Transcriptomic Analysis of Stem Cells Treated with Moringin or Cannabidiol: Analogies and Differences in Inflammation Pathways

Inflammation is a common feature of many neurodegenerative diseases. The treatment of stem cells as a therapeutic approach to repair damage in the central nervous system represents a valid alternative. In this study, using Next-Generation Sequencing (NGS) technology, we analyzed the transcriptomic profile of human Gingival Mesenchymal Stem Cells (hGMSCs) treated with Moringin [4-(α-l-ramanosyloxy)-benzyl isothiocyanate] (hGMSCs-MOR) or with Cannabidiol (hGMSCs-CBD) at dose of 0.5 or 5 µM, respectively. Moreover, we compared their transcriptomic profiles in order to evaluate analogies and differences in pro- and anti-inflammatory pathways. The hGMSCs-MOR selectively downregulate TNF-α signaling from the beginning, reducing the expression of TNF-α receptor while hGMSCs-CBD limit its activity after the process started. The treatment with CBD downregulates the pro-inflammatory pathway mediated by the IL-1 family, including its receptor while MOR is less efficient. Furthermore, both the treatments are efficient in the IL-6 signaling. In particular, CBD reduces the effect of the pro-inflammatory JAK/STAT pathway while MOR enhances the pro-survival PI3K/AKT/mTOR. In addition, both hGMSCs-MOR and hGMSCs-CBD improve the anti-inflammatory activity enhancing the TGF-β pathway.


Introduction
Many neurodegenerative diseases such as Alzheimer's Disease (AD), Parkinson's disease (PD), Multiple Sclerosis (MS) and Amyotrophic Lateral Sclerosis (ALS) are characterized by neuro-inflammation that leads to cell death [1]. The regenerative potential of stem cells can be used as a therapeutic strategy to repair damaged nervous tissue [2,3]. The human Gingival Mesenchymal Stem Cells (hGMSCs) show immunomodulatory effect and differentiation capacity [4]. Particularly, hGMSCs originate from the neural crest; thus, they can differentiate into neuronal cells [5]. Moreover, hGMSCs are easy to harvest and exhibit anti-inflammatory and antiapoptotic effects [6,7]. Specifically, inflammatory activities are mediated by the release of extracellular vesicles containing anti-inflammatory cytokines such as Transforming Growth Factor beta (TGF-β) and Interleukin 10 (IL-10).
while cells differentiated into adipogenic commitment showed red lipid droplets localized at the cytoplasmic level ( Figure 1D). To confirm the differentiation process, reverse transcription polymerase chain reaction (RT-PCR) were performed. The genes that encode for RUNX-2, ALP, FABP4 and PPARγ were upregulated in differentiated cells when compared to the undifferentiated ( Figure 1E,F). adherent capacity ( Figure 1B). To evaluate the ability to differentiate into osteogenic and adipogenic lineages, cells were stained with Alizarin Red S solution and with Oil Red O solution at the end of the induction period. Osteogenic differentiated hGMSCs showed the red positive calcium deposits ( Figure 1C), while cells differentiated into adipogenic commitment showed red lipid droplets localized at the cytoplasmic level ( Figure 1D). To confirm the differentiation process, reverse transcription polymerase chain reaction (RT-PCR) were performed. The genes that encode for RUNX-2, ALP, FABP4 and PPARγ were upregulated in differentiated cells when compared to the undifferentiated ( Figure 1E,F). (F) RT-PCR of adipogenic markers, as FABP4 and PPARγ. Scale bar: 10 µm. ** p <0.01.

Effects on Morphology and Viability of MOR and CBD Treatments
The hGMSCs were treated with MOR at 0.5 µM and with CBD at 5 µM for 48 h. The in vitro biological features were evaluated using Confocal Laser Scanning Microscopy (CLSM). Untreated and treated hGMSCs were cultured in tissue culture dishes and they exhibited a similar fibroblast-like morphology and plastic-adherence to the substrate. MOR and CBD treatment did not lead to a cell morphological modification under CLSM observation ( Figure 2B,D) when compared to the untreated cells (hGMSCs-CTRL, Figure 2A

Effects on Morphology and Viability of MOR and CBD Treatments
The hGMSCs were treated with MOR at 0.5 µM and with CBD at 5 µM for 48 h. The in vitro biological features were evaluated using Confocal Laser Scanning Microscopy (CLSM). Untreated and treated hGMSCs were cultured in tissue culture dishes and they exhibited a similar fibroblast-like morphology and plastic-adherence to the substrate. MOR and CBD treatment did not lead to a cell morphological modification under CLSM observation ( Figure 2B,D) when compared to the untreated cells (hGMSCs-CTRL, Figure 2A,C). The hGMSCs treated with MOR and CBD showed a similar proliferation rate of the untreated cells as demonstrated by 3-(4,5-dimethylthiazolyl-2)-2,5-diphenyltetrazoliumbromide (MTT) assays at 24, 48 and 72 h of culture ( Figure 2E).

Transcriptomic Analysis
The differently expressed genes in hGMSCs-MOR or hGMSCs-CBD against the hGMSCs-CTRL were analyzed by means of Reactome database. In particular, we focused on the pathways triggered by the pro-inflammatory TNF-α, IL-1, IL-6 ( Table 1) and the anti-inflammatory TGF-β (Table 2). In addition, Figure 3 represents a heatmap for the aforementioned pathways in which all the genes represented in Table 1 and in Table 2 are included. The heatmap shows how TNF-α signaling is downregulated asynchronously in the treatments. Moreover, hGMSCs-CBD downregulate most of the genes involved in the IL-1 pathway while hGMSCs-MOR is not able to downregulate its receptor. Furthermore, the IL-6 signaling is totally downregulated by CBD in the pro-inflammatory JAK/STAT pathway while MOR enhances the pro-survival effects of PI3K/AKT/mTOR. Finally, most of the genes in the TGF-β pathway are upregulated in both hGMSCs-MOR and hGMSCs-CBD.
In the analysis, six genes that belong to IL-1 pathway are expressed. MYD88, MAP3K7 and SQSTM1 are downregulated while MAP3K3 is upregulated in both the treatments. IL1R1, CHUK are upregulated in hGMSCs-MOR and downregulated in hGMSCs-CBD.
Moreover, the table counts 8 genes of the IL-6 signaling among which IL6ST, STAT3 and PIK3CD are downregulated while AKT1 is upregulated in both the treatments. PIK3CA, PIK3CB are downregulated in hGMSCs-MOR and upregulated in hGMSCs-CBD while TYK2 is oppositely deregulated. MTOR is upregulated in hGMSCs-MOR but there is no statistical relevant deregulation in hGMSCs-CBD.
In addition, Figure 3 represents a heatmap for the aforementioned pathways in which all the genes represented in Table 1 and in Table 2 are included. The heatmap shows how TNF-α signaling is downregulated asynchronously in the treatments. Moreover, hGMSCs-CBD downregulate most of the genes involved in the IL-1 pathway while hGMSCs-MOR is not able to downregulate its receptor. Furthermore, the IL-6 signaling is totally downregulated by CBD in the pro-inflammatory JAK/STAT pathway while MOR enhances the pro-survival effects of PI3K/AKT/mTOR. Finally, most of the genes in the TGF-β pathway are upregulated in both hGMSCs-MOR and hGMSCs-CBD.

UBA52
Ubiquitin A-52 residue ribosomal protein fusion product 1 0.14 −0.92 The genes represented in the column Gene are associated to the name retrieved in HUGO Gene Nomenclature Committee website (column Name). Fold Change hGMSCs-MOR shows the fold change obtained by Log 2 (hGMSCs-MOR/hGMSCs-CTRL) while Fold Change hGMSCs-CBD highlights the fold change obtained by Log 2 (hGMSCs-CBD/hGMSCs-CTRL). The Q-value hGMSCs-MOR and Q-value hGMSCs-CBD were used to choose the level of significance (<0.05). All the fold change values are rounded to the second decimal digit while the Q-values are approximated with scientific notation. The genes represented in the column Gene are associated to the name retrieved in HUGO Gene Nomenclature Committee website (column Name). Fold Change hGMSCs-MOR shows the fold change obtained by Log 2 (hGMSCs-MOR/hGMSCs-CTRL) while Fold Change hGMSCs-CBD highlights the fold change obtained by Log 2 (hGMSCs-CBD/hGMSCs-CTRL). The Q-value hGMSCs-MOR and Q-value hGMSCs-CBD were used to choose the level of significance (<0.05). All the fold change values are rounded to the second decimal digit while the Q-values are approximated with scientific notation.

Discussion
Inflammatory responses play a central role in the ethiopathogenesis of many neurodegenerative diseases and stem cell-based therapies could provide an alternative approach to support the regeneration of the damaged tissues [22]. Numerous studies have shown that the hGMSCs regulate the immune system by the inhibition of the proliferation of T and B cells [23], reducing apoptosis and influencing the production of pro-inflammatory and anti-inflammatory cytokines [24].

Protein Expression
Western blotting analysis was performed in order to confirm the transcriptomic data. In Figure 4, protein-specific bands were reported. In particular, NF-kB reduction was shown in hGMSC-MOR and in hGMSCs-CBD when compared to the in hGMSCs-CTRL. The TNF-α showed a decreased level in hGMSCs-MOR while it was overexpressed in hGMSCs-CBD. At the same time, TGF-β1 showed an increased expression in hGMSCs-MOR and in hGMSCs-CBD ( Figure 4A,B).

Discussion
Inflammatory responses play a central role in the ethiopathogenesis of many neurodegenerative diseases and stem cell-based therapies could provide an alternative approach to support the regeneration of the damaged tissues [22]. Numerous studies have shown that the hGMSCs regulate the immune system by the inhibition of the proliferation of T and B cells [23], reducing apoptosis and influencing the production of pro-inflammatory and anti-inflammatory

Discussion
Inflammatory responses play a central role in the ethiopathogenesis of many neurodegenerative diseases and stem cell-based therapies could provide an alternative approach to support the regeneration of the damaged tissues [22]. Numerous studies have shown that the hGMSCs regulate the immune system by the inhibition of the proliferation of T and B cells [23], reducing apoptosis and influencing the production of pro-inflammatory and anti-inflammatory cytokines [24].
In this study, hGMSCs were pre-treated with MOR (0.5 µM) or CBD (5 µM), two phytocompounds extensively studied for their antioxidant [25] and anti-inflammatory [17,26] properties. Specifically, the dose of CBD for the treatment of the hGMSCs was chosen based on previous studies in which we have already demonstrated that it is the lowest dose that effectively exerts anti-inflammatory action [20,25]. In contrast, hGMSCs-MOR were treated with a low concentration of MOR because our group showed that higher doses do not change the transcriptomic profile of the hGMSCs [13]. The MOR or CBD treatment of hGMSCs does not induce cytotoxicity as highlighted in MTT ( Figure 2E) and does not modify the morphology of the cells (Figure 2B,D). Here, we focused on the pathways mediated by TNF-α, IL-1, IL-6 and TGF-β.
TNF-α is a pro-inflammatory cytokine that interacts with TNF receptor 1 (TNFR1), expressed in most tissues [27]. In our transcriptome, the TNFRSF1A gene, which encodes for TNFR1, is downregulated in hGMSCs-MOR and upregulated in hGMSCs-CBD. The result matches the Western blot analysis (Figure 4) that shows a reduced level of TNF-α in hGMSCs-MOR but not in hGMSCs-CBD suggesting a drop in TNF-α signaling after MOR treatment. When TNF-α binds TNFR1, it induces the activation of different biological pathways mediated by complexes I and II [28]. Both the complexes are activated from the Receptor-Interacting Protein Kinase 1 (RIPK1) protein, that along with TNFR1, plays the main role in the biological response [29]. RIPK1 is encoded by the RIPK1 gene that is downregulated in hGMSCs-MOR and upregulated in hGMSCs-CBD. In hGMSCs-MOR, the simultaneous downregulation of TNFRSF1A and RIPK1 hinders complexes I and II activation and strengthens the hypothesis of a reduction in the intracellular response.
In particular, complex I actives inflammation, tissue degeneration, cell proliferation or survival. As a result of the inflammatory process, RIPK1 seems to promote the pro-inflammatory response by NF-kB or the cascade activation of kinases [30]. Specifically, RIPK1 needs LUBAC complex to exert its activity, which adds ubiquitin chains to RIPK1. SHARPIN encoded by the SHARPIN gene is an additional component of LUBAC that linearly ubiquitinates IKK-γ, the regulatory subunit of the Inhibitor of kB Kinase (IKK) complex [31]. In this way, SHARPIN and RIPK1-ubiquitinated activate the IKK complex leading to the phosphorylation of the IκBα proteins. This phosphorylation triggers the degradation of IκBα. Consequently, the NF-κB is liberated from IκBα and translocates into the nucleus where it exerts its action [32]. Therefore, the propagation of the inflammatory signal needs the functional activation of both SHARPIN and RIPK1 that are alternatively downregulated in hGMSCs-MOR and hGMSCs-CBD. According to this result, the Western blot analysis shows reduced levels of NF-κB in both the treatments (Figure 4). CHUK gene, expressed in our transcriptome, is downregulated in hGMSCs-CBD and upregulated in hGMSCs-MOR. CHUK, that encodes for IKKα, is another component of the IKK complex [33] likely involved in cell differentiation [34]. In the inflammatory response, the upregulation of CHUK in hGMSCs-MOR is interesting since it acts as part of the canonical IKK complex in the conventional pathway of NF-κB activation [35]. The complex IKK is also modulated by RACK1 protein, encoded by GNB2L1 gene that in our transcriptomes is upregulated in hGMSCs-MOR and downregulated in hGMSCs-CBD. RACK1 is a scaffold protein that can regulate cell growth and mediate stress signals. In particular, GNB2L1 gene is deregulated similarly to CHUK in both the treatments suggesting the inhibition of NF-kB because of the direct interaction of RACK1 with IKK [36].
Moreover, CLIP3 is an adaptor protein involved in recruiting CYLD into the TNFR1 signaling that facilitates CYLD-mediated deubiquitination of RIPK1 in TNF-α signaling [37]. Our results also show that CLIP3, that encodes for CLIP3 proteins, is downregulated in both the treatments. Instead, USP21 and CYLD are downregulated in hGMSCs-MOR and upregulated in hGMSCs-CBD. USP21 encodes for the ubiquitin-specific peptidase 21 while CYLD encodes for CYLD lysine 63 deubiquitinase. They modulate negatively the TNF-α signaling removing the ubiquitins from the RIPK1, as previously described [38]. Noteworthy, the upregulation of both USP21 and CYLD in hGMSCs-CBD should lead to a lower amplification of the signal mediated by the TNF-α. Moreover, in our transcriptome, the genes encoding for the ubiquitins are UBB, RPS27A, UBA52 and UBC. UBB is downregulated in hGMSCs-MOR and upregulated in hGMSCs-CBD while the other ubiquitins are deregulated in the opposite manner. The ubiquitins act as a scaffold for the activation of the complex I and their overall downregulation mediated by hGMSCs-CBD could be useful to reduce the chronic inflammation stimuli [39]. This is another mechanism that shows how CBD exerts its anti-inflammatory action negatively regulating the TNF-α pathway.
RIPK1-ubiquinated can also promote the activation of map kinases by phosphorylation of TAK1 protein, encoded by MAP3K7 gene that is downregulated in hGMSCs-MOR and hGMSCs-CBD. TAK1 is a precursor protein of the Mitogen Activated Protein Kinase (MAPK) pathway and intriguing the downregulation of MAP3K7 reduces the activation of the cascade that leads to inflammation [40].
RIPK1 is also implied in the apoptosis without recruiting NF-kB when it is associated in the complex II. It recruits the caspase 8 and, finally, promote the cell death [41]. The caspase 8 is an initiator of the apoptosis but it plays also an essential role in the inflammation [42]. The downregulation of CASP8 in both the treatments and the MTT assay ( Figure 2E) suggest the reduction of the cell death in hGMSCs-MOR and hGMSCs-CBD.
In parallel, a new scenario hypothesizes the activation of NF-kB by the Signal Peptidase Like 2A (SPPL2A) protein. Indeed, SPPL2A is an intramembrane protease that seems to promote the inflammatory response in activated dendritic cells by cleaving the TNF-α [43]. It is encoded by SPPL2A gene and it is downregulated in hGMSCs-MOR and upregulated in hGMSCs-CBD. The lower expression in hGMSCs-MOR suggests the reduction of the inflammation mediated by NF-kB.
IL-1 is a family of cytokines including IL-1α and IL-1β that are active in inflammatory processes [44]. IL-1 triggers its effects after the interaction with the membrane receptor Interleukin 1 Receptor type 1 (IL1R1) encoded by IL1R1 gene that is upregulated in hGMSCs-MOR and downregulated in hGMSCs-CBD. The interaction between IL-1 and IL1R1 causes a conformational change of the receptor that determines its association with the adaptive protein MyD88. MyD88, encoded by the MYD88 gene, is a signal transducer that promotes the subsequent NF-κB-mediated transcription [45]. Our results show that both hGMSCs-MOR and hGMSCs-CBD downregulate the expression of the gene MYD88 with consequent reduction in signal transduction. Similarly to TNF-α signaling, the IL-1 pathway modulates MAP3K7 (TAK1) and CHUCK (IKKα) genes. Therefore, the downregulation of MAP3K7 in both our experimental groups matches the downregulation of MYD88 and it suggests the reduction NF-kB activation. In addition, our transcriptomic profiles show that SQSTM1 gene is downregulated both in hGMSCs-MOR and hGMSCs-CBD whereas MAP3K3 gene is upregulated in both the treatments. SQSTM1 encodes for p62, a multifunctional protein that acts as an adaptive scaffold/protein associated with Mitogen-Activated Protein Kinase Kinase Kinase 3 (MAP3K3) [46]. MAP3K3 is encoded by the MAP3K3 gene and it is a component of a protein kinase cascade that mediates activation of NF-kB [47]. Therefore, the downregulation of the SQSTM1 gene in both experimental groups leads to less formation of the MAP3K3/p62 complex and, consequently, it contributes to reduce NF-kB effects.
IL-6 is a pro-inflammatory cytokine that contributes to host defense by means of the stimulation of pro-inflammatory responses and immune reactions [48]. When the IL-6 receptor (IL6R) is in complex with the Interleukin 6 Signal Transducer (IL6ST) protein, it can mediate the signal transduction. IL6ST is encoded by IL6ST that in our results is downregulated both in hGMSCs-MOR and hGMSCs-CBD. IL6ST selectively engages a member of the Janus Kinase family (JAK) [49] like TYK2. TYK2 gene, that encodes for TYK2, is upregulated in hGMSCs-MOR but downregulated in hGMSCs-CBD. TYK2 is a tyrosine kinase that phosphorylates IL6ST creating a binding site for the transcription factor STAT3. STAT3 is encoded by the STAT3 gene and it is downregulated in both the treatment. As soon as STAT3 binds IL6ST, it is phosphorylated by TYK2 [50]. The phosphorylation of STAT3 promotes its dimerization and translocation into the nucleus where it starts transcription. The simultaneous downregulation of IL6ST, TYK2 and STAT3 strongly suggests the inhibition of the JAK/STAT pathway in hGMSCs-CBD. In parallel, IL-6 can also trigger the activation of the non-classic PI3K/AKT/mTOR pathway. Specifically, the MOR pretreatment of hGMSCs downregulates PIK3CA, PIK3CB, and PIK3CD that encode for the catalytic subunits of PI3K, restricting the activation of this cascade pathway by IL-6. On the contrary, the CBD pretreatment downregulates only PIK3CD but upregulates PIK3CB and PIK3CD. Nevertheless, both the treatments upregulate AKT1. It encodes AKT Serine/Threonine Kinase 1 that promotes cell survival by inhibition of apoptosis [51]. Interestingly, MTOR gene, that encode mTOR, is upregulated in hGMSCs-MOR whereas in hGMSCs-CBD there is not any statistically significant change in its expression level. The role of mTOR is quite important because it is implied in cell proliferation and neurogenesis [52].
The hGMSCs-MOR and hGMSCs-CBD modulate the anti-inflammatory pathway mediated by TGF-β that is involved in proliferation, differentiation, migration, cellular homeostasis and neuronal survival [53]. The FURIN protein, encoded by the FURIN gene, promotes the functional activation of the TGF-β by the cleavage of its N-terminal [54]. In our results, FURIN is upregulated in both treatments. The activated TGF-β binds the TGF-β Receptor type I (TGFβRI) and type II (TGFβRII), encoded by TGFBR1 and TGFBR2 genes, respectively [55]. Our results show that TGFBR1 is upregulated in both the treatment while TGFBR2 is upregulated only in hGMSCs-MOR. Interestingly, the simultaneous upregulation of FURIN, TGFBR1 and TGFBR2 genes makes to hypothesize the complete trigger of the signaling in hGMSCs-MOR whereas the downregulation of TGFBR2 suggests a partial trigger in hGMSCs-CBD. Moreover, the Western blot analysis in Figure 4 shows higher level of TGF-β1 in both treatments. FKBP1A is a cis-trans peptidyl-prolyl isomerase that prevents TGFβRI phosphorylation by TGFβRII [56]. It is encoded by the FKBP1A gene that is upregulated in hGMSCs-MOR and downregulated in hGMSCs-CBD. When TGF-β binds TGFβRII, TGFβRI complexes with TGFβRII that auto-phosphorylates itself and trans-phosphorylates TGFβRI. TGFβRI phosphorylates SMAD3 that acquires affinity towards SMAD4 protein that finally translocates into the nucleus in which it promotes transcription [57]. SMAD3 and SMAD4 (SMADs) are signal transducers proteins encoded, respectively, by SMAD3 and SMAD4. SMAD3 is downregulated in hGMSCs-MOR and in hGMSCs-CBD while SMAD4 only in hGMSCs-CBD. Even if SMAD3 is downregulated in both the treatments, the overexpression of SMAD4 in hGMSCs-MOR suggests that the SMAD3 phosphorylation could occur. Inside the nucleus, the transcriptional activity of the SMADs is mediated by WWTR1 [58]. WWTR1 is encoded by the WWTR1 gene and it is downregulated in hGMSCs-MOR and hGMSCs-CBD. The modulation of the TGF-β signal can be differently mediated by NCOR1, SMURF2, STRAP, PP1, XPO1 proteins. NCOR1, SMURF2 and STRAP proteins are, respectively, encoded by NCOR1, SMURF2 and STRAP genes. NCOR1 is a nuclear receptor corepressor that stops the transcription activity of SMADs [59]. SMURF2 is an E3 ubiquitin-protein ligase that adds ubiquitin on SMADs, among which SMAD3, promoting their degradation [60]. STRAP is a Serine-threonine kinase receptor-associated protein that binds the receptors TGFβRI and TGFβRII hindering the propagation of the signal [61]. Both SMURF2 and STRAP genes are upregulated in hGMSCs-MOR and hGMSCs-CBD while NCOR1 only in hGMSCs-CBD. The signal induced by TGF-β can be also negatively modulated by Protein Phosphatase 1 (PP1) that removes the phosphate groups from TGFβR1 [62]. In particular, our transcriptomes express two catalytic subunits encoded by PPP1CA and PPP1CB genes and a regulatory subunit encoded by PPP1R15A gene. PPP1CA is upregulated in hGMSCs-MOR and downregulated in hGMSCs-CBD whereas PPP1R15A and PPP1CB are modulated in reverse. The conclusion of the signal is also mediated by the exportin XPO1 that promotes the translocation of the transcripts from the nucleus to the cytosol. XPO1 is encoded by XPO1 gene that is upregulated in both the treatments. Interestingly, the upregulation of XPO1, STRAP and SMURF2 genes matches with the upregulation of FURIN and TGFBR1 genes both in hGMSCs-MOR and hGMSCs-CBD. It suggests the conclusion of the signal propagation by the treated cells in order to avoid prolonged stimuli.

Ethics Statement
The study followed the guidelines of the Helsinki Declaration (2013). The stem cells were collected under the approval of the Ethical Committee (number 266/17.04.2014) at the Medical School, "G. d'Annunzio" University, Chieti, Italy. The three recruited healthy individuals signed the informative consent form before tissue collection.

Compounds Isolation and Purification
The MOR was extracted from the seeds of Moringa oleifera at the CREA-AA (Bologna Laboratory), as previously described by Brunelli et al. [63]. MOR was purified by anion exchange and size exclusion chromatography. The MOR was purified by High-performance liquid chromatography (HPLC) obtaining a purity of about 99% (based on peak area value) [64]. CBD was extracted by flowerheads of Cannabis sativa L. and it was obtained from a greenhouse cultivation at the Council for Research and Experimentation in Agricolture-Research Center for Industrial Crops (CREA-CIN) in Rovigo, using established methods previously described by Taglialatela et al. [65]. The isolation and purification of cannabinoid was done in accordance with their legal status (Authorization SP/106 23/05/2013 of the Ministry of Health, Rome, Italy). The extracted CBD had a purity >99% and it did not show any trace of THC.

Cell Isolation
The hGMSCs have been obtained from the gingival tissue collected during a routine oral surgery procedure as previously described [66]. Healthy patients scheduled to remove the third molar for orthodontic purposes were enrolled in the study. In order to exclude most keratinocytes resident in the gingival, the gingival specimens were completely de-epithelialized with a scalpel. The connective tissues were grinded and after washed several times with phosphate buffered saline (PBS; EuroClone, Milan, Italy). Tissue explants were cultured in tissue culture dishes with Mesenchymal Stem Cells Growth Medium-Chemically Defined (Lonza, Basel, Switzerland). The medium was replaced with a fresh one every two days. After two weeks of culture, cells spontaneously migrated from tissue explants. All experiments were performed with cells at 2nd passage. After the isolation process, cells were treated with MOR (0.5 µM) or CBD (5 µM) for 48h.

Cell Characterization
To evaluate the mesenchymal features of hGMSCs, the cytofluorimetric detection and mesengenic differentiation were performed. The cytofluorimetric analysis was assayed as previously described [67]. Expression of Sox-2, Oct3/4, CD13, CD14, CD29, CD34, CD45, CD73, CD90, and CD105 was evaluated on hGMSCs. The analysis was performed by using FACStarPLUS flow cytometry system and the FlowJo™ software (TreeStar, Ashland, OR, USA). To assess the ability to differentiate into osteogenic and adipogenic commitment, hGMSCs were maintained under osteogenic and adipogenic conditions for 21 and 28 days, respectively, as reported by Pizzicannella, et al. [68]. To evaluate the formation of mineralized precipitates and lipid vacuoles, after the differentiation period, alizarin red and Oil red O staining were performed on undifferentiated and differentiated cells. Inverted light microscopy Leica DMIL (Leica Microsystem, Milan, Italy) was used for samples observations. To validate the ability to differentiate into osteogenic and adipogenic lineages, the expression of RUNX-2, ALP, FABP4, and PPARγ were evaluated by RT-PCR as reported by Pizzicannella, et al. [68]. Commercially available TaqMan Gene Expression Assays (RUNX-2 Hs00231692_m1; ALP Hs01029144_m1; FABP4 Hs01086177_m1; PPARγ Hs01115513_m1) and the Taq-Man Universal PCR Master Mix (Applied Biosystems, Foster City, CA, USA) were used according to standard protocols. Beta-2 microglobulin (B2M Hs99999907_m1) (Applied Biosystems) was used for template normalization. Real-time PCR was performed in three independent experiments, and duplicate determinations were carried out for each sample.

Cell Culture and Treatments
The hGMSCs were cultured in monolayer using DMEM-High Glucose (SIGMA-ALDRICH, Co.) containing 10% fetal bovine serum (FBS) (Sigma-Aldrich Co. Ltd). The cells were grown in a logarithmic phase at 37 • C in a moisturized atmosphere of 5% CO2 and 95% air. The hGMSCs grown until at 70-80% confluence were incubated for 48h in the medium containing 5 µM CBD (dissolved in 0.1% DMSO) or 0, 5 µM of MOR (dissolved in 0.1% DMSO). The hGMSCs treated with 0.1% DMSO were included as controls. After 48 h incubation, cells were harvested for the other analysis.
The experiment was repeated in triplicates.

Cell Viability Assay
The proliferation ability of hGMSCs cultured with or without MOR (0.5 µM) or with CBD (5 µM) for 48 h, were determined using the 3-(4,5-dimethylthiazolyl-2)-2,5-diphenyltetrazoliumbromide (MTT) method. At the end of the treatment, cells were incubated with medium containing MTT (0.5 mg/mL; Sigma-Aldrich) at 37 • C for 4 h. Living cells are able to reduce yellow tetrazole to purple insoluble formazan. The formed formazan crystals were dissolved in acid isopropanol at 37 • C for 1 h. The absorbance level of MTT cleavage was directly proportional to the number of viable cells and indirectly indicate proliferation rate after MOR or CBD treatment. Supernatants were read at 650 nm wavelength using a microplate reader (Synergy HT, BIoTek Instruments, VT).

Morphological Analyses
To evaluate the morphological features in untreated and treated cells, samples were processed for immunofluorescence detections. Untreated hGMSCs and treated hGMSCs with MOR and CBD, were cultured on eight-well chamber slides and then were fixed using 4% paraformaldehyde diluted in 0.1 M sodium phosphate buffer (PBS, Lonza). After the fixation step, cells were permeabilized with 0.5% Triton X-100 in PBS for 10 min, followed by blocking with 5% non-fat dried milk in PBS for 30 min [69]. Cells were incubated by Alexa Fluor 488 phalloidin green fluorescence conjugate (1:400, Molecular Probes, Invitrogen, Eugene, OR, USA) in order to mark the cytoskeleton actin. After immunofluorescence labelling cells were washed and incubated with TOPRO (1:200, Molecular Probes) for 1 h at 37 • C for nuclei evaluation [70]. Samples were observed under a Zeiss LSM800 confocal system (Zeiss, Jena, Germany).

RNA Extraction and NGS Analysis
The extraction of the total RNA was performed with the Reliaprep RNA cell Miniprep System (Promega, Madison, WI, USA). The total RNA was fragmented according to TruSeq RNA Access library kit protocol (Illumina, San Diego, CA, USA). The SuperScript II Reverse transcriptase (Invitrogen, Carlsbad, CA, USA) was used to synthesize the cDNA. In order to facilitate adaptor ligation in the next step, the 3'end of the cDNA were adenylated and then ligation of indexing adaptors was performed. The PCR was used to amplify the library and the AMPure XP beads (Beckman Coulter, Brea, CA) for the clean-up. PCR (15 cycles of 98 • C for 10 s, 60 • C for 30 s and 72 • C for 30 s) was performed to select those DNA fragments that have adapter molecules on both ends and to amplify the amount of DNA in the library. After validation of the library, the first hybridization step (18 cycles of 1 min incubation, starting at 94 • C, and then decreasing 2 • C per cycle) was performed using exome capture probes. Then, to capture probes hybridized to the target regions, streptavidin coated magnetic beads were used; then, the enriched libraries were eluted and prepared for a second round of hybridization. The second hybridization was required to ensure high specificity of the capture regions. A second capture with streptavidin-coated beads was performed, the enriched libraries were eluted and cleaned up by AMPure XP beads. The second PCR amplification step was performed by ten cycles. The library was quantified by qPCR KAPA Library Quantification Kit-Illumina/ABI Prism ® (Kapa Biosystems, Inc., Wilmington, MA, USA). Agilent High Sensitivity DNA Kit on a Bioanalyzer was used for validation of libraries. The size range of the DNA fragments was measured to be in the range of 200-650 bp and peaked around 250 bp. The library was finally normalized at 12 pM and sequenced using MiSeq Reagent Kit v3 by Illumina MiSeq Instrument in single read mode. The transcriptome profile was aligned against the reference genome "UCSC hg19" and the fold change of the differentially expressed genes of each treatment against hGMSCs-CTRL were evaluated by Cufflinks package. The experiment was repeated in triplicate.

Bioinformatics Analysis and Database Inspection
The biological role of differently expressed gene were inspected taking advantage of the Reactome database from the online website (https://reactome.org/). In particular, we focused on the pathways "TNF signaling", "Interleukin-1 signaling", "Interleukin-6 signaling" and "Signaling by TGF-beta Receptor Complex" in order to study TNF-α, IL-1, IL-6 and TGF-β pathways.

Conclusions
The pretreatment of hGMSCs with CBD or MOR has an anti-inflammatory activity enhancing the TGF-β pathway, involved in the regulation of multiple biological processes, including inflammation. Concerning the pro-inflammatory pathways, TNF-α signaling is reduced from the beginning in hGMSCs pretreated with MOR as confirmed by Western blot analysis, while CBD exerts its activity after the process starts.
Contrarily, the treatment with CBD selectively inhibits the IL-1 pathway while MOR is less efficient. Furthermore, IL-6 signaling is modulated by CBD that downregulates the pro-inflammatory JAK/STAT and by MOR that enhances the PI3K/AKT/mTOR pathway, which leads to cell survival and proliferation.