Cynandione A alleviates neuropathic pain through spinal microglial interleukin-10/β-endorphin expression following α7 nicotinic acetylcholine receptor activation

Cynandione A, an acetophenone isolated from Cynanchum Wilfordii Radix, exhibits antihypersensitivity effects in neuropathic pain. This study sought to explore the target molecule and mechanisms underlying cynandione A mechanical antiallodynia, particularly related to the spinal glial expression of IL-10/β-endorphin, cAMP/PKA/p38/CREB signaling and α7 nicotinic acetylcholine receptor (α7 nAChR) activation. attenuated cynandione A-induced mechanical antiallodynia and spinal or microglial expression of IL-10 and β-endorphin. Cynandione A stimulated spinal phosphorylation of the transcription factor STAT3, which was inhibited by methyllycaconitine, H-89 and the IL-10 antibody. The STAT3 inhibitor NSC74859 weakened cynandione A-induced mechanical antiallodynia and spinal expression of β-endorphin. the IL-10/STAT3 signaling, following α7 nAChR activation.


Introduction
As a negative regulator, interleukin (IL)-10 is primarily produced by Th2 cells, activated B cells, monocytes, macrophages and glial cells [1], and regulates pleiotropic effects in in ammation and immunoregulation [2,3]. IL-10 exhibits remarkable neuroprotective and antinociceptive effects in the central nervous system [4]. It has been demonstrated that IL-10 inhibited proin ammatory cytokine production, attenuated thermal hyperalgesia induced by chronic sciatic nerve constriction [5] and enhanced morphine analgesia [6]. Local injection of IL-10 eliminated mechanical hyperalgesia caused by carrageenan in hindpaw of rats [7] and its trigeminal ganglia injection decreased trigeminal neuropathic pain induced by infraorbital nerve constriction in rats [8]. Intrathecal injection of IL-10 produced mechanical antiallodynia and thermal antihyperalgesia through spinal microglial expression of β-endorphin [9]. Furthermore, the spinal glial IL-10 and β-endorphin pathway has been revealed to be associated with the antinociceptive effects of electroacupuncture and the agonists of the glucagon-like peptide-1 (GLP-1) receptor and G proteincoupled receptor 40 (GPR40) in rodent models of neuropathic pain induced by spinal nerve ligation [10][11][12].
Nicotinic acetylcholine receptors (nAChRs), highly expressed in skeletal muscle and the nervous system, are known as pentameric ligand-gated ion channels with a total of 17 subunits (α1-10, β1-4, γ, δ, and ε) identi ed to date [13]. They are formed from various combinations of the subunits and the homomeric α7 nAChRs is one of the major subtypes found in the mammalian central nervous system [14]. α7 nAChRs are expressed not only in neurons but also in non-neuronal cells such as astrocytes, microglia, oligodendrocyte precursor cells and brain endothelial cells [15][16][17]. They have attracted more attention recently because their mechanisms of action are involved in neuroin ammation [18], neurodegenerative diseases such as Alzheimer's disease [19] and neuroprotection [20,21]. Down-regulation of spinal α7 nAChR expression is observed in spared nerve injury-and chronic sciatic nerve constriction-induced neuropathic pain [22]. Furthermore, activation of α7 nAChRs attenuates in ammatory pain, postoperative pain, neuropathic pain and bone cancer pain [23][24][25][26][27][28]. It was suggested that activation of α7 nAChRs blocked neuropathic pain through promoting expression of the antiin ammatory cytokine IL-10 and inhibiting the expression of proin ammatory cytokines such as IL-1β, IL-6, and TNF-α [29,30]. It was also recently revealed that activation of α7 nAChRs produced antinociception in rat models of neuropathic pain and bone cancer pain via spinal microglial pathway of IL-10 and β-endorphin [31,32].
Cynanchum Wilfordii has long been used in the East Asia countries especially in China, Korea and Japan as a traditional herb medicine for the treatment of insomnia, anxiety, anemia, senescence and various geriatric diseases [33][34][35]. The acetophenone cynandione A, one of the most active ingredients isolated from Cynanchum Wilfordii, exhibits antiin ammatory activity by protecting mice against endotoxin shock [36]. Cynandione A also attenuated glutamate-induced cytotoxicity and mitigated ischemic injuries in rats with cerebral ischemia [37] and protected cultured cortical neurons from toxicity induced by hydrogen peroxide, L-glutamate and kainate [38]. In addition, intrathecal injection of cynandione A markedly stimulated spinal β-endorphin expression and alleviated mechanical allodynia and thermal hyperalgesia in neuropathic rats. Pretreatment with the microglial activation inhibitor minocycline, the β-endorphin antibody and µ-opioid receptor antagonist totally attenuated cynandione A-induced mechanical antiallodynia. These results suggest that cynandione A alleviates neuropathic pain through upregulation of spinal microglial expression of β-endorphin [39]. However, little is known about the target molecule and upstream mechanisms underlying cynandione A-induced β-endorphin expression and subsequent antinociception, although its antinociception in neuropathic rats was blocked by the α7 nAChR antagonist methyllycaconitine in our preliminary experiment.
In this study, we explored the target molecule and upstream mechanisms underlying cynandione Ainduced spinal microglial IL-10 and β-endorphin expression and mechanical antiallodynia in neuropathic rats and cultured primary microglial cells. Our results demonstrated that cynandione A produces mechanical antiallodynia in neuropathic pain through spinal microglial IL-10 expression via the cAMP/PKA/p38/CREB signaling and subsequent β-endorphin expression via the IL-10/STAT3 signaling, following α7 nAChR activation.

Chemicals and reagents
Cynandione A was extracted, isolated and puri ed by Institute of Botany, Jiangsu Province and Chinese Academy of Sciences, with ≥ 98% purity determined by 1 H-NMR and HPLC. The rabbit β-endorphin antiserum and recombinant rat IL-10 antibody were purchased from Abcam (Cambridge, UK) and R&D systems (Minneapolis, USA), respectively. The adenylate cyclase inhibitor 2,5-dideoxyadenosine (DDA) and the PKA activation inhibitor H-89 were obtained from Santa Cruz Biotechnologies (Santa Cruz, Canada), while the p38 activation inhibitor SB203580, CREB activation inhibitor KG501 and STAT3 activation inhibitor NSC74859 were purchased from Selleck Chemicals (Houston, TX, USA), Sigma-Aldrich (St. Louis, MO, USA) and Medchem Express (Boston, USA), respectively. The α7 nAChR antagonist methyllycaconitine citrate was purchased from APEx BIO (Houston, USA). Cynandione A was dissolved in 10% dimethyl sulfoxide (DMSO) and 20% polyethylene glycol (PEG400) in 0.9% normal saline for intrathecal injection and dissolved in 0.1% DMSO for cell culture. All other drugs or reagents were dissolved or diluted in normal saline.

Animals
Male adult (160-180 g body weight) and 1-day-old (sex unidenti ed) neonatal Wistar rats were obtained from the Shanghai Experimental Animal Institute for Biological Sciences (Shanghai, China). The adult animals were housed (3-4 per cage) in the Shanghai Jiao Tong University Experimental Animal Center (Shanghai, China) in room temperature (22 ± 2℃) under light conditions of a 12/12 h reversed light-dark cycle (7:00 a.m.-7:00 p.m.). They received food and water ad libitum, accustomed to the laboratory environment for 3-4 days before surgery. The research protocols were approved by the Animal Care and Welfare Committee of Shanghai Jiao Tong University and carried out in accordance with the animal care guidelines of the US National Institutes of Health.

Primary cultures of microglia
Primary microglial cells were isolated from the spinal cords of 1-day-old neonatal rats as previous [11].
The isolated spinal cords were minced and incubated in 0.05% trypsin in the incubator for 7 minutes. The digestion was terminated using the DMEM supplemented with 10% (v/v) fetal bovine serum (FBS), penicillin (100 U/mL) and streptomycin (100 mg/mL). The glia cells were then plated into 75-cm 2 tissue culture asks (1x10 7 cells/ ask) that were precoated with poly-L-lysine (100 mg/mL) and cultured at 37°C in a 5% carbon dioxide incubator. After 8 days of culture, microglial cells were prepared as oating cell suspensions by shaking the asks at 260 rpm for 2 hours. Unattached microglial cells were removed by washing with serum-free DMEM and centrifuged with 300 g for 15 minutes, and transferred into new 12-or 24-well plates for further study.

Western blot
Protein supernatants, mechanically homogenized from the spinal lumbar enlargements (L3-L5) and cultured microglia, were lysed in the RIPA lysis buffer which contained 1% protease inhibitor PMSF and 1% phosphorylase inhibitors cocktail A/B (Biotool, Houston, USA). The protein solution was denatured at 100℃ after adding 5 x of the SDS protein loading buffer and separated in 10% of the sodium dodecyl sulfate--polyacrylamide gel electrophoresis (SDS-PAGE) with the PAGE gel electrophoresis and then transferred to a polyvinylidene uoride membrane using the electrophoretic method. The membrane was then blocked in 5% skim milk powder dissolved in 1 x TBS containing 0.1% Tween 20 (TBS-T). After blocking for 1 hour, the membrane was incubated with the primary antibody against p-STAT3 (phosphor-Stat3, Ser727) (1:1000; Cell Signaling Technology, Danvers, MA, USA), p-PKA (1:2000; Abcam, Cambridge, UK), p-p38 (1:1000; Cell Signaling Technology), p-CREB (1:1000; Cell Signaling Technology), and GAPDH (1:5000; Protein Tech Group, Chicago, USA) overnight at 4℃ with gently shaking. After collecting the primary antibody and washing the membrane four times with TBS-T (10 minutes once) for the second day, the membrane was incubated with the corresponding second antibody diluted in 5% skim milk powder dissolved in 1x TBS-T (1:10000), including the goat anti-rabbit IgG (IRDye 800-conjugated, Cell Signaling Technology) and goat anti-rat IgG (IRDye 700-conjugated, Cell Signaling Technology) for 1 hour at 37℃ with gently shaking. The second antibody was collected and the membrane was washed with 5% skim milk powder dissolved in 1x TBS-T for 4 times (10 minutes once). The Odyssey Infrared Imaging system (Li-Cor Biosciences, Lincoln, NE, USA) was used to detect the protein bands, which was analyzed and quanti ed by using the Image J program (National Institutes of Health, Bethesda, MD, USA). The relative expression of each target protein was obtained after normalization to the GAPDH level as previously reported. The experiments were repeated at least three times.

Immuno uorescence staining
Double immuno uorescence labeling of IL-10 and β-endorphin with cellular biomarkers of microglia, astrocytes and neurons was performed in the spinal cord and visualized under a TCS SP8 confocal microscope (Leica Microsystems, Wetzlar, Germany) as described previously [44]. Pentobarbitalanesthetized rats (40 mg/kg) were subjected to the intracardial perfusion with 100 mL of normal saline to ush blood of the whole body, followed by 60 mL of 4% paraformaldehyde (w/v) to x cells in the spinal cord. The rats were sacri ced and the spinal lumbar enlargements (L3-L5) were isolated and xed in 4% buffered paraformaldehyde for 18 hours and dehydrated in gradient sucrose solutions (10-30%) at 4°C. Tissues were entrapped and frozen in the OCT-freeze tissue medium (Leica Microsystems) and cut into 30-μm sections. The frozen sections were blocked by 10% goat serum (v/v) and 0.5% X-100 (v/v) in phosphate-buffered saline (PBS) at room temperature for 1 hour after washing in PBS and baking in a 37°C-dryer machine for 30 minutes to prevent tissue detachment. The sections were then incubated with the IL-10 antibody (1:100; goat polyclonal; R&D Systems) or β-endorphin antiserum (1:100; rabbit polyclonal; Phoenix Pharmaceuticals) with the primary antibodies against cellular biomarkers at 4°C for 24 hours. The cellular biomarker antibodies included Iba-1 for microglia (1:100; mouse monoclonal; Millipore, USA), GFAP for astrocytes (1:100; mouse polyclonal; Millipore), and NeuN for neurons (1:60; mouse polyclonal; Millipore). The IL-10 or β-endorphin staining was visualized with the Alexa Fluor-555conjugated donkey anti-goat secondary antibody (1:200; Invitrogen, California, USA) or Alexa Fluor-555-conjugated goat anti-rabbit secondary antibody (1:200; Invitrogen), while the Alexa Fluor-488-conjugated goat anti-mouse secondary antibody (1:200; Invitrogen) was used to detect the cell biomarkers.
For quanti cation of the intensity of IL-10-, β-endorphin-, Iba-1-, GFAP-and NeuN-positive cells, photomicrographs of the medial three-fourths of the dorsal horn (laminas I-III) and cultured microglial cells (see below) were taken under a confocal microscope with ×10 or ×30 magni cation. An investigator blinded to the experimental groups measured the positively-stained surface area using the computerassisted image analysis program, Image J software (National Institutes of Health, USA). The background uorescence was excluded and only immuno uorescence intensity measurements from positive-staining areas were included by low-and high-threshold setup. For the colocalization analysis, the colocalization nder of the Image J software was used to generate merged images in which colocalized pixels appeared as white.
For the staining and quanti cation, the cultured primary microglial cells were seeded on the poly L-lysinecoated round coverslips which were placed in the 24-well plates (1×10 4 /well) and cultured overnight. The cells were then washed with PBS 2 hours after cynandione A treatment and xed in 4% paraformaldehyde for at least 1 hour, then the coverslips moved in the 12-well plates were incubated in 10% goat serum (v/v) and 0.5% X-100 (v/v) in PBS for blocking in room temperature for 1 hour. The rest procedures followed the above protocol of the frozen spinal sections except that cultured microglial cells were also stained with the nucleic dye reagent 2-(4-amidinophenyl)-6-indolecarbamidine dihydrochloride (DAPI, 0.1 μg/mL; Beyotime Biotechnology, Shanghai, China) for 3 minutes.

IL-10/β-endorphin measurements
The contralateral and ipsilateral spinal lumbar enlargements (L3-L5) were isolated from neuropathic rats 1 hour after intrathecal drug injection and homogenized at 4000 rpm for 15 seconds with a homogenizer (Fluko Equipment, Germany) in 10 mM Tris-HCl (5 mL/1 g of tissue) and centrifuged at 4000 rpm at 4°C for 15 minutes. In addition, cultured primary microglial cells were placed in 24-well plates (1x10 5 cells/well) and washed once with 1 mL/well of warm PBS, twice with 1 mL/well of warm DMEM containing 2 mg/mL BSA and 15 mmol/L N-(2-hydroxyethyl) piperazine-N-2-ethanesulfonic acid and then incubated with 100 μM cynandione A for 2 hours according to the previous study [39]. The cell culture supernatant was collected and further centrifuged at 5000 rpm for 10 minutes at 4°C, and then aspirated to a new tube. The total protein concentrations in the spinal cord homogenates were measured using the standard bicinchoninic acid protein assay (Beyotime Biotechnology, Shanghai, China). The levels of IL-10 (eBioscience, California, USA) and β-endorphin (Phoenix Pharmaceuticals, USA) were measured using the commercial uorescent immunoassay kits. A microplate reader (Multiskan MK3; Thermo Labsystems, Vantaa, Finland) and a uorescence microplate reader (Thermo Labsystems, Grand Rapids, Wood Ohio, USA) were used to measure the relative uorescence values and the concentrations of IL-10/β-endorphin were calculated by a calibration curve performed at the same time. The assays were validated with the linear range of 1-500 and 1-100 pg/mL for the IL-10 and β-endorphin, respectively.

Rat models of neuropathic pain and intrathecal catheterization
Intrathecal catheterization was performed in rats according to the previously described protocol [45]. Brie y, a 20-cm catheter (PE-10: 0.28 mm inner diameter and 0.61 mm outer diameter, AniLab Software & Instruments Co., Ningbo, China) was inserted into the lumbar level of the spinal cord under inhaled iso urane anesthesia. The other end of the PE-10 catheter was inserted subcutaneously to the neck and xed. The spinal nerve ligation procedure was performed at the same time just after intrathecal catheterization. The left L5 and L6 spinal nerves were carefully isolated and tightly ligated with 6-0 silk sutures. The lumbar fascia and skin were sewed by a 4-0 resorbable polyglactin suture after nerve ligation. The rat returned to its single home cage after surgeries for recovery. Only rats with no major motor impairments and signi cant unilateral allodynia to mechanical stimulation (hindpaw withdrawal thresholds in the operated side < 8 g), and with both hindpaws of immediately limp and feeble after intrathecal injection of 10 μL of 4% lidocaine followed by 15 μL saline ush were chosen for subsequent experiments. The total of 334 neuropathic rats underwent different drug tests during 1-2 weeks after spinal nerve ligation.

Mechanical allodynia assessment
The hindpaw withdrawal threshold to mechanical stimuli was measured using a 2290CE electrical von Frey hair (IITC Life Science Inc., CA, USA) according to the previous study [46]. Neuropathic rats were acclimatized to the Plexiglas box on a metal grid for at least 30 minutes. An examiner blinded to the treatment groups performed the behavior testing using a 2290 CE electrical von Frey hair (IITC Life Science, Woodland Hill, CA, USA). The withdrawal thresholds were evoked in both contralateral and ipsilateral hindpaws using the von Frey hair while the rat stood on a metal grid. The increments of force were applied to stimulate the footpad until the rat suddenly withdrew its hindpaw. The lowest force evoking a withdrawal response was considered the threshold, which was averaged from triplicate measurements at a 1-min interval.

Data statistical analysis
Data were exhibited as means ± SEM. Two-tailed and unpaired Student t-test, and one-way or repeated measures two-way ANOVA were applied to generate statistical signi cance values. The post-hoc Student-Newman-Keuls test was used when the effect of the drug (dose) (for one-way ANOVA, the factor was drug [dose]; for two-way ANOVA, the factors were drug [dose], time, and their interaction) was statistically signi cant. The statistical analysis was performed using GraphPad Prism (Version 7.0, GraphPad Software, San Diego, CA, USA). Probability values were considered statistically signi cant at 5% level.

Results
Cynandione A produced mechanical antiallodynia in neuropathic pain and speci cally stimulated spinal microglial expression of IL-10 and β-endorphin We have previously demonstrated that intrathecal injection of cynandione A dose-dependently attenuated mechanical allodynia and thermal hyperalgesia in neuropathic pain, with E max values of 57% and 59% maximum possible effect (MPE) and ED 50 values of 14.9 and 6.5 µg, respectively [39]. In this stduy, we rst con rmed its spinal mechanical antiallodynic effect at 100 µg, an approximately ED 90 of cynandione A. Two groups of spinal nerve ligated neuropathic rats (n = 6 per group) received single intrathecal injection of 10 µL of the vehicle (10% DMSO and 20% PEG400 in saline) or 100 µg of cynandione A. The withdrawal thresholds in both contralateral and ipsilateral hindpaws of the vehicle-treated control rats were unchanged during the 4 hours of observation. Intrathecal injection of cynandione A did not signi cantly alter withdrawal thresholds in the contralateral hindpaws, but time-dependently inhibited mechanical allodynia in the ipsilateral hindpaws with the peak effect at 1 hour by 48% MPE and duration of approximately 4 hours after injection (P < 0.05, by repeated measures two-way ANOVA followed by the post-hoc Student-Newman-Keuls test; Fig. 1A).
Additional two groups neuropathic rats (n = 6 per group) that received the same intrathecal treatments as above were sacri ced 1 hour after injection (peak time of the antiallodynic effect). The spinal cords were collected and homogenized to detect the gene and protein expression of IL-10, β-endorphin and dynorphin A by using qRT-PCR and immunoassay kits, respectively. As shown in Fig. 1B-1D, intrathecal injection of cynandione A (100 µg) speci cally stimulated spinal mRNA expression of IL-10 and POMC (P < 0.05, by unpaired and two-tailed Student t-test) but not PDYN. Moreover, intrathecal cynandione A also signi cantly stimulated spinal protein expression of IL-10 and β-endorphin (P < 0.05, by unpaired and twotailed Student t-test; Fig. 1E, 1F).
To further con rm the speci c spinal microglial expression of IL-10 and β-endorphin, their double immuno uorescence labeling was performed with cellular biomarkers of microglia (Iba-1), astrocytes (GFAP) or neurons. Two groups of neuropathic rats (n = 5 per group) received intrathecal injection of 10 µL of the vehicle or 100 µg of cynandione A. The rats were sacri ced 1 hour after injection and the spinal cords were collected for immunostaining. There was no signi cant difference of the double IL-10/Iba-1 immunostaining between contralateral and ipsilateral spinal dorsal horns observed under × 10 or × 30 magni cations. Cynandione A treatment signi cantly enhanced the double IL-10/Iba-1 immunostaining in both contralateral and ipsilateral spinal dorsal horns compared to the vehicle control ( Fig. 2A-2F). In contrast, intrathecal cynandione A injection did not enhance the double IL-10/GFAP immunostaining (2G-2L) or IL-10/NeuN immunostaining (2M-2R). Quantitatively using confocal microscope with × 30 magni cation, cynandione A in the contralateral and ipsilateral dorsal horn I-III laminate signi cantly increased the double immuno uorescence intensity of IL-10/Iba-1 by 8.8-fold and 9.5-fold, respectively (p < 0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test; Fig. 2S), but not the double immuno uorescence intensity of IL-10/GFAP (Fig. 2T) or IL-10/NeuN (Fig. 2U).
Furthermore, the speci c stimulatory effect of intrathecal cynandione A on β-endorphin expression was also demonstrated in microglia but not astrocytes or neurons in the spinal dorsal horn (Fig. 3A-3R). Quantitative measurement indicated that cynandione A increased the double immuno uorescence intensity of β-endorphin/Iba-1 in the contralateral and ipsilateral dorsal horn I-III laminate by 9.3-fold and 10.2-fold, respectively (p < 0.05. by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test; Fig. 3S), but not the double immuno uorescence intensity of β-endorphin/GFAP (Fig. 3T) or βendorphin/NeuN (Fig. 3U).
Cynandione A stimulated IL-10 and β-endorphin expression in primary spinal microglia We have previously demonstrated that treatment with cynandione A (3, 10, 30, 100 and 300 µM) concentration-dependently stimulated POMC and β-endorphin expression in cultured primary microglia (but not astrocytes or neurons), with EC 50 values of 38.8 and 20.0 µM, respectively [39]. In this study we rst test its stimulatory effects on IL-10 and β-endorphin at 100 µM, an approximately EC 80 of cynandione A, in cultured primary microglial cells originated from neonatal rats. As previously reported [9], the cultured cells were collected two hours later and digested to detect the gene expression of IL-10, βendorphin and dynorphin A by using qRT-PCR, while the cell culture medium was collected to detect the protein expression of IL-10 and β-endorphin by using commercial uorescent immunoassays. As shown, treatment with cynandione A in microglia signi cantly upregulated the mRNA expression of IL-10 and POMC (P < 0.05, by two-tailed and unpaired Student t-test; Fig. 4A, 4B), but not PDYN (Fig. 4C). In addition, cynandione A treatment upregulated the IL-10 and β-endorphin levels in the cell culture medium (P < 0.05, by two-tailed and unpaired Student t-test; Fig. 4D, 4E).
The stimulatory effect of cynandione A on IL-10 and β-endorphin was further assessed in cultured primary microglia by using single and double immuno uorescence labeling of IL-10 or β-endorphin with Iba-1 and the nuclear staining reagent DAPI. Compared with the vehicle control, treatment with cynandione A (100 µM) signi cantly enhanced the IL-10 expression re ected in single or double immunostaining under × 30 magni cations (Fig. 5A-5H). Quantitatively, cynandione A treatment signi cantly increased the double immuno uorescence intensity of IL-10/Iba-1 by 8.99-fold (p < 0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test; Fig. 5I).
Cynandione A produced mechanical antiallodynia in neuropathic pain through spinal microglial IL-10 expression and subsequent β-endorphin expression We further explored the causal relationship between spinal microglial expression of IL-10/β-endorphin and mechanical antiallodynia in neuropathic pain. Four groups of neuropathic rats (n = 6 per group) received intrathecal injection of saline (10 µL), the IL-10 neutralizing antibody (2 µg) or β-endorphin antiserum (1:10) followed by intrathecal injection of the vehicle (10 µL) or cynandione A (100 µg) 30 minutes later. The withdrawal thresholds in the contralateral and ipsilateral hindpaws were measured 1 hour after injection. Intrathecal injection of cynandione A inhibited mechanical allodynia in the ipsilateral hindpaws, which was nearly completely blocked by the pretreatment with intrathecal injection of the IL-10 antibody or β-endorphin antiserum (P < 0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test; Fig. 6A). Pretreatment with intrathecal injection of the IL-10 antibody or β-endorphin antiserum did not signi cantly affect the baseline mechanical thresholds in the ipsilateral hindpaws as previously reported [10,39].
The above four groups of neuropathic rats were sacri ced immediately after the completion of the behavior testing and the spinal cords were collected and homogenized to detect the gene and protein expression of IL-10 and β-endorphin by using qRT-PCR and uorescent immunoassays, respectively. As shown in Fig. 6B, intrathecal cynandione A speci cally stimulated spinal mRNA expression of IL-10, which was not signi cantly reduced by intrathecal injection of the IL-10 neutralizing antibody or βendorphin antiserum. On the other hand, cynandione A also stimulated spinal mRNA expression of POMC, which was completely attenuated by the pretreatment with intrathecal injection of the IL-10 antibody (P < 0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test) but not the β-endorphin antiserum (Fig. 6C). In addition, as shown in Fig. 6D and 6E, intrathecal injection of cynandione A stimulated IL-10 and β-endorphin expression; pretreatment with the IL-10 antibody neutralized IL-10 secreted and inhibited the expression of β-endorphin (P < 0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test). However, pretreatment with intrathecal injection of the β-endorphin antiserum neutralized β-endorphin (but not IL-10) secreted (P < 0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test).
Furthermore, microglial cells from neonatal rats were treated with the IL-10 antibody (4 µg/mL) or βendorphin antiserum (1:300) for 0.5 hours before cynandione A (100 µM) treatment over 2 hours. As shown in Fig. 6F and 6G, treatment with cynandione A stimulated the mRNA expression of IL-10 and POMC in cultured primary microglial cells. Pretreatment with the IL-10 antibody did not signi cantly alter cynandione A-stimulated IL-10 mRNA expression but completely reduced its stimulation on POMC expression (P < 0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test). On the other hand, pretreatment with the β-endorphin did not signi cantly alter the mRNA expression of IL-10 or POMC.
Cynandione A produced mechanical antiallodynia in neuropathic pain through activation of spinal α7 nAChRs In order to illustrate whether cynandione A-induced mechanical antiallodynia was through activation of α7 nAChRs, four groups of neuropathic rats (n = 6 per group) received intrathecal saline (10 µL) or the speci c α7 nAChR antagonist methyllycaconitine (10 µg) [47] 30 minutes later followed by intrathecal the vehicle (10 µL) or cynandione A (100 µg). As shown in Fig. 7A, intrathecal injection of cynandione A in the ipsilateral hindpaws produced time-dependent mechanical antiallodynia, which was nearly completely inhibited by intrathecal injection of methyllycaconitine (p < 0.05, by repeated measures two-way ANOVA followed by the post-hoc Student-Newman-Keuls test), although it did not signi cantly alter baseline mechanical thresholds in both contralateral and ipsilateral hindpaws.
Another four groups of neuropathic rats (n = 6 per group) that received the same intrathecal treatments as above were sacri ced to obtain spinal cords 1 hour after the last injection to detect the gene and protein expression of IL-10 and β-endorphin. As exhibited in Fig. 7B and 7C, intrathecal injection of cynandione A remarkably increased the mRNA expression of IL-10 and POMC in both contralateral and ipsilateral spinal cords, whereas intrathecal methyllycaconitine was not effective in reducing baseline IL-10 and POMC mRNA expression. However, pretreatment with intrathecal methyllycaconitine entirely blocked cynandione A-stimulated mRNA expression of IL-10 or POMC (p < 0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test). In addition, intrathecal injection of cynandione A in both contralateral and ipsilateral cords also stimulated the expression of IL-10 and β-endorphin, which was entirely attenuated by the pretreatment with intrathecal methyllycaconitine (p < 0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test; Fig. 7D, 7E).
Cultured primary microglial cells were pretreated with the vehicle or methyllycaconitine (100 nM [48]) 30 minutes later followed by cynandione A (100 µM) over 2 hours. The cultured cells and culture medium were collected to measure the gene and protein expression of IL-10 and β-endorphin. As shown in Fig. 7F and 7G, treatment with cynandione A in cultured primary microglial cells stimulated the mRNA expression of IL-10 and POMC, which was totally blocked by the pretreatment with methyllycaconitine (p < 0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test), although it did not signi cantly alter baseline expression of IL-10 or POMC. In addition, cynandione A treatment also stimulated the expression of IL-10 and β-endorphin, which was completely inhibited by pretreatment with methyllycaconitine (p < 0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test; Fig. 7H, 7I).
Cynandione A stimulated spinal microglial expression of IL-10 and β-endorphin through the cAMP/PKA/p38/CREB signaling To explore whether the cAMP/PKA/p38/CREB signaling was responsible for cynandione A-induced mechanical antiallodynia in neuropathic pain and spinal expression of IL-10/β-endorphin, two groups of neuropathic rats (n = 6 per group) received intrathecal injection of saline (10 µL) or the speci c adenylyl cyclase inhibitor DDA (20 µg, dosage of which was based on the previous study [49]) 30 minutes later followed by intrathecal injection of cynandione A (100 µg). Intrathecal injection of cynandione A in the ipsilateral hindpaws produced time-dependent mechanical antiallodynia, which was entirely blocked by the pretreatment with intrathecal DDA (p < 0.05, by repeated measures two-way ANOVA followed by the post-hoc Student-Newman-Keuls test), although it did not signi cantly alter the baseline mechanical thresholds (Fig. 8A). Additional three groups of neuropathic rats (n = 6 per group) that received the same intrathecal treatments as above except for addition of one control group were sacri ced 1 hour after the last injection. The spinal cord was collected and homogenized to detect gene expression of IL-10 and βendorphin by using qRT-PCR. As shown in Fig. 8B and 8C, intrathecal injection of cynandione A stimulated spinal mRNA expression of IL-10 and POMC, which was completely reduced by the pretreatment with intrathecal DDA (p < 0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test). In addition, cultured primary microglial cells were pretreated with saline or DDA (100 µM [49]) 30 minutes before cynandione A treatment (100 µM) for 2 hours. Treatment with cynandione A in microglial cells stimulated the mRNA expression of IL-10 and POMC, which was inhibited by the pretreatment with DDA (p < 0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test; Fig. 8D, 8E).
In addition, four groups of neuropathic rats (n = 6 per group) received intrathecal injection of saline (10 µL) or the speci c PKA activation inhibitor H-89 (5 µg, dosage of which was based on the previous study [50]) 30 minutes before intrathecal injection of the vehicle (10 µL) or cynandione A (100 µg). As shown in Furthermore, three groups of neuropathic rats (n = 6 per group) received intrathecal injection of saline (10 µL) or the speci c p38 mitogen-activated protein kinase (MAPK) activation inhibitor SB203580 (10 µg, dosage of which was based on the previous study [51]) 30 minutes later followed by intrathecal injection of the vehicle (10 µL) or cynandione A (100 µg). Mechanical thresholds in both contralateral and ipsilateral hindpaws were measured 1 hour after the last injection. As displayed in Fig. 10A, intrathecal injection of cynandione A produced time-dependent mechanical antiallodynia, which was completely alleviated by the pretreatment with intrathecal SB203580 (p < 0.05, by repeated measures two-way ANOVA followed by the post-hoc Student-Newman-Keuls test; Fig. 10A). The rats were sacri ced immediately after the completion of the behavioral test and spinal cords were collected and homogenized. As exhibited, intrathecal injection of cynandione A stimulated the spinal mRNA expression of IL-10 ( Fig. 10B) and POMC (Fig. 10C) as well as protein expression of IL-10 (Fig. 10D), which was entirely inhibited by the pretreatment with intrathecal SB203580 (p < 0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test). In addition, treatment with cynandione A (100 µM) in cultured primary microglial cells for 2 hours stimulated the mRNA expression of IL-10 and POMC, which was completely reversed by the pretreatment (30 minutes prior to) with SB203580 (50 µM [51]) (p < 0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test; Fig. 10E, 10F).
Lastly, two groups of neuropathic rats (n = 6 per group) received intrathecal injection of saline (10 µL) or the speci c CREB activation inhibitor KG501 (10 µg [51]) 30 minutes later followed by intrathecal injection of cynandione A (100 µg). As presented in Fig. 11A, intrathecal injection of cynandione A produced time-dependent mechanical antiallodynia, which was completely blocked by the pretreatment with intrathecal KG501 (p < 0.05, by repeated measures two-way ANOVA followed by the post-hoc Student-Newman-Keuls test). Additional three groups of neuropathic rats (n = 6 per group) received intrathecal injection of saline (10 µL) or KG501 30 minutes later followed by intrathecal injection of cynandione A (100 µg) and the spinal cords were collected 1 hour after the last intrathecal injection. As shown, cynandione A-stimulated spinal gene expression of IL-10 and POMC was totally attenuated by the pretreatment with KG501 (p < 0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test; Fig. 11B, 11C). Moreover, pretreatment (30 minutes earlier) with KG501 (25 µM [51]) in cultured primary microglial cells completely alleviated 100 µM cynandione A-promoted but not baseline gene expression of IL-10 and POMC (p < 0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test; Fig. 11D, 11E).
Cynandione A stimulated spinal microglial expression of β-endorphin through the IL-10/STAT3 signaling In order to illustrate the role of IL-10/STAT3 signaling in cynandione A-induced spinal microglial βendorphin expression and mechanical antiallodynia, the spinal STAT3 phosphorylation was rst measured. Four groups of neuropathic rats (n = 6 per group) received intrathecal injection of saline (10 µL) or the PKA activation inhibitor H-89 (5 µg) 30 minutes later followed by the vehicle (10 µL) or cynandione A (100 µg). The rats were sacri ced 1 hour after the last injection and the spinal cords were obtained for the detection of the STAT3 phosphorylation using western blot. Intrathecal injection of cynandione A stimulated spinal STAT3 phosphorylation. Pretreatment with intrathecal injection of H-89 did not affect baseline phosphorylation of STAT3, but abolished cynandione A-stimulated STAT3 activation (p < 0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test; Fig. 12A, 12B).
In addition, additional four groups of neuropathic rats (n = 6 per group) that received intrathecal injection of saline (10 µL) or the IL-10 antibody (2 µg) 30 minutes later followed by intrathecal injection of the vehicle (10 µL) or cynandione A (100 µg). The rats were sacri ced 1 hour after the last injection and the spinal cords were obtained to detect the STAT3 phosphorylation. As shown in Fig. 12C and 12D, pretreatment with the IL-10 antibody totally blocked cynandione A-stimulated but not baseline phosphorylation of STAT3 (p < 0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test).
Furthermore, three groups of neuropathic rats (n = 6 per group) received intrathecal injection of saline (10 µL) or the speci c STAT3 activation inhibitor NSC74859 (10 µg, dosage of which was based on the previous study [9]) 30 minutes before intrathecal injection of cynandione A (100 µg). Cynandione A intrathecal injection produced time-dependent mechanical antiallodynia, which was completely inhibited by the pretreatment with intrathecal injection of NSC74859 (p < 0.05, by repeated measures two-way ANOVA followed by the post-hoc Student-Newman-Keuls test; Fig. 12E). Additional three groups of neuropathic rats (n = 6 per group) received the same intrathecal treatments as above. The rats were sacri ced 1 hour after the last injection and the spinal cords were obtained. As exhibited in Fig. 12F, pretreatment with intrathecal NSC74859 totally blocked cynandione A-stimulated spinal POMC expression (p < 0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test). In addition, pretreatment (30 minutes earlier) with NSC74859 (10 µM [9]) in cultured primary microglial cells inhibited 100 µM cynandione A-stimulated but not baseline expression of POMC (p < 0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test, Fig. 12G).
Cynandione A stimulated phosphorylation of PKA, p38, CREB and STAT3 following spinal α7 nAChR agonism In order to illustrate whether cynandione A activated PKA/p38/CREB/STAT3 signaling following α7 nAChR agonism, four groups of neuropathic rats (n = 6 per group) received saline (10 µL) or methyllycaconitine (10 µg) 30 minutes later followed by intrathecal injection of the vehicle (10 µL) or cynandione A (100 µg). The rats were sacri ced 1 hour after the last intrathecal injection and the spinal cords were obtained. Phosphorylation of PKA, p38, CREB and STAT3 was detected in the contralateral and ipsilateral spinal cords using western blot. As shown in Fig. 13A-13D, intrathecal injection of cynandione A stimulated phosphorylation of PKA, p38, CREB and STAT3 in the contralateral and ipsilateral spinal cords. Pretreatment with intrathecal injection of methyllycaconitine entirely blocked cynandione A-promoted but not baseline phosphorylation of PKA, p38, CREB, and STAT3 (p < 0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test).

Discussion
Intrathecal injection of cynandione A has been demonstrated to produce mechanical antiallodynia and thermal antihyperalgesia in neuropathic pain through spinal expression of β-endorphin [39]. Our current study con rms it and further extends that cynandione A induces expression of β-endorphin subsequently through spinal microglial expression of IL-10 and not in reverse. The notion is supported by the following facts: 1) Intrathecal injection of cynandione A stimulated spinal expression of IL-10 and β-endorphin but not dynorphin A, measured by using qRT-PCR and uorescent immunoassay kits. Furthermore, intrathecal cynandione A speci cally stimulated IL-10 and β-endorphin expression in microglia, but not in astrocytes or neurons, in both contralateral and ipsilateral spinal cords of neuropathic rats, directly identi ed by double immuno uorescence staining. 2) Pretreatment with intrathecal injection of the IL-10 antibody in neuropathic rats eliminated cynandione A-induced spinal mechanical antiallodynia and gene expression of POMC but not IL-10. In addition, intrathecal β-endorphin antiserum also attenuated spinal cynandione A-induced mechanical antiallodynia but not gene expression of POMC or IL-10. On the other hand, intrathecal IL-10 antibody neutralized IL-10 secreted and inhibited the expression of β-endorphin, whereas intrathecal β-endorphin antiserum neutralized β-endorphin (but not IL-10) secreted. 3) Cynandione A treatment stimulated the expression of IL-10 and β-endorphin but not dynorphin A in primary cultures of microglia. 4) Pretreatment with the IL-10 antibody in cultured microglial cells completely inhibited cynandione A-stimulated gene expression of POMC but not IL-10, whereas the β-endorphin antiserum failed to affect cynandione A-stimulated gene expression of POMC or IL-10.
α7 nAChRs have been recognized to mediate long-term modi cation of cell functions, in addition to the typical ligand-gated ion channels that evoke cation-selective currents across the plasma membrane.
Sustained stimulation of α7 nAChRs that mostly expressed in the central nervous system induces delayed cellular responses leading to neuroprotection and antinocieption via intracellular signal pathways probably triggered by Ca 2+ in ux, in addition to acute responses [52]. The α7 nAChR activationinduced antinoccieption has been demonstrated in animal models of pain hypersenstivity such as in ammatory pain, neuropathic pain and bone cancer pain [26][27][28]. It has been extensively reported that 7 nAChRs are colocalized on microglial cells as determined by qRT-PCR, western blot, immuno uorescent, and immunohistochemistry analyses and their activation is associated with antinociception and neuroprotection [29,[53][54][55]. More speci cally, it was recently identi ed that the spinal microglial IL-10/β-endorphin pathway mediated the α7 nAChR agonists PHA-543613-, cinobufaginand lemairamin-induced antinociception in bone cancer pain and neuropathic pain [31,32]. Our current study demonstrated that methyllycaconitine signi cantly blocked cynandione A-induced mechanical antiallodynia in neuroapthic rats, and the expression of IL-10 and β-endorphin in the spinal cords and cultured primary microglial cells. These results together reveal that cynandione A produces antinociception through α7 nAChR activation. The results further highlight the broad signi cance of the newly discovered spinal glial IL-10/β-endorphin pathway in pain modulation, which is associated with electroacupuncture-, GLP-1 receptor-, GPR40-and α7 nAChR-induced antinociception [10][11][12]. On the other hand, activation of α7 nAChR induces marked antineuroin amation by inhibiting the expression of proin ammatory cytokines such as IL-1β, IL-6 and TNF-α [53,54]. Cynandione A has been demonstrated to reduce LPS-induced expression of TNF-α, IL-6, IL-1β, nitric oxide and prostaglandin E2 in BV-2 microglia, RAW264.7 macrophages and primary microglia, as well as in septic mice and neuropathic rats [36,39,56]. However, we previously hypothesized that the antineuroin ammatory property of cynandione A unlikely contributed to its antinociception, as the two effects were separately mediated by MAPKdependent and independent mechanisms [39]. In addition, the antineuroin ammatory effects of IL-10 and minocycline were separated from their antinociception in neuropathic pain [9,44].
α7 nAChRs generate speci c and complex Ca 2+ -dependent signals that include adenylyl cyclase, PKA, protein kinase C, Ca 2+ -calmodulin-dependent kinase [57,58], and phosphatidylinositol 3-kinase [52,58], which trigger cell depolarization and turn on various functional switches [59]. It was reported that α7 nAChR and G-protein interaction regulated cAMP levels under LPS treatment in microglia [18]. In addition, nicotine stimulation led to PKA activation and further Raf-1/MEK/ERK1/2 and JAK2/STAT3 signaling through α7 nAChRs in human oral keratonocytes [30]. Speci cally, the serine 365 in the M3-M4 cytoplasmic loop of the α7 nAChR has been demonstrated to be a phosphorylation site of PKA [60], though PKA is customarily activated by G-protein Gαs [61]. In our current study, cynandion A stimulated PKA phosphorelation in a α7 nAChR-dependent manner, and the adenylate cyclase inhibitor DDA and PKA activation inhibitor H-89 blocked cynandione A-induced spinal microglial expression of IL-10 and subsequent β-endorphin expreession. The results suggest that cynandione A stimulates microglial expression of IL-10 following α7 nAChR agonism through the cAMP/PKA signinaling.
As a transcription factor, CREB has been shown to be directly phosphorylated at Ser-133 by activated PKA in a classic way or indirectly through several other kinases particularly MAPKs including p38 [62]. It has been demonstrated that both classic Gs-cAMP/PKA/CREB and alternative Gs-cAMP/PKA/p38/CREB mediate GLP-1 receptor agonism-induced expression of M2 microglial biomarkers Arg 1, CD206, IL-4 and IL-10 [49]. α7 nAChR agonists biochemically characterized the pharmacological induction of CREB phosphorylation which was blocked by the pretreatment with methyllycaconitine [63]. α7 nAChR/Gprotein interaction was reported to attenuate p38 phosphorylation in microglia in a PTX-sensitive manner [18]. Activation of peripheral α7 nAChRs was also postulated to inhibit the CaMKIIα/CREB signalling and block the expression of TNF-α and CGRP, thus contributing to antinociception [64,65]. It has been donmenstrated that α7 nAChR activation increased CREB phosphorylation in brain and improved cognitive function [66], but it is not known whether activation of α7 nAChRs stimultes CREB phosphorylation and relieves pain. Our results reveal that cynandione A-stimulated microglial IL-10/βendorphin expression is subsequently through CREB phosophrelation via p38 (or p38β) activation alternatively. The notion is supported by the following ndings in our current and previous studies [39]: 1) intrathecal cynandione A stimulated α7 nAChR-dependent spinal p38 and CREB activation in neuropathic rats; 2) cynandione A upregulated phosphorylation of MAPKs including p38, ERK1/2 and JNK in cultured primary microglial cells; 3) cynandione A-stimulated IL-10/β-endorphin expression in microglial cells was completely inhibited by the CREB activation inhibitor KG501 and p38 activation inhibitor SB203580 (but not by the ERK1/2 or JNK activation inhibitors). In addition, knockdown of spinal p38β but not p38α using siRNAs also completely blocked cynandione A-induced β-endorphin expression; 4) cynandione Ainduced mechnical antiallodynia was totally attneuated by SB203580 and KG501.
IL-10 has been considered one of the most important antiin ammatory cytokines so far, by binding to the IL-10 receptor-α followed by activating IL-10 receptor-β localized on the cell membrane [67,68] and inhibiting the release of proin ammatory mediators, including TNF-α, IL-1β, IL-6, IL-8, G-CSF, and GM-CSF from monocytes, macrophages and microglia [9,69,70]. Additionally, exogenous IL-10 exhibites antinociception in pain hypersensitivity induced by peripheral nerve injury [10,11,70], paclitaxel treatment [71] and severe traumatic brain injury [72]. It has been demonstated the IL-10/JAK/STAT3 pathway in microglia mediates IL-10-stimulated β-endorphin expression, which is in parallel to its inhibition of expression of neuroin ammatory cytokines through antiin ammatory elements [9,10]. We thus postulate that cynandione A stimulates β-endorphin expression through autocrime IL-10 expression via the STAT3 signaling. Activation of α7 nAChRs indeed led to reducing in ammatory drive through a JAK2-STAT3 pathway that couples with CREB/Irs2/Akt survival signaling in the mouse islets [73]. α7 nAChR activation increased hypothalamic POMC expression by triggering JAK2/STAT3 pathway [74]. Consistently, our current study demonstrated that cynandione A stimulated microglial STAT3 activation which could be blocked by the α7 nAChR antagonist methyllycaconitine, PKA activation inhibitor H-89 and IL-10 neutralizing antibody. Moreover, the STAT3 activation inhibitor NSC74859 also completely attenuated cyanandione A-induced expression of β-endorphin in the spinal cords of neuropathic rats and in cultured primary microglial cells. Declarations QQH and YXW conceived and designed the experiments; QQH, MY, ZYW, HL and JPA performed the experiments; QQH, MY and YXW analyzed the data; YXW and QQH wrote the paper. All authors read and approved the nal manuscript.

Ethics approval
All animal experimental protocols and procedures in this study were reviewed and approved by the Animal Care and Welfare Committee of Shanghai Jiao Tong University and carried out in accordance with the animal care guidelines of the National Institutes of Health. The consent to participate is not applicable.

Consent for publication
Not applicable

Competing interests
The authors declare no competing nancial interests in this work.
Author details 1     collected from 1-day-old neonatal rats, were incubated with cynandione A for 2 hours and they and the culture medium were collected later. The gene and protein expression of IL-10, β-endorphin and dynorphin A was determined using qRT-PCR and enzyme-linked immunosorbent uorescent assays, respectively.
The data are presented as means ± SEM (n=3 per group with two independent repeats). * p<0.05 compared with the vehicle group, by two-tailed and unpaired Student t-test.

Figure 5
Stimulatory effects of cynandione A on the expression of IL-10 (A-I) and β-endorphin (J-R) in cultured primary microglia originated from 1-day-old neonatal rats. Cultured primary microglia were collected 2 hours after cynandione A incubation and immuno uorescence was triply stained with IL-10 (red)/Iba-1 (green)/DAPI (blue) or β-endorphin (red)/Iba-1 (green)/DAPI (blue). The immunolabeled surface areas of IL-10/Iba-1 (I) and β-endorphin/Iba-1 (R) from microglial cells were quanti ed by using the Image J program. Data are presented as means ± SEM (n=3 per group with two independent repeats). * P<0.05 compared with the control group, analyzed by two-tailed and unpaired Student t-test.   shown as means ± SEM (n=6 per group in neuropathic rats or n=3 per group with two independent repeats in cultured primary cells). *, # p<0.05 compared with the control and cynandione A treatment groups, respectively, analyzed by one-way or two-way ANOVA followed by the post-hoc Student-Newman-Keuls test. hour after the last intrathecal injection for the detection of the IL-10 and POMC gene expression using qRT-PCR. Data are shown as means ± SEM (n=6 per group in neuropathic rats or n=3 per group with two independent repeats in cultured primary cells). *, # p<0.05 compared with the control and cynandione A treatment groups, respectively, analyzed by one-way or measures-repeated two-way ANOVA followed by the post-hoc Student-Newman-Keuls test. spinal nerve ligation and cultured primary microglial cells originated from 1-day-old neonatal rats (E, F). Neuropathic rats received intrathecal the vehicle or SB203580 30 minutes followed by intrathecal cynandione A injection. The spinal cords were obtained 1 hour after the last intrathecal injection for the detection of the IL-10 and β-endorphin expression using qRT-PCR and enzyme-linked immunosorbent assays, respectively. Data are shown as means ± SEM (n=6 per group in neuropathic rats or n=3 per group with two independent repeats in cultured primary cells). *, # p<0.05 compared with the control and cynandione A treatment groups, respectively, analyzed by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test.   were obtained 1 hour after the last intrathecal injection for the phosphorylation detection by using western blot. Data are shown as means ± SEM (n=6 per group). *, # p<0.05 compared with the control and cynandione A treatment groups, respectively, analyzed by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test.