Skip to main content
NCBI home page
BMC Cancer logoLink to BMC Cancer
. 2024 Oct 8;24:1243. doi: 10.1186/s12885-024-12999-9

Systematic screening of protein-coding gene expression identified VWF as a potential key regulator in anthracycline-based chemotherapy-exacerbated metastasis of breast cancer

Yawei Zhao 1, Meihui He 1,4, Lianzhi Cui 3, Min Zhang 1, Tianyu Zhao 1, Xuehan Yang 1, Yang Xu 1, Jianhua Dong 1, Kan He 1, Hansi Zhang 1,✉, Li Chen 1,2,✉
PMCID: PMC11462902  PMID: 39379897

Abstract

Background

Breast cancer is the most commonly diagnosed cancer worldwide. Although major treatments represented by chemotherapy have shown effectiveness at the initial period, recurrence and metastasis still occur later after treatments. The alternation of the tumor microenvironment by chemotherapy is confirmed as a trigger of the elevated proliferation and migration of the remaining tumor cells.

Methods

Using bioinformatic methods, differential gene expression analysis was used to determine DEGs between post-chemotherapy and pre-chemotherapy samples of breast cancer patients, followed by survival analysis and ELISA analysis of the potential key genes. An in vitro model of 2 breast cancer cells lines was used to demonstrate the role of VWF in the evasion and migration of breast cancer cells, using cell migration, evasion and wound healing assays, PCR and molecular docking analysis.

Results

19 hub genes were further identified using GO and KEGG pathway analyses and WGCNA. The 5 secreted protein-coding genes with reported carcinogenesis effects (VWF, SVEP1, DPT, ADIPOQ, and LPL) were further analyzed in breast cancer patients and VWF was identified as a potential key regulator in the anthracycline-based chemotherapy-exacerbated metastasis. It was further confirmed that anthracycline-based chemotherapeutics doxorubicin exacerbated VWF upregulation and the evasion and migration of breast cancer cells. Based on molecular docking analysis and previous study, berberine was used as an inhibitor of VWF, and showed an effective inhibition of the doxorubicin-exacerbated VWF upregulation, migration and evasion in breast cancer.

Conclusions

Doxorubicin-exacerbated evasion and migration through VWF upregulation. Berberine as an inhibitor of VWF was able to reversed the doxorubicin-exacerbated VWF upregulation and evasion and migration in breast cancer cells.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12885-024-12999-9.

Keywords: Breast cancer, Chemotherapy, Doxorubicin, VWF, Metastasis

Background

Breast cancer is the leading cause of both incidence and mortality in female malignancies, and became the most commonly diagnosed cancer worldwide since 2020 [1]. The treatment of breast cancer usually contains surgery as the mainstay, supplemented by chemotherapy, radiotherapy and endocrine therapy [2]. As one of the most common treatments of cancer, chemotherapy has long been an effective and well-recognized therapeutics agent. Despite its sufficient killing effect of tumor and the clinical benefit in a short-term after treatment, the metastasis in a long-term concern is still a major cause of death in breast cancer. Recently, new approaches of neoadjuvant chemotherapy has been emerging as a development of traditional chemotherapy, however, studies of long-term outcomes has indicated that it was related to higher risk of recurrence and micrometastatic spread in breast cancer [3]. Anthracycline, including doxorubicin (Dox) and epirubicine is commonly used in systematic and neoadjuvant chemotherapy [4] of breast cancer, a thorough understand of effects of anthracycline-based chemotherapy beyond cytotoxicity is still needed for taking a better advantage of chemotherapy and avoid the possible pro-metastatic effect.

Despite the cytotoxic effect against tumor, accumulating evidence have shown that chemotherapy may inherently be a double-edged sword. Expect for the side effects caused when the drugs damage fast-growing normal cells [10], like anemia, loss of appetite, delirium, double-edged effects were also found in tumor cells themselves, including tumor regrowth and pro-metastatic effects of the remaining cells post-chemotherapy due to wound healing-related mechanisms [5]. Alternation of the tumor microenvironment (TME) is one of the major effects induced by chemotherapy [6].

Different studies concerning different cancers and chemotherapeutics have shown that, at the initial period of the chemotherapy treatment, along with the damage of tumor cells and other tissue, multiply alternation in TME was also triggered, such as the release of C-C motif chemokine ligand (CCL) 11, CCL4, IL-1β, IL-10, IL-6, C-X-C motif chemokine ligand (CXCL) 1, Prostaglandin (PG) E2, etc [7]. These secreted proteins then triggered their downstream signaling which increased the proliferation, migration, angiogenesis, immune evasion or other pro-tumorigenic abilities, thereby triggered the secondary progression of cancer in the later period of chemotherapy treatment [8]. Although this phenomenon was confirmed in breast cancer, among all the altered secretions, which ones may play more crucial role in the anthracycline-based chemotherapy-exacerbated migration in breast cancer is still worth to be explored.

With the extensive development of microarray technology, abundant, massive and complex biological information data has been generated [9]. Using bioinformatics methods, differentially expressed genes (DEGs) between post- and pre-chemotherapy samples of breast cancer patients were determined in our study. And a secreted protein von willebrand factor (VWF) was determined as a potential key regulator of the anthracycline-based chemotherapy-exacerbated migration in breast cancer.

VWF is a secreted multimeric plasma glycoprotein which induces platelet adhesion in hemostasis [10]. Traditionally, it is confirmed to be derived from the platelet, recent studies have shown that VWF is highly expressed in some types of tumor cells such as osteosarcoma, colorectal cancer and hepatocellular carcinoma in both gene and protein levels [11]. And VWF derived from either platelet or tumor cells has shown pro-tumorigenic effects including metastasis and evasion [12]. However, few researches have focused on the anthracycline-based chemotherapy-induced VWF secretion and its effect on cancer progression.

In our study, differential gene expression analysis was used to determine DEGs between post- and pre-chemotherapy samples of breast cancer patients, in order to explore the TME alternation induced by anthracycline-based chemotherapy. 19 hub genes were further identified using Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses and weighted gene co-expression network analysis (WGCNA). The 7 secreted protein-coding genes were further analyzed by literature validation, survival analysis and analysis of serum level in breast cancer patients. And VWF was identified as a potential key regulator in the anthracycline-based chemotherapy-exacerbated metastasis. The pro-metastatic effect of VWF was confirmed in an in vitro model of Dox-treated breast cancer cells. And based on our previous study and a molecular docking validation, we used berberine as an inhibitor of VWF in vitro, and it showed an effective inhibition of the Dox-exacerbated migration and evasion in breast cancer.

Methods

Acquisition and preprocessing of microarray data

Breast cancer-associated dataset GSE21974 in Gene Expression Omnibus (GEO) database (http://www.ncbi.nlm.nih.gov/geo) was used [13]. The GSE21974 dataset contained samples collected from 25 breast cancer patients pre- and post- anthracycline-based chemotherapy, respectively. The samples were profiled using the chip-based platform GPL6480. the probes were filtered without known gene symbols before DEGs analysis, and in order to merge probes, we converted the probe-level expression profiles of the dataset to gene-level expression with collapseRows function [14].

Analysis of DEGs

R software and Bioconductor 3.3.2 (http://www.bioconductor.org/) were used to perform the Microarray data analysis. DEGs between pre- and post-chemotherapy samples of breast cancer patients were identified using R Limma package [15]. The fold change > 2 and P < 0.05 were regarded as the threshold of differential expression.

GO term and KEGG pathway enrichment analyses

GO and KEGG enrichment analyses of DEGs were performed using R ClusterProfiler package [16]. GO enrichment analysis was divided into 3 parts: biological processes (BP), cellular components (CC) and molecular functions (MF). The GO function enrichment and KEGG metabolic pathway enrichment of DEGs and the network modules genes might be the pivotal pathways closely related to chemotherapy of breast cancer [17]. Accordingly, we could further explore the pathways involved in hub genes. P < 0.05 was used as the threshold of significant enrichment pathways.

WGCNA

WGCNA used the idea of systems biology to describe the similarity of gene co-expression and to identify highly correlated gene modules, providing basis for screening biomarkers and targets [18]. In the co-expression network, the connectivity degree of one gene with other genes was defined as intramodular connectivity (IC), Module eigengene (ME) as the first principal component of a module. Module membership (MM) as Pearson’s correlation coefficient between gene expression levels and the first principal component genes. Gene significance (GS) was used to measure the degree of association between genes and clinical information. Module significance (MS) described the average value of GS for all genes in the module.

Identification of hub genes

In this study, hub genes were most closely related to chemotherapy, and they had the highest degree of connection with the other genes in the module [19]. If a gene had distinct character, such as high MM, high IC, or high GS in the network, it could be considered as a hub gene [20]. The co-expression network of module genes was constructed using Cytoscape software.

Verification and survival analysis of hub genes

According to the literature verification and survival analysis of the hub genes, we could predict which key genes might be highly related to breast cancer metastasis and poor overall survival [21]. The PROGgeneV2 database (http://watson.compbio.iupui.edu/chirayu/proggene/database/?url=proggene) was used to identify whether the expression of hub genes could predict patients’ survival [22]. Patients were grouped into high or low gene expression levels based on median gene expression of all. Then Kaplan-Meier (KM) plots were created to describe the correlation between gene expression and overall survival from The Cancer Genome Atlas (TCGA). Survival analysis was conducted using a threshold of P < 0.05. Furthermore, based on the key genes from literature verification, pathways related to breast cancer metastasis were explored by KEGG enrichment analysis.

Cell culture

Human breast cancer cell lines MCF-7 and MDA-MB-231 were obtained from the National Collection of Authenticated Cell Cultures (Shanghai, CHN). Cells were maintained in DMEM (Cat#11965092, Gibco, USA) supplemented with 10% Fetal Bovine Serum (Cat#FB15015, Gibco, USA), 100 IU/ml penicillin, 100 µg/ml streptomycin, within an incubator with 5% CO2 at 37℃.

Cells were treated with Dox (Qilu Pharmaceutical Co., Ltd, CHN) at the concentration of 0.5 µg/ml, and/or Berberine chloride hydrate (Northeast Pharmaceutical Group Co., Ltd., Shenyang, CHN) at the concentration of 5µM for 24 h, and the supernatant was then discarded and cells were rinsed twice with PBS. Fresh medium was replaced then. The remaining cells were used as feeder cells for migration and evasion assay. For wound-healing assay, the cells were incubated for another 24 h, and the conditioned medium was used in the wound-healing assay. All in vitro experiments were replicated in triplicate.

Quantitative real-time PCR (qPCR)

Total RNA was extracted and qPCR was performed as described previously [23]. GAPDH was used as the internal control. Primers are as follows:

ADIPOQ: F: TATCCCCAACATGCCCATTCG,

R: TGGTAGGCAAAGTAGTACAGCC.

VWF: F: AGAAACGCTCCTTCTCGATTATTG,

R: TGTCAAAAAATTCCCCAAGATACAC.

LPL: F: GGAATGTATGAGAGTTGGGTGC,

R: CAATGCTTCGACCAGGGGACC.

SVEP1: F: GCAACTTGGGCGTGGTTA,

R: CACACCGCTGACCTGTGTAA.

GAPDH: F: TGCACCACCAACTGCTTAGC,

R: GGCATGGACTGTGGTCATGAG.

Clinical samples

The serum samples of breast cancer patients underwent with (n = 17) or without (n = 13) chemotherapy were collected for ELISA analysis of VWF level. The work was approved by the Ethics Committee of Jilin University (No. 2021100701). All patients signed written informed consent.

ELISA assay

The ELISA assay of VWF level was carried out using Human VWF ELISA Kit (Cat#ab108918, Abcam, USA) following the manufacturer’s instructions.

Cell migration and evasion assays

Cell migration assay [23, 24] and evasion assay [25] was carried out as previously described. MCF-7 and MDA-MB-231 cells were seeded as receptor cells into the upper chamber of the Transwell system, as the Dox-treated cells mention above served as feeder cells. The 24-well 8.0 μm pore size transwell chambers (Cat#3464, Corning, USA) were used in the migration and evasion assays. The matrigel (Cat#354234, Corning, USA) was used in the evasion assays.

Wound-healing assay

The wound-healing assay were carried out as previously reported [23]. The conditioned medium of the two cell lines mentioned above was used in the experiments. Images at 0 and 24 h post-scratching were taken respectively. The wound healing % = (width of 0 h – width of 24 h)/ width of 0 h × 100%.

Statistical analysis

Statistical analysis was performed by GraphPad Prism 7.0. Results were presented as means ± SD unless indicated otherwise. Significance of difference between groups was assessed by the Student’s t test for single comparisons or the analysis of variance (ANOVA) with the Newman-Keuls tests for multiple comparisons. A value of P < 0.05 was considered statistically significant.

Results

Identification of hub genes associated with anthracycline-based chemotherapy in breast cancer

Based on our previous reports [23, 25] and other studies [17], chemotherapy may act as not only a killer of the tumor cells, but also an initiator of the secondary progression. In order to further understand the possible factors altered by anthracycline-based chemotherapy that may trigger the metastasis in breast cancer, bioinformatics tools were used to identify the hub genes associated with anthracycline-based chemotherapy in breast cancer. Breast cancer-associated dataset GSE21974 from patients pre- and post- anthracycline-based chemotherapy was analyzed. All patients were with histologically diagnosed primary breast cancer (with basal- or non-basal-like subtypes, grade 1–3) [26] (Table S1).

1656 DEGs were analyzed, and 1110 genes were up-regulated and 546 were down-regulated (fold change > 2, P < 0.05) (Fig. 1A). The 1656 DEGs were further analyzed by hierarchical clustering. The general expression patterns in pre- and post-chemotherapy samples were significantly different by TreeView (Fig. 1B). In order to reveal the pathways associated with breast cancer chemotherapy, KEGG enrichment analysis of 1656 DEGs was performed. As shown in Fig. S1A, The DEGs were enriched mainly in cell cycle, peroxisome proliferator-activated receptor (PPAR) signaling pathway, progesterone-mediated oocyte maturation, regulation of lipolysis in adipocytes and focal adhesion. In GO analysis (Fig. S1B), the significant biological functions of DEGs were proteinaceous extracellular matrix, chromosome segregation, mitotic nuclear division, extracellular matrix organization and extracellular structure organization.

Fig. 1.

Fig. 1

Open in a new tab

Identification of anthracycline-based chemotherapy-related hub genes. (A) Volcano plot of the DEGs. DEGs were selected with P < 0.05 and fold change > 2. (B) DEGs were effectively divided into post- and pre- anthracycline-based chemotherapy groups (up-regulated and down-regulated genes represented by red and green). (C) Gene dendrogram obtained by average linkage hierarchical clustering. (D) Heatmap plot of topological overlap in the gene network. Genes in the rows and columns were sorted by the clustering tree. Clusters corresponded to squares along the diagonal. (E) Module significance of the gene modules to chemotherapy

WGCNA was then performed. Before constructing the network, the soft-threshold β = 9 was selected according to the correlation coefficient close to 0.8 (Fig. S2). The 1656 DEGs were then used for network construction. Five gene modules were identified and was assigned with a different color each (Fig. 1C). The heatmap plot of topological overlap in the gene network is depicted (Fig. 1D). The MEs were calculated to evaluate the physiologic significance of the modules and to search for significant associations (Fig.S3). According to the heatmap plot, genes clustered in blue module had the strongest positive correlation with chemotherapy (r = 0.64, P = 5.0E-07) and the MS values of the blue module was the highest among five modules (MS = 0.42, Fig. 1E). Therefore, the subsequent analysis focused on blue module.

To explore the functional implication of DEGs in blue module, the GO functional enrichment was conducted and three main annotated categories were obtained in GO database (Fig. 2A, Table S2). In the biological process, categories of extracellular matrix organization, extracellular structure organization and blood vessel morphogenesis were significantly enriched. In the case of cellular component enrichment, proteinaceous extracellular matrix, extracellular matrix component, and collagen trimer were well marked. Among the categories of molecular function, glycosaminoglycan binding, heparin binding, and sulfur compound binding should also be highlighted. Moreover, the KEGG annotation results of the genes in blue module were classified (Fig. 2B). The top ten significantly enriched pathways are listed in Table S3, such as PPAR signaling pathway, extracellular matrix (ECM)-receptor interaction, protein digestion and absorption, focal adhesion, etc.

Fig. 2.

Fig. 2

Open in a new tab

Identification of hub genes associated with anthracycline-based chemotherapy in breast cancer. The top 10 significantly enriched GO terms (A) and the top 10 KEGG pathways (B) identified in blue module by P < 0.05. (C) Scatter plot of MM vs. GS in blue module showed a high correlation. Cor represented an absolute correlation coefficient of GS and MM; P-value for significance assessment. (D) Construction of a co-expression network of blue module genes. The red nodes represented the hub genes with highest MM. Then followed the yellow and blue nodes represented the genes with median and lower MM. Edges between two nodes showed the interactions between genes

In order to identify the genes most relevant to chemotherapy in breast cancer, GS, MM, and IC of each module were calculated, and blue module exhibited very significant positive correlations (Fig. 2C). A total of 19 hub genes in blue module was identified based on the MM values (MM > 0.9, Table S4). The gene co-expression network in blue module was visualized in Fig. 2D.

VWF may be a key regulator in the anthracycline-based chemotherapy-exacerbated metastasis in breast cancer

It has been confirmed that one important mechanism of chemotherapy-exacerbated metastasis in different tumors including breast cancer is the alternation of the microenvironment through secretion of soluble factors [8, 17, 27]. It intrigued us to explore the possible secreted factors exacerbated by chemotherapy which play a key role in the anthracycline-based chemotherapy-exacerbated migration in breast cancer. Therefore, the subcellular locations of the hub genes were confirmed by Uniprot (https://www.uniprot.org). Among the 19 hub genes upregulated by breast cancer chemotherapy, 7 of them codes for secreted proteins. And 5 among the 7 genes were reported with carcinogenesis effects such as pro-metastatic or proliferation effects [28–32], including VWF, SVEP1, DPT, ADIPOQ, and LPL (Table S4, Fig. 3A). Survival analysis of the 5 potential genes was performed and 4 of them except for DPT indicated a poor prognosis with worse overall survival in breast cancer (Fig. 3B). The expression of the 4 genes were then analyzed in breast cancer cells MCF-7 (with luminal A characteristics as a representative of the most commonly diagnosed subtype of breast cancer) treated by anthracycline-based chemotherapeutics Dox. All 4 genes were significantly upregulated after chemotherapy treatment, with the upregulation level of VWF being the highest (Fig. 3C). The expression of VWF in breast was further analyzed by gene expression profiling interactive analysis (GEPIA) 2 (http://gepia2.cancer-pku.cn), and it was shown that VWF expression was significantly increased in breast cancer compared with normal breast tissue (Fig. 3D). And the serum VWF level in breast cancer patients was significantly increased after anthracycline-based chemotherapy treatments with anthracycline regimen or anthracycline combined taxane regimen (Fig. 3E). Based on these, we proposed that VWF as the putative gene playing a crucial role in the chemotherapy-exacerbated metastasis in breast cancer.

Fig. 3.

Fig. 3

Open in a new tab

VWF may be a potential regulator in the anthracycline-based chemotherapy-exacerbated metastasis in breast cancer. (A) The subcellular locations of the hub genes were confirmed by Uniprot (https://www.uniprot.org). 5 among the 7 genes were reported with carcinogenesis effect such as pro-metastatic or proliferation effects. (B) The overall survival curves illustrated via the Kaplan-Meier plotter, with high (red) and low (green) expression of each hub genes in patients. A threshold Cox P-value < 0.05 was used in the survival curve analysis. **P < 0.01, *P < 0.05. (C) The mRNA expression of the 4 genes were analyzed in breast cancer cells MCF-7 treated by Dox, student’s t test, **P < 0.01. (D) The expression of VWF in breast cancer compared with normal breast tissue was further analyzed by GEPIA2 (http://gepia2.cancer-pku.cn). (E) Serum VWF level in breast cancer patients before and after anthracycline-based chemotherapy was analyzed by ELISA, student’s t test, *P < 0.05. Experiments were replicated in triplicate

Dox exacerbated VWF upregulation and the evasion and migration of breast cancer cells

Studies showed that VWF level in patients with metastatic breast cancer was elevated [10], indicating a potential role of VWF in the metastasis of breast cancer. To further analyze the role of anthracycline-based cytotoxic chemotherapy-induced VWF upregulation in breast cancer metastasis in vitro, two human breast cancer cell lines MCF-7 and MDA-MB-231 were used. In the cohort of patients pre- and post-chemotherapy (Table S1) [26], the characteristics of the patients showed that 7/25 of the molecular subtype were basal, with a triple-negative character, this can be represented by MDA-MB-231 cells in the in vitro experiment. And among the non-basal-like subtypes, luminal A is the most commonly diagnosed subtypes in breast cancer [33], therefore MCF-7 cells with luminal A characteristics was used as a representative of the non-basal-like subtypes. The above cells were treated with Dox for 24 h and the Dox-treated cells were used as feeder cells in the Transwell migration analysis. The conditioned medium of the above cells was used in the wound healing assay. As it was shown in Fig. 4A-D and Figure S4A-D, migration and wound healing assay showed that Dox treatment significantly exacerbated the migration of both cell lines (con group vs. Dox group). Similar results were shown in the evasion assay of the two cell lines (Fig. 4E-F, Fig. S4D-E). And VWF was significantly upregulated post Dox treatment (Fig. 4G and Fig S4G). The above results confirmed that Dox exacerbated the expression of VWF as well as the migration and evasion in breast cancer.

Fig. 4.

Fig. 4

Open in a new tab

Dox-exacerbated VWF upregulation and the evasion and migration of MCF-7 cells. (A-B) Breast cancer cell line MCF-7 was treated with Dox for 24 h and the chemotherapy-treated cells were used as feeder cells in the Transwell migration analysis. The solubilized crystal violet was quantified at the absorbance of 570 nm. One-way ANOVA, *P < 0.05 compared with con, # P < 0.05 compared with Dox. (C-D) The conditioned medium of the above cells was used in the wound healing assay. Images were taken at 0 and 24 h after scratching. Wound healing % were calculated as indicated. One-way ANOVA, *P < 0.05 compared with con, # P < 0.05 compared with Dox. (E-F) The feeder cells in A were used in the evasion assay and the solubilized crystal violet was quantified at the absorbance of 570 nm. One-way ANOVA, *P < 0.05 compared with con, # P < 0.05 compared with Dox. (G) the VWF expression in MCF-7 cells were analyzed by qPCR 24 h after the Dox was replaced by normal medium. One-way ANOVA, *P < 0.05 compared with con, # P < 0.05 compared with Dox. Experiments were replicated in triplicate

Our previous studies showed that berberine reversed the Dox-exacerbated inflammatory TME and inhibited the repopulation and migration of tumor cells [24, 25, 34, 35], which indicated a potential effect of berberine on improving the TME post- anthracycline-based chemotherapy. Based on this, a molecular docking validation between berberine and 3 major domains of VWF (A1-A3) [10] was performed to predict the possibility of berberine targeting of VWF. The binding free energies were − 27.2585 kcal/mol (A1), -35.7945 kcal/mol (A2) and − 29.3415 kcal/mol (A3), all showing strong binding capacity (Fig. 5). This indicated that berberine may be a promising drug targeting VWF. On the other hand, we also explored the effect of berberine on VWF expression and the Dox-exacerbated migration and evasion in breast cancer. Berberine was used in the migration (Fig. 4A-B, Fig S4A-B), wound healing (Fig. 4C-D, Fig S4C-D) and evasion assay (Fig. 4E-F, Fig S4E-F) of the Dox-treated breast cancer cells mentioned above. It was shown that the Dox-induced VWF upregulation was reversed by berberine (Fig. 4G, Fig S4G). And berberine significantly reversed the Dox-exacerbated migration and evasion in both breast cancer cell lines. The above results indicated that Dox-exacerbated VWF upregulation induced the migration and evasion in breast cancer cells, berberine could inhibit the Dox-exacerbated VWF expression and inhibit the migration and evasion post Dox treatment.

Fig. 5.

Fig. 5

Open in a new tab

Molecular docking between berberine and 3 major domains of VWF. (A-C) Molecular docking results between berberine and A1-A3 domains of VWF

Discussion

Breast cancer is the top 1 malignancy in both new cases and deaths among all kinds of cancers in women [1]. Currently, the major treatment of breast cancer is still chemotherapy combined with surgery. Chemotherapy is given either preoperative or postoperative. Anthracycline-based cytotoxic such as Dox are common choices for systemic therapy [4]. Being the major treatment for systemic therapy in almost all kinds of cancers, chemotherapy showed sufficient cytotoxicity against tumor cells. However, more and more studies have also revealed its pro-tumorigenic effects including metastasis and evasion, which may lead to the development of a secondary cancer [8, 36–39]. There is no doubt that chemotherapy is currently the most common and effective way for cancer treatment, but these recent findings also indicated that exploring the underlying mechanism of the exacerbated metastasis potential post chemotherapy and targeting the key regulator may be a promising strategy to preserve the killing effect of chemotherapy and inhibit the exacerbated metastasis afterwards.

Mechanism researches have suggested that the exacerbated metastasis post chemotherapy or radiation may be an inevitable consequence of wound healing process after massive cells death induced by chemotherapy [5, 38–40]. A decade ago, Huang etc. proposed a theory that apoptotic cells activated pathways to promote wound healing and tissue regeneration described it as “phoenix rising“ [41]. Elevated secretion of PGE2 was confirmed to be a promoter of tissue regeneration activated by the caspase cascade in dying cells [41]. And this phenomenon was also confirmed in cancer cells when large number of cells were killed by chemotherapy or radiation [5]. Further studies have shown that chemotherapy could act as a stress and trigger the phenotypic plasticity of tumor cells. This leaded to a stem-like cell state of the tumor cells, which is more aggressive and chemoresistance, with higher ability of migration and proliferation [37]. Cancer cell-derived soluble factors is one of the major regulators of the above progress [27]. Chemotherapy agents induced changes of virous kinds of factors such as PGE2, arachidonic acid, HMGB1 in the TME and ultimately triggered the second progression of cancer [7, 23, 25]. In our study, bioinformatic methods were used to identify the secreted protein-coded hub genes in breast cancer patients post anthracycline-based chemotherapy. And by further literature validation, survival analysis and analysis of serum level of the above genes in breast cancer patients, VWF was predicted as a potential key regulator in the chemotherapy-exacerbated metastasis in breast cancer. To further explore the role of VWF in chemotherapy-exacerbated migration in breast cancer cells, we established the post- anthracycline-based chemotherapy Transwell in vitro system using 2 breast cancer cells. And the results confirmed that VWF is exacerbated after Dox and inhibition of VWF reversed the Dox-exacerbated migration.

VWF is a secreted multimeric plasma glycoprotein which induces platelet adhesion to the endothelium [12]. Its major physiological role is hemostasis and thereby is closely associated with thrombosis [42]. In pathological condition of cancer, studies have shown that VWF was elevated in tumors and associated with worse disease progression [11, 12, 43–45]. In breast cancer, studies have confirmed elevated VWF level in patients with metastatic breast cancer [10], which may support our findings. Traditionally, VWF was reported to be derived from endothelial cells. Recent findings showed that cancer cells also secret VWF, and both endothelial and tumor-derived VWF could increase the metastasis of cancer and promote cancer progression [10, 12, 44]. In our study, VWF was found to be elevated in tumor tissue and in serum compared with normal ones, and positively related with worse overall survival. And in vitro studies confirmed that Dox treatment induced the expression of VWF, which induced the evasion and migration of breast cancer cells. VWF was reported to induce the adhesion of gastric cancer cells by its receptor Glycoprotein (GP) Ibα [44], and GPIbα was also expressed in human breast cancer cells and interacted with VWF [10, 46], indicating that VWF may induce the migration through its receptor GPIbα. Other studies showed that VWF mediated the metastasis of osteosarcoma through NF-κB signaling [11], and the antagonist of GPIbα was shown to inhibit the activation of NF-κB [47]. We didn’t further explore the downstream signaling of VWF, it is possible that VWF induced the migration through interacting with its receptor GPIbα and activating the NF-κB signaling pathway.

In our study, Dox was used as a representative of anthracycline-based cytotoxic chemotherapeutics. Our previous studies and other studies have reported series findings based on the phenomenon of secondary development after chemotherapy. The chemotherapeutic agents included etoposide, carboplatin, gemcitabine, cisplatin and Dox and the types of cancer are various [5, 17, 23, 25, 34, 48]. Mechanically, the above cytotoxic chemotherapeutics can induce the apoptosis of cancer cells and leads to a tumor-promoting TME change during the process, which induces a secondary development of cancer. We regarded this as a common effect of cytotoxic chemotherapeutics on cancer cells. One limitation of our study is that although Dox is proved to elevate the VWF expression, other anthracycline-based cytotoxic chemotherapeutics was not used in the in vitro experiments. Since the 7 secreted factors predicted by the bioinformatic methods were analyzed from patients receiving epirubicine and cyclophosphamide, and based on our previous theory, there is possibility that VWF may also be elevated by other anthracycline-based cytotoxic chemotherapeutics. Another limitation is that we conducted a retrospective cohort study when analyzing the serum VWF level in breast cancer patients. The patients receiving chemotherapy or not were different individuals. It would further improve this experiment by adding VWF levels of following-up the same patient before and after chemotherapy, which can be further explored in future studies.

Our previous studies showed that berberine reversed the chemotherapy-exacerbated inflammatory TME and inhibited the repopulation and migration of tumor cells [24, 25, 34, 35], which indicated a potential effect of berberine on improving the TME post-chemotherapy. In our study, we performed a molecular docking analysis which predicted high binding capacity between berberine and 3 major domains of VWF. This indicated that berberine may be targeting VWF and therefore may affect the interaction of VWF and its downstream factors. Studies have also shown that berberine inhibited the plasma VWF in coronary microembolization mice [49], indicating the inhibition VWF secretion by berberine. Using berberine as an inhibitor of VWF, we further confirmed that anthracycline-based chemotherapy-exacerbated VWF upregulation and the invasion and migration of breast cancer cells.

Conclusions

In conclusion, we identified 19 hub genes associated with anthracycline-based chemotherapy in breast cancer. By further literature validation and survival analysis, we showed that VWF was a potential key regulator of the anthracycline-based chemotherapy-exacerbated evasion and migration. And in vitro experiments further confirmed the anthracycline-based chemotherapeutics Dox-exacerbated evasion and migration. VWF was significantly upregulated post Dox treatment. Based on previous studies and molecular docking we used berberine as an inhibitor of VWF, and showed that berberine was able to reversed the Dox-exacerbated VWF upregulation and evasion and migration in breast cancer cells.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1 (4.7MB, doc)

Acknowledgements

Not applicable.

Abbreviations

Dox

Doxorubicin

TME

Tumor microenvironment

CCL

C-C motif chemokine ligand

CXCL

C-X-C motif chemokine ligand

PG

Prostaglandin

DEGs

Differentially expressed genes

VWF

Von willebrand factor

WGCNA

Weighted gene co-expression network analysis

GEO

Gene Expression Omnibus

GO

Gene Ontology

KEGG

Kyoto Encyclopedia of Genes and Genomes

BP

Biological processes

CC

Cellular components

MF

Molecular functions

IC

Intramodular connectivity

ME

Module eigengene

MM

Module membership

GS

Gene significance

MS

Module significance

KM

Kaplan-Meier

TCGA

Cancer Genome Atlas

qPCR

Quantitative real-time PCR

PPAR

Peroxisome proliferator-activated receptor

ECM

Extracellular matrix

GEPIA

Gene expression profiling interactive analysis

GP

Glycoprotein

Author contributions

Y.Z: Conceptualization, Writing - Original Draft, Visualization, Funding acquisition; M.H: Writing - Original Draft, Formal analysis; L.C: Resources, Data Curation; M.Z.: Investigation, Methodology; T.Z.: Formal analysis; X.Y.: Supervision; Y.X.: Validation; J.D.: Data Curation; K.H.: Conceptualization, Funding acquisition; H.Z.: Writing - Review & Editing, Conceptualization; L.C.: Writing - Review & Editing, Conceptualization, Supervision.

Funding

This study was funded by Education Department of Jilin Provincial (No. JJKH20231237KJ, JJKH20221103KJ), National Natural Science Foundation of China (No. 82204444), Bethune Program of Jilin University (No. 2023B35), China Postdoctoral Science Foundation (No. 2022M711311).

Data availability

The datasets used and/or analyzed in the study are included in the article/Supplementary Material. Further inquiries can be directed to corresponding authors.

Declarations

Ethics approval and consent to participate

The work was approved by the Ethics Committee of Jilin University (No. 2021100701). All patients signed written informed consent.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Contributor Information

Hansi Zhang, Email: hansi_z@163.com.

Li Chen, Email: chenl@jlu.edu.cn.

References

  • 1.Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global Cancer statistics 2020: GLOBOCAN estimates of incidence and Mortality Worldwide for 36 cancers in 185 countries. Cancer J Clin. 2021;71:209–49. [DOI] [PubMed] [Google Scholar]
  • 2.Gradishar WJ, Moran MS, Abraham J, Aft R, Agnese D, Allison KH, et al. Breast Cancer, Version 3.2022, NCCN Clinical Practice guidelines in Oncology. J Natl Compr Cancer Network: JNCCN. 2022;20:691–722. [DOI] [PubMed] [Google Scholar]
  • 3.Volmer L, Koch A, Matovina S, Dannehl D, Weiss M, Welker G, et al. Neoadjuvant chemotherapy of patients with early breast Cancer is Associated with increased detection of disseminated Tumor cells in the bone marrow. Cancers. 2022;14:635. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Gradishar WJ, Anderson BO, Balassanian R, Blair SL, Burstein HJ, Cyr A, et al. NCCN guidelines insights breast Cancer, Version 1.2016. J Natl Compr Cancer Network: JNCCN. 2015;13:1475–85. [DOI] [PubMed] [Google Scholar]
  • 5.Huang Q, Li F, Liu X, Li W, Shi W, Liu FF, et al. Caspase 3-mediated stimulation of tumor cell repopulation during cancer radiotherapy. Nat Med. 2011;17:860–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Shaked Y. The pro-tumorigenic host response to cancer therapies. Nat Rev Cancer. 2019;19:667–85. [DOI] [PubMed] [Google Scholar]
  • 7.McLeary F, Davis A, Rudrawar S, Perkins A, Anoopkumar-Dukie S. Mechanisms underlying select chemotherapeutic-agent-induced neuroinflammation and subsequent neurodegeneration. Eur J Pharmacol. 2019;842:49–56. [DOI] [PubMed] [Google Scholar]
  • 8.Gandhi S, Chandna S. Radiation-induced inflammatory cascade and its reverberating crosstalks as potential cause of post-radiotherapy second malignancies. Cancer Metastasis Rev. 2017;36:375–93. [DOI] [PubMed] [Google Scholar]
  • 9.Goswami CP, Nakshatri H. PROGgeneV2: enhancements on the existing database. BMC Cancer. 2014;14:970. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Goh CY, Patmore S, Smolenski A, Howard J, Evans S, O’Sullivan J, et al. The role of Von Willebrand factor in breast cancer metastasis. Transl Oncol. 2021;14:101033. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Eppert K, Wunder JS, Aneliunas V, Kandel R, Andrulis IL. Von Willebrand factor expression in osteosarcoma metastasis. Mod Pathol. 2005;18:388–97. [DOI] [PubMed] [Google Scholar]
  • 12.Patmore S, Dhami SPS, O’Sullivan JM. Von Willebrand factor and cancer; metastasis and coagulopathies. J Thromb Haemost. 2020;18:2444–56. [DOI] [PubMed] [Google Scholar]
  • 13.Gupta GP, Massague J. Cancer metastasis: building a framework. Cell. 2006;127:679–95. [DOI] [PubMed] [Google Scholar]
  • 14.Itzel T, Scholz P, Maass T, Krupp M, Marquardt JU, Strand S, et al. Translating bioinformatics in oncology: guilt-by-profiling analysis and identification of KIF18B and CDCA3 as novel driver genes in carcinogenesis. Bioinformatics. 2015;31:216–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Jalkanen S, Karikoski M, Mercier N, Koskinen K, Henttinen T, Elima K, et al. The oxidase activity of vascular adhesion protein-1 (VAP-1) induces endothelial E- and P-selectins and leukocyte binding. Blood. 2007;110:1864–70. [DOI] [PubMed] [Google Scholar]
  • 16.Jeong YJ, Bong JG, Park SH, Choi JH, Oh HK. Expression of leptin, leptin receptor, adiponectin, and adiponectin receptor in ductal carcinoma in situ and invasive breast cancer. J Breast Cancer. 2011;14:96–103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Karagiannis GS, Pastoriza JM, Wang Y, Harney AS, Entenberg D, Pignatelli J et al. Neoadjuvant chemotherapy induces breast cancer metastasis through a TMEM-mediated mechanism. Sci Transl Med 2017; 9. [DOI] [PMC free article] [PubMed]
  • 18.Keklikoglou I, Cianciaruso C, Guc E, Squadrito ML, Spring LM, Tazzyman S, et al. Chemotherapy elicits pro-metastatic extracellular vesicles in breast cancer models. Nat Cell Biol. 2019;21:190–202. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Kim K, Yang J. Decision-making process related to treatment and management in Korean women with breast cancer: finding the right individualized healthcare trajectory. Appl Nurs Res. 2017;35:99–105. [DOI] [PubMed] [Google Scholar]
  • 20.Koch M, Wiese M. Accessing cancer metabolic pathways by the use of microarray technology. Curr Pharm Des. 2013;19:790–805. [PubMed] [Google Scholar]
  • 21.Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9:559. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Letellier S, Garnier JP, Spy J, Bousquet B. Determination of the L-DOPA/L-tyrosine ratio in human plasma by high-performance liquid chromatography. Usefulness as a marker in metastatic malignant melanoma. J Chromatogr B Biomed Sci Appl. 1997;696:9–17. [DOI] [PubMed] [Google Scholar]
  • 23.Zhao Y, He M, Cui L, Gao M, Zhang M, Yue F, et al. Chemotherapy exacerbates ovarian cancer cell migration and cancer stem cell-like characteristics through GLI1. Br J Cancer. 2020;122:1638–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Zhao Y, Yang X, Zhao J, Gao M, Zhang M, Shi T, et al. Berberine inhibits chemotherapy-exacerbated ovarian cancer stem cell-like characteristics and metastasis through GLI1. Eur J Pharmacol. 2021;895:173887. [DOI] [PubMed] [Google Scholar]
  • 25.Zheng X, Zhao Y, Jia Y, Shao D, Zhang F, Sun M et al. Biomimetic co-assembled nanodrug of doxorubicin and berberine suppresses chemotherapy-exacerbated breast cancer metastasis. Biomaterials 2021; 271. [DOI] [PubMed]
  • 26.Stickeler E, Pils D, Klar M, Orlowsk-Volk M, Zur Hausen A, Jäger M, et al. Basal-like molecular subtype and HER4 up-regulation and response to neoadjuvant chemotherapy in breast cancer. Oncol Rep. 2011;26:1037–45. [DOI] [PubMed] [Google Scholar]
  • 27.El-Badawy A, Ghoneim MA, Gabr MM, Salah RA, Mohamed IK, Amer M, et al. Cancer cell-soluble factors reprogram mesenchymal stromal cells to slow cycling, chemoresistant cells with a more stem-like state. Stem Cell Res Ther. 2017;8:254. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Libby EF, Frost AR, Demark-Wahnefried W, Hurst DR. Linking adiponectin and autophagy in the regulation of breast cancer metastasis. J Mol Med (Berl). 2014;92:1015–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Xie C, Mondal DK, Ulas M, Neill T, Iozzo RV. Oncosuppressive roles of decorin through regulation of multiple receptors and diverse signaling pathways. Am J Physiology-Cell Physiol. 2022;322:C554–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Takeuchi T, Suzuki M, Kumagai J, Kamijo T, Sakai M, Kitamura T. Extracellular matrix dermatopontin modulates prostate cell growth in vivo. J Endocrinol. 2006;190:351–61. [DOI] [PubMed] [Google Scholar]
  • 31.Prieto D, Oppezzo P. Lipoprotein lipase expression in chronic lymphocytic leukemia: New insights into Leukemic Progression. Molecules 2017; 22. [DOI] [PMC free article] [PubMed]
  • 32.Chen L, Liu D, Yi X, Qi L, Tian X, Sun B, et al. The novel miR-1269b-regulated protein SVEP1 induces hepatocellular carcinoma proliferation and metastasis likely through the PI3K/Akt pathway. Cell Death Dis. 2020;11:320. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Allison KH. Molecular pathology of breast cancer: what a pathologist needs to know. Am J Clin Pathol. 2012;138:770–80. [DOI] [PubMed] [Google Scholar]
  • 34.Zhao Y, Cui L, Pan Y, Shao D, Zheng X, Zhang F et al. Berberine inhibits the chemotherapy-induced repopulation by suppressing the arachidonic acid metabolic pathway and phosphorylation of FAK in ovarian cancer. Cell Prolif 2017; 50. [DOI] [PMC free article] [PubMed]
  • 35.Zhang F, Jia Y, Zheng X, Shao D, Zhao Y, Wang Z, et al. Janus nanocarrier-based co-delivery of doxorubicin and berberine weakens chemotherapy-exacerbated hepatocellular carcinoma recurrence. Acta Biomater. 2019;100:352–64. [DOI] [PubMed] [Google Scholar]
  • 36.Dagogo-Jack I, Shaw AT. Tumour heterogeneity and resistance to cancer therapies. Nat Rev Clin Oncol. 2018;15:81–94. [DOI] [PubMed] [Google Scholar]
  • 37.Martins-Neves SR, Cleton-Jansen AM, Gomes CMF. Therapy-induced enrichment of cancer stem-like cells in solid human tumors: where do we stand? Pharmacol Res. 2018;137:193–204. [DOI] [PubMed] [Google Scholar]
  • 38.Jiang M-J, Gu D-N, Dai J-J, Huang Q, Tian L. Dark side of cytotoxic therapy: Chemoradiation-Induced Cell Death and Tumor Repopulation. Trends Cancer 6: 419–31. [DOI] [PubMed]
  • 39.Wang Y, Zhao M, He S, Luo Y, Zhao Y, Cheng J et al. Necroptosis regulates tumor repopulation after radiotherapy via RIP1/R IP3/MLKL/JNK/IL8 pathway. J Exp Clin Cancer Res 38: 461. [DOI] [PMC free article] [PubMed]
  • 40.Jiang M-J, Chen Y-Y, Dai J-J, Gu D-N, Mei Z, Liu F-R et al. Dying tumor cell-derived exosomal mir-194-5p potentiates survival and repopulation of tumor repopulating cells upon radiotherapy in pancreat ic cancer. Mol Cancer 19: 68. [DOI] [PMC free article] [PubMed]
  • 41.Li F, Huang Q, Chen J, Peng Y, Roop DR, Bedford JS, et al. Apoptotic cells activate the phoenix rising pathway to promote wound healing and tissue regeneration. Sci Signal. 2010;3:ra13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Rietveld IM, Lijfering WM, le Cessie S, Bos MHA, Rosendaal FR, Reitsma PH, et al. High levels of coagulation factors and venous thrombosis risk: strongest association for factor VIII and Von Willebrand factor. J Thromb Haemost. 2019;17:99–109. [DOI] [PubMed] [Google Scholar]
  • 43.Liu Y, Wang X, Li S, Hu H, Zhang D, Hu P, et al. The role of Von Willebrand factor as a biomarker of tumor development in hepatitis B virus-associated human hepatocellular carcinoma: a quantitative proteomic based study. J Proteom. 2014;106:99–112. [DOI] [PubMed] [Google Scholar]
  • 44.Yang AJ, Wang M, Wang Y, Cai W, Li Q, Zhao TT, et al. Cancer cell-derived Von Willebrand factor enhanced metastasis of gastric adenocarcinoma. Oncogenesis. 2018;7:12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Yang X, Sun HJ, Li ZR, Zhang H, Yang WJ, Ni B, et al. Gastric cancer-associated enhancement of Von Willebrand factor is regulated by vascular endothelial growth factor and related to disease severity. BMC Cancer. 2015;15:80. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Oleksowicz L, Dutcher JP, Deleon-Fernandez M, Paietta E, Etkind P. Human breast carcinoma cells synthesize a protein immunorelated to platelet glycoprotein-Ib alpha with different functional properties. J Lab Clin Med. 1997;129:337–46. [DOI] [PubMed] [Google Scholar]
  • 47.Luo SY, Li R, Le ZY, Li QL, Chen ZW. Anfibatide protects against rat cerebral ischemia/reperfusion injury via TLR4/JNK/caspase-3 pathway. Eur J Pharmacol. 2017;807:127–37. [DOI] [PubMed] [Google Scholar]
  • 48.Kurtova AV, Xiao J, Mo Q, Pazhanisamy S, Krasnow R, Lerner SP, et al. Blocking PGE2-induced tumour repopulation abrogates bladder cancer chemoresistance. Nature. 2015;517:209–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Zhang Y, Ma XJ, Guo CY, Wang MM, Kou N, Qu H, et al. Pretreatment with a combination of ligustrazine and berberine improves cardiac function in rats with coronary microembolization. Acta Pharmacol Sin. 2016;37:463–72. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1 (4.7MB, doc)

Data Availability Statement

The datasets used and/or analyzed in the study are included in the article/Supplementary Material. Further inquiries can be directed to corresponding authors.

Articles from BMC Cancer are provided here courtesy of BMC

RESOURCES