Skip to main content
NCBI home page
Biomicrofluidics logoLink to Biomicrofluidics
. 2024 Apr 5;18(2):024106. doi: 10.1063/5.0188861

Epithelial and mesenchymal phenotypes determine the dynamics of circulating breast tumor cells in microfluidic capillaries under chemotherapy-induced stress

Rong Du 1,2,1,2, Xiaoning Han 1,3,1,3, Linhong Deng 1,3,1,3,a), Xiang Wang 1,3,1,3,a)
PMCID: PMC10998713  PMID: 38585003

Abstract

Circulating tumor cells (CTCs) with different epithelial and mesenchymal phenotypes play distinct roles in the metastatic cascade. However, the influence of their phenotypic traits and chemotherapy on their transit and retention within capillaries remains unclear. To explore this, we developed a microfluidic device comprising 216 microchannels of different widths from 5 to 16 μm to mimic capillaries. This platform allowed us to study the behaviors of human breast cancer epithelial MCF-7 and mesenchymal MDA-MB-231 cells through microchannels under chemotherapy-induced stress. Our results revealed that when the cell diameter to microchannel width ratio exceeded 1.2, MCF-7 cells exhibited higher transit percentages than MDA-MB-231 cells under a flow rate of 0.13 mm/s. Tamoxifen (250 nM) reduced the transit percentage of MCF-7 cells, whereas 100 nM paclitaxel decreased transit percentages for both cell types. These differential responses were partially due to altered cell stiffness following drug treatments. When cells were entrapped at microchannel entrances, tamoxifen, paclitaxel, and high-flow stress (0.5 mm/s) induced a reduction in mitochondrial membrane potential (MMP) in MCF-7 cells. Tamoxifen treatment also elevated reactive oxygen species (ROS) levels in MCF-7 cells. Conversely, MMP and ROS levels in entrapped MDA-MB-231 cells remained unaffected. Consequently, the viability and proliferation of entrapped MCF-7 cells declined under these chemical and physical stress conditions. Our findings emphasize that phenotypically distinct CTCs may undergo selective filtration and exhibit varied responses to chemotherapy in capillaries, thereby impacting cancer metastasis outcomes. This highlights the importance of considering both cell phenotype and drug response to improve treatment strategies.

INTRODUCTION

Aggressive tumors can release a multitude of malignant cells into the bloodstream.1 However, only a small fraction (0.1%) of these circulating tumor cells (CTCs) survive after dislodging from tumors and subsequently form the “seeds” responsible for colonizing distant sites.2 This survival challenge arises from the immune system's elimination mechanisms and the fluid shear stress (FSS) exerted by blood flow.3 Even if the CTCs overcome these obstacles, their transport through small capillaries remains crucial for metastasis. Their relatively larger cell size compared to the luminal diameter of blood capillaries (3–12 μm) can lead to the mechanical stress-induced destruction of CTCs. It also results in physical occlusion in the microcirculation, thereby impeding their translocation to distant sites.4,5 Moreover, once CTCs travel to distant organs through the circulatory system, they may again encounter occlusion within capillaries and interact with endothelial cells lining the blood vessels, subsequently prone to extravasation and initiating successful metastatic growth.6 Therefore, understanding the transit and retention behavior of CTCs within the capillary region provides critical insights into delineating the CTCs' metastatic cascade and their response to therapeutic intervention.

Cancer metastasis, including circulation within blood vessels, represents a dynamic interplay between passive migration induced by external pressure and active migration driven by cytoskeletal activities.7,8 This process may also involve factors such as ion channel dynamics and the localization of mitochondria.9,10 Both active and passive migration processes of CTCs through capillaries are intricately influenced by the diverse mechanical properties of the cells, including size, stiffness, deformability, and surface friction.11,12 These properties are not only cell-type-specific but also significantly dictated by the cell's particular phenotype.4,13,14 For instance, epithelial and mesenchymal breast tumor cells exhibit different mechanical properties influencing their ability to travel through blood vessels.15 Studies have demonstrated that CTCs derived from breast cancer patients comprise a heterogeneous population of cells with distinct epithelial and/or mesenchymal phenotypic characteristics, indicating the occurrence of an epithelial-to-mesenchymal transition (EMT).16–18 During EMT, cancer cells lose their epithelial characteristics, such as cell–cell adhesion, and gain mesenchymal traits, allowing them to move away from the original tumor and invade other tissues. This process is pivotal for the transformation of early-stage, localized tumors into aggressive, metastatic cancers.19 However, the specific impact of CTCs' phenotype or EMT status on their transit through capillaries with varying diameters remains unclear.

In addition to phenotypic distinctions, the transit behavior and survival of CTCs through capillaries are inevitably affected by anticancer drugs like tamoxifen and paclitaxel, integral components of breast tumor chemotherapy used either as endocrine therapy targeting estrogen receptor-positive tumor cells or as broad cytotoxic agents.20 Yet, it remains unknown whether and how these drugs alter the transit dynamics of CTCs through capillaries, especially considering the varied chemoresistance levels exhibited by CTCs in different EMT statuses.21,22 Therefore, there is considerable interest and clinical significance in investigating the transit behavior of both epithelial and mesenchymal CTCs in capillaries of various diameters, both with and without exposure to different chemotherapeutic drugs.23

However, investigating the in vivo dynamics of CTCs within capillaries remains challenging, despite reports of observed CTC transit through small blood vessels in zebrafish embryos and mouse brains.24 Microfluidic technology, in contrast, emerges as a valuable tool for fabricating microchannels narrower than cell diameters. This provides a controlled experimental platform to mimic capillaries with precise geometric and mechanical parameters resembling the microcirculation environment.25,26 Previous research has employed microfluidic models, including single-channel systems simulating blood vessels, to study the impact of FSS on transit behavior and survival of CTCs.27–29 Nevertheless, these models often exhibit limitations in examining CTCs of various sizes traversing microchannels with various widths. Furthermore, they frequently neglect to consider the physiological conditions of CTCs entrapped within capillaries under chemotherapy-induced stress. Understanding these conditions is essential for comprehending how CTCs survive, adapt, and potentially evade treatment, which, in turn, affects the efficacy of cancer therapies and the progression of metastasis.

Therefore, we introduced a novel microfluidic device comprising an array of microwells connected to microchannels on a single chip. The unique design of the microwells facilitated the entrapment of individual cells in front of the microchannels. The microchannels, designed to sequentially alter their widths from 5 to 16 μm, replicated the in vivo capillaries with varying sizes. Utilizing this microfluidic chip, we conducted experiments using epithelial (MCF-7) and mesenchymal (MDA-MB-231) breast cancer cells as a model system, assessing their transit through the microchannels in the presence and absence of tamoxifen and paclitaxel. Furthermore, we evaluated the survival and proliferation of these cells when entrapped in front of the microchannels. Our results revealed that the CTC behaviors including transit, physiological status, viability, and proliferation were dependent on their phenotypes and chemotherapy-induced stress. These findings may contribute to deepening our understanding of the pathophysiological events surrounding CTCs as they travel within capillaries, potentially providing insights into how CTCs might be differentially considered in terms of capacity and treatment of their metastasis.

MATERIALS and METHODS

Design and fabrication of the microfluidic chip with varying channel widths

The mask layout of the microfluidic device was designed using the LayoutEditor software (Juspertor, Germany). The microfluidic chip consists of an inlet, a pre-filter for removal of debris and cell clusters, 8 main channels housing a total of 216 test units, and an outlet. A silicon master with patterns at a depth of 50 μm was fabricated using deep reactive ion etching (GeSiM, Germany). Poly dimethyl siloxane (PDMS) microstructures were prepared by casting a 10:1 (v/v) mixture of Sylgard 184 elastomer and curing agent (Dow Corning, USA) against the master. After thermally curing at 65 °C for 4 h, the PDMS device was treated with oxygen plasma (PT-5S, Sanhe Boda Electromechanical Technology, China) and placed onto a glass slide to seal the channels.

Preparation and morphological observation of cells in suspension

Human breast cancer MCF-7 and MDA-MB-231 cells were purchased from the American Type Culture Collection. The cells were maintained in Dulbecco's Modified Eagle Medium (Gibco, Thermo Fisher Scientific) supplemented with 10% fetal bovine serum (Gibco) at 37 °C and 5% (v/v) CO2. Before experiments, the cells were detached by trypsinization from the culture dish and then aspirated and resuspended in culture media to a concentration of 105 cells per ml. For morphological observation, the suspended cells were stained with Hoechst 33258 (HO, Sigma-Aldrich) and sedimented on a μ-Slide 8-well chamber (Ibidi, Germany). The stained cells were then imaged and measured for cell and nucleus diameter using ImageJ.

Assessment of cell transit in microchannels of the microfluidic chip

A custom-made reservoir was inserted into the inlet, and a syringe controlled by a PID syringe pump (Cetoni, Germany) was connected to the outlet of the microfluidic chip. To remove air and prevent cell adhesion in the microchannels, the microchannels were flushed with 75% ethanol and primed with 0.2% F-127 (Sigma-Aldrich) for 3 h. Afterward, the F-127 solution was replaced with phosphate-buffered saline (PBS). Then, the chip was mounted onto a motorized stage (Prior Scientific, USA) of an Axio Observer 7 fluorescence microscope (Carl Zeiss MicroImaging, Germany). The microscope was equipped with a 20 × /0.4NA objective and an ORCA-Flash 4.0 camera (Hamamatsu, Japan). The experimental setup was maintained at a controlled temperature of 37 °C with 5% CO2. The flow velocity in the main channels was assessed by particle image velocimetry (PIV) analysis. Polystyrene beads (10 μm in diameter, Sigma-Aldrich) suspended in PBS were infused into the inlet. The movement of the beads in the main channels was captured using exposure times of 200 and 400 ms for loading flow rates of 1.5 and 0.5 μl/min, respectively. The velocity was determined by measuring the displacement of the beads over time.

Once the system was prepared, cell samples (105 cells/ml) were introduced into the reservoir. Cells were driven into the microchannels by gravity-induced flow and subsequently captured by microwells. To generate fluid shear stress, the syringe connected to the outlet was controlled by software, allowing accurate withdrawal of fluid to achieve a predefined volumetric flow rate (0.5 or 1.5 μl/min) within the microfluidic channel. For monitoring purposes, bright-field multi-position live-cell imaging was conducted using a 20 min/frame acquisition rate. Each microwell containing a cell was recorded using the ZEN software (Carl Zeiss MicroImaging) once fluid shear stress was initiated. For drug treatments, either 250 nM tamoxifen or 100 nM paclitaxel (all from Sigma-Aldrich) was used. The transit percentage was determined by quantifying the ratio of cells successfully traversing the microchannels to the total number of cells initially confined within microwells. Residence time was defined as the duration elapsed from the entrapment of a cell within a well to its subsequent passage through the microchannel.

Measurement of PDMS and cell stiffness by atomic force microscopy

The PDMS stiffness was determined using a NanoWizard 3 atomic force microscopy (AFM) system (JPK Instruments, Germany) in water at 22 °C. A cantilever (k = 0.03 N/m, MLCT-O10, Bruker, Germany) affixed with a 20 μm spherical polystyrene bead (Sigma-Aldrich) was used for the indentation experiment. Force–distance curves were collected across grids of 5 × 5 measurement points. For cell stiffness measurement, MCF-7 or MDA-MB-231 cells cultured on 3-cm Petri dishes were treated with 250 nM tamoxifen or 100 nM paclitaxel for 12 h respectively. Following the treatment, cells were trypsinized, suspended in CO2-independent medium, and carefully transferred onto an F-127-coated Petri dish. The AFM measurements were performed using the NanoWizard 3 setup mounted on a Zeiss inverted microscope with a controlled temperature of 37 °C. An AFM cantilever (k = 0.03 N/m) glued with a 10 μm spherical polystyrene bead was used. To initiate the indentation, the center of a cell was indented using a trigger force of 1.3 nN at a constant rate of 1.0 μm/s. The cantilever was carefully brought into contact with the cell surface and then retracted. The resulting force–distance curves were analyzed using a Hertzian model fit appropriate for a spherical indenter. The Young's modulus (E) for both PDMS and cells was determined using the JPK data processing software.

Determination of mitochondrial membrane potential and intracellular ROS levels

The mitochondrial membrane potential (MMP) and intracellular ROS level in cells entrapped in front of the microchannels were assessed using JC-1 and CM-H2DCFDA, respectively. A decline in the red/green fluorescence intensity ratio of JC-1 indicates mitochondrial depolarization. CM-H2DCFDA reacts with intracellular glutathione and other thiols, resulting in green fluorescence upon generation of ROS. Briefly, following 12 h of drug treatment or exposure to a high flow rate, the medium in the inlet was removed. Subsequently, 100 μl of staining solutions containing 0.2 μg/ml JC-1 or 10 μM CM-H2DCFDA (both from Thermo Fisher Scientific) were infused into the inlet and allowed to incubate for 30 min. After rinsing with DMEM, fluorescence images were captured using the Axio Observer 7 fluorescence microscope equipped with the 20 × /0.4NA objective. The green and red fluorescence intensity of JC-1 and the green fluorescence intensity of H2DCFDA were quantified using ImageJ.

Assessment of cell proliferation and viability

To determine the level of cell proliferation, the number of single cells that divided into two daughter cells in microwells during 12 h of imaging was counted. The proliferation rate was calculated as the ratio of the number of dividing cells to the number of total cells. To determine cell viability in the microfluidic device, after treatment, the medium with a staining buffer containing 2 μM Calcein-AM and 1 μM propidium iodide (PI) (Solarbio Science & Technology, China) was pumped into the chip from the reservoir. After incubation for 15 min, the cells were imaged by a microscope equipped with a 20 × /0.4NA objective in the green (488 nm) and red (570 nm) channels. The death percentage was calculated by dividing the number of red-labeled cells by the number of total imaged cells.

Statistical analysis

Each experiment was carried out with at least five groups of independent samples. Values are expressed as mean ± SD. The transit percentage, proliferation rate, and death percentage were analyzed using the t-test (SPSS Statistics 21, IBM, USA). The transit time, cell and nucleus diameter, Young's modulus, mitochondrial membrane potential, and ROS level were analyzed using the Mann–Whitney U test (SPSS). p < 0.05 indicates statistical significance. The box plots represent the median value (central bar) and the data distribution (bounding box). The violin plots show the median values (middle lines) and quartiles (side lines).

RESULTS

Photographic verification of the microfluidic device and movement of breast cancer cells in microchannels

The PDMS microfluidic device comprises eight parallel main channels, each connecting a series of 56 test units as shown in [Fig. 1(a)]. The magnified view of the image shows a test unit consisting of a 50 μm long microchannel to mimic a blood capillary, a microwell for single-cell capture at a microchannel entrance, and two bypass channels facilitating the uninterrupted flow of remaining cells downstream [Fig. 1(b)]. To replicate the typical diameters of blood capillaries, the widths of the microchannels varied as 5, 8, 12, or 16 μm. In addition, the stiffness of the PDMS was measured at 8.8 ± 0.8 kPa, closely matching the reported stiffness of blood vessels, which is approximately 10 kPa.30 Precise flow control was achieved by applying negative pressure at the outlet using a computer-controlled pump. When a freshly prepared suspension of live MCF-7 or MDA-MB-231 cells was injected into the inlet reservoir, cells larger than the microchannel width were trapped within the microwells [Fig. 1(c)]. Simultaneously, other cells flowed through the bypass channels, ensuring a random distribution of cells with varied sizes in the array of microwells. As most of the microwells became occupied, the loading flow rate was raised to 0.5 μl/min. It generates a local flow velocity of 0.13 ± 0.03 mm/s in the main channel, as determined by PIV analysis. This velocity falls within the reported range of blood flow rates in capillary beds (0.01 to 1.5 mm/s).31 The persistent flow stress induced the passage of MCF-7 or MDA-MB-231 cells through the microchannels [Fig. 1(c)]. It began with an initial cell deformation determined by cell stiffness, which allowed cells to enter and successfully pass through microchannels.

FIG. 1.

FIG. 1.

Open in a new tab

Verification of the microfluidic device and movement of breast cancer cells through microchannels. (a) Bright-field image of the microfluidic device, showing essential components (inlet, outlet, and main channels) and their configuration. (b) Magnified view of a test unit consisting of a microwell, a microchannel, and two bypass channels. Scale bar = 50 μm. (c) Movement of an MCF-7 or MDA-MB-231 cell from initial capture in a microwell to complete transit through a microchannel under a loading flow rate of 0.5 μl/min (blue arrow). Scale bars = 20 μm.

Transit behaviors of breast cancer cells in microchannels

To investigate the transit behaviors of the two types of breast cancer cells through microchannels of different widths, we conducted time-lapse imaging to monitor individual cell movements in the microfluidic chip for over 12 h. At a low loading flow rate of 0.5 μl/min, MDA-MB-231 cells exhibited a significantly higher transit percentage through the 16 μm microchannel compared to MCF-7 cells [Fig. 2(a)], suggesting that the transit percentage of cells depended on both the width of the microchannel and the phenotype of the tumor cells. Given the substantial heterogeneity in cell and capillary diameters in vivo, our analysis focused on the ratio of cell diameter to microchannel width (D/W). D/W values less than 1.0 were excluded from the study as cells of this size could pass through almost without resistance due to their smaller diameter than that of the microchannel. Our analysis indicates that when D/W was between 1 and 1.2, i.e., the cells were slightly larger than the microchannels, transit percentages were relatively independent of the EMT phenotype of cells [Fig. 2(b)]. As D/W increased to 1.2–1.4 and beyond, transit percentages of both cell types significantly declined. This pattern suggests that cells with D/W < 1.2 constituted the primary population of circulating tumor cells capable of transiting through capillaries. However, when D/W exceeded 1.2, the transit percentage of MCF-7 cells was significantly higher than that of MDA-MB-231 cells. It implies that cells of different EMT phenotypes exhibited disparate passage rates in capillaries under low-flow conditions.

FIG. 2.

FIG. 2.

Open in a new tab

Transit percentages and residence time of breast cancer cells through microchannels. (a) Transit percentages of MCF-7 and MDA-MB-231 cells through microchannels with varied widths (5–16 μm) at a flow rate of 0.5 μl/min. (b) Transit percentages of the two types of cells with different cell diameters to microchannel width (D/W) ratios at a flow rate of 0.5 μl/min. (c) Elevated FSS (1.5 μl/min) induced increased transit percentages for both cell types. (d) Residence time for MCF-7 and MDA-MB-231 cells passing through microchannels at distinct loading flow rates (0.5, 1.5 μl/min, respectively).

To evaluate the impact of flow shear stress on the transit behavior of tumor cells, we replicated the aforementioned experiment using a higher loading flow rate (1.5 μl/min), resulting in a flow velocity of 0.50 ± 0.05 mm/s in the main channel. Consequently, the transit percentages of both MCF-7 cells and MDA-MB-231 cells increased across all D/W ratios [Fig. 2(c)]. Particularly, mesenchymal MDA-MB-231 cells might possess enhanced capabilities in traversing narrow capillaries under the higher FSS. Additionally, we quantified the residence time required for cells to pass through the microchannels, serving as a metric reflective of cell deformability. There were no substantial differences in residence time between MCF-7 and MDA-MB-231 cells [Fig. 2(d)]. With the application of a higher loading flow rate, the residence time significantly decreased for both tumor cell types. This observation suggests that MCF-7 and MDA-MB-231 cells likely exhibit comparable levels of deformability.

Effects of chemotherapeutic agents on the transit of breast cancer cells in microchannels

Tamoxifen and paclitaxel are commonly used chemotherapeutic agents for breast cancer treatments.20 Clinical observations have revealed that the serum concentration of tamoxifen in breast cancer patients ranges from 8.7 to 134.4 ng/ml with a daily dose of 20 mg,32 and that of paclitaxel ranges from 80 to 280 nM, depending on the exposure time.33 To investigate the impact of these drugs on the transit of breast cancer cells in capillaries, we subjected the cells in the chip to either 250 nM (92 ng/ml) tamoxifen or 100 nM (85 μg/ml) paclitaxel for 12 h. Subsequently, we evaluated their transit percentage and residence time. For MCF-7 cells, paclitaxel treatment resulted in a reduced transit percentage of cells across all size ranges, whereas tamoxifen exhibited an inhibitory effect only on cells with a D/W ratio of 1–1.2 [Fig. 3(a)]. Conversely, in the case of MDA-MB-231 cells, only paclitaxel treatment showed a reduction in the transit percentage of cells with a D/W ratio of 1–1.2 [Fig. 3(b)]. This suggests that tamoxifen specifically inhibits the transit of epithelial MCF-7 cells, while paclitaxel exerts a broader effect on both epithelial and mesenchymal cells traversing capillaries. Moreover, we found that tamoxifen treatment led to increased residence time of MCF-7 cells in microchannels [Fig. 3(c)], suggesting that tamoxifen probably decreased the deformability of MCF-7 cells.

FIG. 3.

FIG. 3.

Open in a new tab

Effects of chemotherapeutic agents on the transit of breast cancer cells in microchannels. (a) and (b) Influence of tamoxifen (250 nM) or paclitaxel (100 nM) treatment on the transit percentages of MCF-7 and MDA-MB-231 cells through microchannels. (c) Effects of tamoxifen and paclitaxel on the residence time of MCF-7 and MDA-MB-231 cells in microchannels. Measurements were acquired under a flow rate of 0.5 μl/min. (d)–(f) Cell size, nucleus diameter, and stiffness of MCF-7 and MDA-MB-231 suspension cells treated with 250 nM tamoxifen or 100 nM paclitaxel treatment for 12 h.

To further elucidate the underlying mechanism by which drugs alter cell transit percentages, we examined variations of cell size, nucleus diameter, and Young's modulus of the breast cancer cells in response to the drug treatments. The results indicated that neither tamoxifen nor paclitaxel significantly affected cell size or nucleus diameter, regardless of the tumor cells' phenotype [Figs. 3(d) and 3(e)]. This suggests that changes in cell size do not contribute to the drug's effect on the cell transit through microchannels. Moreover, Young's modulus measurement showed that, in the absence of drug treatment, MCF-7 cells were considerably softer than MDA-MB-231 cells with Young's modulus values of 114 ± 52 and 228 ± 94 Pa, respectively [Fig. 3(f)]. This stiffness difference could account for the higher transit percentage of MCF-7 cells compared to MDA-MB-231 cells with a D/W ratio larger than 1.2, as depicted in Fig. 2(b). Interestingly, both tamoxifen and paclitaxel treatments increased the stiffness of MCF-7 cells, with values changing to 234 ± 120 and 252 ± 92 Pa, respectively [Fig. 3(f)]. This increase in stiffness may explain the observed decrease in the transit percentage of MCF-7 cells, as illustrated in Fig. 3(a). However, none of these drug treatments had any effect on the stiffness of the MDA-MB-231 cells, despite paclitaxel being able to reduce the transit percentage of these cells with a D/W ratio of 1–1.2 as shown in Fig. 3(b). This suggests that the mechanisms underlying the observed effects of tamoxifen and paclitaxel on cell transit percentages are complex and may involve factors beyond alterations in cellular stiffness.

Physiological and survival status of breast cancer cells entrapped at microchannel entrances

Since the entrapment of CTCs at the front end of capillaries is a common phenomenon and is closely associated with CTC extravasation,6 it is crucial to assess the physiological state of cells entrapped at the microchannel entrances under various stress conditions. Mitochondrial membrane potential (MMP) serves as a sensitive indicator of mitochondrial activity and is intricately linked to the survival and differentiation of CTCs.34 Using the fluorescence probe JC-1 to measure MMP [Fig. 4(a)], we observed a significant decrease in MMP levels (JC-1 red/green ratio) in MCF-7 cells, but not MDA-MB-231 cells, in response to the anticancer drugs and higher FSS treatments [Fig. 4(b)]. This indicates that MCF-7 cells are more susceptible to both chemical and physical stresses. Furthermore, we used the H2DCFDA fluorescence probe to measure ROS levels in entrapped MCF-7 or MDA-MB-231 cells under different treatment conditions [Fig. 4(c)]. Aberrant accumulation of ROS in a cell signifies mitochondrial damage due to intracellular redox imbalance, which can impair mitochondrial function.35 Quantitative analysis revealed a significant increase in ROS intensity in entrapped MCF-7 cells following tamoxifen treatment, but not after paclitaxel or high FSS treatment [Fig. 4(d)]. Notably, MDA-MB-231 cells consistently maintained relatively low ROS levels regardless of the treatment they underwent. These results suggest that during occlusion in capillaries, mesenchymal MDA-MB-231 cells exhibit higher resistance to chemical and physical stresses compared to epithelial MCF-7 cells.

FIG. 4.

FIG. 4.

Open in a new tab

Effects of tamoxifen, paclitaxel, and higher FSS on the intracellular MMP and ROS levels in the breast tumor cells entrapped at microchannel entrances. (a) and (b) Representative images and the red/green fluorescence ratio of JC-1 in entrapped MCF-7 and MDA-MB-231 cells without (control) or with the treatment of tamoxifen (250 nM), paclitaxel (100 nM), or higher FSS (1.5 μl/min). (c) and (d) Representative images and relative fluorescence intensity of H2DCFDA in entrapped MCF-7 and MDA-MB-231 cells with the same treatments as in panels (a) and (b). Scale bars = 20 μm.

Since abnormal MMP and ROS levels can lead to cell death, we investigated the viability of cells at microchannel entrances using Calcein/PI staining [Fig. 5(a)]. Statistical analysis reveals similar viability of both cell types under slow flow rates. However, MCF-7 cells, unlike MDA-MB-231 cells, exhibited a significant increase in death percentage in response to paclitaxel, higher FSS, and tamoxifen in rank order [Fig. 5(b)]. During the 12-hour imaging period, we observed a few cells undergoing division into daughter cells [Fig. 5(c)]. Counting the dividing cells in the microwells under different treatment conditions, we found that the division rate of MCF-7 cells was significantly reduced by both tamoxifen and paclitaxel, whereas the division of MDA-MB-231 cells was only inhibited by paclitaxel [Fig. 5(d)]. These results suggest that although under low FSS and drug-free conditions, entrapped breast cancer cells could survive and proliferate at similar rates regardless of their EMT phenotype. However, tamoxifen/paclitaxel (chemical stress) and high FSS (physical stress) were detrimental to the survival and/or proliferation of breast cancer cells to varying extents. Therefore, it suggests that in capillaries, epithelial breast cancer cells are more susceptible to chemotherapy than their mesenchymal counterparts.

FIG. 5.

FIG. 5.

Open in a new tab

Effects of tamoxifen and paclitaxel and higher FSS on the viability and proliferation of breast tumor cells entrapped at microchannel entrances. (a) Representative images of fluorescence of Calcein/PI in entrapped MCF-7 and MDA-MB-231 cells. (b) Death percentages of occluded MCF-7 and MDA-MB-231 cells treated with tamoxifen (250 nM), paclitaxel (100 nM), or higher FSS (1.5 μl/min). (c) Representative time-lapse images showing the division of an MCF-7 or MDA-MB-231 cell into two daughter cells (arrowheads) at microchannel entrances. (d) Division rate of entrapped MCF-7 and MDA-MB-231 cells treated with tamoxifen, paclitaxel, or higher FSS. Scale bars = 20 μm.

DISCUSSION

The EMT status of CTCs significantly impacts their ability to invade and extravasate through blood vessels.4,36 However, the movement and entrapment of epithelial and mesenchymal-like CTCs in capillaries, especially under chemotherapy-induced stress, remain largely unknown. In this study, we developed a microfluidic device comprising microchannels with varying widths to mimic the physical conditions of blood capillaries. We investigated the transit behaviors of epithelial MCF-7 and mesenchymal MDA-MB-231 cells through capillary-like structures while the cells were exposed to two common breast cancer drugs, namely, tamoxifen and paclitaxel. Specifically, we observed that when the cell diameter to channel width ratio (D/W) exceeded 1.2, the transit percentage of MCF-7 cells was higher than that of MDA-MB-231 cells. This observation corroborates a recent report that the EMT status of CTCs was heterogeneous across different vascular compartments of hepatocellular carcinoma,37 emphasizing the crucial role of the EMT status of CTCs in determining their movement within capillaries. Interestingly, we also found that the differences in transit percentage through microchannels between MCF-7 cells and MDA-MB-231 cells at a low flow rate were eliminated at a higher flow rate. This observation implies that a high flow rate facilitates the dissemination of mesenchymal CTCs through capillaries. It is important to note that the interaction between the endothelial lining and CTCs should also be considered, as it significantly influences the behaviors of CTCs in blood vessels.31

During cancer treatment, a crucial question is whether the transit of CTCs is affected by the anticancer drugs, because cancer patients often undergo chemotherapy as part of their long-term treatment strategy. Our results revealed that tamoxifen treatment led to a reduction in the transit percentage of the MCF-7 cell when D/W < 1.2, whereas it had no significant effect on MDA-MB-231 cells. Conversely, paclitaxel could inhibit the transit of both cell types. This discrepancy can be attributed to the fact that tamoxifen specifically targets the estrogen receptors highly expressed in MCF-7 cells, while paclitaxel non-specifically targets rapidly proliferating tumor cells in both the MCF-7 and MDA-MB-231 populations. Hence, targeted chemotherapy approaches may selectively inhibit a specific subpopulation of CTCs within the heterogeneity of CTCs, leaving the remainder unaffected. These unaffected cells may subsequently develop into new tumors with inherent drug resistance. Moreover, mechanical properties of cells, including stiffness and friction, play a crucial role in determining their transit through capillaries.11,14 Our analysis indicated similar cell fractions assessed by residence time for both types of cells. However, AFM measurements revealed that MCF-7 cells were softer than MDA-MB-231 cells, partially resulting in a higher transit percentage of MCF-7 cells. Notably, despite observing no change in cell size, both tamoxifen and paclitaxel treatment increased the stiffness of MCF-7 cells, which may explain the observed decrease in the transit percentage of these cells following drug treatment. Consequently, these findings underscore that CTCs with distinct EMT statuses exhibited diverse transit behaviors through capillaries in response to chemotherapy, which was partly due to the modulation of their stiffness.

CTCs in the bloodstream face potential cell death induced by FSS.38,39 If they manage to survive this FSS-triggered cell death, they often become occluded within narrow capillaries and vessel branch sites, which is a pivotal step for their homing and metastasis potential.24,40,41 Our findings indeed confirm that over 90% of both types of cells with a D/W ratio greater than 1.2 are highly prone to occlusion within microchannels. If entrapped CTCs survive for a certain duration, they may have a chance to invade through the endothelial layer of blood vessels.42 Through measuring MMP and ROS levels in cells entrapped at the microchannel entrances, we observed that the physiological stage of MCF-7 cells was unstable compared to that of MDA-MB-231 cells when exposed to anticancer drugs. Consequently, the viability and proliferation of MCF-7 cells exhibited decreased, particularly in response to tamoxifen. However, the viability of entrapped MDA-MB-231 cells remained unaffected by paclitaxel. This phenomenon could be attributed to their potential adoption of a strategy to transition into a resting state by ceasing proliferation, as observed in this study. These results indicate that in capillaries, epithelial MCF-7 cells exhibit lower stability than mesenchymal MDA-MB-231 cells in response to broad-spectrum anticancer drugs. This is consistent with a study showing that mesenchymal-type CTC sublines are more resistant to chemotherapy than their epithelial-type counterparts in BALB/c mice.21 Therefore, our finding reveals a “survival of the fittest” selection strategy that favors the persistence of mesenchymal CTCs during chemotherapy, which may ultimately contribute to the metastasis cascade and cancer relapse.

CONCLUSIONS

In summary, we developed a microfluidic chip platform designed for the study of the behaviors of circulating tumor cells with diverse EMT phenotypes under chemotherapy-induced stresses through capillaries. Our findings reveal that the transit of CTCs through capillaries depends on not only the size ratio between the cell and the capillary but also the specific EMT phenotypes of the cells. Moreover, flow rate and the presence of anticancer drugs also significantly influence the transit behaviors of CTCs in the microvascular environment, displaying EMT status-dependent patterns. Furthermore, mesenchymal cells exhibit a heightened resistance to both chemical and physical stresses in comparison to their epithelial counterparts. This resistance potentially confers a survival advantage upon mesenchymal CTCs under chemotherapy. Thus, these findings provide valuable insights into the intricate dynamics governing various breast cancer cell types within capillaries and corresponding potential strategies for eliminating CTCs using anticancer drugs.

ACKNOWLEDGMENTS

This work was supported by the National Natural Science Foundation of China (Grant No. 31972922 to X.W., Grant No.12272063 to L.D.). We thank Lei Liu and Yan Pan for their assistance with the AFM experiments.

Contributor Information

Linhong Deng, Email: mailto:dlh@cczu.edu.cn.

Xiang Wang, Email: mailto:wangxiang@cczu.edu.cn.

AUTHOR DECLARATIONS

Conflict of interest

The authors have no conflicts to disclose.

Author Contributions

Rong Du: Data curation (equal); Formal analysis (equal); Investigation (equal); Methodology (equal). Xiaoning Han: Data curation (equal); Formal analysis (equal); Investigation (equal); Methodology (equal); Project administration (supporting). Linhong Deng: Funding acquisition (equal); Resources (equal); Supervision (equal); Writing – review & editing (equal). Xiang Wang: Conceptualization (equal); Formal analysis (equal); Funding acquisition (equal); Investigation (equal); Project administration (equal); Resources (equal); Supervision (equal); Writing – original draft (lead); Writing – review & editing (equal).

DATA AVAILABILITY

The data that support the findings of this study are available within the article.

REFERENCES

  • 1.Poudineh M., Sargent E. H., Pantel K., and Kelley S. O., “Profiling circulating tumour cells and other biomarkers of invasive cancers,” Nat. Biomed. Eng. 2, 72–84 (2018). 10.1038/s41551-018-0190-5 [DOI] [PubMed] [Google Scholar]
  • 2.Thery L., Meddis A., Cabel L., Proudhon C., Latouche A., Pierga J. Y., and Bidard F. C., “Circulating tumor cells in early breast cancer,” JNCI Cancer Spectr. 3, pkz026 (2019). 10.1093/jncics/pkz026 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Massague J. and Obenauf A. C., “Metastatic colonization by circulating tumour cells,” Nature 529, 298–306 (2016). 10.1038/nature17038 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Kurma K. and Alix-Panabieres C., “Mechanobiology and survival strategies of circulating tumor cells: A process towards the invasive and metastatic phenotype,” Front. Cell Dev. Biol. 11, 1188499 (2023). 10.3389/fcell.2023.1188499 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Varotsos Vrynas A., Perea Paizal J., Bakal C., and Au S. H., “Arresting metastasis within the microcirculation,” Clin. Exp. Metastasis 38, 337–342 (2021). 10.1007/s10585-021-10109-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Kienast Y., von Baumgarten L., Fuhrmann M., Klinkert W. E., Goldbrunner R., Herms J., and Winkler F., “Real-time imaging reveals the single steps of brain metastasis formation,” Nat. Med. 16, 116–122 (2010). 10.1038/nm.2072 [DOI] [PubMed] [Google Scholar]
  • 7.Chaffer C. L. and Weinberg R. A., “A perspective on cancer cell metastasis,” Science 331, 1559–1564 (2011). 10.1126/science.1203543 [DOI] [PubMed] [Google Scholar]
  • 8.Zhang X. and Mak M., “Biophysical informatics approach For quantifying phenotypic heterogeneity In cancer cell migration In confined microenvironments,” Bioinformatics. 37, 2042–2052 (2021). 10.1093/bioinformatics/btab053 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Stroka K. M., Jiang H., Chen S. H., Tong Z., Wirtz D., Sun S. X., and Konstantopoulos K., “Water permeation drives tumor cell migration in confined microenvironments,” Cell 157, 611–623 (2014). 10.1016/j.cell.2014.02.052 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Desai S. P., Bhatia S. N., Toner M., and Irimia D., “Mitochondrial localization and the persistent migration of epithelial cancer cells,” Biophys. J. 104, 2077–2088 (2013). 10.1016/j.bpj.2013.03.025 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Byun S., Son S., Amodei D., Cermak N., Shaw J., Kang J. H., Hecht V. C., Winslow M. M., Jacks T., Mallick P., and Manalis S. R., “Characterizing deformability and surface friction of cancer cells,” Proc. Natl. Acad. Sci. U. S. A. 110, 7580–7585 (2013). 10.1073/pnas.1218806110 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Zhang X., Chan T., and Mak M., “Morphodynamic signatures of MDA-MB-231 single cells and cell doublets undergoing invasion in confined microenvironments,” Sci. Rep. 11, 6529 (2021). 10.1038/s41598-021-85640-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Hosseini K., Frenzel A., and Fischer-Friedrich E., “ EMT changes actin cortex rheology in a cell-cycle-dependent manner,” Biophys. J. 120, 3516–3526 (2021). 10.1016/j.bpj.2021.05.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Peralta M., Osmani N., and Goetz J. G., “Circulating tumor cells: Towards mechanical phenotyping of metastasis,” iScience 25, 103969 (2022). 10.1016/j.isci.2022.103969 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Wu Y., Sarkissyan M., and Vadgama J. V., “Epithelial-mesenchymal transition and breast cancer,” J. Clin. Med. 5, 13 (2016). 10.3390/jcm5020013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Polioudaki H., Agelaki S., Chiotaki R., Politaki E., Mavroudis D., Matikas A., Georgoulias V., and Theodoropoulos P. A., “Variable expression levels of keratin and vimentin reveal differential EMT status of circulating tumor cells and correlation with clinical characteristics and outcome of patients with metastatic breast cancer,” BMC Cancer 15, 399 (2015). 10.1186/s12885-015-1386-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Brown H. K., Tellez-Gabriel M., Cartron P. F., Vallette F. M., Heymann M. F., and Heymann D., “Characterization of circulating tumor cells as a reflection of the tumor heterogeneity: Myth or reality?,” Drug Discov. Today 24, 763–772 (2019). 10.1016/j.drudis.2018.11.017 [DOI] [PubMed] [Google Scholar]
  • 18.Hapach L. A., Carey S. P., Schwager S. C., Taufalele P. V., Wang W., Mosier J. A., Ortiz-Otero N., McArdle T. J., Goldblatt Z. E., Lampi M. C., Bordeleau F., Marshall J. R., Richardson I. M., Li J., King M. R., and Reinhart-King C. A., “Phenotypic heterogeneity and metastasis of breast cancer cells,” Cancer Res. 81, 3649–3663 (2021). 10.1158/0008-5472.CAN-20-1799 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Thiery J. P., Acloque H., Huang R. Y., and Nieto M. A., “ Epithelial-mesenchymal transitions in development and disease,” Cell 139, 871–890 (2009). 10.1016/j.cell.2009.11.007 [DOI] [PubMed] [Google Scholar]
  • 20.Waks A. G. and Winer E. P., “Breast cancer treatment: A review,” JAMA 321, 288–300 (2019). 10.1001/jama.2018.19323 [DOI] [PubMed] [Google Scholar]
  • 21.Liu X., Li J., Cadilha B. L., Markota A., Voigt C., Huang Z., Lin P. P., Wang D. D., Dai J., Kranz G., Krandick A., Libl D., Zitzelsberger H., Zagorski I., Braselmann H., Pan M., Zhu S., Huang Y., Niedermeyer S., Reichel C. A., Uhl B., Briukhovetska D., Suarez J., Kobold S., Gires O., and Wang H., “Epithelial-type systemic breast carcinoma cells with a restricted mesenchymal transition are a major source of metastasis,” Sci. Adv. 5, eaav4275 (2019). 10.1126/sciadv.aav4275 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Williams E. D., Gao D., Redfern A., and Thompson E. W., “Controversies around epithelial-mesenchymal plasticity in cancer metastasis,” Nat. Rev. Cancer 19, 716–732 (2019). 10.1038/s41568-019-0213-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Gensbittel V., Krater M., Harlepp S., Busnelli I., Guck J., and Goetz J. G., “Mechanical adaptability of tumor cells in metastasis,” Dev. Cell 56, 164–179 (2021). 10.1016/j.devcel.2020.10.011 [DOI] [PubMed] [Google Scholar]
  • 24.Follain G., Osmani N., Azevedo A. S., Allio G., Mercier L., Karreman M. A., Solecki G., Garcia Leon M. J., Lefebvre O., Fekonja N., Hille C., Chabannes V., Dolle G., Metivet T., Hovsepian F., Prudhomme C., Pichot A., Paul N., Carapito R., Bahram S., Ruthensteiner B., Kemmling A., Siemonsen S., Schneider T., Fiehler J., Glatzel M., Winkler F., Schwab Y., Pantel K., Harlepp S., and Goetz J. G., “Hemodynamic forces tune the arrest, adhesion, and extravasation of circulating tumor cells,” Dev. Cell 45, 33–52.e12 (2018). 10.1016/j.devcel.2018.02.015 [DOI] [PubMed] [Google Scholar]
  • 25.Mak M., Reinhart-King C. A., and Erickson D., “Elucidating mechanical transition effects of invading cancer cells with a subnucleus-scaled microfluidic serial dimensional modulation device,” Lab Chip 13, 340–348 (2013). 10.1039/C2LC41117B [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Liu X., Fang J., Huang S., Wu X., Xie X., Wang J., Liu F., Zhang M., Peng Z., and Hu N., “Tumor-on-a-chip: From bioinspired design to biomedical application,” Microsyst. Nanoeng. 7, 50 (2021). 10.1038/s41378-021-00277-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Landwehr G. M., Kristof A. J., Rahman S. M., Pettigrew J. H., Coates R., Balhoff J. B., Triantafillu U. L., Kim Y., and Melvin A. T., “Biophysical analysis of fluid shear stress induced cellular deformation in a microfluidic device,” Biomicrofluidics 12, 054109 (2018). 10.1063/1.5063824 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Xia Y., Wan Y., Hao S., Nisic M., Harouaka R. A., Chen Y., Zou X., and Zheng S. Y., “Nucleus of circulating tumor cell determines Its translocation through biomimetic microconstrictions and Its physical enrichment by microfiltration,” Small 14, e1802899 (2018). 10.1002/smll.201802899 [DOI] [PubMed] [Google Scholar]
  • 29.Dash S. K., Patra B., Sharma V., Das S. K., and Verma R. S., “Fluid shear stress in a logarithmic microfluidic device enhances cancer cell stemness marker expression,” Lab Chip 22, 2200–2211 (2022). 10.1039/D1LC01139A [DOI] [PubMed] [Google Scholar]
  • 30.Huveneers S., Daemen M. J., and Hordijk P. L., “Between Rho(k) and a hard place: The relation between vessel wall stiffness, endothelial contractility, and cardiovascular disease,” Circ. Res. 116, 895–908 (2015). 10.1161/CIRCRESAHA.116.305720 [DOI] [PubMed] [Google Scholar]
  • 31.Follain G., Herrmann D., Harlepp S., Hyenne V., Osmani N., Warren S. C., Timpson P., and Goetz J. G., “Fluids and their mechanics in tumour transit: Shaping metastasis,” Nat. Rev. Cancer 20, 107–124 (2020). 10.1038/s41568-019-0221-x [DOI] [PubMed] [Google Scholar]
  • 32.Kisanga E. R., Gjerde J., Guerrieri-Gonzaga A., Pigatto F., Pesci-Feltri A., Robertson C., Serrano D., Pelosi G., Decensi A., and Lien E. A., “Tamoxifen and metabolite concentrations in serum and breast cancer tissue during three dose regimens in a randomized preoperative trial,” Clin. Cancer Res. 10, 2336–2343 (2004). 10.1158/1078-0432.CCR-03-0538 [DOI] [PubMed] [Google Scholar]
  • 33.Zasadil L. M., Andersen K. A., Yeum D., Rocque G. B., Wilke L. G., Tevaarwerk A. J., Raines R. T., Burkard M. E., and Weaver B. A., “Cytotoxicity of paclitaxel in breast cancer is due to chromosome missegregation on multipolar spindles,” Sci. Transl. Med. 6, 229ra243 (2014). 10.1126/scitranslmed.3007965 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Park H. A., Brown S. R., and Kim Y., “Cellular mechanisms of circulating tumor cells during breast cancer metastasis,” Int. J. Mol. Sci. 21, 5040 (2020). 10.3390/ijms21145040 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Zorov D. B., Juhaszova M., and Sollott S. J., “Mitochondrial reactive oxygen species (ROS) and ROS-induced ROS release,” Physiol. Rev. 94, 909–950 (2014). 10.1152/physrev.00026.2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Lin D., Shen L., Luo M., Zhang K., Li J., Yang Q., Zhu F., Zhou D., Zheng S., Chen Y., and Zhou J., “Circulating tumor cells: Biology and clinical significance,” Signal Transduction Target Theor. 6, 404 (2021). 10.1038/s41392-021-00817-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Sun Y. F., Guo W., Xu Y., Shi Y. H., Gong Z. J., Ji Y., Du M., Zhang X., Hu B., Huang A., Chen G. G., Lai P. B. S., Cao Y., Qiu S. J., Zhou J., Yang X. R., and Fan J., “Circulating tumor cells from different vascular sites exhibit spatial heterogeneity in epithelial and mesenchymal composition and distinct clinical significance in hepatocellular carcinoma,” Clin. Cancer Res. 24, 547–559 (2018). 10.1158/1078-0432.CCR-17-1063 [DOI] [PubMed] [Google Scholar]
  • 38.Hope J. M., Bersi M. R., Dombroski J. A., Clinch A. B., Pereles R. S., Merryman W. D., and King M. R., “Circulating prostate cancer cells have differential resistance to fluid shear stress-induced cell death,” J. Cell Sci. 134, jcs251470 (2021). 10.1242/jcs.251470 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Moose D. L., Krog B. L., Kim T. H., Zhao L., Williams-Perez S., Burke G., Rhodes L., Vanneste M., Breheny P., Milhem M., Stipp C. S., Rowat A. C., and Henry M. D., “Cancer cells resist mechanical destruction in circulation via RhoA/actomyosin-dependent mechano-adaptation,” Cell Rep. 30, 3864–3874.e6 (2020). 10.1016/j.celrep.2020.02.080 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Paul C. D., Bishop K., Devine A., Paine E. L., Staunton J. R., Thomas S. M., Thomas J. R., Doyle A. D., Miller Jenkins L. M., Morgan N. Y., Sood R., and Tanner K., “Tissue architectural cues drive organ targeting of tumor cells in zebrafish,” Cell Syst. 9, 187–206.e16 (2019). 10.1016/j.cels.2019.07.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Pepona M., Balogh P., Puleri D. F., Hynes W. F., Robertson C., Dubbin K., Alvarado J., Moya M. L., and Randles A., “Investigating the interaction between circulating tumor cells and local hydrodynamics via experiment and simulations,” Cell. Mol. Bioeng. 13, 527–540 (2020). 10.1007/s12195-020-00656-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Perea Paizal J., Au S. H., and Bakal C., “Squeezing through the microcirculation: Survival adaptations of circulating tumour cells to seed metastasis,” Br. J. Cancer 124, 58–65 (2021). [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.

Data Availability Statement

The data that support the findings of this study are available within the article.

Articles from Biomicrofluidics are provided here courtesy of American Institute of Physics

RESOURCES