Curcumin suppresses metastasis of triple-negative breast cancer cells by modulating EMT signaling pathways: An integrated study of bioinformatics analysis

Abstract

This study aimed to use bioinformatics approaches for predicting the anticancer mechanisms of curcumin on triple-negative breast cancer (TNBC) and to verify these predictions through in vitro experiments. Initially, the Cell Counting Kit-8 (CCK8) assay was employed to rigorously investigate the influence of curcumin on the proliferative capacity of TNBC cells. Subsequently, flow cytometry was employed to meticulously assess the impact of curcumin on cellular apoptosis and the cell cycle regulation. Transwell assays were employed to meticulously evaluate the effect of curcumin on the motility of TNBC cells. RNA sequencing was conducted, followed by Gene Ontology and Kyoto Encyclopedia of Genes and Genomes enrichment analyses of differentially expressed genes, aiming to elucidate the potential anticancer mechanisms underlying curcumin’s effects. To thoroughly elucidate the interactions among multiple proteins, we constructed a protein–protein interaction (PPI) network. Finally, the expression levels of several key proteins, including fibronectin, mTOR, β-Catenin, p-Akt, Akt, N-Cadherin, p-S6, and S6, were assessed using the western blot. The CCK8 assay results showed that curcumin significantly inhibited the proliferation of Hs578T and MDA-MB-231 cells. Flow cytometry results showed that curcumin induced apoptosis in these cells and arrested the cell cycle at the G2/M phase. Additionally, Transwell assay results showed that curcumin effectively reduced the motility of Hs578T and MDA-MB-231 cells. Enrichment analysis of RNA sequencing data showed that the mechanism of action of curcumin was significantly associated with signaling pathways such as pathways in cancer, focal adhesion, and PI3K-Akt signaling pathways. Subsequently, we constructed a protein–protein interaction network to elucidate the interactions among multiple proteins. Finally, Western blotting analysis showed that curcumin significantly decreased the expression levels of key proteins including Fibronectin, mTOR, β-Catenin, p-Akt, Akt, N-Cadherin, p-S6, and S6. Curcumin exhibits its therapeutic potential in TNBC by modulating multiple signaling pathways. It may inhibit the epithelial-mesenchymal transition process by downregulating the expression of proteins involved in the mTOR and PI3K-Akt signaling pathways, thereby suppressing the motility of TNBC cells. These findings provide experimental evidence for considering curcumin as a potential therapeutic strategy in the treatment of TNBC.

1. Introduction

Breast cancer is the most prevalent malignancy among women worldwide. In 2020, about 2.3 million new cases of breast cancer were diagnosed.^[1,2] Triple-negative breast cancer (TNBC), a subtype of breast cancer, is particularly known for its aggressiveness, defined by the negative of estrogen receptor, progesterone receptor, and human epidermal growth factor receptor 2.^[3–5] TNBC is characterized by high heterogeneity, a high recurrence rate, a high rate of metastasis, and a poor prognosis, with fewer than 30% of patients surviving beyond 5 years.^[6,7] The distinct biological properties of TNBC, along with the absence of efficient and specific therapeutic targets and drugs, highlight the urgency for comprehensive research in this domain.^[4,5,8]

TNBC poses substantial clinical challenges, primarily due to the absence of specific therapeutic targets, which accounts for the currently limited therapeutic efficacy.^[9] This situation underscores the urgent need for the development of innovative therapeutic strategies.^[10] This study concentrates on curcumin, a yellow phenolic pigment derived from turmeric, renowned for its anti-inflammatory, antioxidant,^[11] and immunomodulatory properties.^[12] We specifically explore its potential therapeutic impacts on TNBC. Recent research has shown that curcumin is crucial in treating several tumors by modulating typical cell biological effects such as cell proliferation, apoptosis, cell cycle, and metastasis.^[13,14] Thus, curcumin may play a significant role in the initiation and progression of various cancers, including breast, lung, and liver cancers, through affecting multiple signaling and molecular pathways, such as Rb, P53, mitogen-activated protein kinase, phosphatidylinositol 3-K (PI3K)/protein kinase B, and NF-kappaB (nuclear factor kappa B cells, NF-κB).^[15–18]

Although previous studies have demonstrated curcumin’s capacity to inhibit cell proliferation and invasion in human TNBC MDA-MB-231 cells,^[19] the underlying molecular mechanisms are still being explored. This study seeks to bridge this gap in knowledge by employing both bioinformatics and experimental approaches to investigate the effects of curcumin on cell motility in TNBC cell lines, specifically Hs578T and MDA-MB-231 cells.

This study is significant for its exploration of curcumin’s potential clinical applications in treating TNBC. By elucidating the antitumor mechanisms of curcumin in TNBC cells, it establishes a crucial foundation for developing innovative therapeutic strategies. Additionally, the integration of bioinformatics with in vitro experimental methods in this research not only enhances the understanding of breast cancer biology but also paves the way for more effective and personalized treatment approaches.

2. Materials and methods

2.1. Cell culture and curcumin stimulation

Human triple-negative breast cancer Hs578T cells and MDA-MB-231 cells (maintained by the Institute of Medical Oncology, Southwest Medical University) were cultured in DMEM high glucose medium containing 10% (v/v) FBS at 37 °C and 5% CO₂. Curcumin (purchased from MedChemExpress) was dissolved using dimethyl sulfoxide (DMSO) to a final concentration of 50 mmol/L and stored in portions at −20 °C. Curcumin-stimulated Hs578T cells and MDA-MB-231 cells were used as the curcumin group, while cells treated with an equivalent amount of DMSO formed the control group. The final concentration of DMSO in the above treatment solution was <0.1% (v/v).

2.2. Cell viability assay

Cell viability was determined by the Cell Counting Kit-8 (CCK-8; APExBIO, USA). In this experiment, cells were seeded in 96-well plates at a density of 4 × 10³ cells per well. The concentration-dependent effects of curcumin on cell proliferation were evaluated by treating cells with various concentrations (0, 5, 10, 20, 40, and 50 μmol/L) for 24 hours to assess its inhibitory effect. The time-dependent assay was conducted to evaluate the inhibitory effects on cell proliferation at different time points (24 hours, 48 hours, 72 hours), following treatment with the selected drug concentration. Prior to the assay, the CCK8 working solution was prepared. The medium in the 96-well plate was removed, and 100 μL of the CCK8 working solution was added to each well. The plate was then incubated at 37 °C in the dark for 1 to 4 hours. Subsequently, the Optical Density (OD) at 450 nm was measured using a spectrophotometer, enabling the calculation of the drug’s inhibition rate on cell proliferation. Five replicate wells were set up for each concentration, and 3 validated biological experiments were performed. The cell proliferation inhibition rate was calculated according to the formula, and cell viability (%) = [(A curcumin − A blank)/ (A control − A blank)] × 100%.

2.3. Flow cytometry analysis

The Annexin V-PE/7-AAD Apoptosis Detection Kit (KeyGEN BioTECH, China) was used for dual staining (Annexin V-PE/7-AAD) to determine the apoptosis ratio in cells. Hs578T cells were cultured in 6-well plates and treated with 20 μmol/L of curcumin for 24 hours. Similarly, MDA-MB-231 cells were cultured under the same conditions and treated with 25 μmol/L of curcumin for 24 hours. After collecting the cell precipitate as described in the instruction manual, resuspend the cells in 100 μL of 1× binding buffer. Add 1 μL of Annexin V-PE and 5 μL of 7-AAD to the suspension, and allow the reaction to proceed for 10 minutes at room temperature in the dark. Then add 400 μL of 1× binding buffer, and analyze the cells using flow cytometry (BD FACSVerse™) within 1 hour. Cell cycle distribution was analyzed using a Cell Cycle Assay Kit (KeyGEN BioTECH, China). Cells were first harvested using trypsin, then fixed in 70% cold ethanol and stored at 4 °C overnight. After washing away the fixative with PBS, the cells were stained with a preprepared PI/RNase A staining solution. Allow the reaction to proceed at room temperature, protected from light, for 30 to 60 minutes before analysis using the flow cytometry (BD FACSVerse™). The results of biological experiments of 3 independent experiments show Mean ± Standard Deviation.

2.4. Transwell assay

For the migration assay, cells in the logarithmic growth phase were harvested, digested, and prepared into a single cell suspension. The number of cells in the suspension was adjusted, setting the concentration of Hs578T cells at 2 × 10⁵ cells/mL and MDA-MB-231 cells at 3 × 10⁵ cells/mL. For the control group, prepare the cell suspension in serum-free medium for use in the upper chamber, and add 500 µL of culture medium supplemented with 20% FBS to the lower chamber. The curcumin group should follow the same protocol, with the only distinction being the incorporation of the specified drug concentrations: treating Hs578T cells with 20 μmol/L of curcumin and MDA-MB-231 cells with 25 μmol/L of curcumin. The cell suspensions were inoculated with 100 µL in the upper chamber of a 24-well Transwell plate (Corning) with a pore size of 8.0 μm. After 24 hours of curcumin treatment, the medium in the upper chamber was discarded, and the untransferred cells in the upper chamber of the small chamber were gently wiped away with a moistened cotton swab. Methanol was fixed for 10 minutes, 0.1% crystalline violet was stained for 10 minutes, and PBS was washed twice. The cells were air-dried, photographed, and counted under 5 random fields of view using an inverted fluorescence microscope (Leica, Germany). For the invasion assay, the Matrigel matrix gel (BD Biosciences, USA) was dissolved at 4 °C. It was subsequently diluted to achieve a final concentration of 1.0 mg/mL in prechilled serum-free medium. The other steps are identical to those of the migration assay.

2.5. RNA extraction, library construction, and sequencing

Hs578T cells were harvested and given a curcumin or DMSO treatment before being rinsed twice with cold PBS. A total of 6 samples were collected, including 3 samples from the curcumin group (M1, M2, M3) and 3 samples from the control group (C1, C2, C3). Total RNA was obtained using Trizol reagent (Thermofisher, 15596018) following the steps in the instructions. Quantitative and purity analysis of total RNA was performed with Bioanalyzer 2100 and RNA 6000 Nano LabChip Kit (Agilent, CA, 5067-1511). RNA samples with RIN numbers ＞7.0 were used to construct sequencing libraries. Two rounds of mRNA purification were performed using Dynabeads Oligo (dT) (Thermo Fisher, CA). Fragment the purified mRNA into short fragments (Magnesium RNA Fragmentation Module [NEB, cat.e6150, NJ] under 94 °C 5–7 minutes). Next, the RNA fragments were reverse transcribed with SuperScript™ II Reverse Transcriptase (Invitrogen, cat. 1896649, CA). The U-labeled second-strand DNA was then synthesized using E. coli DNA polymerase I (NEB, cat.m0209, NJ), RNase H (NEB, cat.m0297, NJ), and dUTP Solution (Thermo Fisher, cat.R0133, CA). The blunt ends of each strand were then given an A-base to help ligate them to the indexed adapters. Each adaptor has a T-base overhang for attaching to the DNA fragments with their A-tailed ends. The fragments were ligated to dual-index adapters, and AMPureXP beads were used for size selection. The U-labeled second-stranded DNAs were treated with the heat-labile UDG enzyme (NEB, cat. m0280, NJ) before the ligated products were amplified with PCR under the following conditions: initial denaturation at 95 °C for 3 minutes; 8 cycles of denaturation at 98 °C for 15 seconds, annealing at 60°C for 15 seconds, and extension at 72 °C for 30 seconds; and finally, final extension at 72 °C. The resultant cDNA libraries had an average insert size of 300 ± 50 bp. Finally, we used an Illumina Novaseq^TM 6000 (LC-Bio Technology CO., Ltd., Hangzhou, China), to carry out the 2 × 150 bp paired-end sequencing (PE150) following the vendor’s suggested procedure.

2.6. Data preprocessing

Reads obtained from the sequencing machines include raw reads containing adapters or low-quality bases, which will affect the following assembly and analysis. Thus, to get high-quality clean reads, reads were further filtered by Cutadapt (https://cutadapt.readthedocs.io/en/stable/, version:cutadapt-1.9). The parameters were as follows: (1) removing reads containing adapters; (2) removing reads containing polyA and polyG; (3) removing reads containing more than 5% of unknown nucleotides (N); (4) removing low-quality reads containing more than 20% of low quality (Q-value ≤ 20) base. The sequence quality was verified using FastQC (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/, 0.11.9), including the Q20, Q30, and GC content of the clean data. After that, a total of G bp cleaned, paired-end reads were produced. The raw sequence data have been submitted to the NCBI Gene Expression Omnibus (GEO) datasets with accession number <GEO accession> or NCBI Short Read Archive (SRA) with accession number <SRA accession>.

2.7. Differentially expressed genes (DEGs) analysis

Gene differential expression analysis was performed by DESeq2 software between 2 groups (and by edgeR between 2 samples). The genes with the q-value below 0.05 and absolute fold change ≥ 1 were considered DEGs. Differentially expressed genes were subjected to enrichment analysis of Gene Ontology (GO) functions and the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway.

2.8. GO and KEGG enrichment analyses

The DEGs were subjected to GO and KEGG pathway analysis to categorize enriched genes functionally and phenotypically (DOE-L and DUE-L). As a result, 2164 differentially expressed bases, including 464 DOE-L and 1700 Due-L, were enriched for analysis. Ontologies in GO include molecular function, cellular component, and biological process. The fundamental unit of GO is the GO-term. Each GO term is associated with a specific ontology. All GO terms that are significantly enriched in DEGs when compared to the genome background are identified by GO enrichment analysis. Pathway analysis contributes to a better understanding of genes’ biological functions. The primary public pathway-related database is KEGG. Because KEGG is the primary public pathway-related database, KEGG analysis of biological pathways was also performed. GO’s top 20. The top 20 GO terms and pathways were sorted according to their P values.

2.9. Protein–protein interaction (PPI) network construction

We examined the protein-protein interaction network in DEGs using the Search Tool for Retrieval of Interacting Genes/Proteins (STRING). Total completed the PPI network construction twice. STRING was subjected to the following conditions: (1) network type: physical subnetwork, (2) minimum required interaction score: highest confidence (0.900), and (3) hide network disconnected nodes. After screening the sequencing data with absolute log2(Fold Change) ≥ 1.05, the DEGs that met the criteria were entered into STRING to build the first PPI network. The Cytohubba plugin was used to calculate the extent of each node. Then, using the KEGG pathway section of the STRING analysis results, a second PPI network was built by selecting proteins with low false discovery rates that were relevant to the STRING analysis.

2.10. Western blot analysis

Total protein was extracted from Hs578T and MDA-MB-231 cells using RIPA lysate (P0013B, Beyotime, China), and the concentration was determined using a BCA protein assay kit (P0009, Beyotime, China). 50 micrograms of the protein samples were separated on 8% SDS-PAGE gels and then transferred onto PVDF membranes. The membranes were subsequently blocked with 5% bovine serum albumin (4240, BioFroxx, Germany) for 1 hour. The membrane was incubated with the specific primary antibody (1:1000; CST, USA) at 4 °C overnight. Following this, it was washed 3 times with TBST and then incubated with a fluorescent secondary antibody on a shaker at room temperature for 2 hours. After an additional 3 TBST washes, the PVDF membrane was visualized using the ODYSSEY CLx system (LI-COR Biosciences, NE).

2.11. Statistical analysis

Data from all samples were showed as Mean ± Standard Deviation after 3 measurements. Statistical analysis was conducted using GraphPad Prism 9.5.1 software to compare differences among various groups. The t-test was used for comparisons between 2 groups, while one-way ANOVA was utilized for comparisons among multiple groups. A P-value ＜ .05 is considered to indicate a statistically significant difference.

2.12. Ethics approval statement

Because the experiment does not involve animals and clinical research, this study does not need to be approved by moral and ethical clerks.

3. Results

3.1. Curcumin inhibited human TNBC Hs578T cell and MDA-MB-231 cell viability in a time- and concentration-dependent manner

After treating human triple-negative breast cancer Hs578T and MDA-MB-231 cells with different concentrations of curcumin (0, 5, 10, 20, 40, 50 μmol/L) for 24 hours, the CCK8 assay results showed that compared with the control group, the different concentrations of curcumin (5, 10, 20, 40, and 50 μmol/L) could inhibit cell viability in a concentration-dependent manner with statistically significant (Fig. 1A, C). Further, we treated Hs578T cells with 20 μmol/L of curcumin and MDA-MB-231 cells with 25 μmol/L of curcumin. We observed the inhibitory effects of curcumin on the proliferation of these 2 cells at different times (24 hours, 48 hours, and 72 hours). The data demonstrated that curcumin could effectively inhibits the viability of both Hs578T and MDA-MB-231 cells in a time-dependent manner (Fig. 1B, D). In subsequent cellular experiments, Hs578T cells were treated with 20 μmol/L of curcumin, while MDA-MB-231 cells were treated with 25 μmol/L curcumin.

Figure 1.

Concentration-dependent, time-dependent inhibition of human Hs578T cell and MDA-MB-231 cell proliferation by Curcumin. (A) Curcumin concentration-dependent (0, 5, 10, 20, 40, 50 μmol/L) inhibited the proliferation of Hs578T cells. (B) Curcumin time-dependently (24 hours, 48 hours, and 72 hours) inhibits the proliferation of Hs578T cells. (C) Curcumin concentration-dependent (0, 5, 10, 20, 40, 50 μmol/L) inhibited the proliferation of MDA-MB-231 cells. (D) Curcumin time-dependently (24 hours, 48 hours, and 72 hours) inhibits the proliferation of MDA-MB-231 cells. *P < .05, ****P < .0001.

3.2. Curcumin induced apoptotic ratio and arrested the cell cycle at the G2/M phase

To analyze the potential mechanism of curcumin’s antiproliferative effects, flow cytometry was employed to detect the effects of 20 μmol/L curcumin on the apoptosis of Hs578T cells and 25 μmol/L curcumin on the apoptosis of MDA-MB-231 cells. Compared with the control group, the curcumin group showed a significant increase in the early apoptotic ratio (5.36 ± 2.26% vs 12.24 ± 1.77%) and the total apoptotic ratio (16.30 ± 1.05% vs 38.46 ± 4.57%) of Hs578T cells (Fig. 2A). Similarly, the MDA-MB-231 cells also showed a similar trend (Fig. 2C). We further examined the effect of curcumin on the cell cycle distribution of TNBC Hs578T and MDA-MB-231 cells using flow cytometry. Results showed a significant increase in the G2/M phase in the curcumin group (23.85 ± 4.13) % compared to the control group (12.32 ± 3.00) % (Fig. 2B). This suggests that curcumin can effectively inhibits the proliferation of Hs578T cells at the G2/M phase. Similarly, the MDA-MB-231 cells also showed a similar result (Fig. 2D). Additionally, we tested the expression of Cleaved caspase-3 and p21 in the different groups. The results were consistent with the conclusions drawn above.

Figure 2.

Flow cytometry analysis. Hs578T cells were treated with DMSO for 24 hours as the control group and 20 μmol/L curcumin for 24 hours as the curcumin group. MDA-MB-231 cells were treated with DMSO for 24 hours as the control group and 25 μmol/L curcumin for 24 hours as the curcumin group. (A) Curcumin induced apoptosis in Hs578T cells. Phase Q2 indicates late apoptosis, and phase Q4 indicates early apoptosis. The total number of cells in the Q2 and Q4 phases was used to calculate the apoptosis rate. (B) Curcumin-induced cell cycle arrest in Hs578T cells. (C) Curcumin induced apoptosis in MDA-MB-231 cells. (D) Curcumin-induced cell cycle arrest in MDA-MB-231 cells. Comparison with the control group: *P < .05, **P < .01. Data represent the Mean ± SD from 3 independent experiments. DMSO = dimethyl sulfoxide.

3.3. Curcumin inhibits the motility and invasion capability of Hs578T cells and MDA-MB-231 cells

To further explore curcumin’s effect on cell motility and invasion, we employed the Transwell assay in human triple-negative breast cancer Hs578T and MDA-MB-231 cells. The motility results showed that curcumin significantly reduced the number of transmembrane Hs578T cells (128.10 ± 5.07) compared to the control group (212.30 ± 2.91) (Fig. 3A). The invasion assay showed similar results, that curcumin significantly reduced the number of transmembrane Hs578T cells (116.50 ± 2.44) compared to the control group (217.60 ± 4.19) (Fig. 3B). Similar results were observed in MDA-MB-231 cells. Figure 3C shows the results of the motility assay for MDA-MB-231 cells, while Figure 3D shows the results of the invasion assay for MDA-MB-231 cells. These results suggest that curcumin possesses anti-invasive and anti-migratory properties in TNBC Hs578T and MDA-MB-231 cells.

Figure 3.

Effect of treatment with 20 μmol/L curcumin for 24 hours on the migration of Hs578T cells and treatment with 25 μmol/L curcumin for 24 hours on the migration of MDA-MB-231 cells (×100). (A) A Transwell assay detected the inhibitory effect of Curcumin on Hs578T cell migration. (B) A Transwell assay detected the inhibitory effect of Curcumin on Hs578T cells invasion. (C) A Transwell assay detected the inhibitory effect of Curcumin on MDA-MB-231 cell migration. (D) A Transwell assay detected the inhibitory effect of Curcumin on MDA-MB-231 cells invasion. Experiments compared with controls: ****P < .0001, data represent the Mean ± SD from 3 independent experiments, and the average of the cell numbers of 5 different fields of view were taken for each independent experiment for the curcumin and control groups, respectively, for statistical purposes.

3.4. Quality control of data and samples

The raw data generated by sequencing were preprocessed and data filtered to obtain > 97.50% of valid data (Table 1). The proportion of bases with high quality (quality score ≥ 20) was >99.95%, and the proportion of bases with high quality (quality score ≥ 30) was >98.00%, indicating that the quality of RNA sequencing was high enough for subsequent bioinformatics analysis. Table 2 shows the mean Pearson correlation coefficients, accompanied by their respective standard deviations, for each of the 6 groups. Higher mean Pearson correlation coefficients mean more robust linear relationships among the groups, while lower standard deviations mean a greater degree of consistency in these relationships. It is important to note that these statistics were computed excluding the perfect self-correlation of each group (a correlation coefficient of 1), thereby ensuring a more accurate depiction of inter-group dynamics. In addition, the expression of 28,484 genes was annotated. PCA and calculation of Pearson correlation coefficients between samples revealed very similar and relatively good biological data reproducibility for samples within groups. At the same time, there were more significant differences between groups (Fig. 4A and B).

Table 1.

Data after filtration.

Sample	Raw data		Valid data		Valid ratio (reads)	Q20%	Q30%	GC content%
	Read	Base	Read	Base
C1	41013104	6.15G	40014430	6.00G	97.56	99.99	98.04	51
C2	41992610	6.30G	41019696	6.15G	97.68	99.98	98.06	51
C3	40591422	6.09G	39609526	5.94G	97.58	99.98	98.06	50
M1	41370368	6.21G	40519030	6.08G	97.94	99.99	98.22	53
M2	40081428	6.01G	39204918	5.88G	97.81	99.98	98.01	52
M3	40446692	6.07G	39552000	5.93G	97.79	99.98	98.05	52

Table 2.

Mean Pearson correlation coefficient for each group.

Group	Mean Pearson correlation coefficient	Standard deviation
C1	0.92	0.07
C2	0.92	0.07
C3	0.91	0.07
M1	0.91	0.07
M2	0.91	0.08
M3	0.93	0.06

Figure 4.

Analysis of biological samples and DEGs. (A) Correlation analysis of sample. (B) PCA (principal component analysis). (C) The volcano plot of DEGs. (D) The heatmap of DEGs. The functional enrichment of DEGs was investigated according to 3 categories of GO (biological processes [E], molecular functions [F], and cellular components [G]) and the KEGG pathway analyses of DEGs (H). DEGs = differentially expressed genes, GO = gene ontology, KEGG = Kyoto Encyclopedia of Genes and Genomes.

3.5. Identification of DEGs

A total of 2164 DEGs were screened (|log2fc|≥1, P < .05 and q < 0.05), including 464 up-regulated and 1700 down-regulated genes. The volcano plot and heat map of DEGs are shown in Figure 4C and D.

3.6. GO database and KEGG database enrichment analysis of DEGs

2164 DEGs (464 DOE-L and 1700 DUE-L) obtained were subjected to GO and KEGG enrichment analysis to functionally and characteristically classify the DEGs. GO has 3 ontologies: biological process, molecular function, and cellular component. For biological process, genes are significantly enriched in the GO terms such as cell cycle, cell division, DNA replication, mitotic cytokinesis, protein phosphorylation, positive regulation of cell migration, regulation of G2/M transition of the mitotic cell cycle, and regulation of cell motility (Fig. 4E). For molecular function, genes are significantly enriched in the GO terms such as protein binding, nucleotide binding, DNA binding, DNA helicase activity, helicase activity, hydrolase activity, chromatin binding, double-stranded DNA binding, and transferase activity (Fig. 4F). For cellular component, genes are significantly enriched in the GO terms such as nucleoplasm, nucleus, cytoplasm, chromosome, nuclear replication fork, chromatin, nuclear body, and chromosome (Fig. 4G). Therefore, these results suggest that genes regulating cell cycle, cell proliferation, cell migration, protein metabolism, signal transduction, and protein-protein binding are significantly enriched. Also, KEGG pathway analysis demonstrates that genes involved in proteoglycans in cancer, regulation of actin cytoskeleton, Pathways in cancer, PI3K-Akt signaling pathway, p53 signaling pathway, cell cycle, transcriptional misregulation in cancer, TNF signaling pathway (Fig. 4H). The top 20 GO terms and pathways were sorted according to their P values.

3.7. PPI network analysis of the DEGs

PPI networks of DEGs were used to construct the STRING online database and visualized by Cytoscape software. First, the sequencing data were screened by setting the filtering condition as |log2(FC)| ≥ 1.05. A total of 1977 DEGs were screened and entered into the STRING database for PPI network construction. There were 595 edges and 1756 nodes in the first constructed PPI network, and the degree of each node was calculated using the Cytohubba plug-in to calculate the genes at the top8 of the PPI network (Fig. 5A). Then, referring to the KEGG pathways section in the analysis results, the top 4 pathways were ranked according to the false discovery rate from the smallest: Proteoglycans in cancer, Pathways in cancer, and Regulation of Then, we entered 81 proteins in Pathways in cancer from the sequencing data into STRING, performed the second PPI network construction, and used the Cytohubba plug-in to calculate the degree of each node to calculate the top8 genes in the PPI network (Fig. 5B). According to the purpose of our study, the KEGG pathways part of the analysis results was explored for related proteins, such as PI3K-Akt signaling pathway, Breast cancer, Regulation of actin cytoskeleton, Focal adhesion, and mTOR signaling pathway.

Figure 5.

Protein–protein interaction networks of differentially expressed genes. PPI enrichment P-value: <1.0e−16. (A) The first PPI network and the top8 hub genes. (B) The second PPI network and the top8 hub genes. PPI = protein–protein interaction.

3.8. Effect of curcumin treatment on the proliferation, invasion, and migration-related proteins of Hs578T cells and MDA-MB-231 cells

To further explore the molecular mechanisms of curcumin in TNBC Hs578T and MDA-MB-231 cells, we used the Western blot technique to analyze key proteins associated with cell proliferation and migration. The results showed that, compared to the control group, curcumin treatment significantly reduced the protein expression levels of Fibronectin, mTOR, β-Catenin, p-Akt, Akt, N-Cadherin, p-S6, and S6 in both Hs578T and MDA-MB-231 cells. Figure 6A and B show the results for Hs578T cells, while Figure 6C and D show the results for MDA-MB-231 cells.

Figure 6.

Effect of Curcumin on the expression of proliferation invasion-related proteins in Hs578T cells and MDA-MB-231 cells. (A) Effect of curcumin on the proliferation-related proteins of Hs578T cells. (B) Effect of curcumin on the migration-invasion-related proteins of Hs578T cells. (C) Effect of curcumin on the proliferation-related proteins of MDA-MB-231 cells. (D) Effect of curcumin on the migration-invasion-related proteins of MDA-MB-231 cells.

4. Discussion

Curcumin, an efficacious anticancer compound isolated from turmeric rhizomes, mediates its therapeutic effects through the modulation of various cellular pathways. These include the suppression of tumor metastasis, angiogenesis, and inflammation, alongside the modulation of apoptosis, cell cycle progression, and resistance to multiple drugs. The role of curcumin as a cancer chemopreventive agent has been rigorously studied in various cancer models.^[20–24] Recent studies have highlighted curcumin’s antiproliferative and antimotility effects on breast cancer cells, yet the intricate molecular mechanisms underlying these actions require further elucidation.^[25–27] Therefore, exploring curcumin’s influence on cell motility, proliferation, cell cycle, and apoptosis, particularly in TNBC, and uncovering its molecular underpinnings are of paramount clinical relevance.

In this study, our objective is to elucidate the antitumor properties of curcumin, focusing on its potential mechanisms of action and therapeutic efficacy. Employing gene enrichment analysis, incorporating both Gene Ontology and Kyoto Encyclopedia of Genes and Genomes pathways, on RNA sequencing data, we determined that curcumin predominantly acts through pathways such as Pathways in cancer, Focal adhesion, PI3K-Akt signaling pathway, and cell cycle. The results of our enrichment analysis showed a prominent ranking for the pathways in cancer signaling pathway. According to these findings and in conjunction with our existing experimental results, we have decided to further explore the epithelial-mesenchymal transition (EMT) signaling pathway. EMT is a multifaceted process where epithelial cells undergo transformation, losing cell-cell adhesion and acquiring a migratory, invasive mesenchymal phenotype.^[28–32] This dynamic and reversible process sees cancer cells traversing the EMT spectrum, encompassing epithelial, partial EMT, and mesenchymal states, and is implicated in various aspects of cancer malignancy, including tumor invasion, metastasis, and treatment resistance.^[33,34] During EMT, cells progressively relinquish epithelial characteristics and acquire mesenchymal traits, enhancing their migratory and invasive abilities.^[35]

In exploring cancer biology, understanding the interaction between the PI3K-Akt-mTOR signaling pathway and EMT is of significant importance. Akt activation has been shown to enhance the activity of key transcription factors directly involved in EMT, such as Snail, Slug, and Twist.^[31,36]Furthermore, mTOR activation significantly affects the expression of proteins related to the cytoskeleton and cell migration, thus facilitating EMT. This has crucial significance for the invasiveness and metastasis of cancer cells.^[37,38] The PI3K-Akt-mTOR pathway’s activation stems not only from intrinsic cellular mechanisms but is also influenced by extracellular signals like growth factors and cytokines, which are vital in modulating EMT.^[39]Therefore, the regulatory role of this pathway in EMT within tumor cells is a critical focus of our research.

The CCK8 assay indicated that curcumin markedly reduces the cellular activity of TNBC cells. In parallel, Western Blot analysis revealed that curcumin lowers the expression of key proteins, including N-Cadherin, Fibronectin, β-Catenin, p-Akt, Akt, mTOR, p-S6, and S6, in Hs578T and MDA-MB-231 cells. These proteins play crucial roles in the EMT process and its related signaling pathways, particularly the PI3K/Akt/mTOR pathway. The results imply that curcumin’s inhibitory effect on the motility of TNBC cells may stem from the simultaneous downregulation of these signaling pathways, impacting the EMT signaling cascade. Additionally, further experiments are required to assess the cytotoxic effects of curcumin on normal cells and to explore the anticancer efficacy of curcumin in vivo.

5. Conclusion

Curcumin demonstrates potential anticancer properties in the treatment of TNBC. The findings suggest that the inhibitory effect of curcumin on the motility of TNBC cells could be attributed to the concurrent downregulation of specific signaling pathways, through influencing the EMT signaling process. This study employs a comprehensive approach that integrates bioinformatics analysis with in vitro experimental methodologies, providing substantial evidence supporting curcumin’s potential in breast cancer therapy.

Acknowledgments

This work was supported by Sichuan Science and Technology Program (No. 2022YFS0623).

Author contributions

Conceptualization: Lin Gan.

Data curation: Ze Chen, Pinjun Lu, Menghan Li.

Formal analysis: Menghan Li, Qing Zhang, Lin Gan.

Funding acquisition: Tao He, Lin Gan.

Investigation: Ze Chen, Pinjun Lu.

Methodology: Ze Chen, Menghan Li, Lin Gan.

Project administration: Tao He, Lin Gan.

Resources: Lin Gan.

Software: Ze Chen, Menghan Li, Qing Zhang.

Supervision: Qing Zhang, Tao He, Lin Gan.

Validation: Ze Chen, Pinjun Lu.

Visualization: Ze Chen.

Writing – original draft: Ze Chen.

Writing – review & editing: Tao He, Lin Gan.