Role of Glycine Receptor α3 in the Phosphorylation of Extracellular Signal-Regulated Kinase by Prostaglandin E2

Background: Glycine receptors (GlyRs) play a key role in the processing of inammatory pain. We used Adeno-associated virus (AAV) for GlyRα1/3 gene transfer in F11 neuron cells, and investigated the effects and roles of pAAV-GlyRα1/3 on cell cytotoxicity and the prostaglandin E 2 (PGE 2 )-mediated inammatory response. Methods: pAAV-GlyRα1 and pAAV-GlyRα3 recombinant vectors were constructed, and cell viability was measured following pAAV-GlyRα1/3 transfection. The activation of mitogen-activated protein kinase (MAPK) inammatory signaling and neuronal injury marker activating transcription factor 3 (ATF-3) were evaluated by western blotting; the level of cytokine expression was measured by ELISA. Results: We found that pAAV/pAAV-GlyRα1/3 transfection slightly, but not signicantly, increased cell viability and induced extracellular signal-regulated kinase (ERK1/2) phosphorylation and ATF-3 activation. However, the transfection reagent lipofectamine signicantly increased cell death and induced ERK1/2 phosphorylation and ATF-3 activation. More importantly, the PGE 2 -induced ERK1/2 phosphorylation in F11 cells was repressed by the expression of pAAV-GlyRα3 and administration of an EP 2 inhibitor (PF-04418948), GlyRαs antagonist (strychnine), and protein kinase C inhibitor (G06983). Conclusions: PGE 2 -induced ERK1/2 phosphorylation can be modulated by GlyRα3. In addition, no changes in the levels of tumor necrosis factor (TNF)-α, interleukin (IL)-1β, or IL-6 of the F11 cells were observed 6 hours after lipopolysaccharide (LPS) or complete Freund's adjuvant (CFA) treatment. These data suggest that delivering pAAV-GlyRα3 The present the Though PGE 2 -induced ERK1/2 phosphorylation in F11 neuron cells. Antagonists of prostaglandin EP 2 receptor, PKC, and glycine receptor can inhibit PGE 2 -induced ERK1/2 phosphorylation. This study found pAAV/pAAV-GlyRα1/3 transfection does not induce cell cytotoxicity, ERK1/2 phosphorylation, or ATF-3 activation in neuron cells. Furthermore, pre-transfected pAAV-GlyRα3 can signicantly suppress PGE 2 induced ERK1/2 phosphorylation. We suggest that PGE 2 -induced ERK1/2 phosphorylation can be modulated by GlyRα3. pAAV/pAAV-GlyRα1/3 recombinant vectors could be a good model for studying glycine-receptor function in neuron cells and are probably suitable for animal infection models.

mechanism of GlyRs remain poorly understood. In our rst experiment, we explored downstream proteins and signals mediated by GlyRs.
Adeno-associated virus (AAV) is a member of the parvovirus family that can be used to infect humans in clinical trials, and in experimental animal models. [19] The use of AAV vectors for gene therapy in human clinical trials has shown promise, as AAV generally causes a very mild immune response and long-term gene transfer, and there have been no reports of disease. For tissue tropism, the existence of a variety of serotypes makes AAV gene therapy more attractive, since they differ in infectivity rates and tissue speci city. Previous studies demonstrated extensive and effective transduction of the brain and spinal cord after AAV8 injection. [20,21] The F11 neuron cell line possesses many properties seen in nociceptive DRG neuronal cells. The transient transfection e ciency is about 50% for F11 neuron cells. [22,23] Compared with other cell lines, such as HEK293, F11 neuron cells are excellent proxies for the responses of real neuron cells. Therefore, we used AAV8 for GlyRα1/3 gene transfer in F11 neuron cells, and investigated the effects and roles of pAAV-GlyRα1/3 on cell cytotoxicity and the PGE 2 -mediated in ammatory response.

Materials & Methods
Cell culture & bacteria F11 cells was purchased from European Collection of Authenticated Cell Cultures (ECACC, 08062601) and was cultured in DMEM supplemented with 10% FBS (both Invitrogen) in a humidi ed atmosphere at 37 ℃ with 5% CO2. Escherichia coli DH5α strain was cultured at 37°C in Luria-Bertani (LB) broth medium supplied with kanamycin (50 ug/ml) and was used as a host in transformation. The institutional review board of Kaohsiung Medical University, Kaohsiung, Taiwan approved gene recombinant experiment in this study (KMU-106076).
The correct recombinant clones which containing pAAV-GlyRα1, or pAAV-GlyRα3 were store at -80℃ for further use.

Cell Viability assay
The effect of pAAV, pAAV-GlyRα1 or pAAV-GlyRα3 as well as PGE2 on F11 cells was determined by MTT assay. F11 cells (7 x10 4 cells/well) were seeded in the 24 well plate, after 24 hours culture, 2 ug pAAV, pAAV-GlyRα1, pAAV-GlyRα3 or lipofectamine only was used to transfect F11 cells for 48 hours. In addition, F11 cells treated with Prostaglandin E2 (PGE2, 100 uM) for 60 mins was included in the experiment. In the end, cell viability was assessed by the MTT assay kit (abcam) according to the manufacturer's protocol. Brie y, remove culture medium and MTT reagent was added into each well, incubated for 2~6 hours at 37ºC, remove MTT reagent then DMSO was added and incubated for 5 minutes. Supernatant was collected and absorbance was measured at OD 550~600 nm. F11 cells without transfection and lipofectamine treatment was used as control.

Western blots
To survey the effect of PGE2, or recombinant pAAV-GlyRα1 or pAAV-GlyRα3 on inducing pERK phosphorylation and activating transcription factor 3 (ATF-3) activation, 2 ug pAAV, or pAAV-GlyRα1 or pAAV-GlyRα3 was selected to transfect F11 cells (3 x10 5 cells/well), which seeded into the 6 well plate, for 48 hours. Serum free medium was replaced for another 24 hours, then F11 cells was treated with PGE2 (100 uM) for 5, 15, 30 and 60 mins. F11 cells treated PGE2 alone or F11 cells transfected with pAAV, or pAAV-GlyRα1 or pAAV-GlyRα3 alone was also included in the present study. To investigate the pathway of PGE2 induced pERK phosphorylation, a glycine receptor antagonist Strychnine, EP2 receptor antagonist PF-04418948 was added for 24 hours. One hours after PGE2 administration cell pellets were harvested and pERK phosphorylation was measured. In western blotting, seeded cells as described above were harvested and were homogenized in RIPA lysis buffer (50 mM Tris pH 7.4, 150 mM NaCl, 1 mM EDTA, 0.1% SDS, 1% NP-40, and 0.5% sodium deoxycholate) containing protease inhibitor cocktail (Roche, Germany). The protein concentration was determined using a Bio-Rad Protein Assay Kit (Bio-Rad, Hercules, CA, USA). 20 ug of total protein was loaded into 8% (w/v) sodium dodecyl sulfatepolyacrylamide gels and transferred to polyvinylidene uoride membranes (Millipore, Bedford, MA, USA).

Measurement of cytokines by ELISA
To investigate the possible role of glycine receptors on modulating lipopolysaccharides (LPS) or complete Freund's adjuvant (CFA) induced cell in ammatory reaction. According to the transfection e ciency assay described above, 2 ug pAAV, pAAV-GlyRα1 or pAAV-GlyRα3 was selected to transfect F11 cells (3 x10 5 cells/well), which seeded into the 6 well plate, for 48 hours. Serum free medium was replaced for another 24 hours, then was treated with LPS (100 ng), or CFA (100 ng) for another 6 hours.

Statistical analysis
Results are presented as mean ± SE. Analytical statistics were performed using the SPSS (version 20) software package. Statistical signi cance was calculated by nonparametric Mann-Whitney U test and for pair-wise comparisons only. In some cases, ANOVA followed by Scheffe multiple post hoc test were used. Differences were considered statistically signi cant at *p < 0.05, **p<0.01, ***p<0.001. pAAV/pAAV-GlyRα1/3 transfection does not induce cytotoxicity The time schedule for measuring the transfection e ciency of pAAV-GlyR α1/pAAV-GlyR α3 is shown in Fig. 2A. Increased transfection e ciency was found after long-duration pAAV-GlyRα1 (Fig. 2B) or pAAV-GlyRα3 transfection (Fig. 2C). Therefore, transfection with 2 ug pAAV, pAAV-GlyRα1, or pAAV-GlyRα3 for 48 hours was used in our study, including for the MTT assay. Figure 3A shows the time schedule for measuring cell viability in the MTT assay. The viability of F11 cells transfected using pAAV, pAAV-GlyRα1, pAAV-GlyRα3, or lipofectamine alone, as well as those treated with PGE 2 , was evaluated. Compared with the control, cell viability was signi cantly decreased in cells transfected with pAAV, pAAV-GlyRα1, or pAAV-GlyRα3, as well as those treated only with lipofectamine. However, cell viability was lower (but not signi cantly) in the pAAV and pAAV-GlyRα1/3 groups compared with the lipofectamine-treated group (Fig. 3B). This indicates that decreased cell viability in pAAV-or pAAV-GlyRα1/3-transfected groups was caused by the lipofectamine transfection reagent. However, cytotoxicity caused by pAAV/pAAV-GlyRα1/3 transfection cannot be ruled out completely. Furthermore, cell toxicity caused by PGE 2 treatment was not found in this study (Fig. 3B).

Blocking The Glycine Receptor Inhibits Pge-induced Perk1/2 Phosphorylation
Strychnine is an inhibitor of postsynaptic GlyRs. We examined whether the EP 2 or glycine receptor was responsible for PGE 2 -induced ERK1/2 phosphorylation in F11 cells. Figure 7A shows the time schedule for application of the EP 2 -and glycine-receptor antagonists, PF-04418948 and strychnine, respectively.
The transfection reagent Lipofectamine 2000 can induce cell damage. [24] We used lipofectamine in our transfection experiments and found signi cantly decreased cell viability following pAAV or pAAV-GlyRα1/3 transfection, caused by the reagent. This indicates that pAAV/pAAV-GlyRα1/3 is safe for neuron cell transfection in vitro and should be suitable for use in animal transfection. PGE 2 has a short lifespan (usually less than 2.5 hours). [25] Similar to previous studies, exogenous PGE 2 was shown to directly induce ERK1/2 (but not p38) phosphorylation in DRG neurons [18] and other nonneuron cell types [18] , [26] Zhao et al. [27] found a PGE 2 -dependent, ERK1/2-regulated microglia-neuron signaling pathway that mediated the microglial component of pain maintenance.
We found that the ERK phosphorylation in pAAV-GlyRα3 transfection was caused by lipofectamine. Furthermore, our in vitro results indicated that exogenous pAAV-GlyRα3 administration can suppress PGE 2 -induced ERK1/2 phosphorylation in F11 neuron cells. Other studies have shown that speci c GlyR subtypes play a key role in different diseases, for example, the GlyRα1 subtype is associated with tumorigenesis and alcoholism, [28,29] and the GlyRα3 subtype is associated with in ammatory hyperalgesia. [1,9,12] PKC-dependent phosphorylation of p38 and ERK have been reported. [30] The phosphorylation of ERK occurs in a PKA-independent manner. [31] Conversely, Chen et al. showed that PKA stimulated p38 and ERK phosphorylation in breast adipose broblasts. [26] Furthermore, a previous study found that pERK1/2 increased signi cantly in DRG neurons 8 hours after PGE 2 exposure; co-treatments of PGE 2 with inhibitors of pan-PKA, pan-PKC, and ERK/mitogen-activated protein kinase (MAPK) signi cantly suppressed PGE 2 -induced IL-6 expression. [18] Our study showed that pAAV-GlyRα3 transfection or administration of glycine-receptor antagonist (strychnine) can downregulate PGE 2 -induced ERK1/2 phosphorylation in F11 neuron cells. There are two possible explanations for this. First, ERK1/2 phosphorylation is also controlled by GlyRs through the PKA/PKC-dependent pathway. [9,26,32] Second, the glycine-receptor structure changes during strychnine binding, as does the internal domain binding site of the PGE 2 -dependent PKA/PKC pathway. [33,34] A previous study showed that GlyRa3 architecture changed after strychnine binding. For an agonist or antagonist to bind with and affect the state of the channel, the signal must be transduced across the extracellular domain and transmembrane domain interface. [34] Further studies are needed to elucidate the GlyR signaling pathways and identify additional potential molecular targets for in ammatory pain inhibition.
The pro-in ammatory cytokines TNF-α [35] and IL-6 [18,36] may be upregulated in DRG neurons after peripheral nerve injury. A previous study used a partial sciatic-nerve ligated model to test IL-6 in DRG neurons, and found upregulation of IL-6 in DRG neurons following nerve injury. [18] Furthermore, increased IL-6 expression shifted from small-to medium-and large-sized damaged DRG neurons. Nerve injury models have also induced neuronal cell death, thereby inducing more pro-in ammation cytokines, such as IL-6. [18,27] Both LPS [37][38][39] and CFA [40,41] are strong in ammatory mediators and can induce PGE 2 synthesis in animal models. Our results showed that the expression of TNF-α, IL-1ß, and IL-6 did not change in F11 cells treated by LPS/CFA or transfected with pAAV, pAAV-GlyRα1, or pAAV-GlyRα3. Hashemian et al. also used F11 neuron cells and found that LPS induced modest increases in IL-6 and COX2 expression, but did not induce signi cant TNF-α expression. [42] A previous ex vivo study treated cultured sensory ganglion explants with a stabilized, long-acting PGE 2 analog (dmPGE 2 ) and showed that after high-dose dmPGE 2 (100 µM) treatment, IL-6 expression increased signi cantly at 20 and 24 hours [18]. However, sensory ganglion explants were used, which differ from the F11 neuron cell line chosen for the present study. Except for neuron cells, sensory ganglion explants include satellite glial cells, which are surrounded and can modulate neuron cell function.
Satellite glial cells are also important immune regulators and can produce in ammatory mediators, such as prostaglandins, IL-6, and TNF-α. In addition, in the nervous system, it has been suggested that cytokines are secreted by peripheral immune cells, microglia, astrocytes, and neurons.

Declarations
Ethics approval and consent to participate: The institutional review board of Kaohsiung Medical University, Kaohsiung, Taiwan approved gene recombinant experiment in this study (KMU-106076).
Consent for publication: Not applicable.
Availability of data and material: The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.