Serum-deprived differentiated neuroblastoma F-11 cells express functional dorsal root ganglion neuron properties

Cell cultures

F-11 cells (mouse neuroblastoma N18TG-2 x rat DRG, ECACC Cat#08062601 RRID:CVCL_H605; Platika et al., 1985) were seeded at 60,000 cells/35 mm dish and were maintained without splitting in low serum medium for almost 2 weeks to induce differentiation. The complete composition of the medium was: Dulbecco’s modified Eagle’s medium (Cat#D6546; Sigma-Aldrich, St. Louis, MO, USA), 2 mM glutamine (Sigma-Aldrich, St. Louis, MO, USA), 1% fetal bovine serum (FBS, Cat# F2442; Sigma-Aldrich, St. Louis, MO, USA). The cells were incubated at 37 °C in a humidified atmosphere with 5% CO₂, receiving fresh medium twice per week. F-11 cells maintained in 10% FBS medium were used as control. Morphological and functional analysis were performed at 10–14 days of differentiation, whereas cells growing in the non-differentiation medium underwent the same analysis at 7 days to avoid dramatic cell death due to confluence.

Immunofluorescence

F-11 cells were plated at a density of 60,000 cells on coverslips pre-treated with gelatin from porcine skin (Sigma-Aldrich, St. Louis, MO, USA). Cells were then maintained in DMEM medium supplemented with 1% or 10% serum. After 10 and 7 days, respectively, differentiated and undifferentiated cells were fixed for 10 min in 3.7% paraformaldehyde in phosphate-buffer saline (PBS), permeabilized for 4 min with 0.1% Triton X-100 in PBS, and stained with monoclonal anti-NeuN/Fox3 produced in mouse primary antibody (1:150, Cat#MAB-94161; Immunological Sciences, Rome, Italy). Incubation with secondary antibody Cyanine3 goat anti-mouse IgG (H+L) (1:200, Cat#A10521, RRID: AB_2534030; Thermo Fisher, Waltham, MA, USA) was maintained for 45 min. After washout in 1× PBS, DAPI (Sigma-Aldrich, St. Louis, MO, USA) was added at the final concentration of 1 μg/ml in 1× PBS and incubation was maintained for 10 min at room temperature. After washing, the slides were mounted and photographed using an A1RNikon (Nikon, Tokyo, Japan) laser scanning fluorescence confocal microscope at 40× magnification. A total of 16–20 z-stack images from 10 different fields for each condition were taken. For this analysis three coverslips of cells maintained in 10% serum and three coverslips of cells in 1% serum were prepared. Transmission images were captured with a Leica TCSSP2 confocal microscope equipped with a 100×/PH3 oil immersion objective. The images were acquired from three cultures of cells maintained in 10% serum and from three cultures of cells in 1% serum.

Functional analysis by the patch-clamp technique

Functional characterization of the electrophysiological properties of F-11 cells was performed by the patch-clamp technique in the whole-cell configuration. For the experiments, culture media were replaced by a standard extracellular solution which contained (mM): NaCl 135, KCl 2, CaCl₂ 2, MgCl₂ 2, hepes 10, glucose 5, pH 7.4. The standard pipette solution contained the following (mM): potassium aspartate 130, NaCl 10, MgCl₂ 2, CaCl₂ 1.3, EGTA 10, hepes 10, pH 7.3. Recordings were acquired by the pClamp8.2 software (pClamp, RRID:SCR_011323) and the MultiClamp 700A amplifier (Axon Instruments; Molecular Devices, LLC., San Jose, CA, USA). Resting membrane potential (V_rest) and APs were monitored in the current-clamp mode. Cells that did not exhibited spontaneous firing were depolarized with 1 s-long current pulses under conditioning hyperpolarization at −75/−80 mV to verify their capability to generate repetitive spiking. In the voltage-clamp mode, series resistance errors were compensated for a level of up to 85–90%. Sodium (I_Na) and potassium (I_K) currents were recorded by applying a standard protocol: starting from a holding potential of −60 mV, cells were conditioned at −90 mV for 500 ms and successively tested by depolarizing potentials in 10 mV-increments, from −80 to +40 mV. Na⁺ and K⁺ currents were isolated by applying 0.3–1 μM tetrodotoxin (TTX) or 10 mM tetraethylammonium (TEA) in the bath. To determine current densities (I_Na and I_K), the maximal inward and outward current intensities were respectively chosen for Na⁺ and K⁺ currents. Na⁺ channel biophysical properties of activation and inactivation were studied by using a pipette solution containing (mM): CsF 105, CsCl 27, NaCl 5, MgCl₂ 2, EGTA 10, hepes 10, pH 7.3. The voltage-dependence of activation was determined by the above mentioned protocol, whereas for the inactivation properties the protocol consisted in a conditioning step with amplitude from −105 to 0 mV and duration of 600 ms, followed by a test at −10 mV.

Ether-à-go-go-related gene (ERG) potassium currents (I_erg) were recorded by using an extracellular solution containing a higher K⁺ concentration (40 mM), and by imposing a standard protocol which, starting from an holding potential of −60 mV, conditioned the cell membrane for 15 s from −80 to +20 mV (20 mV increments) and successively hyperpolarized at −120 mV to evoke the tail current. Normalized tail currents were interpolated with a Boltzmann function to obtain the activation curve. To study the voltage-dependence of ERG channel inactivation, a stimulation protocol was applied which, starting from a holding potential of 0 mV, applied voltage steps with duration of ~200 ms and amplitude from +20 mV to −140 mV. ERG potassium current was analyzed by using WAY-123,398 (1 µM) as selective blocker and CsCl (5 mM) as further inhibitor.

Ba²⁺ currents through voltage-dependent Ca²⁺ channels were recorded under conditions which suppressed Na⁺ and K⁺ currents: the culture medium was replaced by an external solution composed by TEA-chloride 130 mM, BaCl₂ 10 mM, MgCl₂ 1 mM, hepes 10 mM, TTX 1 μM, glucose 10 mM, pH 7.3; the internal pipette solution contained CsCl 140 mM, MgCl₂ 4 mM, hepes 10 mM, EGTA 10 mM, Na-ATP 2 mM, pH 7.2. To examine the current-voltage relationship (IV) of I_Ba, the cells were depolarized with increasing 10 mV steps from −60 to +50 mV. Moreover, to analyze the contribution of high threshold (I_Ba(high)) and low threshold (I_Ba(low)) currents, cells were stimulated by two different protocols: I_Ba(high) were elicited by depolarizing pulses to 0 mV for 150 ms from an holding potential of −80 mV, whereas for I_Ba(low) the holding potential was imposed at −90 mV and the test potential was clamped at −50 mV. Addition of 200 µM CdCl₂ (Sigma-Aldrich, St. Louis, MO, USA) in the extracellular solution confirmed that I_Ba flowed through voltage-dependent calcium channels, and nifedipine (5 µM, Sigma-Aldrich, St. Louis, MO, USA) was used to isolate the component of I_Ba through L-type voltage-dependent calcium channels.

Acetylcholine (ACh) currents were evoked by applying ACh (1 mM, Sigma-Aldrich, St. Louis, MO, USA) and were inhibited by the nicotinic ACh receptor antagonist d-tubocurarine (DTC, 1 µM, Sigma-Aldrich, St. Louis, MO, USA). Glutamate (1 mM, Sigma-Aldrich, St. Louis, MO, USA), CNQX (6-cyano-7-nitroquinoxaline-2,3-dione, 10 µM, Tocris, Bristol, UK) and AP5 ((2R)-amino-5-phosphonovaleric acid, 40 µM, Tocris, Bristol, UK) were bath applied to evaluate the functional expression of α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) and N-methyl-D-aspartic acid (NMDA) receptors, respectively. To allow NMDA receptor activation, Mg²⁺-free extracellular solution was used.

To study acidic condition responses mediated by transient receptor potential vanilloid 1 (TRPV1) non-selective cation channel and acid-sensing ion channels (ASIC), cells were superfused with a solution containing (mM): NaCl 135, KCl 2, CaCl₂ 2, MgCl₂ 2, MES 10, glucose 5, pH 5, or pH 6. To investigate the expression of proteins involved in nociception, capsaicin (CAPS, 3 µM, Sigma-Aldrich, St. Louis, MO, USA) and substance P (SP, 2 µM, Sigma-Aldrich, St. Louis, MO, USA) were applied.

Data analysis

Patch-clamp experiments were performed on a minimum of two and on a maximum of twenty independent cultures for each condition. For the analysis, Origin 8 (RRID:SCR_014212; Microcal Inc., Northampton, MA, USA) and Excel (Microsoft, Redmond, WA, USA) were routinely used. Data are presented as mean ± s.e.m. Mean comparisons were obtained using the unpaired t-test or the non parametric Mann–Whitney test. The number of responsive cells in the two conditions was compared using the χ² test. The significance level was set for p < 0.05.

Neuronal differentiation of neuroblastoma F-11 cells

After 12–14 days in 1% FBS medium, F-11 cells stained positively for the neuronal nuclear protein NeuN (Fig. 1) and about 50% of the culture was characterized by neuronal networks of cells exhibiting typical neuronal morphology. When 1% FBS cultures were analyzed by the patch-clamp technique, only cells with neuronal morphology showed electrophysiological properties characteristic of mature neurons (Fig. 2). These cells, defined as “differentiated cells” throughout the article, compared to cells maintained in 10% FBS medium (“undifferentiated cells”), had more hyperpolarized resting membrane potentials (V_rest: −50.5 ± 1.9 mV vs. −17.1 ± 3.8 mV), and exhibited increased sodium and potassium current densities (for I_Na: 114 ± 10.2 pA/pF vs. 42.5 ± 15 pA/pF; for I_K: 181.4 ± 17.9 pA/pF vs. 40.9 ± 5.5 pA/pF). Moreover, a significantly higher percentage of cells was able to fire induced or spontaneous APs. Cells endowed with APs were 85% in differentiating conditions vs. 13% in control conditions (χ² test); moreover cells with spontaneous spiking reached 61% vs. 18% (χ² test) (Figs. 2E and 2F). Therefore, we investigated in the differentiated cells the presence of ion channels expressed in DRG neurons.

Expression of voltage-dependent sodium and potassium channels in differentiated cells

Sodium currents were fast and completely blocked by 1 μM TTX, indicating that differentiated F-11 cells did not express TTX-resistant sodium currents which are conversely present in some classes of DRG neurons. Activation and inactivation properties were consistent with those of TTX-sensitive currents characterized in small DRG neurons by Cummins & Waxman (1997) (for activation: V_1/2 = −22 ± 0.5 mV, k = 6.2 ± 0.4 mV, n = 5; for inactivation: V_1/2 = −68 ± 2 mV, k = 5 ± 1 mV, n = 7) (Figs. 3A and 3B). Potassium current kinetic and voltage-dependence (Fig. 3A) were consistent with delayed rectifier potassium currents. Potassium current amplitude was reduced of 84% ± 1% by 10 mM TEA administration (n = 17). F-11 cells also expressed ERG potassium current I_erg (Figs. 3E–3G), as already referred for undifferentiated F-11 cells in Faravelli et al. (1996) and for cells differentiated in retinoic acid by Chiesa et al. (1997). In our conditions, I_erg current density significantly increased in differentiated compared to undifferentiated cells (42 ± 9 pA/pF, n = 8, vs. 14 ± 2 pA/pF, n = 14). Thus voltage-dependence of activation and inactivation were also compared. V_1/2 and k values for activation and inactivation in undifferentiated cells were: −29.7 ± 2.4 and 9 mV (n = 6); −64.8 ± 4.4 and 18 mV (n = 7), respectively. Activation properties did not change in differentiated cells and were V_1/2 = −32 ± 3 mV and k = 5 mV, n = 12. Concerning the voltage-dependence of inactivation, V_1/2 was −56 ± 3 mV and k was 12 mV, n = 4. I_erg currents from both undifferentiated and differentiated cells were almost completely inhibited by WAY-123,398 (block fractions were respectively, 77% ± 4%, n = 6, and 71% ± 6%, n = 9).

Moreover, they were also blocked by 5 mM Cs²⁺ (mean inhibition was 70% for both undifferentiated and differentiated cells). Cs²⁺ had no effects on the outward potassium components, which conversely were almost completely inhibited by 10 mM TEA (Fig. 3H).

Barium currents through voltage-dependent calcium channels

In DRG neurons all the different types of voltage-dependent calcium channels have been described, but low voltage-activated calcium currents (I_Ca(low)) have been identified only in small and medium diameter cells (Scroggs & Fox, 1992). To verify the functional expression of calcium channels in F-11 cells, whole-cell Ba²⁺ currents were recorded under conditions which suppressed Na⁺ and K⁺ currents, by adding TTX and TEA in the extracellular solution, and CsCl in the patch-pipette (see Materials and Methods for details). The I–V curve was determined by measuring the peak current evoked at potentials from −60 and +50 mV. It showed a peak between −20 and −10 mV in undifferentiated cells (n = 9) and around −10 mV in differentiated cells (n = 8) (Figs. 4A and 4B). To discriminate between high threshold (I_Ba(high)) and low threshold-activated currents (I_Ba(low)), two different protocols were applied as described in Materials and Methods. Test at 0 mV from a holding potential of −80 mV could evoke Ba²⁺ currents in 15 out of 21 undifferentiated cells (74%); the mean current amplitude was 107 ± 26 pA (current density 3 ± 0.4 pA/pF) (Fig. 4C). In differentiated cells currents had mean amplitude of 203 ± 44 pA (current density 5 ± 1 pA/pF, recorded in 16 out of 24 cells, Fig. 4C). Test at −50 mV from a holding potential of −90 mV evoked responses in neither undifferentiated nor differentiated cells (13 and 10 cells tested, respectively), indicating that low voltage-activated Ca²⁺ channels were not expressed. Cadmium application (200 µM) completely blocked I_Ba in all the differentiated cells (n = 10) and in all the undifferentiated cells (n = 11). The L-type voltage-gated calcium channel antagonist nifedipine (5 µM) blocked high threshold currents equivalently in differentiated (63% ± 7% of block, 7 out of 7 cells) and undifferentiated cells (51% ± 11% of block, 5 out of 5 cells) (Figs. 4D–4F).

Capsaicin

Capsaicin, the pungent ingredient of the hot chili pepper, is the agonist of the transient receptor potential vanilloid 1 (TRPV1) non-selective cation channel, highly expressed in DRG sensory neurons (Goswami et al., 2006; Masuoka et al., 2017). In undifferentiated cells, 3 μM CAPS did not evoke any response (n = 12). In differentiated cells, it induced appreciable inward currents (≥20 pA) in 13 out of 62 cells (21%). Mean current amplitude was 41 ± 9 pA (current density: 1.2 ± 0.3 pA/pF, n = 13). The effects of CAPS are shown in Figs. 5A–5E. In our study, no correlation between responses and cell capacitance was evident (mean capacitance of responsive cells: 38.2 ± 2.7 pF; mean capacitance of non-responsive cells: 38.2 ± 4.4 pF).

Substance P

Substance P modulates the excitability of sensory neurons in pain pathways. In DRG neurons it can increase or decrease excitability, by modulating ligand-gated channels including P2X3 ATP receptors, TRPV1 CAPS receptors and ASIC3 channels, as well as several types of voltage-gated channels (sodium, calcium and potassium channels, and the hyperpolarization-activated I_h current). In our experiments, undifferentiated cells did not respond to SP (n = 12). On the contrary, in differentiated cells, SP depolarized the cells of about 12 mV (from −48 ± 4 to −36 ± 4 mV, n = 9). No cell tested underwent hyperpolarization in presence of the substance. When tested in the voltage-clamp mode, SP promoted small inward currents in 11 cells. Mean current amplitude calculated for currents ≥20 pA was 24.5 ± 3.3 pA (n = 4) (Figs. 5F–5J). Responses to SP were recorded in 79% of cells (15 out of 19) overall. No correlation between responses to SP and cell capacitance was evident.

Acidic solutions

Acid-sensing channels are expressed by neurons throughout the nervous system and are involved in acidotoxicity related to several pathological conditions and the perception of pain. F-11 cells maintained at holding potential of −70 mV responded to the application of acidic (pH 5 and pH 6) solutions with fast inward currents (Figs. 5K–5T) which were recorded in <40% of undifferentiated cells (pH 6: 31%, 5 out of 16 cells, mean current: 8.6 ± 1.8 pA; pH 5: 37%, 6 out of 16 cells, mean current: 11.6 ± 2.8 pA). On the contrary, they were evoked in all the differentiated cells with a mean current amplitude of 1,021 ± 181 pA (n = 22) at pH 6 and of 931 ± 131 pA (n = 32) at pH 5.

Acetylcholine and nicotinic acetylcholine receptors

Functional nicotinic acetylcholine receptors (nAChRs) have been described in heterogeneous populations of dissociated rat and mouse DRG neurons (Smith et al., 2013) and they are known to be involved in pain modulation. The pathway in which nAChRs operate to modulate pain is actually of great interest since it has been suggested that the anti-allodynic effect of their agonists may have a peripheral component (Rueter et al., 2003). In undifferentiated cells, ACh evoked currents in 14 out of 14 cells, with mean amplitude 43.6 ± 12 pA at −70 mV. In differentiated F-11 cells, currents were recorded in 89% of cells and displayed a significantly higher mean amplitude (136 ± 35 pA, n = 34 out of 38). In differentiated cells the effects of ACh administration was also evident on the resting membrane potential (mean depolarization: +44% ± 8%, n = 20). Since ACh-evoked currents were completely inhibited by 1 μM DTC, they were consistent with nAChRs (Figs. 6A–6D). Approximately the same percentage (70–80%) of rat DRG primary neurons was referred to express functional nAChRs in Genzen, Van Cleve & McGehee (2001).

Glutamate receptors

The localization of AMPA, Kainate, and NMDA receptor subunits has been demonstrated in rat DRGs by immunohistochemistry and in situ hybridization histochemistry, suggesting that the glutamatergic system plays an important role in the primary sensory afferent systems (Sato et al., 1993). In our experiments 1 mM glutamate was effective on only 1 out of 13 undifferentiated cells, but on 28 out of 33 differentiated cells (85%). In the current-clamp mode, it depolarized differentiated cells of 27% ± 5% (n = 14), and in voltage-clamp recordings it evoked currents with amplitude ranging from 20 to 227 pA (mean amplitude: 77 ± 25 pA, n = 10) (Figs. 6E–6H). Currents were inhibited by 71% ± 9% during CNQX and AP5 co-application. However, glutamate administration in the Mg²⁺-free extracellular solution and at the holding potential of −40 mV evoked no AP5-sensitive currents, demonstrating that differentiated F-11 cells expressed non-NMDA receptors prevalently.