Most commonly the material of carbon nanotubes obtained by arch discharge or laser evaporation of graphite or by chemical decomposition of carbonaceous vapor is a powder that should be purified, filtered from undesirable impurities (such as amorphous carbon, defective nanotubes, catalyst particles), as well as functionalized, in some cases, to obtain nanotube properties required for the specific application. However in this case, preparation of solutions is needed. It is well-known that nanotubes can be dissolved in most polar and nonpolar solvents, so to dissolve SWNTs obtained in the arc discharge, it is necessary to split bunches and bundles of them which were formed in the process of growth. The most effective way to do that is additional functionalizetion of nanotubes by surfactant materials (SM), which allows a stable colloidal solution of single nanotubes of any concentration to be achieved. These SM solutions are ideal to form thin percolated films on any surfaces. By regulating nanotube concentration in the solution one can regulate the thickness of the film to be formed on the surface. As we have shown previously [13], the nanotube orientation and density on the surface can be controlled while using different deposition techniques: dipping, pulling, centrifugation, sputtering.

To prepare conductive composites based on nanotubes one can use their mixtures with proteins as a natural polymer providing a biocompatible substrate and improved adhesion of cells [12]. Albumin is one of the most important transport proteins, which regulates metabolism functions in cell growth and tissue regeneration. In this research we suggest using BSA protein molecules as a surfactant to form a biocompatible conductive layer on the cover-slip. Previously, a composite consisting of nanotubes and albumin was examined as a scaffold-forming material in implants in cartilaginous tissue regeneration [14,15].

2.5 mg of 99.5 mass % SWNT provided by A. V. Krestinin (Institute of Problems of Chemical Physics of RAS, Moscow, Russia) were placed into a BSA aqueous solution (10 mg of BSA in 5 mL of water) and dispersed in a Branson B300 (Branson Ultrasonics, Danbury, CT, USA) ultrasonic bath (34 kHz, 50 W) for 10 hours. Typical length of nanotubes was less than 1000 nm after ultrasonic treatment. Nevertheless the CNT combines in bundles up to 5 mm long and about 10 nm in diameter. Cover-slips, 24 × 24 mm in size, 0.13 - 0.17 mm thick were preliminary mechanically washed with cotton in 2-propanol, then they were kept in 2-propanol for 15 minutes and placed in ultrasonic bath. On one of the surfaces of the cover-slip two gold contact pads, 30 nm thick, were formed by magnetron sputter deposition (Emitech K575X, Quorum Technologies, Ringmer, UK). About 25 mL of nanotube solution in albumin were applied onto this surface with a microdispenser and a thin film was made to cover the whole slip surface by rodcoating method, then the film dried for 15 minutes at 40˚C. To improve the film adhesion and conductance the structure was annealed in the air at a temperature of 150˚C for 2 minutes. Figure 1 shows the photography of a cover-slip with the current-conductive SWNT/BSA coating formed. The resistance of the structures investigated in the work ranged from 100 kOhm to 10 MOhm, but in the experiment for proliferation estimation at low stimulation voltage we’ve used a set of samples with close resistance (several MOhm).

It should be noted that the samples of SWNT/BSA film obtained not only possessed conductance but also transparency within the visible range at a level of 85% - 90%.

To carry out experiments with electrical stimulation of

cells, a custom system for supplying electrical signals which comprises a 6-well plate, a signal generator and an oscilloscope has been developed. The plate cover has a breadboard with a connector fixed on it. Over each of these 6 wells two sharp needle electrodes with diameter 0.3 mm made of a 40KHNM-VI biocompatible alloy with a gold electrolytic coating (Doriva, Moscow, Russia) were placed. The electrodes were connected in pairs in 2 rows of 3 wells each to provide lead output to the signal generator connector. On the one hand, this ensured biological compatibility with the medium, on the other, the contact area of the whole surface of the needle with the cultural medium was much smaller than the area of the modified cover-slip (~40 times), which minimized the current contribution through the medium. In each set we’ve used 5 wells for experiment and one for control (Figure 2). To apply voltage we used a RIGOL DG1022 generator (Rigol Technologies Inc., USA).

The generator can be set to arbitrarily shaped stimulatory signal, as well as to produce two independent signals in parallel. Due to the modified cultural plate design, the system was able to send a signal directly to the cells through the SWNT/BSA film. The voltage generator was connected with the plate by means of a flexible cable with the core diameter of 0.1 mm, which could be easily placed in the CO₂ incubator. Figure 2(b) shows the schematic diagram of the connection.

For each experiment we used five cover-slips covered by SWNT/BSA film: two slips—to minimize statistical error for proliferation measurements, one—for morphology investigation, one—for microscopy analysis, one was cultured without voltage applied and used as control. Also there was one pristine cover slip with voltage applied through the medium.

HEF—human embryonic fibroblast cells were provided by the Tissue Cultures Laboratory of Ivanovskiy Institute of Virology, a Federal State Institution of the Ministry of Health and Social Development (Moscow, Russia), were cultured in the electrical stimulation unit (Figure 2(a)). Linbro 6-well plates (MP Biomedicals, Solon, OH, USA) were used as a culture dish. The unit, plates, and coverslips were sterilized in 70% ethanol solution for 10 minutes with the follow-on 20-minute ultraviolet treatment. Immediately prior to the incubation cover-slip samples were rinsed in the Eagle culture medium for 10 min to eliminate contaminants. After removing this medium the HEF cells in the concentration of 10⁵ cells/mL were added into each well in the amount of 1 mL of the Eagle medium with 10% FBS (fetal bovine serum) and incubated for 24 hours in a thermostat with 5% CO₂ at 37˚C.

After a 24 hour incubation with no voltage applied, a

signal of 5 pulses was sent with 5 ms pulse duration, 5 ms spacing between pulses, 1 s interval between pulse groups. The main pulse characteristics were chosen with previous experiments in electrical stimulation of growing fibroblasts with standard two carbon electrode unit taken into consideration [16]. We have done 3 sets of experiments with different amplitudes 10, 50, 100, 200, 500, 5000 mV applied to the cells. Since the resistance of wires and electrodes is relatively small, the main voltage drop through SWNT/BSA film was expected. Despite the fact that the resistance of the fabricated SWNT/BSA film samples varied, in the experiment, for proliferation estimation we used samples of the same order of resistance (from 3 to 10 MOhms). The electrical stimulation lasted for 48 hours. In addition, we conducted an experiment where the current was applied to needle electrodes placed into culture medium with cover-slips in the absence of a conductive substrate. As expected, the size of the electrodes was relatively small for a significant current to leak through the medium.

The proliferation was defined by means of a modified MTT assay. For this purpose, the Eagle medium with 10% FBS was removed after electrical stimulation, the cover-slips were moved to a clean plate, where 1 mL of the Eagle medium and 200 ml of MTT solution (with the initial concentration of 5 mg/mL) were added. The coverslips were incubated in the thermostat with CO₂ at 37˚C for 4 hours. Then the medium with MTT removed and 1 mL of DMSO was added into each well to dissolve MTT-formazan reduced by the cells. Cell precipitates with MTT-formazan on glass samples were resuspended for 5 min and the solution absorbance was measured by a Titertek Multiscan Plus photometer (Flow Laboratories, Helsinki, Finland) on a 492 nm wavelength. To achieve that, 1 mL of DMSO with the dissolved formazan was transferred to a 96-well plate, with a 100 mL per well dosage. A structure with gold electrodes and a SWNT/ BSA film applied, which was in one of the wells of a 6- well plate, but without voltage supply was used as a control.

The cover-slips with a SWNT/BSA coating were investigated with atomic force microscope Solver P47 (NT-MDT, Moscow, Russia) in a semi-contact mode before and after the cell growth. For AFM measurements samples were removed from culture medium, fixed in 2.5% glutaraldehyde for 30 minutes and washed in the phosphate buffer saline (PBS) 2 - 3 times, 2 minutes each, then dehydrated in 50%, 70% and 96% solution of ethanol for 2 min in each one. Some of the samples after electrical stimulation were stained with azur-eosine for morphology study.