The Role of Apoptotic Genes and Protein-Protein Interactions in Triple-negative Breast Cancer

Abstract

Background/aim: Compared to other breast cancer types, triple-negative breast cancer (TNBC) has historically had few treatment alternatives. Therefore, exploring and pinpointing potentially implicated genes could be used for treating and managing TNBC. By doing this, we will provide essential data to comprehend how the genes are involved in the apoptotic pathways of the cancer cells to identify potential therapeutic targets. Analysis of a single genetic alteration may not reveal the pathogenicity driving TNBC due to the high genomic complexity and heterogeneity of TNBC. Therefore, searching through a large variety of gene interactions enabled the identification of molecular therapeutic genes.

Materials and methods: This study used integrated bioinformatics methods such as UALCAN, TNM plotter, PANTHER, GO-KEEG and PPIs to assess the gene expression, protein-protein interaction (PPI), and transcription factor interaction of apoptosis-regulated genes.

Results: Compared to normal breast tissue, gene expressions of BNIP3, TNFRSF10B, MCL1, and CASP4 were downregulated in UALCAN. At the same time, BIK, AKT1, BAD, FADD, DIABLO, and CASP9 was down-regulated in bc-GeneExMiner v4.5 mRNA expression (BCGM) databases. Based on GO term enrichment analysis, the cellular process (GO:0009987), which has about 21 apoptosis-regulated genes, is the top category in the biological processes (BP), followed by biological regulation (GO:0065007). We identified 29 differentially regulated pathways, including the p53 pathway, angiogenesis, apoptosis signaling pathway, and the Alzheimer's disease presenilin pathway. We examined the PPIs between the genes that regulate apoptosis; CASP3 and CASP9 interact with FADD, MCL1, TNF, TNFRSRF10A, and TNFRSF10; additionally, CASP3 significantly forms PPIs with CASP9, DFFA, and TP53, and CASP9 with DIABLO. In the top 10 transcription factors, the androgen receptor (AR) interacts with five apoptosis-regulated genes (p<0.0001; q<0.01), followed by retinoic acid receptor alpha (RARA) (p<0.0001; q<0.01) and ring finger protein (RNF2) (p<0.0001; q<0.01). Overall, the gene expression profile, PPIs, and the apoptosis-TF interaction findings suggest that the 27 apoptosis-regulated genes might be used as promising targets in treating and managing TNBC. Furthermore, from a total of 27 key genes, CASP2, CASP3, DAPK1, TNF, TRAF2, and TRAF3 were significantly correlated with poor overall survival in TNBC (p-value <0.05); they could play important roles in the progression of TNBC and provide attractive therapeutic targets that may offer new candidate molecules for targeted therapy.

Conclusion: Our findings demonstrate that CASP2, CASP3, DAPK1, TNF, TRAF2, and TRAF3 were substantially associated with the overall survival rate (OS) difference of TNBC patients out of a total of 27 specific genes used in this study, which may play crucial roles in the development of TNBC and offer promising therapeutic interventions.

Keywords: Breast cancer; apoptosis; bioinformatics; triple-negative breast cancer.

Figure 1. Expression of apoptosis-regulated genes in breast cancer (BC). (A) The overall effect of 27 apoptosis-regulated genes in invasive breast carcinoma. Kruskall-Wallis nonparametric test is used to compare measurements across patient groups for overall analyses. (B) Differential gene expression analysis in tumor, normal,] and metastatic tissues of the 27 apoptosis genes in breast invasive carcinoma patients according to TNMplot.

Figure 2. The apoptosis-regulated genes in tumor and normal breast tissue using UALCAN heatmap expression. (A) The difference in expression between the invasive breast tumor and normal breast tissue. (B) the differences in expression between tumor and normal breast tissue in TNBC.

Figure 3. Apoptosis gene expression profile based on the UALCAN databases. The difference between normal breast tissue (NBT) and TNBC is shown as *p<0.05, **p<0.01, ***p<0.001, and ns: nonsignificant.

Figure 4. Apoptosis gene expression profile based on the bc- GeneExMiner v4.5 mRNA expression (BCGM) databases. The difference between normal breast tissue (NBT) and TNBC is shown as ****p<0.0001, ns: nonsignificant. The values are expressed as mean±standard deviation.

Figure 5. Gene ontology (GO) analysis. (A) Biological process, (B) molecular function, (C) cellular components, and (D) protein class. The color of the bar graph represents the specific cellular activities, while the height represents the number of genes associated with each process.

Figure 6. PANTHER pathway analysis. The height of the bar indicates the number of genes on the corresponding pathways. Colors indicate the specific pathways. After inserting 27 genes, we found 29 various pathways which directly or indirectly contributed to apoptosis.

Figure 7. Enriched biological processes, molecular function, and cellular components. Following analysis of the gene array data, significantly enhanced processes are displayed as an interactive hierarchical clustering tree with ShinyGO. (A) Biological processes, (B) molecular function, and (C) cellular components that share a lot of genes tend to group. Larger dots indicate more significant p-values. Following examination of the genes, enriched processes that were similar were highlighted and marked.

Figure 8. The 27-apoptosis regulation protein-protein interactions (PPIs). Network analysis from the PPI network showed TNBC. (A) Minimum required interaction score: we set up the highest confidence at 0.9000 using Cytoscape. The nodes’ sizes indicated the count of interactions (degree) in the PPI network. There are 188 edges and 27 nodes in the circular structure of the protein-protein interaction, with an average node degree of 13.9 and an average local clustering coefficient of 0.76. The red line indicates the presence of fusion evidence; the green line is neighborhood evidence; the blue line is co-occurrence evidence; the purple line is experimental evidence; the yellow line is text mining evidence; the light blue is database evidence; the black line is co-expression evidence. (B) clustered PPIs. Twenty-seven genes were predicted from a cluster analysis using STRING. Associations within the same cluster are indicated by solid and dotted lines. Different colors denote various types of interactions.

Figure 9. Gene-gene interaction network among the apoptosis-regulated genes predicted by GeneMANIA. Different colors indicate various networks of gene-gene interactions. A total of 27 genes were input shown in big blue black color circle, and 20 related genes, shown in the small blue black color circle, were enriched in this network based on shared protein domains with 47 total genes, 0 attributes, and 1411 total links.

Figure 10. Overall survival (OS) analysis of 27hub genes in TNBC patients by KM Plotter (n=278). (A) CASP2, (B) DAPK1, (C) TNF, (D) TRAF2, (E) APONON/CASP3, (F) TRAF3, (G) AKT1, (H) CED4/APAF1, (I) FADD, (J) BAD, (K) TNFRSF10A, (L) DFFA, (M) TRAIL-R2/TNFRSF10B, (N) OPG-TNFRSF11B, (O) TNFRSF21, (P) BAG1, (Q) ACTB, (R) DIABLO, (S) BAX, (T) BCL10, (U) BNIP3, (V) BIK, (W) CASP4, (X) GADD45A, (Y) MCL1, (Z) TP53, (AA) CASP9.

Figure 11. Breast cell cancer selected protein expression. (A-F) Immunohistochemistry showing the expression of theTRAF2, TRAF3, DAPK1, TNF, CASP2 and CASP3 proteins in normal breast tissue and breast cancer as shown on the Human Protein Atlas website.