Quantitative impedance-based characterization of breast cancer cell migration and metastatic potential

Abstract

Background

See More

Cellular impedance-based assays offer a sensitive, label-free, and non-destructive method to continuously monitor cells in real time, allowing the assessment of both kinetics and degree of migration for breast cancer cells. A scratch assay is one of the most commonly used methods for testing cell migration in a two-dimensional (2D) monolayer culture. Traditional methods to evaluate 2D cancer migration commonly use image analysis to determine the rate of wound closure over a set of timepoints as an indicator of migratory/metastatic potential for cancer cells. An impedance-based assay system was employed towards establishing a modified wound healing assay technique that can measure wound coverage and therefore, 2D cancer migration continuously. This method can also be used to measure a variety of cell characteristics, including proliferation and epithelial barrier integrity.

Results

Using the Maestro Z Live-cell Analysis System by Axion Biosystems, cell spread, related to single cell morphology, and cell proliferation were observed for multiple breast cancer cell lines. A distinct quantifiable difference in the behavior of aggressive triple-negative breast cancer cells (HCC1806, MDA-MB-231), compared to less aggressive luminal MCF7 cells was determined. With an established assay method, cells were then treated with pro-inflammatory cytokine leptin, which plays a crucial role in metabolism and epithelial to mesenchymal transition (EMT), to verify assay sensitivity. The effects of leptin concentration in media were measurable for MCF7 and HCC1806 cells, and cell barrier integrity was significantly higher in the luminal MCF7 cells as compared to the more aggressive triple-negative cell lines. Cell migration to close a physical wound was measured over 36 hours, with the modified wound healing assay providing quantifiable evidence that the more aggressive breast cancer cells migrated to close the gap.

Conclusions

This work validates the use of cellular impedance-based assay systems to evaluate multiple cell characteristics. In a single experiment, cell spread, cell proliferation, cell-cell barrier integrity, and 2D cell migration were able to be quantified. These findings parallel previously published data for cell migration of the specific cell lines used, while highlighting the role of leptin in cancer behavior. Overall, the potential for a bioelectronic impedance assay system was demonstrated and its validity in effectively detecting and quantifying cell behaviors was proven.

Background

Breast cancer is the second most lethal cancer in women. When detected early, the rate of breast cancer survival is high (99%), however that rate decreases significantly (27%) once the cancer has metastasized to other parts of the body, such as the brain, lungs, bone, and liver [1]. Unfortunately, effective methods to detect metastatic cancer are limited. If not identified at the original breast cancer diagnosis, detection of metastatic breast cancer will not happen until after the cancer has spread to other parts of the body, resulting in delayed treatment and often, more advanced/detrimental cancer cases. Thus, early identification and determination of metastatic cancer could significantly alter therapeutic approaches worldwide.

There have been significant strides towards the identification and detection of metastatic breast cancer through the investigation of genomic, proteomic, and metabolomic profiling of samples to identify potential biomarkers for breast cancer metastasis correlated to clinical data. In these instances, there is reliance on multiple analytical methods for sample analysis (i.e. nuclear magnetic resonance (NMR) and mass spectrometry for metabolomics), which can be labor intensive and subject to delayed information processing since correlating clinical data may not always be readily obtained. Key factors known to influence breast cancer metastasis are receptor subtype, environment, and motility. Therefore, a method that allows for evaluation of these factors, while yielding an earlier, more reliable, and time-efficient approach for identification of metastatic cells is needed. This research validates the use of a novel bioelectronic assay to characterize breast cancer cell motility and response to biochemical stimuli in vitro.

Breast cancer metastasis remains one of the leading causes of death for women with cancer, accounting for 70% of breast cancer deaths, largely because methods to predict metastasis are limited. Metastasis remains difficult to predict, in part, due to cancer heterogeneity. Three subtypes of breast cancer can be determined by their receptor status of the estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor-2 (HER2). This determines treatment and is correlated to how aggressive/metastatic a cancer may be. The most aggressive type of breast cancer is triple negative cancer (ER-, PR- and HER2-). This subtype has a high chance of metastasis; therefore, this research includes two different human breast cancer cell lines that are defined as triple negative.

In addition to receptor status, environmental factors can also influence a cancer cell’s behavior. Proinflammatory cytokines like leptin, Interleukin (IL)-1β, IL-6, or tumor necrosis factor (TNF)-α encourage blood vessel formation (angiogenesis) towards the tumor site, a primary step in the metastatic cascade [2, 3]. Leptin, in particular, can bind to breast cancer cells and trigger proliferation, invasion, and migration through oncogenic signaling pathways like mitogen-activated protein kinase (MAPK), phosphatidylinositol 3-kinase (PI3K), and janus kinase/signal transducers and activators of transcription (JAK/STAT) [4, 5]. Additionally, leptin levels have been shown to be significantly higher in breast cancer cases with lymph node metastasis, compared to no metastasis [6]. In this work, we aim to establish the utility of a bioelectronic impedance-sensing assay as an alternate in vitro approach applicable for studying cell migration and assessing metastatic potential of cancer cells by demonstrating quantifiable effects of leptin concentration on breast cancer aggressiveness, separate from leptin-induced immune response [7, 8] in vitro.

A wound closure assay, also known as a wound healing assay or scratch assay, is an example of an in vitro assay to study cancer metastasis. In this method, cancer cells are seeded at the bottom of a well plate and grown to confluence. Once confluent, a physical scratch through the cancer cell monolayer is made, creating a gap. Over a set period, the cancer cells will migrate towards each other to close the gap or “heal” the wound. The distance traveled by the cancer cells and the time it takes to close the gap is viewed as a cancer cell’s migratory potential and can be used to determine metastatic likelihood. While wound healing assays are a common and effective tool in evaluating breast cancer cell migration, they rely heavily on periodic imaging or video recording for data collection, which require time-intensive processing and can limit throughput precision [9].

To overcome these limitations, impedance-based techniques, such as Electric Cell-substrate Impedance Sensing (ECIS), offer a compelling alternative for monitoring cell migration and behavior in real time [10, 11]. Bioelectronic systems, like ECIS, are quickly gaining popularity as a label-free, non-invasive, and non-terminal method that continuously measures changes in electrical impedance to quantify several aspects of cell behavior. In this technique, a small alternating current (I) is applied across an electrode configuration at the bottom of a tissue culture surface. This results in a potential (V) across the electrodes which is measured by the ECIS instrument. The impedance (Z) is then determined by Ohm’s law where Z = V/I [12]. By tracking impedance changes as cells adhere, proliferate, and migrate, ECIS provides continuous, quantitative data without the need for frequent imaging.

As cells adhere onto the electrode-coated surface and proliferate, they act as insulators, increasing impedance (Fig. 1). Alterations in impedance reflect key cellular characteristics, including surface coverage, morphology, and adhesion, as shown in Fig. 1B. When cell function shifts in response to the environment (e.g. growth factors, cytokines, hypoxia, cytotoxic agents), so do the impedance readings. At higher alternating current (AC) frequencies, the current travels via transcellular pathways (Fig. 1B), providing information about cellular surface coverage and morphology. Conversely, at lower AC frequencies, the current flows via paracellular pathways. To quantify this, trans-endothelial/epithelial electrical resistance (TEER) is widely used as a robust and established method for assessing barrier function in epithelial and endothelial monolayer cultures [13, 14]. This measurement is particularly relevant in cancer studies, where loss of barrier integrity reflects the degradation of tight junctions during tumor cell migration and invasion through both epithelial layers and vascular endothelium [15,16,17].

Given its advantages, ECIS is increasingly being adopted in cancer research. Several commercial systems are available from companies like Applied BioPhysics and Axion Biosystems, while many researchers also develop custom setups tailored to their experimental needs [18,19,20,21]. Individual systems often lack cross-study comparability, therefore this work utilizes the commercially available Maestro Z system (Axion Biosystems) [22, 23], ensuring consistency and reproducibility. Although ECIS has been used predominantly with breast cancer cells to quantify drug interactions [19, 24,25,26], a system that integrates multiple behavioral readouts using ECIS can significantly advance efforts to diagnose and characterize metastatic potential in breast cancer.

To better identify metastatic cancer cells more efficiently in 2D culture, we propose to leverage cell impedance for a “smart” in vitro modeling approach to interrogating the relationship between specific attributes of cancer cells and their metastatic potential. Such an approach provides a modern, quantitative framework for studying cancer cell migration and metastasis and, because of its 96-well plate format, small numbers of cells are utilized, making it practical for clinical settings where biopsies of human tissues do not typically yield a large quantity of cells for culturing [27]. Specifically, we first determined the feasibility of a quantitative impedance-based assay to characterize cancer cell proliferation and migration. The assay approach was then applied to measure the effects of leptin on breast cancer cell behavior to demonstrate assay sensitivity and understand how migration is affected by varying leptin concentrations. Overall, the approach established in this work allows for the collection of an array of data within a single experiment, streamlining the process of cancer cell characterization.

Fig. 1

Cell impedance concepts. (A) Electrical circuit of an animal cell, where R_o is resistance of the surrounding media, R_m is the resistance of the membrane, C_m is the capacitance is the cell membrane, and R_i is the resistance of the protoplasm. The lipid bilayer of the cell membrane works in both a capacitive and resistive role, which both contribute to impedance. The flow of electrical signals through a cell monolayer at different alternating current (AC) frequencies. At low frequencies, current primarily travels through paracellular pathways (red arrow), reflecting barrier integrity and providing information about cell phenotype or “what kind” of cell. At high frequencies, current predominantly passes transcellularly (blue arrows), correlating with cell coverage and “how many” cells there are. (B) Applications of cell impedance assays for monitoring various cellular behaviors. As cells cover the electrode surface, impedance increases. Different factors can be measured, depending on frequency, allowing for many practical applications of this technology. Created with BioRender.com

Methods

Reagents and cell lines

Poly-D-lysine lyophilized powder (Sigma Aldrich) was dissolved in sterilized deionized (DI) water to form a 21.3 µM solution. This solution was stored at -20˚C and thawed on ice when needed. To prepare leptin, the adipokine used as a cell stimulant, human leptin (Sigma Aldrich) was reconstituted in sterilized DI water to make a 0.5 µg/mL solution and 1.5 µg/mL solution. Leptin solutions were stored at 4 °C for short term storage and  -20˚C for longer storage.

MCF7, HCC1806, and MDA-MB-231 cells were purchased from American Type Culture Collection (ATCC). Characteristics of the cells include: MCF7 cells (metastatic adenocarcinoma, ER+, PR+); HCC1806 cells (triple negative ER-/PR-/HER2-), African American donor); and MDA-MB-231 (triple negative (ER-/PR-/HER2-), Caucasian donor). All three cell line populations were expanded as a monolayer in 75 cm² flasks (Falcon) at 37˚C and 5% CO₂ atmosphere in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (R&D Systems) and 1% penicillin/streptomycin (Gibco), referred to as DMEM-Complete. Three days before data collection began, cell media was changed every day and 200 µL of stimulant was added. This stimulant was either sterile DI water for the control groups, 0.5 µg/mL leptin for the normal concentration groups, or 1.5 µg/mL leptin for the high concentration groups. Cells were confirmed to be free of mycoplasma using a commercially available kit (InvivoGen).

Seeding cells onto CytoView plates

Experiments for this work were performed according to the plan shown in Fig. 2. Before cells were added to the CytoView-Z 96-well electrode plates (Axion), the culture surfaces were treated with 50 µL of the poly-D-lysine solution (21.3 µM) and incubated at room temperature for 1 h in a biosafety cabinet. Afterward, each well was washed with 50 µL of Phosphate Buffered Saline (PBS, Gibco) twice and allowed to dry in the biosafety cabinet. Cells were transferred from the 75 cm² flasks using 0.5% trypsin-EDTA (Gibco). Before seeding the cells on the electrode plate, a Media Only Baseline test was performed in the AxisZ software of the Maestro Z (Axion) by adding 100 µL of DMEM-Complete to each well of the plate and evenly distributing 8 mL of sterile room temperature DI water to the reservoirs on the CytoView-Z plates (shown in Supplementary Fig. 1A). The plate was docked in the Maestro Z instrument, and the Media Only Baseline was measured. Once the baseline was measured, the plate was transferred to a biosafety cabinet and the cells were added to the CytoView-Z plate. MDA-MB-231 cells were plated at a concentration of 75,000 cells/cm², while MCF7 and HCC1806 cells were plated at 85,000 cells/cm² to account for varying growth rates and cell sizes. A volume of 4 µL of the stimulant was added to each well and the plate was allowed to rest in a biosafety cabinet at room temperature for 1 h to allow cells to attach to the culture surface and avoid edge effects. Lastly, the plate was docked in the Maestro Z and cultured at 37˚C and 5% CO₂. Impedance and Barrier Index measurements are automatically collected upon plate engagement.

Evaluating cell proliferation and barrier integrity

Cell growth was monitored by tracking impedance through the AxisZ software associated with the Maestro Z. At 24 h, the plate was undocked, and the cells were visually checked for confluence using an EVOS FLc microscope (Invitrogen). Shown in Fig. 2, all medium was changed to a low-serum medium (DMEM supplemented with 2% fetal bovine serum and 1% penicillin/streptomycin) to halt growth. A volume of 4 µL of the stimuli treatment was added to each well. The CytoView-Z plate was redocked in the Maestro Z and cells continued to culture for 12 h while their growth ceased. Impedance readings were collected on Axion’s AxisZ Software approximately every minute and all data exported to Excel at the end of experiment. Barrier Index measurements were collected simultaneously with impedance during cell proliferation. Barrier integrity was then calculated in AxisZ.

Fig. 2

Schematic of the experimental flow. Created with BioRender.com

Modified quantitative wound healing assay

The CytoView-Z plate was undocked from the Maestro Z and placed in a biosafety cabinet. To create a physical wound in each well of the plate, a multi-channel micropipettor fitted with micropipette tips was used to gently scratch the plate surface in a vertical line. The media was replaced by fresh low-serum media and supplemented with 4 µL of stimuli treatment. The plate was redocked in the impedance system. After 36 h, the experiment was terminated.

Transwell migration assay

CellTracker Green (Invitrogen) was used to stain each cell line immediately prior to seeding the cells on the inserts. Transwell inserts of 24-mm diameter and 8-µm porosity (Corning Costar) were used to evaluate migration potential of each cell line. DMEM-Complete was added to the bottom wells and each cell line was seeded on individual inserts at 100,000 cells/cm² using a low-serum medium. Plates were cultured and evaluated at 6, 12, 18, and 24 h. To image cell nuclei, 2 drops of NucBlue™ Live ReadyProbes™ Reagent (Hoechst 33342, Invitrogen) per mL of media were added to each well and allowed to incubate at room temperature for 20 min. After incubation, each transwell insert was gently rinsed with PBS and transferred to a well with fresh PBS. The top of each insert was gently swabbed to remove the non-migrated cells. Using the Cytation1 Cell Imaging Multi-Mode Reader (Agilent BioTek), each well was imaged in 4 locations, and the nuclei were counted using Gen5 Cell Imaging & Microscopy Software (BioTek). Object count was performed using a threshold of 9000, minimum object size of 9 microns, and maximum object size of 35 microns. The number of nuclei was used to quantify the number of migrated cells per image for each cell type.

Traditional wound healing assay

MCF7, HCC1806, and MDA-MB-231 cells were seeded onto a standard 96-well plate using DMEM-Complete at the same density as the modified wound healing assay. CellTracker Green was used to stain the cells immediately prior to seeding the cells on the plate. At 24 h, the cells were visually checked for confluence using an EVOS FLc microscope (Invitrogen). To halt growth, all medium was changed to a low-serum medium. The plate was returned to the incubator and cells continued to culture for 12 h while their growth ceased. After 36 h of culture, nuclei were stained, and cell monolayers were scratched.

To stain the nuclei, NucBlue (Invitrogen) was added to a stock of DMEM complete at a concentration of 2 drops per mL. Spent media was removed from the 96-well plate and 100 µL of NucBlue-DMEM solution was added to each well and incubated in a biosafety cabinet for 20 min at room temperature. To create a physical wound in each well, a multi-channel micropipettor fitted with micropipette tips was used to gently scratch the plate surface in a vertical line. The media was replaced by fresh low-serum media and the plate was imaged using a Cytation 1 (Agilent Biotek). Each well was imaged in a two-by-three montage that was stitched into one image using Agilent Biotek’s Gen5 software. The plate was imaged every 12 h for 36 h to visualize wound closure over time.

Gene and protein expression evaluation

RNA-seq gene expression data was obtained from the Cancer Dependency Map (DepMap) Project, a Broad Institute resource compiling genetic and molecular profiles across diverse human cancer models [28,29,30]. Expression values (log₂[TPM + 1]) were used to compare EMT-related gene expression with impedance measurements from this study. A panel of 10 EMT-associated genes was analyzed: CDH1, DSP, OCLN, TWIST1, VIM, ZEB1, FN1, SNAI1, and CDH2. Results were visualized as a robust z-score heatmap generated in Morpheus (Broad Institute), and hierarchical clustering of cell line expression was performed using Pearson correlation and average linkage [31].

To assess leptin receptor (Ob-R) expression, cells were cultured as monolayers and lysed on ice for 45 min with RIPA buffer (Thermo Scientific) and periodic agitation. Lysates were centrifuged at 10,000 × g for 10 min, and the supernatant was stored at − 80 °C. Ob-R protein levels were measured using a human leptin receptor ELISA (Assay Genie, HUFI00192).

Data processing and statistical analysis

Impedance data was exported via Microsoft Excel, where the rate of wound closure was calculated. Statistical analyses were calculated using GraphPad Prism 3.9. Statistical significances of ECIS data were determined via ordinary one-way ANOVA. Transwell migration significance was calculated using an unpaired parametric t-test.

Results

Determination of quantitative impedance assay feasibility

Preliminary evaluation of the efficacy of cell impedance for monitoring changes in breast cancer cell behavior was performed. Stained images demonstrating the representative morphology of the three tested breast cancer cell lines, MCF7, HCC1806, and MDA-MB-231 cells, are shown in Fig. 3A. Each of the three cell lines were cultured using the Maestro Z Impedance Assay System, with a high frequency impedance, measured across the electrodes of each well.

Initial evaluation of the data for control cell samples without any treatment showed a distinct impedance profile for each breast cancer line, where variations in Impedance and Barrier Integrity were detected during adhesion, proliferation, and wound healing phases of the culture period (Fig. 3B and C, respectively). The graphs obtained show high frequency impedance averaged across the set of wells for each cell type. Key points during the culture period are reflected in the graphs, where the baseline impedance for each cell line is indicated at 0 h, and an increase in impedance is observed as cells proliferate up to 36 h. Interruptions in the culture process, such as media change at 24 h or the initiation of wounds for a scratch assay at confluence (36 h) are reflected by an immediate decrease in impedance, and subsequent increase, depending on cell response.

For the tested cell lines, as shown in Fig. 3B, during the first six to eight hours, the MDA-MB-231 and HCC1806 cell impedance values plateaued. Impedance measurements during the first 36 h of proliferation showed that the MCF7 cells reached the highest impedance for the tested cells, and the MDA-MB-231 cells maintained the lowest impedance for the entire culture period. The MCF7 cells had an average impedance of 49.5 Ω at confluency, while the HCC1806 and MDA-MB-231 cells measured as 34.1 Ω and 11.8 Ω respectively. The Barrier Integrity of each breast cancer cell monolayer was measured during proliferation by quantifying the ratio between low frequency (1 kHz) resistance and high frequency (41.5 kHz) resistance. In addition to the highest impedance, the MCF7 cells also had the highest Barrier Integrity among the three cell lines tested, which remained the highest during the proliferation and wound healing assay phases (Fig. 3C). These functional differences align with reported EMT-associated gene expression (Fig. 3D), where MCF7 cells exhibit and retain epithelial-like characteristics, the MDA-MB-231 cells are strongly mesenchymal, and the HCC1806 cells fall between these extremes.

Fig. 3

Imaging and impedance monitoring of breast cancer cell dynamics. (A) Breast cancer cells (green) and their nuclei (blue) stained to visualize morphology. (B) Application of impedance assay system for simultaneously monitoring Impedance and (C) Barrier Integrity during culture. Cells were allowed to attach to plate and proliferate for 36 h (yellow background). Then, each well was physically scratched, and wound closure was observed for another 36 h (purple background). (D) Heatmap of EMT-associated gene expression, sourced from the DepMap Portal. Pearson correlation and average linkage were used for hierarchical clustering of cell lines

Fig. 4

Wound closure over 36 h. (A) Normalized high frequency (41.5 kHz) impedance of breast cancer cells without any treatment. To visualize wound closure, impedance was normalized to the value of each well pre-scratch. (B) Bar graph of wound closure, measured by normalized impedance. Graph shows normalized impedance levels directly after creating a wound and 12, 24, and 36 h after. The grey dotted line represents the point in which the impedance matches the pre-scratch impedance. Significance between groups is denoted by asterisks: * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. n = 6 wells per condition

Confirmation of quantitative wound healing assay protocol

When cells reached confluence at 36 h, the wound healing assay was initiated by making a physical scratch within each well, and the impedance was measured for an additional 36 h as cells either closed the created wound or not depending on cell type (Supplementary Fig. 1B). Impedance was normalized to the point prior to creating a wound. Immediately after the scratch a decrease in impedance was observed for each cell type, as indicated in Fig. 4A. As shown in Fig. 4B, both triple-negative cell lines were able to close the created wounds within 36 h, if not sooner, with the impedance for the HCC1806 and MDA-MB-231 cells reaching or exceeding the impedance values measured before the scratch. The less aggressive MCF7 cells showed very little migration, as there was little to no change in the measured MCF7 impedance over the 36 h. These observations were consistent with cell migration measured via a transwell migration assay and a traditional wound healing assay performed for each cell line (Supplementary Fig. 2).

Fig. 5

Impedance and barrier integrity across the full experimental timeline. (A) High-frequency impedance (41.5 kHz) and (B) barrier integrity measurements for three breast cancer cell lines treated with leptin at 0, 10, or 30 ηg/mL. Cells were seeded and allowed to attach and proliferate for 36 h (yellow background). A scratch wound was then created, and wound closure was monitored for an additional 36 h (purple background). n = 6 wells per condition

Application of impedance assay protocols to evaluate cell behavior

The high frequency impedance for each cell type treated with either a normal leptin concentration (10 ηg/mL) or high leptin concentration (30 ηg/mL) was measured and compared to control samples (Fig. 5A). In addition, the barrier integrity of each breast cancer cell monolayer was measured simultaneously (Fig. 5B). For each cell type, the effects of leptin on cell behavior were assessed by monitoring cells during adhesion, proliferation and wound healing phases of culture. For the MCF7 cells, their impedance and barrier integrity across all three leptin levels was significantly higher than either triple-negative cell line. Each cell line had a significant change in impedance with response to leptin (Fig. 6 A-B), however there was no consistent pattern of response to leptin across all three cell lines (Fig. 6 C).

Fig. 6

Cell spread is more pronounced in elongated breast cancer cell lines. (A) High frequency impedance recorded in the first 12 h of the experiment to visualize HCC1806 and MDA-MB-231 cell spread. Plate was removed from recording device at t = 6.5 h (dashed line), resulting in a data artifact. (B) High frequency (41.5 kHz) impedance recorded in the first 36 h of the experiment to visualize cell spread and proliferation. Media was changed at t = 24 h (dashed line). (C) Average impedance (41.5 kHz) at t = 36 h. * indicates a statistically significant comparison where p > 0.05. ** indicates a statistically significant comparison where p > 0.01. *** indicates a statistically significant comparison where p > 0.001. **** indicates a statistically significant comparison where p > 0.0001. (n = 6)

At high cell density (t = 36 h), MCF7 cell barrier integrity is influenced by concentration of leptin (Fig. 7A). Additionally, a significant difference in barrier integrity across all three control groups was measured. This is linked to cell-cell adhesion and shows a significantly higher barrier integrity for the less aggressive MCF7 cells in comparison to the more aggressive triple negative cell lines (Fig. 7B).

Fig. 7

Breast cancer cell monolayer barrier integrity. A) Cell-cell barrier integrity recorded in the first 36 h of the experiment. Media was changed at t = 24 h (dashed line). (B) Average barrier integrity at t = 36 h. Significance between groups is denoted by asterisks: * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. n = 6 wells per condition

Discussion

Here, we developed a protocol for impedance-based data collection that evaluates multiple cancer cell characteristics in a single experiment. This method is sensitive, label-free, and non-destructive, making it a suitable option for small cell populations. Cell surface adhesion, proliferation, response to stimuli, migration, and barrier integrity were all measured within a single experiment, providing knowledge of how each cell line compares with one another and how leptin affects these behaviors.

Determination of quantitative impedance assay feasibility

Cell spread and proliferation were evaluated using high frequency (41.5 kHz) impedance. At this frequency, the impedance of the cell membrane is relatively low. Thus, most of the current couples capacitively through the cell membranes. In other words, the current is traveling in a transcellular manner, providing information about the cell layer, such as confluency and surface coverage where the cells attach to the electrode surface and spread out. In the first 12 h of culture, the impedance of each triple-negative breast cancer cell group plateaued. This was later identified as cell spread as the cells spread out and attach onto the substrate, a phenomenon that is well documented using ECIS [32,33,34]. While MCF7 cells have a cobblestone-like morphology, the triple-negative breast cancer cell lines are more elongated in shape. Therefore, it is indicated that the HCC1806 and MDA-MB-231 cell lines exhibited a plateau in their impedance readings as the cells work to fully attach to the plate before proliferating.

Once cells began to proliferate, the MCF7 cells grew quickly and tended to pack tightly together due to their more epithelial phenotype as indicated by reported gene expression data. Because of this, a higher number of MCF7 cells can cover the same surface area as either of the triple-negative breast cancer cell lines. This is reflected in the impedance reading at t = 36 h, when cells are at their highest density. This alone can give us insight into the growth patterns of each cell line and how cells might behave in vivo.

Endothelial and epithelial cells express tight junctions, allowing them to link tightly with their neighboring cells to form a selectively permeable barrier. A cancer cell’s increase in invasiveness is often associated with a transition from epithelial-like cells, that maintain a cuboidal shape and adhere to the basement membrane, to more elongated mesenchymal-like cells [35, 36]. There is an observed tradeoff between the cells’ proliferation and invasiveness for this process, the epithelial-to-mesenchymal-transition. This change occurs on an epigenetic level where markers like E-cadherin, desmoplakin, and keratin are associated with an epithelial-like state and n-cadherin, vimentin, and fibronectin are associated with a mesenchymal-like state [37]. A degradation of tight junctions is associated with the transition to a more mesenchymal-like state, leading to decreased barrier integrity and a proclivity for migration [38].

At lower frequencies, electrical signal passes through paracellular pathways, measuring the integrity and permeability of a cell monolayer. While this measurement, referred to as trans-epithelial electrical resistance (TEER) typically requires a fully confluent monolayer to accurately measure barrier function, the Maestro Z normalizes this data to cell surface coverage, measured with high frequency (41.5 kHz) resistance (Eq. 3.1). As a result, this is a unitless measure.

$$\begin{array}{l}\:Barrier\:Integrity\\=\frac{measured\:resistance\:at\:1\:kHz\:current\:\left[{\Omega\:}\right]}{measured\:resistance\:at\:41.5\:kHz\:current\:\left[{\Omega\:}\right]}\end{array}$$

(3.1)

This difference in barrier integrity is observable across all three cell lines in Fig. 3C where highly cuboidal and non-migratory MCF7 cells maintain a high normalized TEER value (i.e. barrier integrity). The more aggressive HCC1806 and MDA-MB-231 cells have a much lower barrier integrity, which is associated with their morphology, EMT-status, and aggressiveness.

Confirmation of quantitative wound healing assay protocol

Cell migration is essential for many physiological processes including embryonic development, wound repair, angiogenesis, and tumor metastasis [39]. Cancer cell migration is one of the first steps in the metastatic cascade and is commonly used to evaluate cancer cell aggressiveness.

With high frequency (41.5 kHz) cell impedance being directly related to cell surface coverage, we can use this value to determine cell migration as a form of wound closure. Impedance was normalized to the value prior to creating a wound, meaning that the wound would be effectively covered once the cell value has reached 1.0 again. Over 36 h of wound closure, the luminal MCF7 cells did not show any measurable difference in impedance and did not migrate to close the wound at all. This is in contrast to the triple-negative breast cancer cell lines, which displayed an increase in impedance over time as they worked to close the wound. The MDA-MB-231 cells migrated the most and quickest of all three breast cancer cell lines, and eventually began to die back, resulting in a drop in impedance between 24 h and 36 h of migration. The non-migratory tendencies of the luminal MCF7 cell line have been mentioned previously in literature [40,41,42,43,44]. Luminal cell lines are comparably more differentiated and have a lower propensity for migration due to a higher number of tight junctions [45] and MCF7 cells have higher rates of claudin-1 than their triple-negative counterparts [38, 46]. Our data correlates with published information, thus validating the efficacy of our modified wound healing approach; however, we also assessed migration potential of our specific cell populations using a more traditional transwell migration assay. Our transwell migration data mirrors our impedance-measured wound closure where the MCF7 and HCC1806 cells exhibit no migration over 24 h, and the MDA-MB-231 cells do. Notably, the MDA-MB-231 groups displayed immediate and rapid migration with both migration assays. All of this further proves the efficacy of impedance-based cell monitoring to measure cell migration.

Application of impedance assay protocols to evaluate cell behavior

The sensitivity and efficacy of the quantifiable assay approaches developed here was tested by treating cultured cells with exogenous leptin added to their culture media at varying levels. It was expected that high levels of leptin, commonly associated with diseased states, would impact cell proliferation and migration, particularly for the more aggressive cell lines tested. Impedance and barrier integrity measurements in response to leptin treatment yielded quantitative and reproducible measures, with a coefficient of variation below 0.25 across all groups (Supplemental Table 1), demonstrating that this assay is sensitive to subtle changes in cell behavior. However, the effects of leptin varied among cell lines.

Because MCF7 cells have a cobblestone morphology and grow tightly packed, their impedance across all three leptin levels was significantly higher than either triple-negative cell line. Each cell line had a significant change in impedance with response to leptin, however there was no consistent pattern of response to leptin across all three cell lines. Leptin has previously been shown to have an effect on cell breast cancer cell proliferation [47,48,49], though there is no distinct pattern in response across cell lines. In this work, analysis of leptin receptor (Ob-R) expression (Supplementary Fig. 3) revealed that MCF7 cells expressed significantly higher levels of Ob-R compared to both TNBC lines. This may explain why MCF7 cells showed the strongest response to leptin, HCC1806 cells showed minimal response, and MDA-MB-231 cells showed none. These findings suggest that differences in Ob-R expression contribute to the varied responses observed. Additionally, leptin was shown to have an effect on cell barrier integrity across all three cell lines, but particularly with the luminal MCF7 cells. Previous studies have shown that leptin can have an effect on E-cadherin expression within breast cancer cells [50, 51]. Since MCF7 cells have a higher expression of E-cadherin, this may explain why their barrier integrity is affected more by leptin expression.

For the modified wound healing assay, since high frequency impedance corresponds to electrode surface coverage, it was used to track wound closure. When treated with leptin it was expected that high leptin concentrations would yield increased cell migration. However, while there were differences observed between cell types and across all MCF7 cells tested, there was no measurable difference in wound closure as a response to leptin (Supplementary Fig. 4).

Leptin was shown to have no effect on cell migration in this work, although leptin has previously been shown to increase cell migration and invasion [52, 53]. This outcome may be resultant from the concentrations of leptin used within this study in comparison to other studies. Concentration levels were determined based on a literature review of serum leptin levels [6, 54,55,56,57,58,59], however leptin concentrations within the fatty tumor microenvironment are likely higher. Additionally, the effect of leptin on breast cancer migration in vivo involves not only oncogenic signaling within the cancer cell but can lead to an inflammatory response in macrophages and T-cells in the surrounding microenvironment. By decoupling the leptin-induced cancer cell and immune cell responses, this may understate the effect that leptin plays in breast cancer progression.

Performance and operational issues

Overall, this experimental method proved simple and effective in providing a large collection of data on many different cell characteristics. ECIS is non-invasive and non-terminal allowing us to evaluate cell spread, proliferation, barrier integrity, and cell motility with the same cell population. This is practical in a clinical setting where there is a small population of cells (e.g. derived from a biopsy).

The primary operational issue of this adapted wound healing assay is creating the wound without damaging the electrode embedded within each well. This is due to the difficulty of creating an even distribution of weight across a row of wells with the multichannel pipettor. While some wells had damaged electrodes from the pipette tip, others did not show a full scratch. In later iterations of this experiment, the user applied a more even distribution of weight by holding the pipettor at its base to perform the scratch. In future work, we plan to test higher precision wound-making tools, like the AutoScratch Wound Making Tool by Agilent.

Another consideration is that removing the plate from the impedance instrument during an experiment causes transient spikes in the recorded data. Because cells are highly sensitive to environmental perturbations, temporary changes in behavior may result from removal from the controlled environment. While such fluctuations may not be apparent with conventional imaging or biochemical assays, impedance measurements are highly sensitive and readily capture these variations. Awareness of this effect is important for both experimental planning and data interpretation. With these minor adaptations, this method enables a straightforward wound healing assay that requires minimal maintenance while providing a broader and more comprehensive view of cell behavior and migration over time.

Conclusion

Breast cancer metastasis remains one of the leading causes of death in women with cancer. While wound healing assays are a common way to evaluate 2D cell migration, traditional methods rely on image processing. This method is not only more computationally intense but is less quantitative. The methods outlined above propose a modified wound healing assay that can track wound closure, as cells migrate to close the gap, with ECIS. Impedance spectroscopy is a label-free, non-terminal way to continuously observe cell behavior throughout the entire experiment. Within this experiment, we were able to measure cell response to leptin (a proinflammatory adipokine), track cell spread and proliferation, quantify cell barrier integrity, and determine 2D cell migration via a modified wound closure assay. We found leptin to be a contributing factor in cancer cell proliferation and barrier integrity, however no variation in migration was recorded between our leptin conditions. This method allowed us to characterize cancer cell behavior and will be applied to more mammary cell lines in the future to establish a library of data on established breast cancer cell lines.