Reduced Expression of SFRP1 is Associated with Poor Prognosis and Promotes Cell Proliferation in Breast Cancer: An Integrated Bioinformatics Approach

Breast carcinoma is the most frequent form of malignancy in women globally. Exploration of the breast cancer genome tiled the way for the validation of novel cancer biomarkers and to explore various mechanisms involved in the progression of carcinogenesis. The purpose of the research is to find an identification of potential gene linked to breast cancer (BC) progression and prognosis. Three datasets (GSE71053, GSE61724, and GSE36295) were downloaded from the Gene Expression Omnibus (GEO) database. An integrated analysis of several gene expression profile datasets was used to find differentially expressed genes (DEGs) in BC and normal breast tissue samples. Protein–protein interaction (PPI) network was used to verify hub genes associated with the pathogenesis and prognosis of BC. The functional enrichment and pathway analysis was performed by FunRich and cBioPortal. The expression pattern was assessed using COSMIC, GEPIA2, and BC-GenExMiner. The results revealed that among the hub genes, Secreted Frizzled-related protein 1 (SFRP1) was a negative regulator of the Wnt pathway in breast cancer. Loss of SFRP1 may result in abnormal cellular proliferation, migration, and invasion, which may trigger cancer cells, leading to progression of the disease, poor prognosis, and therapy resistance. Lastly, the Kaplan–Meier plotter online database demonstrated that expression levels of the SFRP1 gene were related to lower survival. The findings of this research would provide some directive significance for further investigating the diagnostic and prognostic biomarker to facilitate the molecular targeting therapy of breast cancer; SFRP1 expression may be effective as a novel prognostic biomarker in early breast cancer.


Introduction
Breast cancer (BC) is the most frequent cancer in women and the second leading cause of cancer death. Breast cancer is associated with several risk factors, including long-term fertility, hormonal contraceptive usage, lack of physical activity, and alcohol intake; nonetheless, its etiology and pathophysiology are not fully determined. Several genes and cellular processes are implicated in the genesis and progression of BC [1].
The molecular mechanisms underlying the development and progression of BC tumors are unclear. As a result, determining the development of disease and key signaling pathways is vital for developing more efficient diagnostic and treatment strategies. Bioinformatics analysis has been used to promote oncology research in recent years, providing a basis for better disease prevention, early detection, and therapy [2]. Using bioinformatics tools, we can now screen and identify essential genes comprehensively. Identification of potential genes and pathways linked to BC carcinogenesis and disease prognosis will not only to the discovery of new diagnostic biomarkers and treatment targets but also helps in elucidating the underlying molecular mechanisms [3].
Microarrays are especially useful for detecting differentially expressed genes (DEGs) since they can detect gene expression on a global level rapidly. Gene chips are a type of microarray that allows for high-throughput gene expression studies with excellent sensitivity, selectivity, and reliability [4]. Microarrays have produced a huge amount of data, and the majority of that data have already been uploaded and preserved in publicly available databases searches. This insight could help researchers understand better the molecular pathways causing BC [5].
In this study, we tried to identify novel indicators for prognosis in BC patients as well as the prospective therapeutic targets for this disease. An integrated analysis of DEGs involved in BC will reveal more information about the BC mechanism. These findings could serve as the basis for the development of future BC diagnostic and therapeutic tools.

Gene Expression Profile Data
GSE71053, GSE61724, and GSE36295 gene expression datasets were screened out based on gene expression omnibus (GEO) datasets, a public repository for data storage containing microarray data (http://www.ncbi.nlm. nih.gov/geo/). The DEGs in BC samples were compared with normal samples using the Limma tool in R language [6]. DEGs were calculated using the following criteria: jlog2FCj C 1 and adjust P value \ 0.05.

Functional Enrichment Analysis of DEGs
FunRich was used for functional enrichment and interaction network analysis of genes and proteins to elucidate the underlying biological processes and molecular activities of DEGs [7]. Meanwhile, P-value \ 0.05 was defined as the cut-off criterion.

Protein-Protein Interaction (PPI) Analysis
To construct a PPI network, DEG protein products were matched to the search engine for retrieving the interacting genes database (STRING, https://stringdb.org/cgi/input.pl), using a confidence score C 0.9 as the cut-off criterion. The PPI network was visualized using the Cytoscape software [8].

Pathway Analysis
Breast invasive carcinoma datasets (TCGA, PanCancer Atlas) encompassing 1084 samples were chosen from the Cancer Genomics Portal (cBioportal) (http:// www.cbio portal.org) to investigate gene alterations and activities of hub genes in breast cancer. We developed a group using cBioportal to display hub genes in the context of biological interactions derived from public pathway databases [9].

Analysis of Genetic Alterations of Hub Genes
Somatic mutation information from COSMIC (https://can cer.sanger.ac.uk/cosmic) was utilized to examine hub gene alterations in breast cancer [10].

Analysis of Expression Level and Correlation Analysis
The gene expression profiling interactive analysis (GEPIA, http://gepia.cancer-pku.cn/index.html) was used to analyze the hub gene's expression level and correlation. It examines tumor and normal differential expression and was used to show the expression of hub genes in BC and normal tissues. The link between these hub genes was then shown using a boxplot [11].

Survival Analysis
The hub gene's prognosis values were calculated using the Kaplan-Meier plotter (KM plotter, http://kmplot.com/ana lysis/) mRNA BC database. According to this software, the relapse-free survival (RFS) and overall survival (OS) information were based on GEO, TCGA, and EGA database. To assess the relationship between gene expression and survival, the hazard ratio (HR) with 95% confidence intervals and log-rank P value were calculated and plotted [13].

Identification of Differentially Expressed Genes (DEGs)
Three gene expression profiles were selected, and Table 1 shows the comprehensive information about 115 breast cancer and 21 normal tissue samples in the included datasets. A total of 24 DEGs comprising 19 down-regulated and 5 up-regulated genes were retrieved after the integrated analysis of three GEO datasets. Figure 1 shows the volcano plot of the DEGs.

Functional Enrichment Analysis of DEGs
FunRich software was used to perform enrichment analysis for up-and down-regulated DEGs after gene integration. Cell division, mitotic nuclear division, kinesins, aurora B signaling, FOXM1 transcription network, signaling by aurora kinases, and M phase signaling were found to be notably abundant in up-regulated DEGs and down-regulated DEGs (Fig. 2).

PPI Network Construction and Analysis of Interrelations Between Pathways
The STRING database was used to construct a PPI network. A total of 23 nodes and 136 edges were mapped in the PPI network with a local clustering coefficient of 0.739 and a PPI enrichment P-value \ 1.0e -16 . Figure 3 depicts the information for the PPI network constructed in string and visualized using cytoscape. Gene ontology (GO) analysis showed that the DEGs are involved in the biological process (BPs) and cellular components such as meiotic sister chromatid cohesion and centromeric, actomyosin contractile ring assembly, centrosome separation, mitotic spindle midzome assembly, and regulation of mitotic centrosome separation. Moreover, the GO molecular function (MFs) analysis showed that the DEGs are mainly involved in ATP binding, carbohydrate derivative binding, and anion binding. The DEGs were mainly enriched in pathways such as mitotic prometaphase, M phase, cell cycle, mitotic, and resolution of sister chromatid cohesion.

Pathway Analysis of DEGs Generated by cBioPortal
We analyzed the influence of DEGs on biological pathways in the breast cancer dataset containing 1918 samples. Among the DEGs, SFRP1 involved in the WNT signaling pathway is the down-regulated gene. The Wnt signaling pathway's abnormal activation is linked to the formation of solid tumors including breast cancer. Analysis of breast carcinoma revealed similar frequencies of SFRP1 loss in breast cancer (101%, respectively). Figure 4 depicts the loss of SFRP1 gene activity and its modulation of the Wnt pathway.

Mutational Analysis of SFRP1
The SFRP1 mutations were evaluated in 602 samples from patients with breast cancer. Out of 602 samples, the major types of mutation were found to be a missense substitution (123 samples), synonymous substitution (68 samples), nonsense substitution (10 samples), inframe deletion (10 samples) followed by frameshift deletion (2 samples) (Fig. 5a).

Validation of Down-Regulation of SFRP1 mRNA in Breast Cancer Tissues in TCGA Database
Using the GEPIA (Gene Expression Profiling Interactive Analysis) tool, we compared the mRNA expression of SFRP1 between breast cancer and breast tissues. The

Clinicopathological Relevance of SFRP1 Expression in Breast Cancer Patients
Next, the bc-GenExMiner database was used to determine the association between SFRP1 expression and clinicopathological variables in patients with BC. The results demonstrated that SFRP1 mRNA expression was negatively associated with estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor-2 (HER-2) status as shown in Fig. 6.

Survival Analysis and the Prognostic Value of SFRP1
To investigate the prognostic value of SFRP1, the survival analysis was conducted by the K-M plotter platform. Figure 7 shows the K-M survival curves for the SFRP1 gene (HR = 0.91, P = 0.069). It was found that the SFRP1 gene was the risky gene for prognosis with HR [ 1 and P \ 0.01. Higher expression of SFRP1 predicts shorter survival times for BC patients.

Discussion
Despite advances in treatment, breast cancer remains the most common malignant tumor in women worldwide, with the highest rate of increase in prevalence. The understanding of breast cancer's molecular pathways is critical for its diagnosis, treatment, and prognosis. The use of DNA microarray gene expression profiles to investigate DEGs involved in cancer has yielded useful diagnostic and medical applications [14].
In the present study, three gene expression profile datasets (GSE71053, GSE61724, and GSE36295) from the GEO database were retrieved and analyzed. The DEGs were identified using the 'limma' R package. The common DEGs were filtered out, and 24 hub genes were identified. GO function and pathway enrichment analysis was performed to further analyze the mechanisms of action of these DEGs. These DEGs were associated with the GO BP terms such as cell division, mitotic nuclear division, kinesins, aurora B signaling, FOXM1 transcription network, signaling by aurora kinases, M phase, meiotic sister chromatid cohesion and centromeric, actomyosin contractile ring assembly, centrosome separation, mitotic spindle midzome assembly, regulation of mitotic centrosome separation and response to ATP binding, carbohydrate derivative binding, and anion binding as molecular functions terms. Furthermore, the pathways of DEGs were mainly enriched in mitotic prometaphase, M phase, cell cycle, mitotic, and resolution of sister chromatid cohesion.
Of the 24 genes, SFRP1 (Secreted Frizzled Related Protein 1) gene that is closely associated with breast cancer Wnt signaling pathway was identified. One of the most essential mechanisms controlling cell physiologic activities such as division, multiplication, and adhesion is the Wnt/bcatenin signaling pathway [15]. Wnt ligand bind to Frizzled proteins and lipoprotein receptor-related proteins 5 and 6 receptors initiates signaling in normal circumstances. Then, as a transcription cofactor with T-cell factor/lymphoid enhancer factor, b-catenin aggregates and modulates the transcription of genes involved such as c-myc and cyclin D1 [16]. Abnormal activation of the Wnt/b-catenin signaling pathway is a common occurrence in malignancy, and also the abnormal methylation state of Wnt antagonists including such Dickkopf proteins, Wnt inhibitory factor1, and SFRPs may contribute to it [17]. SFRP1, a member of the SFRP family, can inhibit Wnt/-catenin signaling by interfering with Wnt-receptor associations via an N-terminal cysteine-rich domain similar to Frizzled proteins. SFRP1 is hypermethylated and down-regulated in breast cancer [18]. SFRP1 hypermethylation and down-regulation are also associated with poor prognosis in breast tumors [19]. Moreover, SFRP1 is associated with tumor chemotherapy, and some antitumor drugs inhibit cell growth through the re-expression of SFRP1. However, new research has revealed that SFRP1 may also be strongly expressed in carcinomas and enhance tumor development or migration, in breast cancer [20].
Our study demonstrates that SFRP1 is down-regulated in breast cancer patients which were analyzed in GEPIA and bc-GenExMiner database. Kaplan-Meier analysis showed that patient's low SFRP1 expression had significantly poorer survival rates. These findings imply that the level of SFRP1 expression can predict patient prognosis and could be used as a novel therapy target for personalized patient treatment. As a result, SFRP1 may be linked to breast cancer pathogenesis and could be used as a diagnostic biomarker for the disease.

Conclusion
Using an integrated analysis approach of different cohort profile datasets, the current study discovered possible candidate gene SFRP1 and the pathway involved in BC progression. These findings could contribute to a greater understanding of the biological mechanisms behind BC and the development of a possible biomarker. As a result, more research with higher patient cohorts is needed to validate the findings of this study. To characterize the precise roles of the identified gene, in vivo and in vitro examination of gene and pathway interaction is required, which might help to confirm gene functions and reveal the mechanisms behind BC.
Review and editing the manuscript. SS: Conceptualization, supervision, and editing the manuscript. All authors have read and approved the manuscript.
Funding Not applicable.

Declarations
Conflict of Interest The authors declare that they have no conflict of interest.

Consent for Publication Not applicable.
Ethical Approval Not applicable.