Use of reagents for detecting protein markers in the preparation of tumor diagnostic products

Abstract

The application belongs to the technical field of biological medicine, and particularly relates to a tumor cell secreted blood protein marker for tumor screening and application thereof. The marker combination provided by the application has excellent diagnostic efficiency in tumor screening. It has been verified that the marker combination can achieve a higher degree of differentiation in total tumor patients and early stage patients, and the sensitivity and specificity thereof are superior to those of traditional protein markers. Meanwhile, the marker combination has stable classification performance in breast cancer, liver cancer, lung cancer and colorectal cancer, indicating that the marker combination has the potential for cross-cancer application and is suitable for general cancer screening scenarios. The detection sample is peripheral blood, which is convenient to take, has small trauma, is convenient for repeated detection and dynamic follow-up, and has strong operability and popularization value in large-scale population screening and clinical auxiliary diagnosis.

Description

Technical Field

[0001] This invention belongs to the field of biomedical technology, specifically relating to the application of reagents for detecting protein biomarkers in the preparation of tumor diagnostic products. Background Technology

[0002] Cancer has become a major global public health burden. Early detection and diagnosis of cancer are crucial for reducing mortality and improving survival outcomes. Currently, the clinical diagnosis of malignant tumors still relies on imaging examinations such as computed tomography (CT), magnetic resonance imaging (MRI), and tissue biopsy. Although imaging examinations offer high spatial resolution, their detection rate for early-stage, small tumor lesions is limited, and they involve issues such as radiation exposure or high costs. Histopathological biopsy is the gold standard for tumor diagnosis, but as an invasive procedure, it carries the potential risk of tissue damage and is not convenient for continuous monitoring of disease progression. Due to the difficulty in early detection of tumors, it is challenging to accurately identify tumors in the early stages of the disease using tissue biopsy.

[0003] Blood biomarker testing offers advantages such as being minimally invasive, readily available, and repeatable, and has broad application prospects. Among different types of blood biomarkers, protein biomarkers have mature detection methods, relatively low cost, and a high degree of standardization, and are currently the most widely used type of blood biomarker in clinical practice. Traditional protein-based tumor biomarkers have been widely used in clinical tumor screening and diagnosis, but their sensitivity and specificity are limited. For example, alpha-fetoprotein (AFP) has a sensitivity of less than 50% for early screening of primary liver cancer. Although CA125 can be used for ovarian monitoring, it may also be abnormally elevated during menstruation, in cases of inflammation, and in benign ovarian diseases, leading to a high false-positive rate. Therefore, there is an urgent need to develop new, high-precision tumor-specific blood biomarkers.

[0004] The occurrence and progression of tumors is a multi-stage, multi-factor driven biological evolutionary process. Normal cells undergo a series of molecular and biological behavioral changes and ultimately transform into malignant tumor cells. Although different cancer types have different origins and tissue backgrounds, epithelial-derived tumors exhibit certain commonalities during their evolution, such as persistent activation of inflammatory signals, enhanced de novo nucleotide synthesis, reversal of intracellular and extracellular pH gradients, and significantly increased cell proliferation capacity. Studies have shown that tumor cells secrete functional proteins in their early stages. These secreted proteins can stably enter the bloodstream, becoming detectable signals earlier than circulating tumor DNA (ctDNA) and circulating tumor cells (CTCs). Therefore, screening for stable and detectable tumor-secreted proteins that cross cancer types based on common characteristics of tumor evolution holds promise for improving the accuracy and clinical feasibility of blood biomarkers for tumor screening and diagnosis. Summary of the Invention

[0005] This invention addresses the shortcomings of existing tumor early screening methods in terms of sensitivity and specificity by proposing a pan-cancer detection molecular biomarker based on tumor-secreted proteins and its applications.

[0006] To achieve the above-mentioned objectives of this invention, the technical solution adopted by this invention is as follows: A first aspect of the present invention provides the use of reagents for detecting protein biomarkers in the preparation of tumor diagnostic products, said protein biomarkers including FEN1, TUBB3, and UBE2C; In some embodiments of the present invention, the marker includes one of a1) to a3): a1) FEN1, TUBB3, UBE2C, CBX3, and MSH2; a2) FEN1, TUBB3, UBE2C, CBX3 and MCM5; a3) FEN1, TUBB3, UBE2C, MSH2 and PCNA.

[0007] In some embodiments of the present invention, the marker includes one of b1) to b3): b1) FEN1, TUBB3, UBE2C, CBX3, MSH2 and NME1; b2) FEN1, TUBB3, UBE2C, CBX3, MCM5 and MSH2; b3) FEN1, TUBB3, UBE2C, BOLA2, MSH2 and NME1; b4) FEN1, TUBB3, UBE2C, CBX3, PCNA and MSH2.

[0008] In some embodiments of the present invention, the marker includes one of c1) to c3): c1) FEN1, TUBB3, UBE2C, CBX3, MCM5, MSH2 and NME1; c2) FEN1, TUBB3, UBE2C, BOLA2, CBX3, MSH2 and NME1; c3) FEN1, TUBB3, UBE2C, CBX3, MSH2, NME1 and PCNA.

[0009] In some embodiments of the present invention, the reagents for detecting protein biomarkers are selected from reagents of one or more detection methods from the group consisting of: chemiluminescence, immunofluorescence, protein chip, proteometry, immunohistochemistry, patch tracing based on labeling technology, Western blotting, and enzyme-linked immunosorbent assay (ELISA).

[0010] In some embodiments of the present invention, the samples for product testing include blood samples.

[0011] In some embodiments of the present invention, the blood sample includes whole blood, serum, plasma, and peripheral blood.

[0012] In some embodiments of the present invention, the tumor includes breast cancer, liver cancer, lung cancer, or colorectal cancer.

[0013] In some embodiments of the present invention, the tumor is classified according to disease progression, and the tumor includes early-stage tumors.

[0014] In some embodiments of the present invention, the product includes a test kit, a test chip, or a test strip.

[0015] In some embodiments of the present invention, the risk of diagnosing tumors increases when the expression level of the protein biomarker is upregulated relative to the reference population.

[0016] In some embodiments of the present invention, the reference population includes healthy individuals or patients with other diseases who do not have cancer.

[0017] In some embodiments of the present invention, the up-adjustment includes a Log2 multiple ≥ 0.585.

[0018] A second aspect of the present invention provides a method for constructing a tumor prediction and diagnostic model, comprising the following steps: Model construction was performed using the expression levels of protein markers from the first aspect of the invention.

[0019] In some embodiments of the present invention, the model construction algorithm includes at least one of logistic regression, deep learning, linear discriminant analysis, support vector machine, random forest, and recursive partitioning tree.

[0020] A third aspect of the present invention provides a system for predictive diagnosis of tumors, the system comprising a computational device for predicting and diagnosing tumors based on the detection results of the expression levels of protein markers in the application according to the first aspect of the present invention.

[0021] In some embodiments of the present invention, the detection results include protein level results or RNA level results.

[0022] In some embodiments of the present invention, the system further includes any one or more of the following: 1) Detection result collection device, also known as detection result input device, may specifically be one or more of the following: mouse, keyboard, touch screen display, one or more buttons, one or more switches, one or more triggers, etc. 2) Diagnostic result output device, also known as diagnostic result display device, can specifically be one or more of the following: liquid crystal display (LCD), light-emitting diode (LED) display, plasma display, projection display, touch screen display, etc. 3) Diagnostic result sending device, which can send the results of distinguishing whether the subject is in a strong or weak risk group to an information communication terminal device that can be viewed by the patient or medical staff.

[0023] The beneficial effects of this invention are: The protein biomarker combination provided by this invention exhibits excellent diagnostic efficacy in tumor screening. Verification has shown that this biomarker combination achieves high discrimination among all cancer patients and early-stage patients, with an AUC value exceeding 0.9. Its sensitivity and specificity are superior to traditional protein biomarkers, particularly demonstrating a significantly improved detection capability for early-stage tumors, potentially overcoming the shortcomings of existing screening methods in terms of insufficient sensitivity at early stages. Furthermore, this biomarker combination shows stable classification performance in breast cancer, liver cancer, lung cancer, and colorectal cancer, indicating its potential for cross-cancer application and suitability for pan-cancer screening scenarios. The test sample is peripheral blood, which is convenient and minimally invasive, facilitating repeated testing and dynamic follow-up, making it highly operable and valuable for large-scale population screening and clinical auxiliary diagnosis. In addition, the biomarker combination allows for flexible selection of different numbers of genes according to actual needs, achieving a balance between cost and performance, and providing multiple feasible technical pathways for clinical translation. Attached Figure Description

[0024] The present invention will be further described below with reference to the accompanying drawings and embodiments, wherein: Figure 1The classification performance and expression characteristics of candidate pan-cancer blood biomarkers in tumor and non-tumor tissues were studied, where: A represents the AUC value of the candidate gene set used to distinguish between tumor and normal tissues; B represents the expression changes of candidate pan-cancer blood biomarkers in tumor and non-tumor tissues compared with corresponding normal tissues. P.adjust < 0.001, P.adjust < 0.01, P.adjust < 0.05.

[0025] Figure 2 To validate the protein expression results of the sample set, where: A represents the protein levels of candidate pan-cancer blood markers in the serum of healthy controls and cancer patients; B represents the protein levels of candidate pan-cancer blood markers in the serum of healthy controls and early-stage cancer patients (Stage 0, I & II). P < 0.001, P < 0.01, P < 0.05.

[0026] Figure 3 The AUC results of different biomarker combinations in the validation set, where: A is the AUC value of the candidate pan-cancer blood biomarker used to distinguish between cancer patients and healthy controls; B is the AUC value of the candidate pan-cancer blood biomarker used to distinguish between early-stage cancer patients and healthy controls. Detailed Implementation

[0027] The following will describe the concept and technical effects of the present invention clearly and completely with reference to embodiments, so as to fully understand the purpose, features and effects of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are all within the scope of protection of the present invention.

[0028] Example 1: Screening of pan-cancer blood biomarkers Bulk RNAseq expression data of 17 epithelial-derived tumors and their corresponding normal tissues were collected from the Xena database (Table 1), and bulk RNAseq expression data of 21 non-cancerous chronic inflammatory diseases and their corresponding normal tissues were collected from the GEO database (Table 2). Subsequently, the gene-encoded protein products were predicted to be secreted proteins using the deep learning framework BloodProST, and genes with a prediction score greater than 0.9 were selected as initial candidate genes.

[0029] Table 1. Statistical analysis of tumor and normal tissue samples

[0030] Table 2. Statistical analysis of samples from non-cancerous diseases and normal tissues.

[0031] Furthermore, through a systematic review of literature and databases, common biological characteristics related to tumor development and progression, along with their corresponding gene sets, were compiled, including cell cycle, hypermutation, de novo nucleotide synthesis, sialic acid metabolism, cell polarity, cytoplasmic Fenton reaction, and uncoupling proteins. The intersection of these common tumor evolution characteristic gene sets with candidate gene sets was used to obtain candidate genes for further screening and validation of subsequent tumor hematologic biomarkers.

[0032] Furthermore, differential expression analysis was performed between diseased tissues and corresponding normal tissues using the DESeq2 R package to screen for genes significantly upregulated in each disease tissue. The screening criteria were Log2Fold change > 0.585 and p.adjust < 0.05. Subsequently, for the gene set corresponding to each type of tumor-related biological process, the genes were ranked by comprehensively considering their significant upregulation ratio in tumor tissues and non-tumor inflammatory disease tissues, their average expression level in tumor tissues, and their upregulation fold in tumor tissues. The top three genes with tumor-specific high expression were selected as the candidate gene set. If there were fewer than three genes with tumor-specific high expression for a certain process in tumor tissues, all tumor-specific high-expression genes for that process were selected as candidate genes.

[0033] Furthermore, after stratifying all samples according to category labels (tumor / normal), the samples were randomly divided into test and validation sets in a 3:7 ratio, and a logistic regression classification model was constructed to evaluate the classification performance of different combinations of candidate gene sets in various tumor types.

[0034] Based on the number of genes in the gene ensemble, the number of tumor-related biological processes included, and the classification AUC score, four gene sets that showed good classification performance in multiple tumors were selected: 1) BOLA2, FEN1, MCM5, MSH2, NME1, TUBB3; 2) BOLA2, CBX3, FEN1, NME1, TUBB3; 3) BOLA2, FEN1, MCM5, NME1, TUBB3; 4) FEN1, MCM5, NME1, TUBB3, and UBE2C were combined and their union was taken to obtain 9 candidate pan-tumor blood biomarker genes: BOLA2, CBX3, FEN1, MCM5, MSH2, NME1, PCNA, TUBB3, and UBE2C.

[0035] AUC statistical results and expression characteristics of candidate pan-tumor hematologic marker genes in tumor and non-tumor tissues, such as Figure 1 As shown.

[0036] Example 2 Validation of Potential Pan-Cancer Diagnostic Molecular Markers Based on Tumor Secreted Proteins To verify the clinical application value of the blood biomarkers screened in this invention, 45 healthy controls and 45 patients each of colorectal cancer, breast cancer, lung cancer, and liver cancer were recruited as the validation set. All subjects were cancer patients or healthy individuals who had undergone clinical diagnosis and treatment at Shengjing Hospital of China Medical University (Table 3).

[0037] The inclusion criteria for the healthy population control group were: a. Individuals undergoing physical examinations who do not exhibit any clinical symptoms related to tumors; b. No history of malignant tumors; c. Complete blood count, blood biochemistry, and liver and kidney function are basically normal.

[0038] The inclusion criteria for cancer patients are: a. Patients with lung cancer, liver cancer, breast cancer, or colorectal cancer who were initially diagnosed at Shengjing Hospital affiliated with China Medical University and were pathologically confirmed; b. Has not undergone any anti-tumor treatment such as surgery, radiotherapy, or chemotherapy; c. Exclude patients with multiple primary tumors through CT, MRI, ultrasound, or PET-CT examinations; d. Complete blood count, blood biochemistry, and liver and kidney function were basically normal.

[0039] Exclusion criteria are: sample quality does not meet operational requirements (such as improper sample collection, improper storage, insufficient sample volume, sample leakage, sample contamination, sample abnormality, etc.).

[0040] All participants signed informed consent forms after being fully informed.

[0041] Table 3 Clinical information of cancer patients

[0042] The expression levels of the aforementioned candidate proteins in the serum of each subject were detected using ELISA. The results showed that, compared with the healthy control group, the levels of these candidate proteins in the serum of cancer patients were significantly increased, such as... Figure 2As shown in Figure A. Further analysis revealed that the levels of the aforementioned candidate proteins were also elevated in the serum of early-stage cancer patients (stage 0, I, and II) compared to healthy controls, such as... Figure 2 As shown in Figure B, the candidate protein has potential application value for early tumor detection.

[0043] Furthermore, Stratified 10-fold Cross-Validation combined with a logistic regression classification model was used to evaluate the classification performance of different combinations of the above nine genes in the experimental validation data. Specifically, all samples were stratified according to category labels (tumor / normal) and then randomly divided into 10 subsets to ensure that the ratio of tumor samples to healthy control samples in each subset was consistent with the overall distribution. In each iteration, nine subsets were selected as the training set, and the remaining subset was used as an independent validation set. This process was repeated 10 times to ensure that each subset was evaluated as a validation set once. For cancer types with small sample sizes, the number of folds was automatically adjusted according to the smallest category sample size to ensure the effectiveness of stratification. The training and validation of the model were strictly independent: the mean and standard deviation required for data standardization (Z-score normalization) were calculated only from the training set and directly mapped to the validation set; the logistic regression classifier only performed parameter fitting on the training set, and the validation set samples did not participate in the model training or any intermediate optimization process, thereby effectively avoiding data leakage and ensuring that the evaluation results truly reflected the generalization ability of the model.

[0044] The results are as follows Figure 3 As shown in Table 4, although the expression of the above individual genes varied differentially across different tumors in the screening set ( Figure 1 In the validation set, the single genes and dual gene combinations mentioned above did not show good diagnostic performance for different tumors. Considering classification accuracy, stability, and generalization ability, 10 gene combinations (AUC score > 0.9) were finally selected that could effectively distinguish between overall tumors and healthy controls, as well as early-stage tumors and healthy controls, in the four types of tumor patients. Figure 3 As shown in Table 4. Further intersection analysis of the above 10 gene combinations revealed three common genes: FEN1, TUBB3, and UBE2C, suggesting that these three genes may be key molecules for pan-cancer diagnosis. The three-gene combination constructed based on FEN1, TUBB3, and UBE2C also showed good classification performance in the above four types of tumor patients (AUC>0.84, Table 4).

[0045] Table 4. AUC values ​​of gene combinations used to distinguish between cancer patients and healthy controls.

[0046] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited to the above embodiments, and various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention. Furthermore, the embodiments of the present invention and the features thereof can be combined with each other unless otherwise specified.

Claims

1. Application of reagents for detecting protein biomarkers in the preparation of tumor diagnostic products; The protein biomarkers include FEN1, TUBB3, and UBE2C; The tumors include breast cancer, liver cancer, lung cancer, and colorectal cancer; The risk of diagnosing tumors increases when the expression level of the protein biomarker is upregulated relative to the reference population.

2. The application according to claim 1, characterized in that: The marker includes one of a1) to a3): a1) FEN1, TUBB3, UBE2C, CBX3, and MSH2; a2) FEN1, TUBB3, UBE2C, CBX3 and MCM5; a3) FEN1, TUBB3, UBE2C, MSH2 and PCNA.

3. The application according to claim 1, characterized in that: The marker includes one of b1) to b3): b1) FEN1, TUBB3, UBE2C, CBX3, MSH2 and NME1; b2) FEN1, TUBB3, UBE2C, CBX3, MCM5 and MSH2; b3) FEN1, TUBB3, UBE2C, BOLA2, MSH2 and NME1; b4) FEN1, TUBB3, UBE2C, CBX3, PCNA and MSH2.

4. The application according to claim 1, characterized in that: The marker includes one of c1) to c3): c1) FEN1, TUBB3, UBE2C, CBX3, MCM5, MSH2 and NME1; c2) FEN1, TUBB3, UBE2C, BOLA2, CBX3, MSH2 and NME1; c3) FEN1, TUBB3, UBE2C, CBX3, MSH2, NME1 and PCNA.

5. The application according to claim 1, characterized in that: The reagents for detecting protein biomarkers include reagents selected from one or more detection methods from the group consisting of: chemiluminescence, immunofluorescence, protein chip, proteometry, immunohistochemistry, patch tracing based on labeling technology, Western blotting, and enzyme-linked immunosorbent assay (ELISA).

6. The application according to claim 5, characterized in that: The samples tested for the product include blood samples.

7. A method for constructing a tumor prediction and diagnostic model, comprising the following steps: Model construction was performed using the expression levels of protein biomarkers in the applications described in claims 1-6.

8. The construction method according to claim 7, characterized in that: The model construction algorithm includes at least one of logistic regression, deep learning, linear discriminant analysis, support vector machine, random forest, and recursive partitioning tree.

9. A system for predictive diagnosis of tumors, the system comprising a computational device for predicting and diagnosing tumors based on the detection results of the expression levels of protein markers in the applications of claims 1 to 6.

Images