Use of biomarkers in the classification diagnosis or prognosis prediction of glioblastoma

Abstract

The application discloses biomarkers in glioblastoma classification diagnosis or prognosis prediction, and the biomarkers are PLCH1 genes and EGFR genes. The PLCH1 genes and the EGFR genes in a to-be-detected ex vivo biological sample are detected by using reagents, and the marker genotypes are used for classification diagnosis or prognosis prediction of the glioblastoma. According to the biomarkers, GBM patients are classified, a survival prediction model of the glioblastoma patients is constructed through reasonable assignment of the classification, and the accuracy of the survival rate prediction of the GBM patients is greatly improved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of molecular biology, and particularly relates to the application of biomarkers in the classification diagnosis or prognosis prediction of glioblastoma. BACKGROUND

[0002] Glioma is the most common primary brain tumor in adults. Glioma is divided into astrocytoma, oligodendroglioma and ependymoma according to its histological appearance. The World Health Organization (WHO) divides glioma into grades I, II, III and IV. The higher the grade, the more invasive. Grade IV glioma, i.e. glioblastoma multiforme (GBM), is the most invasive and highest degree of malignancy. The 1-year and 5-year overall survival rates of high-grade glioma are 40% and 10%, respectively. The high mortality rate of glioma is due to its invasiveness and high recurrence rate. Although surgery, radiotherapy and alkylating chemotherapy are used, the heterogeneity of glioma does not improve the survival rate of patients. Significant advances in genomics, transcriptomics and epigenetics have brought new concepts to the classification and treatment of glioma. Therefore, specific tumor progression-related molecular markers as potential therapeutic targets have received much attention in recent years.

[0003] Phospholipase C Eta 1 (PLCH1) is a member of the PLC-Eta family of the phosphatidylinositol-specific phospholipase C (PLC) superfamily, which can cleave phosphatidylinositol 4,5-bisphosphate to produce the second messenger inositol 1,4,5-trisphosphate (IP3) and diacylglycerol (DAG). PLCH1 is a putative protein kinase C and Ca 2 + signal component in neurons and neuroendocrine tissues. The PLCH1 domain consists of a pleckstrin homology (PH), EF-hand, catalytic X and Y domains, and a protein kinase C conserved region 2 (C2). The X and Y domains (~150 and ~115 residues, respectively) fold to form the catalytic site. The X domain is involved in substrate and Ca 2+ binding (essential for catalysis), while the Y domain mainly interacts with the substrate. The C2 domain is essential for catalytic activity and can contain up to 4 Ca 2+ binding sites, which are usually present in proteins that interact with phospholipids. The additional c-terminal region is rich in serine and proline residues, which are believed to play a role in protein-protein interactions.

[0004] Reports on PLCH1 are scarce, but some evidence suggests that PLCH1 abnormalities are associated with this disease. Vasco et al. showed that lipopolysaccharide (LPS)-treated cells no longer express PLCH1, suggesting that the PI-PLC subtype may be involved in the metabolic pathway activated by LPS-induced inflammation. Zhang found that PLCH1 polymorphism is closely associated with squamous cell carcinoma. In 2005, four research groups identified a novel PLC family, h1 and h2, using mammalian genomes. PLCH1 enzymes may act as Ca2 sensors, playing a major role in the formation and maintenance of neuronal networks in the postnatal brain. Kim et al. reported the downregulation of PLCH1 in viral-infected primary non-small cell lung cancer. Epidermal growth factor receptor (EGFR) is a transmembrane tyrosine kinase that can be activated by epidermal growth factor (EGF) and transforming growth factor (TGF-β). α (TGF α EGFR receptor dimer activation occurs via EGFR or other ligands. Upon activation, the EGFR receptor dimer induces autophosphorylation of the receptor's C-terminal tyrosine residues, activating the intracellular MAPK pathway, the PI3K signaling pathway, and STAT transcription factors. Downstream pathways lead to DNA synthesis and cell proliferation. Mutations in EGFR have been reported to promote tumorigenesis. Interestingly, the binding of EGFR and PLCH1 has been observed in membrane yeast two-hybrid assays; however, the regulatory relationship between EGFR and PLCH1 is rarely reported. The binding mechanism of EGFR to PLCH1 and how mutant EGFR activates the pro-tumorigenic function of PLCH1 require further investigation.

[0005] GBM prognostic risk models often use Kaplan-Meier and Cox proportional hazards models to fit the survival time of all GBM patients. However, due to the heterogeneity of tumors—that is, the differences in genotype and phenotype between different individuals with the same malignant tumor or between tumor cells in different sites within the same patient—this heterogeneity can manifest as different genetic backgrounds, such as differences in the quantity and quality of chromosomes, the diversity of cell evolution at different cell types, clinical stages, and differentiation degrees, and even significant differences at the molecular level in homogeneous tumors, reflecting the high complexity and diversity of malignant tumor evolution. Therefore, this method ignores the heterogeneity caused by molecular differences between different tumor individuals. Summary of the Invention

[0006] The technical problems to be solved by this invention are: the difficulty in correctly classifying GBM and the technical problems in prognosis prediction.

[0007] The technical solution of the present invention is: the application of reagents for detecting biomarker genotypes in the preparation of kits for the classification diagnosis or prognosis prediction of glioblastoma, wherein the biomarkers are the PLCH1 gene and the EGFR gene.

[0008] Furthermore, reagents were used to detect the PLCH1 and EGFR genes in ex vivo biological materials, and the genotypes of these biomarkers were used for classification, diagnosis, or prognostic prediction of glioblastoma.

[0009] When both the EGFR genotype and the PLCH1 gene are mutant, it is the first subtype;

[0010] When both the EGFR genotype and the PLCH1 gene are wild-type, it is the second subtype;

[0011] When the EGFR genotype is wild-type and the PLCH1 gene is mutant, it is the third subtype;

[0012] When the EGFR genotype is mutant and the PLCH1 gene is wild-type, it is the fourth subtype.

[0013] Furthermore, the biomaterial is blood or tumor cells.

[0014] This invention also discloses a method for constructing a survival prediction model for glioblastoma patients, comprising:

[0015] (1) Obtain clinical data information on the patient’s age, gender, IDH status, primary or recurrent glioblastoma (PRS), radiotherapy status, chemotherapy status, and genotype combinations of PLCH1 and EGFR genes.

[0016] (2) Using the patient’s clinical data from step (1), calculate the total score (Total Points) for each indicator according to the corresponding score (Points) in the table, and use nomogram to build a survival prediction model to assess the patient’s survival rate.

[0017] The values ​​assigned to each item in the data information are shown in the table below:

[0018]

[0019] In the table, EGFR-mut / PLCH1-H means that the EGFR gene is mutant and the PLCH1 gene is wild-type; EGFR-mut / PLCH1-L means that the EGFR gene is mutant and the PLCH1 gene is mutant; EGFR-wt / PLCH1-H means that the EGFR gene is wild-type and the PLCH1 gene is wild-type; EGFR-wt / PLCH1-L means that the EGFR gene is wild-type and the PLCH1 gene is mutant.

[0020] Compared with the prior art, the present invention has the following beneficial effects:

[0021] 1. This invention is the first to discover that phospholipase C Eta 1 (PLCH1) is a target gene of GBM. Considering the heterogeneity of GBM, this invention analyzes EGFR subtypes of GBM separately, finding that PLCH1 is a tumor suppressor in EGFR wild-type GBM, but a tumor driver in EGFR mutant GBM. Based on PLCH1 expression levels and EGFR gene status, glioblastoma is reclassified as follows: EGFR-wildtype / PLCH1-high, EGFR-wildtype / PLCH1-low, EGFR-mutation / PLCH1-high, EGFR-mutation / PLCH1-low. This new classification is used to address the problems of diagnosis, prognosis prediction, and treatment of GBM patients.

[0022] 2. This invention classifies GBM patients based on biomarkers and constructs a survival prediction model for glioblastoma patients by assigning reasonable values ​​to the classifications, which greatly improves the accuracy of predicting the survival rate of GBM patients. Attached Figure Description

[0023] Figure 1 The survival rate of patients with PLCH1 mutations is reduced: a. Data analysis workflow; b. Mutation sites of patients carrying PLCH1 mutations; c, d. Significantly reduced survival in GBM and RGBM patients after PLCH1 mutations; e. Differences in PLCH1 expression levels; f. Tumor mutation burden of patients.

[0024] Figure 2 Reduced survival in PLCH1 patients after mutation: j. Significantly differentially expressed genes between PLCH1-mut and PLCH1-wt patients; h. Signaling pathways involved in differentially expressed genes between PLCH1-mut and PLCH1-wt patients.

[0025] Figure 3Patients with low PLCH1 expression have significantly reduced survival rates: a. Data analysis workflow; b. Glioma patients with low PLCH1 expression have significantly reduced survival rates; c, d. Comparison of PLCH1 expression levels between low-grade and high-grade gliomas;

[0026] Figure 4 Patients with low PLCH1 expression have significantly reduced survival rates: e, f, g, h Comparison of PLCH1 expression levels in different types of gliomas;

[0027] Figure 5 Patients with low PLCH1 expression have significantly reduced survival rates: i. Experimental validation of PLCH1 expression in glial cells; j. Survival analysis of PLCH1 in CGGA-GBM.

[0028] Figure 6 Patients with low PLCH1 expression had significantly lower survival rates: k,i. Pathways involved by differentially expressed genes in the two groups of patients with high and low PLCH1 expression.

[0029] Figure 7 PLCH1 exhibits a dual role in different EGFR subtype tumors: a. Data analysis workflow; b. In vitro binding assay of EGFR to PLCH1; c. Survival analysis of PLCH1 in different EGFR subtypes; d. PCA analysis of different glioma subtypes; e. Experimental validation of the role of PLCH1 in different EGFR subtypes; f. Schematic diagram of the role of PLCH1 in different EGFR subtypes.

[0030] Figure 8 Application of PLCH1 reconstructed subtypes in prognostic prediction: a. Four new subtypes were formed based on PLCH1 hyperexpression and EGFR mutation; b. The four new subtypes were significantly associated with patient prognosis.

[0031] Figure 9 Application of PLCH1 reconstructed subtypes in prognostic prediction: reconstruction of predictive models for patients based on risk scores and clinical characteristics;

[0032] Figure 10 Application of PLCH1 reconstructed subtypes in prognostic prediction: d. Correlation between PLCH1 and risk score; e. AUC value of patient clinical prognostic prediction model; f. Accuracy of prognostic model in predicting within 1 year. Detailed Implementation

[0033] Unless otherwise specified, the experimental methods used in the following examples are conventional methods. Unless otherwise specified, the experimental materials used in the following examples were all purchased from commercial sources.

[0034] Download transcriptome sequencing data, clinical data, and gene mutation data from the Cancer Genome Atlas Program (TCGA) glioma project, and select 150 patients with these three types of information as our research dataset.

[0035] Raw transcriptome, whole-exome sequencing, and clinical data were used from the Chinese Glioma Genome Atlas (CGGA), totaling 693 cases, as our study dataset. The Whole-Exome Analysis Toolkit (GATK) pipeline identified SNPs / INDELs in WES and RNA-Seq. GATK is a software library that provides a suite of tools for processing human data, including a deep coverage analyzer, quality score recalibrator, local recalibrator, and SNP / INDEL caller. Variation annotation was then performed using annovar, and mutated genes in WES were located using the maftools package in R. Reads from the RNA-Seq data were mapped to a reference transcriptome, and gene counts for each gene or transcript were quantified (RSEM + STAR pipeline). These data were then used for differential expression analysis.

[0036] Using the R maftools package, whole-exome analysis of CGGA revealed 6 cases carrying PLCH1 mutations, with 7 mutation sites: PLCH1-Q326X, PLCH1-R951l, PLCH1-S112N, PLCH1-T1206I, PLCH1-G1644R, and PLCH1-T1677l. These mutations significantly reduced patient survival rates. Figure 1 Patients with PLCH1 mutations (b, c, d) and a survival rate of less than one year showed decreased PLCH1 expression and significantly increased tumor mutational burden (TMB), strongly suggesting that PLCH1 is a tumor suppressor gene. Differential analysis (FDR < 0.05, log2FC > 1 or < -1) was performed between PLCH1 mutant patients (PLCH1-mut) and PLCH1 wild-type patients (PLCH1-wt), yielding 158 differentially expressed genes. The genes with the largest fold changes were: CCL18, NPAS4, NDST3, PRRG3, DCHS2, SLC9A7, ZNF124, MLXIPL, and JPH1. Figure 2The signaling pathways involved by differentially expressed genes mainly include: calcium signaling pathway, neuroactive ligand-receptor interaction, dopaminergic synapse, GABAergic synapse, morphine addiction, circadian entrainment, aldosterone synthesis and secretion, cholinergic synapse, amphetamine addiction, and nicotine addiction. Among these, PLCH1 plays a significant regulatory role in the calcium signaling pathway. Figure 2 (h).

[0037] Transcriptome data from CGGA were analyzed, and TCGA data were used for further validation. Optimal grouping sites were identified using R tinyarray, and survival analysis was performed on low-grade gliomas (LGG). The results showed that patients with low PLCH1 expression had significantly lower survival rates. Figure 3 (b) The comparison of PLCH1 expression levels between low-grade gliomas and high-grade gliomas (GBM) also showed that high-grade gliomas had lower PLCH1 expression levels. Comparison of PLCH1 expression levels among different types of gliomas ( Figure 4 The study (e, f, g, h) found that PLCH1 expression was not related to age or sex. Experiments verified PLCH1 expression in glial cells (…). Figure 5 (i). Using R tinyarray to find the optimal grouping sites, survival analysis was performed on high-grade gliomas (GBM) in CGGA, including 26 patients in the PLCH1 high expression group (PLCH1-H) and 185 patients in the PLCH1 low expression group (i). Figure 5 In the study, differential gene analysis was performed on the two groups of patients (FDR < 0.05, log2FC > 1 or < -1), revealing 1499 upregulated DEGs and 2974 downregulated DEGs. These primarily regulated the upregulated Calcium signaling pathway and Neuroactive ligand receptor interaction signaling pathway, and the downregulated Cytokine receptor interaction and Cell cycle signaling pathways. Figure 6In TCGA-GBM, differentially expressed genes also primarily regulate the above four pathways, but in TCGA, cytokine receptor interaction is also upregulated with increasing PLCH1 levels. This strongly suggests that PLCH1 has an inhibitory effect on gliomas, and that increased PLCH1 levels can regulate Ca ion-related signaling pathways and reduce cell cycle signaling pathways (…). Figure 6 (l).

[0038] CGGA transcriptome data were analyzed for mutation sites using the GATK analysis pipeline. NetworkAnalyst was used to predict PLCH1 binding to EGFR, and experiments verified the direct in vitro binding of PLCH1 to EGFR. Figure 7 (b) Since EGFR is an important target factor in tumors, and EGFR mutations can promote tumor progression and drug resistance, this invention divides CGGA GBM into EGFR-mut and EGFR-wt subtypes, and studies the role of PLCH1 in each subtype. This invention found that PLCH1 remains a tumor suppressor in the EGFR-wt subtype, but a tumor promoter in the EGFR-mut subtype. Figure 7 (c) Using ternary PCA to project different types of patients onto a two-dimensional plane, it was found that the different types of patients were significantly separated, and the separating effect of EGFR was greater than that of PLCH1 expression, indicating that PLCH1 only shows differences in EGFR subtypes. Figure 7 (d). Experiments in GBM cells demonstrated the bidirectional role of PLCH1 in different EGFR isoforms; in the EGFR-mut isoform, PLCH1 knockout led to cell death. Figure 7 (e).

[0039] The taxonomic subtypes of GBM were reconstructed based on EGFR status and PLCH1 expression. Figure 8 (a) The survival rate differences among different subtypes were statistically significant. Figure 8 (b) To quickly and accurately predict patient survival based on EGFR status and PLCH1 expression classification, we used the rms package in R to integrate survival time, survival status, EGFR / PLCH1 subtype, primary or recurrent GBM (PRS), sex, age, radiotherapy (Radio status), chemotherapy (Chemstatus), and IDH mutation. Cox's nomogram survival prediction model assessed the significance of 1-, 3-, and 5-year survival in CGGA GBM patients. Each feature of a patient has a corresponding score, and the sum of the scores of the patient's seven features yields a linear predictive risk score. Figure 9 The predicted risk score line is significantly correlated with actual survival time.Figure 10 (d). The overall C-index of the model was 0.71 (p < 0.0001). ROC analysis was performed using pROC (version 1.17.0.1) in R to obtain the AUC value, which is an important basis for classification prediction. The AUC values ​​of our model at time points of 365, 1095, and 1825 were 0.79, 0.78, and 0.88, respectively. Figure 10 (e).

[0040] This survival prediction model can be applied clinically (corresponding to Table 1). For example, a 50-year-old male (score 10, score 4) with primary GBM (score 0), who is genetically assessed to be a wild-type IDH (score 56), belongs to the EGFR-wt / PLCH1-L subtype (score 47), receives chemotherapy (score 0), does not receive radiotherapy (score 26), has a total score of 143, and has a 1-year survival rate of ~45%.

[0041] Table 1 Scoring table for each clinical feature

[0042]

[0043] The results above indicate that PLCH1 has a tumor-suppressive effect in EGFR wild-type GBM, while it has a tumor-promoting effect in EGFR mutant GBM. The diagnostic value of PLCH1 under different EGFR gene states can determine the severity of GBM. Furthermore, the triggering effect of EGFR mutation on PLCH1 demonstrates the unique regulatory role between EGFR and PLCH1. Classifying GBM based on these two genes allows for more precise classification, thus providing treatment recommendations. The discovery of PLCH1's function also provides a new drug therapeutic target for diseases such as glioblastoma, contributing to a new direction for subsequent drug development and clinical treatment, and possessing significant social value and market application prospects.

[0044] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. The application of reagents for detecting biomarker genotypes in the preparation of a kit for predicting the prognosis of glioblastoma, wherein the biomarkers are the PLCH1 gene and the EGFR gene, the reagents are used to detect the genotypes of the PLCH1 gene and the EGFR gene in ex vivo biological materials, wherein the biological materials are blood or tumor solid cells, and the genotypes of the biomarkers are used to predict the prognosis of glioblastoma. The specific method for predicting the prognosis is as follows: (1) Obtain clinical data information on the patient’s age, gender, IDH gene status, primary or recurrent glioblastoma, radiotherapy status, chemotherapy status, and subtype classification; (2) Using the patient's clinical data from step (1), calculate the total score for each indicator according to the scores in the table below: ， The subtype classification method is as follows: When both the EGFR genotype and the PLCH1 gene are mutant, it is the first subtype; When both the EGFR genotype and the PLCH1 gene are wild-type, it is the second subtype; When the EGFR genotype is wild-type and the PLCH1 gene is mutant, it is the third subtype; When the EGFR genotype is mutant and the PLCH1 gene is wild-type, it is the fourth subtype; (3) The patient’s survival rate was assessed using the nomogram survival prediction model in Figure 9 of the instruction manual.

Images