The iPSC lines used in this study were generated and provided by the iPSC Québec platform, LOEX, Québec, Canada. These iPSCs were derived from urothelial cells isolated from urine samples of three middle-aged men (Table 1). The derived iPSC clones were characterized and quantified to evaluate the iPSC quality using various techniques, such as immunoassays, RT-PCR to assess the expression of iPSC markers (Oct4, Nanog, Dnm3tb, hTERT, and Rex1), analysis of 15 short tandem repeats for genotyping, amelogenin analysis for sex determination, and G-banding for karyotyping. Subsequently, a well-validated iPSC clone from each line was transported to our laboratory for further study. Then, iPSCs were thawed, expanded, and maintained using hESC-qualified matrigel-coated plates (356277, Corning) with mTeSR™1 iPSC medium (85850, STEMCELL Technologies) as per the commercially available protocol from STEMCELL Technologies. The iPSCs were passaged when they reached approximately 70%–80% confluency using 0.5 mM EDTA (Ethylenediaminetetraacetic acid, S311-500, Fisher Scientific) prepared in PBS (10010023, Thermo Fisher Scientific).

All viral vectors were prepared and packaged for infection at the viral vector core facility of the Canadian neurophotonics platform, CERVO Brain Research Centre, Québec, Canada (RRID:SCR_016477). Viral infection experiments were conducted on iPSC-neural cultures using various serotype combinations as detailed in Supplementary Table S1. The primary objective was to optimize viral infection efficiency and assess expression levels. From systematic exploration, it emerged that the most efficient constructs for Ca²⁺ imaging experiments were AAV2/retro_hSyn_GCaMP6f, which is neuron-specific, and AAV2/retro_CAG_GCaMP6f, which exhibits global expression. Furthermore, three different concentrations of genome copies per milliliter (GC/mL) were also examined: 0.5 × 10¹⁰ GC/mL, 1 × 10¹⁰ GC/mL, and 2 × 10¹⁰ GC/mL. Insufficient fluorescence expression was observed at a concentration of 0.5 × 10¹⁰ GC/mL. However, no difference in terms of fluorescence expression or activity level was observed between the concentrations of 1 × 10¹⁰ GC/mL and 2 × 10¹⁰ GC/mL (data not shown). Consequently, a concentration of 1 × 10¹⁰ GC/mL was used for the experiments.

The GECI viral infection of neural cultures was performed 10 days before the live-cell Ca²⁺ imaging experiments. For viral infection, the complete 1 mL of neuronal medium was removed from wells, and 500 μL of fresh neuronal medium with 1 × 10¹⁰ GC/mL was added per well. Following a GECI viral infection incubation period of 4–5 h, an additional 500 μL of neuronal medium was added to maintain a total volume of 1 mL per well. After 48 h, complete medium changes were carried out, followed by 50% medium changes every 2–3 days until the day of experiments.

The neural cultures were perfused with an artificial CSF (ACSF) containing the following (mM): NaCl 135 (S6191, Sigma), KCl 5 (60128, Sigma), CaCl₂ 2 (21097, Sigma), MgCl₂^.6H₂O 2 (M33-500, Fisher Scientific), NaH₂PO₄ 2 (S5761, Sigma), HEPES 20 (BP310-100, Fisher Scientific), and Dextrose 10 (D16-1, Fisher Scientific). The pH was maintained between 7.2 and 7.4, and the osmolarity was 310–320 mOsm/L. The ACSF solution, prepared a day before each experiment and stored at 4°C, was utilized within 24 h. Moreover, L-glutamic acid (30 μM, 56860, Tocris) and TTX (1 μM, Cedarlane Labs) were prepared for additional experiments.

For the Ca²⁺ imaging experiments, two batches of neurodifferentiation were conducted for each of the three iPSC lines, resulting in a total of six batches for every time point. For each time point of the three iPSC lines, 3–4 CSs from the two batches were examined. In preparation for live Ca²⁺ imaging experiments, the neural culture medium was replaced with ACSF buffer and acclimated for 30 min in an incubator before imaging, ensuring optimal conditions for the imaging process. Then, CSs were transferred into a custom-designed perfusion chamber (Bélanger et al., 2019, 2022). During the Ca²⁺ imaging, the perfusion chamber and ACSF buffer temperature were maintained at 37°C while the buffer flow rate was 1 mL/min (Lévesque et al., 2018). The cytosolic Ca²⁺ signals were recorded using a customized multimodal epifluorescence microscope (DHM-T1003, Lyncée Tec) (Bélanger et al., 2018). The light source used for fluorescence excitation is a 120-W lamp (X-Cite 120PC Q, Excelitas) with a FITC filter cube (FITC-5050A-000, Semrock). The epifluorescence signal is detected with a CCD camera (INFINITY3S-1UR, Lumenera) with a resolution of 1,329 × 1,080 pixels. Images were acquired at 10 Hz for 5 min with a magnification of 20× and a 0.70-numerical aperture (NA). A short terminal application of L-glut pulses for 30 s was applied after 180 s of baseline recording to distinguish neurons in a neural culture (Pickering et al., 2008). Throughout the experiment, three fields of view were recorded for each CS, and approximately 30–40 cells were imaged per field of view to assess Ca²⁺ activity. The Ca²⁺ data were acquired and stored in a tiff format for offline analysis.

The cells were fixed for 15 min at room temperature using a 4% paraformaldehyde solution (J19943-K2, Thermo Fisher Scientific). Then, they were washed and permeabilized for 5 min with 0.1% Triton X-100 (AC215682500, Fisher Scientific) in PBS. The cells were then incubated with a mixture of 10% normal goat serum (P131873, Fisher Scientific) and 1% BSA (SH3057402, Fisher Scientific) in PBS to block the nonspecific bindings for 60 min. After that, the cells were incubated overnight at 4°C with the following primary antibodies: NeuN (1:50, mouse monoclonal, MAB377M, Fisher Scientific), GFAP (1:1,000, rabbit polyclonal, 60128, STEMCELL Technologies), MAP2 (1:1,000, chicken polyclonal, ab92434, Abcam), and GFP tag (1:2,000, chicken polyclonal, ab13970, Abcam), which were prepared in blocking solution. The following day, the cells were washed with PBS and incubated with secondary antibodies, Alexa Fluor 488 (1:1,000, ab150077, Abcam), or Alexa Fluor 568 (1:1,000, ab175473, Abcam), based on the primary antibody host species, for 1 h. Then, the cells were washed with PBS and incubated with DAPI solution (4′,6-diamidino-2-phenylindole, dihydrochloride, 1:10,000, D1306, Fisher Scientific) for 3 min to detect all nuclei. Subsequently, the cells were washed with PBS and then immersed in dH₂O until mounted. The CSs were mounted with Fluoromount-G (0100-01, Southern Biotech), and all glass slides were placed in a dark chamber for 24 h to facilitate drying.

The visualization of the targeted molecule was performed with a confocal microscope (LSM 700, Zeiss). From each iPSC’s lines, 1–2 CSs resulting in 3–4 fields of view were randomly captured using a Z-stack image function with an objective of 20× and a NA of 0.50. The maximum intensity projection function was used to get the final image, and all cell counting was performed in the ImageJ software (v1.5.3 t, NIH).

Statistical analysis was performed with GraphPad Prism version 9.4. A T-test (two-tailed, unpaired) was conducted to compare all iPSC-derived neural cell immunoassays and neuronal network activity data between the hSyn and CAG promoter groups.

To explore neuronal development and activity, neural cultures were effectively generated by following distinct stages, starting from iPSCs to neural cells (Figure 1A), using commercially defined protocols, culture media, and supplements (Figure 1B). Afterward, on DIV10 and DIV30, the neural cultures were virally infected with the AAV2/retro_hSyn_GCaMP6f and AAV2/retro_CAG_GCaMP6f constructs, and their fluorescence expression was regularly monitored. Figures 1C,D illustrates the results obtained from the two GECIs AAV2/retro_hSyn_GCaMP6f and AAV2/retro_CAG_GCaMP6f on weeks 3 and 6. Additionally, the CAG promoter groups showed both an earlier onset and a higher intensity of expression compared to the hSyn promoter groups, and all iPSC lines exhibited a similar pattern of fluorescence expression (Supplementary Figure S1). The targeted genes reached an optimum expression level around day 10 after the viral infection.

As presented in the “Materials and methods” section, neural cultures were fixed on week 3 and week 6, on the same day of live neural cell Ca²⁺ imaging, to assess the neuron and glial cell ratio by immunostaining assay. The neurons were identified using the MAP2 protein marker, while the glial cells used the GFAP indicator. In addition, all nuclei were counterstained with DAPI (Figure 2A). The results show that the average percentage of the neuronal population was 52.5% on week 3, decreasing to 47.3% on week 6 [p = 0.24; difference between means (week 6 – week 3) ± SEM = −5.2 ± 3.7]. In contrast, the expression of GFAP was 24.9% on week 3, increasing to 34.7% on week 6 [p = 0.19; difference between means (week 6 – week 3) ± SEM = 9.8 ± 5.7]. Furthermore, the amount of only DAPI-positive cells was 22.6% on week 3 and 21.9% on week 6 [p = 0.9; difference between means (week 6 – week 3) ± SEM = −0.7 ± 7.2; n = 6 CSs/3 iPSC lines] (Figure 2C). The immunostaining with a GFP tag and GFAP marker has been performed on AAV2/retro_hSyn_GCaMP6f and AAV2/retro_CAG_GCaMP6f neural cultures to evaluate the neuronal cells expression efficiency and specificity of the hSyn promoter as well as the CAG promoter on week 3 (Figure 2B). The GFP- and GFAP-positive cells were assessed relative to DAPI staining (Figure 2D). In the AAV2/retro_hSyn_GCaMP6f group, the percentage of total positive cells was as follows: GFP 35.8% (SEM ± 2.6), GFAP 24.8% (SEM ± 2.2), and only DAPI 39.4% (SEM ± 3.0). In the hSyn promoter group, neural cultures did not show GFP and GFAP colocalization, stressing neuronal specificity, whereas the AAV2/retro_CAG_GCaMP6f group showed the following percentage of total positive cells: GFP 50.8% (SEM ± 3.1), GFAP 26.0% (SEM ± 0.9), and only DAPI 23.2% (SEM ± 2.3). The CAG promoter group demonstrated 51.7% (SEM ± 4.9) colocalization of GFP- and GFAP-positive neural cells (Figure 2D). The analysis indicated that the glial cell populations in both groups were quite similar, suggesting that the promoters had no significant effect on neural cell composition. Moreover, the CAG promoter was expressed in both neurons and glial cells, while the hSyn promoter showed high specificity for neurons.

To assess the efficiency of GECI expression induced by hSyn and CAG promoters in neurons, an immunostaining assay was conducted using a GFP tag to identify cells expressing the hSyn or CAG promoter, a NeuN marker to identify neurons, and counterstaining with DAPI to stain all cell nuclei. Under similar conditions, immunostained neural cultures were imaged and analyzed, and representative images of the hSyn promoter (Figure 3A) and CAG promoter (Figure 3B) groups’ neural cultures expressing GFP- and NeuN-positive cells were obtained. The statistical analysis within and between groups showed no significant difference in the expression level of the hSyn promoter group on week 3 (38.1%) and week 6 (34.8%) [p = 0.68; difference between means (week 6 - week 3) ± SEM = −3.3 ± 7.4]. The CAG promoter group also did not show a difference between week 3 (50.8%) and week 6 (55.7%) [p = 0.54; difference between means (week 6 – week 3) ± SEM = 4.9 ± 7.5]. Furthermore, the percentage of NeuN-positive cells on week 3 (50.5%) did not show a difference from week 6 (52.5%) [p = 0.84; difference between means (week 6 - week 3) ± SEM = 1.9 ± 9.2] (Figure 3C). In the hSyn and CAG promoter groups, GFP-positive cells were compared with NeuN-positive cells. The hSyn promoter group showed 75.6% of neurons were positive with GFP on week 3, whereas 66.2% of neurons were positive on week 6 [p = 0.35; difference between means (week 6 – week 3) ± SEM = −9.4 ± 9.0] (Figure 3D). Furthermore, the CAG promoter group analysis showed 65.1% of neurons were positive on week 3 and 69.6% of neurons were positive with GFP on week 6 [p = 0.54; difference between means (week 6 – week 3) ± SEM = 4.5 ± 6.9] (Figure 3D). Notably, there were no significant differences between hSyn and CAG promoter GECI expression on neurons at both time points of analysis (Figure 3D). Therefore, this result suggests that hSyn- and CAG-promoters of GECI express similarly in iPSC-derived neurons.

The neuronal Ca²⁺ activity assessment was divided into five main steps: acquisition and saving (Figures 4A,B), selection of the regions of interest (ROIs), data processing, Ca²⁺ spontaneous event peak identification, and estimation of peak parameters. Specifically, a custom-made algorithm written in MATLAB (version 9.1, MathWorks) was designed to detect the cell soma based on maximum intensity fluorescence, threshold the image, and generate an image delineating the ROIs that can be loaded in ImageJ using the ROI manager. Further automated ROI selections were manually verified and validated to prevent incorrect selections or to add any missing cells if needed (Figure 4C). To adjust the baseline signal of cellular data, multiple background ROIs (areas without cells) were also manually defined within the field of view. The raw ROI intensity traces were then exported back into MATLAB. The intensity time traces for each of the ROIs selected were then stored in comma-separated values (CVS) files. The time traces were analyzed using another MATLAB script adapted from Deneux et al. (2016) that applied a background correction, normalized and smoothed the data, and detected Ca²⁺ events. Concretely, this data processing was used to subtract the background, correct the signal for potential drift or photobleaching, and detect spontaneous Ca²⁺ events.

First, the background correction was performed, and each trace baseline fluorescence level was calculated as the average basal fluorescence intensity. The Ca²⁺ signal intensity of each raw trace was then normalized by calculating ΔF/F as the ratio of the increase in fluorescence (ΔF) to the baseline (F). Next, the signal drift, background, and baseline corrected data were saved for further analysis (Figures 4E,F). In the final step, an automatic fit event correction was performed to identify the Ca²⁺ spontaneous events. Additionally, all the automatic Ca²⁺ signal fit events were manually reviewed and corrected if necessary. The neuronal Ca²⁺ spontaneous activity was considered a signal if it was ≥2× the baseline mean standard deviation values (Figures 4G–I). Each data set was evaluated under identical experimental conditions as described earlier during the analysis process. The last step consists of estimating single-event parameters. To achieve this, each detected Ca²⁺ event was fitted using the following 4-parameter equation:

where Amp , τ decay , τ rise , and t 0 represent, respectively, the amplitude, the decay time, the rise time, and the time of each event. The parameters were exported in xlsx files for comparison.

Practically, 180-s Ca²⁺ trace data were considered to evaluate the neuronal network activity. After 180 s of live Ca²⁺ imaging, L-glut was applied to the neural culture to identify the neurons (Figure 4D). Only cells that had a clear response to L-glut were kept in the analysis (Figure 4D) and were assumed to be neurons (Supplementary Figure S2). Indeed, L-glut has been frequently used by researchers in live Ca²⁺ imaging experiments to recognize the neuronal cells in mixed cortical cultures in vitro. Specifically, L-glut application induces substantial Ca²⁺ influx in the neurons due to the vast number of L-glut-responsive receptors, mainly NMDA receptors, in comparison to non-neuronal cells, including glial cells such as astrocytes which may show only a transient Ca²⁺ influx (Pickering et al., 2008).

The spontaneous Ca²⁺ events in neurons were assessed and compared between the hSyn and CAG promoter groups. The representative Ca²⁺ traces from all three iPSC lines (Ctl9062, Ctl9001, and Ctl5439) show neuronal spontaneous activity in 3-week-old neurons (Figures 5A,B). The neuronal Ca²⁺ intensity signals, represented as ΔF/F values, for the hSyn promoter group (Figure 5A) and the CAG promoter group (Figure 5B), illustrate the neuronal Ca²⁺ events of individual neurons within a single field of view. Notably, our findings depict that neuronal Ca²⁺ spontaneous events significantly decline over time. In the hSyn promoter group, neurons showed higher spontaneous activity at 3.61 events per minute on week 3, which declined to 1.89 events per minute on week 6 [p = 0.002; difference between means (week 6 – week 3) ± SEM = −1.72 ± 0.45] (Figure 5C). The CAG promoter group also showed a similar trend. On week 3, spontaneous neuronal Ca²⁺ activity was 3.81 events per minute, whereas on week 6, it was 1.33 events per minute [p = 0.0001; difference between means (week 6 – week 3) ± SEM = −2.48 ± 0.35] (Figure 5D). However, no significant differences were found within the three iPSC lines at both time points in both virally infected groups (data not shown). Conversely, when hSyn and CAG promoter groups were compared on week 3 [p = 0.69; difference between means (hSyn week 3 – CAG week 3) ± SEM = 0.20 ± 0.49] and on week 6 [p = 0.07; difference between means (hSyn week 6 – CAG week 6) ± SEM = −0.55 ± 0.29], no significant differences can be observed (Figures 5E,F). Hence, our data confirmed that spontaneous neuronal network activity is higher in the early phase (week 3) of neuronal development compared to the late phase (week 6). Furthermore, both hSyn and CAG promoter-based GCaMP6f enabled accurate assessment of the spontaneous frequency of Ca²⁺ in neuronal cells. We also tested TTX, a blocker of Na⁺ channels, in some experiments to validate that spontaneous Ca²⁺ events are mainly associated with Na⁺ channels (Supplementary Figure S2). We observed that in the presence of TTX, spontaneous Ca²⁺ events were significantly reduced or abolished, and they reappeared after washing it out (Supplementary Figure S2), which suggests that Ca²⁺ events presumably result from voltage changes through the depolarization of neurons (Chong and Ruben, 2008).

The neuronal spontaneous Ca²⁺ events were further analyzed and compared in terms of amplitude, rise time, and decay time of signals (Figure 6A) obtained from the hSyn and CAG promoter groups. The amplitudes of spontaneous Ca²⁺ events were compared within and between both groups on week 3 and week 6. Within the hSyn promoter group, the average amplitude of spontaneous Ca²⁺ events was 0.31 a.u. on week 3 and 0.28 a.u. on week 6 [p = 0.439; difference between means (week 6 – week 3) ± SEM = −0.4 ± 0.5] (Figure 6B), while in the CAG promoter group, the average amplitude was 0.57 a.u. at week 3 and 0.55 a.u. at week 6 [p = 0.864; difference between means (week 6 – week 3) ± SEM = −0.02 ± 0.11] (Figure 6C). Although there were no significant differences in Ca²⁺ event amplitude within each group between week 3 and week 6, the hSyn promoter group displayed a significantly lower average amplitude at both time points. On week 3, the hSyn promoter group amplitude was 0.31 a.u., whereas the CAG promoter group amplitude was 0.57 a.u. [p = 0.002; difference between means (hSyn week 3 – CAG week 3) ± SEM = 0.26 ± 0.07] (Figure 6D). At week 6, the hSyn promoter group amplitude was 0.28 a.u. and the CAG promoter group amplitude was 0.55 a.u. [p = 0.02; difference between means (hSyn week 6 – CAG week 6) ± SEM = 0.28 ± 0.11] (Figure 6E). In contrast, the rise time of spontaneous Ca²⁺ events was similar in both groups and showed no significant change between the two time points. The hSyn promoter group rise time was 0.23 s on week 3 and 0.24 s on week 6 [p = 0.82; difference between means (week 6 – week 3) ± SEM = 0.01 ± 0.03] (Figure 6F). Similarly, the CAG promoter group rise time was 0.27 s on week 3 and 0.24 s on week 6 [p = 0.29; difference between means (week 6 – week 3) ± SEM = −0.03 ± 0.03] (Figure 6G). The spontaneous Ca²⁺ event rise time between hSyn and CAG promoter groups did not differ from each other on week 3 [p = 0.19; difference between means (hSyn week 3 – CAG week 3) ± SEM = 0.04 ± 0.03] (Figure 6H) and on week 6 [p = 0.82; difference between means (hSyn week 6 – CAG week 6) ± SEM = 0.01 ± 0.03] (Figure 6I). The decay time of spontaneous Ca²⁺ events within groups did not show differences between the two time points in either group. In the hSyn promoter group, the decay time was 2.37 s on week 3 and 2.29 s on week 6 [p = 0.69; difference between means (week 6 – week 3) ± SEM = −0.09 ± 0.21] (Figure 6J). Meanwhile, in the CAG promoter group, the decay time was 2.71 s on week 3 and 2.81 s on week 6 [p = 0.75; difference between means (week 6 – week 3) ± SEM = 0.09 ± 0.29] (Figure 6K). Moreover, there were no differences in the decay time between the hSyn and CAG promoter groups on week 3 [p = 0.25; difference between means (hSyn week 3 – CAG week 3) ± SEM = 0.34 ± 0.28] (Figure 6L). However, on week 6, the hSyn and CAG promoter groups showed a significant difference in decay time [p = 0.03; difference between means (hSyn week 6 – CAG week 6) ± SEM = 0.52 ± 0.22] (Figure 6M). The neuronal spontaneous Ca²⁺ event kinetics data suggest that the properties of the Ca²⁺ events remained relatively stable within the hSyn and CAG promoter groups throughout the studied neurodevelopmental time period, despite the overall changes in Ca²⁺ event frequency. Furthermore, these findings indicate that the CAG promoter may lead to more robust Ca²⁺ signaling compared to the hSyn promoter in the studied neuronal population.