minutes each time) with PBS (pH = 7.4) on a destaining shaker. After washing, the sections were incubated with secondary antibody solution at room temperature for 50 minutes. This process should be performed in the dark. After washing again, 4',6-diamidinyl-2-phenylindole (DAPI) staining reagent (Servicebio, G1012) was added for nuclear staining at room temperature for 10 minutes. After staining, the tissue sections were imaged using a scanner (3DHISTECH, Pannoramic MIDI).

[0130] RC-NIRFR labeled protein standards:

[0131] Taking STAT3 labeling as an example, STAT3 (5 µM, dissolved in phosphate-buffered saline PBS, Meilun Biotechnology, MA0015) was transferred into an Eppendorf tube, and RC-NIRFR (2 mM, dissolved in dimethyl sulfoxide DMSO) was added, ensuring a 1:1 molar ratio of RC-NIRFR to STAT3. The solution was then incubated in a shaker (at 37°C and 50°C, respectively) for 2 hours. AKT1, ERK, RAS, and Vimentin were labeled with RC-NIRFR using the same protocol (at 37°C and 50°C, respectively). The effectiveness of RC-NIRFR in labeling the target proteins was verified by SDS-PAGE (12.5%). The gel after electrophoresis was scanned using a scanner (Azure, Sapphire) at 800 nm, with specific fluorescence intensities and a pixel size of 100 µm. The fluorescence signal was quantitatively analyzed using ImageJ software.

[0132] In summary, RC-NIRFR can detect reactive cysteine ​​signature proteins overexpressed in tumor tissues, laying the foundation for tumor identification. HARP-C was developed by integrating RC-NIRFR with 3D tissue electrophoresis. Through spatial mapping of proteomic cysteine ​​activity, HARP-C rapidly (within 45 minutes) and accurately (100% specificity, 100% sensitivity) diagnosed tumor-positive areas in clinical breast, thyroid, cervical, and lung cancer surgical specimens, demonstrating the great potential of proteomic cysteine ​​activity in tumor tissue diagnosis.

[0133] Intraoperative identification of high-risk patients remains a critical unmet need in cancer management. HARP-C addresses this issue by rapidly detecting reactive cysteine ​​signature proteins and standardizing signal analysis, enabling personalized intraoperative prognostic assessment. In breast cancer, TSI, a prognostic indicator reflecting tissue cysteine ​​reactivity, significantly optimizes patient stratification. Importantly, this invention discloses a strong correlation between TSI and breast cancer molecular subtypes (P < 0.0001), highlighting that highly reactive cysteine ​​can serve as a reliable prognostic factor, playing a key regulatory role in cancer progression. The improved risk stratification provided by HARP-C will guide more informed perioperative clinical decisions, filling molecular information gaps during the critical postoperative window. This will facilitate closer follow-up and examination of patients, including providing previously unnecessary examinations (such as breast cancer bone scans) for low-risk patients. It will also further promote consideration of intensive therapy or clinical trial enrollment, potentially improving prognosis. Future expansion of the patient cohort will further validate the prognostic predictive capabilities of HARP-C.

[0134] The above description of the embodiments is only for the purpose of helping to understand the method and core ideas of the present invention. It should be noted that those skilled in the art can make several improvements and modifications to the present invention without departing from the principles of the present invention, and these improvements and modifications also fall within the protection scope of the claims of the present invention.

Claims

1. A tumor grading detection system based on whole-proteome cysteine ​​reactivity analysis, characterized in that, include: In situ labeling module: Tumor-selective dyes containing halogen substituents and / or alkynyl groups are used to react in situ with tumor characteristic proteins containing active cysteine ​​in tissue sections to specifically label the tumor characteristic proteins; Histological separation module: Separates labeled tumor proteins by molecular weight into layers and counts the fluorescence signal intensity of each layer; Processing and Analysis Module: By statistically summing the fluorescence signal intensities of each layer, the reactivity of cysteine ​​is reflected, and a proteomics-level cysteine ​​reactivity map with position resolution is constructed. Tumor grading module: The average fluorescence intensity of each layer and the highest signal layer are normalized to obtain the normalized intensity T. The component-specific risk curve is plotted with the layer number as the abscissa. The area under the curve is the normalized intensity T, denoted as TSI. TSI is positively correlated with the degree of tumor malignancy and is used as the basis for tumor grading.

2. The tumor grading detection system based on whole-proteome cysteine ​​reactivity analysis according to claim 1, characterized in that, The tumor-selective dye is selected from RC-NIRFR or RC-NIRFR-alkyne: RC-NIRFR; RC-NIRFR- alkyne。 3. The tumor grading detection system based on whole-proteome cysteine ​​reactivity analysis according to claim 1, characterized in that, The tissue section was a section of cancerous tissue; The cancer mentioned is one or more of the following: breast cancer, cervical cancer, lung cancer, bile duct cancer, and thyroid cancer.

4. The tumor grading detection system based on whole-proteome cysteine ​​reactivity analysis according to claim 1, characterized in that, The in-situ marking module includes the following steps: S1) Frozen embedding of cancerous tissue; S2) Cut the frozen-embedded tissue into tissue sections; S3) The tissue slices are attached to the surface of the microporous array sample plate; S4) Protein lysis and labeling were performed by co-incubating cell tissue lysate with dye.

5. The tumor grading detection system based on whole-proteome cysteine ​​reactivity analysis according to claim 4, characterized in that, The micropore array sample plate is made of transparent material and contains 10 to 100 micropores per square centimeter.

6. The tumor grading detection system based on whole-proteome cysteine ​​reactivity analysis according to claim 4, characterized in that, The cell tissue lysis buffer was RIPA lysis buffer.

7. The tumor grading detection system based on whole-proteome cysteine ​​reactivity analysis according to claim 1, characterized in that, The histological separation module includes: a cathode plate, a microporous array template, a separation gel medium, an anode plate, an electrophoresis tank, and a constant voltage power supply.

8. The tumor grading detection system based on whole-proteome cysteine ​​reactivity analysis according to claim 1, characterized in that, The histological separation module specifically involves performing sodium dodecyl sulfate-polyacrylamide gel electrophoresis on the labeled proteins.

9. The tumor grading detection system based on whole-proteome cysteine ​​reactivity analysis according to claim 1, characterized in that, The processing and analysis module specifically includes: After electrophoresis, the corresponding gel bands were cut and digested with trypsin and chymotrypsin. The processed samples were then analyzed by liquid chromatography-tandem mass spectrometry.

10. The tumor grading detection system based on whole-proteome cysteine ​​reactivity analysis according to claim 1, characterized in that, The tumor grading module specifically includes: (i) Quantitative T-intensity measurement of the specified tumor-positive electrophoretic components; (ii) Normalize the intensity of each layer to T-normalized intensity to enable cross-case comparison; (iii) Construct component-specific risk curves from the normalized intensity of T; (iv) The area under the risk curve is used as the final TSI value for tumor grading.

Images