Physiological Expression of Ion Channel Receptors in Human Periodontal Ligament Stem Cells

The etiopathogenesis of neurodegenerative diseases is characterized by the death of neurons. Human periodontal ligament stem cells (hPDLSCs), coming from neuronal crest, can potentially become neuronal cells because of their embryologic origin. In this study, we performed an RNA-seq analysis of hPDLSCs in order to determine whether their transcriptomic profile revealed genes encoded for ion channel receptors. Next, each found gene was enriched by the information of pathways stored in the Reactome database. Our results show that the hPDLSCs express GABBR1 and GABBR2, CHRNA1, GRINA genes, respectively associated with GABAB, NMDA and nACh receptors. In particular, the two subunits of GABAB receptor are expressed in hPDLSCs. Further, the proteic extract for GABABR1, GABABR2 and AChRα1 confirmed their expression in hPDLSCs. Our results show that hPDLSCs express physiologically genes associated with ion channel receptors maintaining multipotent features which are useful for neurogenesis.


Introduction
The embryonic head development, together to dental structure formation, is founded on a complex and delicate process directed by a specific, hierarchically-defined genetic program [1]. Teeth are ectodermal appendages whose development is influenced by reciprocal interactions between the surface epithelium (ectoderm) and the underlying neural crest-derived mesenchyme. These interactions are arbitrated by several conserved signaling pathways, such as: Wnt, bone morphogenetic protein (BMP), fibroblast growth factor (FGF), sonic hedgehog (SHH), and ectodysplasin A (EDA) [2]. The neural crest (NC), the embryonic structure present in vertebrates, is indicated as a "fourth germ layer"; it is constituted from multipotent cells, named neural crest cells (NCCs), which are able to differentiate into a surprising array of adult structures. NCCs are a transient multipotent stem cell population that originally appears at the junction of the epidermal and neural ectoderm by reciprocal interactions between these tissues and by signals from the mesoderm [3]. During neurulation process, NCCs begin to delaminate from the dorsal region of the neural epithelium, going toward epithelial-mesenchymal transition [4]. Subsequently, they migrate into their destination in the embryo to generate various groups of cells, including: sensory, autonomic, and enteric ganglia within the peripheral nervous system; and the adrenal medulla, melanocytes, in addition to skeletal, connective, adipose, and endocrine cells [5]. Actually, several studies indicate that NCCs play a crucial role in embryonic development, representing a likely stem cell population that may be an innovative approach 4 • C in the dark. The hPDLSCs were stained using the following antibodies: anti-CD29, anti-CD44, anti-CD45, anti-CD105 (Ancell, Bayport, MN, USA); anti-CD14 (Milteny Biotec, Bergisch Gladbach, Germany); anti-CD73, anti-CD90, anti-OCT3/4, anti-Sox2 and anti-SSEA4 (Becton Dickinson, BD, San Jose, CA, USA); anti-CD34 (Beckman Coulter, Fullerton, CA, USA). After incubation, the cells were acquired with a flow cytometer (FACSCalibur TM ; Becton Dickinson, BD, San Jose, CA, USA). Data were analyzed by the FlowJo software v8.8.6 (TreeStar, Ashland, OR, USA).

Mesengenic Differentiation
The hPDLSCs were induced to osteogenic and adipogenic differentiation commitment, as previously reported by Cianci et al. [19]. Briefly, at the end of differentiation process, the cells were stained with Alizarin Red S (Sigma-Aldrich, Milan, Italy) and Adipo Oil Red (Lonza, Basel, Switzerland) solution to evaluate the osteogenic and adipogenic differentiation, respectively, and observed by means of light microscopy, Leica DMIL (Leica Microsystem, Milan, Italy) [20].

Statistical Analysis
Data were analyzed using GraphPad Prism 6.0 (GraphPad Software, La Jolla, CA, USA). The statistical analyses were performed with the ANOVA test. The p-value threshold used to validate the analysis is 0.05.

RNA Isolation and Real Time-PCR Analysis
To assess osteogenic and adipogenic differentiation ability of PDLSCs, the total RNA was isolated using the Total RNA Purification Kit (NorgenBiotek Corp., Ontario, CA, USA) according to the manufacturer's instructions. The RNA was extracted from a pool of 5 × 10 5 hPDLSCs. M-MLV Reverse Transcriptase reagents (Applied Biosystems) were used to generate cDNA. Real-Time PCR was carried out with the Mastercycler ep realplex real-time PCR system (Eppendorf, Hamburg, Germany). hPDLSCs expression of Runt-related transcription factor-2 (RUNX-2) and AlkalinPhospatase (ALP) was evaluated after 28 days in osteogenic differentiated culture, the expression of Fatty Acid Binding Protein 4 (FABP4) and Peroxisome Proliferator-Activated Receptor γ (PPARγ) were analysed after 28 days of adipogenic differentiation culture. Commercially available TaqMan Gene Expression Assays (RUNX-2 Hs00231692_m1; ALP Hs01029144_m1; FABP4 Hs01086177_m1; PPARγ Hs01115513_m1) and the TaqMan Universal PCR Master Mix (Applied Biosystems, Foster City, CA, USA) were used according to standard protocols. Beta-2 microglobulin (B2M Hs99999907_m1) (Applied Biosystems, Foster City, CA, USA) was used for template normalization [21]. RT-PCR was performed in three independent experiments; duplicate determinations were carried out for each sample.

Scanning Electron Microscopy
The hPDLSCs at second passage were fixed for 4 h at 4 • C in 4% glutaraldehyde in 0.05 M phosphate buffer (pH 7.4), and subsequently dehydrated in increasing ethanol concentrations and then critical-point dried. Samples were then mounted onto aluminum stubs and gold-coated in an Emitech K550 (Emitech Ltd., Ashford, UK) sputter-coater before imaging by means Zeiss SEM EVO 50 (Zeiss, Jena, Germany) [22].

RNA-Seq Preparation
The Reliaprep RNA cell Miniprep System (Promega, USA) was used to extract the total RNA from all of the samples. Each of them was treated with 0.1% of DMSO. The TruSeq RNA Access library kit protocol (Illumina, San Diego, CA, USA) was followed. From the total RNA, 40 ng were extracted and fragmented through a thermal cycler for 8 min at 94 • C. The achieved fragments, >200 nt, were used to synthesize a first strand of cDNA using the activity of the reverse transcriptase SuperScript II (Invitrogen, Carlsbad, CA, USA). The Second Strand Marking Master Mix was used to obtain the double strand of the cDNA. The incubation ran for 1 h at 16 • C. The reaction mix was eliminated by AMPure XP beads purification. Fragment adhenilation was performed at 3 ends in order to ligate the complementary adapters, thereby avoiding huge chimera development. The identification of the samples and their preparation for flow cell hybridization was performed by the ligation of Adapter-Indexes to the fragment of cDNA. A first step of clean-up was run before the PCR amplification (denaturation: 30 s at 98 • C; 15 cycles: 10 s at 98 • C, 30 s at 60 • C, 30 s at 72 • C; extension: 5 min at 72 • C). The regions of interest are selected and enriched via a first reaction of hybridization (10 min at 95 • C, un minute of incubation per 18 cycles, 90 min at 58 • C). The procedure allowed us to mix the exome capture probes with the cDNA library. Starting from 200 ng of each library, the result is a pool of libraries which were differently indexed. The purification of the pool was obtained by streptavidin conjugated magnetic beads. Then, subsequent hybridization and purification were performed. A second PCR amplification was done following the same protocol as that of the first one, except for the cycles (just 10 instead of 15). The process ended with a clean-up. The Bioanalyzer instrument (Agilent High Sensitivity DNA Kit, Richardson, TX, USA) was used to validate the quality of the library. The validation of the quantity was made using Real-Time PCR (KAPA Library Quantification Kit-Illumina/ABI Prism@) (Kaba Biosystem, Inc. Wilmington, MA, USA). A denaturation step through 2N NaOH was performed, and then it was diluted until it reached a concentration of 12 pM. The MiSeq Instrument (Illumina, San Diego, CA, USA) was used to sequencing (by single read setting) using MiSeq Reagent Kit v3 for 150 cycles.

RNA-seq Data Processing and Gene Analysis
The MiSeq instrument provides multiplexed samples in "bcl" format. The CASAVA software (version 1.8) was used to convert these data in demultiplexed "Fastq" files, one for each sample. Then, the quality of the reads was checked using the fastQC software. The reads were trimmed of the adapters by Trimmomatic [24]. Finally, the reads are aligned to the reference genome "homo sapiens UCSC hg19" using STAR RNA-seq aligner [25]. All the transcripts which were common among the samples were then collected together using Cufflinks [26]. The set of genes that transcribed the found transcripts were used for the study. The whole gene list was inserted as file in the "Analysis" tools section of the Reactome database. From this list, a subset of 64 genes included in the "Neuronal System" pathways were studied. A deeper inspection of each of these genes showed that SAT1 was identified by a UNIPROT [27] entry that is related to SLC38A1 gene, and P3H3 is identified by a UNIPROT linked to GNB3. Finally, SAT1, P3H3 and HSPA8 genes are excluded from the analysis. By a literature inspection, were searched for genes encoded for proteins associated with ion channel receptors and present in this list of genes. GABAergic, Glutamatergic and Cholinergic pathway associations were found. The genes in the GABAergic subpathways are expressed in GABAergic "GABA receptor activation" and "GABA synthesis, release, reuptake and degradation", while for Glutamatergic ones, "Glutamate binding, activation of AMPA receptors and synaptic plasticity", "Activation of kainate receptors upon glutamate binding", "Glutamate Neurotransmitter Release Cycle", "Astrocytic Glutamate-Glutamine Uptake and Metabolism", "Activation of NMDA receptors and postsynpatic events", as well for Cholinergic "Acetylcholine Neurotransmitter Release Cycle" and "Activation of Nicotinic Acetylcholine Receptors".

Western Blot Analysis
The proteins were collected from hPDLSCs as previously described [9]. In parallel, the proteins from DPK-SKDF-H were also collected and used as a control. Protein concentrations were measured using Bradford assay (Bio-Rad Laboratories, Inc., Hercules, CA, USA). Twenty-five µg of proteins were heated for 5 min at 95 • C and resolved by SDS-polyacrylamide gel electrophoresis (SDS-PAGE), and then transferred onto a PVDF membrane (Amersham Hybond, GE Healthcare Life Sciences, Milan, Italy). Membranes were blocked with 5% skim milk in Phosphate Buffered Saline (PBS) for 1 h at room temperature. Afterwards, the membranes were incubated overnight at 4 • C with the following primary antibodies: GABAB R1 (1:500; Santa Cruz Biotechnology, Dallas, TX, USA), GABAB R2 (1:500; Santa Cruz Biotechnology, Dallas, TX, USA) and AChRα1 (1:500; Santa Cruz Biotechnology, Dallas, TX, USA). After washing the membranes with PBS 1X, they were incubated with horse radish peroxidase (HRP)-conjugated anti-mouse or anti-rat IgG secondary antibodies (1:2000; Santa Cruz Biotechnology, Dallas, TX, USA) for 1 h at room temperature. The membranes were also incubated with an antibody for glyceraldehyde 3-phosphate dehydrogenase (GAPDH) HRP Conjugated (1:1000, Cell Signaling Technology, Danvers, USA). The bands were analyzed through an enhanced chemiluminescence system (Luminata Western HRP Substrates; Millipore, Burlington, MA, USA), and the protein bands were acquired through ChemiDoc™ MP System (Bio-Rad Laboratories Inc., Hercules, CA, USA).

MTT Assay Analysis
In order to access to the proliferation rate of the hPDLSCs, an MTT assay analysis was performed. The proliferative rate was detected at 24 h, 48 h, 72 h and after 1 week of culture, as shown in the Figure 1B. The proliferation rate grew according to a logarithmic pathway.

Mesengenic Differentiation and Genetic Evaluation
To evaluate osteogenic differentiation, hPDLSCs were stained with an Alizarin Red S solution. Calcium precipitates were detected in hPDLSCs culture after 28 days of incubation ( Figure 1F). The adipogenic differentiation was evaluated using lineage Oil-red O solution staining; cells showed lipid droplets at the cytoplasmic level ( Figure 1G). To confirm the osteogenic and adipogenic commitment, RT-PCR was performed. RUNX2, ALP, PPARγ and FAPB4 markers were expressed in hPDLSCs maintained under differentiation ( Figure 1C,D). Both mesengenic differentiations showed statistically significant differences between undifferentiated and differentiated cells.

Scanning Electron Microscopy Analysis and Immunofluorescence
Glass adherent cells showed at SEM analysis a fibroblast like morphological shape with ovoidal nuclei and two or more nucleoli are visible. Long cytoplasmic processes make contact with neighboring cells (Figure 2A). Confocal microscopy observation showed the same morphological aspects ( Figure 2B). Immunofluorescence of the hPDLSCs was performed in order to observe the expression of specific neurogenic markers. It showed a positivity for nestin at cytoplasmic level and for p75 localized at the nuclear level ( Figure 2C,D).

Reactome and Pathway Analysis
The NGS analysis returns files in Fastq format containing the sequenced reads of our cDNA pool. The genes obtained after the NGS analysis were studied by the "Analysis" tool in the Reactome database that recognizes 2905 genes. Sixty-four of these genes are part of "Neuronal System" pathway. The analysis of hPDLSCs transcriptomic profile reveals genes related to ion channel receptors for GABA B (GABBR1 and GABBR2), acetylcholine (CHRNA1) and glutamate (GRINA) ( Table 1). Genes are distributed as shown in Fig 3. The pathways related to GABA neurotransmitters are "GABA receptor activation" and "GABA synthesis, release, reuptake and degradation" (Figure 3A). Among those, 3 genes encoded for Adenylate Cyclase proteins act as Signal Transductors in the "Inhibition of adenylate cyclase pathways" that is directly linked to "Neuronal system". ( Figure 3B). The pathways related to glutamate are collected in "Glutamate binding, activation of AMPA receptors and synaptic plasticity", "Activation of kainate receptors upon glutamate binding", "Glutamate Neurotransmitter Release Cycle", "Astrocytic Glutamate-Glutamine Uptake and Metabolism" and "Activation of NMDA receptors and postsynpatic events" ( Figure 3C). Among those genes, 17 are involved in "Signal Transduction" (Figure 3D). The pathways related to acethylcoline are inside "Acetylcholine Neurotransmitter Release Cycle" and "Activation of Nicotinic Acetylcholine Receptors" (Figure 3E). None of the genes in these pathways is involved also in "Signal Transduction" (Figure 3F). Table 1.
Genes of our dataset included in Cholinergic, GABAergic, Glutamatergic REACTOME pathways.

Western Blot Results
Western Blot analysis was used to confirm the expression of the GABA B and Acetylcholine receptors. As shown in Figure 4, the proteic extract of hPDLSCs expresses GABA B R1, GABA B R2 and AChRα1. In contrast, the proteic extract of DPK-SKDF-H does not express these receptors.

Discussion
The etiopathogenesis of the neurodegenerative diseases is linked to the death of cells in nervous tissue. Their incidence is growing and stem cells are recently receiving attention from researchers because they can potentially replace damaged nervous cells. In particular, the hPDLSCs are becoming important because there is no ethical concern associated with their usage, they are easy to retrieve, derive from neural crest origin, and they can differentiate into mesengenic and neurogenic lineages. [28][29][30]. Indeed, the hPDLSCs cultured in basal medium spontaneously express neural markers as Nestin and GAP-43. Nestin is a protein belonging to class VI of intermediate filaments, expressed in stem/progenitor cells in the mammalian CNS during development [31,32]. In order to better understand the multipotent state of hPDLSCs, their physiological expression of ion channels receptors was investigated. The transcriptomic profile of the hPDLSCs was retrieved by three control individuals taking advantage of NGS technology. The RNA-seq analysis reveals 4378 genes express. In order to investigate all the genes involved in the nervous system, a database containing biological pathways (Reactome) was used. The "Neuronal System" pathway counts 64 genes. The hPDLSCs physiologically express the genes GABBR1 and GABBR2, GRINA, CHRNA1 related respectively to GABA B , NMDA and nACh receptors ( Figure 5). For this reason, we studied all of the subpathways in which they are involved. A Western Blot analysis of hPDLSCs confirmed the expression of the GABA B R1, GABA B R2 and AChRα1 proteins. In order to verify whether the expression of GABBR1 and GABBR2 is specific to the hPDLSCs, the lineage of fibroblasts DPK-SKDF-H was used as control. Our results showed that DPK-SKDF-H does not express proteins for these receptors (Figure 4). GABBR1 and GABBR2 are indirectly involved with ion channel membrane proteins through the guanine nucleotide-binding proteins (G proteins). In hPDLSCs, GNG10 and GNB1 genes are expressed, that encode for two subunits of G proteins. In particular, GNG10 encodes for a protein that directly interacts with GABBR2 subunit [33]. The G proteins complex promotes the flow of potassium and calcium ions and acts as an inhibitor for Adenylate Cyclase proteins. The hPDLSCs express the ADCY3, ADCY4 and ADCY9 genes, that encode respectively for Adenylate Cyclase type 3, type 4 and type 9 that catalyze ATP in cyclic adenosine monophospate (cAMP) [34]. Moreover, the hPDLSCs express also ABAT and GLUL genes that encode for proteins involved in the catabolism of GABA. ABAT encodes for the GABA-transaminase enzyme that converts GABA into succinic semialdehyde. The GLUL gene encodes for the glutamine synthetase enzyme that synthesizes L-glutamine in L-glutamate. SLC1A3 and SLC38A1 encode for carrier proteins. Specifically, the SLC1A3 gene encodes for amino acid transporter 1 protein, that is involved in the transport of the glutamate [35], whereas SLC38A1 encodes for the sodium-coupled neutral amino acid transporter 1, that has high affinity for glutamine [36]. In hPDLSCs, we also find RIMS1 and STX1A that are involved in the exocytosis of GABA. RIMS1 gene, that belongs to RAS superfamily, is involved in calcium channel modulation preparing the active zone of the synapsis for exocytosis [37]. The syntaxin-1A, protein encoded by STX1A, is characterized by a SNARE (SNAp REceptor) domain that allows the fusion to take place of the vesicles with the presynaptic membrane [38]. The hPDLSCs express also genes related to glutamate. Glutamate is a neurotransmitter that can be intercepted by metabotropic or ionotropic receptors. neurotransmitter. In the figure, all the genes that were found in hPDLSCs are highlighted. For a better reading, the genes ADCY3, ADCY4 and ADCY9 are clustered in "ADCY". The genes PLCB1 and PLCB3 are grouped in "PLCB". The genes CALM1 and CALM2 are bundled in "CALM".
Our results show that hPDLSCs only express the GRINA gene, a transmembrane NMDA subunit associated protein that binds the glutamate. Further, PLCB1 and PLCB3 are two genes that encode for Phospholipase C Beta proteins that allow to release the diacylglycerol (DAG), another secondary messenger [39]. The hPDLSCs express also several genes which encoded proteins are mainly located in the postsynaptic neurons. These proteins are involved in the maturation of the synapses and in the trafficking of the neurotransmitter. In hPDLSCs they are encoded by: AP-2 complex (AP2A1, AP2A2, AP2M1, AP2S1), MYO6, NSF, EPB41L1, CALM1, CALM2, DLG3, DLG4, MDMD2. The AP-2 complex, supported by the unconventional myosin VI (encoded by MYO6), reduces the excitatory stimuli by endocytosis of AMPA [40]. The NSF gene encodes for the Vesicle-Fusing. ATPase binds the subunit GluA2 of the AMPA receptor in order to facilitate the fusion of the membrane and the vesicle by the SNARE domain [41]. The trafficking of the receptor AMPA is also regulated by Band 4.1-like protein 1, encoded by EPB41L1 gene, that handles the long-term potential in the synapses mainly by palmitoylation and phosphorylation on GluR1 subunit [42]. Calmodulins are calcium-modulated proteins encoded by genes like CALM1 and CALM2. They are involved in the calcium homeostasis carrying it towards the endoplasmic reticulum or out of the cell [43]. DLG3 encodes for the protein Disks Large homolog 3 that stabilizes NMDA receptor [44] while DLG4 encodes for the protein Disks Large homolog 4 that interacts also with AMPA receptor [45]. The hPDLSCs express also MDM2, encoding for an E3 ubiquitin ligase, that indirectly causes the AMPA receptor endocytosis [46]. The transcriptomic profile of hPDLSCs also reveals a gene encoded for proteins located on presynaptic neurons like PPFIA1 and UNC13B. The PPFIA1 gene, encoded for Liprin-Alpha-1 protein, belongs to the protein family of liprins that is localized closed to cell focal adhesions. Liprins are involved in functional plasticity, the formation of the synapses and in axon guidance [47]. UNC13B is characterized by a SNARE domain that allows the docking to take place among the vesicle and the membrane aimed at the release of glutamate [48]. Only CHRNA1 gene, encoding for one of the five subunit of nAChr, is expressed in cholinergic synapses [49].

Conclusions
The RNA-seq analysis performed in this study reveals the expression of genes that encode for ionotropic and metabotropic receptors, along with proteins located in presynaptic and postsynaptic neurons. The hPDLSCs are known to differentiate in different neuronal cells. In our results, the expression of GABA B R1, GABA B R2 and AChRα1 suggests that a subpopulation of hPDLCSs differentiate in neurons.