Electrical stimulation

For the electrical stimulation (ES), we used a self-build “Mobini chamber” (Fig. 1A). This chamber was first described for DC ES experiments by Mobini et al. [15]. The chamber fits on top of a standard 6-well-plate. L-shaped platinum electrodes are submerged in the culture medium. Three wells are stimulated in series (stimulation), while the other three wells containe electrodes without any applied electrical stimuli (control) (Fig. 1B). The stimulation setup consists of a stimulator (ISO-STIM-II, npi electronic GmbH, Tamm, Germany) powered by a DC power supply (VOLTCRAFT FSP-1132, Conrad Electronic SE, Hirschau, Germany) and driven by a function generator (GW Instek MFG-2230 M, dataTec AG, Reutlingen, Germany). The stimulation chamber was placed in an incubator at 37 °C, 5% CO2. Voltages were monitored using an oscilloscope (digital oscilloscope DS1054, Rigol Technologies Inc., Portland, USA) and the data were recorded on a computer (Fig. 1C,D). For AC stimulation, we applied current-controlled 10 ms rectangular biphasic pulses with 6 mA and 10 ms pulse width for 2 h at 20 Hz under standard culture conditions directly after cell seeding (Fig. 1E).

Fig. 1

Electrical stimulation setup. A Bottom view of electrode lid (Mobini chamber) in a standard 6-well-plate. B Schematic view of Mobini chamber electrode setup and wiring. The symbols with lightning bolt indicate electrical stimulation of cells (red) or cell-free medium (yellow). In the parallel control approach no current was applied (symbols without lightning). C View of the stimulation setup with 1) Mobini chamber, 2) incubator, 3) circuit board, 4) power supply, 5) oscilloscope, 6) stimulator, 7) function generator and 8) laptop for data acquisition. D Schematic view of the ES setup and wiring. The function generator creates an “on-signal” at 20 Hz frequency. This is used as a trigger for the stimulator to produce the isolated biphasic 6 mA pulses. The resulting applied voltage across the three stimulated wells is measured with an oscilloscope (Ch1 = channel 1, Ch2 = channel 2, TTL trigger = timed transistor-transistor logic trigger). E Diagram of the applied current over time

Cyclic voltammetry

Electrochemical reactions at the platinum electrodes in cell culture medium were studied via cyclic voltammetry (CV) using a potentiostat (Autolab PGSTAT204, Metrohm AG, Herisau, Switzerland). To perform the CV characterisation, we fabricated a dummy device with two L-shaped Pt electrodes (wire of 0.6 mm diameter, 22 mm distance between electrodes) in a well of a 6-well plate. Therefore, all measurements correspond to a single well. The measurements were performed inside an incubator at standard cell culture conditions, 37 °C and 5% CO2. The three-electrode configuration was used. The working electrode and counter electrode were assigned arbitrarily to any of the electrodes of the stimulation device, as they are equivalent. An Ag/AgCl electrode (RE-1B, Biologic, Seyssinet-Pariset, France) with 6 mm diameter was used as a reference electrode. 4 ml fresh DMEM (Dulbecco's Modified Eagle Medium) was loaded to the system as electrolyte for each measurement. CV tests were performed within the voltage window (− 0.6 V to 0.9 V), where no reactions associated with water hydrolysis have been observed. Moreover, additional CV measurements have been carried out within the extended window (− 1.6 V to 2 V) obtained from voltage measurements. In all cases, cycles were performed at a sweep rate of 0.1 V/s and 10 mV step width. Several cycles were run until the system was stable. To determine the CV range, voltages were measured every 100 μs during 100 cycles (5 s) during stimulation. These measurements were carried out in the two electrodes configuration, using the same dummy device employed in the CV characterisation (Additional files Fig. A1).

Electric field simulation and electrochemical impedance spectroscopy

We closely followed the methods described by Zimmermann et al. for the simulation of the applied electric field [31]. Briefly, a calibrated equivalent circuit model comprising the impedance of the electrode-electrolyte interface and the cell culture medium impedance was used to predict the stimulation pulses and the voltage drop across the medium in three wells in series. The calibration data were obtained from electrochemical impedance spectroscopy (EIS) spectra. We conducted the EIS on a single well at different input voltages (10 mVrms to 1 Vrms). The impedance spectra were recorded using a Reference 620 potentiostat (GAMRY instruments, Warminster, PA, USA). The conductivity of the cell culture medium DMEM was measured with a handheld conductivity meter (SevenGo Duo SG23 with probe InLab 751-4 mm, Mettler Toledo, Gießen, Germany) and was measured to be 1.46 Sm−1 at 35.5 °C. The probe had been calibrated with conductivity standards in the same range beforehand. The electric field was computed by the finite element method (FEM) using NGSolve 6.2.2102 [32], which is an open-source library for higher-order FEM built on top of the mesh generator NETGEN [33]. The voltage drop across the medium was used to set the boundary conditions in the FEM model.

Cell culture and AC stimulation

Human osteoblast-like cells MG-63 (CRL1427™, ATCC, Manassas, USA) were cultured in DMEM GlutaMAX (Gibco, Thermo Fisher Scientific) with 10% fetal bovine serum (FBS Superior; Sigma Life Science, Thermo Fisher Scientific) and 1% gentamycin (5 mg/ml, Ratiopharm, Ulm, Germany). Cells were used from passages 5 to 30 [34]. For AC stimulation experiments, 20,000 cells/cm2 were seeded on glass coverslips in DMEM without pyruvate (Gibco, Thermo Fisher Scientific) with 10% FBS and 1% gentamycin and electrically stimulated (see above) for 2 h under standard culture conditions at 37 °C and 5% CO2 (ES cells). DMEM without pyruvate was used, as pyruvate is a scavenger of hydrogen peroxide [35] and we wanted to examine ROS.

To study medium-mediated effects, DMEM without pyruvate was stimulated as described above (ES medium) for 2 h. Cells were seeded in the ES medium shortly after stimulation and incubated for 2 h before further analysis. Cells were cultured up to 24 h after stimulation. To study long lasting effects of ES medium, the stimulated medium was stored at 37 °C, 5% CO2 for 7 days before cell seeding. For assessment of metabolic activity, cell amount, antioxidant concentration and calcium ion imaging additional experiments were conducted with 1 mM pyruvate (Sigma-Aldrich, Merck KGaA, Darmstadt, Germany) in the medium. As a control, unstimulated medium was supplemented with H2O2 (Fluka Analytical, Thermo Fischer Scientific) to a concentration of 10 μM. After ES the electrodes of the “Mobini chamber” were subsequently washed with A. dist., isopropanol (Walter-CMP GmbH & Co. KG, Kiel, Germany) and sterilised in UV light for at least 10 min each.

pH-value and temperature

To assess the impact of ES on DMEM (without pyruvate), the pH value was measured after 2 h ES. Medium containing the electrodes (without ES) served as controls. The pH meter and pH probe with sensor tip (SI series, Sentron Europe BV, Leek, Netherlands) were used outside the incubator. The temperature of the culture medium was measured using an infrared camera (ThermaCAM™, FLIR Systems AB, Danderyd, Sweden). The camera was placed inside the incubator underneath the culture plate. Measurement points were positioned in the stimulation and control samples near the electrodes. Images of the bottom of the wells were taken every 5 min during treatment.

Scanning electron microscopy and energy dispersive X-ray

Cells and medium were electrically stimulated as described above. Directly after stimulation or 2 h incubation in ES medium, cells were washed in phosphate buffered saline (PBS, Sigma-Aldrich, Merck KGaA, Darmstadt, Germany) and fixed with 2.5% diluted glutardialdehyde solution (Sigma-Aldrich, Merck KGaA). Fixed samples were washed with sodium-phosphate buffer (0.1 M), dehydrated in an ascending series of acetone and critical point dried (Emitech K850, Quorum Technologies LTD, East Sussex, UK). Samples were mounted on Al-SEM-carrier with adhesive conductive carbon tape (PLANO GmbH, Wetzlar, Germany) and coated with carbon (> 20 nm, CCU-010 HV, safematic GmbH, Zizers, Switzerland) or with gold (20 nm, EM SCD 004, BALTEC, Balzers, Liechtenstein) to reduce accumulation of electrostatic charge. Samples were analysed by a field emission scanning electron microscope (FE-SEM, MERLIN® VP Compact, Carl Zeiss AG, Oberkochen, Germany). Scanning electron microscopy images were taken from the selected regions with accelerating voltage 5 kV, working distance 5.1–5.5 mm. To image the electrodes, the “Mobini chamber” was mounted on a SEM-carrier with adhesive conductive Al-tape (PLANO GmbH). SEM-images were taken from the selected regions. Representative areas or interesting areas of the samples were analyzed for elemental composition with an energy dispersive X-ray detector (XFlash 6/30, Co. Bruker Corporation, USA) by the QUANTAX ESPRIT Microanalysis software (version 2.0). The applied detector and magnification are given in the explanations of the figures.

Cell count and viability

Metabolic activity of adherent cells 24 h after ES treatment was assessed using CellTiter 96® AQueous One Solution Cell Proliferation Assay (MTS; Promega GmbH, Walldorf, Germany). The tetrazolium compound is reduced by NADPH dependent dehydrogenases into a coloured formazan product. The amount of coloured product is therefore dependent on the metabolic activity of viable cells. To compare whether the metabolic activity of the viable cells was changed after ES, the amount of adherent cells was determined with crystal violet staining and used to normalise MTS absorbance.

As described before, 20,000 cells/cm2 were AC stimulated for 2 h and incubated at 37 °C, 5% CO2 for 24 h. Subsequently, non-adherent cells were acquired and cell count and viability were assessed using the NucleoCounter® NC3000™ with the Via1 cassette and Viability and Cell Count Assay (ChemoMetec, Allerod, Denmark). Cells were acquired and stained automatically with DNA binding acridine orange and 4′,6-diamidino-2-phenylindole (DAPI). Acridine orange is cell permeable and used to count all cells, while DAPI, which binds to the minor grove of adenine-thymine-rich regions of DNA [36], cannot diffuse across intact cell membranes and is used to stain and count dead cells in this assay. Adherent cells were incubated in MTS assay reagent diluted 1:6 in medium without pyruvate for 2 h. Absorbance was measured at 492 nm with a reference wavelength of 620 nm with a plate reader (anthos Mikrosysteme GmbH, Friesoythe, Germany). Afterwards, the cells were fixated with methanol (Sigma-Aldrich, Merck KGaA) for 10 min and stained with crystal violet (Neisser solution II, Carl Roth GmbH + Co. KG, Karlsruhe, Germany) for 15 min. Cells were washed with distilled water and the staining was extracted using 33% acetic acid (JT Baker, Avantor, Radnor, PA, USA). The absorbance was measured at 620 nm with the plate reader (anthos Mikrosysteme GmbH).

Reactive oxygen species (ROS)

After 2 h ES, H2O2 was quantified in DMEM without pyruvate using the fluorimetric hydrogen peroxide assay kit (Sigma-Aldrich, Merck KGaA) according to manufacturer’s protocol. H2O2 is thereby detected via the reaction with a red peroxidase substrate through horseradish peroxidase and the generation of a fluorescent product. Fluorescence was measured with the fluorescence reader (ex/em: 540/590 nm) (infiniteM200, Tecan Group Ltd., Männedorf, Switzerland) before and after stimulation and H2O2 concentration was calculated from standard curve.

Intracellular ROS were detected using the 20,7′-dichlorofluorescein diacetate (DCFDA) assay (Abcam, Cambridge, UK) according to manufacturer’s protocol. DCFDA is a membrane-permeable compound that can enter the cells. Inside, cytoplasmic esterases deacetylate the compound into the non-fluorescent DCF. The dye is then oxidised by intracellular ROS, which makes it highly fluorescent. For our experiments, 1 × 106 cells/ml were stained in suspension with 20 μM DCFDA for 30 min in the dark at 37 °C. Afterwards, cells were seeded in DMEM without pyruvate and then directly AC stimulated for 2 h. Fluorescence intensity was measured at 485 nm with a fluorescence reader (infiniteM200) before and after stimulation. The fluorescence of non-adherent cells was measured using flow cytometry (FACSCalibur, BD Biosciences, Ann Arbor, MI, USA). To differentiate ROS signal locations, adherent cells were stained after ES with 5 μM CellROX green (ROS in nucleus and mitochondria), 5 μM CellROX orange (ROS in cytoplasm) (Invitrogen GmbH, Darmstadt, Germany) and 9 μM Hoechst (Life Technologies, Thermo Fisher Scientific) for 30 min at 37 °C in the dark. 180 μM H2O2 served as positive control. After incubation the samples were washed with PBS and covered with medium without pyruvate. Images were acquired at random positions in the well and 10 cells per image were analysed using the Zen blue software (LSM 780, 40x/1 W Plan-Apochromat water objectiv, Carl Zeiss Microscopy GmbH, Jena, Germany; Software acquisition: Zen black 2.1 V 11.0.0.190; Software analysis: Zen 2.3 blue edition, Carl Zeiss Microscopy).

Antioxidant assay

Intracellular ROS levels are tightly regulated by antioxidants such as ascorbate, glutathione and catalase [37]. In cell culture medium, 1 mM pyruvate is often added, which serves as ROS scavenger. Unspecific antioxidant concentration in cells was measured using the antioxidant assay kit (Cayman chemical, Ann Arbor, MI, USA). In short, the combined antioxidant ability of the sample to suppress the oxidation of ABTS (2,2-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) by metmyoglobin is compared to a Trolox standard curve.

Stimulated cells were collected in cold assay buffer (Cayman chemical) using a cell scraper. The cells were homogenised by sonication at 35 kHz (Sonorex RK 100, Bandelin electronic, Berlin, Germany) for 5 min. Samples were then centrifuged at 10,000 x g at 4 °C for 15 min (Eppendorf centrifuge 5417R, Eppendorf AG, Hamburg, Germany). Subsequently, the supernatant was collected and stored at − 80 °C. For the assay, samples were thawed and pipetted into a 96-well-plate. ABTS and metmyoglobin were added and the reaction was started by the addition of H2O2. Samples were incubated for 5 min and absorbance was read at 750 nm with a fluorescence reader (infiniteM200, Tecan Group Ltd., Männedorf, Switzerland).

Aquaporins (AQP)

For general AQP expression, cells were cultured in DMEM with 1 mM pyruvate and subsequently processed. MG-63 cells were washed with PBS, trypsinised, and centrifuged (Centrifuge 5810 R, Eppendorf AG) for 5 min at 200 x g. The cells were then fixed with 4% PFA (in PBS, Sigma-Aldrich) at room temperature (RT). Cells were permeabilised with 0.1% Triton X-100 (Sigma-Aldrich) for 10 min at RT. After washing, incubation with 2% FBS (FBS Superior; Sigma Life Science, Thermo Fisher Scientific) in PBS was performed for 1 h to prevent non-specific binding of the primary antibodies. Subsequently, cells were washed and incubated with the primary antibodies at the given dilutions overnight at 4 °C (Table 1). The following day, cells were washed and incubated with secondary antibodies for 30 min at RT. Excess antibody was removed and cells were resuspended in 400 μl PBS and stored at 4 °C until analysis. Flow cytometry was performed using the FACSCalibur and Cellquest Pro software (4.0.1). FlowJo software (10.4.2; BD Biosciences) was used for analysis.

Table 1 Antibodies for aquaporin staining

Images of AQP 1 were acquired after fixation with ice-cold methanol (99.9%, Sigma Aldrich) for 10 min at 4 °C and permeabilised with Triton X-100 for 10 min at RT. Incubation with FBS (2% in PBS) was performed for 1 h, followed by addition of primary AQP 1 antibody (Table 1) and incubating overnight in a wet chamber at 4 °C. The samples were washed again with PBS and secondary antibodies were added for 30 min at RT. Fluoroshield mounting medium with DAPI (Fluoroshield™, Sigma-Aldrich) was added to the bottom of the cell container and covered with a coverslip (Menzel GmbH, Braunschweig, Germany). Samples were then stored at 4 °C until analysis with the LSM 780 confocal laser scanning microscope (Carl Zeiss). Images were taken with the Plan-Apochromat 63x oil immersion objective.

Intracellular calcium ion (Ca2+) imaging

To evaluate the effects of AC stimulation on MG-63 s’ intracellular Ca2+ levels, 20,000 cells/cm2 were seeded in 6-well glass bottom plates (Cellvis, Mountain View, CA, USA). Cells were stimulated for 2 h directly after seeding as described before. After the stimulation, adherent cells were stained with 5 μM Ca2+ sensitive dye Fluo-3 acetoxymethyl ester (AM) (Life Technologies, Thermo Fisher Scientific) for 30 min in HEPES (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid) buffer (hypotonic and isotonic 1:1, Table 2) [38]. Afterwards, cells were resuspended in isotonic HEPES buffer. For assessment of ES medium on cellular Ca2+ levels, DMEM without pyruvate alone was stimulated for 2 h. MG-63 cells were stained with 5 μM Fluo-3/AM in HEPES buffer and seeded in ES medium. Fluorescence images were acquired using a confocal laser scanning microscope (LSM 780, Carl Zeiss Microscopy) with a Plan-Apochromat 63x oil immersion objective (Carl Zeiss Microscopy) and the Zen software (Zen 2.3 SP1 FP3 black, Carl Zeiss Microscopy). Fluorescence intensity was measured using the Zen software (Zen 2.3 blue edition, Carl Zeiss Microscopy). In well-spread cells the Fluo-3 dye and thus the calcium signal is spread over a greater area than in spherical cells. Therefore, bright field images of the LSM were used to determine the cell shape and area. Mean fluorescence intensity (MFI) of Fluo-3 was measured over the whole cell area (Fig. 2).

Table 2 HEPES buffer compositionFig. 2

Image analysis of intracellular Ca2+: Mean fluorescence intensity (MFI) of MG-63 osteoblasts stained with Ca2+ sensitive dye Fluo-3. Cell shape was determined using light microscopy images from LSM. Then, MFI of Fluo-3 was measured over the whole cell area. (LSM 780, Carl Zeiss, scale bars 20 μm)

Statistical analysis

Statistical analysis was performed using Graph Pad Prism 7.02 (GraphPad Software Inc., USA). The graphs show mean values with the standard error of mean (s.e.m). Data was tested for Gaussian distributions using the Shapiro-Wilk normality test. For comparison of multiple samples One-way ANOVA was used and for data, grouped by two or more factors, the two-way ANOVA was applied, both with Bonferroni post-hoc-tests. Where paired data was available for all data points, paired tests were used (RM-ANOVA). Wilcoxon matched-pairs signed rank test was used to test multiple, non-parametric, paired samples. Statistical tests are described in the figure captions. Significance levels are depicted as *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001 and ****p ≤ 0.0001.