Background
Recent advancements in neoadjuvant chemotherapy (NAC) have broadened its applications beyond large aggressive tumors in breast cancer treatment. Candidates for preoperative systemic therapy are inoperable breast cancer, and operable breast cancer in selected patients who desire breast conservation with large primary tumor relative to breast size, who have HER2-positive disease and triple-negative breast cancer greater than clinical T2 or clinical N1. Current NAC strategies are highly personalized, considering the cancer subtype, stage, and molecular characteristics, to optimize treatment efficacy. Preferred regimens of NAC for HER2-negative breast cancer are dose-dense adriamycin and cytoxan followed by paclitaxel or weekly paclitaxel, and those for HER2-positive breast cancer are paclitaxel + trastuzumab, paclitaxel + carboplatin + trastuzumab (TCH), and TCH + pertuzumab. Moreover, these approaches might be integrated with both conventional therapies and immunotherapies to align with each patient’s unique cancer profile [
1,
2]. The response to NAC remains a key prognostic indicator in several studies. Moreover, these developments in NAC protocols have notably increased the feasibility of breast-conserving surgeries, marking a significant shift in the breast cancer management landscape [
3‐
5].
Clinically established therapeutics are highly effective in treating breast cancer; however, they present significant challenges. One major challenge is the limited therapeutic options for triple-negative breast cancer (TNBC), which is often resistant to chemotherapy and cross-resistant to other antitumor agents, suggesting multidrug resistance. In addition, the development of drug resistance to anthracycline + cyclophosphamide + taxane (ACT) combination chemotherapies undermines their therapeutic potential [
6]. Moreover, some tumors relapse rapidly after NAC and surgery. Thus, a better understanding of the molecular mechanisms underlying drug resistance is urgently required to effectively treat TNBC.
Extracellular vesicles (EVs) are mediators of cell-to-cell communication, are surrounded by lipid bilayers, and are released from living cells. EVs carry real-time molecular information about their cell of origin, such as nucleic acids, proteins, and lipids. Among body fluid-derived EVs, tumor-derived EVs comprise vesicles released by highly heterogeneous breast cancer cells [
7]. Therefore, the more accurately tumor-derived EVs can be isolated, the more accurately they can reflect the pathophysiological characteristics and behaviors of tumor cells [
8]. An analysis of tumor-derived EVs collected from tumor tissues, especially at the early stages of drug resistance development, may offer insights into the possibility of screening and monitoring cancers, including breast cancer, and provide appropriate treatment options for patients in terms of precision medicine [
9].
In this study, we developed several stable drug-resistant TNBC cell lines in vitro using a stepwise treatment strategy with chemotherapeutic agents for 28 weeks. We aimed to identify highly expressed biomarkers within EVs released from these cell lines and evaluate the clinical feasibility of EV-based assessments for predicting drug response in breast cancer patients receiving NAC.
Methods
Cell lines and anticancer drugs
The human TNBC cell lines HCC1395 (CRL-2324), MDA-MB-231 (HTB-26), and MDA-MB-468 (HTB-132) were purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA). All cell lines were grown in RPMI-1640 (22400-089, Gibco, Carlsbad, CA, USA) supplemented with 10% heat-inactivated fetal bovine serum (FBS, 12483-020, Gibco), and 1% penicillin-streptomycin (15140-122, Gibco), and maintained in a humidified atmosphere of 5% CO2 in air at 37 °C.
Doxorubicin hydrochloride (Adriamycin, or Anthracycline chemotherapy drug, D4000) and docetaxel (Taxotere, or Taxane chemotherapy drug, D1000) were purchased from LC Laboratories (Woburn, MA, USA), and cyclophosphamide monohydrate (Cytoxan, NSC-26271) was purchased from Selleck Chemicals (Houston, TX, USA).
Induction of chemotherapy resistance in breast cancer cells
Drug-resistant sublines of each TNBC cell line were derived from each original parental cell line by continuous exposure of low to high doses of anticancer drugs (1/120 IC50, 1/90 IC50, 1/60 IC50, 1/30 IC50, 1/10 IC50, and IC50) for over 6 months. Each parental cell line was treated with anthracycline + cyclophosphamide (AC, 1:10 molar ratio) for 72 h, repeated four times. Subsequent treatments with Taxotere (T) proceeded in the same manner, and this process was defined as one cycle. The cells were maintained for 6 months while the drug concentration was gradually increased and allowed to recover for an additional month.
Evaluation of drug-resistant activity and growth rate
Sensitivity to chemotherapeutic drugs was measured with a colorimetric assay using MTT (M2003, Sigma-Aldrich, St. Louis, MO, USA). Following the treatment of cells with serial dilutions of AC or T for 72 h, MTT was added to each well and incubated for 4 h at 37 °C. Methanol was then added to each well and mixed for 30 min on an orbital shaker. The absorbance was recorded at 570 nm with a correction wavelength of 690 nm using a NanoDrop 3000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). IC50 values were calculated using Prism v9.0.0 (GraphPad, San Diego, CA, USA).
To measure the growth rate of the derived cell lines, three groups of drug-resistant sublines and their parental cell line counterparts (1 × 105) were seeded in 6-well plates and allowed to attach to the well surface. After attachment, the cells were counted every 24 h for 96 h using a cell counter (TC20, Bio-Rad, Hercules, CA, USA). The trypan blue exclusion assay was used to determine the number of viable cells in the cell suspension and evaluate the doubling time of the cells.
RNA sequence analysis
The concentration and quality of total RNA were checked using a Qubit 2.0 fluorometer (Thermo Fisher Scientific). Total RNA (10 ng) was used to prepare strand-specific barcoded RNA libraries using the Ion AmpliSeq
™ Transcriptome Human Gene Expression Kit (Thermo Fisher Scientific) following the manufacturer’s protocol. The Ion AmpliSeq Transcriptome Human Gene Expression Kit was designed for simultaneous targeted amplification of over 20,000 human genes using a single primer pool. A short amplicon (approximately 110 bp) was obtained from each target gene. AmpliSeq sequencing data were obtained using the Torrent Mapping Alignment Program optimized for Ion Torrent
™ sequencing data to align raw sequencing reads against a custom reference sequence set containing all transcripts targeted by the AmpliSeq kit [
10,
11].
Transcriptome analysis
Differentially expressed genes (DEGs) were identified using DEGSeq (version 1.48.0) with p-values of < 0.05 and|fold changes|>2.5 threshold [
12]. The function of each DEG was annotated based on the Biological Process Gene Ontology gene set (MSigDB collections, C5, BROAD Institute, Cambridge, MA, USA) and the Gene Reference into Function (GeneRIF, The National Center for Biotechnology Information) database. Volcano plots, bar plots, Venn diagrams, and heat maps were generated using ggplot2 (version 3.3.5) and the Complex Heatmap (version 2.10.0) software. All statistical analyses and visualizations were performed using R (version 4.1.2) and R Studio environment (release 077589bc).
Flow cytometry for EV surface profiling
The surface profiles of cells and EVs were analyzed using flow cytometry (FACS LSR Fortessa system, Becton Dickinson, Franklin Lakes, NJ, USA) and specific antibodies against surface proteins. Cell samples were incubated for 30 min at 4 °C in the dark with one test dose of drug resistance detection antibodies (anti-MDR1, anti-MRP1, and anti-BCRP) and rinsed twice with FACS buffer to prevent excessive reactions; fluorescein isothiocyanate (FITC)-labeled secondary fluorescent antibodies were used to detect fluorescence signals.
Breast cancer-derived EVs bound to 3-µm microbeads (SPHERO
™ Streptavidin Coated Particles, SVP 30 − 5, Spherotech Inc., Lake Forest, IL, USA) that were conjugated with biotinylated breast cancer-targeting antibodies (anti-EpCAM, anti-ITGA2, and anti-ITGAV), during incubation to enable the isolation of breast cancer-derived EVs, following protocols in a previous study [
13]. After isolation from tumor tissue, the drug resistance detection antibody was then immobilized with a fluorescent detection antibody. Detailed schematics are presented in Additional File 1: Figure
S1a and
S1b.
Confocal microscopy
To stain actin filaments, Alexa Fluor 594-conjugated phalloidin (A12381, Thermo Fisher Scientific) was used according to the manufacturer’s protocol, and DAPI (Vectashield, H-1200, Vector Laboratories, Inc., Newark, CA, USA) was used to stain cell nuclei. To confirm the presence of EVs, the captured EVs were detected using a primary PE-Cy7-labeled antibody against the general EV marker, CD63. EVs were also immobilized with a primary drug resistance detection antibody against MDR1 and a secondary FITC-labeled fluorescent antibody. Fluorescence images were obtained using a Zeiss LSM 700 confocal microscope (Carl Zeiss). The target and detection antibodies used in this study are listed in Additional File 1: Table
S1.
Transmission electron microscopy (TEM)
A drop of the EV sample was fixed in 2% glutaraldehyde-2% paraformaldehyde and placed on a Formvar carbon-coated grid for 15 s. Droplets were removed using filter paper, and a drop of 1% uranyl acetate was added for 15 s for negative staining, removed using filter paper, and washed with a drop of distilled water. The dried grids were observed using a transmission electron microscope (JEM-1011, JEOL, Tokyo, Japan) at an acceleration voltage of 80 kV equipped with a MegaView III CCD camera.
Clinical characteristics of the participants
Clinical samples were obtained from subjects who visited Severance Hospital in South Korea in accordance with the guidelines of the independent Ethics Committee at the College of Medicine Yonsei University (IRB No. 4-2020-1292). Informed consent for the use of blood samples for research purposes was obtained from all patients. Pre-operative plasma samples were collected from the same patient before anesthesia. The criteria for subject eligibility in the analysis included (1) a confirmed pathological diagnosis of breast cancer, (2) collection of blood samples post-NAC and during the pre-operative period, and (3) hemolysis assessed before the isolation of EV to evaluate plasma sample quality. Details of the 36 individuals are shown in Additional File 1: Table
S2.
Receiver operating characteristic (ROC) analysis
ROC analysis of drug-resistant EV markers was performed on data from 20 patients with TNBC using MedCalc version 20.014 (MedCalc Software Ltd., Ostend, Belgium). We used univariate ROC analysis for each marker to obtain the area under the ROC curve (AUC) and evaluate the diagnostic power of drug-resistant EV marker combinations. Optimal criterion values were calculated by considering not only sensitivity and specificity, but also disease prevalence and costs of various decisions [
14]. After performing a univariate ROC analysis on each combination of drug-resistant EV markers, we chose the “Best” combination with the highest AUC and the lowest standard error of AUC. Based on these calculations, we developed a combinatorial predictive score composed of MDR1, MRP1, and BCRP (combi-3) that was utilized to predict drug response in patients with breast cancer. Statistical analyses were performed using a one-way analysis of variance or Welch’s t-test.
Discussion
Within the past few decades, the standard method of evaluating the effectiveness of NAC has been to identify any residual tumor in surgical specimens; however, this is not possible before surgery. Approximately 10% of breast tumors do not respond to NAC, and resistant breast tumors, particularly TNBC, lead to disease progression and poor prognosis. Accordingly, to reduce disease progression, it is necessary to predict which patients will not respond to standard treatments [
20]. Circulating EVs in the blood and intracellular communication vesicles with a lipid bilayer ranging in size from 50 to 300 nm are essential for tumorigenesis, development, progression, and metastasis [
21]. Multiple technologies have been developed for tumor-derived EV detection and analysis (e.g., immunoaffinity-based capture) and have greatly advanced our understanding of tumor characteristics through liquid biopsy, despite the absence of tumor tissues. EVs can shuttle bioactive molecules such as proteins and a wide variety of genetic materials from one cell to another, leading to molecular transformations in recipient cells [
22‐
25]. There may be an interplay or synergy between tumor cells in the acquisition of drug resistance via EV exchange. Therefore, we suggest a method that focuses on the role of EVs in multidrug resistance for early therapeutic response prediction and therapy monitoring.
Our study revealed the dynamics of the epigenetic changes that lead to drug resistance after chemotherapy in a cell line model. Various molecular mechanisms are involved in chemoresistance; among these, we found significant changes in cancer stemness in drug-resistant TNBC models. For example, dysregulated TGF-β and Wnt signaling pathways may affect the overall progression to malignancy (Additional File 1: Figure
S6). They are also known to play a positive role in promoting drug-resistant properties in the cancer stem cell (CSC) population [
26,
27]. Furthermore, the cytotoxicity-induced morphological changes can be related to drug resistance. Cytoskeletal reconstruction induces biological changes in cancer cells with drug resistance [
28]. After long-term treatment with chemotherapeutic drugs, actin stress fibers change, with a distinct feature showing the migratory dynamics of cell spreading [
29]. All these phenotypic features can be predominantly attributed to their tumorigenic potential and multidrug resistance. Moreover, we focused mainly on the overexpression of ABC transporters, which results in drug resistance, a characteristic feature of CSC. Accumulating evidence from numerous studies, including ours, indicates that high expression of transmembrane proteins of this superfamily, such as MDR1, MRP1, and BCRP, is found in breast cancer, particularly in breast CSCs [
30].
Despite the clear relevance of ABC transporters, which play a critical role in the development of multidrug resistance, clinical approaches for assessing these proteins in the development of drug resistance have not been successful and are yet to be elucidated [
31,
32]. However, with the availability of the latest technology for EV analysis, we recommend re-evaluating the role of ABC transporters within EVs, not focusing on those in tumor tissues. Two notable singularities of our study are worth discussing. First, we found that a certain percentage of cells were stably resistant to NAC when a chemotherapeutic drug was administered continuously (Fig.
4a). A superficial explanation for this is that drug resistance may be induced by more than simple genetic alterations; dysregulation of major epigenetic factors in breast cancer may play a more important role. This hypothesis is consistent with several reports on the epigenetic control of CSCs and their influence on tumorigenesis, development, and responsiveness to therapy [
33,
34]. Second, by using the enrichment of tumor-derived EVs, a higher expression of drug efflux transporters was observed in drug-resistant EVs than in wild-type EVs (Fig.
4e) [
35‐
37]. This has been suggested earlier and may have clinical significance, as it indicates that resistant clones may induce bursts of ABC transport in EVs to enhance drug resistance.
Although a few studies have utilized miRNAs, EV concentrations, and cancer antigens to monitor and predict response to treatments [
38‐
40], our study is a novel attempt to use tumor-derived EVs to assess ABC transporters isolated from the plasma of patients with breast cancer who are treated with NAC. In a retrospective study using plasma from 36 patients, considerable differences in the expression of MDR1, MRP1, and BCRP were detected between patients with pCR and those with residual tumors. The combination of these three parameters offered acceptable sensitivity, specificity, and accuracy in predicting the effectiveness of NAC through cumulative ROC analysis. Apart from the lack of understanding regarding the role of EVs in cancer stem cells and drug resistance, we suggest that this method is a better strategy for repeatedly screening molecular information to predict therapeutic responses. However, three minor limitations of this study merit further discussion. First, we performed NGS analysis for established drug-resistant cell models compared to the wild-type to identify drug resistance genes and validated their expression in EVs isolated from breast cancer patients undergoing NAC. This was due to our concerns regarding the challenges of accurately separating tumor-derived EVs and ensuring sufficient nucleic acid material for reliable NGS analysis. Further research is required to precisely separate tumor-derived EVs and analyze their contents. Second, there was a shortage of sufficient sample sizes to conduct significant cohort studies with high statistical validity. Another issue is the lack of longitudinal studies analyzing changes in EV markers over time in patients. Future studies will advance to larger patient groups and address how EVs crosstalk with other tumor cells of different phenotypes to obtain drug-resistant characteristics, in addition to MDR1, MRP1, and BCRP.
Conclusions
We established drug-resistant TNBC cell lines and investigated the genes involved in drug transport, such as MDR1, MRP1, and BCRP, which were highly expressed in resistant cell lines compared to that in their wild-type counterparts. The expression of MDR1, MRP1, and BCRP in breast cancer-derived EVs increased after in vitro chemotherapeutic treatment, particularly in drug-resistant cell lines. To investigate the clinical significance of drug-resistant EV markers in patients with TNBC receiving NAC, patients with residual tumors were found to have higher expression levels of MDR1, MRP1, and BCRP in EVs than in patients with a pathological complete response. Integrated analysis of MDR1, MRP1, and BCRP expression showed significant differences according to tumor response, not only in TNBC but also in other subtypes. In conclusion, our findings demonstrate that this EV marker combination could be a useful predictor for discriminating breast cancer patients with residual disease from those with no residual disease after NAC, especially in TNBC.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit
http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (
http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.