Optimization of electrical stimulation parameters for MG-63 cell proliferation on chitosan/functionalized multiwalled carbon nanotube films

Abstract

Over the past decade, electrical stimulation and conductive materials have been investigated for their effects on cell behaviors. Although numerous studies have been conducted in these fields, fewer studies have been dedicated to the combination of electrical stimulation and conductive materials. Additionally, precise and optimal protocols for electrical stimulation are yet to be determined. In the current study, the in vitro effects of electrical stimulation on MG-63 cell lines have been studied. The cells were cultured on pseudo conductive chitosan/functionalized multiwalled carbon nanotubes films for better guidance of electrical current to the cells. These composite films with different contents of f-MWCNT (0.0%, 0.05%, 0.1%, 0.2%, and 0.4% w/v) were fabricated and characterized. The stimulations were conducted by means of an electrical stimulation device which was designed and fabricated for the purpose of this investigation. Different electrical signaling regimes which were designed by MiniTab software were then applied to the cells in order to find the optimum regime for augmenting cell proliferation. The results of MTT assays after 3, 7 and 10 days of electrical stimulation were used to determine the optimum regime for cell proliferation. Additionally, alkaline phosphatase assay, alizarin red staining and scanning electron microscopy were conducted and the results were analyzed. Finally, 1 volt, 0.4% f-MWCNT content, 50 Hz frequency and 5% duty cycle were determined as optimum electrical stimulation parameters for MG-63 cell proliferation long term.

---

Introduction

The ultimate goal of tissue engineering is to create effective and biocompatible tissue replacements. Different kinds of biochemical and biophysical stimulations support the development of tissue structure and function. Thus, understanding and using these kinds of stimulations for improving the quality of engineered tissues are essential for us. Nowadays, chemical, mechanical, electrical, material-based (such as topographical, scaffolds and etc.) and magnetic cues are known tools in developing tissues and organs in vitro. Amongst these cues, electrical stimulation has become an active research field in nerve, cardiac and musculoskeletal tissue engineering. Additionally, conductive substrates have attracted much attention in recent years. Simultaneous utilization of conductive materials and electrical stimulation can create a great opportunity for tissue engineering applications.^1–8

It is a well-known fact that the human body generates electrical fields and currents. In 1983, Foulds and Barker identified the 10 to 60 mV electrical potential in different parts of the body. Endogenous electrical fields play an essential role in controlling cell behaviors such as cell morphology, cell elongation, gene expression, proliferation and migration. The response of different biological systems to endogenous and exogenous electrical stimulations indicates that a proper electrical field can be an effective tool for controlling and regulating tissue and cell homeostasis. In vitro studies on a wide range of cells such as mesenchymal stem cells, bone and cartilage cells, nerve cells and heart cells have shown that electrical stimulation can affect important cell behaviors such as cell adhesion, proliferation, differentiation, and directed migration.^9–14

Aside from electrical stimulation, a determinative step in bone tissue engineering is proper material selection for fabricating a substrate. Biodegradable polymers are amongst the most promising materials for cell attachment, proliferation and phenotype retention. It has been shown that chitosan which is a natural biodegradable polymer can be used successfully as a single material, or in combination with other polymers and materials such carbon nanotubes in bone tissue engineering.^15–21

Application of conductive materials is potentially beneficial for electrical applications. Carbon nanotubes (CNTs) exhibit a wide range of electrical properties and this electrical conductivity can be advantageous for bone tissue engineering. Multiwalled carbon nanotubes (MWCNTs) are comprised of several graphite concentric cylindrical shells with a 0.3 to 0.4 nm gap. MWCNT-containing materials are widely used in electrical stimulation studies and they can accelerate bone formation and regeneration and also elevate the rate of ECM proteins and growth factors synthesis, and enhance osteogenic markers expression.^{10,15,22–24}

It would be beneficial to improve the outcomes of a bone tissue engineering study at least by considering characteristics of in vivo studies that mimic bone tissue engineering models. Such characteristics include the role of electricity which plays an important role in bone tissue formation and healing. In order to apply naturally inspired electrical stimulation, voltage, frequency and pulse width are amongst the most determinative parameters. It has been shown that outcomes of cell electrical studies, especially outcome of migration, orientation, and gene expression heavily depend on these parameters. Despite numerous efforts in the field of electrical stimulation, there are no completely meaningful and successive studies in this field because different signaling ranges have not been carefully compared; there are diverse protocols and regimes with different voltages, currents and even duty cycles and especially this fact is more evident for a specific tissue such as the bone tissue. For example, Hu et al. applied fields in the range of 35–350 mV cm⁻¹, Hronik-Tupaj et al. applied 20 mV cm⁻¹ fields, Dubey et al. considered 0.33 to 8 V cm⁻¹ fields and McCullen et al. even applied 1–10 V cm⁻¹ fields. Although, all of these studies yielded better results with electrical stimulation as compared to the control samples, selecting an optimum regime for practical purposes require extensive studies and optimization for signaling parameters. The purpose of this study was to fill this void and find an optimum signaling protocol and optimum values for stimulation parameters.^25–29

In the current project, effect of electrical stimulation has been studied from the aspect of bone cell proliferation. A novel electrical stimulation device was designed and fabricated in order to increase stimulation precision, pace and easiness. This setup included a main body, 316 stainless steel electrodes and a software-controlled electronic part.

MG-63 cells were stimulated on nonconductive chitosan films and pseudo-conductive composite films and the latter was used for better current conductivity to the cells. The conductivity in the films was provided by functionalized multiwalled carbon nanotubes (f-MWCNT) in their composite structure. These films were characterized by FTIR, tensile test and electrical conductivity measurement.

Because of the great number of required tests and the time-consuming test procedure, Minitab software (MiniTab Inc., State College, PA, USA) was used to minimize the number of tests and to analyze test outcomes. MG-63 cells were cultured for different periods and stimulation was applied for a certain amount of time each day based on a schedule during the culture period. The results of proliferation were then assessed by MTT proliferation assay and the analysis was performed using the Minitab software. Additionally, alkaline phosphatase (ALP) assay, alizarin red staining, and Scanning Electron Microscopy (SEM) were used for further investigation.

Experimental

Materials

Medium molecular weight chitosan (molecular weight = 210 000 Da, deacetylation degree = 80%), cetylpyridinium chloride, and 3-[4,-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) were supplied from Sigma, USA. Acetic acid and glutaraldehyde were purchased from Merck, Germany. Carboxylic acid functionalized multi-walled carbon nanotubes (f-MWCNT) were supplied by Neutrino Corporation, Iran. Human osteosarcoma cells, MG-63, were supplied by the National Cell Bank of Iran (NCBI), Pasteur Institute of Iran (NCBI, C555).

Preparation of f-MWCNT/chitosan films

Based on our previous evaluations for scaffolds with different contents of f-MWCNT, it was shown using f-MWCNT content more than 0.5% w/v is not practical because of structural integration loss (data not published). In this project, films with f-MWCNT contents of 0.01, 0.1, 0.2 and 0.4 w/v% were fabricated by simple evaporation method for further evaluation and more importantly to be used in electrical stimulation tests. Chitosan films were prepared as the control samples.

F-MWCNTs were dispersed ultrasonically in 1 v/v% acetic acid for 30 minutes. Ultrasonic dispersion beyond this would have helped the dispersion of f-MWCNTs but it could also inflict damage to the structure of f-MWCNTs and undermine the role of f-MWCNTs in our project. Chitosan was added slowly to the dispersion while it was on a magnetic mixer (250 rpm, 4 hours). Then, the solution was poured in to the wells of a 24-well plate which was covered by a non-adhesive layer. The plate was placed in a chemical incubator (37 °C) for 24 hours.

Films were neutralized in 0.05 M NaOH for 45 minutes. Then, they were sterilized by soaking them in ethanol in three steps: (1) first in 70% v/v ethanol for 30 min, (2) then, in 50% v/v ethanol for 30 min, and (3) finally in 25% v/v ethanol for 30 min. Then, the films were washed by PBS.

Characterization of f-MWCNT/chitosan scaffolds

Tensile test. The tensile properties of the samples were evaluated by using a modified version of tissue tensile machine designed and fabricated in a previous study.³⁰ The thickness of the rectangular specimens with dimensions of 5 × 10 mm² was measured using a digital micrometer. The mechanical properties of the samples were determined by uniaxial tensile testing machine with a 50 N load cell and an extension rate of 1 mm min⁻¹. The films were hold between two clamps and they were pulled by one clamp. Each sample was stretched destructively and the tensile strength and the stress–strain curve were obtained. Each test was repeated three times.

FTIR analysis. For Fourier transform infrared spectroscopy (FTIR) analysis, the films grinded in to powder. The main purpose of this test was to evaluate the bond formation between carboxyl groups of f-MWCNTs and amino groups of chitosan. Thus, only two films were analyzed (with 0% and 0.2% of f-MWCNT), because the fabrication processes were the same for the samples with different f-MWCNTs contents.

Electrical conductivity measurement. The electrical conductivity of the scaffolds was measured similarly to a previous work (data not published). Hydrated films were placed between two aluminum conductive sheets and they were fixed in a nonconductive substrate. The two conductive substrates were connected to a high precision multimeter and the resistance of the samples was measured with high precision. The distances between the conductive sheets and samples' dimensions were measured and the conductivities of the samples were calculated using eqn (1). Each test was performed 3 times.

σ = L/(R × A) (1)

In vitro cytotoxicity. MTT cytotoxicity assay was conducted based on ISO 10993-5 standard in order to evaluate the cytotoxicity of the films. 0.1 grams of each sample was cut and 1 ml of RPMI cell culture medium was added for every 0.1 g of sample.

Exposure process for each sample was conducted for 7 days. After those periods, the cell culture medium was extracted from the samples. For the MTT assay, MG-63 cells were incubated for 24 hours in 96-well plates. Then, film extracts were added to the wells with five repeats. Cell culture medium was added to MG-63 cells in five wells as controls. Cells were incubated for another 24 hours. Then, the culture medium was removed and 5 mg ml⁻¹ MTT solution was added to each well. After four hours, MTT solution was removed and isopropanol solution was added to each well. Finally, solution absorbance was quantified using a microplate reader at 570 nm.

Direct cell culture. In addition to the previous MTT assay, direct cell culture was conducted to ensure that the film surfaces are suitable for MG-63 cell proliferation.

Chitosan/f-MWCNT films (0%, 0.01%, 0.1%, 0.2% and 0.4% f-MWCNT content) were neutralized and sterilized. Then, 50 000 cells per cm² were placed on the scaffolds. RPMI cell culture medium containing 10% FBS was used for cell culture. After 3, 7 and 10 days, cells were trypsinized and they were transferred to a 96-well plate. 50 000 cells were cultured in five wells as controls.

After 24 hours of incubation, culture medium was removed and 5 mg ml⁻¹ MTT solution was added to each well. After 4 hours, MTT solution was removed and isopropanol solution was added to dissolve formazan product. Finally, the absorbance of this solution was quantified using a microplate reader at 570 nm. The absorbance was normalized to the control absorbance and the cell population was measured.

Electrical stimulation setup

Electrical section. An electrical stimulation setup was designed and fabricated in order to apply rectangular pulse electrical signals for in vitro cell studies (Fig. 1). This setup enabled us to simultaneously apply different electrical fields to 16 separate samples. Additionally, this setup gave us the capability to apply wide ranges and specific values of stimulation parameters accurately.

Fig. 1 Electrical stimulation setup with the ability to apply 16 different signaling regimes for 16 different samples.

The block diagram of our setup is illustrated in Fig. 2. USB card 4704 was used to send computer commands to the circuit. This card has the ability to create voltages between 0 and 5 V. In order to prevent using 16 USB cards, an expanding circuit was used to obtain a setup with 16 output channels (Fig. 3). LF389 module was the main component of the expanding circuit which delivers the voltage to the capacitor in which the voltage is stored and it is discharged over a long period of time.

Fig. 2 The block diagram of the electrical stimulation setup.

Fig. 3 The expanding circuit of the electrical stimulation device which converts an output voltage in to 16 output voltages.

In this stage, 16 obtained voltages were in the 0–5 volts range and after passing through two multiplexers, the voltage range is extended to include higher voltages (up to 25 V) and an acceptable accuracy was obtained for very low and very high voltages. Low or high voltage is chosen by a command which is sent from the digital circuit to the multiplexer and then, the voltage is transferred in to two outputs. One of these outputs is inverted by a subtraction circuit to obtain a signal to be used for the creation of pulsed signal. Then, these output signals are transferred to two multiplexers. LF389 was used in these multiplexers. The multiplexers require selectors and in this case, the selector signals originate from the section named invert (Fig. 2) which is connected to oscillator circuit. Two buffers which are placed after these two multiplexers, decrease undesired signal noise and increase the accuracy.

All of the modules which were used in the system work with currents in the range of microamperes. Since the resistance of the cell culture medium is relatively low, these modules would break down in our range of voltages because of higher amperes. Thus, we used a voltage follower right medium to amplify the current before the signal enters the culture (Fig. 4).

Fig. 4 The voltage follower which was used for current amplification.

In the voltage follower circuit, transistor base voltages are provided by 15 and −15 V power supplies and a voltage with higher current which is equivalent to the input voltage is generated. Since the dissipated power of this circuit is relatively high because of the significant difference between the output and input currents, a fan is required for cooling the transistors.

This previously mentioned oscillator circuit regulates the combination of the two initial output signals in order to create a signal with desired frequency and duty cycle. There are two comparators in the oscillator circuit which compare N_A and N_B with N_counter and determine the pulse length and the distance between two pulses, respectively. N_A and N_B are created by the digital circuit.

The selection of oscillator depends on the lowest required N_A. Since in our experiments, 150 Hz frequency and 0.01% duty cycle create the lowest value of N_A (0.0067 microseconds) and it is equivalent to 1.5 MHz, we used a 4 MHz oscillator with a 22 bit counter which covers the minimum and maximum necessary values. Finally, the feedback circuit was necessary to send all of the output values to the computer for further processing. The expanding circuit (Fig. 2) collects the output values of the feedback circuits and sends the results of the 16 channels to the device via one channel.

Physical section. The physical section of the setup consist of a body with 16 holders made of poly(methyl methacrylate) (PMMA) which is biologically neutral and sterilizable. Two parallel L-shaped electrodes (316L stainless steel which is commonly used for medical applications and has been proposed by Tandon et al. as a promising material for electrical stimulation purposes³¹) are devised in every well in order to apply electrical fields. A holder lid (PMMA) serves for setting and holding the electrodes 3 cm apart and it also prevent contaminants from entering the culture medium during the stimulation period.

Experimental design. In the current study, CNT content, voltage, frequency, and duty cycle of the applied signals were considered as variables. Our relatively wide ranges for variables were chosen with regards to previously applied range of parameters in the literature (Table 1).

Table 1 Selected values for voltage, frequency, duty cycle, and CNT content parameters

Parameter	Value #1	Value #2	Value #3	Value #4	Value #5	Value #6
Voltage (V)	0.05	0.15	0.5	1	2	4
Frequency (Hz)	1	10	50	—	—	—
Duty cycle (%)	1%	5%	20%	—	—	—
Film CNT content (%)	0.00	0.20	0.40	—	—	—

---

The f-MWCNT percentages were chosen under 0.5% w/v based on a previous research by our group (data not published). At first, films with 0 (control), 0.01, 0.1, 0.2 and 0.4% of f-MWCNT were fabricated and after film characterization, films with 0 (control), 0.2 and 0.4% of f-MWCNT were chosen for further evaluation. Based on the number of different parameters, 182 tests should have been conducted and considering the necessity for test repeats, this huge amount of tests was not applicable. Thus, Minitab® 17.1.0 software which is a prominent software for experimental design was used to design efficient numbers of experiments and after performing the tests, the results were analyzed by the same software to choose an optimum signaling regime and the best film f-MWCNT content. In the minimization of the tests, the software combined each parameter value with the different values of the other parameters, so that the specified parameter value is at least once tested with every possible value of other parameters. The 18 designed tests by this software are presented in Table 2.

Table 2 Experiments designed using Minitab software

Test number	Voltage (V)	CNT content (%)	Duty cycle (%)	Frequency (Hz)
Test 1	0.05	0.00	1	1
Test 2	0.05	0.20	5	10
Test 3	0.05	0.40	20	50
Test 4	0.15	0.00	5	1
Test 5	0.15	0.20	20	10
Test 6	0.15	0.40	1	50
Test 7	0.5	0.00	1	10
Test 8	0.5	0.20	5	50
Test 9	0.5	0.40	20	1
Test 10	1	0.00	20	50
Test 11	1	0.20	1	1
Test 12	1	0.40	5	10
Test 13	2	0.00	20	10
Test 14	2	0.20	1	50
Test 15	2	0.40	5	1
Test 16	4	0.00	5	50
Test 17	4	0.20	20	1
Test 18	4	0.40	1	10

---

Electrical stimulations. Eighteen composite films (according to Table 2) were neutralized and sterilized for every test series and each sample was placed in a 6 cm plate and 50 000 cells were place on the sample as described in the Section 2.3.5.

Each part of the electrical stimulation setup was sterilized based on its material composition. As a result, the main body and the doors of the setup which were made of poly(methyl methacrylate) (PMMA) were sterilized by UV radiation, and electrodes, screws, nuts and washers were autoclaved and assembled after sterilization under sterile conditions.

After 1 day of initial cell culture, the electrical stimulation was applied. The signaling regimes were applied 30 minutes each day with predetermined parameters from Table 2 for 5 days.

MTT proliferation assay. Cell population was assessed at 3, 7 and 10 days of cell culture. After culture periods, the cells were trypsinized and certain and equal amount of cell suspension from each sample was poured in to 5 wells of a 96-well plate and they were cultured for 24 hours. In the next step, culture medium was removed from the wells and 0.1 ml of 5 mg ml⁻¹ MTT solution was added to each well. After four hours, MTT solution was removed and 0.1 isopropanol was added to each well. Finally, solution absorbance was quantified using a microplate reader at 570 nm.

Alkaline phosphatase assay. The osteoblast activity was assessed by measuring the alkaline phosphatase (ALP) production of MG-63 cells. The level of ALP was determined on days 7 and 14 after the start of stimulation. The MG-63 cells were lysed with 0.5% Triton X-100. The centrifuged lysates were added to the p-nitrophenyl phosphate solution. The OD which indicates the release of p-nitrophenol from p-nitrophenyl phosphate was measured at 405 nm. The absorbance results were normalized by using eqn (2).

Alizarin red assay. Calcium minerals deposition of MG-63 cells was quantitatively measured by alizarin red staining. On days 7 and 14, the scaffolds were washed three times with PBS and fixed in 70% ethanol at 4 °C for 1 hour. Then, the samples were washed twice with deionized water and they were soaked for 15 minutes in alizarin red stain (2% (w/w) in deionized water).

In the next step, the samples were washed three times with deionized water and stain was extracted by incubating the samples in a 10% cetylpyridinium chloride solution for 1 hour at room temperature. The absorbance of the obtained stain was measured at 540 nm using a microplate reader. The results were normalized using eqn (2).

Scanning electron microscopy (SEM). After determining the optimum value for each parameter, all the parameters except the intended parameter are set to optimum value in order to evaluate the direct effect of that parameter. For the first set of tests, 0.4% f-MWCNT content, 50 Hz frequency and 5% duty cycle were used for the first six samples. 0.05, 0.15, 0.5, 1, 2 and 4 volts were the applied voltages for this set of tests. For the next three tests, 1 volt, 50 Hz and 5% were the fixed values for voltage, frequency and duty cycle. 0.0%, 0.2% and 0.4% f-MWCNT content were used in these samples. In the third set of samples, 1, 10 and 50 Hz frequencies were used in different samples, while 1 volt, 0.4% f-MWCNT content, and 5% duty cycle were the constant parameters. Different duty cycles (1%, 5%, and 20%) were used in the last set of tests and 1 volt, 0.4% f-MWCNT content and 50 Hz frequency were used for these tests.

These stimulations were also conducted 30 minutes each day for 5 days. The conditions of the cells were evaluated by SEM 5 days after the conclusion of stimulations. First, the culture medium was removed and the cells were washed with PBS. Then, 4% v/v glutaraldehyde was added to each plate. After 30 minutes of fixation, glutaraldehyde was removed and the fixed cells were again washed with PBS. Finally, each sample was sputter coated with gold and visualized by scanning electron microscope (SEM, Seron Technology AIS-2100, South Korea).

Statistical analysis. Each test was conducted 3 times except for the MTT assay, in which the test was performed five times for each sample. For statistical analysis, we used one-tailed t-test. A p-value of less than 0.05 was considered as statistically significant.

Results and discussion

Tensile test

Fig. 5 shows the results of ultimate stress of the films. A significant enhancement in the ultimate stress of the films can be easily observed from the figure as the f-MWCNT content increases up to 0.1%, but increasing f-MWCNT content from 0.2 to 0.4% resulted in a significant decrease in ultimate stress. The results of Young's modules were inconclusive because of the weak elasticity of the films.

Fig. 5 Ultimate stress of the chitosan/CNT films. * indicates significant difference (p < 0.05).

FTIR analysis

Fig. 6 shows the FTIR spectra for the control and 0.2% f-MWCNT samples. In the control sample spectrum, the peak at 3428 cm⁻¹ refers to hydroxy group in the chitosan structure. 2852 cm⁻¹ and 2923 cm⁻¹ peaks were in the range of C–H bond. Carbonyl group resulted in the peak at 1639 cm⁻¹. The peak that exists at 666 cm⁻¹ indicated the existence of type 1 amino group.

Fig. 6 FTIR spectrum for the control and the 0.2% f-MWCNT samples.

In the 0.2% f-MWCNT sample spectrum, all of the mentioned peaks are present with insignificant changes in wavelengths. One of the main differences is the appearance of a new peak at 1261 cm⁻¹. This peak may be referred to two bonds: (1) carboxyl group in the form of ketone bond, (2) C–N bond resulting from bond formation between chitosan amino group and f-MWCNTs carboxyl group. Both of these two potential cases indicate the bond formation between chitosan and f-MWCNTs via amino and carboxyl functional groups.

The other evident difference in absorption refers to amino group which occurs at 803 cm⁻¹ indicating type 2 amino group. This difference also supports this fact that a bond between chitosan and f-MWCNTs has formed.

Electrical conductivity measurement

Fig. 7 shows the results of electrical conductivity measurement. As expected, conductivity was enhanced as the f-MWCNT content increased. This enhancement was more evident in higher percentages of f-MWCNTs. As it can be seen, the increase in electrical conductivity was exponential because by increasing f-MWCNT content in the film structure, these f-MWCNTs can form longitudinal and transverse joints which can intensify the increase in the overall conductivity of the films.

Fig. 7 Electrical conductivity of the films with different f-MWCNT content. By increasing the f-MWCNT content, conductivity enhanced exponentially. * indicates significant difference (p < 0.05).

In vitro cytotoxicity

MG-63 cell line was cultured on films with different contents of f-MWCNTs. The results are shown in Fig. 8. Despite numerous concerns over the use of CNTs for tissue engineering applications, films with up to 0.4% f-MWCNTs exhibited no visible cytotoxicity. These results are of high importance because using a film in vitro requires biocompatibility.

Fig. 8 Cell viability of MG-63 cells on the films as a function of f-MWCNT content, measured by MTT assay which represents active mitochondrial activity of living cells. No significant cytotoxicity was observed.

Direct cell culture

The results of cell population on days 3, 7 and 10 of cell culture are shown in Fig. 9. The rates of cell proliferation were higher on the films which contain f-MWCNTs and there was no visible negative effect on cell population.

Fig. 9 MG-63 cell population on days 3, 7 and 10. * indicates significant difference (p < 0.05).

Based on the results of film characterizations (Sections 3.1 through 3.5), 0.2% f-MWCNT, 0.4% f-MWCNT and 0.0% f-MWCNT (as control samples) were chosen for further evaluations. The electrical conductivity was a crucial property for the purpose of this study and it was significantly higher in 0.2% f-MWCNT and 0.4% f-MWCNT samples. Additionally, 0.01% f-MWCNT and 0.1% f-MWCNT samples had no significant effect on cell population as compared to 0.0% f-MWCNT, whereas the cell populations were significantly augmented on 0.2% f-MWCNT and 0.4% f-MWCNT samples on days 7 and 10. Although the mechanical properties were lower in 0.2% f-MWCNT and 0.4% f-MWCNT samples, this slight decline was acceptable since the importance of mechanical properties was somewhat lower for the purpose of this study.

Electrical stimulations

MTT proliferation assay. As previously described, in order to minimize the necessary number of tests, the necessary tests were designed by Minitab software and all of the designed experiments were carefully performed and the results were exported to the software. Finally, the analytical graphs for each parameter were obtained using the software. The presented result for each parameter value is the result of software's statistical analysis of all of the conducted tests which include that value, while other results of the conducted tests were taken into account; finally, a single value is presented in the analytical graphs. The analytical graphs for each parameter on days 3, 7 and 10 are shown in Fig. 10, 11 and 12, respectively.

Fig. 10 Minitab software analytical graphs for day 3 results. Graphs A, B, C and D show the effects of voltage, f-MWCNT content, frequency and duty cycle on cell proliferation, respectively. The dotted line indicates the mean population. (A) 0.05, 0.15, 0.5, 1, 2 and 4 volts are the values for numbers 1–6, respectively. (B) 0, 0.2 and 0.4% f-MWCNT are the values for numbers 1, 2 and 3, respectively. (C) 1, 10 and 50 Hz are the values for numbers 1, 2 and 3, respectively. (D) 1, 5 and 20% are the values for numbers 1, 2 and 3, respectively.

Fig. 11 Minitab software analytical graphs for day 7 results. Graphs A, B, C and D show the effects of voltage, f-MWCNT content, frequency and duty cycle on cell proliferation, respectively. The dotted line indicates the mean population. (A) 0.05, 0.15, 0.5, 1, 2 and 4 volts are the values for numbers 1–6, respectively. (B) 0, 0.2 and 0.4% f-MWCNT are the values for numbers 1, 2 and 3, respectively. (C) 1, 10 and 50 Hz are the values for numbers 1, 2 and 3, respectively. (D) 1, 5 and 20% are the values for numbers 1, 2 and 3, respectively.

Fig. 12 Minitab software analytical graphs for day 10 results. Graphs A, B, C and D show the effects of voltage, f-MWCNT content, frequency and duty cycle on cell proliferation, respectively. The dotted line indicates the mean population. (A) 0.05, 0.15, 0.5, 1, 2 and 4 volts are the values for numbers 1–6, respectively. (B) 0, 0.2 and 0.4% f-MWCNT are the values for numbers 1, 2 and 3, respectively. (C) 1, 10 and 50 Hz are the values for numbers 1, 2 and 3, respectively. (D) 1, 5 and 20% are the values for numbers 1, 2 and 3, respectively.

As it can be observed from Fig. 10, increasing the voltage up to 1 volt resulted in higher cell population, but increasing the voltage beyond 1 volt led to lower rate of proliferation. This was not evident until 2 volts, but cell population decreased noticeably in 4 volts. Despite this drop, cell population in 4 volts signaling was similar to the results of the control samples (without stimulation) (Fig. 9), but considering the severe drop for 4 volts stimulation, we can safely assume that the rate of proliferation beyond 4 volts will become substantially less than the control rates of proliferation. This decrease in proliferation will intensify due to start of medium hydrolysis and corrosion of the electrodes at high voltages and the cell death rate can even become higher than the rate of proliferation. Based on the results of the day 3, 1 volt was the best voltage which is equal to an electrical field with a magnitude of approximately 0.33 V cm⁻¹ (considering the 3 cm distance between the electrodes). It should be mentioned that in some studies, higher values for electrical fields have been considered the optimum values (for example, Dubey et al.²⁸ and McCullen et al.²⁹). However, in the present study, we used a pseudo conductive substrate which increases the delivered amount of current to the cells. Additionally, it should be noted that low electrical fields have not been used in most of the papers and in these papers, it is merely concluded that very high values are not desirable. The exact opposite of this notion has been observed in other articles; in other words, only low electrical fields were investigated (for example, Hronik-Tupaj et al.²⁷ and Castells-Sala et al.³²). Interestingly, in a study which investigated both high and low values for electrical fields, 0.35 V cm⁻¹ was introduced as the most efficient value, which is close to our optimal electrical field.²²

Fig. 10B shows cell population variations for different f-MWCNT contents. The use of f-MWCNTs enhanced the results of cell proliferation. By comparing these results with the results of Fig. 9, it can be seen that the effect of f-MWCNT content on cell population was not sensible when electrical stimulation is not applied. The rate of proliferation increased substantially when 0.2% f-MWCNT film was used. Although the rate of proliferation decreased when 0.4% f-MWCNT content was used, this result cannot be considered completely undesirable as it was still higher than the rate of proliferation in the sample without the addition of f-MWCNT.

For evaluating and comparing the rate of proliferation in different frequencies, we refer to Fig. 10C. It is obvious that by increasing the frequency, the results of cell proliferation improved. Due to the numerous required experiments for this study, we did not examine frequencies higher than 50 Hz and we do not know up to which point this growing trend will continue. Dubey et al. obtained acceptable results even for frequencies up to 150 Hz.²⁸ However, we can conclude that applying lower frequency which was used in the literature^29,32 is not suitable for bone cell proliferation and higher frequencies up to 50 Hz would yield better results.

Finally, the analytical graph for the duty cycle parameter in Fig. 10D is addressed. Increasing the duty cycle up to 5% resulted in higher rate of proliferation, but increasing the duty cycle beyond this limit had a drastic negative effect on cell population. Although in the literature, duty cycles less than 1% (approximately 2 ms pulse width with moderate frequency) has been used,³² we have shown that higher values have resulted in better rate of proliferation.

By comparing Fig. 10 and 11, it is clear that the results of day 7 were fairly similar to the results of day 3. However, there were some differences between the graphs. In the voltage analytical graph, the upward trend continued until 1 volt in day 7 but there was a decrease for 2 volts which indicated long term negative effects for stimulations with higher voltages. Although the result of this voltage was still higher than the average line, this drop cannot be neglected.

According to the results of f-MWCNT content (Fig. 10B vs. 11B), the result of 0.4% f-MWCNT film improved substantially on day 7 and it became close to the results of 0.2% f-MWCNT film. Despite the fact that the cell population for 0.4% f-MWCNT sample was still less than the 0.2% f-MWCNT sample, the rate of cell proliferation was better from day 3 to day 7 in the 0.4% f-MWCNT sample. This improvement indicated that f-MWCNTs have positive long-term effects and by increasing the stimulation periods to 7 days, higher contents of f-MWCNT resulted in higher rate of proliferation, which is supported by the results of previous studies and can be contributed to the resolution of carbon nanotubes problems over the course of time.^33,34

For frequency and duty cycle parameters, we refer to Fig. 11C and D and compare them with Fig. 10C and D, respectively. There was a strong similarity between the figures. Although when these results are compared with the voltage parameter results, it was observed that the effects of different values of frequency and duty cycle parameters was decreased in long-term culture as compared to the voltage parameter.

In general, it can be said that electrical stimulation yielded better results than control samples without any electrical signaling. The only exception was the application of 4 volts on day 7 which resulted in lower cell population than the control samples.

The results of electrical stimulation on day 10 (Fig. 12) were also similar to the previous days, but we can notice some changes in the voltage graph. As it could be anticipated, the results of applying high voltages were undesirable in long-term. Even for 2 volts, for which the results of previous days were higher than the average, we observed a considerable drop in the proliferation in long term and cell population on day 10 was close to the cell population of control samples (without stimulation).

On the other hand, the results of 0.4% f-MWCNT samples were even superior to 0.2% f-MWCNT samples on day 10. The better induction of electrical stimulation in higher percentages of f-MWCNTs can help cell proliferation. Additionally, on day 10, the negative impacts of f-MWCNTs were alleviated, which makes the positive effects of carbon nanotubes more prominent.^33,34

In the two other graphs (Fig. 12C and D), there are no notable differences in comparison to the previous days and the results were almost identical to the previous days.

Alkaline phosphatase assay. As it can be observed from the results of day 7 (Fig. 13), ALP production was increased in higher voltages and contrarily to the results of cell proliferation, where the increase in cell proliferation was stopped in 1 volt, the increase in ALP production continued until 4 volts.

Fig. 13 Minitab software analytical graphs for day 7 results. Graphs A, B, C and D show the effects of voltage, f-MWCNT content, frequency and duty cycle on alkaline phosphatase activity, respectively. The dotted line indicates the mean absorbance. (A) 0.05, 0.15, 0.5, 1, 2 and 4 volts are the values for numbers 1–6, respectively. (B) 0, 0.2 and 0.4% f-MWCNT are the values for numbers 1, 2 and 3, respectively. (C) 1, 10 and 50 Hz are the values for numbers 1, 2 and 3, respectively. (D) 1, 5 and 20% are the values for numbers 1, 2 and 3, respectively.

For the different percentages of f-MWCNTs, we can see an upward trend as the f-MWCNTs content increases. Again, this is in contrary to the results of cell proliferation, where the cell proliferation on 0.4% f-MWCNT samples was less than the cell proliferation on 0.2% f-MWCNT samples.

Considering Fig. 13C, the effects of frequency parameter on ALP production were negligible. As the frequency increases from 10 Hz to 50 Hz, the ALP production was slightly increased, but this increase was not noticeable.

As for the duty cycle parameter, again there were differences between the ALP production and cell proliferation results. There was a direct relationship between ALP production and duty cycle but there is no maximum duty cycle value for the increase in ALP production.

The general trends of day 7 and day 14 were relatively similar, but in Fig. 14A and D, the slopes were less than Fig. 13A and D. Considering the fact that by increasing the voltage and duty cycle, a continuous increase in ALP production can be observed, it can be concluded that the increase in ALP production had a direct relationship with the value of applied energy during electrical stimulation. Although the cell proliferation capability was lower in high voltages and duty cycles, the cells were capable of producing more ALP. It can be concluded that the greater values of voltage and duty cycle are more proper for differentiation purposes. Additionally, one can notice the lack of substantial effects of frequency on ALP production.

Fig. 14 Minitab software analytical graphs for day 14 results. Graphs A, B, C and D show the effects of voltage, f-MWCNT content, frequency and duty cycle on alkaline phosphatase activity, respectively. The dotted line indicates the mean absorbance. (A) 0.05, 0.15, 0.5, 1, 2 and 4 volts are the values for numbers 1–6, respectively. (B) 0, 0.2 and 0.4% f-MWCNT are the values for numbers 1, 2 and 3, respectively. (C) 1, 10 and 50 Hz are the values for numbers 1, 2 and 3, respectively. (D) 1, 5 and 20% are the values for numbers 1, 2 and 3, respectively.

On the other hand, despite the decrease in cell proliferation capability in higher percentages of f-MWCNT on days 3 and 7, the ALP production was greater in higher contents of f-MWCNTs. In other words, although the number of viable cells decreased, the differentiation ability was augmented.

On day 14, we can clearly observe that the effects of CNT percentage parameter have become more prominent than other effective stimulation parameters. Unlike the results of proliferation in stimulated samples, in which the optimum samples continued to show better proliferation even after the end of signaling on day 5 as if the effects of stimulation were maintained, by comparing the increased amount of ALP production form day 7 to day 14, it is fairly evident that the effects of stimulation gradually faded out in high values of voltage and duty cycle. This diminished effect indicated that the effects of electrical stimulation after the stimulation period were not maintained for ALP production unlike the results of cell proliferation and were more distinguished during the excitation period in comparison to the time period after the end of stimulation. Unlike the effects of other parameters, the effects of f-MWCNT content were maintained after the stimulation periods, since the effective parameter was still affecting the cells after the stimulation period.

Alizarin red assay. The results of mineral deposition which were assessed by qualitative alizarin red staining are quite similar to the results of ALP activity (Fig. 15). It was shown that the using higher voltages and duty cycles, and increasing the f-MWCNTs content led to an evident elevation in mineralization on 7th day.

Fig. 15 Minitab software analytical graphs for day 7 results. Graphs A, B, C and D show the effects of voltage, f-MWCNT content, frequency and duty cycle on mineral deposition, respectively. The dotted line indicates the mean absorbance. (A) 0.05, 0.15, 0.5, 1, 2 and 4 volts are the values for numbers 1–6, respectively. (B) 0, 0.2 and 0.4% f-MWCNT are the values for numbers 1, 2 and 3, respectively. (C) 1, 10 and 50 Hz are the values for numbers 1, 2 and 3, respectively. (D) 1, 5 and 20% are the values for numbers 1, 2 and 3, respectively.

In contrast to the results of ALP assays, in which the frequency effects on the level of ALP activity were not noticeable, higher frequencies led to slightly higher level of mineralization.

The results of quantified alizarin red staining in day 14 were almost identical to day 7, but the gap between the results of different values for voltages and duty cycles have been reduced in comparison to the f-MWCNT values. Similarly to the results of day 7, increasing the f-MWCNT content had a strong impact on the level of mineralization (Fig. 16).

Fig. 16 Minitab software analytical graphs for day 14 results. Graphs A, B, C and D show the effects of voltage, f-MWCNT content, frequency and duty cycle on mineral deposition, respectively. The dotted line indicates the mean absorbance. (A) 0.05, 0.15, 0.5, 1, 2 and 4 volts are the values for numbers 1–6, respectively. (B) 0, 0.2 and 0.4% f-MWCNT are the values for numbers 1, 2 and 3, respectively. (C) 1, 10 and 50 Hz are the values for numbers 1, 2 and 3, respectively. (D) 1, 5 and 20% are the values for numbers 1, 2 and 3, respectively.

It is worth mentioning that like the results of ALP assays, the effects of electrical stimulation on mineralization were not maintained with same strength as during the stimulation period.

Scanning electron microscopy (SEM). SEM images of the samples are shown in Fig. 17. Fig. 17a–f are the different samples for investigating the effects of the voltage parameter on cell aggregation. In Fig. 17c and d for 0.5 and 1 volt, cell aggregations with higher densities are observed. These results conform to the rate of cell proliferation on day 10 in which 1 volt led to best results. Addition of f-MWCNT to the film structure led to better cell contact (Fig. 17i vs. 17g). In Fig. 17i with the highest f-MWCNT content, better cell network formation is observed. In Fig. 17j–l, cell elongation gradually improved as the frequency was elevated. As for the duty cycle parameter, cell aggregations were in better conditions in moderate duty cycle (Fig. 17m–o).

Fig. 17 SEM micrographs of osteoblast-like cells after electrical stimulation. (a–f) Effects of 0.05, 0.15, 0.5, 1, 2 and 4 volts, respectively (0.4% f-MWCNT, 50 Hz, 5% duty cycle). (g–i) Effects of 0.0%, 0.2% and 0.4% f-MWCNT content, respectively (1 volt, 50 Hz, 5% duty cycle). (j–l) Effects of 1, 10 and 50 Hz frequencies, respectively (1 volt, 0.4% f-MWCNT, 5% duty cycle). (m–o) Effects of 1, 5, and 20% duty cycles, respectively (1 volt, 0.4% f-MWCNT, 50 Hz).

Conclusion

The focus of this study was to investigate the effects of electrical stimulation on cell behaviors, particularly on cell proliferation and we identified an optimum electrical stimulation regime for osteoblast-like cell proliferation.

Based on film characterizations, 0.2% and 0.4% w/v f-MWCNT samples were chosen as superior samples. Overall, electrical stimulation with 1 volt (0.33 V cm⁻¹), 0.4% f-MWCNT content, 50 Hz frequency and 5% duty cycle is determined as the optimum signaling regime for bone cell proliferation purpose. However, 0.2% f-MWCNT content led to better short-term results.

The results of ALP production and alizarin red staining collectively showed that higher values of voltage, duty cycle and f-MWCNT percentage resulted in enhanced ALP production and calcium deposition. As for the frequency parameter, different frequencies did not yield in considerable changes in ALP production, but increasing the frequency slightly elevate the level of mineralization.

We successfully developed an electrical stimulation device for this study. The in vitro applicability of our electrical stimulation setup has been demonstrated in this study. Additionally, our fabricated setup can be used for in vivo experiments with slight modification of the physical part and there is a promising opportunity to continue this research in future in vivo studies.

Acknowledgements

The authors of this paper would like to thank Mr Rahim Faturechi, Dr Mahdi Hasani Sadrabadi, Dr Mohammadreza Tahriri, Mr Reza Karimi, Ms Fatemeh Mohabbatpour and the staff of National Cell Bank of Iran.

References

1. R. Balint, N. J. Cassidy and S. H. Cartmell, Tissue Eng., Part B, 2012, 19, 48–57.
2. L. Ghasemi-Mobarakeh, M. P. Prabhakaran, M. Morshed, M. H. Nasr-Esfahani, H. Baharvand, S. Kiani, S. S. Al-Deyab and S. Ramakrishna, J. Tissue Eng. Regener. Med., 2011, 5, e17–e35.
3. A. Llucià-Valldeperas, B. Sanchez, C. Soler-Botija, C. Gálvez-Montón, C. Prat-Vidal, S. Roura, J. Rosell-Ferrer, R. Bragos and A. Bayes-Genis, J. Tissue Eng. Regener. Med., 2015, 9, E76–E83.
4. Z. Li, G. Yang, H. Huang, Z. Xiao and X. Ren, Chin. J. Cell. Mol. Immunol., 2015, 31, 1200.
5. J. G. Hardy, R. C. Sukhavasi, D. Aguilar, M. K. Villancio-Wolter, D. J. Mouser, S. A. Geissler, L. Nguy, J. K. Chow, D. L. Kaplan and C. E. Schmidt, J. Mater. Chem. B, 2015, 3, 8059–8064.
6. S. Ahadian, J. Ramón-Azcón, H. Chang, X. Liang, H. Kaji, H. Shiku, K. Nakajima, M. Ramalingam, H. Wu, T. Matsue and A. Khademhosseini, RSC Adv., 2014, 4, 9534–9541.
7. C. J. Bettinger, R. Langer and J. T. Borenstein, Angew. Chem., Int. Ed., 2009, 48, 5406–5415.
8. I. Martin, D. Wendt and M. Heberer, Trends Biotechnol., 2004, 22, 80–86.
9. S. Meng, M. Rouabhia and Z. Zhang, Electrical stimulation in tissue regeneration, INTECH Open Access Publisher, 2011.
10. P. Petrov, P. Mokreva, I. Kostov, V. Uzunova and R. Tzoneva, Carbohydr. Polym., 2016, 140, 349–355.
11. D. M. Thompson, A. N. Koppes, J. G. Hardy and C. E. Schmidt, Annu. Rev. Biomed. Eng., 2014, 16, 397–430.
12. J. H. Lee, W. Y. Jeon, H. H. Kim, E. J. Lee and H. W. Kim, Biomaterials, 2015, 53, 358–369.
13. R. Balint, N. J. Cassidy, L. Araida Hidalgo-Bastida and S. Cartmell, J. Biomater. Tissue Eng., 2013, 3, 212–221.
14. R. Mishra, D. B. Raina, M. Pelkonen, L. Lidgren, M. Tägil and A. Kumar, Int. J. Biol. Sci., 2005, 11, 1325.
15. J. Venkatesan and S. K. Kim, Mar. Drugs, 2010, 8, 2252–2266.
16. B. S. Harrison and A. Atala, Biomaterials, 2007, 28, 344–353.
17. A. De la Paz Orozco, F. Jiménez Vega, S. A. Martel-Estrada, A. Hernández Aguilar, M. E. Mendoza-Duarte, M. C. Chavarría-Gaytán, C. A. Rodríguez-González and I. Olivas-Armendáriz, Open J. Regener. Med., 2016, 5, 14.
18. J. Yan, T. Wu, Z. Ding and X. Li, Carbohydr. Polym., 2016, 136, 1288–1296.
19. P. Gupta and D. Lahiri, Neural Regener. Res., 2016, 11, 1062.
20. A. K. Mahanta and P. Maiti, Chitin and Chitosan for Regenerative Medicine, Springer, India, 2016, pp. 123–149.
21. R. LogithKumar, A. KeshavNarayan, S. Dhivya, A. Chawla, S. Saravanan and N. Selvamurugan, Carbohydr. Polym., 2016, 151, 172–188.
22. W. W. Hu, Y. T. Hsu, Y. C. Cheng, C. Li, R. C. Ruaan, C. C. Chien, C. A. Chung and C. W. Tsao, Mater. Sci. Eng., C, 2014, 37, 28–36.
23. Z. J. Han, A. E. Rider, M. Ishaq, S. Kumar, A. Kondyurin, M. M. M. Bilek, I. Levchenko and K. K. Ostrikov, RSC Adv., 2013, 3, 11058–11072.
24. G. Kaur, R. Adhikari, P. Cass, M. Bown and P. Gunatillake, RSC Adv., 2015, 5, 37553–37567.
25. N. W. S. Kam, E. Jan and N. A. Kotov, Nano Lett., 2008, 9, 273–278.
26. W. L. Stoppel, D. L. Kaplan and L. D. Black, Adv. Drug Delivery Rev., 2016, 98, 135–155.
27. M. Hronik-Tupaj, W. L. Rice, M. Cronin-Golomb, D. L. Kaplan and I. Georgakoudi, Biomed. Eng. Online, 2011, 10, 9.
28. A. K. Dubey, S. D. Gupta and B. Basu, J. Biomed. Mater. Res., Part B, 2011, 98, 18–29.
29. S. D. McCullen, J. P. McQuilling, R. M. Grossfeld, J. L. Lubischer, L. I. Clarke and E. G. Loboa, Tissue Eng., Part C, 2010, 16, 1377–1386.
30. R. Faturechi, A. Hashemi and N. Abolfathi, J. Med. Eng. Technol., 2014, 38, 411–415.
31. N. Tandon, A. Marsano, R. Maidhof, L. Wan, H. Park and G. Vunjak-Novakovic, J. Tissue Eng. Regener. Med., 2011, 5, e115–e125.
32. C. Castells-Sala, B. Sanchez, L. Recha-Sancho, V. Puig, R. Bragos and C. E. Semino, Annual International Conference of the IEEE Engineering in Medicine and Biology Society, 2012, pp. 5658–5661.
33. R. A. MacDonald, B. F. Laurenzi, G. Viswanathan, P. M. Ajayan and J. P. Stegemann, J. Biomed. Mater. Res., Part A, 2005, 74, 489–496.
34. P. Newman, A. Minett, R. Ellis-Behnke and H. Zreiqat, Nanomedicine, 2013, 9, 1139–1158.

---