Screening Biophysical Sensors and Neurite Outgrowth Actuators in Human Induced-Pluripotent-Stem-Cell-Derived Neurons

Under these conditions we observed a large and significant increase in fluorescence ratio from pH 6–8 (Figure 3E). Thus, SNARF-5F-AM successfully detects changes in intracellular pH in live neurons and can be used to monitor the intracellular pH of neurons while testing various conditions of stimuli, injury, regeneration, and other environmental changes. Changing extracellular pH can be used as a modality for changing the intracellular pH of neurons albeit to a very small extent.

Neurons are metabolically active cells and reactive oxygen species (ROS) are largely byproducts of cellular metabolism [109,110,111]. Baseline levels of ROS are important in neurons for regulating the signaling involved in differentiation, polarization, and synaptic plasticity, while excessive ROS accumulation results in oxidative damage to neurons [109,110,111]. Hence, monitoring intracellular ROS in neurons is important to determine neuronal health. To monitor intracellular reactive oxygen species (ROS) in hiNSC-derived neurons, we used a Peroxy Orange 1 (PO1) sensor (tested at 1–5 µM and used at 5 µM for all experiments) [112]. Once inside the cell, PO1 responds to ROS and emits a fluorescence with an intensity proportional to intracellular ROS. To test the sensor, we used media with 500 µM hydrogen peroxide (H2O2). After incubating the neurons in the media with H2O2 for 10 min, we observed a significant increase in PO1 signal (Figure 3F). Thus, PO1 successfully detects changes in intracellular ROS in live neurons and can be used to monitor the metabolic health of neurons under various conditions of stimuli, injury, regeneration and other environmental changes.

Overall, the pH and ROS sensors can be used to monitor the cell state in live neurons. These sensors can be used in any combination with the morphology dyes and the bioelectric sensors to monitor neuronal dynamics based on specific experimental variables. Altogether, the panel of morphology, bioelectric, and cell state sensors characterized above serve as powerful tools for monitoring intracellular biophysical dynamics in live neurons.

Wounding involves nerve injury in addition to many other aspects, and neuronal regeneration and neurite outgrowth is an important determinant of proper wound healing [45,50,51,52,53,54,113]. Scratch injury assays are routinely performed to study injury and subsequent recovery [56]; thus, we sought to establish methods to characterize bioelectric states during this process. Confluent hiNSC differentiated into neurons over ten days were subjected to mechanical scratching (Figure 4A). The scratch location was monitored and imaged till day 11 post-scratch (Figure 4A–F). On day 1 post-scratch, sparse small projections were given out into the scratch by neurons lining the scratch (Figure 4B). By day 4 post-scratch, we observed that the small projections had grown into small, localized network of neurite outgrowths in the scratch connecting neighboring neurons (Figure 4C). By day 8 post-scratch, they had grown into large neurite networks within the scratch with thick neurite fibers traversing across the scratch injury (Figure 4D). By day 11 post-scratch, it had grown into an extensive dense network of neurite outgrowth and thick neurite fibers covering the entire scratch injury (Figure 4E). The live morphology sensor Calcein Green AM showed that these neurons, neurite networks, and fibers were alive and healthy (Figure 4E). Given the technical problems with multiday repeated use of Calcein Green AM as discussed above, we used the well-tolerated Calcein Red-Orange AM over ten days post-scratch to track and quantify neurite outgrowth (Figure 4G–I). We observed a marked increase in neurite density within the scratch injury between day 2 and day 10 post-scratch. This scratch injury assay and neurite density quantification was used to test the effect of various actuators/stimuli on neuronal regeneration and recovery postinjury.

Neurotransmitter signaling is both a canonical transduction step downstream, and a trigger upstream, of voltage dynamics, not only in neurons but also in non-neural cells [114,115,116,117]. Much work has gone into serotonergic endpoints in bioelectric signaling [118,119,120], but the effects of neurotransmitters on the repair by hiNSC derivatives are largely unknown. Thus, we used our system to examine functional impacts of a number of neurotransmitters, starting with acetylcholine, which is known to affect the central and peripheral nervous system as well as non-neural tissues through its ionotropic (direct influence on membrane voltage) and metabotropic (complex long-term effects) receptors [121,122].

To test the effect of acetylcholine on neurite outgrowths after injury, we first checked if hiNSC-derived neurons had cholinergic neurons. Immunostaining for choline acetyl transferase showed cholinergic neurons distributed evenly throughout the cultures (Figure 5A). Neurons were then mechanically injured as mentioned above and left untreated (controls) or treated with three different concentrations of acetylcholine (250 µM, 500 µM, 1 mM) for 2 days. On day 2 post-scratch, the 250 µM and 500 µM acetylcholine-treated scratch injuries were indistinguishable from the controls (Figure 5B). However, 1 mM acetylcholine significantly reduced the number of neurite outgrowth and neurite density (Figure 5B–D). Contrarily, day 10 post-scratch showed a significant increase (in comparison to the controls) in overall neurite density in the acetylcholine treatment group, which peaked at 500 µM and then fell back a bit at 1 mM (Figure 5E–G). Although the increase in neurite density was striking (Figure 5F,G), we observed this increase mainly in the number of neurite fibers running parallel to the scratch edge. The neurite outgrowths away from the scratch surface (perpendicular to the scratch) were the same if not reduced in comparison to the controls (Figure 5F,G). Thus, acetylcholine seems to show a biphasic effect on neurite outgrowth with a suppression of neurite outgrowth on day 2 post-scratch but an enhanced neurite density by day 10 post-scratch with neurite fibers mainly arranged parallel to the scratch, a sign of impaired nerve repair and regeneration.

To test the effect of serotonin on neurite outgrowths after injury, we first checked if hiNSC-derived neurons had serotonergic neurons. Immunostaining for serotonin reuptake transporter showed bundles of serotonergic neurons within the cultures (Figure 6A). Neurons were then mechanically injured as mentioned above and left untreated (controls) or treated with three different concentrations of serotonin (250 µM, 500 µM, 1 mM) for 2 days. On day 2 post-scratch, the 250 µM serotonin-treated scratch injuries were indistinguishable from controls (Figure 6B). However, 500 µM and 1 mM serotonin significantly increased the number of neurite outgrowth and neurite density (Figure 6B–D). Analogously, day 10 post-scratch showed a significant dose-dependent increase (in comparison to the controls) in overall neurite density in the serotonin treatment group (Figure 6E,G). The increase in neurite outgrowth was very striking (Figure 6F,G), with much longer neurite outgrowths (perpendicular to the scratch surface) and very few fibers parallel to the scratch surface. Such extensive neurite outgrowth pattern is known to be conducive to nerve regeneration and wound healing [45,50,51,52,53,54,113,123]. Thus, serotonin exposure significantly enhances neurite outgrowths and serotonin treatment may aid in nerve repair and regeneration postinjury.

To test the effect of GABA on neurite outgrowths after injury, we first checked if hiNSC-derived neurons had GABAergic neurons. Immunostaining for glutamate decarboxylase showed GABAergic neurons within the cultures (Figure 7A). Neurons were then mechanically injured as mentioned above and left untreated (controls) or treated with three different concentrations of GABA (250 µM, 500 µM, 1 mM) for 2 days. Both on day 2 post-scratch and day 10 post-scratch, there was no discernable effect of GABA at any of the concentrations tested (Figure 7B–G). The neurite outgrowth density and patterns were indistinguishable from the controls. Thus, although GABAergic neurons are present in culture, targeting them using GABA does not affect the neurite outgrowths and hence nerve repair and regeneration. However, the loss of function (blockade) of native GABA signaling could be a target in future studies.

Wound environments in general have elevated pH which is detrimental to wound healing, and acidic pH has been shown to promote wound healing by various means such as altering protease activity, enhancing epithelialization, and angiogenesis [105,106,107,108]. To test the effect of pH on nerve repair and regeneration, hiNSC-derived neurons were mechanically injured as mentioned above and left untreated (controls) or incubated in regular culture media with pH adjusted to 6, 7, and 8 for 2 days. On day 2 post-scratch, neurons in pH 7 and pH 8 media had scratch injuries indistinguishable from those of the controls (pH 7.4) (Figure 8A). However, neurons in acidic pH 6 media showed a significantly higher neurite density (Figure 8A–C). However, the increase in neurite outgrowth pattern was mainly in the form of increased fibers parallel to the scratch surface, which is characteristic of impaired nerve repair and regeneration (Figure 8C). In contrast, on day 10 post-scratch, neurons exposed to pH 6 media were not different from the controls (pH 7.4); however, neurons exposed to both pH 7 and pH 8 showed a significant decrease in neurite density and outgrowth (Figure 8D–F). Thus, overall, it seems that acidic pH is better for neurite outgrowth and an elevated pH impairs nerve repair and regeneration. This is in line with studies [105,106,107,108] showing that an acidic wound environment promotes better wound healing. Interestingly, the controls at a physiological pH of 7.4 seemed to violate this trend at day 10, which may be due to an absence of pH shock on day 2 (going from pH 7.4 to pH 7.4 as opposed to going from pH 6 to pH 7.4) or a homeostatic sweet spot at a physiological pH of 7.4. Nonetheless, the modulation of pH could be used as a strategy to boost nerve repair and regeneration post-injury.

Here, we characterized a number of reagents that allow the dynamic monitoring of cell morphology, membrane voltage, ion levels, and cell activity such as pH and reactive oxygen species, tracked responses to injury and reinnervation, and showed in a quantitative manner the ability of bioelectrical actuators (neurotransmitters and pH) to manipulate nerve repair and regeneration. Although our neurite scratch density measurements were very successful in detecting and quantifying changes in neurite outgrowths, they did not seem to capture the differences in patterns of neurite outgrowth. Hence, additional quantitative measures such as computing the neurite length and direction might have to be added for capturing the changes in the patterns of neurite outgrowths [123,124]. Ultimately, these hiNSC-derived neurons can be cocultured with glia, astrocytes, oligodendrocytes or with muscles, skin cells, etc., [65,66] to further study the bioelectric and neurotransmitter signaling during the complex interactions of these cells under various stimuli and experimental conditions. Future work on the biophysics of cultured stem cells will continue to enrich efforts to decipher the workings of cellular collectives during wound healing, regeneration, and other environmental challenges, with many biomedical applications.

We thank Rakela Colon for general lab assistance, David Kaplan for the generous gift of the hiNSC, and Julia Poirier for assistance with the manuscript.

The following are available online at https://www.mdpi.com/article/10.3390/cells11162470/s1, Figure S1: Calcium spiking behavior in hiNSC-derived neurons, Figure S2: NeuO stains non-neural M0 macrophage cells, Figure S3: MQAE live stain fails to label hiNSC derived neurons, Figure S4: Quantitative determination of neurite outgrowth from scratch-injured neurons, Movie S1: hiNSC-derived day 10 neurons stained with Fluo4 AM showing baseline Ca2+ dynamics in neurons, Table S1: Ionic composition (in mM) of extracellular solutions with changing Na+ ion concentration and physiological levels in green, Table S2: Ionic composition (in mM) of extracellular solutions with changing K+ ion concentration and physiological levels in green, Table S3: Ionic composition (in mM) of extracellular solutions within changing Na+ and K+ ion concentration for generating various resting membrane potentials with physiological levels in green.

V.P.P. and B.G.C. performed the experiments; V.P.P., B.G.C. and M.L. designed the experiments and interpreted data; V.P.P. and M.L. wrote the manuscript together. All authors have read and agreed to the published version of the manuscript.

All data generated or analyzed during this study are included in this article and its Supplementary Materials and are available from the corresponding author upon request.

We gratefully acknowledge the support of the Allen Discovery Center program through The Paul G. Allen Frontiers Group (12171), and the Defense Advanced Research Projects Agency (DARPA), Army Research Office, under Cooperative Agreement no. W911NF-18-2-0104, and the Department of Interior, Award no. D20AC00003.

All data generated or analyzed during this study are included in this article and its Supplementary Materials and are available from the corresponding author upon request.