The P19 cell line is an effective and inexpensive model for investigating the effect of various chemical agents on neurons’ growth and differentiation in the laboratory. It provides a method of investigating neurogenesis in vitro in the absence of any surrounding tissues or growth factors (Babuška et al. 2010). In the present study, we evaluated neurogenic differentiation and patterning in P19 cells by inducing neural differentiation in the two groups of EBs obtained by hanging drop and aggregation methods. Marshall et al., Blumberg et al., and Blumberg et al. showed that RA morphogen, in addition to neuralization, causes caudalization and dorsalization of neurons in the neural tube and embryonic stem cells (Blumberg 1997; Marshall et al. 1992). Consistent with their findings, the results of our research revealed that P19 teratocarcinoma embryonic stem cells were capable of successfully differentiating into neural cells in vitro with minimal serum. Invert microscopy confirmed that a very wide network of nerve-like cells was formed around the EBs. The resulting nerve cells had bipolar to multipolar cell bodies, extensive cytoplasmic inclusions, large and euchromatin nuclei, and granular cytoplasm (Fig. 4). These results were parallel to the morphological studies of differentiated nerve cells derived from the P19 cell line (Esmaeili et al. 2006).
The evaluation of molecular markers in our research also showed that 4 days after treatment, RA led to a decrease in the expression of the Oct4 factor, a stemness marker and the occurrence of neural phenotype with the expression of neural-specific factors Nestin, TUB-βIII, and Map2 in P19 cells (Fig. 5).
It has been confirmed that during development (in vivo), RA as a morphogen, at the same time as neural induction, plays an essential role in neural tube modeling, or rather in patterning (Novitch et al. 2003). Thus, we also evaluated and compared the identity of nerve cells derived from EBs in the two groups, (a) hanging drop and (b) aggregation methods, in the rostral-caudal axis.
Studying chick embryos, Jonas Muhr concludes that the nervous tissue initially formed in the embryo has the characteristics of cranial neurons (forebrain). Moreover, the resulting nerve cells acquire the characteristics of posterior neurons of the neural tube (midbrain, hindbrain, and spinal cord) in the sixth hour from the start of gastrulation that are influenced by factors such as RA or FGFs secreted from paraxial mesoderm, (Muhr et al. 1999). Sagha et al. showed that treating EBs with 1 µM RA induces the expression of Hoxb4, Hoxc5, and Hoxc8 genes parallel to the Hb9 marker; thereby causing the differentiation of motor neurons in the posterior part of the hindbrain and upper cervical spinal cord (Sagha et al. 2013).
Consistently, we also found that the mRNA levels of all three Hoxb4, Hoxc5, and Hoxc8 genes in the hdEBs + RA group significantly increased compared to the control (Cont a) and negative control groups. However, in the AggEBs + RA group, compared to the control (Cont b) and negative control groups, only the Hoxc5 mRNA level increase was significant. (Fig. 6).
In their research on chicken embryos, which confirms our results, Liu, J. P. et al. declare that RA interacts with FGFs in a concentration gradient-dependent manner, establishing spatial profiles of Hox genes to identify motor neuron subtypes. This means that in the presence of a high concentration gradient of RA and low amounts of FGF, the neurons of the posterior segment of the hindbrain express Hoxb4 and the motor neurons of the cervical spinal cord express Hoxc5 and Hoxc8 (Liu et al. 2001).
In our research, the Hox10 mRNA level did not increase significantly in any of the hdEBs + RA and AggEBs + RA groups compared to their respective control groups (Cont a and Cont b) (Fig. 6). According to a research study conducted by Lance-Jones et al., it can be concluded that RA does not play a role in the differentiation of lower thoracic and lumbar spinal cord neurons, the expression of which depends on the presence of FGF. Hox10 is highly expressed in lumbosacral spinal cord motor neurons (Lance-Jones et al. 2001).
An important research question that led us to carry out the present research was that we expected a wider or more neural differentiation in hanging drop EBs compared to Aggregation. Because studies have shown that compared to aggregation EBs, hanging drop EBs have more complete interactions and connections that lead to neurogenesis (Messana et al. 2008). Nevertheless, such a result was not obtained in our study.
As illustrated in Fig. 4, during the first stage of differentiation and at the end of day 8 after neural differentiation had been induced, both hanging drop EBs and aggregation EBs exhibited very similar cell morphology, processes, and neural networks. Additionally, no significant difference existed between the EBs of the hanging drop and aggregation groups in terms of neural identity markers.
Even in the control groups (Cont a and Cont b), which were the internal controls for the hanging drop and aggregation EBs, respectively, a weaker expression of Nes, Tub βIII, and MAP2 markers was observed (Fig. 5). This finding indicated that in the absence of exogenous nerve inducer (RA) alongside a minimal serum (FGF 3%), a lower percentage of P19 cells differentiates into neural progenitors or neurons. According to studies, this may occur for two reasons: (1) reducing or removing serum from the culture medium increases neural differentiation since serum inhibits neural differentiation, whereas serum augments mesodermal derivative differentiation (Okada et al. 2004); and (2) researchers such as Okada et al., Lake et al., and Rathjen et al., revealed that spontaneous EBs differentiation occurs without any inducer. A variety of cells differentiated by the origin of all three embryonic germ layers, including cardiac muscle cells, skeletal muscle cells, hematopoietic cells, and nerve cells, are observed. However, the percentage of nerve cells formed is either very low or none (Lake et al. 2000). Therefore, the neuronal gene expression profiles of the two groups of nerve cells resulting from neurogenic differentiation induced by RA appear to be similar.