CN112063720B

CN112063720B - Osteosarcoma prognosis marker and prognosis evaluation model

Info

Publication number: CN112063720B
Application number: CN202011004518.4A
Authority: CN
Inventors: 杨孟恺; 华莹奇; 蔡郑东
Original assignee: Shanghai First Peoples Hospital
Current assignee: Shanghai First Peoples Hospital
Priority date: 2020-09-22
Filing date: 2020-09-22
Publication date: 2022-09-27
Anticipated expiration: 2040-09-22
Also published as: CN112063720A

Abstract

The invention relates to osteosarcoma prognostic markers and a prognostic evaluation model. The invention integrates the overall survival data of osteosarcoma patients with TARGET and GSE21257 which are independent databases, screens and confirms that glycolysis related genes P4HA1, ABCB6 and STC2 are osteosarcoma prognosis markers, constructs a prognostic risk assessment model based on P4HA1, ABCB6 and STC2, verifies that the model HAs higher accuracy through a training set and a verification set, and further analyzes and finds that the risk score of the model is also obviously related to metastasis and a tumor immune microenvironment. The invention is helpful for better predicting the prognosis of osteosarcoma patients, and can bring great help for guiding the clinical treatment of the disease.

Description

Osteosarcoma prognosis marker and prognosis evaluation model

Technical Field

The invention relates to disease prognosis evaluation, in particular to an osteosarcoma prognosis marker and a prognosis evaluation model.

Background

Osteosarcoma is a malignant tumor which is usually found in children and adolescents and is originated from bone tissues, and has the characteristics of easy recurrence and easy metastasis, wherein lung metastasis is the most important cause of death of osteosarcoma patients. The preferred treatment scheme for osteosarcoma patients is new adjuvant chemotherapy combined with surgery, which greatly improves the 5-year survival rate of patients, but once the patients have lung metastasis, the 5-year survival rate is still lower than 20%. Therefore, the mechanism of osteosarcoma generation and transfer is explored from the molecular level, a foundation is laid for early diagnosis and prognosis evaluation of osteosarcoma, and the method has important scientific and social significance. An effective osteosarcoma patient prognosis marker is searched, and a basis can be provided for further diagnosis and treatment of osteosarcoma patients.

The metabolic reprogramming capability of tumor cells provides enough energy for rapid proliferation, and the increase of aerobic glycolysis is one of the important metabolic reprogramming characteristics of tumor cells with stronger migration capability. Glycolysis not only provides energy for cancer cells, but also provides the necessary precursors for the biosynthesis of cancer cells. Metabolites of glycolysis, such as glucose-6-phosphate, dihydroxyacetone phosphate, and the like, respectively participate in the pentose phosphate pathway and intracellular lipid synthesis, and have important significance for cancer cell growth. Several studies have found that an increase in aerobic glycolysis is also closely related to the development of metastases in osteosarcoma patients. With the application of gene sequencing in osteosarcoma, many biomarkers are used to construct risk assessment models for predicting the prognosis of osteosarcoma patients, but whether the glycolysis-related genes can predict the prognosis of osteosarcoma patients remains to be studied.

ABCB 6: ATP-binding cassette (ABC) transporters are a large superfamily of transmembrane proteins with specific amino acid sequences and ATP binding domains that facilitate the transport of a variety of biological compounds, such as polypeptides, drugs, steroids, bile acids, phospholipids and ions, across membranes via ATP-dependent pathways. The human genome encodes 48 ABC transporters, which are divided into 7 subfamilies according to their function and homology: ABCA to ABCG. ABCB6 is a member of the ABCB subfamily, is 842 amino acids long, and functions by forming homodimers with itself. ABCB6 is widely distributed in cells, can be positioned in mitochondria, lysosomes, Golgi apparatus and other organelles, and can play different roles. The research shows that the ABCB6 gene mutation can cause Lan blood group negativity and is also related to ocular defect and familial pseudohyperkalemia.

P4HA 1: prolyl 4-hydroxylase subunit α 1(prolyl 4-hydroxyase subbunit α 1, P4HA1) is the rate-limiting subunit of prolyl 4-hydroxylase and is required for the synthesis of various types of collagen. Research shows that P4HA1 can regulate the synthesis and secretion of collagen in fibroblast, so as to change the components of extracellular matrix and influence the biological behaviors of tumor adhesion, migration and the like. However, P4HA1 is still less studied in tumors.

STC 2: stanniocalcin (STC) is a class of glycoprotein hormones with a variety of biological functions. Human STC comprises both members of STC1 and STC2, both of which are associated with tumors. Research shows that STC2 promotes the growth of tumor and avoids apoptosis mainly by protecting tumor cells from hypoxia, promoting the generation of new blood vessels of tumor cells, promoting the proliferation, invasion and metastasis of tumor cells, inhibiting immune response and the like. STC2 is a new tumor marker and needs to be further studied.

The above ABCB6, P4HA1 and STC2 all belong to glycolysis-related genes.

Patent document CN108410986A, published japanese patent No. 2018.08.17, discloses that CDH6 is closely related to the overall survival and prognosis of osteosarcoma patients, and may be a biomarker for osteosarcoma prognosis. Patent document CN109628593A, published japanese 2019.04.16, discloses that expression of CD24 is associated with prognosis of clinical osteosarcoma patients, and that cases with high expression of CD24 may have higher risk of recurrence or metastasis in the future, and may be used for evaluating prognosis of osteosarcoma. However, the glycolysis related genes ABCB6, P4HA1 and STC2 are not reported for prognosis of osteosarcoma at present.

Disclosure of Invention

The invention aims to provide a novel osteosarcoma prognosis marker and a prognosis evaluation model aiming at the defects in the prior art.

In a first aspect, the invention provides application of one or more of glycolysis related genes or proteins ABCB6, P4HA1 and STC2 as markers in preparation of osteosarcoma prognosis reagents or kits.

As a preferred embodiment of the present invention, the prognosis is selected from any one or several of the following:

a) predicting the survival rate of an individual with the osteosarcoma for a certain time;

b) predicting survival time of individuals with osteosarcoma;

c) predicting the possibility of the occurrence of metastasis of an individual with osteosarcoma;

d) predicting the tumor microenvironment of individuals with osteosarcoma.

More preferably, the tumor microenvironment in the individual with the osteosarcoma is selected from the group consisting of a predictive immune score and the number of naive B cells in the tumor immune microenvironment.

As another preferred example of the invention, ABCB6, P4HA1 and STC2 are combined for prognosis of osteosarcoma.

More preferably, the specific method is as follows: detecting the expression amount of individual ABCB6, P4HA1 and STC2 genes in tumor tissues, and then substituting the expression amounts into a risk score formula: the risk score (0.946 × ABCB6 gene tumor tissue expression level) + (0.413 × P4HA1 gene tumor tissue expression level) + (0.435 × STC2 gene tumor tissue expression level) was calculated.

As another preferred example of the present invention, the expression level of a gene or protein is measured for prognosis.

More preferably, the method for detecting the expression level is selected from RT-PCR, real-time fluorescent quantitative PCR, gene chip, high-throughput sequencing or immunological detection.

In a second aspect, the present invention provides a kit for prognosis evaluation of osteosarcoma, which comprises a reagent for detecting the expression level of at least one gene selected from the group consisting of ABCB6, P4HA1 and STC 2.

As a preferred example of the present invention, the kit contains a reagent for detecting the expression amounts of ABCB6, P4HA1 and STC 2.

More preferably, the kit contains instructions describing: detecting the expression quantity of a specific gene in the tumor tissue, and substituting the expression quantity into a risk scoring formula: and calculating the risk score (0.946 × ABCB6 gene tumor tissue expression level) + (0.413 × P4HA1 gene tumor tissue expression level) + (0.435 × STC2 gene tumor tissue expression level).

The invention has the advantages that:

1. the invention integrates the overall survival data of osteosarcoma patients with TARGET and GSE21257 independent databases, screens and confirms that glycolysis-related genes P4HA1, ABCB6 and STC2 are osteosarcoma prognosis markers, constructs a risk assessment model based on P4HA1, ABCB6 and STC2, verifies the accuracy of the model through a training set and a verification set, and further analyzes and finds that the risk score of the model is obviously related to metastasis and tumor immune microenvironment. The invention is helpful for better predicting the prognosis of osteosarcoma patients, effectively evaluating the prognosis risks of different patients, finding the transfer tendency of the patients in advance, intervening in advance, and bringing great help for guiding the clinical treatment of the disease, thereby improving the 5-year survival rate of the osteosarcoma patients.

2. The risk assessment model has high prediction accuracy and universality, is obviously superior to other osteosarcoma prognosis methods in the prior art, and is also obviously superior to other models constructed in the research process.

Drawings

FIG. 1: glycolytic gene sets were found to be enriched in the transfer group samples by GSEA enrichment analysis.

FIG. 2 is a schematic diagram: a prediction model of glycolysis-related genes is constructed in the test set. A: transfer and non-transfer group differential gene heatmaps; b: single factor COX analysis of forest maps; c: constructing a multifactor COX analysis model; d: a distribution of risk scores; e: survival status of patients with different risk scores; f: the expression profiles of 3 genes in the model were predicted.

FIG. 3: and (5) verifying the relation between the prediction model and the overall survival rate and the transfer character of the patient in a test set. A: the Kaplan-Meier survival curve shows the survival rate difference between the high and low risk groups; B. c: the ROC curve shows the accuracy of single-gene prediction of 1-year and 3-year survival rates in the risk model and the model respectively; d: relationship between risk score and patient metastatic trait.

FIG. 4 is a schematic view of: the relationship between the predictive model and the overall patient survival and metastatic behavior is validated in a validation set. A: the Kaplan-Meier survival curve shows the survival rate difference between the high and low risk groups; b: the ROC curve shows the reliability of the prediction model after prediction; c: a relationship between risk score and patient metastatic trait; d: a distribution of risk scores; e: survival status of patients with different risk scores; f: the expression profiles of 3 genes in the model were predicted.

FIG. 5: the relationship between the risk prediction model and the osteosarcoma tumor immune microenvironment. A. B: the test set and the verification set respectively have higher immune score difference between the low risk groups; C. d: the test set and the verification set respectively compare the survival rate difference of patients with higher and lower immune scores; E. f: respectively comparing the relationship between the risk score and the ratio of naive B cells in the tumor immune microenvironment in a test set and a verification set; G. h: the Kaplan-Meier survival curves show the difference in overall survival of B cells from the patients in the test set and validation set.

Detailed Description

The following detailed description of the present invention is provided in connection with the accompanying drawings.

Example 1

Method and device

1. Data acquisition

mRNA expression profile data and clinical information of osteosarcoma patients are downloaded from a TARGET database respectively to serve as a training set, and mRNA expression profile data and clinical information of osteosarcoma patients are downloaded from a GEO database GSE21257 to serve as a verification set.

2. Screening for prognostic markers

The patients were divided into two groups of metastatic and non-metastatic patients in the training set according to the clinical information of the patients, and the glycolytic related gene set of the patients in the metastatic group was found to be obviously enriched (P <0.05) by GSEA gene enrichment analysis (figure 1). We then selected 198 glycolytic related genes from this gene set for subsequent studies, and found that 23 of the 198 genes were differentially expressed in the metastatic and non-metastatic groups by comparing gene expression in the metastatic and non-metastatic groups using P <0.05 as the threshold for screening for differentially expressed genes, followed by further screening by Kaplan-Meier survival analysis to find that 10 of the above 23 genes were expressed in association with the overall survival rate of osteosarcoma patients. Then we performed a one-way COX regression analysis to find that 6 genes were associated with osteosarcoma prognosis (P <0.05), and finally screened 3 genes associated with osteosarcoma patient prognosis as risk markers by a multi-way COX regression model (a, B and C in fig. 2).

3. Model building

Based on the linear correlation of the regression coefficient coef generated by the multifactor COX regression model and the risk marker gene, a prediction model of three glycolysis related genes with prognosis correlation is established.

4. Risk scoring formula

And (4) dividing the sample into two groups according to the median of the risk score, wherein the group with high risk is higher than the median, and the group with low risk is lower than the median of the risk score, namely (0.946 multiplied ABCB6 gene tumor tissue expression level) + (0.413 multiplied P4HA1 gene tumor tissue expression level) + (0.435 multiplied STC2 gene tumor tissue expression level).

5. Evaluating the relationship between risk model and overall survival rate in test set and verifying the accuracy of the model

By drawing the survival rate difference of a Kaplan-Meier curve compared with a high-risk group and a low-risk group, drawing an ROC curve of a patient by using an R package 'survivvalROC', calculating the area under the curve, and verifying the accuracy of the model.

6. Verifying model accuracy in verification set

Firstly, performing batch correction on mRNA expression profile data in a test set and a verification set by using an R package 'sva', dividing samples in the verification set into a high-risk group and a low-risk group according to a median of risk scores in the test set, drawing survival rate difference of a Kaplan-Meier curve compared with the high-risk group and the low-risk group, drawing a ROC curve of a patient by using the R package 'survivvalROC', calculating the area under the curve, and further verifying the accuracy of the model in the verification set.

7. Risk score relationship analysis with other factors

First, the risk score was determined to be significantly correlated with prognosis, and then the relationship of the risk score to metastasis and tumor immune microenvironment was compared.

Second, result in

1. Obtaining a predictive risk model from multifactor Cox regression analysis

And obtaining a risk model constructed by 3 genes and a risk scoring formula from multi-factor Cox regression analysis. The calculation is as follows: the risk score is (0.946 × ABCB6 gene tumor tissue expression level) + (0.413 × P4HA1 gene tumor tissue expression level) + (0.435 × STC2 gene tumor tissue expression level). All 3 genes are negatively associated with survival and play a role in increased risk, as shown in table 1.

TABLE 1. coefficients for 3 genes in the model

Gene	coef	HR	HR.95L	HR.95H	p-value
						ABCB6	0.946	2.575	1.345	4.931	0.004
P4HA1	0.413	1.512	1.074	2.123	0.018
						STC2	0.435	1.545	1.137	2.097	0.005

2. Training set risk scoring

Osteosarcoma samples in the training set were divided into high-risk group and low-risk group by using the median calculated risk score (1.03) as a threshold. Kaplan-Meier curve analysis revealed that the survival rate was lower in the lower risk group (P <0.001) in patients in the high risk group (A in FIG. 3). The areas under the ROC curves for predicting 1-year and 3-year survival rates were 0.884 and 0.790, respectively, further illustrating the good predictive power of the model (B and C in fig. 3). Meanwhile, the distribution of risk scores, survival information and gene expression heat maps in the training set is verified (D, E and F in figure 2), and the risk scores of patients in a high-risk group are higher, the survival number is lower, and the expression content of 3 risk markers is higher.

3. Validating a risk model in a validation set

To verify the universality of the risk model, the median of the risk scores in the training set is used as the threshold of the verification set, and the verification set is divided into a high risk group and a low risk group. Kaplan-Meier curve analysis found lower survival in the high risk group patients in the lower risk group (P ═ 0.018), consistent with that found in the training set (a in fig. 4). The areas under the ROC curves for predicting 1-year and 3-year survival rates were 0.740 and 0.759, respectively, indicating that the model also has good predictive power in the validation set (B in fig. 4). The relationship between the validated centralized risk score, survival information and gene expression heatmap was also subsequently validated (D, E and F in FIG. 4), and patients in the high risk group were also found to have higher risk scores, lower survival numbers and higher expression levels of 3 risk markers.

4. Relationship between Risk score and metastasis

We found that the transfer rate was higher in the high risk group patients, whether in the test set (D in fig. 3) or the validation set (C in fig. 4). The risk model is proved to be capable of well predicting the capability of the osteosarcoma patient to generate metastasis, so that the possibility is provided for early diagnosis of the osteosarcoma metastasis.

5. Relationship of Risk score to tumor microenvironment

The immune score for each sample was calculated by the ESTIMATE algorithm in the R software and found to be lower for the high risk group in both the test and validation sets (a and B in fig. 5). The proportion of tumor infiltrating immune cells was calculated for each sample by CIBERSORT algorithm in R software and correlation analysis showed that the risk score was positively correlated with the proportion of naive B cells in both the test and validation sets (E and F in fig. 5).

Example 2

The embodiment provides a kit for osteosarcoma prognosis, which contains a reagent for detecting the expression levels of ABCB6, P4HA1 and STC2 genes. The kit comprises an operation instruction, and is described as follows: according to the conventional method, the tumor tissue expression quantity of ABCB6, P4HA1 and STC2 genes is detected by a high-throughput sequencing method, and then the expression quantity is substituted into a risk score formula: the risk score is calculated as (0.946 × ABCB6 gene tumor tissue expression level) + (0.413 × P4HA1 gene tumor tissue expression level) + (0.435 × STC2 gene tumor tissue expression level), and the higher the risk score is, it indicates that: the shorter the survival, the lower the survival over a certain length of time, the higher the risk of metastasis, the lower the immune score of the tumor microenvironment, and the greater the number of naive B cells in the tumor immune microenvironment.

Example 3

The embodiment provides a kit for osteosarcoma prognosis, which contains a reagent for detecting the expression level of at least one gene in ABCB6, P4HA1 and STC2, in particular to a gene chip for detecting the expression level of the gene. The higher the expression level of the gene to be tested, the following expression levels are shown: the shorter the survival, the lower the survival over a certain length of time, the higher the risk of metastasis, the lower the immune score of the tumor microenvironment, and the greater the number of naive B cells in the tumor immune microenvironment.

The above description is only a preferred embodiment of the present invention, and it should be noted that, for a person skilled in the art, several modifications and additions can be made without departing from the method of the present invention, and these modifications and additions should also be considered as the protection scope of the present invention.

Claims

1. Application of a reagent for detecting the expression levels of ABCB6, P4HA1 and STC2 genes in preparation of osteosarcoma prognostic reagent or kit.

2. Use according to claim 1, wherein the prognosis is selected from any one or more of:

b) predicting survival time of individuals with osteosarcoma;

d) predicting the tumor microenvironment of an individual with osteosarcoma;

the tumor microenvironment in the individual with the osteosarcoma is selected from the group consisting of a predictive immune score and the number of naive B cells in the tumor immune microenvironment.

3. The use according to claim 1 or 2, wherein the kit contains instructions describing: detecting the expression quantity of a specific gene in the tumor tissue, and substituting the expression quantity into a risk scoring formula: risk score = (0.946 × ABCB6 gene tumor tissue expression) + (0.413 × P4HA1 gene tumor tissue expression) + (0.435 × STC2 gene tumor tissue expression), and the risk score is calculated.