Inhibitory Effect to Smooth Muscle Cells Adhesion and Proliferation from Oscillating Electric Fields by Nanogenerators

Abstract

A common complication of the removal of atherosclerotic plaques or thrombi deposits to restore blood flow is restenosis. It is known that the excessive adhesion and proliferation of smooth muscle cells (SMCs) is the primary reason for restenosis. In this study, we conducted an in vitro study to show that a weak oscillating electric field (EF) generated by a mechanically-driven nanogenerator could prohibit SMC adhesion and proliferation on a substrate surface. Our results revealed a decrease in cell numbers when the oscillating EF was introduced underneath the substrate. The cell coverage was found dependent on the EF strength and oscillating frequency, where higher EF strength and frequency yielded stronger inhibitory effect. Comparing to the control, this reduction in cell coverage reached up to a 54% under the optimal EF parameters. This inhibitory effect was attributed to the EF-induced surface charge oscillation, which weakened the electrostatic interaction between the cell membrane and substrate. Our discovery suggests a potential for self-powered anti-restenosis solutions by integrating NG-induced oscillating EF with biomedical device surfaces.

Graphical Abstract

Introduction

Atherosclerosis is the leading cause of vascular disease worldwide.¹ It is characterized by the buildup of plaques in the inner lining of the artery wall resulting in the thickening and hardening of arteries.² The atherosclerotic plaques are composed of lipids, dead cells, and inflammatory cells, all encapsulated by a fibrous cap made up of smooth muscle cells, collagen, macrophages, and other immune cells.³ These plaques typically form in the high-pressure arterial system, especially where arteries branch, as blood flow becomes turbulent in these areas, triggering a pro-inflammatory response.⁴ Plaques with thick fibrous caps are considered stable, however when the fibrous cap becomes too thin, the plaque can rupture and subsequently lead to heart attacks and stoke. For symptomatic patients with occlusive or unstable plaques, vascular interventions such as angioplasty and stenting are employed to unblock the atherosclerotic plaque and/or thrombi deposits and restore blood flow to downstream organs and tissues.^5–7 While these vascular interventions are effective in restoring blood flow, their long-term efficacy is limited by a commonly occurred process called restenosis or the re-narrowing of arteries.⁸

The development of restenosis is primarily driven by the excessive adhesion and proliferation of smooth muscle cells (SMCs) at the site of intervention.⁹ As schematically shown in Figure 1a, this process begins with endothelial injury and inflammation.¹⁰ During angioplasty or stent placement, the endothelial lining of the artery is often damaged, triggering an inflammatory response.¹¹ The combination of endothelial injury and inflammation result in activation of SMCs, the predominant cell types residing in the medial layer.¹² Once activated, these normally quiescent SMCs migrate from the media to the intima (the innermost layer of the vessel) and proliferate, and produce extracellular matrix (ECM). The accumulation of proliferating SMCs and ECM causes the treated arteries to thickened and narrow, leading to a reduction in the diameter of the artery and thereby reducing blood flow.

figure 1.

experimental setup for anti-restenosis testing. (a) schematic of restenosis by smcs after vessel wall injury. (b) schematic showing ef experiment setup. (c) images of substrates used for movas cell culture: (i) image of prepared electrode substrates; (ii) image of connected substrate in cell culture dish inside incubator, (d) representative voltage output measurement from the external nanogenerator.

Several preventive and therapeutic strategies have been developed to prevent or treat restenosis.¹³ Drug-eluting stents (DES) slowly release anti-migratory and anti-proliferative drugs that inhibit SMC proliferation and migration, thereby reducing the incidence of restenosis.^{14, 15} Drug-coated balloons (DCB) deliver drugs directly to the arterial wall during angioplasty, helping to suppress SMC activity that leads to restenosis.^{16, 17} Targeting inflammation using of corticosteroids or specific anti-inflammatory agents is also shown to mitigate the early inflammatory response.¹⁸

Oscillating electric fields (EFs) have been shown to influence cellular behaviour, including cell adhesion, migration, and proliferation.^{19, 20} Recently, we discovered that a low-frequency and weak alternative EF could effectively prevent the attachment of biological species.²¹ While the EF used in this study had a much lower intensity than those used conventionally, it exhibited a significant anti-attachment effect by disturbing the surface electrical double layer, which subsequently disrupts the electrostatic force between the organics and substrate. Considering the electrostatic attraction is a common mechanism for cells to form an initial attachment to surfaces, we hypothesize that a low intensity alternating EF may prevent SMCs from adhering to the endothelium, and thereby inhibiting the subsequent buildup of SMCs. To test this hypothesis, we developed a device that produces a weak oscillating EF in vitro via a mechanically-driven nanogenerator (NG).^22–25 The optimal parameters for the oscillating electric fields, including frequency and amplitude were investigated to achieve the desired inhibitory effects on SMCs. This study revealed a novel drug-independent anti-restenosis mechanism that could be harnessed in implantable vascular devices.

Results and Discussion

Since SMC proliferation is a critical step contributing to restenosis (Figure 1a), we designed and constructed an in vitro experimental setup to study the effects of EFs on SMCs proliferation. As schematically shown in Figure 1b, a pair of insulated Au interdigitated electrodes were deployed to apply the EF to the substrate surface. The electrode pair has a spacing of 100 μm with a finger thickness of 100 μm and length of 1 cm, covering a total area of 4×1 cm² (Figure 1c-i, Figure S1). This design of the electrodes ensures uniform distribution of the electric field across the cell-seeded substrate, allowing for precise modulation of cellular activity. The entire electrode and substrate surface was covered by a thin and flexible film of Polydimethylsiloxane (PDMS) with inherent biocompatibility and chemical stability, allowing to prevent current leakage and provide a benign surface for cell growth. The optical transparency of PDMS contributed to clear microscopy and cell staining imaging. The two leads of the electrodes were connected to an external NG that was driven by mechanical forces to provide controlled voltage outputs. At the time of cell seeding, these PDMS coated substrates were placed in 6 cm cell culture plates and covered with culture media containing a suspended SMC line, Mouse Vascular Smooth Muscle Cell (MOVAS) (Figure 1c-ii). During the test, the external NG was pressed by linear actuator at controlled force and frequency (Figure S2) and generate regular output voltages peaks. Figure 1d shows a typically voltage output profile used in the study with a peak-to-peak voltage of 3.99 V at a frequency of 1 Hz. This voltage was directly applied to the interdigitated electrodes, inducing an oscillating EF at the substrate surface with specific frequencies and amplitudes.

As shown in Figure 2a, specific areas were delineated on the electrode substrate surface to identify the influences from the NG-induced oscillating EF. The area in between the electrode fingers was labelled as “Middle” (M), while the regions on the substrate without electrodes buried underneath (i.e. outside the influence of the EF) was labelled as “Outside” (O). To better understand the spatial distribution of the EF, finite element simulation was used to model the EF distribution inside and outside of the electrodes covered area. As shown in Figure 2b, the EF had a uniform distribution in the area in between the electrodes, which also had the highest strength. In the region outside the electrode zone, the EF strength decreases monotonically, and reached only 10% of the strength in the middle of area at a distance of 4 mm away from the electrodes. Therefore, analysing the cell coverage results could directly reflect the EF strength influences.

figure 2.

cell staining and imaging for data collection and coverage analysis. (a) image of interdigitated electrode detailing location of imaging. (b) finite element simulation showing the ef distribution on and in vicinity of the electrode pair. (c) images of giemsa stained movas cells taken from different areas on the electrode substrates when the ng-induced ef was present (top) and without the ef (bottom). (d) graph comparing the cell coverage percentages at different areas between stained images and showing the cell coverage expressed as percentage of the total area analyzed on the slide.

This electrode substrate was connected to the actuated NG and placed in a 6 cm cell culture dish and 1.0 ×10⁶ MOVAS cells suspended in 5mL of culture media were added on top. The same cell culture conditions were used without connecting to the NG and was therefore in the absence of EF to serve as a control. After the 24 h incubation period, the cells on the surface of the substrates were stained for imaging. As the cells on the surface were distributed nearly as a monolayer, we used the coverage by the stained cells to represent the number of cells growing on the substrate (Figure S3). Representative stained cell images at different locations from both experiment and control substrates are shown in Figure 2c. In the middle area, where the EF was strongest, there was a noticeable reduction in cell coverage compared to the same area on the control substrate without the EF influence. This suggests that the EF effectively inhibited cell adhesion and proliferation in this central region. Moreover, as we moved away from the center towards the edge of the electrode covered area (O1) and further to 0.40 cm (O2) and 0.20 cm (O3) away from the electrode edge, the cell coverage increased significantly. This gradient of cell density increase exhibited a good accordance to the reduction of EF strength from the center to the outside of the electrode-covered area. Image analyses quantified the average cell coverage at each area (Figure 2d). It clearly showed that all the four selected areas in the control substrate had insignificant variations in cell coverage. The slight cell coverage gradients were likely due to cell crowding and contact inhibition from the middle area. These results suggest that in the absence of EF, the cell growth was not affected or decreased. It appeared that cells are uniformly distributed without EF. In general, all the four selected areas in the experiment substrate had a significantly lower cell coverage. Particularly, we could observe a sharp decrease from 72.69% to 45.48% when comparing the middle area of the control slides and the ones exposed to EF after 24h incubation. Along the edge of electrodes (O1), a decrease of 32.61% was observed between the control and experiment substrates, which was similar to the O2 area (32.05% decrease). The cell coverage difference was reduced to 36.52% on the edge area (O3), where the EF strength was the weakest. The increase of cell coverage indicates a diminishing influence from the EF as moving away from the electrodes. These observations corroborated our hypothesis that localized oscillating EFs can modulate cell attachment and prevent excessive smooth muscle cell proliferation.

pH examination of both control and EF groups showed that the pH values of the cultural media were not influenced by the application of EF (Figure. S4), and therefore the cell proliferation was not caused by pH change. We further investigated the influence of the applied voltage amplitude on MOVAS surface coverage. The applied voltage was adjusted by switching the load resistor in the circuit, where the higher resistance yielded a higher voltage applied to the cell culture. Figure 3a shows the measured voltage profile with the peak-to-peak voltages of 1.73V, 3.19V and 3.99V while maintaining a constant frequency of 1Hz. Stained images revealed that all the substrates at different voltage showed a similar variation in cell density change following the EF strength distribution (Figure S5, S6, S7). To quantify this location dependence, the cell coverage at four selected distances from the center of the electrodes were quantified by image analyses (Figure 3b). Similarly, all the substrates with applied EF had a lower cell coverage compared to the control substrates. A clear increasing trend in cell coverage could be observed as the location moved from the middle to the outer regions of the substrate, where the influence of the EF was weaker. This gradient in cell distribution highlights the spatial specificity of the electric field’s inhibitory effects on MOVAS proliferation under all different voltages. This location-dependent effect further supports the hypothesis that the EF has a localized impact on cell adhesion that is dependent on its strength.

figure 3.

cell coverage and applied voltage relationship (a) voltage output measured from the ng with 1v, 3v and 4v for experimental cell culture conditions. (b) corresponding cell coverage analysis of movas cultured on electrodes with no applied voltage (control). (c) cell coverage analysis of movas cultured on electrodes with 1v, 3v and 4v of applied voltage to test the ef influences. (d) comparison of cell area coverage (normalized by the control) at the middle area under different applied voltage.

Considering the middle area has the most representative cell inhibitory effect under EF influences, the cell coverages in this area under different voltages were collected and compared together with the control (Figure 3c). By normalizing the coverage on the control substrate to 100%, it could be clearly observed that higher voltages resulted in more significant reductions in cell coverage. Under 1V, the relative cell coverage dropped to 78.58%. As the voltage increased to 3 V and 4 V, the relevant cell coverages further reduced to 65.50% and 62.56%, respectively. There seemed to be a saturation effect at a relatively high voltage of 3V and 4V, the relevant cell coverages further reduced to 65.50% and 62.56%, respectively. There seemed to be a saturation effect at a relatively high voltage of 3V and 4V, indicating that further increasing the voltage may not introduce a greater influence in further decreasing the cell coverage.

The relationship between oscillating frequency and cell coverage was further explored using the same experimental setup. Figure 4a shows the voltage output profiles that were used in the frequency-relationship test, where 0.5 Hz, 1 Hz and ~7 Hz were applied to cover the normal frequency range that could be accessible from human body. To maximize the EF effect, the voltage outputs were set at the saturation value near 4V. It should be noted that the NG output intensity was directly related to the driving frequency. Therefore, a much higher voltage amplitude ~21.60 was obtained at the highest frequency of 7 Hz. As these amplitudes was beyond the saturation voltage, we assume the amplitude contribution would be minimal. Image analysis of the cell coverage on different electrode-covered/uncovered areas for the control and experimental groups is illustrated in Figure 4b and Figure 4c, respectively. The quantified graphs in Figure 4c demonstrate a pronounced trend where escalating the frequency of the applied EF corresponds to a substantial diminution in cell coverage on the substrate surface. At the lower frequencies of 0.5Hz and 1Hz, the relative cell coverage was observed to decline to 39.61% and 40.66%, respectively. This reduction became more pronounced as the frequency was increased to 7Hz, where cell coverage is further diminished to 28.76%. This inverse correlation between frequency and cell coverage suggests that higher frequencies of EF would exert a more potent inhibitory effect on cell adhesion and proliferation. From the direct comparison of the middle area between all conditions (Figure 4d), the frequency-dependent inhibitory effect could be clearly observed. As the frequency increased from 0.5 Hz to 1 Hz, the average cell coverage difference was minimal (1.76%), which fell within the range of error, indicating that a small change of frequency of 0.5 Hz might not introduce significant EF influences. In contrast, when the frequency was increased to 7 Hz, an appreciable decrease to 46.12% cell coverage was observed comparing to the control groups. This is a 14.68% and 16.44% decrease from the 0.5 Hz and 1 Hz conditions, respectively. This result suggests that higher frequencies disrupt the cell’s ability to adhere and proliferate on the substrate more effectively than lower frequencies.

figure 4.

cell coverage and frequency relationship. (a) voltage output measured from the ng at 0.5hz, 1hz and 7hz for experimental cell culture conditions. (b) corresponding cell coverage analysis of movas cultured on electrodes with no applied voltage (control). (c) cell coverage analysis of movas cultured under 0.5hz, 1hz and 7hz stimulations to test the frequency influences. (d) comparison of cell area coverage (normalized by the control) at the middle area under different applied frequencies.

Based on the fundamental of cell adhesion principle, we suggest the anti-adhesion effect could be attribute to the surface charge disturbance due to the presence of oscillating EF. It is known that MOVAS cell membranes are charged negatively. When adhere to a substrate surface, it prefers positive surface charge to form initial bonding through electrostatic interaction. As schematically illustrated in Figure 5a, when a voltage is applied to the two electrodes by a NG, an EF is induced at the area in between. This EF drives the redistribution of surface charge in the electrical double layer, i.e. positive and negative ions move toward the negative and positive electrodes, respectively. This surface charge redistribution can lead to opposite surface-cell membrane interactions. At the positive charge accumulation area, the negatively charged cell membrane experiences enhanced attraction force to the substrate; while at the negative charge accumulation area, the cell membrane experiences repulsive force instead. As the charges reverse due to EF oscillation, the electrostatic force experienced by the cell membrane reverses accordingly. This constant switching of attractive and repulsive force prevents a stable interaction of cell membrane with the substrate. Therefore, it inhibits stable adhesion and proliferation of the MOVAS on the EF-covered surfaces.

figure 5.

proposed mechanism underlying the inhibitory effect of oscillating efs on cell adhesion control. (a) schematic showing oscillating electric field movas repelling mechanism via disturbance of electric double layer in liquid media. (b) representative images of (i) the control group, cultured cells over exposed electrodes (no pdms coverage) in control conditions or (ii) the experimental group, cells cultured over exposed electrodes and subjected to current exposure, and (c) corresponding cell coverage analysis.

It should also be noted that similar EF signals have been used for promoting cell migration and proliferation, for example for wound healing applications.²⁶ This effect may seem counterintuitive to our discovery in this work. We highlight that the key difference lies in the specific parameters and configuration of the EF. Our inhibitory phenomenon relies on the EF in between a pair of electrodes that are insulated from the cell environment. There was no electric current flow through the cell environment. On the contrary, the setup for promoting cell proliferation requires an electric current flow through the cell culture media to stimulate cell activities.^{20, 27–29} To further prove the difference between these two effects, we conducted a set of experiments to compare the distinguishing influences from electric potential and electric current. Using the same cell culture setup, we utilized electrodes without the PDMS insulating coating. As such, the electrodes were completely exposed to the cell culture media, where electric current could be induced through the media when an EF was introduced to the electrode. As shown in Figure 5b, (ii) under the electric current stimulation, the cell coverage was obviously higher than that on the (i) control surface, where there was no EF applied. Quantitative analysis revealed that the cell coverage increased by 48.95% as compared to the control (Figure 5c). This result was opposite to the inhibitory results we observed from insulating electrodes, confirming the opposite contribution from current and potential stimulations. Therefore, it was proved that cells directly exposed to electric current were promoted to spread and propagation while cells under EF without current going through were inhibited to adhesion and proliferation. In addition, many studies have demonstrated that direct electric current stimulation promotes actin cytoskeleton reorganization and focal adhesion turnover, facilitating the fibroblasts migration and extracellular matrix allignments.^30–33 Some research also attributed the EF effects on cells to different membrane potentials.^{33, 34} These contributions are important to provide further quantitative understanding of the EF’s inhibitory effects we observed on SMCs. However, our current EF setup is limited to the observations of only the cells surfaces by a stereomicroscope. New setup that allows unblocked optical imaging, such as using transparent electrodes, would be recommended in future studies to advance the understanding and application of this intriguing effect.

Conclusion

In summary, we performed an in vitro study to investigate the influence of oscillating EF on the adhesion and proliferation of MOVAS cells. By applying an oscillating EF from an externally connected NG to the electrode pairs underneath the cell culture environment, we demonstrated a significant reduction of cell coverage of cells at the electrode-covered area where the EF remained a uniform distribution with the highest strength. The influences of applied voltage amplitude and frequency on MOVAS surface coverage were both investigated. It was discovered that higher voltages resulted in more significant reductions in cell coverage, which might saturate at ~4 V. Increasing the frequency of the applied EF was found correlating with reduction in cell coverage as well. The mechanism underlying this inhibitory effect is believed to involve the interference of the EF with the electrostatic interaction between the cell membrane and substrates. Oscillating EF likely disrupts the stable surface charge distribution, and thereby weakens the initial cell attachment that is necessary for cellular adhesion and proliferating. It should be noted that additional cellular-level quantifications, including This discovery could lead to the development of self-powered devices that repel cells without the need for drugs or surgical interventions, alleviating complications associated with restenosis. More broadly, by leveraging the ability of EFs to modulate the cell-substrate interaction, these effects may also be designed to prevent unwanted bacteria or other cells from adhesion, thereby enhancing the performance and longevity in diverse clinical settings.

Experimental Section

Au electrode substrate preparation:

To make working electrodes, 50nm of Au was evaporated over 5nm Ti sticky layer onto glass slides using stainless steel masks that resulted in 100μm wide interdigitated electrodes spaced 100 μm apart. Au wire contacts were attached to each side of the connected fingers using copper tape. These electrodes were then covered with a ~450 μm-thick layer of PDMS. The PDMS surface was treated by oxygen plasma using a PE-200 Oxygen Plasma Surface Treatment and Etching System (O₂ flow rate of 8 cc/min, t = 20 s, RF = 100 W) to turn the surface to hydrophilic for better cell adhesion.

Cell culture setup:

1.0 ×10⁶ Murine Vascular Smooth Muscle cells purchased from American Type Culture Collection (ATCC, CRL-2797) were cultured in 6-cm dishes (Corning 430166). Four dishes had the UV-sterilized experimental electrode substrates, where the EF was applied from an externally connected NG. The triboelectric NG device comprised three layers of materials, namely two electrodes (80 μm, 40×40 mm) and a Fluorinated Ethylene Propylene (FEP) film (50 μm). The copper electrode (on the stationary side) acted as the positive triboelectric material, and also served as the induction electrode. The FEP film was attached to the other copper electrode (on the moving side) as a negative triboelectric material. The FEP film could generate negative charge on the surface after contact with top copper electrodes due to different electron adsorption capacity. When the device was continuously driven by a motor, the two layers of material go through a process of continuous contact and separation, through which the device produced electrical pulses to the cell cultural system. The moving side of the NG was secured on the actuator’s arm that was placed 3cm away from a 3D printed base that held the stationary contact of the NG. The computer-controlled actuator cyclically moved the NG at a speed of 30, 60 or 420 RPM to control the frequency. The NG output was measured by a multimeter (DMM 6500, Keithley, internal resistance 10 MΩ) after related to the substrate electrodes. The output voltage was controlled by connecting a resistor in parallel with the electrode pair, where a larger resistance yielded a higher voltage amplitude on the electrodes. Resistors of a resistance of 330 kΩ, 680 kΩ and 1.5 MΩ were used, which achieved a 1V, 3V and 4V output respectively. Four dishes contained PDMS-coated electrode substrates but without the EF were used as controls. All the dishes were cultured in modified DMEM containing 4.5 g L−1 D-glucose (Thermo Scientific, 11965118) supplemented with 10% fetal bovine serum (FBS), 100 U mL−1 penicillin, and 100 U mL−1 streptomycin and incubated for a total of 24h. After treatment all plates were stained with GIEMSA stain to observe the internal structure of the cells for ease of localization and quantification.

Cell staining protocol:

The plates were removed from the incubator and the culture media was removed and the surface of the substrates were washed with PBS. The cells were then fixed with 4% paraformaldehyde for 15 min at room temperature and rinsed again with PBS. Afterwards, 5ml of GIEMSA stain (Sigma-Aldrich GS500–500ml) was applied over each slide for 10 minutes. The slides were then washed with PBS and dried in an oven at 55°C. Stained slides were stored at 4°C until time of imaging.

Optical Microscopy imaging:

All slides were categorized into three sections and imaged using a Nikon ECLIPSE Ti optical microscope. One section showcases three images covering the top area, another section displays three images occupying the bottom area, and the final section features an image covering the middle area with the electrodes. Each of these images was captured by the microscope at a resolution of 1280 × 1024 pixels.

pH measurements:

Each dish was filled with 5 mL of high-glucose Dulbecco’s Modified Eagle Medium (DMEM, Thermo Fisher) without cells. In the EF group, the electrodes were connected to an electric circuit to generate an electric field, while the control group was left unconnected. All dishes were incubated at 37°C with 5% CO₂ for 24 hours. pH measurements were taken at baseline and after 24 hours of incubation. Baseline pH was measured using a 5 mL aliquot of high-glucose DMEM collected at the start of the experiment. After 24 hours, 5 mL samples were collected from each dish for pH measurement. All measurements were performed using a pH/mV meter (FiveEasy Plus FP20, Mettler Toledo) in accordance with the manufacturer’s instructions. The results were analyzed using a Two-Way Analysis of Variance (ANOVA) model with multiple comparisons performed in PRISM software (GraphPad Software).

Image processing and analysis:

The pixels in the capture images were utilized as coordinates to accurately locate the section under analysis (Figure S8). Frames were chosen according to their position on the x and y axes, as well as their width and height dimensions. The frames used were 600 × 420 for areas without electrodes, while 200 × 600 and 125 × 960 were designated for rectangles that avoid electrode shadows. ImageJ macro was used after identifying the desired frame, to execute the following series of commands: crop the image, convert it to 16-bit RGB [Image > Type > RGB Stack, 16-Bit], and launch the threshold adjustment to be manually manipulated until obtaining the most accurate percentage of cell area coverage. The threshold adjustment consists of two intensity values that convert everything below into white and everything above into black, allowing a red surface over the desired area. Once the surface covering the MOVAS cells is selected, the area coverage is measured obtaining, the mean, and the % Area value.

NG Output measurement:

To measure the output voltage from the NG, both sides of the NG electrodes were taped with Cu foils as electrodes. Under continuous compressive force, the output was collected by a multimeter (DMM 6500, Keithley, internal resistance 10 MΩ) and then plotted using Origin software.

Scanning Electron Microscopy imaging:

The size and dimensions of the evaporated electrodes was confirmed through SEM imaging using a Zeiss LEO 1550VP SEM at a 3.00kV accelerating voltage and a working distance of 8mm.

The main text of the article should appear here with headings as appropriate.

Data availability

The data supporting the findings of this study are available within the Article and its Supplementary Information, as well as from the corresponding author upon reasonable request.