Spinal sigma-1 receptor activation increases the production of d -serine in astrocytes which contributes to the development of mechanical allodynia in a mouse model of neuropathic pain

Spinal sigma-1 receptor activation increases the production of d -serine in astrocytes which contributes to the development of mechanical allodynia in a mouse model of neuropathic pain

Accepted Manuscript Title: Spinal sigma-1 receptor activation increases the production of D-serine in astrocytes which contributes to the development ...

10MB Sizes 0 Downloads 37 Views

Accepted Manuscript Title: Spinal sigma-1 receptor activation increases the production of D-serine in astrocytes which contributes to the development of mechanical allodynia in a mouse model of neuropathic pain Author: Ji-Young Moon Sheu-Ran Choi Dae-Hyun Roh Seo-Yeon Yoon Soon-Gu Kwon Hoon-Seong Choi Suk-Yun Kang Ho-Jae Han Hyun-Woo Kim Alvin J. Beitz Seog-Bae Oh Jang-Hern Lee PII: DOI: Reference:

S1043-6618(15)00187-5 http://dx.doi.org/doi:10.1016/j.phrs.2015.08.019 YPHRS 2916

To appear in:

Pharmacological Research

Received date: Revised date: Accepted date:

3-3-2015 5-8-2015 20-8-2015

Please cite this article as: Moon Ji-Young, Choi Sheu-Ran, Roh Dae-Hyun, Yoon SeoYeon, Kwon Soon-Gu, Choi Hoon-Seong, Kang Suk-Yun, Han Ho-Jae, Kim Hyun-Woo, Beitz Alvin J, Oh Seog-Bae, Lee Jang-Hern.Spinal sigma-1 receptor activation increases the production of D-serine in astrocytes which contributes to the development of mechanical allodynia in a mouse model of neuropathic pain.Pharmacological Research http://dx.doi.org/10.1016/j.phrs.2015.08.019 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Title page Spinal sigma-1 receptor activation increases the production of D-serinein astrocytes which contributes to the development of mechanical allodyniain a mouse model of neuropathic pain Running title: The modulation of Sig-1Rs on D-serine release Ji-Young Moon1a,Sheu-Ran Choi2a, Dae-Hyun Roh3, Seo-Yeon Yoon4, Soon-Gu Kwon2, Hoon-Seong Choi2, Suk-Yun Kang1, Ho-Jae Han2, Hyun-Woo Kim5, Alvin J. Beitz6, SeogBae Oh4, Jang-Hern Lee2* 1

KM FundamentalResearch Division, Korea Institute of Oriental Medicine, Daejeon 305-811,

Republic of Korea 2

Department of Veterinary Physiology, BK21 PLUS Program for Creative Veterinary Science

Research, Research Institute for Veterinary Science and College of Veterinary Medicine, Seoul National University, Seoul 151-742, Republic of Korea 3

Department of Maxillofacial Tissue Regeneration, Kyung Hee University School of Dentistry, Seoul

130-701, Republic of Korea 4

Pain Cognitive Function Research Center,Department of Brain and Cognitive Sciences College of

Natural Sciences, Dental Research Institute and Department of Neurobiology and Physiology, School of Dentistry, Seoul National University, Seoul 110-749, Republic of Korea 5

Department of Physiology, Institute of Brain Research, Chungnam National University Medical

School, Daejeon 301-747, Republic of Korea 6

Department of Veterinary and Biomedical Sciences, College of Veterinary Medicine, University of

Minnesota, St Paul, MN 55108, USA a

Ji-YoungMoon and Sheu-RanChoi contributed equally to this study. (co-first authors)

* Corresponding author: Jang-Hern Lee Postal address: Room #913, Bldg #85, College of Veterinary Medicine, Seoul National University, Daehak-dong, Gwanak-gu, Seoul 151-742, Republic of Korea Tel: +82-2-880-1272,Fax: +82-2-885-2732 E-mail: [email protected]

1

Graphical abstract

2

3

Abstract We have previously demonstrated that activation of the spinal sigma-1 receptor (Sig-1R) plays an important role in the development of mechanical allodynia (MA) via secondary activation of the N-methyl-D-aspartate (NMDA) receptor. Sig-1Rs have been shown to localize to astrocytes, and blockade of Sig-1Rs inhibits the pathologic activation of astrocytes in neuropathic mice. However, the mechanism by which Sig-1R activation in astrocytes modulates NMDA receptors in neurons is currently unknown. D-serine, synthesized from Lserine by serine racemase (Srr) in astrocytes, is an endogenous co-agonist for the NMDA receptor glycine site and can control NMDA receptor activity. Here, we investigated the role of D-serine in the development of MA induced by spinal Sig-1R activation in chronic constriction injury (CCI) mice. The production of D-serine and Srr expression were both significantly increased in the spinal cord dorsal horn post-CCI surgery. Srr and D-serine wereonlylocalized to astrocytesin the superficial dorsal horn,while D-serine was also localized to neurons in the deep dorsal horn. Moreover, we found that Srr exists in astrocytes that express Sig-1Rs. The CCI-induced increase in the levels of D-serine and Srr was attenuated by sustained intrathecal treatment with the Sig-1R antagonist, BD-1047 during the induction phase of neuropathic pain. In behavioral experiments, degradation of endogenous D-serine with DAAO, or selective blockade of Srr by LSOS, effectively reduced the development of MA, but not thermal hyperalgesia in CCI mice. Finally, BD-1047 administration inhibited the development of MA and this inhibition was reversed by intrathecal treatment with exogenous D-serine.These findings demonstrate for the first time that the activation of Sig-1Rs increases the expression of Srrand D-serine in astrocytes.The increased production of D-serine induced by CCI ultimately affects dorsal horn neurons that are involved in the development of MA in neuropathic mice.

4

Chemical compounds studied in this article: L-serine O-sulfate potassium salt (PubChem CID: 5951); N-[2-(3,4-dichlorophenyl)ethyl]-Nmethyl-2-(dimethylamino) ethylamine dihydro-bromide (PubChem CID: 188914); D-serine (PubChem CID: 71077)

Abbreviations: BD-1047, N-[2-(3,4-dichlorophenyl)ethyl]-N-methyl-2-(dimethylamino) ethylamine dihydrobromide; [Ca2+]i, cytoplasmic Ca2+ concentration; CCI, chronic constriction injury; CNS, central nervous system; DAAO, D-amino acid oxidase; ER, endoplasmic reticulum;GFAP, glial fibrillary acidic protein;i.t., intrathecal; LEHA, L-erythro-3-hydroxyaspartate; LSOS, Lserine O-sulfate potassium salt; MA, mechanical allodynia; MAPK, mitogen-activated protein kinase; NECK, neck region; NMDA, N-methyl-D-aspartate; nNOS, neuronal nitric oxide synthase; NO, nitric oxide; NP, nucleus proprius; GluN1, NMDA receptor GluN1 subunit; pGluN1, phosphorylated NMDA receptor GluN1 subunit; p-p38, phosphorylation of p38 MAPK; PWF, paw withdrawal frequency; SDH, superficial dorsal horn; Sig-1R, sigma non-opioid intracellular receptor 1; Srr, serine racemase; TH, thermal hyperalgesia Key words; Sig-1R, D-serine, Astrocyte, Mechanical allodynia, Neuropathic pain

5

1. Introduction Peripheral neuropathic pain is a serious health problemand affects between 6 and 8% of the general population with certain groups like diabetic patients exhibiting percentages as high as 26%. Two of the hallmarks of neuropathic pain are mechanical allodynia (MA, lowering of response threshold to non-noxious tactile stimuli) and thermal hyperalgesia (TH, an increased response to a noxious thermal stimulus). We have previously provided evidence for a novel central mechanism underlying the development of MA by demonstrating that activation of the spinal sigma non-opioid intracellular receptor 1 (Sig-1R) and N-methyl-Daspartate (NMDA) receptor are critical to the induction of MA in chronic constriction injury (CCI) rats [1]. In addition, spinal Sig-1R-mediated nociceptive action is mediated by an increase in neuronal nitric oxide synthase (nNOS) activity, which is associated with an nitric oxide (NO)-induced increase in PKC-dependent phosphorylated GluN1 (pGluN1) expression [2]. Recently, we have shown that the Sig-1Rs are located in spinal astrocytes and that intrathecal treatment with the Sig-1R antagonist, BD-1047 during the induction phase (days 0-3 after CCI surgery) reduced the expression of glial fibrillary acidic protein(GFAP), a key feature of astrocyte activation, via p38 mitogen-activated protein kinase (MAPK) phosphorylation in a mouse CCI model [3]. However, the precise mechanisms by which Sig1R-mediated astrocyte activation ultimately affects neuronal activity including NMDA receptor activation remain unclear. D-serine is an endogenous ligand for the glycine site of the NMDA receptor, which modulates NMDA receptor mediated neurotransmission [4]. Activation of NMDA receptors requires binding of glutamate at the glutamatebinding site, but also of a co-agonist glycine or D-serine at their glycine site for the efficient opening of the receptor. D-serine is synthesized by serine racemase (Srr), which converts L- to D-serine. In the brain, D-serine is predominantly synthesized in astrocytes [5], and the Ca2+-dependent release of D-serine controls NMDA receptor-dependent long-term potentization [6]. Evidence to date supports 6

the hypothesis that astroglial D-serine is involved in pain mechanisms. Intrathecal administered fluorocitrate, an astrocyte inhibitor, or D-amino-acid oxidase (DAAO), which catalyzes the oxidative deamination of D-amino acids, inhibited tetanic sciatic stimulationinduced MA in rats [7]. In addition, intrathecal injection of the Srr inhibitors, LSOSand LEHA, decreased wind-up potentiation in an arthritic pain model [8]. Despite findings reported in these studies, the involvement of D-serine in the spinal cord in the development of MA and TH in CCI mice remains unknown. Based on these findings we hypothesized that Sig-1R modulation of astrocytes plays an important role in the induction of MA through a Dserine signaling and examined this hypothesis inan animal model of neuropathic pain. The present study was designed to examine:(1) whether there are CCI-induced increases in the expression of D-serine and Srr in the lumbar spinal cord dorsal horn; (2) whether the selective inhibition of spinal cord D-serine with LSOS or DAAO suppresses the development of CCI-induced MA or TH; (3) whether the inhibition of Sig-1Rs can modulate the CCI-induced increase in D-serine or Srr expression; and (4) whether exogenous D-serine can reverse BD-1047’s blockade of MA in CCI mice.

7

2. Materials and methods 2.1. Animal preparation Male ICR mice (20-25 g) were purchased from the Laboratory Animal Center of Seoul National University (Seoul, Republic of Korea). The experimental protocols for animal usage were reviewed and approved by the SNU Animal Care and Use Committee and conform to NIH guidelines (NIH publication No. 86-23, revised 1985).The mice had free access to food and water and were maintained in temperature and light controlled rooms (232°C, 12/12h light/dark cycle with lights on at 08:00) for at least 1 week prior to beginning an experiment. A CCI of the common sciatic nerve was performedaccording to the method described by Bennett and Xie[9] with a minor modification. Briefly, mice were anesthetizedwith 3% isoflurane in a mixture of N2O/O2 gas.The left sciatic nerve was exposed and 3 loose ligatures of 6-0 silk were placed around the nerve. Sham surgery was performed by exposing the sciatic nerve in the same manner, but without ligating the nerve. The total number of mice used in this study was 156.

2.2. Drug administration The following drugs were used:D-amino acid oxidase (DAAO, an endogenous Dserine degrading enzyme; 0.001, 0.01, 0.1 U, Sigma–Aldrich, St. Louis, MO, USA); L-serine O-sulfate potassium salt (LSOS, a Srr inhibitor; 1, 3, 10 nmol, Santa Cruz Biotechnology Inc., CA,

USA);

N-[2-(3,4-dichlorophenyl)ethyl]-N-methyl-2-(dimethylamino)

ethylamine

dihydro-bromide (BD-1047, a Sig-1R antagonist; 100 nmol, Tocris Cookson Ltd, Bristol, UK); D-serine (50, 500 nmol; Sigma–Aldrich, St. Louis, MO, USA).The dose of BD-1047 used was based on that used in previous study from our laboratories showing that this dose produces maximal effects with no detectable side-effects [1]. The doses of DAAO and LSOS used in the present study were selected based on doses previously used in the literature [8,10]. The drugs were dissolved in 5 µl of saline and injectedintrathecally. 8

For intrathecal administration, we used the modified method of direct transcutaneous intrathecal injection in mice[11].Mice were briefly anesthetized with 3% isoflurane in a mixture of N2O/O2 gas to prevent any handling-induced stress and to allow more accurate injection of drugs. Intrathecal administration of drugs was made into the L5-L6 intervertebral space using a 50 µl Hamilton syringe connected to a 30-gauge needle. The flick of the tail was considered indicative of a successful intrathecal administration. Intrathecal injection of the drugs was performed twice a day on postoperative days 0-3 (induction period). Control groups received injection of only vehicle. Examination of spinal cords from these animals following euthanasia indicated no external damage to the cords due to the injection procedure.

2.3. Behavioral testing The degree of mechanical allodynia (MA) was measured using a0.16g von Frey filament (North Coast Medical, Morgan Hill, CA) applied to left hind paws as described in previous study from our laboratories[12]. The number of paw withdrawal responses to 10 von Frey applications was recorded and expressed as a percent paw withdrawal response frequency (PWF, %). To assess thermal hyperalgesia (TH), sensitization to noxious heat stimulation was examined with the plantar paw-flick latency test developed by Hargreaves et al.in rats [13] as modified for use in the mouse [14].Mice were placed into a plastic chamber on the glass floor and a radiant heat source was positioned under the floor beneath the hind paw. Paw withdrawal response latency (PWL, sec) was measured byusing a plantar analgesia meter (IITC Life Science Inc., Woodland Hills, CA).Cutoff time in theabsence of a response was set at 20 s. Behavioral tests were performed 1 day before CCI or sham surgery in all mice to obtain normal baseline values of withdrawal response to mechanical and thermal stimuli. Animals were tested again at 1, 3, 6, 9, 14 and 21 days following CCI or sham surgery.Mechanical allodynia testing was performed first and this was followed by thermal hyperalgesia testing in all animals used in this study.Animals were randomly assigned to 9

experimental and control groups and all behavioral analyses were performed blindly.

2.4. Western blotting analysis Mice were deeply anesthetized and euthanized at several time points after CCI or sham surgery. The mouse spinal cords were removed and collected to examine possible increases in Srr expression. The ipsilateral dorsal horn from each CCI or sham mousewas subsequently processed for western blot analysis according to the method detailed in previous reports from our laboratories [2,12].The membranes were blocked with 5% skim milk for 1 hr at room temperature (RT) and incubated at 4°C overnightwith a primary antibody specific for Srr (1:1000,rabbit anti-serine racemase antibody,sc-48741, Santa Cruz) or -actin (1:5000,mouse anti--actin antibodyloading control, sc-47778, Santa Cruz).The membranes were washed and primary antibodies were detected using goat anti-rabbit or anti-mouse IgG conjugated to horseradish peroxidase.The bands were visualized with enhanced chemiluminescence (Amersham Biosciences; Buckinghamshire, UK). The positive pixel area of specific bands was measured with a computer-assisted image analysis system and normalized against the corresponding β-actin loading control bands. The mean value of the ratioin sham injuryanimals was set at 100%. Thus, the percent change relative to the sham surgery condition was then calculated for each time-point in each group.

2.5. Immunohistochemistry To confirm the specificityof the D-serine antibody immunoreactivity, we performed a pre-absorption test in which the antibody wasmixed withD-serine-conjugated to BSA (0.1, 1 orwith 10mg of peptide/ml of diluted primary antibody, MBS358268, Mybiosource) overnight at 4°C prior to staining. Moreover, the specificity of the anti-Srr antibody used in the present study was previously confirmedin the brain of Srr knockout mice in which there was an absence of immunostaining in brain tissue from Srr knockout mice[15]. 10

Mice were perfusedtranscardially with fixative containing 4% paraformaldehyde in 0.1M phosphate buffer (pH 6.9) at several time points post-CCIandthe spinal cord was removed and sectionedbased on the methods detailed in previous reports from our laboratories [2,12]. Transverse spinal cord sections were incubated in blocking solution for 1h at RT and then incubated for 48h at 4°C with one of the following primary antibodies: preabsorbed anti-D-serine antibody,rabbit anti-D-serine antibody (1:500, ab6472, Abcam)or rabbit anti-Srrantibody (1:500, sc-48741, Santa Cruz). Following incubation, tissue sections were washed and incubated for 2h at RT in secondary antibodies. Alexafluor 488 or 555 antirabbit IgG (1:200, Invitrogen) antibody was used as the secondary antibody. Double-immunofluorescence labeling was used to identify which cell types in the spinal cord dorsal horncontainedD-serine and Srr in CCI and control animals. For double immunofluorescence staining, floating sections were first incubated for 48h at 4°C with rabbit anti-D-serine or rabbit anti-Srr antibody. After washing with TPBS, the sections were then incubated for 2h at RT with Alexafluor 488-conjugated anti-rabbit IgG antibody (1:200). After washing, slices were incubated for 48h at 4°Cwith GFAP (1:1000, mouse anti-GFAP antibody, Chemicon; to identify astrocytes), neuronal-specific nuclear protein (NeuN) (1:1000, mouse anti-NeuN antibody, Millipore; to identify neurons) or Iba-1 (1:500, goat anti-Iba-1 antibody, Abcam; to identify microglia) followed by Alexafluor 555 anti-mouse or anti-goatIgG secondary antibody (1:200) for 2h at RT. For double immunofluorescence staining of Srr and Sig-1Rs, primary antibody specific for Sig-1R (1:1000, rabbit anti-OPRS1 antibody, ab53852,Abcam)and Srr (1:500, mouse anti-serine racemase antibody, sc-365217, Santacruz) were used. The slides were viewed undera confocal microscope (Fluoview4.3; Olympus). To analyze images, three to five spinal cord sections from the L4-5 lumbar spinal cord segments were randomly selected from each animal, and were analyzed using a computerassisted image analysis system (Metamorph version 7.7.2; Molecular Devices Corporation, 11

PA).The average number ofD-serine-immunoreative (ir) cells from each animal was obtained and these values were averaged across each group and presented as group data. To maintain aconstantthresholdforeachimage and to compensate for subtlevariabilityof theimmunostaining, weonly counted cells that were at least 45% brighter than the average level of each image after background subtraction and shading correction. We quantified immunostaining in the following three dorsal horn regions: 1) the superficial dorsal horn (SDH, laminae I and II); 2) the nucleus proprius (NP, laminae III and IV); and 3) the neck region (NECK, laminae V and VI) as previously described [3]. All analytical and quantitative procedures described above were performed blindly without knowledge of the experimental conditions.

2.6. Statistical analysis All values are expressed as the mean  SEM. Statistical analysis was performed using Prism 5.0 (Graph Pad Software, San Diego, USA). Repeated measurestwo-way ANOVA was performed to determine overall differences in the time-course of all nociceptive behavioral tests. One-way ANOVAwas used to determine differences across all experimental groups (immunohistochemistry and Western blot assay). Post-hoc analysis was performedusing the Bonferroni’s multiple comparison test in order to determine the P-value among experimental groups. P<0.05 was considered statistically significant.

12

3. Results 3.1. CCI-induced changes in the levels of D-serine in the dorsal horn of neuropathic mice In this study, we utilized an anti-D-serine antibody to stain mouse lumbar spinal cord sections following CCI surgery. The specificity of the anti-D-serine antibody was tested using a pre-absorption test in which the antibody was incubated with a serial concentration of Dserine antigen (0.1, 1 or 10 mg/ml). As shown in fig.1A, D-serine-immunoreactivity was not detected in any of the spinal sections processed with the anti-D-serine-antibody pre-absorbed with a maximal concentration (10 mg/ml) of D-serine-conjugated to BSA. This lack of immunofluorescence demonstrates the specificity of the anti-D-serine antibody for D-serine. To investigate the potential roles of D-serine in CCI mice, we performed an immunohistochemical analysis using an anti-D-serine antibody on mouse lumbar spinal cord sections at 0, 1, 3 and 7 days after CCI surgery. The levels of D-serine in the ipsilateralsuperficial dorsal horn (SDH, laminae I and II)and nucleus proprius (NP, laminae III and IV) regions were significantly increased from days 1 to 3 post-CCI surgery as compared with that of the day 0 group (fig.1B). This increasedexpression peaked at day 3 post-surgery and was only significantly different in the SDH region at day 7 as compared with that of the day 0 group. The left hand panel in figure 1B shows representative photomicrographs of L4-5 spinal cord sections demonstrating D-serine-ir cells at 0, 1, 3 and 7 days after CCI surgery. An illustration depicting the location of the different spinal cord regions (SDH, NP and NECK) analyzed in this study is shown in the upper left panel of these photomicrographs.

3.2. Cellular distribution of D-serinein the dorsal horn of neuropathic mice Double staining was performed at day 3 post-CCI surgery to determine which cell types express D-serine in the dorsal horn of CCI mice. Anti-D-serine antibody was used in 13

combination with antibodies specific for astrocytes (GFAP), neurons (NeuN) or microglial cells (Iba-1). Double-labeling experiments with D-serine and GFAP revealed that D-serine immunostaining was found within astrocytes in the superficial dorsal horn(fig.2A). By comparison, in the deep dorsal horn, D-serine was found primarily associated with neurons rather than astrocytes (fig.2B). There was no evidence of D-serine staining in Iba-1 positive microglia (fig.2C).

3.3. CCI-induced changes in serine racemase expression and cellular distribution in the dorsal horn of neuropathic mice Because D-serine is primarily generated by conversion from L-serine via the enzyme, serine racemase (Srr), we performed a western blot analysis to determine the levels of Srr expression at 0, 1, 3 and 7 days after CCI surgery. The expression of Srr in the ipsilateral dorsal horn was significantly increased on day 1 following CCI as compared with that of the day 0 groupand slowly decreased thereafter(fig.3A). By 7 days post-CCI, Srr expression was restored to normal pre-CCI values, and no statistical significance was evident when compared with Srr expression

in the dorsal horn of the day 0 group. Next

double

immunocytochemicalstaining was performed at day 1 post-CCI surgery to determine which cell types express Srr in the ipsilateral dorsal horn in CCI mice. The anti-Srr antibody was used in combination with antibodies specific for astrocytes (GFAP), neurons (NeuN) or microglial cells (Iba-1). Double-labeling with Srr and GFAP revealed a distinct colocalization of Srr within astrocytes (fig.3B). In contrast, there was no evidence of Srr staining in neurons or Iba-1 positive microglia. Thus, the increase in D-serine immunostainingin the superficial dorsal horn of CCI mice temporally correlated with the selective increase inSrr expression in dorsal horn astrocytes.

3.4. Effects of intrathecal administration of DAAO, an endogenous D-serine degrading 14

enzyme or LSOS, a serine racemase inhibitor, on the development of CCI-induced MA and TH To investigate the contribution of D-serine to the development of pain behaviors during the induction phase of CCI-induced neuropathy, we intrathecally administered the endogenous D-serine degrading enzyme, DAAO or the Srr inhibitor, LSOS on postoperative days 0-3. The effects of DAAO and LSOS on MA and TH were examined in CCI mice from days 0 to 21 post-surgery. Repeated daily, intrathecal treatment with either DAAO (0.001, 0.01 or 0.1 U) or LSOS (1, 3 or 10 nmol) significantly reduced the CCI-induced increase in paw withdrawal response frequency (PWF, %) to innocuous mechanical stimuli (MA), as compared with vehicle-treated CCI mice (fig.4A and C). Following the termination of these repeated injections on day 3, this suppressive analgesic effect was sustained throughout the 21-day experimental period following CCI surgery. On the other hand, the CCI-induced decrease in paw withdrawal response latency (PWL, seconds) to heat stimuli (TH) was not affected by repeated intrathecal treatments with DAAO or LSOS throughout the 21-day testing period (fig.4B and D).

3.5. Effects of intrathecal BD-1047, DAAO or LSOS administration on the increase in D-serine expression in CCI mice To determine whether the CCI-induced increase in the expression of D-serine was induced by Sig-1R activation during the induction phase and whether the analgesic effects of DAAO and LSOS as illustrated in figure 4 resulted from a decrease in spinal D-serine levels, we performed immunohistochemistry analysis on day 3 post-CCI surgery. Sustained intrathecal administration of the Sig-1R antagonist, BD-1047 (l00 nmol, CCI+BD), the endogenous D-serine degrading enzyme, DAAO (0.1 U, CCI+DAAO) or the Srr inhibitor, LSOS (10 nmol, CCI+LSOS) on postoperative days 0-3 significantly attenuated the CCIinduced increase in the levels of D-serine in the SDH and NP regions of the dorsal horn as 15

compared to the vehicle treated group (fig.5A). The doses of BD-1047 used in the present study were based on those used in previous studies from our laboratories showing maximal effects with no detectable side-effects [3,12]. Representative photomicrographs of L4-5 spinal cord sections illustrating D-serine-ir cells in the sham group (SHAM), saline treated CCI group (CCI+VEH), BD-1047treated CCI group (CCI+BD),DAAO treated CCI group (CCI+DAAO) and LSOS treated CCI group (CCI+LSOS) are shown in figure 5B.

3.6. Effects of intrathecal BD-1047 administration on serine racemase expression and colocalization of serine racemase in Sig-1R-immunoreactive cells in CCI mice To determine whether the CCI-induced increase in Srr expression was regulated by Sig-1R activation during the induction phase, we performed western blot analysis on day 1 after CCI surgery. Intrathecal treatment with BD-1047 during the induction phase significantly reduced the CCI-induced increase in Srr expression (fig.6A). The relative pixel area (%) of Srr and β-actin expression was significantly reduced by BD-1047 treatment during the induction period as compared with mice in the vehicle-treated CCI group. Because we previously reported that Sig-1Rs are located in astrocytes in the spinal cord in CCI mice [3], we performed double staining on day 1 with an anti-Srr antibody in combination with antibodies for the Sig-1R. Double staining with the Sig-1R andSrr antibodies showed thatSrr expression occurs in astrocytes that co-contain Sig-1Rs(fig.6B). These results indicate that Srr and Sig-1Rs are co-expressed in astrocytes in the spinal cord superficial dorsal horn in CCI mice.

3.7. Effects of concomitant BD-1047 and D-serine treatment on the development of CCIinduced MA and TH To confirm the potential role of D-serine in the development of MA induced by Sig-1R activation in CCI mice, we intrathecally injected exogenous D-serine in combination of BD16

1047 on postoperative days 0-3. Treatment with BD-1047 reduced the CCI-induced increase in the PWF (%) to innocuous mechanical stimuli (MA), while this same treatment did not affect CCI-induced TH. These effects of BD-1047 treatment are similar to those reported previously by our laboratories [3]. Treatment with exogenous D-serine (50 or 500 nmol in association with BD-1047) restored the CCI-induced MA that was inhibited by BD-1047 administration (fig.7A). However, CCI-induced TH was not affected by concomitant D-serine and BD-1047 treatment (fig.7B).

17

4. Discussion The present study demonstrates two important novel findings. First, CCI-induced sciatic nerve injury increases both serine racemase (Srr) expressionand D-serine production in astrocytes in the lumbar spinal cord dorsal horn, which in turn contributes to the induction of the mechanical allodynia (MA) in these animals. Secondly,the CCI-induced increases in Srr expression and D-serine production are significantly reduced by i.t. administration of the Sig-1R antagonist, BD-1047 in CCI mice. In addition, blockade of Sig-1Rs attenuated the induction of MA and this attenuation is reversed by exogenous D-serine administration. Collectively, these findings suggest that spinalSig-1Rs activation increases Srr expression in astrocytes, resulting in increased production of D-serine,ultimately D-serine contributes to the development of peripheral nerve injury-induced MA in CCI mice. Although NMDA receptors are very diverse in their subunit composition, and their biophysical and pharmacological properties, their activation requires the binding of a coagonist classically thought to be glycine. However, research over the last ten years has provided evidence demonstrating that the amino acid, D-serine, is in fact the preferential coagonist for some synaptic NMDA receptors. D-serineis synthesized by the activation of theSrr enzyme and is now thought to be involved in controlling NMDA receptor activation[4,6].

The

role

of

D-serine

in

regulating

the

NMDA receptors

is

effectivelysupportedbythe close distribution of D-serine withNMDA receptors.It is well established that D-serine is concentrated in gray matter regions of the rat brain that are enriched in NMDA receptors and is localized to astrocytic foot processes and glial elements of the neuropil known to surround the dendrites and spines of neurons [16,17]. In this regard, D-serine released in astrocytescan mediate astroglial-neuronal communication via activation of the neighboring neuronal NMDA receptors. Recent studies from our laboratories have demonstrated that Sig-1R activation increases neuronal NOS activity, which is triggered by extracellular Ca2+ influx into neurons via NMDA receptor activation [2,18]. In addition, Sig18

1Rs are expressed in astrocytes of the lumbar spinal cord dorsal horn in CCI mice [3]. In the present study,Srr and D-serine were found to becolocalized in astrocytes and D-serine productionin lumbar spinal cord dorsal horn astrocytes was increased by Sig-1R activationin CCI mice.In addition, D-serine is also localized to neurons in deeper laminae of the dorsal horn as well as astrocytes in the dorsal horn. These results indicate that spinal Sig-1Rs can modulate the protein expression of Srrat the gene transcription level leading to a concomitantincrease in D-serine production in the same astrocytes in CCI mice. We speculate that the CCI-induced increased in astroglial D-serine production contributes to CCI-induced MA, however it remains to be determined if D-serine is released directly from astrocytes and plays a role as a putative astroglial-derived transmitter influencing nearby neurons in the spinal cord dorsal horn. There aretwopossibilities that Sig-1Rs activationcan induce an increase in D-serine production and release either by elevating gene transcription of Srrenzyme or by increasingthe enzymatic activity of Srr.Sasabe et al. showed that the elevation of Srr expression, initially induced by glial activators, subsequently increased D-serine levels in activated glia in amyotrophic lateral sclerosis [19]. Wu et al. also reported that amyloid βpeptide stimulation elevated D-serine release by increasing expression of Srrmediated by a JNK MAPK-dependent activation of a transcription factor activator protein-1 in cultured microglia [20,21]. We recently reported that Sig-1R can increase p38 phosphorylation and MAPK activation in astrocytes during the induction phase of neuropathic pain in CCI mice [3]. Thus,it is possible that the increased expression of Srr in CCI mice is mediated by a p38dependent activation of transcription factors induced by Sig-1Rs activation. Another possibility is that there is an increase in the activity of the Srr enzyme by Sig-1Rs activation.An increase in intracellular Ca2+ concentration in astrocytes and direct Ca2+ binding to the Srr causes activation of this enzyme, which in turn leads to an increase in Dserine levels [22]. This suggests that Ca2+ could also be an important Srr cofactor. Sig-1Rshas 19

been reported to normally reside at a mitochondrion-associated endoplasmic reticulum (ER) membrane and cause IP3-induced Ca2+ efflux from the ER when cells are stimulated by Sig1R ligands or undergo prolonged stress[23,24]. Thus, these findings suggest a potential model by which the activation of Sig-1Rs in astrocytes regulates Srr expression and/or activation via Ca2+ release leading to an increase in D-serine in these glial cells. However, the precise mechanisms will require further investigation. Peripheral nerve injury induces an intense and prolonged discharge of afferent nociceptive fibers that leads to the release of substance P and excitatory amino acids in the spinal cord dorsal horn. Substance P and glutamate bind to and activate neurokinin-1 and glutamate receptors, respectively, not only on postsynaptic neurons, but also on astrocytes [25]. Activated astrocytes can cause enhancement of the postsynaptic neuronal hyperexcitability via release of gliotransmitters including glutamate and D-serine, which are required for activation of NMDA receptors [26]. Recently, there are several reports indicating that D-serine contributes to the processing of nociceptive transmission in the spinal cord dorsal horn. Guo et al. reported that spinally administered D-serine significantly enhanced the C-fiber responses of wide dynamic range neurons in normal rats [17]. In mutant mice lacking D-amino acid oxidase (DAAO), a D-serine degrading enzyme, the second phase of the formalin-induced licking response was significantly increased and NMDA receptor-mediated excitatory postsynaptic transmission of dorsal horn neurons were potentiated [27]. In addition, i.t. administration ofSrr inhibitor, LSOSor LEHA, decreased wind-up potentiation in an arthritic pain model and these antinociceptive effects were abolished by i.t. injection of Dserine[8]. Furthermore, Dieb et al. reported that spinal astrocytes are involved in the modulation of orofacial post-traumatic neuropathic pain through the secretion of D-serine [28]. These reports are consistent with our data showing that i.t. administration of LSOS or DAAO (which reduce D-serine levels in the spinal cord), significantly suppresses the development of CCI-induced MA. Although these drugs were only administered during the 20

induction phase of neuropathic pain, the suppression of MA development is effectively sustained over the 21-day experimental period following CCI surgery. Similarly, administration of the Sig-1R antagonist, BD-1047 during the induction phase of neuropathic pain (a period when Sig-1R expression is increasing) significantly attenuates the development of MA in CCI animals [1,12,29]. Considering that both Sig-1Rs and Srr expression are only increased during the first few days following peripheral nerve injury, these proteins appear to be important for spinal pain signaling transmission particularly during the induction phase of neuropathic pain development. These results are consistent with several other reports showing that different mechanisms are associated with the early phase versus the late phase of chronic pain [30–32]. Collectively, these results demonstrate that Srr and D-serine only increase during the early stages of neuropathic pain development and we speculate that these increases in Srr and D-serine play a critical role in the initial spinal processing of nociceptive transmission leading to the development of chronic neuropathic pain. On the contrary, several studies have reported that DAAO, a D-serine degrading enzyme, may be a pro-nociceptive factor that contributes to pain sensation. Zhao et al. reported that intraperitoneal or direct i.t. injection of the DAAO inhibitor, sodium benzoate blocked the maintenance of MA in a rat model of spinal nerve ligation as well as formalininduced hyperalgesia in the formalin test [33]. In addition, systemic administration of DAAO inhibitor, which increased D-serine concentrations in brain and plasma,reduced spontaneous neuronal activity in both central and peripheral recordings and alsoattenuated pain behaviors in rat models of neuropathic and inflammatory pain[34].Furthermore, intracerebroventricular administration of D-serine has been shown to significantly and dose-dependently decrease formalin-induced pain behaviors [35]. Recently, it has been suggested that Srr and D-serine are localized to both excitatory and inhibitory neurons in the adult mouse and human forebrain [15]. Under conditions in which cultured microglial cells werestimulated by amyloid β-peptide,the levels of bothSrr and D-serine were elevatedin these cells [20].Thus, 21

D-serine can be synthesized in diverse cell types in accordance with the conditions of stimulation. Although D-serine can be metabolized by DAAO, levels of D-serine were unchanged in the forebrain region in DAAO-deficient mice, suggesting thatother mechanisms are alsoimportant for regulation of D-serine concentrations in this brain area [26,36]. In addition, it is likely that NMDA responses are not fully blocked by administration of DAAO, which suggests that a DAAO-insensitive portion of NMDA mediated responses still exists. Since NMDA receptors require simultaneous binding of both glutamate and glycine for activation it is likely that endogenous glycine might account for activation of the remaining co-agonist sites of the NMDA receptor, which may account for somewhere between 30 to 50% of NMDA activity [37,38]. Furthermore, non-NMDA receptors, as well as the glycine/NMDA receptor have been shown to mediate central synaptic transmission and spinal cord nociceptive processingindicating the complexity of these processes [30]. The cellular distribution of D-serine as well as the degree of contribution of these receptors to nociceptive processingcan changedepending on a number of factors includingthe animal model being examined,thetime course of ongoing nociception associated with the model, thenervous system location being examined and how rapidly the ubiquitin system degrades D-serine[14,29,30]. These differences may cause structurally and functionallydiverse pathophysiological conditions thatcomplicate our mechanistic understanding of the role of Dserine in processing nociceptive signaling. In the present study, spinal Srr inhibition or DAAO administration attenuated the peripheral nerve injury-induced development of MA, but not TH. Under conditions in which Sig-1Rs in astrocytes were inactivated by BD-1047,exogenous D-serine treatment restored MA.Our results are supported by the work of Miraucourt et al.,who reported that intracisternal treatment with the astrocyte inhibitor, fluorocitrate prevented the development of MA induced by removal of glycine inhibition and this effect was restored with intracisternal administration of exogenous D-serine [39].In addition, i.t. injection of a DAAO 22

inhibitor had no effect on radial heat-induced acute nociception norcarrageenan-induced TH in rats [37]. This is consistent with the results of another study showing that i.t. administration of an NMDA receptor antagonist reduces MA but not thermal allodynia in a rodent model of chronic central pain [40]. Peripheral sensitization and maladaptive central changes contribute to the generation and maintenance of MA and TH, with separate mechanisms in different subtypes of allodynia and hyperalgesia. As a result of nerve injury, maladaptive changes occur in cell structure, function, biochemical properties, and connections both at the injury site and in the CNS [41]. Following peripheral nerve injury, the nociceptive nerve fibers are excited and sensitized by inflammatory mediators resulting in an alteration in the transduction properties of these nociceptors, a phenomenon termed ‘peripheral sensitization’ [42,43]. Peripheral nerve injuryinduced TH is mediated by mechanisms associated with peripheral sensitization of the uninjured nerve fibers in whichthe key heat transducer TRPV1 is upregulated causing the nociceptive nerve fibers to become more sensitive to heat stimuli[31,41,44,45]. Peripheral nerve injury-induced activation of primary C-fiber afferents can induce an increase in the excitability of spinal cord neurons resulting in amplification of responsiveness to noxious and innocuous stimuli, a phenomenon known as ‘central sensitization’ [46]. The development of MA depends on changes in spinal processing that causes inputs from mechanoreceptive Aβfibers to be perceived as painful [47,48]. It has been reported that spinal activation of NMDA receptors plays an important role in the central sensitization induced by C-fiber conditioning inputs [30,46,49]. In this respect, Sig-1R-induced D-serine release from dorsal horn astrocytes may be closely related to the CCI-induced development of MA by contributing to the activation of spinal NMDA receptors resulting in central sensitization. Conversely the development of TH appears to be more closely associated with peripheral sensitization. In conclusion, the present study demonstrates that spinal Sig-1Rs activation increases Srr enzyme expression and D-serine production in the lumbar spinal cord dorsal horn 23

astrocytes of CCI mice. Moreover, our results show that D-serine is involved in the development of CCI-induced MA. Thus, we would speculate that the increased production of D-serine in spinal astrocytes induced by Sig-1Rs activation plays an important role in the development of chronic neuropathic pain in CCI mice.However, further studies are necessary to confirm that astrocyte released D-serine contributes directly to the development of CCIinduced MA.

24

Acknowledgements This research was supported by the National Research Foundation (NRF) Grant (2014R1A2A2A01007695) funded by the Korean Government (MSIP), Republic of Korea.

Conflicts of interest All authors declare no conflict of interest.

Author contributions JY Moon and SR Choi contributed to the writing of the article, analysed the data and carried out the experiments. DH Roh and SY Yooncontributed to the construction of the animal model and the revisedarticle. SG Kwon and HS Choi performed the behavioral experiments and histological examinations. SY Kang, HJ Han andHW Kim assisted in molecular biological techniques. AJ Beitz and SB Oh assisted in interpreting the data and writing the manuscript. JH Lee designed the experimental program and assisted in data interpretation and manuscript preparation.

25

References [1] Roh DH, Kim HW, Yoon SY, Seo HS, Kwon YB, Kim KW et al. Intrathecal injection of the sigma-1 receptor antagonist BD1047 blocks both mechanical allodynia and increases in spinal NR1 expression during the induction phase of rodent neuropathic pain. Anesthesiology. 2008;109:879-89. http://dx.doi.org/10.1097/ALN.0b013e3181895a83 [2] Roh DH, Choi SR, Yoon SY, Kang SY, Moon JY, Kwon SG et al. Spinal neuronal NOS activation mediates sigma-1 receptor-induced mechanical and thermal hypersensitivity in mice: involvement of PKC-dependent GluN1 phosphorylation. Br J Pharmacol. 2011;163:1707-20. http://dx.doi.org/10.1111/j.1476-5381.2011.01316.x [3] Moon JY, Roh DH, Yoon SY, Choi SR, Kwon SG, Choi HS et al. sigma1s activate astrocytes via p38 MAPK phosphorylation leading to the development of mechanical allodynia in a mouse model of neuropathic pain. Br J Pharmacol. 2014;171:5881-97. http://dx.doi.org/10.1111/bph.12893 [4] Mothet JP, Parent AT, Wolosker H, Brady RO, Jr., Linden DJ, Ferris CD et al. D-serine is an endogenous ligand for the glycine site of the N-methyl-D-aspartate receptor. Proc Natl Acad Sci U S A. 2000;97:4926-31. http://dx.doi.org/10.1073/pnas.97.9.4926 [5] Panatier A, Theodosis DT, Mothet JP, Touquet B, Pollegioni L, Poulain DA et al. Gliaderived D-serine controls NMDA receptor activity and synaptic memory. Cell. 2006;125:77584. http://dx.doi.org/10.1016/j.cell.2006.02.051 [6] Henneberger C, Papouin T, Oliet SH, Rusakov DA. Long-term potentiation depends on release of D-serine from astrocytes. Nature. 2010;463:232-6. http://dx.doi.org/10.1038/nature08673 [7] Ying B, Lu N, Zhang YQ, Zhao ZQ. Involvement of spinal glia in tetanically sciatic stimulation-induced bilateral mechanical allodynia in rats. Biochem Biophys Res Commun. 2006;340:1264-72. http://dx.doi.org/10.1016/j.bbrc.2005.12.139 [8] Laurido C, Hernandez A, Pelissier T, Constandil L. Antinociceptive effect of rat D-serine racemase inhibitors, L-serine-O-sulfate, and L-erythro-3-hydroxyaspartate in an arthritic pain model. TheScientificWorldJournal. 2012;2012:279147. http://dx.doi.org/10.1100/2012/279147 [9] Bennett GJ, Xie YK. A peripheral mononeuropathy in rat that produces disorders of pain sensation like those seen in man. Pain. 1988;33:87-107. http://dx.doi.org/10.1016/0304-3959(88)90209-6 [10] Ren WH, Guo JD, Cao H, Wang H, Wang PF, Sha H et al. Is endogenous D-serine in the rostral anterior cingulate cortex necessary for pain-related negative affect? J Neurochem. 2006;96:1636-47. http://dx.doi.org/10.1111/j.1471-4159.2006.03677.x 26

[11] Hylden JL, Wilcox GL. Intrathecal morphine in mice: a new technique. Eur J Pharmacol. 1980;67:313-6. http://dx.doi.org/10.1016/0014-2999(80)90515-4 [12] Moon JY, Roh DH, Yoon SY, Kang SY, Choi SR, Kwon SG et al. Sigma-1 receptormediated increase in spinal p38 MAPK phosphorylation leads to the induction of mechanical allodynia in mice and neuropathic rats. Exp Neurol. 2013;247:383-91. http://dx.doi.org/10.1016/j.expneurol.2013.01.004 [13] Hargreaves K, Dubner R, Brown F, Flores C, Joris J. A new and sensitive method for measuring thermal nociception in cutaneous hyperalgesia. Pain. 1988;32:77-88. http://dx.doi.org/10.1016/0304-3959(88)90026-7 [14] Caterina MJ, Leffler A, Malmberg AB, Martin WJ, Trafton J, Petersen-Zeitz KRet al. Impaired nociception and pain sensation in mice lacking the capsaicin receptor.Science. 2000;288:306-13. http://dx.doi.org/10.1126/science.288.5464.306 [15] Balu DT, Takagi S, Puhl MD, Benneyworth MA, Coyle JT. D-serine and serine racemase are localized to neurons in the adult mouse and human forebrain.Cell Mol Neurobiol. 2014;34:419-35. http://dx.doi.org/10.1007/s10571-014-0027-z [16] Schell MJ, Brady RO Jr, Molliver ME, Snyder SH. D-serine as a neuromodulator: regional and developmental localizations in rat brain glia resemble NMDA receptors.J Neurosci. 1997;17:1604-15. http://www.jneurosci.org/content/17/5/1604.long [17] Wolosker H, Blackshaw S, Snyder SH. Serine racemase: a glial enzyme synthesizing Dserine to regulate glutamate-N-methyl-D-aspartate neurotransmission.Proc Natl Acad Sci U S A. 1999;96:13409-14. http://dx.doi.org/10.1073/pnas.96.23.13409 [18] Zhou L, Zhu DY. Neuronal nitric oxide synthase: structure, subcellular localization, regulation, and clinical implications. Nitric Oxide. 2009;20:223-30. http://dx.doi.org/10.1016/j.niox.2009.03.001 [19] Sasabe J, Chiba T, Yamada M, Okamoto K, Nishimoto I, Matsuoka M et al. D-serine is a key determinant of glutamate toxicity in amyotrophic lateral sclerosis. EMBO J. 2007;26:4149-59. http://dx.doi.org/10.1038/sj.emboj.7601840 [20] Wu SZ, Bodles AM, Porter MM, Griffin WS, Basile AS, Barger SW. Induction of serine racemase expression and D-serine release from microglia by amyloid beta-peptide. J Neuroinflammation. 2004;1:2. http://dx.doi.org/10.1186/1742-2094-1-2 [21] Wu S, Barger SW. Induction of serine racemase by inflammatory stimuli is dependent on AP-1. Ann N Y Acad Sci. 2004;1035:133-46. http://dx.doi.org/10.1196/annals.1332.009 [22] Cook SP, Galve-Roperh I, Martinez del Pozo A, Rodriguez-Crespo I. Direct calcium 27

binding results in activation of brain serine racemase. J Biol Chem. 2002;277:27782-92. http://dx.doi.org/10.1074/jbc.M111814200 [23] Su TP, Hayashi T, Maurice T, Buch S, Ruoho AE. The sigma-1 receptor chaperone as an inter-organelle signaling modulator. Trends Pharmacol Sci. 2010;31:557-66. http://dx.doi.org/10.1016/j.tips.2010.08.007 [24] Hayashi T, Su TP. Regulating ankyrin dynamics: Roles of sigma-1 receptors. Proc Natl Acad Sci U S A. 2001;98:491-6. http://dx.doi.org/10.1073/pnas.98.2.491 [25] Watkins LR, Milligan ED, Maier SF. Glial activation: a driving force for pathological pain.Trends Neurosci. 2001;24:450-5. http://dx.doi.org/10.1016/S0166-2236(00)01854-3 [26] Oliet SH, Mothet JP. Regulation of N-methyl-D-aspartate receptors by astrocytic Dserine.Neuroscience. 2009;158:275-83. http://dx.doi.org/10.1016/j.neuroscience.2008.01.071 [27] Wake K, Yamazaki H, Hanzawa S, Konno R, Sakio H, Niwa A et al. Exaggerated responses to chronic nociceptive stimuli and enhancement of N-methyl-D-aspartate receptormediated synaptic transmission in mutant mice lacking D-amino-acid oxidase. Neurosci Lett. 2001;297:25-8. http://dx.doi.org/10.1016/S0304-3940(00)01658-X [28] Dieb W, Hafidi A. Astrocytes are involved in trigeminal dynamic mechanical allodynia: potential role of D-serine. J Dent Res. 2013;92:808-13. http://dx.doi.org/10.1177/0022034513498898 [29] Choi SR, Roh DH, Yoon SY, Kang SY, Moon JY, Kwon SGet al. Spinal sigma-1 receptors activate NADPH oxidase 2 leading to the induction of pain hypersensitivity in mice and mechanical allodynia in neuropathic rats.Pharmacol Res. 2013;74:56-67. http://dx.doi.org/10.1016/j.phrs.2013.05.004 [30] Mao J, Price DD, Hayes RL, Lu J, Mayer DJ. Differential roles of NMDA and nonNMDA receptor activation in induction and maintenance of thermal hyperalgesia in rats with painful peripheral mononeuropathy.Brain Res. 1992;598:271-8. http://dx.doi.org/10.1016/0006-8993(92)90193-D [31] Wagner R, Janjigian M, Myers RR. Anti-inflammatory interleukin-10 therapy in CCI neuropathy decreases thermal hyperalgesia, macrophage recruitment, and endoneurial TNFalpha expression.Pain. 1998;74:35-42. http://dx.doi.org/10.1016/S0304-3959(97)00148-6 [32] Ro LS, Chen ST, Tang LM, Jacobs JM. Effect of NGF and anti-NGF on neuropathic pain in rats following chronic constriction injury of the sciatic nerve.Pain. 1999;79:265-74. http://dx.doi.org/10.1016/S0304-3959(98)00164-X [33] Zhao WJ, Gao ZY, Wei H, Nie HZ, Zhao Q, Zhou XJet al. Spinal D-amino acid oxidase contributes to neuropathic pain in rats. J Pharmacol Exp Ther. 2010;332:248-54. http://dx.doi.org/10.1124/jpet.109.158816 28

[34] Hopkins SC, Zhao FY, Bowen CA, Fang X, Wei H, Heffernan ML et al. Pharmacodynamic effects of a D-amino acid oxidase inhibitor indicate a spinal site of action in rat models of neuropathic pain. J Pharmacol Exp Ther. 2013;345:502-11. http://dx.doi.org/10.1124/jpet.113.204016 [35] Ito M, Yoshikawa M, Ito K, Matsuda M, Jin XL, Takahashi S et al. Antinociceptive effect of intracerebroventricular administration of D-serine on formalin-induced pain. J Anesth. 2014;28:228-34. http://dx.doi.org/10.1007/s00540-013-1708-3 [36] Hamase K, Konno R, Morikawa A, Zaitsu K. Sensitive determination of D-amino acids in mammals and the effect of D-amino-acid oxidase activity on their amounts.Biol Pharm Bull. 2005;28:1578-84. http://dx.doi.org/10.1248/bpb.28.1578 [37] Gong N, Gao ZY, Wang YC, Li XY, Huang JL, Hashimoto Ket al. A series of D-amino acid oxidase inhibitors specifically prevents and reverses formalin-induced tonic pain in rats. J Pharmacol Exp Ther. 2011;336:282-93. http://dx.doi.org/10.1124/jpet.110.172353 [38] Wolosker H, Panizzutti R, De Miranda J. Neurobiology through the looking-glass: Dserine as a new glial-derived transmitter. Neurochem Int. 2002;41:327-32. http://dx.doi.org/10.1016/S0197-0186(02)00055-4 [39] Miraucourt LS, Peirs C, Dallel R, Voisin DL. Glycine inhibitory dysfunction turns touch into pain through astrocyte-derived D-serine. Pain. 2011;152:1340-8. http://dx.doi.org/10.1016/j.pain.2011.02.021 [40] Bennett AD, Everhart AW, Hulsebosch CE. Intrathecal administration of an NMDA or a non-NMDA receptor antagonist reduces mechanical but not thermal allodynia in a rodent model of chronic central pain after spinal cord injury.Brain Res. 2000;859:72-82. http://dx.doi.org/10.1016/S0006-8993(99)02483-X [41] Jensen TS, Finnerup NB. Allodynia and hyperalgesia in neuropathic pain: clinical manifestations and mechanisms. Lancet Neurol. 2014;13:924-35. http://dx.doi.org/10.1016/S1474-4422(14)70102-4 [42] Michaelis M, Vogel C, Blenk KH, Jänig W. Algesics excite axotomised afferent nerve fibres within the first hours following nerve transection in rats. Pain. 1997;72:347-54. http://dx.doi.org/10.1016/S0304-3959(97)00050-X [43] Michaelis M, Vogel C, Blenk KH, Arnarson A, Jänig W. Inflammatory mediators sensitize acutely axotomized nerve fibers to mechanical stimulation in the rat.J Neurosci. 1998;18:7581-7. http://www.jneurosci.org/content/18/18/7581.long [44] Jhaveri MD, Elmes SJ, Kendall DA, Chapman V. Inhibition of peripheral vanilloid TRPV1 receptors reduces noxious heat-evoked responsesof dorsal horn neurons in naive, carrageenan-inflamed and neuropathic rats. Eur J Neurosci. 2005;22:361-70. http://dx.doi.org/10.1111/j.1460-9568.2005.04227.x [45] Campana WM. Schwann cells: activated peripheral glia and their role in neuropathic 29

pain.Brain Behav Immun. 2007;21:522-7. http://dx.doi.org/10.1016/j.bbi.2006.12.008 [46] Ma QP, Woolf CJ. Noxious stimuli induce an N-methyl-D-aspartate receptor-dependent hypersensitivity of the flexion withdrawal reflex to touch: implications for the treatment of mechanical allodynia.Pain. 1995;61:383-90. http://dx.doi.org/10.1016/0304-3959(94)00195-K [47] Field MJ, Bramwell S, Hughes J, Singh L. Detection of static and dynamic components of mechanical allodynia in rat models of neuropathic pain: are they signalled by distinct primary sensory neurones? Pain. 1999;83:303-11. http://dx.doi.org/10.1016/S0304-3959(99)00111-6 [48] Baron R.Peripheral neuropathic pain: from mechanisms to symptoms.Clin J Pain. 2000;16:S12-20. http://www.ncbi.nlm.nih.gov/pubmed/?term=Clin+J+Pain.+2000%3B16%3AS12-20 [49] Woolf CJ, Thompson SW. The induction and maintenance of central sensitization is dependent on N-methyl-D-aspartic acid receptor activation; implications for the treatment of post-injury pain hypersensitivity states.Pain. 1991;44:293-9. http://dx.doi.org/10.1016/0304-3959(91)90100-C

30

Figure legends Figure 1. Immunofluorescent images and graph showing an elevation of D-serine levels in the spinal cord following CCI surgery.(A) The specificity of the D-serine antibody (Ab) was tested using a pre-absorption test with a D-serine antigen (Ag; 0.1, 1 or 10 mg/ml). D-serineimmunoreactivity was not detected in any of the spinal sections of CCI mice 3 days postsurgery that were processed with the anti-D-serine-antibody pre-absorbed with a maximal concentration (10 mg/ml) of D-serine-conjugated to BSA. (B) IF analysis with an anti-Dserine antibody was performed on lumbar spinal cord sections from CCI mice at 0, 1, 3 and 7 days post-surgery. D-serine-immunoreactive (ir) cells in the ipsilateralSDH (laminae I and II)and NP (laminae III and IV) regions were significantly increased from days 1 to 3 postCCI as compared with that of the day 0 group. This increase peaked at day 3 post-surgery and decreased thereafter. Asignificant difference was only detected in the SDH region at day 7 post-surgery compared with that of the day 0 group. Representative photomicrographs of Dserine-ir cells indicate that D-serine is significantly increased in the ipsilateral superficial layers of the dorsal horn, but also increased in the deeper laminae of CCI mice compared to the day 0 groups. Arrows indicate D-serine-ir cells.*P<0.05, **P<0.01 and ***P<0.001 as compared with those of the day 0 group. SDH, superficial dorsal horn; NP, nucleus proprius; NECK, neck region. Scale bar, 200μm.

Figure 2. Immunofluorescent images showing distribution of D-serine in the spinal cord following CCI surgery. (A–C) Double IF labeling was performed with an antibody against Dserine (D-serine, green) and an antibody against GFAP (A), a marker for astrocytes, NeuN (B), a marker for neurons or Iba-1 (C), a marker for microglia (red staining)on lumbar spinal cord sections from CCI mice at 3 days post-surgery. D-serine was colocalized with GFAPpositive astrocytes in the superficial dorsal horn and NeuN-positive neurons in the deepdorsal horn. There was no evidence of D-serine staining in Iba-1 positive microglia. Arrows indicate 31

colocalization of D-serine and GFAP or NeuN.Scale bar, 200μm.

Figure 3. Graph and immunofluorescent images showing an elevation inSrr expression in the spinal cord following CCI surgery and specific localization of Srrto dorsal horn astrocytes. (A) Western blot analysis was performed to determine the levels of Srr expression at 0, 1, 3 and 7 days post-CCI surgery. The expression of Srr in the ipsilateral dorsal horn was significantly increased on postoperative day 1 in comparison to Srr expression on day 0 and slowly decreased thereafter(n= 5 at each time point in the CCI groups). (B) Double staining was performed at day 1 post-CCI surgery to determine which cell types express Srrimmunoreactivity in the ipsilateral dorsal horn in CCI mice. Double-labeling with Srr and GFAP revealed a distinct colocalization of Srr in astrocytes. There was no evidence of Srr staining in neurons or Iba-1 positive microglia. Arrows indicate colocalization of Srr and GFAP.*P<0.05as compared with those of the day 0 group. Scale bar, 200μm.

Figure 4.Endogenous D-serine contributes to the development of MA, but not TH in CCI mice.(A) Repeated daily, intrathecal treatment with the endogenous D-serine degrading enzyme,DAAO (0.001, 0.01 or 0.1 U)on postoperative days 0-3 reduced the CCI-induced increase in PWF (%) to innocuous mechanical stimuli (MA), as compared with vehicletreated CCI mice in a dose dependent manner. After the termination of repeated DAAO injection on day 3, this suppressive analgesic effect of DAAO was sustained throughout the 21-day experimental period following CCI surgery. (B) On the other hand, the CCI-induced decrease in PWL (seconds) to heat stimuli (TH) was not influenced by repeated intrathecal treatment with DAAO throughout the 21-day testing period. (C) Repeated daily treatment with the Srr inhibitor, LSOS (1, 3 or 10 nmol)on postoperative days 0-3 also reduced MA as compared with vehicle-treated CCI mice in a dose dependent manner. (D) Conversely, TH was not affected by LSOS treatment. *P<0.05, **P<0.01 and ***P<0.001 as compared with 32

those of CCI+VEH group.

Figure 5. The CCI-induced increase in spinal D-serine levels is blocked by intrathecal injection of BD-1047, DAAO or LSOS. (A) IF analysis indicated that sustained intrathecal administration of the Sig-1R antagonist, BD-1047 (100 nmol, CCI+BD), the endogenous Dserine degrading enzyme, DAAO (0.1 U, CCI+DAAO) or the Srr inhibitor, LSOS (10 nmol, CCI+LSOS) on postoperative 0-3 significantly attenuated the CCI-induced increase in the number of D-serine-ir cells in the SDH and NP regions compared with the vehicle treated group. (B) Representative photomicrographs of L4-5 spinal cord sections illustrating D-serineir cells in the sham group (SHAM), saline treated CCI group (CCI+VEH),BD-1047 treated CCI group (CCI+BD), DAAO treated CCI group (CCI+DAAO) and LSOS treated CCI group (CCI+LSOS).Arrows indicate D-serine-ir cells.*P<0.05 and ***P<0.001 as compared with those of SHAM group, and #P<0.05 and##P<0.01 as compared with those of the CCI+VEH group.SDH, superficial dorsal horn; NP, nucleus proprius; NECK, neck region. Scale bar, 200μm.

Figure 6. Western blot showing that the CCI-induced increase in Srr expression is blocked by intrathecal injection of the Sig-1R antagonist, BD-1047 and immunofluorescent images illustrating that Srrimmunoreactivity coexists with Sig-1R-ir in dorsal horn cells. (A) Western blot analysis was performed on day 1 after CCI surgery to determine whether the CCIinduced increase in Srr expression is regulated by Sig-1R activation during the induction phase. Intrathecal treatment with BD-1047 during the induction phase significantly reduced the CCI-induced increase in Srr expression as compared with mice in the vehicle-treated CCI group (n= 5 at each groups). (B) Double staining was performed at day 1 post-CCI surgery using an anti-Srr antibody in combination with antibodies for Sig-1R. Double staining with Sig-1R antibody showed that the Srr expression is localized to Sig-1R-ir cells, which 33

presumably are astrocytes based on the data shown in figure 2. These results indicate that Srr and Sig-1R expression occur in the same cells in the spinal cord in CCI mice.Arrows indicate colocalization of Srr and Sig-1R.*P<0.05 as compared with those of SHAM group, and #P<0.05 as compared with those of the CCI+VEH group. Scale bar, 10μm.

Figure 7.Intrathecal injection of D-serine restores the CCI-induced MA that is blocked by BD-1047 administration. (A) Treatment with the Sig-1R antagonist, BD-1047 (BD) reduces the CCI-induced increase in PWF (%) to innocuous mechanical stimuli (MA). Treatment with exogenous D-serine (D-ser; 50 or 500 nmol in combination with BD-1047) restored the CCI-induced MA that was blocked by BD-1047.(B) Conversely, the decrease in PWL (seconds) to heat stimuli was unaffected by repeated intrathecal treatment with BD-1047or concomitant D-serine and BD-1047 treatment. *P<0.05, **P<0.01 and ***P<0.001 as compared with those of the CCI+BD group.

34

Supplementary Figure legends Supplementary Figure 1. Original western blots from which Fig.3A was created illustrating the time course of CCI-induced changes in serine racemase (Srr) expression and the corresponding expression of β-actin in the lumbar spinal cord. (A) Western blot analysis was performed to determine the levels of Srr expression at 0, 1, 3 and 7 days post-CCI surgery. The expression of Srr in the ipsilateral dorsal horn was increased on postoperative days 1 and 3 in comparison to Srr expression on day 0 (n = 3 at each time point in the CCI groups). The yellow arrow indicates the band of Srr proteins. (B) The corresponding β-actin expression is shown in B.

Supplementary Figure 2. Original western blots from which Fig.6A was created illustrating the effect of intrathecal administration of the Sig-1R antagonist, BD-1047 on the CCIinduced increase in serine racemase (Srr) expression. (A) Western blot analysis was performed on day 1 after CCI surgery to determine whether the CCI-induced increase in Srr expression is regulated by Sig-1R activation during the induction phase. Intrathecal treatment with BD-1047 during the induction phase reduced the CCI-induced increase in Srr expression as compared with mice in the vehicle-treated CCI group (n = 3 in each group). The yellow arrow indicates the band of Srr proteins. (B) The corresponding β-actin expression is shown in B. Key words; Sig-1R, D-serine, Astrocyte, Mechanical allodynia, Neuropathic pain

35

36

37

38

39

40

41

42