A set of biomarkers for risk prediction of idiopathic pulmonary arterial hypertension and uses thereof

Abstract

The application belongs to the technical field of biotechnology, and discloses a set of biomarkers for predicting the risk of idiopathic pulmonary arterial hypertension and application thereof. The application integrates expression quantitative trait locus (eQTL) data and single-cell RNA sequencing (scRNA-seq) data sets from peripheral blood and lung tissue and whole genome association analysis (GWAS) summary statistics of idiopathic pulmonary arterial hypertension, uses Mendelian randomization algorithm (MR), genome enrichment analysis and immunofluorescence experiments to identify and accurately locate related genes, and for the first time discovers biomarkers related to the risk of idiopathic pulmonary arterial hypertension, which include at least one of CST7, HLA-B, HLA-E, FKBP1A and UBB. By detecting the above biomarkers, the risk of idiopathic pulmonary arterial hypertension can be effectively diagnosed.

Description

Technical Field

[0001] This invention belongs to the field of biotechnology, specifically relating to a set of biomarkers for predicting the risk of idiopathic pulmonary hypertension and their applications. Background Technology

[0002] Idiopathic pulmonary arterial hypertension (IPAH) is a progressive cardiovascular disease that can ultimately lead to right ventricular failure and death. Current clinical treatments primarily focus on lowering pulmonary artery pressure to prolong survival and improve patients' quality of life. However, due to the complex etiology of IPAH, which includes genetic factors, environmental factors, medication use, and abnormal immune states, despite significant research efforts, the treatment outcomes for IPAH remain poor, with a challenging prognosis, placing a heavy burden on society and families.

[0003] To date, 12 genes with high evidence levels and 5 genes with low evidence levels have been identified as being associated with pulmonary hypertension (PAH), among which bone morphogenetic protein type II receptor (BMPR2) mutations are common in hereditary PAH, accounting for 20% of sporadic cases. Although reports have identified pulmonary vascular endothelium as a key cell type in the promotion of PAH development by BMPR2 mutations, the potential association between BMPR2 deficiency and changes in NK cell homeostasis or disease immunity in PAH patients has received less attention. However, a reduction and dysfunction of natural killer (NK) cells have been preliminarily observed in PAH patients and animal models of the disease. T cells and NK cells may exacerbate inflammation in PAH patients by upregulating CCL5. However, the specific mechanisms by which NK cells regulate PAH patients are not fully elucidated.

[0004] Recent research has proposed using quantitative trait loci (eQTLs) of gene expression as functional mediators to investigate the underlying biological mechanisms of genetics in various diseases, including pulmonary hypertension. Measuring these biomarkers can reflect an individual's health status and may provide new insights into the impact of disease. Furthermore, monitoring downstream gene expression changes is crucial for the development of potential drug targets. Mendelian randomization (MR), a genetic research method that uses genetic variation to explore causal relationships, has been widely used to explore the etiology of complex diseases. By selecting single nucleotide polymorphisms (SNPs) associated with eQTLs as instrumental variables (IVs), the causal effect of gene expression levels on pulmonary hypertension can be inferred, which will help identify novel risk genes and pathogenesis mechanisms, and develop targeted gene therapy drugs. In addition, the rapid development of single-cell sequencing technology in recent years will allow for the exploration of which cell types genes play a major role, making clinical interventions more cell-type specific.

[0005] The link between genetic factors and idiopathic pulmonary arterial hypertension (IPH) is widely accepted in the scientific community, and natural killer cell dysregulation is also a contributing factor to IPH. However, potential risk genes for NK cell-mediated disease remain undiscovered. With breakthroughs in in vivo gene editing technology, identifying these potential risk genes and exploring and developing genetically supported IPH gene and cell therapies is urgently needed. Summary of the Invention

[0006] The first aspect of the present invention aims to provide the application of reagents for detecting biomarkers in the preparation of products for risk diagnosis of idiopathic pulmonary hypertension.

[0007] The second aspect of the present invention is to provide a product.

[0008] A third aspect of the present invention aims to provide a model operation module for risk diagnosis of idiopathic pulmonary hypertension.

[0009] The fourth aspect of the present invention aims to provide a detection system for risk diagnosis of idiopathic pulmonary hypertension.

[0010] The fifth aspect of this invention is to provide an electronic device.

[0011] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A first aspect of the invention provides the use of reagents for detecting biomarkers in the preparation of products for risk diagnosis of idiopathic pulmonary hypertension; said biomarkers include at least one of CST7, HLA-B, HLA-E, FKBP1A, and UBB, said biomarkers being NK cell-derived biomarkers.

[0012] In some embodiments of the present invention, the reagent includes a reagent for quantitatively detecting the content of the biomarker.

[0013] In some embodiments of the present invention, the reagents include reagents for detecting biomarkers at the gene or protein level.

[0014] In some embodiments of the present invention, the reagents for detecting biomarkers at the protein level 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).

[0015] In some embodiments of the present invention, the reagent for detecting biomarkers at the gene level is selected from reagents of one or more detection methods from the group consisting of: DNA sequencing, RNA sequencing, RNA-in situ hybridization, digital PCR, and quantitative real-time PCR.

[0016] In some embodiments of the present invention, the product includes a test kit, a test chip, or a test strip.

[0017] In some embodiments of the present invention, the test sample for the product is NK cells isolated from the peripheral blood of the subject.

[0018] A second aspect of the present invention provides a product comprising a reagent for detecting biomarkers, said biomarkers including at least one of CST7, HLA-B, HLA-E, FKBP1A, and UBB.

[0019] In some embodiments of the present invention, the reagents for detecting biomarkers include reagents for detecting biomarkers at the gene level or protein level.

[0020] In some embodiments of the present invention, the reagents for detecting biomarkers at the protein level 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).

[0021] In some embodiments of the present invention, the reagent for detecting biomarkers at the gene level is selected from reagents of one or more detection methods from the group consisting of: DNA sequencing, RNA sequencing, RNA-in situ hybridization, digital PCR, and quantitative real-time PCR.

[0022] In some embodiments of the present invention, the reagents for detecting biomarkers at the gene level are selected from reagents used for RNA sequencing.

[0023] A third aspect of the present invention provides a model running module for risk diagnosis of idiopathic pulmonary hypertension, the model running module including a computing component; The computing component performs calculations on the biomarker data in the first aspect of the present invention.

[0024] A fourth aspect of the present invention provides a detection system for risk diagnosis of idiopathic pulmonary hypertension, the detection system comprising a parameter acquisition device and a data processing device; The parameter acquisition device quantitatively acquires the relative content of the biomarkers in the first aspect of the present invention in the subject sample; The data processing device determines whether the condition is idiopathic pulmonary hypertension based on data obtained from the parameter acquisition device.

[0025] A fifth aspect of the present invention provides an electronic device including a storage device, a processor, and a computer program stored in the storage device and executable on the processor, wherein the computer program stored in the storage device and executable on the processor includes the detection system of the fourth aspect of the present invention.

[0026] The beneficial effects of this invention are: This invention integrates quantitative trait locus (eQTL) data from peripheral blood and lung tissue with single-cell RNA sequencing (scRNA-seq) datasets and genome-wide association study (GWAS) data for idiopathic pulmonary hypertension (IPH). Using Mendelian randomization (MR), genomic enrichment analysis, and immunofluorescence assays, it identifies and precisely locates relevant genes, and for the first time discovers biomarkers associated with the risk of IPD, including at least one of CST7, HLA-B, HLA-E, FKBP1A, and UBB. Detection of these biomarkers can effectively diagnose the risk of IPD.

[0027] Specifically, multi-omics analysis revealed that CST7, HLA-B, HLA-E, FKBP1A, and UBB are associated with the risk of idiopathic pulmonary hypertension (IPH). Single-cell sequencing analysis showed that CST7, FKBP1A, and UBB play important roles in NK cells in both humans and rats. Immunofluorescence of these proteins in classic limonene rat models and SuHx rat models validated these findings. These discoveries elucidate the pathogenic genes (CST7, HLA-B, HLA-E, FKBP1A, and UBB) and their NK cells in IPH, thus providing intervention targets for in vivo gene editing of pathogenic NK cells. This invention not only reveals novel pathogenic genes in disease-causing NK cells but also identifies new gene and cell therapy targets for in vivo gene editing and clinical intervention of NK cells. Attached Figure Description

[0028] The present invention will be further described below with reference to the accompanying drawings and embodiments, wherein: Figure 1 This document outlines the research design and workflow for this invention (an overall overview of the entire research).

[0029] Figure 2Quality control for single-cell sequencing data; AC represents the quality control of human single-cell sequencing datasets GSE169471 and GSE185479 (A is single-cell quality control, showing the number of cells, genes, and sequencing depth for each sample; B is the left graph showing the relationship between cell sequencing depth and mitochondrial content, and the right graph showing the relationship between sequencing depth and gene number, both of which are positively correlated; C is a feature variance plot showing genes with significant differences between cells); DF represents the quality control of mouse single-cell sequencing dataset GSE154959 (D is single-cell quality control, showing the number of cells, genes, and sequencing depth for each sample; E is the left graph showing the relationship between cell sequencing depth and mitochondrial content, and the right graph showing the relationship between sequencing depth and gene number, both of which are positively correlated; F is a feature variance plot showing genes with significant differences between cells).

[0030] Figure 3 This section describes the grouping and integration of the human PAH single-cell database. A represents the gene Harmony integration diagram; B represents the variance ranking diagram of each principal component (PC); C represents the principal component analysis (PCA) results and the distribution of principal components (PC), with dots representing cells and colors representing samples; D represents the important principal components obtained based on PCA, using the tSNE algorithm to divide the cells into 19 clusters.

[0031] Figure 4 This section describes the grouping and integration of the mouse PAH single-cell database. A represents the gene Harmony integration diagram; B represents the variance ranking diagram of each principal component (PC); C represents the principal component analysis (PCA) results and the distribution of principal components (PC), with dots representing cells and colors representing samples; D represents the important principal components obtained based on PCA, where the cells are divided into 22 clusters using the tSNE algorithm.

[0032] Figure 5 This section describes single-cell sample annotation and gene function analysis; where A represents the cell annotation status of 19 clusters (annotating the 19 clusters into 7 cell types: endothelial cells, monocytes, macrophages, and CD8+). + (T cells, NK cells, epithelial cells, and fibroblasts); B represents the difference in the proportion of the seven cell types between the two groups; C represents the annotation of the marker genes for the seven cell types.

[0033] Figure 6 To perform GO and KEGG enrichment analysis on 140 differentially expressed genes using the Metascape database.

[0034] Figure 7 A PPI network for 140 differentially expressed genes.

[0035] Figure 8Annotations for single-cell samples from the mouse single-cell dataset GSE154959 are provided. A represents the cell annotations for 22 clusters (the 22 clusters are annotated as 9 cell types: epithelial cells, γδT cells (Tgd), endothelial cells, mast cells, natural killer cells (NK cells), monocytes, fibroblasts, macrophages, and neutrophils)); B represents the difference in the proportion of the 9 cell types between the two groups; and C represents the annotations of the marker genes for the 9 cell types.

[0036] Figure 9 The chart shows a scatter plot of CST7 gene and pulmonary hypertension MR analysis, as well as a forest plot of the corresponding CST7 gene SNPs using the Leaveout test. Different colors represent different statistical methods, and the slope of the lines represents the causal effect of each method.

[0037] Figure 10 The scatter plot shows the relationship between the FKBP1A gene and pulmonary hypertension in MR analysis, and the forest plot shows the Leaveout test results for the corresponding FKBP1A gene SNPs. Different colors represent different statistical methods, and the slope of the lines represents the causal effect of each method.

[0038] Figure 11 The scatter plot shows the relationship between the HLA-B gene and pulmonary hypertension in MR analysis, and the forest plot shows the Leaveout test results for the corresponding HLA-B gene SNPs. Different colors represent different statistical methods, and the slope of the lines represents the causal effect of each method.

[0039] Figure 12 The scatter plot shows the relationship between the HLA-E gene and pulmonary hypertension under MR analysis, and the forest plot shows the Leaveout test results for the corresponding HLA-E gene SNPs. Different colors represent different statistical methods, and the slope of the lines represents the causal effect of each method.

[0040] Figure 13 The chart shows a scatter plot of the UBB gene and pulmonary hypertension MR analysis, as well as a forest plot of the corresponding UBB gene SNPs using the Leaveout test. Different colors represent different statistical methods, and the slope of the lines represents the causal effect of each method.

[0041] Figure 14To explore the clinical predictive value of key genes in multi-omics studies; A represents the Pearson correlation among 29 types of immune cells, with blue indicating negative correlation and red indicating positive correlation; B represents the difference in immune cell content between the control group and the disease group, with blue representing the control group and yellow representing the disease group; C represents the correlation between 5 key genes and immune cell content; DH represents the Pearson correlation between 5 key genes and immune factors, representing chemokines (D), immunosuppressants (E), immunostimulants (F), MHC (G), and receptors (H), respectively.

[0042] Figure 15 The correlation analysis is performed between key genes and disease-related genes. Among them, A is an overview of the expression of 5 key genes in various cell types; B is the expression level of 5 key genes in different cells; C is the Pearson correlation analysis between 5 key genes and disease genes, with blue indicating negative correlation and red indicating positive correlation; D is the co-expression of 5 key genes and multiple immune genes at the single-cell level (correlation threshold r>0.5).

[0043] Figure 16 To explore the specific signaling mechanisms associated with these key genes, GO and KEGG pathway analyses were conducted to investigate the specific signaling pathways involved in these five key genes.

[0044] Figure 17 The figures show immunofluorescence staining of typical plexiform lesions and routine indicators of the classic rat MCT model. A shows double immunofluorescence staining of Cst7 and Prf1 in typical PAH plexiform lesions (scale bar: 20 µm); B shows double immunofluorescence staining of Fkbp12 (Fkbp1a) and Ccl4 in typical PAH plexiform lesions (scale bar: 20 µm); C shows double immunofluorescence staining of Ubb and Slpi in typical PAH plexiform lesions (scale bar: 20 µm). The arrows in the figures indicate typical double-stained cells in plexiform lesions.

[0045] Figure 18 The images show the immunofluorescence staining results of typical plexiform lesions in the classic mouse SuHx model. In the figures, A shows double immunofluorescence staining of Cst7 and Prf1 in typical plexiform lesions of PAH (scale bar: 2020 µm); B shows double immunofluorescence staining of Fkbp12 (also known as Fkbp1A) and Ccl4 in typical plexiform lesions of PAH (scale bar: 20 µm); and C shows double immunofluorescence staining of Ubb and Slpi in typical plexiform lesions of PAH (scale bar: 20 µm). Typical double-staining positive cells in the plexiform lesions are marked with arrows.

[0046] Figure 19 This is a potential genetic intervention map of the target gene; where A is the ceRNA network of the key gene; and B is the targeting compound of the key gene in the DGIDB database.

[0047] Figure 20 The area under the ROC curve for the target gene (CST7 gene, FKPB1A gene, HLA-B gene, HLA-E gene, or UBB gene) in the mRNA of peripheral blood of PAH patients. Detailed Implementation

[0048] 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.

[0049] Unless otherwise specified in the examples, the procedures should be performed under standard conditions or conditions recommended by the manufacturer. Reagents or instruments whose manufacturers are not specified are all commercially available products.

[0050] The features and performance of the present invention will be further described in detail below with reference to embodiments.

[0051] Example Materials and Methods 1. Blood sample Both patients and controls were recruited from the Women and Children's Medical Center of Guangzhou Medical University, including a control group and an idiopathic pulmonary arterial hypertension (IPAH) group. All patients were clinically diagnosed with IPAH by two attending physicians, meeting the typical clinical features of IPAH (excluding other diseases), according to the 2022 ESC / ERS guidelines for the diagnosis and treatment of pulmonary hypertension. The control group consisted of healthy children undergoing routine checkups. All patients (5 IPAH patients and 5 controls) were fully informed of their participation in a medical research project and completed the informed consent process before treatment. This study has been approved by the Ethics Review Committee of Guangzhou Women and Children's Medical Center.

[0052] 2. Extraction of peripheral blood mononuclear cells The peripheral blood mononuclear cell extraction protocol follows the established method described on the website as “Isolation of whole mononuclear cells from peripheral blood” (https: / / bio-protocol.org / en / bpdetail?id=939&type=0).

[0053] 3. RNA extraction Using TRIzol TM (Invitrogen) TM 15596026) and Direct-zol TMRNA Miniprep (zymoR2050) extracts RNA from cells.

[0054] 4. RNA Quantification and Quality Control RNA purity and concentration were determined using a NanoDrop One (Thermo, USA) and a Qubit 4.0 Fluorometer (Thermo, USA). RNA integrity was determined using a Qsep100 (Bioptic, China).

[0055] 5. Construction of transcriptome sequencing libraries Sequencing libraries were constructed using the Hieff NGS® Ultima Dual-mode mRNA Library Prep Kit (Cat#12309, China). Library quality was evaluated using a Qubit dsDNA HS Fluorometer (Thermo, USA) and a Bioptic Qsep100 system.

[0056] 6. Sequencing The library was sequenced on MGI SEQ-T7, generating 150bp paired-end reads.

[0057] 7. Data Analysis - Quality Control The raw reads in FastQ format are first processed using FastP v0.23.0. In this step, clean reads are obtained by removing reads containing headers, low-quality reads, and reads with a trimmed length less than 36bp. Q20, Q30, and GC are also calculated. All downstream analyses are based on this high-quality clean data.

[0058] 8. Reads aligned to the reference genome The reference genome and gene model annotation files can be downloaded directly from the Ensembl website. A reference genome index is built using Hisat2 v2.2.1, and the clean data is aligned to the reference genome using Hisat2 v2.2.1 to generate .sam or .bam files containing the alignment results.

[0059] 9. Quantitative analysis of gene expression levels Gene expression levels were quantified for each sample using the featureCounts v2.0.1 software from the subRead package, and read counts for each gene were obtained. The results from all samples were then merged to obtain an expression matrix for all samples. Based on gene length and the read count aligned to that gene, the FPKM (number of fragments per kilobase model per million fragments) and TPM (number of transcripts per kilobase per million transcripts) were calculated for each gene.

[0060] 10. Differential Expression Analysis Batch effects between different batches of samples were removed using the ComBat_seq function in the sva v3.42.0 R package. Differential expression analysis between the two conditions / groups was performed using the DESeq2 v1.34.0 R package. DESeq2 provides a statistical procedure for differential expression of digital gene expression data based on a negative binomial distribution model. The obtained p-values ​​were corrected using the Benjamini and Hochberg methods to control for false detection rates. Corrections were identified by DESeq2. P Genes with a value <0.05 are differentially expressed genes.

[0061] 11. Data Acquisition Exposure data: eQTL data comes from the eQTLGen consortium database (https: / / www.eqtlgen.org). The eQTLGen consortium aims to study the genetic structure of gene expression in blood, containing trait-related single nucleotide polymorphisms (SNPs) from 31,864 individuals to understand the genetic basis of complex traits. The large-scale eQTLGen project is currently in its second phase, focusing on large-scale genome-wide meta-analysis in blood.

[0062] Outcome data: Participants in the outcome-related GWAS studies selected for this invention were primarily of European descent. Pooled outcome data were obtained from the FinnGen biobank database (finn-b-I9_HYPTENSPUL_EXNONE). In the FinnGen study, pulmonary hypertension cases were defined using the corresponding International Classification of Diseases (ICD) codes. There were 125 cases of pulmonary hypertension and 218,667 controls.

[0063] RNA-seq data: The GEO database (https: / / www.ncbi.nlm.nih.gov / geo / ), short for Gene Expression Comprehensive Database, is a gene expression database created and maintained by the National Center for Biotechnology Information (NCBI). Pulmonary arterial hypertension (PAH) related data for GSE169471 were downloaded from the NCBI GEO public database for single-cell association analysis, totaling 11 samples (8 disease samples and 3 control samples). PAH related data for GSE185479 were downloaded from the NCBI GEO public database for single-cell association analysis, totaling 6 samples (3 disease samples and 3 control samples). The Series Matrix File data for GSE131793 was downloaded from the NCBI GEO public database. The annotation platform was GPL6244. Data from 20 PAH patients with complete expression profiles and phenotypic information were extracted (10 disease samples and 10 control samples). 10 locally tested data were also included (5 disease samples and 5 control samples).

[0064] Relevant gene set: The immune-related gene set (immune) used in this analysis was obtained from the GeneCard database (https: / / www.genecards.org).

[0065] 12. Single-cell analysis First, the inventors used the Seurat package to read the expression profile data and filtered out abnormal samples based on their expression (nFeature_RNA > 500 and percent.mt < 5%). Next, the inventors standardized and normalized the data and performed PCA analysis. The optimal number of principal components (11) was determined using the ElbowPlot method. Then, t-SNE analysis was used to observe the positional relationships between different cell clusters. To annotate cell clusters, the inventors used the celldex package. This package can annotate cell clusters with cell types that play important roles in disease development. Finally, the inventors used FindAllMarkers to extract marker genes for each cell subtype from the single-cell expression profile. The logfc.threshold parameter was set to 1, and the min.pct parameter was set to 0.5 to screen for significantly differentially expressed genes. Unadjusted genes were selected. p Genes with a value (p_val_adj) less than 0.05 and an average log2 fold change (|avg_log2FC|) greater than 1 are designated as unique marker genes for each cell subtype.

[0066] 13. Gene functional enrichment analysis Functional annotation of important gene sets was performed using the Metascape database (www.metascape.org) to comprehensively explore the functional relevance of gene sets. Gene Ontology (GO) analysis and KEGG pathway analysis were performed on specific genes. Minimum overlap ≥ 3 and p A value ≤ 0.01 is considered statistically significant.

[0067] 14. PPI Network Construction Gene lists were uploaded to the STRING database (https: / / string-db.org) to explore the PPI network of genes, with a confidence score >0.4 as the screening criterion. The generated PPI network was further visualized using Cytoscape software to analyze the role of genes in the development and progression of pulmonary hypertension.

[0068] 15. Mendelian randomization analysis The MR Base database (http: / / app.mrbase.org / ) contains extensive aggregate statistics from hundreds of GWAS studies. Result IDs filtered through the MR Base database are used to extract relevant causal relationships in eQTLs from the aggregate GWAS data (https: / / gwas.mrcieu.ac.uk / ), and significance thresholds are selected for each gene at each locus. P SNPs with p < 1e-8 were used as potential instrumental variables (IVs). The linkage disequilibrium (LD) among SNPs was calculated. For SNPs with R² < 0.001 (clustering window size = 10,000 kb), only SNPs with p² < 5e-8 were retained. The reliability of the causal relationships was assessed using various statistical methods, including inverse variance weighted regression (IVW, combining meta-analysis with Wald estimation for each SNP), MR-Egger regression (based on the assumption that instrument strength is independent of direct effects (InSIDE)), weighted median regression (allowing for correct estimation of causal relationships even with up to 50% IV invalidity), and weighted modal regression (which has stronger causal effect detection capability, lower bias, and lower Type I error rate compared to MR-Egger regression). (If there is only one SNP in the causal relationship, only the Wald ratio is used). This yielded a comprehensive estimate of the overall impact of all cis- and some cross-regional gene expression on whole blood pulmonary hypertension. Finally, the selected causal relationships were validated and analyzed using a leave-one-out method.

[0069] 16. Analysis of immune cell infiltration The ssGSEA method is a widely used approach for assessing immune cell types in the microenvironment. This method distinguishes 29 human immune cell phenotypes, including T cells, B cells, and NK cells. This invention uses the ssGSEA function in the R package to analyze expression profile data, infer the relative proportions of the 29 immune infiltrating cells, and performs Pearson correlation analysis on gene expression and immune cell abundance.

[0070] 17. GSEA Enrichment Analysis GSEA analysis uses a predefined gene set, ranking key genes according to their differential expression levels in two sample classes, and then examines whether the predefined gene set is enriched at the top or bottom of the ranking list. This invention uses GSEA to compare signaling pathway differences between high-expression and low-expression groups, exploring the molecular mechanisms of core genes in the two patient groups. The number of permutations is set to 1000, and the permutation type is set to phenotype.

[0071] 18. DGIDB Potential Therapeutic Drug Prediction The Drug-Gene Interaction Database (DGIdb) (https: / / www.dgidb.org) is a database of drug targets, genomes, and drug-gene interactions. The inventors searched DGIdb to predict potential drugs or molecular compounds that interact with key genes. The gene-compound interaction network was visualized using Cytoscape software.

[0072] 19. Immunofluorescence staining Immunofluorescence staining followed classic methods; detailed steps can be found at the following website: https: / / www.servicebio.cn / goodsdetail?id=1679. The antibodies used in the experiment included: Santa Cruz Bio:FKBP12(H-5):sc-133067; Ubiquitin(P4D1):sc-8017; MIP-1beta(B7):sc-393441; Perforin 1(A-2):sc-373943; SLPI(A-11):sc-374575; biorbyt CST7:orb101860.

[0073] 20. Animal models All animals were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. These included the classic MCT rat model and the SuHx mouse model. All experimental groups of animals exhibited typical symptoms of pulmonary hypertension. All animals were housed by experienced keepers under SPF-grade conditions, with bedding changed twice weekly and ample food and water provided. The animal handlers were unaware of the grouping procedures.

[0074] 21. Statistical Analysis Reliable MR analysis is based on three premises: (1) the correlation assumption (instrumental variables are closely related to exposure but not directly to the outcome), (2) the independence assumption (instrumental variables cannot be related to confounding factors), and (3) the exclusivity assumption (instrumental variables can only affect the outcome through exposure. If an instrumental variable can affect the outcome through other pathways, then pleiotropic effects are considered to exist). This analysis used R language (version 4.0). All statistical tests were two-tailed. p A value <0.05 was considered statistically significant. For animal experiments, an unpaired t-test was used to determine whether there were statistically significant differences in the evaluation indicators, and statistical calculations were performed using GraphPadPrism software (version 8).

[0075] Experimental results 1. Preliminary processing of single-cell data This analysis included 17 IPAH tissue samples and was validated using a rodent single-cell database (MCT classic rat and SuHx rat). The analysis flowchart provides an overall overview of the entire project, such as... Figure 1 As shown in the figure. Only cells with nFeature_RNA > 500 and percent.mt < 5% in the expression profile were retained for this analysis, totaling 72,499 cells for subsequent analysis. This figure illustrates the gene expression in the sample ( Figure 2 (among A and B), and identified the five genes with the highest standardized variance ( Figure 2 (C), including LDHA, MT-ATP6, CYSTM1, COPS6, and ARHGDIA. Relevant information from the rodent sc-seq database is as follows... Figure 2 As shown in DF. 2. Single-cell sample subtype cluster analysis The inventors performed PCA dimensionality reduction analysis on 20 genes and found that they had different scores in different dimensions. Figure 3 (A). However, when performing PCA and Harmony dimensionality reduction analyses between samples, the overall differences between samples were not significant ( Figure 3 (C). The optimal number of principal components observed through the elbow plot is 11 ( Figure 3 (B), and 19 subtypes were obtained through t-distributed random neighborhood embedding (tSNE). Figure 3 (D). The rodent dataset GSE154959 was also processed using the same analytical methods as in Results 1 and 2. All results are as follows: Figure 4 As shown. 3. Annotation of cell cluster subtypes The inventors first annotated each subtype using the SingleR package in R and manually verified all markers for the 19 clusters. Based on literature review, they identified the subtypes as belonging to seven cell types: endothelial cells, monocytes, macrophages, and CD8+ cells. + T cells, NK cells, epithelial cells, and fibroblasts ( Figure 5 (A). Among them, NK cell subtypes showed the greatest difference ( Figure 5 (B). Construct a dot plot to display the gene markers annotated for all cell types (B). Figure 5 (C). The rodent dataset GSE154959 was also processed using the same method. All results are shown in [link to results]. Figure 8 AC. Finally, FindAllMarkers was used to extract a total of 140 NK cell marker genes from the single-cell expression profile (Table 1).

[0076] Table 1. List of partial marker genes for different cell types

[0077] 4. Genome functional analysis Further pathway analysis was performed on these 140 marker genes. Pathway analysis using the Metascape database showed that these genes were mainly enriched in natural killer cell-mediated cytotoxicity, lytic vacuoles, and lytic granules pathways. Figure 6 In addition, the inventors used Cytoscape software to perform protein-protein interaction network analysis on the genes in the dataset, and the results are as follows: Figure 7 As shown.

[0078] 5. Mendelian randomization analysis of NK cell marker genes To further identify key genes influencing IPAH, the inventors used aggregated statistical data from 218,792 pulmonary hypertension-related samples (control group: 218,667; case group: 125): finn-b-I9_HYPTENSPUL_EXNONE, and extracted causal relationships between 121 pairs of marker genes and outcomes using extract_instruments and extract_outcome_data (Table 2). Furthermore, Mendelian randomization analysis yielded causal relationships between 5 pairs of marker genes and corresponding positive eQTL outcomes. Figures 9-13 IVW p <0.05). The corresponding genes are CST7, HLA-B, HLA-E, FKBP1A, and UBB. The presence of the HLA-E gene (0.349; 0.157) 0.777; p=0.010) may be associated with a lower IPAH risk, while FKBP1A (1.732; 1.002) 2.995; p =0.049), CST7 (1.781; 1.053 3.013; p =0.031), UBB (2.785; 1.077 7.201; p =0.035) and HLA-B (3.658; 1.034) 12.946; p =0.044) was associated with a higher risk of IPAH. The inventors further conducted a sensitivity analysis on the causal relationships of these five genes to assess their reliability. The results showed that excluding any one SNP had no significant impact on the overall error range, indicating that the selected five causal pairs were stable. Figures 9-13 ).

[0079] Table 2. SNP locus information of some marker genes related to causal outcome.

[0080] 6. Multi-omics studies explore the clinical predictive value of key genes. The microenvironment is mainly composed of immune cells, extracellular matrix, various growth factors, and inflammatory factors, with NK cells also involved. By analyzing the relationship between expression levels and immune infiltration, the inventors further explored the potential molecular mechanisms by which the expression levels of key genes affect the progression of IPAH. The results showed that multiple pairs of immune factors were significantly correlated ( Figure 14 (A). Type I interferon response, type II interferon response, and parainflammatory were significantly higher in disease samples than in control samples. Figure 14 (B). The CST7 gene is significantly positively correlated with cell lysis activity and Th1 cells; the HLA-B gene is significantly positively correlated with CCR and mast cells; the HLA-E gene is significantly positively correlated with aDCs and CD8. + T cells showed a significant positive correlation; the FKBP1A gene showed a significant positive correlation with macrophages and neutrophils, while showing a significant negative correlation with B cells and checkpoints; the UBB gene showed a significant positive correlation with CD8. + T cells, cell lysis activity, etc. showed a significant positive correlation. Figure 14 (C). The inventors obtained the correlations between these key genes and different immune factors (including immune regulatory molecules, chemokines, and cell receptors) from the TISIDB database. The results showed that the key genes were significantly correlated with multiple immune factors (C). Figure 14 (DH).

[0081] 7. Studies on the expression levels of disease-related genes The inventors compared the expression levels of five key genes using the GeneCards database (https: / / www.genecards.org / ). Figure 15 The analysis included (A and B) and the correlation scores of the top 20 genes between them. Analysis of the expression levels of multiple genes showed that the expression levels of key genes were significantly correlated with the expression levels of various disease-related genes. Specifically, HLA-E showed a significant positive correlation with RPL5 (cor=0.658), and FKBP1A showed a significant negative correlation with CAV1 (cor=-0.763). Figure 15 (C). The expression of these five key genes and the top ten immune genes ranked by correlation score were analyzed at the single-cell level. Key genes were co-expressed with multiple immune genes at the single-cell level (correlation threshold r>0.5); Figure 15 (D). The positive correlation between key gene expression and cytokine / exhaustion factor scores indicates that as the expression level of the key gene increases, the level of cytokine / exhaustion factors also increases. This association may suggest that the gene plays an important role in regulating cytokine / exhaustion factors and may be involved in a range of biological processes, such as inflammatory responses, immune regulation, and cell proliferation.

[0082] 8. Key gene-related specific signaling pathways Next, the inventors explored the specific signaling pathways associated with these five key genes and analyzed the impact of these key genes on disease progression-related pathways. The inventors screened out some highly significant pathways and presented them separately. GO-enriched pathways in the CST7 gene include mRNA transcription and RNA polymerase II-mediated mRNA transcription. The pathways enriched by KEGG include those involved in acute myeloid leukemia, antigen processing and presentation, etc.; the pathways enriched by GO in the HLA-B gene include molecular chaperone cofactor-dependent protein refolding and de novo protein folding, etc.; the pathways enriched by KEGG include the Fc epsilon RI signaling pathway and glycolysis / gluconeogenesis, etc.; the pathways enriched by GO in the HLA-E gene include cellular responses to cadmium ions and the formation of the stomatal structure of inner ear receptor cells, etc.; the pathways enriched by KEGG include ABC transporters and chemokine signaling pathways; the pathways enriched by the FKBP1A gene include the detection of biostimuli and the negative regulation of mitochondrial tissue, etc.; the pathways enriched by KEGG include Fc gamma R-mediated phagocytosis and glycosylphosphatidylinositol (GPI) anchored biosynthesis; the pathways enriched by GO in the UBB gene include glomerular epithelial development, etc.; the pathways enriched by KEGG include antigen processing and presentation and butyrate metabolism, etc. Figure 16 ).

[0083] 9. Immunofluorescence staining results of lung tissues from classic lily alkaloid model and SuHx model mice validated and revealed the consistency of cellular gene targets. The rat model of limonene and the SuHx mouse model are the most commonly used animal models in IPAH-related research. The inventors followed classic tissue fixation methods and performed immunofluorescence staining according to guidelines. NK cell markers (Prf1, Ccl4, and Slpi) and selected cell genes (Cst7, Fkbp12, and Ubb) were selected for double immunostaining of the tissues. White arrows indicate double-stained cells (…). Figure 17 The SuHx model also showed NK cell infiltration (). Figure 18 (AC). Double staining was observed in typical pathological changes—plexiform lesions, which also verified the inventors' above-mentioned results.

[0084] 10. Analysis of ceRNA networks and targeting compounds of key genes To identify potential gene intervention methods, the inventors conducted the following analysis: They analyzed these five key genes using the miRWalk and ENCORI databases to obtain their potential miRNA and lncRNA intervention targets. First, they extracted the mRNAs associated with these five mRNAs from the miRWalk database. miRNA pairings (a total of 890 miRNAs were obtained). The inventors retained only 72 pairs of mRNAs that could be detected by TargetScan or the miRDB database. miRNA relationships (containing 4 mRNAs and 71 miRNAs). Interacting lncRNAs were then predicted based on these miRNAs, resulting in a total of 1,033 interaction pairs (containing 16 miRNAs and 483 lncRNAs). Finally, a ceRNA network was constructed using Cytoscape (v3.9). Figure 19 (A). Furthermore, screening of the DGIDB database revealed the existence of multiple compounds targeting HLA-B, HLA-E, and FKBP1A in existing small molecule libraries. Figure 19 (B)

[0085] 11. Area under the ROC curve of the target gene in the mRNA of peripheral blood NK cells from PAH patients. To predict the sensitivity and specificity of the target gene, the inventors analyzed the area under the ROC curve based on mRNA in peripheral blood NK cells of clinical PAH patients. Figure 20 The AUCs of genes FKPB1A and HLA-E were both >0.8, indicating high predictive accuracy. While the AUCs of CST7, HLA-B, and UBB were between 0.7 and 0.8, and from... Figure 20 The results suggest that the expression of CST7 and UBB genes differs in a rodent model of pulmonary hypertension (the human leukocyte antigen HLA gene could not be verified in rodents). This indicates that the predictive accuracy of these genes for pulmonary hypertension remains relatively good and reliable.

[0086] 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. The application of reagents for detecting biomarkers in the preparation of risk diagnostic products for idiopathic pulmonary hypertension; wherein the biomarkers include at least one of CST7, HLA-B, HLA-E, FKBP1A and UBB, and the biomarkers are NK cell-derived biomarkers.

2. The application according to claim 1, characterized in that, The reagents include those for quantitatively detecting the content of the biomarkers.

3. The application according to claim 2, characterized in that, The reagents include those for detecting biomarkers at the gene or protein level.

4. The application according to claim 3, characterized in that, The reagents for detecting biomarkers at the protein level 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). or The reagents for detecting biomarkers at the gene level are selected from reagents of one or more detection methods from the group consisting of: DNA sequencing, RNA sequencing, RNA-in situ hybridization, digital PCR, and quantitative real-time PCR.

5. The application according to any one of claims 1-4, characterized in that, The products include test kits, test chips, or test strips.

6. The application according to any one of claims 1-4, characterized in that, The test sample for the product was NK cells isolated from the peripheral blood of the subject.

7. A product comprising a reagent for detecting a biomarker, said biomarker comprising at least one of CST7, HLA-B, HLA-E, FKBP1A, and UBB.

8. A model running module for risk diagnosis of idiopathic pulmonary hypertension, the model running module comprising a computing component; The computing component performs calculations on the data of the biomarkers described in any one of claims 1-6.

9. A detection system for risk diagnosis of idiopathic pulmonary hypertension, the detection system comprising a parameter acquisition device and a data processing device; in, The parameter acquisition device quantitatively acquires the relative content of the biomarkers described in any one of claims 1-6 in the subject sample; The data processing device determines whether the condition is idiopathic pulmonary hypertension based on data obtained from the parameter acquisition device.

10. An electronic device comprising a storage device, a processor, and a computer program stored on the storage device and executable on the processor, wherein the computer program stored on the storage device and executable on the processor comprises the detection system of claim 9.

Images