Bioinformatics screening of breast cancer-related genes and potential drug resea

LIANG Xiao, LI Ya-lan, BAI Hao-tian, YANG Jing, WANG Rui✉

1.School of Pharmacy, Heilongjiang University of Traditional Chinese Medicine, Harbin 150040, China

2.Basic Medical College of Heilongjiang University of Traditional Chinese Medicine, Harbin 150040, China

Keywords:

ABSTRACT Objective:To search the the differentially expressed genes between breast cell carcinoma tissues and normal tissues by using bioinformatics technology, and the potential therapeutic drugs for breast cancer were identified, which can provide reference for clinical immune targeted therapy and drug therapy of breast cancer in the future.Methods:"Breast cancer"was searched by using Gene Expression Omnibus (GEO), and GSE79586 chip data was downloaded.The differentially expressed genes in the control group and the breast cancer model group were screened by using bio-communication technology and subjected to GO function analysis, KEGG pathway analysis, differential gene characteristic expression analysis and protein-protein interaction network (PPI) analysis, and the analysis results were further visualized.Prognosis analysis, related function prediction and immune infiltration analysis were performed using the GEPIA, GeneMANIA, and Timer2.0 databases, respectively.Finally,the compounds with potential therapeutic effects on breast cancer are identified through Connectivity Map (CMap).Western blotting and real-time PCR (RT-PCR) were used to verify the core genes and potential therapeutic agents with the highest correlation in vitro.Results:A total of 3 916 differentially expressed genes including 1 786 up-regulated genes and 2 130 down-regulated genes were screened.GO analysis showed that the differential genes were mainly involved in the positive regulation of phosphorylation, secretory vesicles, racemase and epimerase activities.KEGG analysis showed that differential genes were involved in systemic lupus erythematosus, alcoholism, sticky spots, amoebic dysentery Ras signal pathways and other disease pathways.The characteristic expression analysis of differential genes showed that MEK inhibitors, HSP90 inhibitors and signal transduction pathway kinase inhibitors were drugs similar to the differential genes.PPI results showed that H2AFJ, TFF1, GATA3, FOXA1,and CDH1 were core genes related to breast cancer.Two core genes of H2AFJ and TFF1 with the highest correlation were further selected for GEPIA analysis.The results of the analysis showed that the mRNA expression levels of H2AFJ and TFF1 in breast cancer cells were significantly higher than those in normal tissues, and there was a significant correlation with the pathological staging, overall survival rate and disease-free survival rate of breast cancer patients.H2AFJ and TFF1 may be potential prognostic biomarkers for survival of breast cancer patients.The functions of differentially expressed H2AFJ and TFF1 are mainly related to hormone receptor binding, epithelial structure maintenance and epigenetic negative regulation of genes, chromatin tissue involved in negative regulation of transcription, etc.The results of immune infiltration showed that the expressions of H2AFJ and TFF1 had a significant correlation with the infiltration of macrophages, neutrophils, monocytes, CD4+T, CD8+T, B lymphocytes and other immune cells.CMap results showed that compounds such as Gefitinib,Alpelisib, Sorafenib, and Sunitinib had potential therapeutic effects on breast cancer.Western blot and RT-PCR results showed that H2AFJ and TFF1 were significantly overexpressed in breast cancer cells.Gefitinib significantly inhibited the expression of H2AFJ and TFF1 in breast cancer cells(P＜0.05,P＜0.01).Conclusion:In this study, differentially expressed genes between breast cell carcinoma tissues and normal tissues were screened out by bioinformatics means to further identify key genes and compounds with potential therapeutic effects in the onset process of breast cancer and to further verify the effectiveness of the screened drugs on breast cancer through experiments.It will provide reference for clinical research and development of new drugs against breast cancer in the future in order to develop more effective treatment options.

1.Introduction

Breast cancer (BRCA) is one of the three most common malignant tumors in the world, which greatly affects the physical and mental health of people around the world.The mortality rate of breast cancer ranks first among all kinds of malignant tumors in women,and the incidence rate of early breast cancer ranges from 7% to 18%[1].At present, breast cancer treatment includes local treatment and systemic treatment including surgery, radiotherapy, chemotherapy,hormone and targeted treatment [2-3].About 70-80% of early breast cancer patients can be cured, but some treatment methods for breast cancer do not improve the survival rate due to the invasiveness of tumor cells, but bring psychological pressure to patients.During the radiotherapy process, while killing tumor cells, it will also interfere with the survival of normal cells, bringing secondary damage to patients.In addition, because of the specificity of breast cancer cells, there are great individual differences in the treatment results.At the molecular level, the activation of human epidermal growth factor receptor 2 (HER2), hormone receptor and BRCA mutation are the causes.Although some scholars have explored its molecular mechanism, there are still unknown aspects [4].Therefore, studying the pathogenesis of breast cancer and breast cancer related genes plays an active role in the early diagnosis and treatment of breast cancer, effectively alleviating the current status of clinical breast cancer treatment.In recent years, the bioinformatics analysis technology based on gene chip, which integrates massive biological information through data screening, statistical analysis, pathway analysis and visual analysis of protein interaction network, has been widely used [5-6].And the bioinformatics technology used in cancer research accounts for a large proportion, through which the pathogenesis of breast cancer can be further explored at the molecular level.

In this paper, gene expression database (GEO) is used to search breast cancer gene chip, screen out the genes with different expression between breast cancer model group and control group,and then carry out gene ontology function analysis and Kyoto gene and genome encyclopedia pathway analysis of the differential genes, further build the differential gene PPI protein interaction network, screen and identify key genes and their differential expression, prognosis analysis and immune infiltration.Finally, the differential genes were converted into probe lists, imported into ConnectivityMap Map (CMap) to screen potential compounds, and the core genes and potential therapeutic drugs were verified in vitro.It provides more options for clinical treatment of breast cancer, and provides new research ideas for exploring the molecular mechanism and potential therapeutic drugs of breast cancer combined with bioinformatics analysis and gene chip.

2.Materials and Methods

2.1 Data acquisition

In gene expression database(https://www.ncbi.nlm.nih.gov/geo/)Enter, search and download the GSE79586 gene expression profile of "BRCA".The retrieved GSE79586 gene expression profile is based on the Affymetrix GPL19057 platform (Affymetrix Illumina NextSeq 500).There are 18 breast cancer samples and 18 normal control samples, a total of 36 samples.

2.2 Data Processing and Screening

Group and rename the downloaded sample data into breast cancer group and control group, and import them into BioJustice(https://amp.pharm.mssm.edu/biojupies/)The differential genes were processed and screened.The processed data will be further processed to identify differential genes (DEGs).Principal component analysis(PCA) is a method of transforming original data into another set of variables without linear correlation through orthogonal transformation to achieve the purpose of data dimensionality reduction.It is usually used to extract the main feature components of a group of data [7].The screening conditions for differential genes were set as log 2 Fold Change 2 and P＜0.05 after correction.

2.3 Function and pathway analysis of differential genes

GO analysis, including biological processes (BP), molecular functions (MF) and cellular components (CC), is a widely used method in large-scale functional enrichment research.KEGG is a comprehensive database integrating a large number of data on genome, biological pathway and disease.Using the DAVID database(https://david.ncifcrf.gov/)GO annotation analysis and KEGG pathway enrichment analysis were conducted for the differential genes screened in this study.P＜0.05 and gene count＞5 were recorded as screening conditions in GO analysis; P＜0.05 was recorded as the screening condition of KEGG pathway analysis.

2.4 Visualization of Differential Gene Expression

L1000FWD database(https://maayanlab.cloud/l1 000fwd/)It provides interactive visualization for the expression characteristics of more than 16 000 drugs and small molecule induced genes, and supports coloring the characteristics through different attributes(such as cell type, time point, concentration) and drug attributes, so as to analyze the small molecule function and drug action mode, and infer the affinity of drugs for the above differential genes [8].The obtained drug data were ranked from high to low according to the similarity score.

2.5 Construction of PPI Interactive Network

String database(https://string-db.org/)It is an online analysis website that can visually analyze the interaction between different genes.The differential genes screened in this study were imported into the String database for protein interaction network analysis, set the confidence level (high highconfidence: 0.700), hidden the free nodes, and obtained the gene interaction network.To further screen out proteins that have a greater correlation with breast cancer for subsequent research and analysis.

2.6 Prognostic Analysis of Key Genes

GEPIA (http://gepia.cancer-pku.cn/index.html)It is a newly developed interactive web server for analysis and interactive analysis of tumor and normal gene expression, which is composed of thousands of tumor and normal tissue sample data[9].In this study, the differential expression of genes between tumor and normal tissues was analyzed by GEPIA, and the pathological stage was analyzed.Finally, Kaplan Meier plotter was used(http://kmplot.com/analysis/index.php?p=service&canc er=liver_ Rnaseq)online mapping tool carries out prognostic analysis on the mRNA expression of core genes in breast cancer patients screened above,and it is generally considered that there is a significant difference(P＜0.05)[10].

2.7 Gene Related Function Prediction

GeneMANIA (http://www.genemania.org)It is a website that can search many large and publicly available biological data sets to find relevant genes.These include protein protein, protein DNA and genetic interaction, pathway, reaction, gene and protein expression data, protein domain and phenotype screening overview [11].In this study, the screened core genes were introduced into this platform to analyze the function of the screened key genes and infer their value.

2.8 Relationship Between Gene and Cellular Immune Infiltration

TIMER 2.0 (http://timer.cistrome.org/)The use of the six most advanced algorithms can provide a more comprehensive and flexible estimation of immune invasion level for cancer genome maps(TCGA) or tumor maps provided by users.TIMER 2.0 web server provides the function of analyzing tumor infiltrating immune cells and further visualizing them.By inputting the selected key genes into the scatter map generated by the database, we can intuitively observe their expression and the relationship with breast cancer immune infiltration level [12].

2.9 CMap Analysis

CMap (ConnectivityMap Map) is an important database in the field of pharmacogenomics(https://portals.broadinstitute.org/cmap/) There are more than 1300 compounds in total, which can quickly use gene expression profile data to compare drugs highly related to diseases through the association of genes, diseases and drugs established by gene expression profiles, and summarize the possible action mechanism of drug molecules, thus revealing the relationship between disease causes and potential compounds [13-14].The above screened differential genes are converted into the corresponding probe list after R language processing, and uploaded to the CMap database to compare the differential gene list with the database reference data set, so as to screen potential compounds.

2.10 Cellular Experimental Verification

2.10.1 Main Material

Human breast cancer cell line MDA-MB-436 was provided by the School of Basic Medicine of Heilongjiang University of Traditional Chinese Medicine; Gefitinib (GYZZ J20180014, AstraZeneca Pharmaceutical Co., Ltd.); DMEM culture medium (batch number C11995500BT, Gibco Company, USA); Fetal bovine serum (batch No.C0230, Bovogen Company); Trizol (batch No.10296028);UltraPure Agarose (batch No.16500100); SuperScript III RT reverse transcription kit (batch number 11752050); Sybr qpcr mix(batch number 4472920) was purchased from Invitrogen Company in the United States; H2AFJ antibody (batch number: GTX53601,GeneTex, USA); TFF1 antibody (batch number 5893-100, Biovision Company, USA); GAPDH antibody (lot No.5174P), HRP goat anti rabbit antibody (lot No.7074), and HRP goat anti mouse antibody(lot No.91,196S) were purchased from Cell Signaling Technology;ECL test kit (batch number D412DA006, Shanghai Sangong Biotechnology Technology Service Co., Ltd.).

2.10.2 Instrument

Cell incubator (MMM Group, Germany); Legend Micro 21R Desktop High Speed Refrigeration Centrifuge (Thermo, USA);StepOne Software fluorescent quantitative PCR instrument (Applied Biosystems, USA); DYY-6C electrophoresis apparatus, DYCZ-24K transfer electrophoresis apparatus and WD-9413B gel imaging system were purchased from Beijing Liuyi Company.

2.10.3 Detection mRNA expression levels of H2AFJ and TFF1 by RT-PCR

After MDA-MB-436 cells were treated with different concentrations of sunitinib (0 μmol/L, 10 μmol/L, 15 μμmol/L, 20 μmol·L-1) for 24 h, total RNA was extracted by the Trizol method.The concentration and purity of total RNA were used as indicators for quality detection.RNA was reverse transcribed into cDNA for RT-PCR analysis to detect the expression level of H2AFJ, TFF1 mRNA.After reaction, PCR products were scanned and imaged by 2% agarose gel electrophoresis and automatic gel imaging analysis system.The ratio of the density of the electrophoresis bands of the target gene and the internal reference gene was taken as the relative expression amount of mRNA.The primer sequence is shown in Table 1.

Tab 1 Primer sequences

2.10.4 The expression levels of H2AFJ and TFF1 proteins were detected by Western blot

MDA-MB-436 cells were treated with different concentrations of sunitinib (0 μmol/L, 10 μmol/L, 15 μmol/L, 20 μmol/L) for 24 h,then protein lysate was added to lyse and collect protein samples.Protein samples were separated by 10% or 12% SDS-PAGE electrophoresis, transferred to 5% skimmed milk powder, shaken and sealed, H2AFJ (1:1 000 dilution) and TFF1 (1:1 000 dilution) first antibody were added, and incubated at 4 ℃ for 12 h, After TBST washing the membrane, add secondary antibody (1:2 000 dilution)and incubate it at room temperature for 1 h, then wash the membrane with TBST solution for three times, develop color, image and take photos with ECL kit, and finally analyze the protein bands with ImageJ software and calculate the gray value.

2.10.5 Statistical analysis

GraphPad 8.0.2 software was used for data statistics and analysis.The measurement data was expressed in(±s), and one-way ANOVA was used for comparison between groups.The difference was statistically significant (P＜0.05).

3.Results

3.1 Screening Results of Differential Genes

PCA visual analysis results show the interactive three-dimensional scatter plot of the first three main components of the data (PC1,PC2, PC3), and each point represents a selected RNA seq data sample.It can be seen from Figure 1 that the distance between samples with similar gene expression profiles in three-dimensional space is closer, and the expressions of control group and model group are independent, which can be used for the next analysis.The differentially expressed genes were screened under the condition of P＜0.05.The output of Log2FC value＞2 was the up-regulated gene, and the output of Log2FC＜- 2 was the down-regulated gene.A total of 3916 differentially expressed genes, including 1786 upregulated genes and 2130 down-regulated genes, were screened.The visualization results were presented in the form of volcano map and cluster map (Figure 2).

Fig 1 Schematic diagram of PCA analysis results

Fig 2 Differential gene volcano map and cluster heat map

3.2 GO Function Enrichment And KEGG Pathway Analysis

The DAVID online website was used to analyze the GO and KEGG of 3916 differential genes in this study, and the results were visualized.GO enrichment analysis shows that the main enrichment biological processes (BP) of differential genes are positive regulation of phosphorylation, positive regulation of MAP kinase activity,extracellular matrix organization, and positive regulation of protein phosphorylation, Regulation of cell proliferation; Differential genomic analysis (CC) mainly focused on secretory vesicle,bicellular tight junction, focal adhesion, cytoplasmic vesicle,endoplasmic reticulum lumen, etc; The molecular function (MF) of differential genes was analyzed as racemize and epimerase activity,myosin V binding, Rho GTPase binding, protein kinase activity, and protein homodimerization activity; The KEGG pathway analysis results show that the main enrichment pathways of differentially expressed genes are related to systemic lupus erythematosus,alcoholism, focal adhesions, amebic dysentery and other diseases and Ras signal path.The enrichment process and joint score of each pathway are shown in Table 2 and Figure 3.

Tab 2 KEGG pathway analysis

Fig 3 Results functional GO enrichment analysis of differential gene

3.3 Interactive Visualization Results of Differential Gene Feature Expression

After the differential genes are imported into the L1000FWD online database, the similarity results of small molecules and drug characteristics are obtained, as shown in Table 3.MEK inhibitors,HSP90 inhibitors, signal transduction pathway kinase inhibitors,etc.are drugs with similar characteristics to the differential genes.The data are visualized and displayed in the form of pyrotechnic diagrams, in which different shapes represent different time nodes and different colors represent different cell types.As shown in Figure 4.

Tab 3 Main drugs with similar characteristics with different genes

3.4 Visualization Results of Gene PPI Network

The obtained differential genes were subjected to protein-protein interaction network analysis using a String database to obtain a PPI network map.The graph consists of 140 nodes and 126 links(excluding irrelevant nodes), as shown in Fig.5.It could be seen that the genes closely related to the occurrence and development of breast cancer were mainly H2AFJ, TFF1, GATA3, FOXA1, CDH1,etc.

Fig 4 Visual pyrotechnic map of the expression of differentially expressed genes

Fig 5 Network diagram of differential gene protein interaction

3.5 Analysis on the Expression of Key Genes H2AFJ and TFF1

The mRNA expressions of H2AFJ and TFF1 in BRCA and normal breast cancer cells were compared using the GEPIA database, and the results are shown in Fig.6.The expression levels of H2AFJ and TFF1 in breast cancer cells were higher than those in normal tissues.According to the correlation between the expression differences of H2AFJ and TFF1 and the pathological stages of breast cancer patients, both H2AFJ and TFF1 groups had statistical significance(P＜0.05).As shown in Fig.7, the central loci of H2AFJ and TFF1 were not on the same central axis, and significant changes occurred in the fourth and fifth stages.It can be speculated that H2AFJ and TFF1 play important roles in this stage of progression of breast cancer.

Fig 6 Expression of H2AFJ and TFF1 in breast? cancer cells

3.6 Prognostic Value of Expression of H2AFJ And TFF1 in Breast Cancer Patients

The value of H2AFJ and TFF1 expression in the process of breast cancer was evaluated by GEPIA to further analyze the correlation between H2AFJ, TFF1 and clinical results.The curves of H2AFJ and TFF1 overall survival (OS) and disease free survival (DFS)are shown in Figures 8a and 8b respectively.The results showed that H2AFJ (P=0.008 5) and TFF1 (P=0.003 1) were significantly correlated with short-term disease-free survival (P＜0.05).

3.7 Prediction Results of Differential Gene Function Analysis

Input the genes H2AFJ and TFF1 into the GeneMIANA network database for visualization to obtain the interaction network diagram, as shown in Figure 9.Figure 9 shows the functions of differentially expressed genes of H2AFJ and TFF1, the functions of their similar genes, and the parts of the screened genes that are functionally similar to their similar genes, which are mainly related to the processes of hormone receptor binding, epithelial structure maintenance, RNA polymerase II specific DNA binding transcription factor binding, epigenetic negative regulation of genes,gene silencing, chromatin tissue involved in negative regulation of transcription, etc.The proportion of physical interaction, gene interaction and pathway is 76.59%, 3.17% and 1.88% respectively.

Fig 7 Correlation between the expression of H2AFJ and TFF1 and tumor stage in breast cancer patients

3.8 Immune Cell Infiltration of H2AFJ And TFF1 In Breast Cancer Patients

The TIMER2.0 database was used to explore the correlation between H2AFJ and TFF1 and immune cell infiltration, and H2AFJ and TFF1 were input into the database.The results are shown in Figures 10 and 11.The expression of TFF1 was positively correlated with the infiltration of CD+8T lymphocytes and monocytes in breast cancer, negatively correlated with the infiltration of CD+4T lymphocytes and neutrophils, but not significantly correlated with B cells and macrophages.The expression of H2AFJ was positively correlated with B cells in breast cancer, negatively correlated with CD+8, CD+4, macrophages, neutrophils, and not significantly correlated with monocytes.

Fig 8 Prognostic value of H2AFJ and TFF1mRNA expression in breast cancer cells (GEPIA)

Fig 9 Cluster analysis of h2afj, TFF1 and their similar genes

Fig 10 Correlation between H2AFJ and immune cell infiltration

Fig 11 Correlation between TFF1 and immune cell infiltration

3.9 CMap Analysis

The score range of CMap is-11.A positive score of the analysis result indicates that the drug can promote the development of the disease, and there is a positive correlation between the two; A negative score of the analysis results indicates that the drug has an inhibitory effect on the development of the disease, and there is a negative correlation between the two; The greater the absolute value of the score, the higher the correlation between the drug and the disease, and the greater the percentage, indicating the stronger the inhibitory effect of the drug [15].The list of differential gene probes processed by R language was imported into the CMap database,and five compounds with negative scores and the highest correlation strength were screened, as shown in Table 4.Compounds including Gefitinib, Alpelisib, Sorafenib, Sunitinib, everolimu, etc.have inhibitory effects on gene expression of breast cancer.

Tab 4 The top 5 potential therapeutic drugs for breast cancer in CMAP list

3.10 The Expression Analysis of H2AFJ and TFF1 mRNA in Breast Cancer

In order to further verify the data mining results, this experiment used RT-PCR to detect the expression changes of H2AFJ and TFF1 mRNA in human breast cancer cell MDA-MB-436 treated with different concentrations of gefitinib, as shown in Figure 12.Compared with the control group, the expression of H2AFJ and TFF1 mRNA in each dose group of gefitinib decreased significantly(P＜0.05, P＜0.01).

3.11 Analysis of H2AFJ and TFF1 Protein Expression in Breast Cancer

The changes of H2AFJ and TFF1 protein expression in human breast cancer cells MDA-MB-436 treated with different concentrations of gefitinib are shown in Figure 13.Compared with the control group, the expression of H2AFJ and TFF1 mRNA in each dose group of gefitinib is significantly reduced (P＜0.05, P＜0.01).

Fig 12 Effect of gefitinib on the expression of H2AFJ and TFF1 mRNA in MDA-MB-436 cells

Fig 13 Effects of gefitinib on the expression of H2AFJ and TFF1 proteins in MDA-MB-436 cells

4.Discussion

Breast cancer, including LuminaA, LuminaB, Her2 and Basal subtypes, is a complex and heterogeneous disease.It is also the most common invasive cancer among women in the world.Although the overall survival rate of breast cancer has improved significantly in recent studies, it is still one of the main causes of female death[16].The occurrence and development of cancer are closely related to gene mutation.In recent years, with the rise of gene therapy strategies, more and more proto oncogenes and tumor suppressor genes have been tested as new targets for cancer treatment, and gene changes also play an important role in the development of breast cancer [17].Therefore, it is of great significance to study the pathogenesis of breast cancer from the perspective of genes.Bioinformatics is to explore the pathogenesis of various diseases,including cancer, from the genetic level, providing a new method for further exploring the pathogenesis of breast cancer and identifying new and more effective targets, and also complementing the clinical invention of treatment methods for breast cancer.

In this study, we used GSE79586 gene chip of GEO database to conduct bioinformatics analysis of differentially expressed genes between breast cancer tissue and normal tissue, and screened a total of 3916 differentially expressed genes, including 1786 up-regulated genes and 2130 down regulated genes.Through GO function analysis and KEGG pathway analysis, the enrichment function and pathway involved by differential genes were identified.At the same time, the PPI protein interaction network was constructed,and visual analysis was carried out.It was clear that the core gene H2AFJ closely related to breast cancer cells was compared with TFF1, Yao [18] and other gene arrays for a group of breast tumors.The amplification of candidate sites was confirmed in breast cancer and cell lines through quantitative polymerase chain reaction and fluorescence in situ hybridization, The results showed that only H2AFJ was overexpressed in breast cancer with 12p13 amplification, and H2AFJ was a newly presumed oncogene in breast cancer.TFF1 is a member of the TFF family of trefoil factor family domain proteins, which is mainly constitutively expressed in gastric epithelium.TFF1 gene is the first mammalian member identified in the TFF family.Siu [19] and others analyzed it with microscope and flow cytometry, and confirmed that TFF1 is closely related to the cell membrane of breast cancer cell MCF-7.TFF1 gene may be a potential gene of breast cancer.Through GEPIA, GeneMANIA,TIMER2.0 and other databases, the differential expression of H2AFJ and TFF1, prognostic value and immune cell infiltration were further clarified.The results showed that H2AFJ and TFF1 mRNA were significantly expressed in breast cancer cells than in normal tissues, which may be potential prognostic markers for the survival of breast cancer patients.Then CMap was used to screen potential compounds and provide candidate therapeutic drugs for clinical prevention and treatment of breast cancer.The results showed that Alpelisib, by inhibiting the PI3K enzyme subunit coding protein and targeting the phosphatidylinositol-3-kinase of breast cancer to catalyze PIK3CA gene mutation, had a significant role in the treatment and prognosis of women with advanced or metastatic breast cancer, and could significantly extend the disease-free survival of breast cancer patients [20-21].Both sunitinib and sorafenib belong to small molecule tyrosinase inhibitors.They are multi target anti angiogenesis TKIs that act on vascular endothelial growth factor receptor (VEGFR), platelet-derived growth factor receptor (PDGFR),etc.Some studies have shown that they are significantly related to the prognosis of breast cancer, and can promote the metastasis of breast cancer models [22-23].The clinical efficacy of sorafenib alone could not reach the expected value.Currently, sorafenib combined with chemotherapy drugs are mostly used to treat advanced breast cancer.The disease control of sorafenib combined with paclitaxel in the first-line treatment of advanced breast cancer is better than that of paclitaxel alone combined with placebo.Among them,gefitinib is a potential therapeutic drug for breast cancer with the highest correlation score in CMap analysis.Therefore, gefitinib was subsequently selected for subsequent in vitro experiments to verify the expression of selected core genes in breast cancer.The results confirmed the effectiveness of gefitinib in the treatment of breast cancer.Therefore, potential compounds including gefitinib, sunitinib,and apelixi are expected to become potential therapeutic drugs for breast cancer.

To sum up, this experiment used bioinformatics combined with gene chip to explore the differentially expressed genes in breast cancer, and made a preliminary analysis of them, to obtain the signal pathways and functions that the differentially expressed genes mainly participate in, and to further explore its potential therapeutic drugs and verify the screened potential therapeutic drugs and core genes through in vitro experiments.The results further confirmed that gefitinib, a potential therapeutic drug, can effectively inhibit the expression of core genes H2AFJ and TFF1 in breast cancer MDAMB-436 cells, providing a theoretical basis for the clinical treatment and mechanism research of breast cancer in the future.

Author′s contribution

The design idea of this paper is provided by Professor Wang Rui and Yang Jing; Liang Xiao is responsible for target screening,pathway enrichment, network construction, CMap drug screening,data visualization and experiment; Li Yalan and Bai Haotian are responsible for literature review and sorting.Conflict of interest:There is no conflict of interest in this article.