Introduction

Conducting polymers are an arising interest in the field of tissue engineering as they can deliver electrochemical as well as electromechanical stimulation to cells. From those polypyrrole (PPy) and poly(3,4-ethylenedioxythiophene) (PEDOT) are the most investigated for biomedical applications owing to their good biocompatibility in vivo and in vitro.8 PPy is intensively investigated for bone9,37,38,46 and neural applications41,50 due to its easy modification with bioactive agents in ambient conditions and highly adjustable properties, such as surface charge and topography17,18,44,46 whereas PEDOT studies concentrate more on neural electrodes and nerve grafts1–5,13,20,21,35,42 mostly owing to PEDOT’s higher electrical conductivity and stability compared to PPy.8 In addition to bone and neural tissue engineering, PPy has so far been studied as bioactive coatings to improve osseointegration,11 in biosensors,7 drug delivery systems50 and actuators.34 Regards to the comprehensive research supporting PPy’s use in bone tissue engineering, we chose PPy and evaluated the effects of the two most potential bioactive dopants, hyaluronic acid (HA) and chondroitin sulfate (CS) in the PPy films.

In electrochemical polymerization of PPy, charged biomolecules, such as negatively charged glycosaminoglycans (GAGs), can be incorporated into the structure by doping when PPy is electrochemically polymerized by oxidation. Dopants play an important role in mediating the electric charges between the PPy chains.23 In addition, the surface topography and mechanical properties of PPy can also widely be altered by the choice of dopant.17,18

CS and HA are GAGs commonly found in the extracellular matrices (ECMs) of most animal tissues. CS is a major proteoglycan component in organic matrix of the bone and is involved in the mineralization of the bone tissue whereas HA takes part in various cellular processes, such as ECM organization and metabolism.43 Both GAGs are reported to support osteogenic differentiation of mesenchymal stem cells (MSCs) in scaffold structures in vitro.28,53

As regards integrating HA and CS into PPy surfaces, HA doped PPy (PPy–HA) has been studied with mouse bone marrow derived MSCs resulting in promoted osteogenic differentiation46 and with MC3T3-E1 osteoblasts confirming cell differentiation on the surface.45 In addition, we recently were the first to report an excellent attachment, proliferation and early osteogenic differentiation of hASCs on chemically synthetized PPy–CS coating in non-woven polylactide fiber scaffolds.38 As both biomolecules are potential dopants for PPy coating in osteogenic applications, a systematic comparison is required to understand their benefits and differences with human MSCs.

Inherent electrical currents and fields are essential in terms of the growth and remodeling of bone tissue. This was first demonstrated by Fukada and Yasuda, who reported bone formation under tension when positive charge was dominating, and the opposite in case of a negative charge and compression.15 This remark led to the development of electrical stimulation (ES) devices for treating severe bone defects.22

Even though ES has been acknowledged as a bone treatment method for several decades, the exploitation of ES to MSCs in bone tissue engineering has been studied only recently.25,26,31,36 These studies have shown that various types of ES can be applied to improve osteogenic differentiation and proliferation of MSCs, yet no specific parameters for the efficient differentiation of MSC towards mature osteoblasts have so far been identified. In addition, most of the above mentioned studies exploited inert conductive substrates; hence a conductive coating with bioactive molecules and topographical cues may yield interesting synergy mimicking the natural environment in the bone tissue more closely. We therefore wanted to evaluate hASC spreading, proliferation and osteogenic differentiation on PPy surfaces under ES with our novel ES device developed in-house. To the best to our knowledge, this is the first paper to systematically compare HA and CS doped PPy coatings for hASCs.

Materials and Methods

Polypyrrole Synthesis

Pyrrole (Sigma-Aldrich, St. Louis, USA) of 0.07 mL and 1 mg of HA from Streptococcus equi (Sigma-Aldrich) or CS A from bovine trachea (Sigma-Aldrich) were added per 1 mL of water. PPy–HA and PPy–CS films were grown electrochemically on a sputter-coated polyethylene-naphthalate film (PEN)/Au films (125 µm Dupont Teonex®), with 50 nm Au-coating (VTT Technical Research Center of Finland) as a working electrode, platinum mesh as a counter electrode and Ag/AgCl as a reference electrode. Constant potential of 1.0 V was applied to the films until 300 mC cm−2 polymerization charge had passed the cell. The stimulation plates and plate covers were sterilized by gamma irradiation (BBF Sterilisationsservice GmbH, Kernen, Germany) with an irradiation dose of >25 kGy that has not been reported to significantly alter the conductivity of the films.12,54

Surface Characterization of Polypyrrole Film

The surface morphology and roughness (R a) values of the PPy films were characterized by an atomic force microscope (AFM; Park Systems XE-100, Korea) in both dry and wet conditions due to the significant water absorption and hence swelling phenomenon of wet PPy films in physiological conditions.39,48 PPy–HA and PPy–CS films were incubated for 4 days in osteogenic medium (OM) containing 250 mM ascorbic acid 2-phosphate (Sigma-Aldrich), 5 nM dexamethasone (Sigma-Aldrich) and 10 mM b-glycerofosfate (Sigma-Aldrich) supplemented to maintenance medium consisting of Modified Eagle Medium/Ham’s Nutrient mixture F-12 (DMEM/F-12 1:1 Invitogen), 10% fetal bovine serum (FBS; Invitrogen), 1% l-glutamine (GlutaMAX I; Invitrogen) and 1% antibiotics/antimycotic (100 U mL−1 penicillin, 0.1 mg mL−1 streptomycin; Invitrogen).

To distinguish the swelling effect from the typical polysaccharide doped PPy nodular morphology,17,18,39,45,47 and in order to image the nanoscopic details of the soft films,17,38 dried samples were analyzed using non-contact AFM (Park Systems XE-100) in air using silicon probe ACTA-905M (Applied NanoStructures, Inc.) with a nominal resonance frequency of 300 kHz, spring constant 40 N m−1 and tip radius <10 nm. Images of 5 × 5 µm2 were acquired with a scan rate of 0.5 Hz. Prior to imaging, the sample surfaces were carefully rinsed with deionized water and dried in ambient air.

The films pre-incubated in OM were imaged in Dulbecco’s phosphate-buffered saline (PBS; Lonza Biowhittaker, Switzerland) using a contact mode AFM. Silicon nitride probes HYDRA-6R100N (Applied Nanostructures, Inc., Santa Clara, USA), with a nominal force constant 0.28 N m−1 and a tip radius of curvature <8 nm were applied. Areas of 20 × 20 µm2 were scanned at 7 and 20 nN force set points for PPy–HA and PPy–CS surfaces respectively. Images of 12 × 12 µm2 were acquired with the scan speed of 1 Hz.

R a value analysis of the raw 512 × 512 pixel data was conducted. R a values for 10 randomly chosen 4 × 4 µm2 subareas were calculated using Park Systems XEI 1.7.5 image analysis software. AFM image data was 4th order plane fitted to show the nanoscale details of the PPy–HA and PPy–CS surfaces.

Characterization of Electrical Properties of the Films

Electrochemical impedance spectroscopy (EIS) and cyclic voltammetry measurements were taken from 1 cm2 of doped PPy films on gold Mylar that was acting as working electrode. Platinum mesh acted as counter electrode and Ag/AgCl (3.0 M NaCl) as reference electrode. Impedance measurements were recorded by using CH 660D Electrochemical Analyzer/Workstation (CH Instruments, Austin, USA). Impedance spectra were obtained from 100 mHz to 100 kHz using an AC amplitude of ±200 mV. All EIS measurements were performed at their resting potential ranging from +70 to +195 mV vs. the reference electrode to prevent destruction in the films. Average impedance and the standard deviation at 10 and 100 Hz were calculated from three samples per film type.

Cyclic voltammetry of the films was recorded in PBS by CH 660D Electrochemical Analyzer/Workstation in similar electrode setup as impedance recordings. Measurements were performed at the scan rate of 50 mV s−1 within the range of −0.6 to 0.5 V.

After the cell culture experiments, the through plane electrical conductivity of PPy films was monitored in air using a simple two-wire test setup (Fluke 170 multimeter, Washington, USA). The top and bottom electrodes applied were a round Au contact electrode (contact area of 16 mm2) and the PEN/Au film, respectively.

Isolation and Culture of Human Adipose Stem Cells

The adipose tissue was obtained from tissue harvests from surgical procedures on three female donors with an average age of 54 ± 12 years in the Department of Plastic Surgery, Tampere University Hospital. The tissue harvesting and the use of hASCs were conducted in accordance with the Ethics Committee of the Pirkanmaa Hospital District (R03058). The hASC isolation method was presented earlier by Haimi et al.24. Briefly, the samples from adipose tissue were digested with collagenase type I (1.5 mg (ml; Invitrogen, California, USA). After centrifugation and filtration, the isolated hASCs were maintained and expanded in T-75 cm2 polystyrene flask (Nunc, Roskilde, Denmark). The experiments were repeated three times, each time using different donors.

Before the cell seeding, PPy–CS and PPy–HA films were rinsed with PBS and pre-treated with maintenance medium at 37 °C for 48 h. PPy–CS and PPy–HA coated plates were cell seeded at passage 3 with a density of 16000 cells cm−2. Cells were allowed to attach for 24 h before initiation of ES. On the first day of stimulation the maintenance medium was replaced by OM and medium was changed twice a week.

Flow Cytometric Surface Marker Expression Analysis

Cells were characterized by a fluorescence-activated cell sorter (FACSAria; BD Biosciences, Erembodegem, Belgium) at passage 1 after primary culture in T-75 flasks. This was described earlier by Lindroos et al. 33. Monoclonal antibodies were used against the following surface markers: CD14, CD19, CD49d-PE, CD90-APC, CD106-PECy5 (BD Biosciences); CD45-FITC (Miltenyi Biotech, Bergisch Gladbach, Germany); CD34-APC, HLAABC-PE, HLA-DR-PE (Immunotools GmbH, Friesoythe, Germany) and CD105-PE (R&D Systems Inc., MN, USA). Analysis was performed on 10,000 cells per sample and unstained cell samples were used to compensate the background autofluorescence levels.

Biphasic Electrical Stimulation of Human Adipose Stem Cells

The stimulation plate assembly was earlier described by Pelto et al.39 As an exception to the previous setup, PPy film was polymerized on the bottom electrode. ES was performed in a cell culturing incubator (37 °C, 5% CO2). Samples were stimulated for 4 h a day for 14 days with a biphasic electric current (BEC) of ±0.2 V amplitude, 2.5 ms pulse width and 100 Hz pulse repetition frequency. Non-stimulated samples acted as controls in each film type. The shortest vertical distance between the top (Fig. 1: 1) and the bottom (Fig. 1: 2) electrodes immersed in each well was 2 mm. The measured steady state direct current after the 2.5 ms pulses was in the range of 40–50 µA cm−2, corresponding to a cell impedance of 5 kΩ. The direction of the current was perpendicular to the PPy films.

Figure 1
figure 1

(a) Schematic illustration of the stimulation device geometry side projection. Top (1) and bottom (2) electrodes were made of gold coated PEN film. Cells were seeded to the bottom electrode (4) coated with PPy layer

Full size image

Cell Attachment and Viability

Cell attachment and viability were evaluated qualitatively using live/dead staining (Molecular Probes, Eugene, USA). Cells were incubated in PBS-based dye solution containing 0.5 µM of CellTracker™ Green (5-chloromethylfluorescein diacetate, CMFDA; Molecular Probes) and Ethidium homodimer-1 (EthD-1; Molecular Probes) at room temperature for 45 min. Samples were examined with a fluorescence microscope (Olympus IX51, Olympus Finland PLC, Vantaa, Finland). Standard polystyrene (PS) culturing plates (Nunc, Roskilde, Denmark) served as a positive control for the cell viability and morphology evaluation.

Cell Proliferation

Cell proliferation was studied with CyQuant® Cell Proliferation Assay Kit (Molecular Probes). The experiment was performed according to the manufacturer’s protocol. Briefly, on the day of the analysis, samples were carefully washed with PBS and cells were suspended in 0.1% Triton-X 100 buffer (Sigma-Aldrich) in PBS and stored at −70 °C until analysis. After thawing, 20 µl of three parallel samples was mixed with CyQuant® GR dye and lysis buffer. The fluorescence was measured with a microplate reader (Victor 1420 Multilabel Counter, Wallac, Turku, Finland) at 480/520 nm.

Osteogenic Differentiation

Alkaline phosphatase (ALP) activity was determined using an ALP Kit (Sigma-Aldrich) according to the manufacturer’s protocol. The ALP activity was determined from the same Triton-X 100 lysates as in the cell proliferation assay. The samples were incubated with 50% alkaline buffer solution (2-amino-2-methyl-1-propanol, 1.5 mol L−1, pH 10.3; Sigma-Aldrich) and 50% of stock substrate solution (p-nitrophenyl phosphate; Sigma-Aldrich) at 37°C for exactly 15 min. To stop the reaction 1.0 mol L−1 sodium hydroxide was added. The intensity of the color was measured at 405 nm using a microplate reader (Victor 1420).

The mineralization of the ECM was studied with Alizarin Red staining. Samples were rinsed with PBS and fixed in ice cold 70% ethanol (Altia Corporation, Helsinki, Finland) for 60 min at room temperature. Samples were then rinsed with distilled water before the addition of 2% Alizarin Red solution (pH 4.2; Sigma-Aldrich) for 5 min. After incubation, samples were rinsed three times with distilled water and once with 70% ethanol. Samples were then incubated in cetylpyridium chloride (Sigma-Aldrich) for 3 hours. Supernatant was pipetted in triplicate on a 96-well plate (Nunc, Roskilde, Denmark) and absorbance measured at 544 nm using a microplate reader (Victor 1420).

Statistical Analysis

The statistical analyses of R a values, DNA content, ALP activity and mineralization were performed with SPSS, version 19. R a values of the films were analyzed with Student’s t test and the equal variance assumption was checked by Levene’s Test. A one way analysis of variance (ANOVA) with Bonferroni post hoc correction was used to determine the effect of the PPy coating and ES on DNA content, ALP activity and mineralization. The effect of culture duration on proliferation and ALP activity was analyzed using Student’s t test for independent samples. The cell culture experiments were repeated three times with three parallel samples, each repetition using a different hASC donor. The data from the three experiments were combined and presented as mean ± standard deviation (SD) and the results were considered statistically significant when p < 0.05.

Results

Surface Characterization of Polypyrrole Film

The measured R a values of the wet and the dry PPy films are presented in Table 1. PPy–HA films had significantly higher R a values than PPy–CS in PBS, whereas dry PPy–HA film had significantly lower R a values when measured with non-contact AFM in air.

Table 1 R a values of non-contact AFM data of air-dried PPy films and contact mode AFM of wet films in PBS as well as impedance at 100 Hz frequency
Full size table

The non-contact AFM data (Figs. 2a and 2b) show the nanoscopic details of the dry PPy–CS (Fig. 2a) and PPy–HA (Fig. 2b) films. PPy–CS surface texture (Fig. 2a) consisted mainly of nodules, sized 40–50 nm in diameter and 5–10 nm in measurable height, which were organized into a porous web. The surface texture of PPy–HA (Fig. 2b) consisted of larger nodules, 150–160 nm in diameter and 30–35 nm in height. In both films, protrusions were sparsely distributed. The protrusions in the PPy–CS and PPy–HA films were similar in size, 500 nm in diameter and 100 nm in high. The nanoscopic texture was only observed in the dry samples and was not resolvable by AFM in PBS.

Figure 2
figure 2

Surface topography of dry PPy–CS (a) and PPy–HA (b) films imaged using non-contact AFM in air. The sparsely distributed protrusions are seen as white areas. Scanned area is 5 × 5 µm and Z-scale 20 nm/division. (c) Surface topography images of wet PPy–CS and (d) PPy–HA films imaged with contact AFM in PBS. 800–1000 nm nodules are seen as white areas. Scanned area is 12 × 12 µm and Z-scales of the images are 40 nm/division

Full size image

Imaging in PBS (contact mode) revealed a strongly undulating morphology of both films consisting of uniformly spread small 800–1000 nm nodules (Figs. 2c and 2d) and sparsely spread 4–10 µm circular protrusions covered with nodules (not visible in the flattened image data in Fig. 2).The nodules were typically of 100–150 and 200–300 nm in height for the PPy–HA (Fig. 2d) and PPy–CS (Fig. 2c) respectively. It is noteworthy that the hills caused by the undulating morphology as well as the protrusions were significantly higher than the nodules, typically 400–500 and 600–800 nm for the PPy–CS and PPy–HA films respectively. The protrusions in the PPy–CS were circular whereas those in PPy–HA were more oval. Hence the R a values obtained from the AFM in PBS based on raw imaged data were not only representative of the nodules’ height and shape but of the height of the protrusions.

Electrical Properties of the Films

The impedance of PPy–HA and PPy–CS films did not vary significantly yet PPy–CS showed slightly higher impedance values at 10 and 100 Hz (Table 1). Both films showed similar trend in the impedance spectra (Fig. 3). Both films showed well-defined voltammetric profiles though PPy–CS had slightly higher electrochemical activity and doping level compared to PPy–HA, as evidenced by the integrated surface areas covered by the respective voltammograms (Fig. 4). The conductivity of the films was confirmed to be in similar 10−3 S cm−1 levels as before the experiment when measured in air (data not shown).

Figure 3
figure 3

Impedance spectra for PPy films recorded in PBS

Full size image
Figure 4
figure 4

Cyclic voltammogram of PPy–CS and PPy–HA in PBS

Full size image

Flow Cytometric Surface Marker Expression Analysis

Surface marker expression of hASCs was characterized by flow cytometric analysis. The cells used in this study expressed the surface markers CD73, CD90 and CD105 as shown in Table 2. Moderate expression was expressed by CD34, CD49d, HLA-ABC and HLA-DR whereas no expression was detected in CD14, CD19, CD45 and CD106. According to the results, hASCs expressed several of the specific antigens that verify the mesenchymal origin of hASCs.33

Table 2 Surface marker expression of the undifferentiated hASCs after primary culture in maintenance medium
Full size table

Cell Attachment and Viability

Cell attachment and viability were determined by live/dead staining which revealed extensive clustering of hASCs on PPy–HA surfaces as shown in Figs. 5e, 5f, 5g and 5h. In contrast, hASCs on PPy–CS were homogenously spread already after 1 week of culture (Figs. 5a, 5b, 5c, and 5d). This homogenous monolayer of hASCs was only evident on PPy–CS, since hASCs cultured on PS (Fig. 5i and 5j) were also sparsely spread on day 7. The increase in cell number over time was most obvious on PS, which had decidedly fewer cells than PPy–CS at both time points. As regards BEC, no notable differences in cell number were seen between the control and stimulated group with neither of the dopants. The majority of the cells were viable on both PPy films at both time points in all experimental groups.

Figure 5
figure 5

Live/dead images of hASCs on PPy–CS controls (a, b) and BEC stimulated group (c, d), PPy–HA controls (e, f) and BEC stimulated group (g, h) and PS (i, j) on days 7 and 14. Scale bar is 500 μm

Full size image

Cell Proliferation

Cell proliferation was evaluated quantitatively by measuring the total DNA content (Fig. 6). In the control group, the cell number on PPy–CS was significantly higher than PPy–HA on day 7. The increase in cell number on PPy–CS followed a similar trend at both time points under BEC compared with PPy–HA. However, no further significant differences were found. Consistently with the live/dead staining, no significant differences in cell number were found between the control and the stimulation group. The cell number increased significantly over time with both PPy–CS and PPy–HA without stimulation. In the stimulated group, only PPy–CS showed a significantly increasing cell number.

Figure 6
figure 6

Relative DNA content of hASCs cultured for 7 and 14 days on PPy–CS and PPy–HA substrates with and without BEC. PPy–CS had significantly higher DNA content than did PPy–HA in control. The cell number increased significantly over time in the control group, on both PPy–CS and PPy–HA samples. In the stimulated group, only PPy–CS showed significantly increasing cell number. The results are expressed as mean ± SD and *p < 0.05

Full size image

Osteogenic Differentiation

Both PPy–HA and PPy–CS supported hASC early osteogenic differentiation (Fig. 7) and no significant differences were detected between these two different material groups. However, ALP activity increased significantly over time only in the PPy–CS with and without stimulation. No significant differences between BEC and control groups were detected.

Figure 7
figure 7

Relative ALP activity of hASCs cultured for 7 and 14 days on PPy CS and PPy–HA substrates with and without BEC. ALP activity increased significantly over time on PPy–CS control and BEC group. The results are expressed as mean ± SD

Full size image

Under BEC, PPy–CS triggered significantly higher mineralization compared to PPy–HA (Fig. 8), whereas in the control group no significant differences were detected. Consistently with proliferation and ALP activity, no significant differences were found between BEC and the control group. One donor line did not show reliably detectable mineralization at the 14 day time point, therefore only data from two other repetitions are shown.

Figure 8
figure 8

Relative ECM mineralization of hASCs cultured for 14 days on PPy CS and PPy–HA substrates with and without BEC. PPy–CS had significantly higher mineralization under BEC compared to PPy–HA. The results are expressed as mean ± SD and *p < 0.05

Full size image

Discussion

It has recently been reported that nanoscopic roughness in combination with microscale roughness are essential for the adhesion, proliferation and osteogenic differentiation of cells of mesenchymal origin.19 Moreover, strong cell–ECM interactions enhanced by specific surface characteristics have an important role in osteogenic commitment among MSCs,14 whereas proliferation is required for effective occupation of the scaffold.

As a main finding, our results demonstrate that CS was a superior dopant to HA by triggering significantly higher proliferation of hASCs after 1 week of culture. Both PPy films supported early osteogenic differentiation of hASCs yet PPy–CS showed significantly higher mineralization than PPy–HA in the stimulation group. Importantly, hASCs cultured on PPy–CS showed typical MSC morphology and already formed a homogenous monolayer on the PPy–CS film by day 7, whereas PPy–HA triggered clustering of the cells leading to detachment from the film. Aggregation of cells is an undesired effect in osteogenic applications and more characteristic of chondrogenic differentiation as it is one of the earliest signs of chondrogenesis. This suggests that PPy–HA could be a potential scaffold coating candidate for chondrogenic applications.55

As both dopants are commonly used macromolecules in bone tissue engineering, the reason why PPy–HA caused clustering of hASC most probably lies in the physical surface properties, such as the morphology, hydrophilicity and elasticity of the PPy–HA surface. Even though the non-uniform microscopic morphology was present in both PPy–CS and PPy–HA wet films, the swollen protrusions were significantly higher in the PPy–HA, consequently reflecting the measured R a values in contact mode imaging. Similar differences between PPy–HA and PPy–CS morphology have also been reported by Gelmi et al.17 In addition, Gilmore et al. compared thin and thick PPy films, formed by different polymerization charges, and examined various dopants. They concluded that PPy–HA did not support myoblast adhesion or differentiation and the elevated R a values were one of the parameters correlating with the poor performance of PPy–HA. Greater surface thickness of PPy–HA compared to other films, such as PPy–CS, was deduced to lead to the development of greater nodules.18 Moreover, smooth surface morphology polymerized with low current densities (100–700 µA cm−2) was demonstrated to be the key parameter for MC3T3-E1 osteoblast adhesion on PPy–HA surfaces, whereas those polymerized with higher current densities (~1 mA cm−2) exhibited more irregular surface and did not ensure good cell adhesion.45 The current density in our study stayed under 700 µA cm−2 implying that surface roughness should be within suitable ranges in means of cell adhesion during the ES.

The optimal AFM imaging conditions in PBS for the PPy–CS (force setpoint 20 nN) and PPy–HA (7 nN) films were different from those measured in air, indicating that PPy–HA was softer than PPy–CS in wet state. This was also supported by the occasional adhesion of the AFM tip to the PPy–HA surface. Comparison of AFM images in wet and dry state also showed that both PPy–CS and PPy–HA films absorbed significant amounts of PBS, resulting in swelling, large dimensional changes and softening of the films to hydrogel-like materials. Earlier studies with similar films have estimated the swelling percent to be 11–25%.39,46,48 In contrast, Gelmi et al. did not detect any significant swelling of the PPy–HA and PPy–CS films.17 Furthermore, the nanoscopic textures of the dry samples were significantly different from those of the wet samples. The characteristic nodular nanomorphology was detected for both PPy films in PBS consistent with earlier reports on PPy–CS and PPy–HA.17,18,47

Hills caused by undulating morphology may result from the local detachment of the material from its metallic substrate, which has been previously observed for electropolymerized PPy–HA films on Au microelectrodes upon repeated electrochemical cycling. However, the characteristic pattern of the microscopic circular protrusions has not been observed on microelectrodes of other substrates.39

Regarding electrical properties of the films, the choice of dopant did not pose drastically different impedance either at 10 or at 100 Hz, yet the decrease in impedance was more evident for PPy–CS from 10 to 100 Hz. In contrast to our study, Gilmore et al. has earlier compared impedance of electrochemically grown thick PPy–HA and PPy–CS films reporting substantially higher impedance for PPy–HA at 10 Hz.18 Cyclic voltammetry showed slightly higher electrical activity and doping level of PPy–CS. Higher acidic strength of dopants has been shown to enhance charge carrier (bipolaron) formation in the PPy chains, hence, the stronger acidity of the sulfonic acid groups in CS compared to less acidic carboxylic acid groups in HA could potentially yield higher doping ratio in PPy–CS.32

The selected parameters of the ES in our study were based on earlier study with ASCs in osteogenic applications.25 According to Hammerick et al., ALP activity increased in mouse ASCs in response to pulsed electric field stimulation. We chose a pulsed BEC waveform since it is expected to prevent the accumulation of charged proteins and keep the pH of media at steady levels.31 In addition to our present work, only one study has examined the effect of a combination of PPy and ES on osteogenic differentiation for conducting a study with SAOS-2 cells.37 Interestingly, mineralization in our study was significantly higher in PPy–CS compared to PPy–HA in BEC group whereas no differences were detected in control group. Mineralization slightly increased in both groups, being more substantial in PPy–CS, yet not significant. This could suggest synergistic effects of PPy–CS and BEC. The reason why no significant effect of ES on ALP activity could be detected in our study is not clear. One possible reason might be donor variation in means of osteogenic differentiation, which has been reported with MSCs exposed to dexamethasone.6 In our study, one patient had clearly higher ALP activity than the two others on day 14 (data not shown). In addition, the standard deviation of the stimulated groups was, in general, slightly higher than in the control groups, which may have prevented the detection of some differences between the groups. The reason for greater deviation could be attributable to synergy between BEC and PPy coating, such as overoxidation of PPy during high electrochemical potentials that leads to the loss of conductivity,10 and de-doping or ion exchange between medium and PPy.16,27 Moreover, several factors, such as protein deposition from the culture medium and their interactions with redox reactions of the PPy surface27 or local detachment of PPy film may have affected on the homogenicity of the electrical field. To minimize the bias, three parallel samples were used in all of the assays in every patient lineage. Despite the slight increase in mineralization under BEC, more systematic screening of BEC parameters is needed to find an effective ES pattern for ASCs.

Even though PPy is considered biocompatible by several in vivo studies, the longest implantation period has only been six months,29,40,51,52 PPy is not inherently biodegradable which may pose challenges in its use in tissue engineering. It may be possible that PPy coatings do not fully erode during their use of time and therefore it needs further evaluation of the long term effects in the body. What comes to the inevitable erosion products, PPy nanoparticles have shown to be less cytotoxic compared to silver or TiO2 nanoparticles that are common wear products of orthopedic implants.30,49

Conclusion

PPy–CS supported the proliferation and homogenous spreading of hASCs significantly more than PPy–HA. Both PPy–CS and PPy–HA supported early osteogenic differentiation of hASCs over time but mineralization of hASCs was significantly greater with PPy–CS under BEC. PPy–HA is not recommended for osteogenic applications as it promotes the clustering and detachment of hASCs. This is most probably due to different surface properties since PPy–HA had a significantly rougher and softer surface than did PPy–CS in wet conditions. BEC stimulation showed no significant effects on hASC proliferation or osteogenic differentiation. This is the first study to report on the suitability of PPy–CS for bone tissue engineering applications.