There is a wide variety of different methods and protocols for differentiation of pluripotent stem cells (PSCs) into neurons in vitro (Chambers et al., 2009; Zhang et al., 2013; Galiakberova and Dashinimaev, 2020). Globally, two main approaches to 2D generation of neurons from PSCs can be distinguished. The first one is differentiation of PSCs into neurons through the neural stem cell (NSC) stage, and the second one is direct differentiation of PSCs into neurons.

One of the most popular methods of the first approach is the DUAL SMAD inhibition method developed by Chambers and colleagues in 2009 (Chambers et al., 2009). This method is based on inhibition of two pathways of SMAD signaling activation (Lefty/Activin/TGFβ- and BMP-signaling pathways) in PSCs using small molecules (Smith et al., 2008; Yu et al., 2008; Chambers et al., 2009). This inhibition results in the differentiation of PSCs into neuroectoderm cells and their derivatives - neural stem cells. The obtained NSCs can be cultured for some number of passages or directed to terminal differentiation. This approach in a sense mimics the step-by-step processes of neurogenesis. As a result, neural cultures can be obtained, in which, however, contain not only neurons, but also various types of neural precursors as well as glial cells (Chambers et al., 2009; Sudina et al., 2024). In addition, despite the simplicity of this approach, it is quite time-consuming.

The approach with direct differentiation of PSCs into neurons involves exogenous overexpression of transcription factors that determine neuronal fate, such as Neurogenin 2 (NGN2) (Zhang et al., 2013; Lin et al., 2021). Protocols for exogenous overexpression of the NGN2 gene can use plasmids, viruses, synthetic mRNA or CRISPR genome editing technologies to obtain constitutive, inducible or temporary expression. The best known and used protocol is lentiviral transduction of PSCs with the TetON system for tetracycline-regulated transgene expression (Urlinger et al., 2000; Zhang et al., 2013). Induction of temporary overexpression of NGN2 in PSCs leads to their rapid differentiation into neurons without delaying the cells at the precursor stage. At the same time, the resulting neural cultures are mainly composed of neurons and do not contain glial cells (Zhang et al., 2013). Despite the rapidity and efficiency, this method is labor-intensive due to the need to obtain transgenic PSCs expressing NGN2 as part of the TetON system.

Although both of the aforementioned neuron generation approaches are quite actively used and studied, researchers do not always have an understanding of which approach should be used for specific tasks. In addition, each approach has many different protocols and modifications, which may affect the final outcome of these approaches. In this study, we compared the transcriptomic profiles of neural cultures derived from the same induced pluripotent stem cell (iPSC) line using two different approaches: an approach with differentiation of neurons through the stage of NSCs derived from iPSCs by DUAL SMAD inhibition method (N-DSi), and an approach with exogenous NGN2 overexpression based on lentiviral transgene delivery within a tetracycline-activated system (iN-NGN2).

We have compared the two most popular approaches to differentiate iPSCs into neurons: the approach with differentiation through the stage of NSCs derived from iPSCs using a method with DUAL SMAD inhibition, and the approach with induction of overexpression of exogenous NGN2 by lentiviral delivery of a tetracycline regulated transgene expression system.

We previously found that neural cultures spontaneously differentiated from iPSC-derived NSCs using a method with DUAL SMAD inhibition are highly heterogeneous which, however, contain cells expressing key neuronal markers. Nevertheless, the composition of heterogeneous cultures was influenced by the duration of NSCs cultivation as well as the iPSC lineage, the source of NSCs (Galiakberova et al., 2023).

Neural cultures obtained from iPSCs by overexpression of exogenous NGN2 were visually more homogeneous in cellular composition. Most of the cells of such culture had specific calcium activity and demonstrated the presence of ionotropic glutamate receptors (Galiakberova et al., 2022).

To compare the approaches, we analyzed the expression of marker genes by quantitative RT-PCR as well as the data of total transcriptome profiles of 14-day neural cultures derived from the same iPSC-KYOU line but by two different approaches. iN-NGN2, neural cultures obtained by overexpression of the NGN2 gene in iPSC-KYOU; N-DSi, neural cultures spontaneously differentiated from NSCs (derived from iPSC-KYOU using DUAL SMADi). Although it is known that the minimum maturation period for neurons produced by DUAL SMAD inhibition is 21–28 days (Telias et al., 2014), which is almost twice as long as for neurons induced by NGN2 overexpression (14 days) (Zhang et al., 2013), a single 14-day maturation period was chosen for comparative analysis. This decision was made in order to ensure comparability of conditions, as well as to emphasize the difference in the properties of neuron cultures during the same maturation period.

The results of quantitative RT-PCR demonstrated that both neural cultures differed in the relative expression levels of neuronal markers from the iPSC-KYOU from which both neural cultures were derived. However, the level of relative expression of NSC and glial cell markers was higher in N-DSi than in iN-NGN2, and the level of neuronal markers was lower. NGN2 expression was upregulated in N-DSi but not in iN-NGN2. This is consistent, since NGN2 is expressed in neuronal precursors and promotes their differentiation, but is not expressed in postmitotic neurons (Hulme et al., 2022).

An interesting result is the expression of ZIC3A and ZIC3B genes in N-DSi, comparable to that in iPSCs. ZIC3A and ZIC3B are isoforms of the transcription factor ZIC3 (Bedard et al., 2011). Expression of this transcription factor may indicate cell stemness, as its activity is required during early embryonic development (gastrulation, neurulation) (Herman and El-Hodiri, 2002), and in mouse embryonic stem cells, ZIC3 maintains their pluripotency (Lim et al., 2007; Yang et al., 2019). It is also known that ZIC3 is expressed in cells of the developing forebrain. It promotes the proliferation of neuron precursors and is also involved in their differentiation at various stages of brain development (Inoue et al., 2007; Winata et al., 2015). Thus, the expression of the isoforms of the ZIC3 gene once again emphasizes the presence of a significant proportion of proliferating stem cells in 14-day neural cultures obtained using the DUAL SMAD inhibition method.

The results of analyzing transcriptome profiles also showed significant differences between neural cultures. Metabolic and signaling pathway enrichment analysis showed that the transcriptional profile of N-DSi compared to iN-NGN2 was significantly enriched in pathways and processes related to proliferation, cell cycle maintenance, and WNT and TGFβ signaling pathways. At the same time, the transcriptional profile of iN-NGN2 was found to be enriched in various processes related to chemical synapses, neuronal plasticity, signal transduction, etc.

Compared to iN-NGN2, N-DSi cultures contained a large number of NSCs and neuronal precursors of different stages in addition to neurons. Moreover, the expression of various astroglial and oligodendrocytic genes, as well as marker genes of some other cell types, such as fibroblasts, endothelial, and vascular leptomeningeal cells, was detected. Which further confirms the heterogeneity of cultures obtained through the NSC stage.

It is generally known that NSCs derived from PSCs using a method with DUAL SMADi are capable of giving rise to both neuronal and glial branches (Chambers et al., 2009; Perriot et al., 2018; Lam et al., 2019). Based on this, various protocols have been developed to obtain astroglia from such NSCs (Perriot et al., 2018; Shimbo et al., 2020). In our experiment, we guided NSCs to spontaneous differentiation, which allowed cells to differentiate in different directions: both neuronal and glial. However, there is evidence that even when NSCs are directed to differentiate into neurons, glial cells are found in cultures (Fujimori et al., 2017; Nilsson et al., 2021). Astroglial cell markers were barely expressed in iN-NGN2, which also replicates the results of other protocols with induction of exogenous NGN2 overexpression (Chen et al., 2020; Lin et al., 2021; Schörnig et al., 2021). On the one hand, the presence of glial cells in the culture of neurons may interfere with some studies or tests, especially if a pure culture of neurons is required. On the other hand, it is known that glial cells support neurons not only in vivo, but also in vitro (Pyka et al., 2011; Clarke and Barres, 2013). Astroglia promotes better adhesion of neurons and their survival, and also secretes various neurotrophic factors (BDNF, NGF, GDNF) that contribute to synaptogenesis and neurophysiological maturation of neurons (Pfrieger, 2010; Tang et al., 2013). In this context, the method of obtaining neurons through DUAL SMAD inhibition has an undeniable advantage.

In addition to astroglia, other cell types are found in neuron cultures obtained using the DUAL SMAD inhibition approach. One of the studies aimed at obtaining dopaminergic neurons reported that the obtained neural cultures contained not only glial but also vascular leptomeningeal cells in addition to dopaminergic neurons and their immediate precursors (Nilsson et al., 2021). In iN-NGN2, however, markers of iPSCs and mesenchymal stem cells were detected. Such non-targeted cell populations were previously detected in a study where neurons were obtained by the same approach (Lin et al., 2021). And in another study, fibroblasts were detected (Schörnig et al., 2021). In general, such collateral non-neural fibroblast-like cell types can be expected when differentiating iPSCs derived from fibroblasts by reprogramming with pluripotency factors (Buckberry et al., 2023). In neurons derived by NGN2 overexpression, a neural stem cell population may be present as a bystander lineage (Vainorius et al., 2023), but in protocol we used the addition of AraC promotes the elimination of all actively proliferating cells, including NSCs.

Also, iN-NGN2 neural cultures, in contrast to N-DSi, demonstrated increased expression of marker genes of cholinergic and glutamatergic neurons, but expression of markers of GABAergic neurons was not observed. Interestingly, iN-NGN2 demonstrated not only cortical fate but also the fate of peripheral sensory neurons. All these results are consistent with the literature (Schuurmans et al., 2004; Nickolls et al., 2020) including those obtained by analyzing the transcriptomes of single neural culture cells derived by NGN2 overexpression (Lin et al., 2021; Schörnig et al., 2021). In contrast, some markers of GABAergic neurons were expressed in N-DSi, suggesting that method-based protocols with DUAL SMADi are more suitable for obtaining GABAergic neurons (Liu et al., 2013; Vigont et al., 2021).

Оur study has several limitations. First, only a single iPSC line was used. While iPS-KYOU is a common iPSC line, many others are widely used, and the reported effects could differ depending on the cell background. In the future, to confirm these results, it is necessary to compare neural cultures obtained by two approaches from several different iPSC lines. Secondly, for the generation of neuronal cultures, we utilized two methods routinely used in our laboratory. For the DUAL SMAD inhibition method, we used a popular commercial kit from Thermo (Gibco) with a composition that is not publicly disclosed. Using modified protocols of similar approaches could lead to different results. Thirdly, we analyzed neural cultures with a fairly short maturation period of 14 days. Probably, when comparing neural cultures with a longer maturation stage, we also could get different results. For example, Rosa et al. obtained completely different results in the context of neuron maturity, where they compared 3-6-month-old neural cultures obtained by the DUAL SMAD inhibition approach (the protocol differed from that in this article) with 14-19-day cultures of NGN2-induced neurons (Rosa et al., 2020). Fourth, it should be noted that the set of marker genes we chose cannot fully prove the presence of certain cell types, because often a cell type is characterized not by one or two markers, but by a whole set of genes and the degree of their expression. Moreover, some of them may mark other cell types as well.

Nevertheless, based on the data obtained, it can be stated that the obtained iN-NGN2 cultures are more homogeneous compared to N-DSi. Since this study employed bulk transcriptome analysis, which aggregates data from the entire cell population, the direct assessment of neuronal maturity under the DUAL SMADi protocol was associated with a risk of insufficient sensitivity due to the potential dilution of transcripts from mature neurons. To validate the interpretation of our data, we leveraged our previously published scRNA-seq results (Galiakberova et al., 2023), obtained from an identical culture under the same conditions. It was determined that at the corresponding stage of differentiation, the proportion of mature neurons in the population did not exceed 5% (Supplementary Figure 3). This allowed us to extrapolate the findings and interpret changes in marker expression in the bulk transcriptomic data as indicators of the neuronal maturation process itself, rather than merely a consequence of shifts in cellular composition.

Thus the results of our pilot study showed that neural cultures derived from iPSCs using the DUAL SMAD inhibition method through the NSC stage (N-DSi) were more heterogeneous with neurons comprised only a small proportion (on the 14th day of maturation) compared to neural cultures obtained from iPSCs via exogenous NGN2 overexpression (iN-NGN2). Furthermore, the expression levels of mature neuronal markers were significantly higher in iN-NGN2 compared to N-DSi. Nevertheless, residual iPSC-derived cells are likely present in iN-NGN2. Although cortical excitatory (glutamatergic) neurons predominate among the neurons in the neural cultures obtained by the two different approaches, the other minor neuronal types differ between the methods. For example, the DUAL SMAD inhibition approach produces neurons of different neurotransmitter specification: GABAergic, serotonergic, and dopaminergic. In contrast, NGN2-induced neural cultures do not express markers of these types; instead, markers indicating the presence of cholinergic neurons and, in addition, peripheral sensory neurons with unspecified neurotransmitter or functional properties are detected. Based on these preliminary results, it can be concluded that, at least for the studied iPSC line (iPS-KYOU), each of the two approaches yields neural cultures with a distinct spectrum of cell types. This variability should be considered when selecting a protocol for deriving neurons from iPSCs.