The isolation procedure was done according to the animal experimentation license 22.1/354/3/2011 (issued by the National Food-chain Safety Office) and was executed as described before (Markó et al., 2011). Briefly, timed pregnant B6;129S-Gt(ROSA)26Sor^{tm32(CAG-COP4∗H134R/EY FP)Hze}/J mice (Madisen et al., 2012) were sacrificed on 17.5 post-conception day by intraperitoneal injection of 100 μg/g ketamine (CP-Pharma mbH, Germany) and 10 μg/g xylazine (CEVA-Phylaxia, Hungary). Brains of the embryos were aseptically removed, telencephali were isolated and placed into sterile phosphate buffered saline (PBS). Under stereomicroscope the meninges were removed and the tissue was manually cut into small pieces (∼1 mm³). The pieces were suspended in DMEM (Sigma) and triturated with Pasteur-pipette. The cell suspension was filtered through a nylon mesh with pore diameter of 45 μm. The cells were sedimented by centrifugation (120 g; 10 min) and the pellet was quickly re-suspended in DMEM/F12 (1/1) medium (Sigma) containing 1% B27 (Gibco). The cells (6 × 10⁶ cells/cm²) were plated into 60 mm Petri-dishes (Falcon) coated with a synthetic AK-cyclo[RGDfC] peptide (Markó et al., 2008) at 0.25 μg/cm² surface density. The culture medium was supplemented with EGF (20 ng/ml; Peprotech), as the only exogenous growth factor in addition to insulin present in B27. The culture medium was changed every second day and the cells were harvested by trypsinization when reached confluency. After 3–4 passages, cultures were composed by virtually homogeneous populations of proliferating radial glia-like cells (Markó et al., 2011). The radial glial nature of cultivated cells was checked by PCR studies on gene expression according to Markó et al. (2011) and by immunocytochemical investigations of cell-type specific markers (see Supplementary Material S1).

Radial glia-like cells containing loxpStoploxpChR2(H134)-EYFP construct were grown in 4-well plates (5 × 10⁴ cells/well) and were transfected with 1 μg pTurbo-Cre plasmid [Cre expression driven by chicken-β-actin promoter (Hug et al., 1996)] using Lipofectamin-Plus reagent (Invitrogen-Thermo Fisher Scientific) according to the manufacturer’s instructions. After the cells were transfected with pTurbo-Cre the cultures were protected from light using aluminum foil shielded culture dishes and a microscope (Zeiss Axiovert 200M) with blue-light filter excluding light with λ: 380–500 nm.

On the 3rd post-transfection day, enhanced Yellow Fluorescent Protein (eYFP)-positive and negative cells were separated using fluorescence activated cell sorting. About 1 - 1.5 × 10⁶ cells were harvested by trypsinization and collected in 2 ml “sorting buffer” (1 mM EDTA and 0.4% bovine serum albumin in PBS). Aliquots of the cell suspensions were introduced into the FACS instrument (BD FACSAria II; BD Biosciences) in order to check the intactness of the cell preparation and to set the optimum parameters for separating eYFP positive and negative cells (FSC and SSC analyses; gate settings; see Supplementary Material S2). Using “high” flow rate (about 65 μl/min; flow grade 10), about 3 % of intact single cells showed high fluorescence and about 78% no fluorescence at all in FITC (fluorescein isothyocyanate) channel. Highly fluorescent and non-fluorescent cells were collected and cultivated separately.

Neuronal differentiation of radial glia-like neural stem cells were induced as described before (Markó et al., 2011). Briefly, EGF was withdrawn from the medium of confluent cultures of RGl cells. Confluent cultures were washed with DMEM/F12 (1/1) medium (Sigma) containing 1% B27 (Gibco), and were grown further in this medium without EGF supplementation. Half of the medium was changed every second day.

Radial glia-like cells or RGl-derived neuronal cultures were fixed with 4% (w/v in PBS) paraformaldehyde (Taab) for 20 min at room temperature, then permeabilized with 0.1% Triton-X 100 (Sigma) for 5 min. For blocking non-specific antibody binding, the fixed, permeabilized cultures were incubated with 2% bovine serum albumin (Sigma) in PBS (BSA-PBS) for 1 h at room temperature. First-layer antibodies (Table 1) were diluted in BSA-PBS. Fixed cells were incubated with the primary antibodies at 4°C, overnight.

After washing with PBS, the cultures were incubated at room temperature for 1 h with fluorochrome-conjugated secondary antibodies: anti-mouse-Alexa-594 and anti-rabbit-Alexa-488 (Invitrogen) diluted in PBS to 1/1000. After washing, the preparations were mounted with Mowiol (Calbiochem) containing 10 μg/ml bisbenzimide (Hoechst 33258; Sigma). Immunocytochemical staining was analyzed with a Zeiss Axiovert 200M inverted fluorescent microscope.

Using the AxioVision 4.8 (Zeiss) program, cells were outlined and the area and fluorescence-intensity of the outlined spots were measured. Fluorescence intensity values were related to 1 μm² cell areas. Averages and standard deviations were calculated from data of 16–25 randomly selected non-induced and 5-day-induced cells (see Supplementary Material S3).

In vitro whole-cell patch clamp recordings were performed from ChR2 expressing or ChR2 non-expressing RGl and RGl-derived neuronal cultures grown on glass coverslips. Cells were visualized with the aid of DIC (differential interference contrast microscopy) optics and EYFP expressing cells were visualized with the Andor Spinning Disk Confocal system (Andor Technology Ltd) mounted on a Nikon FN1 upright microscope. During recordings coverslips were placed into a submerged type recording chamber and perfused (∼2 ml/min, 35–36°C) with a solution containing the following constituents (in mM): NaCl 145, KCl 3, CaCl₂ 2, MgCl₂ 1, D-glucose 10, HEPES 10, osmolality 300 mmol/kg, and bubbled with carbogen (95% O₂, 5% CO₂). 4–10 MΩ recording pipettes were fabricated from borosilicate glass capillaries (1.5 mm O.D.) with a P1000 puller (Sutter Instruments) and filled with intrapipette solution containing (in mM): KCl 130.0, CaCl₂ 0.5, MgCl₂ 2.0, EGTA 5.0, HEPES 10.0. (pH 7.2 with KOH). For membrane current recording, cells were voltage clamped at -65 mV membrane potential with an MultiClamp 700B Amplifier (Molecular Devices) and were stimulated with 50 ms blue light laser pulses (λ: 488 ± 20 nm), gradually increasing the laser power. Resting states and responses to the stimulation were recorded with pClamp 10.3 software (Molecular Devices) and the data were offline analyzed with Clampfit 10.3 software.

For video microscopy, non-induced and neuronally differentiated RGl cells were grown in 35 mm Petri dishes with special optical bottom (high glass bottom μ-Dish; Ibidi). Cultures were kept at permanent 37°C temperature and in 5% CO₂ and 95% air atmosphere within a mobile mini-incubator (Szabó et al., 2011) attached to the microscope stage. Time-lapse recordings were performed with Zeiss Axiovert 200M inverted fluorescence microscope. For light stimulation and fluorescence images, an epifluorescence filter set (band-pass filters for excitation λ: 470 ± 40 nm and for emission λ: 525 ± 50 nm) was used. The cells were stimulated with light for 300 ms in every 5 min with a light intensity of 0.13 mW/ mm². Images were acquired with a 10x objective at the end of each illumination period (in every 5 min) for 12 h, both in phase contrast and epifluorescence optical modes. Light-stimulation and video microscopy was started exactly 24 h and 120 h after the withdrawal of EGF (e.g., the onset of induction) in case of 1-day and 5-day induced populations, respectively.

The effects of light stimulation on cell move were investigated in cultures of non-induced RGl cells, and on 1-day and 5-day induced cells, e.g., on the zero, on the 1st and on 5th days of neuronal differentiation induced by EGF withdrawal (Table 2).

At the end of the experiments, cells were fixed with 4% (w/v in PBS) paraformaldehyde (Taab) for 20 min at room temperature and stored in PBS at 4°C for immunocytochemical analysis.

Time-lapse images were opened in WTrack program (Gönci et al., 2010) and 20 cells were selected in every starting image for movement tracking. As selection criteria, the tracked cells should not leave the microscopic field during the 12-h recording period. A not moving reference point was also selected on each microscopic field in order to correct artifacts due to potential shifts of the frame. The movements of each selected cell were tracked by clicking on the center of the cell and the coordinates were recorded from image to image by WTrack program. The cell coordinates were corrected with the changes in the coordinates of the reference point. The real displacement was determined according to the magnification and the image resolution: 1 pixel corresponded to 0.645 micrometer. Using the corrected cell coordinates, the distances made by cell centers from frame to frame were calculated according to the two dimensional Euclidean distance

where p_i and p_i+1 are the positions in the actual moment and 5 min later, respectively and 0.645 coefficient was used to calculate distances (d) in micrometer. The resulting data showed the displacement in μm/5 min. Summarizing the frame-to-frame distances

the total distance of cell displacement in 12 h was determined for each tracked cell (n = 60 for each differentiation stage, and n = 20 for ChR2-expressing but not illuminated cells).

Statistical analyses were performed with R statistic programming (R Core Team, 2016). The graphs were plotted with ggplot2 package. In each case p < 0.05 (^∗), p < 0.01 (^∗∗), and p < 0.001 (^∗∗∗) were considered statistically significant.

The μm values of total distances made by each cell during 12 h (Figure 5) were categorized as smaller than 200 μm, between 200 and 400 μm and longer than 400 μm. The data were plotted and statistical significance was determined with Pearson’s Chi-squared test in each contingency table, then paired wise comparisons were performed by post hoc Fisher test, where p-values were adjusted with Bonferroni method. The electrophysiological responses to illumination (Figure 4) and the total distance data (Figure 6) were plotted as boxplots and the significances were determined with Wilcoxon–Mann–Whitney test. The significances in the distribution of total distances made by 5-day induced ChR2+ and ChR2- cells in 12 h with or without illumination (Figure 7) were determined by Kruskal–Wallis rank-sum test and Dunn’s test.

In vitro neurogenesis by optogenetically modified radial glia-like (RGl) neural stem cells (Markó et al., 2011) (see Supplementary Material S1) provided the model for studying the motility responses of differentiating neural cells to ionic stimulation and to compare the effects of stimulation in different stages of RGl cell differentiation. For the presented studies, RGl cells were prepared from the forebrain of channelrhodopsin expressing transgenic (Madisen et al., 2012) mouse embryos on the 17.5 post-conception day. RGl cells carrying the floxed construct for channelrhodopsin-eYFP were grown in AK-c[RGDfC]-coated dishes in serum-free conditions. The cells displayed spindle-like epithelial morphology and radial glia-like immunmarkers (Markó et al., 2011) (Supplementary Material S1) similarly to wild-type RGl cells. When reaching confluency, the cells were transfected with pTurbo-Cre plasmid (Hug et al., 1996) and were let to grow for another week, when about 5% of RGl cells showed eYFP-fluorescence (Figure 1A). The cells were collected by trypsinization and eYFP-ChR2-expressing cells were separated from non-expressing cells by FACS (see Supplementary Material S2). eYFP-ChR2-positive (Figure 1B) and negative (Figure 1C) cells were propagated in separate cultures and were characterized further (Figures 1D–G).

ChR2-expressing cells were protected from light by growing in aluminum foil covered dishes, fed in a semi-dark sterile hood and examined with blue light filter in daily microscopic investigations. Under these conditions, morphological and immunohistochemical differences were not revealed between ChR2-expressing and non-expressing cells (Figure 2).

Both ChR2-expressing (Figures 2A–L) and non-expressing cells (Figures 2M–P) displayed nestin, RC2 and Pax6 immunoreactivity. To compare the neurogenic capacity of ChR2-expressing and non-expressing RGl cells, mixed (non FACS-separated) cultures containing both ChR2-expressing and non-expressing cells were induced by withdrawal of EGF from the medium. Induction of neuronal differentiation resulted in cell aggregation, and a number of cells died regardless of ChR2 expression as it was found previously in induced cultures of wild-type RGl cells (Markó et al., 2011). βIII-tubulin-positive neuronal precursors appeared among both, eYFP-positive ChR2-expressing and eYFP- negative (ChR2-non-expressing) cells by the 5th day of induction (Figure 3) indicating that under the same culturing and induction conditions, both ChR2-expressing and non-expressing cells could generate neurons. In comparison either to “wild-type” (Markó et al., 2011) or to ChR2 construct carrying but not expressing RGl cells, no changes were revealed in the morphological or immunohistochemical features of ChR2 expressing RGl cells. The ChR2 expression did not prevent the neuron production by RGl cells. The effect of illumination and all motility studies were carried out on cultures of FACS-separated cells which either expressed the ChR2-eYFP construct, or were ChR2-negative according to the absence of detectable eYFP fluorescence. The intensity of eYFP fluorescence measured in non-induced RGl stem cells and in RGl-derived neurons (Supplementary Material S3) indicated that the expression was not lost in the course of neuronal induction.

Illumination with blue (λ: 488 nm) laser light was used to open light-gated channelrhodopsin cation-channels (Zhang et al., 2010) and thus, to modify the ion distribution in non-induced RGl stem cells, and in differentiating progenies on the 1st and 5th days after EGF withdrawal.

In order to find the right parameters for light stimulation, the appropriate intensity of laser illumination was determined by patch-clamp assays (Figure 4) and by checking the survival of cells illuminated in every 5th minutes for 12 h. Repeated illumination at an intensity equal or higher than 0.25 mW/mm² caused important cell decay by the end of the 12-h exposure (Supplementary Material S4), while light stimulation below 0.2 mW/mm² did not cause cell damages and did not impair neuronal differentiation.

For 12-h-long motility assays, the cells were repeatedly illuminated with blue light with 0.13 mW/mm² light intensity. This intensity elicited measurable inward cation currents (Figure 4), and did not cause decay of either non-induced RGl cells or RGl-derived neuronal precursors. Cultured cells in optical culture dishes were placed in a mini-incubator stuck to the microscope table and were illuminated for 300 ms in every 5 min. Images in both phase contrast and epifluorescence optical modes were taken at the end of each 300 ms illumination period.

The centers of individual cells were tracked on consecutive frames of the time-lapse video images and the cell positions were determined using the WTrack program (Gönci et al., 2010) (Supplementary Material S5). The x, y coordinates of individual cell positions were related to a non-moving reference point and the move of labeled cell centers were calculated from frame to frame. The trajectories of individual cells were plotted (Supplementary Material S5) and the total distances made by cell centers were calculated by summarizing the lengths of 5-min displacements, starting from the position on the first image (t₀) and ending with the positions at the end of the 12th hour (t₁₄₃). The displacement of the cell center was accepted as migration if the tracked cell center moved twice as far as the longest dimension of the cell during the 12-h observation period. As the average length of RGl cells did not exceed 100 μm, displacement was accepted as migration if the cell center moved as far as 200 μm as a minimum in 12 h.

The move of ChR2-expressing and non-expressing (control) cells was analyzed in non-induced, 1-day induced, and 5-day induced populations (Table 2).

Displacement data obtained on illuminated ChR2-expressing and non-expressing (control) cells demonstrated that cell displacement decreased with the advancement of neuronal differentiation in both ChR2-expressing and non-expressing cells (Figures 5, 6). The number of cells migrating longer than 400 μm in 12 h (Figure 5) and the median of displacement distances (Figure 6) decreased by the 5th day of induction. In ChR2-negative populations, a mild reduction of motility was detected as soon as 24 h after the onset of induction, and the displacement activity of process-bearing cells with neuron-specific tubulin immunoreactivity was markedly reduced on the 5th day of induction. The light-sensitive cells, however, displayed different displacement-activity if exposed to repeated light stimulation. Illumination slightly (but not significantly) increased the displacement activity of non-induced ChR2-expressing RGl stem cells (Figures 5, 6). On the first day of neuronal induction, the number of ChR2-expressing cells with longer migratory routes increased further (Figure 5) and the average distance of cell displacement was significantly higher than that in ChR2- populations (Figure 6). On the 5_th day of induction, however, light stimulation markedly reduced the displacement activity: the migratory routes made by illuminated ChR2-expressing cells were significantly shorter than those of ChR2 non-expressing or ChR2-expressing but not illuminated cells (Figure 7).

The data clearly demonstrated that light-induced inward cation fluxes modified the motility of developing neural cells on a differentiation stage-dependent way.

The migration activity of 5-day induced ChR2-expressing cells decreased even without any direct exposure to blue light (Figure 7) in comparison to ChR2-non-expressing cells. In response to blue light illumination, however, the displacement activity of ChR2-expressing cells, was more significantly reduced. The data indicate that physiologically effective amount of ions could move through ChR2 channels even in the absence of targeted illumination with blue light. Stimulation with blue light, however, caused significantly enhanced inward cation currents and modified significantly the displacement activity of the cells.