NSC-34 cells are a cell line obtained by the hybridization of embryonic spinal cord cells and neuroblastoma cells from mice. They are enriched with motor neurons, so this cell line could be used as a model for a developing motor neuron system [20]. We used it in an undifferentiated state in order to understand whether our substance of analysis, CBNR, can influence the neurogenesis process. To our knowledge, this is the first time that CBNR has been administrated in an in vitro model: since NSC-34 cells are a model of developing neurons and cannabinoids proved to be useful in neurogenesis, we tried to understand whether CBNR can play a role in the generation of new cells or in differentiation and which receptor could be involved in such process.

After exposing the cells to different concentrations of CBNR, we assessed their viability and concluded that the substance is not toxic, at least until reaching the concentration of 50 µM for 24 h, as Figure 2 shows. After assuring the safety of CBNR, a transcriptomic analysis followed.

Cannabinoids proved to be able to induce neurogenesis and neuronal differentiation, so the classical markers of these processes were inspected. We clearly did not expect to see a full shift of our cells to a mature-like stage, since normally it takes days of exposure to retinoic acid to induce a complete motor neuron-like phenotype [21]. However, initial signals and activation of genes involved in neuronal differentiation might be already present. Nestin, encoded by Nes, as well as Vimentin, encoded by Vim gene, are markers of neuronal progenitor cells [22] and, for this reason, the decrease of Vim at 50, 20 and 7.5 µM pointed toward a differentiating state of the cells. A 7.5 µM concentration of CBNR is the only dose that activates two important markers of neuronal differentiation, Neurod1 [23] and Map2. It is important to mention that MAP2 is one of the markers used to identify differentiated NSC-34 [21]. On the other hand, the other dose seemed to increase the expression of Synaptophysin, encoded by the gene Syp, which is a marker widely used to identify synapses [24] and mature neurons, that decreases only at 7.5 µM. This evidence suggests that neuronal differentiation signals are active, so the possible pathways that CBNR can activate to induce this change were investigated.

CBNR is similar to CBG, thus it is reasonable to think that it may act on the same receptors. As can be seen from Table 2, every dose of CBNR upregulates Cnr1 gene, encoding for CB1 [25]. Western Blot analysis shown in Figure 3a provided us another indication about CB1: it seems that increasing the dose decreases the amount of receptor, even though it always remains above the level of the control. This is likely due to the saturation and subsequent desensitization of the receptor, which might be internalized and degraded when the dose of the substance—and consequently the stimulation of the receptor—increases. NSC-34 cell line does not express Cnr2, so no conclusion could be given about a possible affinity and interaction between CBNR and the Cannabinoid Receptor 2 (CB2), encoded by the last-mentioned gene [26].

CB1 stimulation can lead to neuronal differentiation using Akt pathway [17]. Since the activity of Akt is measured in its phosphorylation, Western Blot was performed and the peak of activity, as demonstrated in Figure 3, seems to be at 10 µM. According to previous evidence [17], the phosphorylation of ERK should decrease during differentiation, thus this part of Akt pathway was also inspected and, as shown in Figure 3, phosphorylation of ERK seems to decrease at 20 and 5 µM, while it is surprisingly high at 7.5 µM. The almost constant downregulation of the substrate of mTOR, the ribosomal protein S6, identified by the gene Rps6, indicates that the mTOR pathway also seems not to be followed. This is in accordance to previous observation regarding neuronal differentiation mediated by CB1, since its phosphorylation seems to also be constantly under the level of the control, as seen in the Western Blot reported in Figure 3 [17]. It is possible that different doses trigger neuronal differentiation with a strength proportional to the concentration. A dose of 5 µM seems too low to initiate complete neuronal differentiation signals in only 24 h and the high level of CB1 may indicate that the receptor is stimulated by the presence of CBNR, but the activity of Akt is just below the level of control. However, CB1 already bears an effect on p-ERK, since it is strongly inhibited and this might represent a very early stage of differentiation. Other evidence suggests that the cells exposed to 5 µM are at early stage of differentiation: it emerges that Cdc42, encoding for Cell Division Cycle 42, plays an important role during development, during a knockdown experiment Cdc42 demonstrated to be important during the early development of dendritic spines [27]. The fact that 5 µM of CBNR upregulates Cdc42 suggests that at this dose the cells are at an early stage of development toward the mature neuron stage. Rbfox2, essential for the proper development of motor function [28], is upregulated at 5, 7.5 and 10 µM, so it will reinforce the evidence of development ongoing. There is another signal that plays an important role in the early stage of neuronal specification: this is calcineurin (CaN), encoded from the genes Ppp3ca, Pppcb and Ppp3cc. CaN is essential for neuronal induction and thus is required during the early specification of neuroectoderm as a result of antagonizing BMPs signaling [29]. As can be seen from Table 1, Pppcb is overexpressed at 5, 7.5 and 10 µM and decreases at the higher dose, 50 µM, as though its activity is no longer necessary. Cdk5r2 is a gene that encodes p39, a protein that influences the activity of CDK5, involved in neurodevelopment along with p35, encoded by the gene Cdk5r1 [30]. However, it is important to mention that p35 can be cleaved in p25 by Calpain, encoded from the gene Capn1, that in our experimental sets is upregulated at all doses except 10 µM. p25 is more stable and induces overactivation of CDK5, which is detrimental to its deactivation. In particular, it seems able to lead to neurodegeneration [31]. What is important to notice is that upregulation of p39 was found in cortical differentiating neurons, being more resistant to calpain cleavage, while p35 expression remained steady. The role of p39 in development lies in the capacity to induce axonal extension [32] and in our analyses there was no upregulation of Cdk5r2 at 5 µM, while it started to be upregulated from 7.5 µM until 50 µM. A pathway whose blocking is correlated with neuronal differentiation is Notch1/Rho-associated protein kinase (ROCK) pathway: inhibition of Notch1 through knock-out resulted in neuronal differentiation in NSC-34 and blocking of its downstream effector ROCK1 promoted, in addition to differentiation of the progenitors [33]. In our experiment, Rock1 is downregulated at 10, 20 and 50 µM, as if the increasing dose could in some way trigger its downregulation. Again, it can be seen that treating the cells with 5 µM of CBNR does not downregulate Rock1. The second isoform of ROCK is ROCK2, encoded from the gene Rock2, which also showed a kind of dose-dependent trend to downregulation, starting from upregulation at 5 µM and decrease to downregulation at 50 µM. It is involved in different processes, such as apoptosis since it can influence the activity of caspase-3, but our interest is focused on its downregulation that correlates with positive axon outgrowth. This capacity seems mediated by Limk1, encoding for LIM Domain Kinase 1, which is downregulated at 5, 7.5 and 10 µM, not deregulated at 20 µM and upregulated at 50 µM [34]. Rock2 and Limk1 seem to follow a kind of dose-dependent trend, where increasing the dose seems to increase the expression of Limk1, allowing it to interact with Rock2. Another piece of evidence regarding the capacity of CBNR to stimulate differentiation lies in the downregulation of Camk2b at all doses. The downregulation of this gene, encoding for Calcium/Calmodulin dependent protein kinase II beta (Camk2β), was correlated with an increase in length of dendrites and an increased arborization of neurites, indicating the role of Camk2β in pruning of neurites [35]. It is important to mention that pruning is equally as important as growth. However, at such an early stage there is no need to selectively retract or remodel neuronal branches.

CB1, using Akt pathway, could also activate Cyclic AMP-Responsive Element-Binding Protein 1 (CREB) in order to promote the synthesis of brain neurotrophic factor (BDNF) [36], known to promote neuron survival [37]. Bdnf is upregulated at 50 and 20 µM. BDNF is also known to promote cannabinoid-mediated differentiation in neurons [38].

All these results suggest that CBNR has the capacity to induce neuronal differentiation of NSC-34 cells using CB1 receptor and Akt pathway. This effect appears to be dose-dependent, since the signals seem to follow a trend where higher doses can trigger more signals than lower doses. However, it has to be pointed out that, once 20 µM of CBNR is reached, we do not observe a faster or more efficient differentiation: at 50 µM there is no expression of more mature neuron markers than at 20 µM. The dose of 20 µM is where the simultaneous increase of Akt phosphorylation and decrease of ERK phosphorylation can be observed, according to the CB1-mediated differentiation [17]. Ppp3cb, Cdc42 and Rbfox2, signals that are usually expressed at the early stages of development, are present below 20 µM, as though this dose might have already pushed the cells to a more mature stage than the lower doses, as suggested by the highest expression of Syp.

It can be concluded that CBNR can induce neuronal differentiation in NSC-34 cells after only 24 h and the optimal dose appears to be 20 µM. Lower doses can still trigger neuronal differentiation, but 20 µm seems to demonstrate the most efficient balance between the different signals.