Electrical stimulation (ES) has been recognized to play important roles in regulating cell behaviors. A microfluidic device was developed for the electrical stimulation of single cells and simultaneous recording of extracellular field potential (EFP). Each single cell was trapped onto an electrode surface by a constriction channel for ES testing and was then driven to the outlet by the pressure afterward. This design allows the application of ES on and detection of EFP of single cells continuously in a microfluidic channel. Human cardiomyocytes and primary rat cortex neurons were tested with specific ES with the device. Each cell's EFP signal was detected and analyzed during the ES process. Results have shown that after applying specific ES on the excitable single cells, the cells evoked electrical responses. In addition, increased secretion of glutamic acid was detected from the stimulated neurons. Altogether, these results indicated that the developed device can be used to continuously apply ES on and accurately determine cell responses of single cells with shorter probing time. The throughput of the measurement can achieve 1 cell per minute, which is higher than the traditional ES methods that need culturing cells or manually positioning the cells onto the electrode surface. Before and after the application of ES, the cell viability had no significant change. Such a device can be used to study the biological process of various types of cells under electrical stimulation.

Results from recent studies of the influence of electrical stimulation (ES) on regulating cell behaviors have contributed to many biomedical research progresses. For example, the obtained knowledge has improved the understanding of ES based clinical treatments and helped the development of a variety of electrobioreactor for tissue-engineering application.1–3 ES has been exerted on different kinds of stem cells, such as cardiac stem cells, embryonic stem cells, human MSCs (mesenchymal stem cells), and iPSCs (induced pluripotent stem cells), which showed the ability of ES to direct the differentiation of various stem cells.4,5 Electrical stimulation has also been used in neurons for decades for modulating the activity of individual neurons within a neural network.6 The extracellular electrical recording method enables the simultaneous stimulation and recording of excitable cells with large populations for days or months without causing mechanical damage to the neuron's plasma membrane.7–9 

There has been an urgent need to investigate single cell electrical stimulation and responses. Important results have been revealed during recent years using bulk cell ES. For example, many studies have demonstrated that an extracellularly applied electrical field could significantly enhance neurite outgrowth10 and improve nerve regeneration.11,12 In addition, ES has been found to effectively promote the functional maturation of cardiomyocytes derived from stem cells.13,14 However, due to the variances of ES applied to each cell or heterogeneity in the cell population, it is very challenging to clarify the relationship between defined electrical stimulation signals and specific cell responses using current bulk measurement approaches. Such a relationship is critical for exploiting ES as a powerful tool to manipulate cells. Recently, microfluidic methods have been studied for cell electrical stimulation, and multiple devices were invented for this purpose. These microfluidic methods often utilize microelectrode array for both electrical stimulation and response recording.15–19 However, in order to position cell or cell clusters onto the electrode surface, these devices required the manual positioning of cells onto a micropiette,16 stopping the flow when a cell was detected,18,19 or using a barrier dam to trap cells.17 These positioning methods have low efficiency, low accuracy, or low throughput. For instance, while the device reported by Klauke et al.17 can stop a single cell on the microelectrode surface embedded in a cavity, the cell was needed to be taken out from the cavity after ES and signal recording in order to analyze the next cell, which was a lengthy and low throughput. Recently, we developed a device with microfluidic methods that can apply electrical stimulation on cell clusters suspended in microfluidic channels and detect their electrical responses.20 While the device does not need manually positioning of the cells onto the electrode surface, the device is only suitable for the analysis of cell clusters.

Here, we demonstrated a microfluidic device with the ability to apply ES signals on single cells and record the extracellular electric signals. Each single cell can be easily positioned onto the measurement electrode surface with a constriction channel for ES application and simultaneous extracellular field potential (EFP) measurement. After ES, the cell can be driven by the pressure to move toward the outlet and be collected; the next cell is ready to be stimulated and recorded in a continuous flow. As a result, the device can continuously exert ES on single cells and record the EFP with a higher throughput than the traditional ES methods that need culturing cells or manually positioning the cells onto the electrode surface.10–17 Excitable cells, including human cardiac myocytes (hCMs) and primary rat cortex neurons (rCNs), were used to demonstrate the utility of this device.

Human cardiac myocytes (hCMs), myocyte growth medium kit, cryo-SFM, and detach kit were obtained from PromoCell GmbH (Heidelberg, Germany). Primary rat cortex neurons (rCNs), B-27TM plus neuronal culture system, LIVE/DEAD viability/cytotoxicity kit, and AmplexTM red glutamic acid/glutamate oxidase assay kit were purchased from ThermoFisher Scientific (Walkersville, MD, USA). Trypsin/EDTA and antibiotic−antimycotic solution were obtained from Gibco (Carlsbad, CA, USA). Penicillin-streptomycin was obtained from Sigma-Aldrich (St. Louis, MI). Poly-d-lysine coated 12-well plate was purchased from Greiner Bio-One (Monroe, NC, USA). Cell scraper and cell culture grade 1× phosphate buffered saline solution were obtained from Corning (Manassas, VA, USA). All the materials were used as received from the manufacturers.

The design of the microfluidic device for single cell electrical stimulation is shown in Fig. 1. The device has 3 main fluidic channels; each has dimensions of 500 μm in width and 50 μm in height. A constriction microchannel was fabricated between the main channels in order to trap and locate the single cell on top of a measurement electrode. The dimensions of the constriction channel were 10 μm in height, 8 μm in width, and 15 μm in length. A pair of 200 μm wide stimulation electrodes was fabricated to apply ES signals. Another pair of electrodes with a 50 μm width was located on both sides of the constriction channel and worked as measurement electrodes to record the EFP. 3 reservoirs with 1.5 mm diameter access holes were fabricated along with the channels: an inlet reservoir (Inlet) for loading the floating cells, an outlet reservoir (Outlet2) to collect the cells after ES, and another outlet reservoir (Outlet1) for collecting the fluidic medium.

FIG. 1.

Schematic of the microfluidic device for single cell ES and FP recording. (a) Illustration of the device design for single cell electrical stimulation and FP measurement. (b) Test procedures for ES and measurement: (i) loading single cells from the inlet and trapping it on top of the measurement electrodes, (ii) applying electrical stimulation on and measuring the response of each single cell via the measurement electrodes. (iii) After ES measurement, applying pressure on Outlet1 to push the cell to and collect the cell at Outlet2. (c) Picture of the device. (d) Microscopic photo of the measurement region, including the measurement electrodes and the constriction microchannel. (e) Microscopic images of the cardiomyocyte trapped by the constriction channel (measurement electrodes were intentionally removed to take clear images of cell trapping).

FIG. 1.

Schematic of the microfluidic device for single cell ES and FP recording. (a) Illustration of the device design for single cell electrical stimulation and FP measurement. (b) Test procedures for ES and measurement: (i) loading single cells from the inlet and trapping it on top of the measurement electrodes, (ii) applying electrical stimulation on and measuring the response of each single cell via the measurement electrodes. (iii) After ES measurement, applying pressure on Outlet1 to push the cell to and collect the cell at Outlet2. (c) Picture of the device. (d) Microscopic photo of the measurement region, including the measurement electrodes and the constriction microchannel. (e) Microscopic images of the cardiomyocyte trapped by the constriction channel (measurement electrodes were intentionally removed to take clear images of cell trapping).

Close modal

The microfluidic device was fabricated with the standard soft lithography process. A two-layer master mold with the structure of microchannels and reservoirs made of SU8 (5 & 2025, MicroChem) was created using a two-step photolithography method described by Han et al.21 The fabrication process is as follows: First, a 10 μm thick SU8-5 layer was spin coated on a silicon wafer. A photolithograph step was then utilized to define the pattern of the constriction channel. Next, a second layer of SU8-2025 (40 μm thick) was spin coated on top of the first layer and it underwent a photolithograph step to obtain the wide channel and reservoir structures. Subsequently, polydimethylsiloxane (PDMS) was used to form the microchannel with the mater mold. The microelectrodes were fabricated with the 100 nm/10 nm gold/titanium coating on a glass slide (1 × 3 in.2, EMF Corporation). Photolithography and wet etching with KI/I2 gold etchant (Sigma-Aldrich) were used to fabricate electrodes. Finally, an oxygen plasma treatment (200 mTorr, 50 W, 50 s) was used to activate the surfaces of PDMS and the glass. After the treatment, the PDMS structure was bonded to the glass slide. A surface profilometer (Dektak 150, Veeco Instrument) was used to measure the dimensions of the constriction microchannel. The dimensions were 14.67 ± 1.34 μm (length), 9.23 ± 0.95 μm (depth), and 6.87 ± 0.74 μm (width).

Note that the electrodes were platinized to reduce their impedance before bonding the glass slide with the PDMS microchannels. Platinization is the process of depositing a layer of platinum black on the gold electrodes. The electrodeposition of platinum black usually results in a rough surface and largely increases the effective area of electrodes. Before the platinization, the electrode surface was rinsed by methanol. Electroplating of platinum black on the gold electrodes was conducted in the platinizing solution (YSI 3140 Platinizing Solution, water solution of 5% chloroplatinic acid and 0.1% lead acetate). A 5 V DC voltage was applied across the two measurement electrodes for 10 s. Then, a negative DC voltage (−5 V) was applied on the electrodes for another 10 s. After the platinization, we scanned the impedance spectrum of the measurement electrodes using a Gamry Reference 600 potentiostat (Gamry Instruments, Warminster, PA, USA). With the myocyte growth medium, we measured the impedance of the electrodes as a function of AC frequency. Results showed that the electrode impedance was largely reduced at lower frequencies (<1 kHz) in which the dominant frequencies of the FP signals fall. The measurement result and other detailed information are shown in Fig. S2 in the supplementary material. Electrodeposition was conducted before we bonded the glass slide with the PDMS microchannels. Because of the small thickness of the gold and black Pt (∼0.1 μm), the glass slide and the PDMS channel bonded well; the microchannel remained sealed. No leakage was found in our tests.

hCMs were cultured following the manufacturer's instruction using the myocyte growth medium kit. rCNs were cultured using the B-27TM plus neuronal culture system containing 1% penicillin-streptomycin solution. Both cells were maintained in cell culture incubator at the temperature of 37 °C with 5% CO2. A medium change was performed every other day for the hCMs and every day for the rCNs. The neurons (rCNs) were cultured in vitro for 1 day prior to the testing. Because of the short culture time, we did not observe any formation of cell-cell connection and neurites in the cultured neurons. hCMs at passages 6–9 and rCNs at passage 1 were used for all the testing.

Cells suspended in the medium were introduced into the device via the inlet reservoir at a concentration of 20–50 cells per μl. We used a flow controller (Flow-EZ, Fluigent, France) to apply a constant pressure of 3 kPa through the inlet reservoir to control the cell flow. The single cell was trapped and located on top of the left measurement electrode by the constriction microchannel. Single cardiomyocytes and neurons with diameters from 10 μm to 50 μm should be captured on the electrode surface. This was confirmed by experimental observations as shown in Fig. 1(e). The constriction dimension can be modified to position single cells with diameters out of this size range.

The flow was approximately stopped once a single cell was trapped by the constriction channel and blocked the constriction channel. This can be confirmed by observing a flow rate reduction by the flow controller and also by optical observation from the microscope, the applied pressure was removed by setting the pressure to zero. With the low cell concentration (about 20 cells/μl), only one cell could be trapped and positioned on the surface of the measurement electrodes. Then, we applied the ES signals on cells through a pair of wide electrodes (200 μm wide) and measured the extracellular field potential signals from the trapped cell through a pair of measurement electrodes (50 μm wide). The extracellular field potentials (EFP) reflect the electrical activity of cardiac and nervous cells or tissues based on the transmembrane currents in the extracellular medium. When a cell is electrically stimulated and invokes changes of transmembrane ion pumps and voltage-gated channels, a transient imbalance in ion concentration forms across the cell membrane. The caused transient alteration of the transmembrane currents will induce response EFP signals and could be quantitatively recorded by this device to reflect the cell activity. Note that all cells loaded into the channel traveled through the stimulation region (channel region between the pair of stimulation electrodes) and were electrically stimulated. Because of the use of the current pulses as ES signals, all cells presented in the stimulation region experienced the same electrical stimulations.

The device was designed to detect the EFP signals of the single cell trapped by the construction channel and located on top of a measurement electrode. The FP magnitude is inversely proportional to distances between each cell and measurement electrodes. Hence, the FP magnitude from the cells that were away from the measurement electrode was too weak to detect. By controlling the low concentration of the cell (about 20 cells/μl), the chance that multiple cells situating on or very close to the measurement electrode is small. From the microscope observations during the tests, we typically saw that only one cell was trapped and situated on the measurement electrode. This can be confirmed by a microscope image, as shown in Fig. 1(e). As a result, only the FP signals from the cell tapped on top of the electrode were detected.

While a cell may have experienced some stress and deformation due to trapping, the trapping was unlikely to cause the cell to generate detectable electrical activities. This was confirmed by the following test: we trapped one cell outside the constriction channel and let it situate on the measurement electrode for 1 min without applying ES. Then, we measured the electric signals from the cells. We did not observe any detectable FP signals from the trapped cell without electrical stimulations. Subsequently, we applied ES (0.4 mA, 1 Hz, 0.5 ms pulse width) to the same cell for 1 min. With ES, we measured a clear FP signal from the cell (nearly identical to the FP responses shown in Fig. 2). 10 different single cells were trapped and measured separately (with and without applying ES). No FP signal was detected when the cells were not electrically stimulated, while we did detect FP signals when they were stimulated. This test proved that the cell trapping did not cause a noticeable change in the cells' electrical activity.

FIG. 2.

Detected field potential signals from human cardiac myocytes electrically stimulated in the device. (a) hCM responses under different ES amplitudes (0.2/0.4/0.6 mA, 0.5 Hz pulses with a 0.5 ms pulse width. (b) Microscopy phase contrast images of single hCMs in suspension (Scale: 50 μm), taken before cells were loaded into the device. (c) Magnitude and occurrence of hCM responses under different ES frequencies.

FIG. 2.

Detected field potential signals from human cardiac myocytes electrically stimulated in the device. (a) hCM responses under different ES amplitudes (0.2/0.4/0.6 mA, 0.5 Hz pulses with a 0.5 ms pulse width. (b) Microscopy phase contrast images of single hCMs in suspension (Scale: 50 μm), taken before cells were loaded into the device. (c) Magnitude and occurrence of hCM responses under different ES frequencies.

Close modal

The entire system was settled in a Faraday cage to reduce the electromagnetic interferences/noises. Details of the Faraday cage can be found in Ref. 20. During the test, we used an upright microscope (PSM 1000, Motic) equipped with a video camera (QICAM 12 bit, QIMAGING) to observe the cell positioning/trapping. To apply the ES signals applied on cells, we used a waveform generator (33600A, KEYSIGHT) and input circuits to generate the desired signals and exert the signals via the stimulation electrodes. We then recorded and digitized the EFP signals from cells through a data acquisition board (PCI-6133, National Instrument, USA). A custom LabVIEW program was set for the data acquisition at a sampling rate of 1.5 MHz. The circuits used to generate the current pulses and capture the response signals were the same as the circuits described in our prior ES study on cell clusters.20 Here, we used current-controlled signals with a magnitude of 0.1–0.8 mA, similar to those used in our prior study,20 which resulted in a 0.3–15 V/cm electrical field. The frequency, pulse width, and duration were from 0.5 to 2 Hz, 0.5 ms, and 1 min, respectively. The ES signals used in this work were comparable to the stimulation signals used by other researchers for studying cell electrical stimulation behaviors, which were selected to ensure that the electrical current (1) will not cause cell hydrolysis and (2) can trigger the cell response.19,23 The equivalent electrical fields used in the tests were also comparable to the electroporation field.22 

The setting of ES is based on the research findings from our previous paper20 and other studies.4,5,18,19,21,24 The voltage and current electric pulses have been the two most common ways to deliver ES to cells or tissues. Publications have shown that current-controlled stimulation has advantages over the voltage-controlled approach.22,24 For the voltage-controlled approach, the stimulation voltage applied on individual cells may vary depending on their location and the impedance between the cell and the electrode. However, current pulses can ensure constant stimulation signals applied on all the cells. Therefore, for our studies, we chose to use current pulses to deliver ES to cells. In our previous study to apply ES to cell clusters made of cardiomyocytes, we conducted extensive preliminary studies to optimize the parameters of the current pulses to trigger the electrical activity of the stimulated cells. We found that the above setting of ES that will not cause significant cell death but can effectively trigger cell electrical response. We also found that similar ES settings had also been used by other groups to stimulate cardiomyocytes and neurons.18,19,22 Taken together, we selected the mentioned ES setting for this study.

Both hardware and software methods were used to remove noises and stimulus artifacts that resulted from the input ES signals from the detected signals. Note that the stimulation signals also induced artifact signals on the measurement electrodes. Because the induced artifacts signals had much larger peak magnitudes than the FP signal from a single cell and occurred regularly with the same frequency as the stimulation signals, the stimulus artifacts were filtered out from FP signals using filter circuits and Matlab algorithm. Details of the measurement circuits for data acquisition and noise filtering can be found in Ref. 20. Details of artifacts elimination are provided in the supplementary material.

Tests of cell viabilities of primary rCNs before and after electrical stimulation were performed with the LIVE/DEAD viability/cytotoxicity kit (Thermo Fisher Scientific, Waltham, MA, USA) using a microplate protocol according to the manufacturer's recommendation. First, we seeded rCNs on poly-d-lysine coated 12-well plate. The cells were incubated in the humidified cell culture incubator (37 °C, 5% CO2) for 48 h. After the incubation, we stained the adherent neurons with calcein AM (1 μM) for live cells and ethidium homodimer (4 μM) for dead cells, followed by harvesting the cells for experiments. The cells were divided into two experimental groups before being loaded into the device. We applied ES to one group, while no ES signals were applied to the other group. ESs used for viability test were 2 Hz pulses with 0.4 mA in magnitude, 0.5 ms in width, and 60 s in duration. Cells before being introduced into the device were tested and used as control groups for each experiment. The fluorescence intensity was measured at excitation/emission of 495/645 nm and 495/530 nm using a Synergy H1 hybrid microplate reader (Bio-Tek Instruments, Winooski, VT, USA). All the cells loaded into the device were collected and tested for cell viability.

The glutamic acid concentration of rCN before and after the electrical stimulation was measured by the AmplexTM red glutamic acid/glutamate oxidase assay kit (Molecular Probes, Eugene, OR, USA). Briefly, the rCNs were detached from the culture plate, passed through the microfluidic device, and were applied the ES (0.1–0.4 mA; 0.5–2 Hz; 0.5 ms pulse width; 60 s period). For the no ES control group, the cells were passed through the microchannel and were collected without applying ES. For both experiment groups, the collected cells were then centrifuged at 100×g for 3 min at 4 °C to collect the supernatant for measurement. After staining the samples with the AmplexTM red glutamic acid/glutamate oxidase assay kit, the fluorescence intensity was measured at excitation/emission of 545/590 nm using a Synergy H1 hybrid microplate reader (Bio-Tek Instruments, Winooski, VT, USA).

The measurement of the glutamic acid concentration change was to prove that the electrical stimulation did trigger electrical activities of excitable neuron cells indicated by the secretion of glutamic acid; thus, the measured FP signals were generated by the stimulated cells rather than artifacts or crosstalk signals.

All quantified results are demonstrated in means ± standard deviation of 3–6 independent samples. Graphpad Prism 5 software (La Jolla, CA, USA) was used for statistical analysis of the results. We performed student's t-test to compare significant differences between experimental groups. We also conducted analysis of variance (ANOVA) and Tukey's posthoc test to compare the differences among all experiment groups. A p-value of less than 0.05 was considered statistically significant.

Human cardiac myocytes (hCMs) were introduced into the microfluidic device and positioned onto the measurement electrodes one by one. Electrical stimulations with different conditions were applied on single hCMs. Extracellular field potential of the cell was recorded at the same time (Fig. 2). Different amplitudes of ES were tested from 0.2 mA to 0.6 mA, as shown in Fig. 2(a). Without stimulation or with stimulation below 0.4 mA, the range of the detected signal was found to be between −5 μV and 5 μV, which was within the noise band from the measurement circuit. When the ES amplitude was above 0.4 mA, the field potential responses were detected by the device. When a single cell was evoked, the detected extracellular potential underwent a rapid drop forming a sharp negative pulse. The potential then raised above 0 V, resulting in a positive pulse. Finally, the potential returned to the baseline of 0 V. The FP signals were clearly identified with waveform patterns. The peak value of the pulses ranged from –60 to 50 μV. The magnitude and occurrence of field potentials generated by hCMs in responses to ES with different magnitudes were recorded and compared in Fig. 2(c). The magnitude of the hCM's FP signals was measured as peak to peak value of each FP signal. Average magnitude was calculated from all FP pulses during a 20 s ES period. The occurrence of the cell responses was defined to represent the average amount of identified FP pulse signals per second in 20 s. The results showed that the hCMs were electrically activated when the ES was stronger than 0.4 mA. Continuing to increase the ES amplitude to 0.6 mA, the FP responses measured from electrodes did not change significantly. The results demonstrate that a magnitude threshold of ES is needed to sufficiently trigger the cell electrical activity. When the excitable cells are triggered for electrical activity, the voltage-gated ion channels embedded in a cell's plasma membrane will open and let the ion exchange happen, causing the detected FP signals. While applying higher ES magnitude (0.4–0.6 mA), the FP signals did not alter much, which was similar to the observation in our previous study.20 The possible cellular mechanism we suspect is that the voltage-gated ion channels do not have the function to further alter its ability for ion exchange after they are triggered by the electric current above the threshold.

To confirm the FP responses of hCMs and study the relation between various stimulation signals and single cell electrical activity, single hCMs introduced into the device were stimulated by ES signals (0.4 mA, 0.5 ms wide pulses with a duration of 60 s) at multiple frequencies (0.5 Hz, 1 Hz, and 2 Hz). The FP signals' magnitude and occurrence were measured and compared in Fig. 3. Results demonstrated that changing input ES signals' frequency from 0.5 Hz to 2 Hz, the cell FP signals' occurrence significantly increased from 0.27 to 1.06 pulse/s on average. These results were consistent with the results from our previous study of ES frequency effect on hCM clusters,20 as shown in Fig. 3(b), which means the effects of different ES frequencies on cell clusters also apply to single cells. The input ES signals' frequency played an important role in controlling hCMs' FP signals. Higher frequency ES seemed to have the ability to further increase the cell electrical activities of single hCMs. Similar observations of the ES frequency effects on cell electrical activities have been reported by other groups.19,22 The cellular mechanisms behind this phenomenon have not been fully elucidated. One possible mechanism we suspect is that a higher frequency of ES increased the occurrence of cell membrane alterations (e.g., open frequency of the voltage-gated ion channels), which may further increase the cell electrical activities. The results showed that the frequency of the ES might have the ability to regulate the level of triggered cells' electrical activities. This could be used to optimize the future bioreactor design and apply controlled ES on cardiac tissue construct in order for the desired electrical property.

FIG. 3.

Effect of stimulation frequencies on responses of single hCMs. (a) hCM responses under different ES frequencies (0.4 mA, 0.5/1/2 Hz pulses with a 0.5 ms pulse width). (b) Magnitude and occurrence of hCM cluster responses under different ES frequencies.20 (c) Magnitude and occurrence of single hCM responses under different ES frequencies.

FIG. 3.

Effect of stimulation frequencies on responses of single hCMs. (a) hCM responses under different ES frequencies (0.4 mA, 0.5/1/2 Hz pulses with a 0.5 ms pulse width). (b) Magnitude and occurrence of hCM cluster responses under different ES frequencies.20 (c) Magnitude and occurrence of single hCM responses under different ES frequencies.

Close modal

Neuron cell ES and FP measurement is a widely used tool to study the function and connectivity of neuronal circuits. To validate the ability to electrically stimulate neurons and detect FP signals of our device, single primary rat cortex neurons (rCNs) were introduced into the microfluidic device and electrically stimulated by input signals (0.4 mA, 0.5 ms wide pulses with a duration of 60 s). FP signals generated from the stimulated single neurons under ES of different frequencies (0.5 Hz, 1 Hz, and 2 Hz) were recorded (Fig. 4). Results also showed a trend of improving cell electrical responses under increasing frequencies of single neurons, similar to the hCMs. The occurrence of the rCN FP signals significantly increased from 0.3 to 1.2 pulse/s on average, when the frequency was increased from 0.5 Hz to 2 Hz. The average magnitude did not change significantly under different frequencies. These results indicate that the frequency of ES is a critical factor in triggering rCNs' FP signals. Higher frequency ES had the ability to further increase the cell electrical activities of single rCNs, which is consistent with the results from hCMs. These two types of cells showed similar trends of field potential signal change when ES conditions were varied.

FIG. 4.

Detected field potential signals from single primary rat cortex electrically stimulated in the device. (a) rCN responses under different ES frequencies (0.4 mA, 0.5/1/2 Hz pulses with a 0.5 ms pulse width). (b) Microscopy images of single rCNs (the scale bar is 50 μm). (c) Magnitude and occurrence of rCN responses under different ES frequencies. * indicates the statistically significant difference between the experiment groups with p-value less than 0.05.

FIG. 4.

Detected field potential signals from single primary rat cortex electrically stimulated in the device. (a) rCN responses under different ES frequencies (0.4 mA, 0.5/1/2 Hz pulses with a 0.5 ms pulse width). (b) Microscopy images of single rCNs (the scale bar is 50 μm). (c) Magnitude and occurrence of rCN responses under different ES frequencies. * indicates the statistically significant difference between the experiment groups with p-value less than 0.05.

Close modal

Note that the FP signals detected in this paper have similar shapes with FP signals measured by other researchers from cardiomyocytes and neurons.16,17,19,25 The signals are also similar to that of cardiomyocyte clusters in our prior study.20 After a cardiomyocyte or a neuron is electrically stimulated, the cell membrane potential rapidly rises to a peak potential, then drops and overshoots to a different polarity, and finally return to resting potential.26 

We conducted cell viability tests to investigate if the applied ES conditions could cause a significant amount of cells to die during experiments. Results shown in Fig. 5(c) demonstrated that optimized ES conditions would not cause a significant number of cells to die after the tests. Percentages of viable cells before and after ES were 84.66 ± 2.37% and 82.16 ± 4.70%, respectively. No statistically significant difference was found between the tested groups in student's t-test analysis. The average time consumed from harvesting the cells to completing the entire ES tests was ∼2 h, showing that this device has the feasibility to apply ES up to 2 h without compromising the cell viability. Thus, this device has the ability to electrically stimulate single cells while not affecting the viability of the cells.

FIG. 5.

Cell viability and secretion of glutamic acid by primary rat cortex neurons before and after the electrical stimulation (ES). (a) and (b) Fluorescence images of rCN stained with calcein AM (green) and ethidium homodimer (red). The scale bar is 50 μm. (c) Percentage of viable cells before and after ES. (d) Secretion of glutamic acid concentration by primary rat cortex neurons with and without the ES (0.4 mA, 0.5 Hz pulses with a 0.5 ms pulse width). (e) Secretion of glutamic acid concentration with a different ES magnitude (0.5 Hz). (f) Secretion of glutamic acid concentration with a different ES frequency (0.4 mA). * indicates the statistically significant difference between the tested experiment groups with p-value less than 0.05.

FIG. 5.

Cell viability and secretion of glutamic acid by primary rat cortex neurons before and after the electrical stimulation (ES). (a) and (b) Fluorescence images of rCN stained with calcein AM (green) and ethidium homodimer (red). The scale bar is 50 μm. (c) Percentage of viable cells before and after ES. (d) Secretion of glutamic acid concentration by primary rat cortex neurons with and without the ES (0.4 mA, 0.5 Hz pulses with a 0.5 ms pulse width). (e) Secretion of glutamic acid concentration with a different ES magnitude (0.5 Hz). (f) Secretion of glutamic acid concentration with a different ES frequency (0.4 mA). * indicates the statistically significant difference between the tested experiment groups with p-value less than 0.05.

Close modal

Analysis of the secretion of glutamic acid from single rCNs under different ES conditions was investigated to verify the ability to trigger electrical activities of excitable cells with ES within the device and also to validate the potential biological applications of this device. Glutamate is the most abundant neurotransmitter in brain and central nervous system, which is involved in all major excitatory brain functions. Many studies have used glutamate as a highly sensitive functional marker for excited neurons.15,27,28 After passing through the microfluidic device with different ES conditions, the cells were collected and centrifuged at 100 × g for 3 min at 4 °C. Then, the samples were stained by red glutamic acid/glutamate oxidase assay kit, and the fluorescence intensity was measured (see Glutamic acid measurement section). The glutamic acid concentration in the supernatant collected from each experiment group with different ES conditions is shown in Figs. 5(d)5(f). The results showed that the electrical stimulation significantly increased the glutamic acid concentration in electrically stimulated cells compared to the control group [Fig. 5(d)]. The concentration of glutamic acid with and without ES was 2.21 ± 0.17 μM and 0.56 ± 0.08 μM, respectively. Results showed that applying different ES magnitude did affect the glutamic acid concentration and high enough magnitude is necessary to achieve sufficient cell activity. It was also shown that the glutamic acid concentration of cells stimulated by different ES frequencies did not have a significant difference, which might be due to the sufficient glutamic acid release from the cells under ES. The neurons released a majority of the glutamic acid after ES in the device even under lower frequency and increasing the frequency would not further increase the glutamic acid concentration in the supernatant. It is widely understood and proved by many studies that changes in ES magnitude would alter the number of activated neurons in the stimulated tissue.27 However, the extent to which different stimulation frequencies and pulse durations affect local neuronal responses remains less explored. Our results indicated the dominant role the pulse magnitude plays as the ES parameter for neuron activation.

We developed a microfluidic device to achieve quick electrical stimulation on single cells in microchannels and simultaneous extracellular field potential recording. Cells can be collected after electrical stimulation for further cell analysis. The device is capable of locating single cells onto the measurement electrode surface, applying controlled ES and recording FP signals, and driving them to the outlet after the measurement. This scheme allows application of ES on many single cells in a row and measurements of cell responses in the continuous flow. Various ES conditions were applied on excitable hCMs and rCNs, and the evoked FP signals were recorded from the single cells with the device. To trigger sufficient cell electrical activities, a magnitude threshold of ES was identified during ES of hCMs. The results also demonstrated that the triggered electrical activities could be regulated with the ES frequency for single hCMs and rCNs. Furthermore, the secretion of glutamate by rCNs electrically stimulated in this device was investigated. The results show that ES with sufficient magnitude applied on neurons within this device can significantly increase the glutamate released from neurons, which indicate the effectiveness of the applied ES to trigger cell functions. This device can be potentially used in many biological applications related to electrical stimulation and will greatly improve the understanding of the relationship between specific ES conditions and triggered cell activities (e.g., cell proliferation, differentiation, etc.). The obtained knowledge plays a critical role for the study of biological mechanism of cell electrical activities.

See the supplementary material for the details of (1) stimulus artifacts elimination, (2) electrodes platinization, and (3) equivalent circuit models of cells.

This work was supported by the National Science Foundation (NSF) of the U.S.A. under Award Nos. ECCS-1625544 and ECCS 1905786. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

1.
A.
Kotwal
and
C. E.
Schmidt
,
Biomaterials
22
,
1055
(
2001
).
2.
M.
Zhao
,
H.
Bai
,
E.
Wang
,
J. V.
Forrester
, and
C. D.
McCaig
,
J. Cell Sci.
117
,
397
(
2003
).
3.
4.
Y. C.
Chan
,
S.
Ting
,
Y. K.
Lee
,
K. M.
Ng
,
J.
Zhang
,
Z.
Chen
,
C. W.
Siu
,
S. K. W.
Oh
, and
H. F.
Tse
,
J. Cardiovasc. Transl. Res.
6
,
989
(
2013
).
5.
D.
Hernández
,
R.
Millard
,
P.
Sivakumaran
,
R. C. B.
Wong
,
D. E.
Crombie
,
A. W.
Hewitt
,
H.
Liang
,
S. S. C.
Hung
,
A.
Pébay
,
R. K.
Shepherd
,
G. J.
Dusting
, and
S. Y.
Lim
,
Stem Cells Int.
2016
,
1718041
.
6.
M.
Hutzler
,
A.
Lambacher
,
B.
Eversmann
,
M.
Jenkner
,
R.
Thewes
, and
P.
Fromherz
,
J. Neurophysiol.
96
,
1638
(
2006
).
7.
D.
Eytan
and
S.
Marom
,
J. Neurosci.
26
,
8465
(
2006
).
8.
L.
Berdondini
,
K.
Imfeld
,
A.
MacCione
,
M.
Tedesco
,
S.
Neukom
,
M.
Koudelka-Hep
, and
S.
Martinoia
,
Lab Chip
9
,
2644
(
2009
).
9.
T. J.
Blanche
,
J. Neurophysiol.
93
,
2987
(
2005
).
10.
C. E.
Schmidt
,
V. R.
Shastri
,
J. P.
Vacanti
, and
R.
Langer
,
Proc. Natl. Acad. Sci. U.S.A.
94
, 8948 (
1997
).
11.
P.
Aebischer
,
R. F.
Valentini
,
P.
Dario
,
C.
Domenici
, and
P. M.
Galletti
,
Brain Res.
436
, 165 (
1987
).
12.
B. F.
Sisken
,
M.
Kanje
,
G.
Lundborg
,
E.
Herbst
, and
W.
Kurtz
,
Brain Res.
485
, 309 (
1989
).
13.
J.-L.
Ruan
,
N. L.
Tulloch
,
M. V.
Razumova
,
M.
Saiget
,
V.
Muskheli
,
L.
Pabon
,
H.
Reinecke
,
M.
Regnier
, and
C. E.
Murry
,
Circulation
134
,
1557
(
2016
).
14.
S.
Baumgartner
,
M.
Halbach
,
B.
Krausgrill
,
M.
Maass
,
S. P.
Srinivasan
,
R. G. A.
Sahito
,
G.
Peinkofer
,
F.
Nguemo
,
J.
Müller-Ehmsen
, and
J.
Hescheler
,
J. Cardiovasc. Pharmacol. Ther.
20
, 104 (
2015
).
15.
S.-Y.
Wu
,
H.-S.
Hou
,
Y.-S.
Sun
,
J.-Y.
Cheng
, and
K.-Y.
Lo
,
Biomicrofluidics
9
,
054120
(
2015
).
16.
W.
Cheng
,
N.
Klauke
,
H.
Sedgwick
,
G. L.
Smith
, and
J. M.
Cooper
,
Lab Chip
6
,
1424
(
2006
).
17.
N.
Klauke
,
G. L.
Smith
, and
J.
Cooper
,
Biophys. J.
91
,
2543
(
2006
).
18.
F. B.
Myers
,
O. J.
Abilez
,
C. K.
Zarins
, and
L. P.
Lee
, in
Proceedings of Annual International Conference on IEEE Engineering in Medicine and Biology Society EMBS
(
IEEE
,
2011
), pp.
4030
4033
.
19.
F. B.
Myers
,
C. K.
Zarins
,
O. J.
Abilez
, and
L. P.
Lee
,
Lab Chip
13
,
220
(
2013
).
20.
L.
Ni
,
P.
Kc
,
E.
Mulvany
,
G.
Zhang
, and
J.
Zhe
,
Biomed. Microdevices
21
, 20 (
2019
).
21.
Y.
Han
,
H.
Wu
,
F.
Liu
,
G.
Cheng
, and
J.
Zhe
,
Anal. Chem.
86
, 9717 (
2014
).
22.
N.
Klauke
,
G. L.
Smith
, and
J.
Cooper
,
Biophys. J.
85
, 1766 (
2003
).
23.
B. C.
Kim
,
S. M.
Jun
,
S. Y.
Kim
,
Y. D.
Kwon
,
S. C.
Choe
,
E. C.
Kim
,
J. H.
Lee
,
J.
Kim
,
J. K. F.
Suh
, and
Y. S.
Hwang
,
Biotechnol. Bioeng.
114
, 903 (
2017
).
24.
G. E.
Loeb
,
C. J.
Zamin
,
J. H.
Schulman
, and
P. R.
Troyk
,
Med. Biol. Eng. Comput.
29
, NS13 (
1991
).
25.
M. E.
Spira
and
A.
Hai
,
Nat. Nanotechnol.
8
,
83
(
2013
).
26.
D.
Purves
,
G.
Augustine
,
D.
Fitzpatrick
,
W.
Hall
,
A.
LaMantia
,
J.
McNamara
, and
L.
White
,
Neuroscience
, 4th ed. (Sinauer Associates,
2008
).
27.
S.
Ahadian
,
J.
Ramón-Azcón
,
S.
Ostrovidov
,
G.
Camci-Unal
,
V.
Hosseini
,
H.
Kaji
,
K.
Ino
,
H.
Shiku
,
A.
Khademhosseini
, and
T.
Matsue
,
Lab Chip
12
,
3491
(
2012
).
28.
I. S.
Kim
,
J. K.
Song
,
Y. L.
Zhang
,
T. H.
Lee
,
T. H.
Cho
,
Y. M.
Song
,
D. K.
Kim
,
S. J.
Kim
, and
S. J.
Hwang
,
Biochim. Biophys. Acta Mol. Cell Res.
1763
,
907
(
2006
).

Supplementary Material