Gene combination for Alzheimer's disease detection and application thereof

Abstract

The invention discloses a gene combination for detecting Alzheimer's disease and application of the gene combination. The gene combination comprises any one or at least two of the following combinations: (1) a gene combination based on mRNA expression change; (2) gene combination based on variable shear change; and (3) a gene combination based on variable polyadenylation change. According to the invention, mRNA (messenger ribonucleic acid) is expressed, and altered splicing is carried out; the invention relates to a method for preparing a polyadenylation system, which comprises the following steps of: preparing a polyadenylation system (AS, AS) and a polyadenylation system (altered polyadenylation; according to the method, three transcriptional regulation and control layers (APA, APA) are subjected to conjoint analysis, so that the understanding of AD pathogenesis is increased, a part of problems related to the AD pathogenesis are answered, the influence of different risk genes on AD can be deeply understood, the auxiliary diagnosis of AD is realized, and a possible disease prevention or treatment method is determined.

Description

Technical Field

[0001] The present invention belongs to the technical field of molecular biology, and relates to a gene combination for Alzheimer's disease detection and its application, in particular to a risk gene combination for Alzheimer's disease based on transcriptional regulatory changes and its application. Background Art

[0002] Alzheimer's disease (AD) is a progressive neurodegenerative disease. The pathological features of AD are senile plaques and neurofibrillary tangles. The main component of senile plaques is extracellular amyloid protein precipitation, and the main component of neurofibrillary tangles is intracellular hyperphosphorylated Tau protein. AD has high clinical heterogeneity. The typical pathological manifestations are progressive memory impairment, cognitive function decline, loss of daily living ability, and at the same time accompanied by neuropsychiatric symptoms such as personality changes, which seriously affect social interaction and daily life. Therefore, it is of great significance to increase the research on AD.

[0003] RNA-seq is a technology that can detect the RNA map across the whole genome, including mRNA expression, alternative splicing (AS), and alternative polyadenylation (APA), etc. At present, it has been widely used in the research of the pathogenesis of AD. RNA-seq can detect the expression patterns of genes that are transcribed, which can enhance our understanding of the disease. For example, an mRNA expression profile of the amygdala and cingulate cortex of AD patients revealed several pathogenic mechanisms related to AD, including chronic inflammation, oxidative stress, mitochondrial dysfunction, impaired proteasome activity, abnormal phosphorylation, and cell cycle signal dysregulation. In addition, RNA-seq can also be used to detect abnormal AS in AD patients. For example, pathological splicing of MAPT will lead to abnormal ratios of 3R-tau / 4R-tau isoforms, and abnormal splicing of APP will lead to an increase in the production of Aβ42. These abnormal regulations will promote the development of AD. APA dysregulation may lead to a decrease in the mRNA expression efficiency of genes COX-2, MAPT, and APP, which may lead to the occurrence of AD.

[0004] Gene expression is a complex process that involves multiple transcriptional regulatory layers such as AS, RNA editing, APA, RNA interference, etc. The coordinated division of labor among these different transcriptional regulatory layers ensures the normal expression of genes. Currently, research on AD mainly focuses on one or two aspects of mRNA expression, AS, and APA. For example, a study including mRNA expression and AS showed that the expression levels of different AS transcripts of APOE in AD brain tissues are related to the use of promoters, and the expression levels of different AS transcripts can reflect the disease progression of AD. Another study using a similar method found that the DBI gene in AD patients tends to express shorter splicing isoforms with retained introns through AS. In another study analyzing mRNA expression and APA, it was found that different APA differences can better distinguish patients from the control group compared with changes in mRNA expression. However, there is currently no study on AD that simultaneously examines the three transcriptional regulatory layers.

[0005] In summary, it is of great significance to conduct a more comprehensive study on the pathogenesis of AD, deeply understand the impact of different risk genes on AD, use it for the auxiliary diagnosis of AD, and determine possible disease prevention or treatment methods. Summary of the Invention

[0006] In view of the deficiencies of the prior art and the actual needs, the present invention provides a gene combination for the detection of Alzheimer's disease and its application, applying the comprehensive study of mRNA expression, AS, and APA at three levels to the research of AD to explore the impact of these three transcriptional layers on the pathogenesis of AD, and at the same time mining the risk genes of AD.

[0007] To achieve the above object, the present invention adopts the following technical solutions:

[0008] In the first aspect, the present invention provides a gene combination for the detection of Alzheimer's disease, and the gene combination includes any one or at least two combinations of the following combinations:

[0009] (1) A gene combination based on mRNA expression changes, which includes: YWHAH, UCHL1, BEX1, STMN2, PEBP1, VSNL1, YWHAG, CALM3, CALM1, and BASP1;

[0010] (2) A gene combination based on AS changes, which includes ENC1, NEFL, NPTN, GABRG2, STXBP1, RTN3, SNAP25, DNM1, SLC4A10, and EPB41L3;

[0011] (3) A gene combination based on APA changes, which includes CALM1, RPL15, FAIM2, SUB1, CLTB, CST3, and SAP18.

[0012] In the present invention, by jointly analyzing the three transcriptional regulatory layers of mRNA expression, AS, and APA, the understanding of the pathogenesis of AD is increased, some questions regarding the pathogenesis of AD are answered, and the effects of different risk genes on AD can also be deeply understood, which can be used for the auxiliary diagnosis of AD and to determine possible disease prevention or treatment methods.

[0013] The present invention explores combinations of AD risk genes respectively based on changes in mRNA expression, AS, and APA, including combinations of AD risk genes based on changes in mRNA expression, where the expression levels of the genes in the gene combination all decrease in AD patients; combinations of AD risk genes based on changes in AS, where the expression levels of the genes in the gene combination all decrease in AD patients; and combinations of AD risk genes based on changes in APA, where the expression levels of the genes in the gene combination all decrease in AD patients.

[0014] In a second aspect, the present invention provides the use of the gene combination for detecting Alzheimer's disease described in the first aspect or a substance for detecting the same in the preparation of a product for detecting Alzheimer's.

[0015] Preferably, the substance for detecting the gene combination for detecting Alzheimer's disease includes any one or a combination of at least two of a substance for detecting changes in the mRNA expression of the gene, a substance for detecting changes in AS of the gene, or a substance for detecting changes in APA of the gene.

[0016] In a third aspect, the present invention provides a product for detecting Alzheimer's disease, and the product includes a substance for detecting the gene combination for detecting Alzheimer's disease described in the first aspect.

[0017] In a fourth aspect, the present invention provides the use of the gene combination for detecting Alzheimer's disease described in the first aspect or a substance for detecting the same in the preparation of a device for detecting Alzheimer's.

[0018] In a fifth aspect, the present invention provides a device for detecting Alzheimer's disease, and the device includes a detection unit and an evaluation unit;

[0019] The detection unit is used to perform the following:

[0020] Detect changes in the mRNA expression, AS, or APA of the gene combination for detecting Alzheimer's disease described in the first aspect in a test sample of a subject;

[0021] The evaluation unit is used to perform the following:

[0022] Judge whether it is positive for Alzheimer's disease according to the results detected by the detection unit.

[0023] Preferably, the detection method in the detection unit includes transcriptome sequencing.

[0024] Preferably, the judgment method in the evaluation unit includes:

[0025] The criteria for Alzheimer's disease positive are as follows: at the mRNA expression level, the expression levels of genes YWHAH, UCHL1, BEX1, STMN2, PEBP1, VSNL1, YWHAG, CALM3, CALM1, and BASP1 all decrease in AD patients; at the AS level, the expression levels of genes ENC1, NEFL, NPTN, GABRG2, STXBP1, RTN3, SNAP25, DNM1, SLC4A10, and EPB41L3 all decrease in AD patients; at the APA level, the expression levels of genes CALM1, RPL15, FAIM2, SUB1, CLTB, CST3, and SAP18 all decrease in AD patients.

[0026] Specifically, the gene YWHAH decreases by at least 0.36 times, the gene UCHL1 decreases by at least 0.37 times, the gene BEX1 decreases by at least 0.43 times, the gene STMN2 decreases by at least 0.45 times, the gene PEBP1 decreases by at least 0.25 times, the gene VSNL1 decreases by at least 0.44 times, the gene YWHAG decreases by at least 0.37 times, the gene CALM3 decreases by at least 0.37 times, the gene CALM1 decreases by at least 0.26 times, and the gene BASP1 decreases by at least 0.38 times; at the AS level, the gene ENC1 decreases by at least 0.48 times, the gene NEFL decreases by at least 0.48 times, the gene NPTN decreases by at least 0.35 times, the gene GABRG2 decreases by at least 0.46 times, the gene STXBP1 decreases by at least 0.37 times, the gene RTN3 decreases by at least 0.34 times, the gene SNAP25 decreases by at least 0.46 times, the gene DNM1 decreases by at least 0.3 times, the gene SLC4A10 decreases by at least 0.41 times, and the gene EPB41L3 decreases by at least 0.26 times; at the APA level, the gene CALM1 decreases by at least 0.26 times, the gene RPL15 decreases by at least 0.22 times, the gene FAIM2 decreases by at least 0.22 times, the gene SUB1 decreases by at least 0.35 times, the gene CLTB decreases by at least 0.17 times, the gene CST3 decreases by at least 0.004 times, and the gene SAP18 decreases by at least 0.08 times.

[0027] In a sixth aspect, the present invention provides the use of the gene combination for detecting Alzheimer's disease described in the first aspect as a target in screening drugs for preventing or treating Alzheimer's disease.

[0028] Preferably, the screening includes determining whether a candidate drug can be used for preventing or treating Alzheimer's disease based on the effects on the gene combination for Alzheimer's disease detection before and after the use of the candidate drug.

[0029] Compared with the prior art, the present invention has the following beneficial effects:

[0030] By jointly analyzing three transcriptional regulatory layers of mRNA expression, AS, and APA, the present invention mines multiple risk genes as a combination, and uses the changes of genes in mRNA expression, AS, and APA as detection indicators for the auxiliary judgment of Alzheimer's disease symptoms. It has the characteristics of high detection accuracy, convenience, and speed, and has important clinical guiding significance for the auxiliary diagnosis of AD-related indicators. It can also be used as a target in screening drugs for treating or preventing AD. Description of the Drawings

[0031] Figure 1 It is a result graph of factors affecting the distribution of temporal lobe mRNA expression data. Among them, Figure A is the PCA graph of temporal lobe mRNA expression data before removing batch effects; Figure B is the multi-factor variance analysis of temporal lobe mRNA expression data before removing batch effects; Figure C is the PCA graph of temporal lobe mRNA expression data after removing batch effects; Figure D is the multi-factor variance analysis of temporal lobe mRNA expression data after removing batch effects; (F test: P<0.1, **P<0.01, ***P<0.001);

[0032] Figure 2 It is a result graph of factors affecting the distribution of temporal lobe AS data. Among them, Figure A is the PCA graph of temporal lobe AS data before removing batch effects; Figure B is the multi-factor variance analysis of temporal lobe AS data before removing batch effects; Figure C is the PCA graph of temporal lobe AS data after removing batch effects; Figure D is the multi-factor variance analysis of temporal lobe AS data after removing batch effects; (F test: P<0.1, **P<0.01, ***P<0.001);

[0033] Figure 3 It is a result graph of factors affecting the distribution of temporal lobe APA data. Among them, Figure A is the PCA graph of temporal lobe APA data before removing batch effects; Figure B is the multi-factor variance analysis of temporal lobe APA data before removing batch effects; Figure C is the PCA graph of temporal lobe APA data after removing batch effects; Figure D is the multi-factor variance analysis of temporal lobe APA data after removing batch effects; (F test: P<0.1, **P<0.01, ***P<0.001);

[0034] Figure 4This is a combined analysis result diagram of the mRNA expression, AS, and APA transcriptional regulation layers in the temporal lobe. Among them, Figure A is a bar chart of the total contribution of all factors and a heat map of the contribution of the top five factors; Figure B is a distribution map of AD patients and healthy controls in Factor 1; Figure C is the weight distribution of the top 10 features with the largest weights screened from the three transcriptional regulation layers of mRNA difference, AS, and APA in Factor 1; Figure D is the KEGG enrichment analysis of the top 10 features with the largest weights screened from the three transcriptional regulation layers of mRNA difference, AS, and APA in Factor 1, (Wilcoxon test: ***P<0.001);

[0035] Figure 5 This is a ROC analysis result diagram. Specific implementation manners

[0036] To further elaborate on the technical means and effects adopted by the present invention, the present invention will be further described below in conjunction with embodiments and drawings. It can be understood that the specific implementation manners described herein are only used to explain the present invention, rather than limiting the present invention.

[0037] For those not specifying specific techniques or conditions in the embodiments, they shall be in accordance with the techniques or conditions described in the literature in this field or in accordance with the product specifications. For reagents or instruments not indicating the manufacturer, they are all conventional products that can be obtained through regular channels.

[0038] Based on transcriptome data (RNA-seq), the present invention comprehensively studies the pathogenesis of AD from three transcriptional regulation layers of mRNA expression, AS, and APA, and simultaneously mines the risk genes of AD. The RNA-seq data is derived from the temporal lobe brain tissues of AD patients and healthy controls in the GEO database of NCBI. The screened data is classified according to the health status, and the classified data is respectively subjected to mRNA expression difference analysis, AS difference analysis, and APA difference analysis, and then the MOFA2 is used to perform a combined analysis on the analysis results of the three levels, so as to screen out the risk genes related to AD.

[0039] Example 1

[0040] Retrieve the RNA-seq data of AD patients and healthy controls from the GEO database of the public database NCBI. The retrieval conditions are set as: ((Homo sapiens[Organism]) AND (Expression profiling by highthroughput sequencing[DataSet Type]) AND (Alzheimer)). Screen the data according to the retrieved results. The requirements are: RNA-seq sequencing is performed on RNA extracted from the temporal lobe of the brains of AD patients or healthy controls. A total of 354 samples are counted, and the phenotypic data of these samples are also counted, including: age, location of the brain bank, CERAD, health status (AD patients / healthy controls), Braak stage, and gender. Then use the KNN algorithm to fill in the phenotypic data. Then collect and preprocess the data, including: use prefetch to download the data, use FastQC for quality control, use prinseq++ to filter the data, and use STAR to align the filtered sequences to the human genome (GRCh38) to generate BAM files. After obtaining the BAM files, perform mRNA differential expression analysis, AS differential analysis, and APA differential analysis on the data respectively.

[0041] Example 2

[0042] mRNA differential expression analysis: Use featureCount to count the BAM files. Before performing mRNA differential expression analysis, first perform PCA analysis and multivariate analysis of variance on the data of the temporal lobe. From the graph ([ Figure 1 ) drawn by PCA, it can be seen that there is a batch effect in the data of the temporal lobe (Figure A), and the results of the multivariate analysis of variance also show that the batch effect has a significant impact on the data of the temporal lobe (Figure B). Therefore, the batch effect needs to be removed before analysis. Use the Combat-seq function in the sva package to remove the batch effect, and perform PCA and multivariate analysis of variance on the data with the batch effect removed again. The results show that the batch effect has been removed ([ Figure 1 in Figures C and D). Use DESeq2 to perform differential gene analysis on the data with the batch effect removed, and screen the top 5000 P adj with the smallest P < 0.05 adj genes for joint analysis of MOFA2.

[0043] Example 3

[0044] AS differential analysis: The LeafCutter software was used to perform AS analysis on the BAM files. When performing AS analysis, the LeafCutter software first quantified the reads spanning each intron to quantify the intron usage across samples, and then clustered them to obtain the read counts. Before performing AS analysis, we first standardized the data using the voom function in the limma package of R language. After normalization, we performed multi-factor variance and PCA analysis on the normalized data and found that there was a batch effect in the AS data ( Figure 2 Figures A and B in Figure 2 Figures C and D in ). Therefore, the removeBatchEffect function in the limma package was used to remove the batch effect, and then multi-factor variance and PCA analysis were used to check whether the batch effect was removed. The results showed that the batch effect had been removed ( adj <0.05, and the top 5000 adj introns with the smallest P values were selected for joint analysis using MOFA2.

[0045] Example 4

[0046] APA differential analysis: We used the bedtools software to convert the BAM files into wig files and used the DaPars 0.9 software for APA analysis. The DaPars 0.9 software first used DaPars_Extract_Anno.py to extract the 3'UTR information of all genes from the genomic annotation file (bed file of GRCh38) to determine the distal polyA sites; then used DaPars_main.py to determine the proximal polyA sites and calculated the relative usage rates of the proximal and distal polyA sites based on the transcript expression levels of the proximal and distal polyA sites, that is, evaluated using the percentage of distal polyA site usage index (PDUI). Before performing APA differential analysis, we first used multi-factor variance and PCA analysis to evaluate the factors affecting the data, and found that there was a batch effect in the APA data ( Figure 3 Figures A and B in ), and then used the removeBatchEffect function in the limma package of R language to remove the batch effect; then multi-factor variance and PCA analysis were used to check whether the batch effect was removed. The results showed that the batch effect had been removed (Figure 3 (Figures C and D). Finally, the Wilcox test was used to analyze the inter-group differences in the data after removing the batch effects, and the P-values were corrected by the Benjamini-Hochberg method. The top 5000 transcripts with P adj < 0.05 in AD patients and healthy controls were selected for joint analysis with MOFA2. adj

[0047] Example 5

[0048] We used MOFA2 for joint analysis of three transcriptional regulatory layers: the top 5000 features (genes, introns, and transcripts) screened from mRNA expression difference analysis, AS difference analysis, and APA difference analysis were used for model training, and the training parameters were set as: drop_factor_threshold = -1, maxiter = 1000, and num_factors = 15. MOFA2 decomposes and reduces the dimension of the input matrix, generating a weight matrix between factors and samples and a weight matrix between factors and features. The 15 factors inferred by MOFA2 represent the main sources of variation in the three transcriptional layers. Generally speaking, APA accounts for the largest proportion in the total variation (R 2 = 0.678), followed by AS (R 2 = 0.556) and mRNA expression (R 2 = 0.457) ( Figure 4 Figure A); among the 15 factors, factor 1 shows the largest difference in interpretation, with the largest contribution from mRNA expression (R 2 = 0.362), followed by AS (R 2 = 0.316) and APA (R 2 = 0.239) ( Figure 4 Figure A). Therefore, we focused on factor 1 and found that there were significant distribution differences between AD patients and healthy controls in factor 1 ( Figure 4 Figure B). To quantify the contribution of each feature to each factor, MOFA2 assigned a weight to each feature, indicating the degree of correlation between the feature and the factor. After Z-score transformation, the weights of each feature were converted to between -1 and 1. We screened the top 10 features with the largest absolute value of weights in each transcriptional regulatory layer of factor 1 for KEGG enrichment analysis ( Figure 4 Figure C), and visualized the top five enriched pathways (P adj < 0.1), and found that these features were mainly related to AD and other neurodegenerative diseases ( Figure 4 Figure D).

[0049] Example 6

[0050] Perform ROC analysis on the top 10 features with the largest weights screened from the three transcriptional regulation layers of mRNA differential, AS, and APA. Adopt K-fold cross-validation, divide the original dataset into 6 parts, then take 5 parts as the training set, and the remaining 1 part as the validation set. Obtain a classification model through the training set. Use the glm function in R language to establish a logistic regression model, and then use the validation set to verify the model. Use the average value of the 6 validation sets as the index for drawing the ROC curve, calculate the AUC value of each ROC curve, and evaluate the prediction performance of the screened biomarkers for AD. It can be seen from the results that the combined features screened from each transcriptional regulation layer have good prediction for AD. The AUC values of the three transcriptional regulation layers of mRNA expression, AS, and APA are 0.848, 0.659, and 0.774 respectively( Figure 5 ). In summary, the present invention has discovered 27 AD risk genes at three transcriptional regulation levels, which are combinations of AD risk genes based on mRNA expression changes, including: YWHAH, UCHL1, BEX1, STMN2, PEBP1, VSNL1, YWHAG, CALM3, CALM1, and BASP1; combinations of AD risk genes based on AS changes, including ENC1, NEFL, NPTN, GABRG2, STXBP1, RTN3, SNAP25, DNM1, SLC4A10, and EPB41L3; and combinations of AD risk genes based on APA changes, including CALM1, RPL15, FAIM2, SUB1, CLTB, CST3, and SAP18, which can be used for the auxiliary diagnosis of AD and determine possible disease prevention or treatment methods.

[0051] The applicant declares that the present invention uses the above embodiments to illustrate the detailed method of the present invention, but the present invention is not limited to the above detailed method, that is, it does not mean that the present invention must rely on the above detailed method to be implemented. Those skilled in the art should understand that any improvement to the present invention, the equivalent replacement of each raw material of the product of the present invention, the addition of auxiliary components, and the selection of specific methods, etc., all fall within the protection scope and disclosure scope of the present invention.

Claims

1. A gene combination for Alzheimer's disease detection, characterized in that, The gene combinations include any one or at least two combinations of the following combinations: (1) Gene combinations based on mRNA expression changes, including: YWHAH, UCHL1, BEX1, STMN2, PEBP1, VSNL1, YWHAG, CALM3, CALM1, and BASP1; (2) Gene combinations based on alternative splicing changes, including: ENC1, NEFL, NPTN, GABRG2, STXBP1, RTN3, SNAP25, DNM1, SLC4A10, and EPB41L3; (3) Gene combinations based on alternative polyadenylation changes, including: CALM1, RPL15, FAIM2, SUB1, CLTB, CST3, and SAP18.

2. Use of the gene combination for Alzheimer's disease detection according to claim 1 or a substance for detecting the same in the preparation of a product for detecting Alzheimer's disease.

3. The use according to claim 2, characterized in that, The substances for detecting the gene combinations for Alzheimer's disease detection include any one or at least two combinations of substances for detecting gene mRNA expression, substances for detecting gene alternative splicing, or substances for detecting gene alternative polyadenylation.

4. A product for detecting Alzheimer's disease, characterized in that, The product includes a substance for detecting the gene combinations for Alzheimer's disease detection described in claim 1.

5. Use of the gene combination for Alzheimer's disease detection according to claim 1 or a substance for detecting the same in the preparation of a device for detecting Alzheimer's disease.

6. A device for detecting Alzheimer's disease, characterized in that, The device includes a detection unit and an evaluation unit; The detection unit is used to perform the following: Detect the mRNA expression, alternative splicing, or alternative polyadenylation of the gene combinations for Alzheimer's disease detection described in claim 1 in a test sample of a subject; The evaluation unit is used to perform the following: Judge whether it is positive for Alzheimer's disease according to the results detected by the detection unit.

7. The device for detecting Alzheimer's disease according to claim 6, characterized in that, The detection method in the detection unit includes transcriptome sequencing.

8. The device for detecting Alzheimer's disease according to claim 6 or 7, characterized in that, The judgment method in the evaluation unit includes: The criteria for being positive for Alzheimer's disease are as follows: at the mRNA expression level, the expression levels of genes YWHAH, UCHL1, BEX1, STMN2, PEBP1, VSNL1, YWHAG, CALM3, CALM1, and BASP1 all decrease in Alzheimer's disease patients; at the alternative splicing level, the expression levels of genes ENC1, NEFL, NPTN, GABRG2, STXBP1, RTN3, SNAP25, DNM1, SLC4A10, and EPB41L3 all decrease in Alzheimer's disease patients; at the alternative polyadenylation level, the expression levels of genes CALM1, RPL15, FAIM2, SUB1, CLTB, CST3, and SAP18 all decrease in Alzheimer's disease patients.

9. Use of the gene combination for Alzheimer's disease detection according to claim 1 as a target in the screening of drugs for preventing or treating Alzheimer's disease.

10. The use according to claim 9, characterized in that, The screening includes determining whether a candidate drug can be used for preventing or treating Alzheimer's disease based on the effects of the candidate drug on the gene combinations for Alzheimer's disease detection before and after use.

Images