Glial cell activation in the spinal cord and dorsal root ganglia induced by surgery in mice

European Journal of Pharmacology 702 (2013) 126–134

Contents lists available at SciVerse ScienceDirect

European Journal of Pharmacology journal homepage: www.elsevier.com/locate/ejphar

Neuropharmacology and analgesia

Glial cell activation in the spinal cord and dorsal root ganglia induced by surgery in mice Asuncio´n Romero, Elizabeth Romero-Alejo, Nuno Vasconcelos, Margarita M. Puig n  ´ Me dica, Hospital del Mar, Universitat Autonoma Department of Anesthesiology, Pain Research Unit, Institut Municipal d’Investigacio de Barcelona, Barcelona, Spain

a r t i c l e i n f o

a b s t r a c t

Article history: Received 5 September 2012 Received in revised form 18 January 2013 Accepted 29 January 2013 Available online 8 February 2013

In rodents, surgery and/or remifentanil induce postoperative pain hypersensitivity together with glial cell activation. The same stimulus also produces long-lasting adaptative changes resulting in latent pain sensitization, substantiated after naloxone administration. Glial contribution to postoperative latent sensitization is unknown. In the incisional pain model in mice, surgery was performed under sevoflurane þremifentanil anesthesia and 21days later, 1 mg/kg of (  ) or (þ ) naloxone was administered subcutaneously. Mechanical thresholds (Von Frey) and glial activation were repeatedly assessed from 30 min to 21days. We used ionized calcium binding adaptor molecule 1 (Iba1) and glial fibrillary acidic protein (GFAP) to identify glial cells in the spinal cord and dorsal root ganglia by immunohistochemistry. Postoperative hypersensitivity was present up to 10 days, but the administration of (  ) but not ( þ ) naloxone at 21days, induced again hyperalgesia. A transient microglia/macrophage and astrocyte activation was present between 30 min and 2days postoperatively, while increased immunoreactivity in satellite glial cells lasted 21days. At this time point, (  ) naloxone, but not (þ ) naloxone, increased GFAP in satellite glial cells; conversely, both naloxone steroisomers similarly increased GFAP in the spinal cord. The report shows for the first time that surgery induces long-lasting morphological changes in astrocytes and satellite cells, involving opioid and toll-like receptors, that could contribute to the development of latent pain sensitization in mice. & 2013 Elsevier B.V. All rights reserved.

Keywords: Postoperative pain Hyperalgesia Glial cell activation Latent pain sensitization Naloxone

1. Introduction Nociceptive hypersensitivity after tissue injury and/or opioid administration has been widely investigated in animal models and man (Ce´le´rier et al., 2001; Hay et al., 2009; Richebe´ et al., 2005; Silverman, 2009). Its relevance is partially related to the fact that in surgical patients, the extent and duration of postoperative pain hypersensitivity seems to be a critical factor contributing to the development of chronic post-surgical pain in genetically predisposed individuals (Kehlet et al., 2006; Macrae, 2008). In general, and depending on the surgery, healing of the wound occurs after a period of days–weeks during which pain and hyperalgesia gradually disappear. After tissue repair, silent (latent) long-lasting plastic adaptations in neuronal/non-neuronal systems remain, and exposure to new stimulus (nociceptive/non-nociceptive) precipitate again hyperalgesia in animal models of nociception (Campillo et al., 2011; Le Roy et al., 2011; Rivat et al., 2007).

n Correspondence to: Department of Anesthesiology, IMIM-Hospital del Mar, Paseo Marı´timo 25, 08003 Barcelona, Spain. Tel.: þ34 93 2483527; fax: þ 34 93 2483617. E-mail addresses: [email protected] (A. Romero), [email protected] (E. Romero-Alejo), [email protected] (N. Vasconcelos), [email protected], [email protected] (M.M. Puig).

0014-2999/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ejphar.2013.01.047

This phenomenon known as latent pain sensitization or longtime pain vulnerability, could be the basis for the progression from acute to chronic pain, but the precise events underlying this transformation are poorly understood. In a rat model of incisional pain, the preoperative administration of nefopam prevented latent pain sensitization (Laboureyras et al., 2009), although it is not yet know if the drug could prevent chronic post-surgical pain in humans. Other reports have shown in rodents that NMDA antagonists (Rivat et al., 2002), or protein kinase M zeta inhibitors (Asiedu et al., 2011) can also block this latent sensitized state, suggesting that different mechanism are involved in the development of persistent pain. Among the multiple potential mechanism implicated in pain and opioid-induced hypersensitivity, spinal cord (SC) glial cell activation has been consistently reported after inflammation, nerve injury (Peters et al., 2010; Raghavendra et al., 2004; Romero-Sandoval et al., 2008) and/or opioid exposure (Garrido et al., 2005; Horvath et al., 2010). It is also interesting that acute opioid administration also activates satellite glial cells (SGCs) in the dorsal root ganglia (DRG), implicating the peripheral nervous system in opioid analgesia (Berta et al., 2012). Animal studies have shown that spontaneous activity in the DRG after peripheral injury and inflammation or manipulation of the DRG, may change the input to the spinal cord modifying spinal neuronal/glial activity (Xie et al., 2009).

A. Romero et al. / European Journal of Pharmacology 702 (2013) 126–134

Since tissue injury and opioid administration each induce glial activation, we hypothesized that surgery performed under remifentanil anesthesia, as routinely done in clinical practice, would induce robust changes in glial cell activity in the postoperative period, correlating with pain hypersensitivity. We also proposed that after the initial injury/activation, latent adaptative changes in glial cells would remain and the exposure to harmful or detrimental stimulus could induce glial re-activation. We used a challenge with naloxone to test for latent pain sensitization, as previously described by our group (Campillo et al., 2011; Romero et al., 2011). Given the increased relevance of the DRG as the first anatomical site where modulation of sensory information occurs, we assessed glial cell activity in the DRG and the dorsal horn of the SC, ipsilateral to the surgery. Our results could be helpful to establish future cellular and anatomical targets to prevent the development of latent pain sensitization after surgery, and the transformation of acute into chronic pain.

2. Material and methods 2.1. Animals Swiss CD1 male mice weighing 25–30 g obtained from Charles-River (CRIFFA, France) were used in all experiments. All procedures and animal handling met the guidelines of the European Communities directive 86/609/EEC regulating animal research. The protocol was approved by the institutional review board of our institution (Comite´ E´tico de Experimentacio´n Animal – Parc de Recerca Biome dica de Barcelona, Spain). All the experimental procedures were carried out in the animal facilities located in the Parc de Recerca Biome dica de Barcelona (accredited by the Association for Assessment and Accreditation of Laboratory Animal Care since June 2010). Animals were housed four per cage with autoclaved poplar soft wood bedding (Souralit S.L., Barcelona, Spain) and maintained in a room under a 12 h light/dark cycle (lights on at 8 AM), at controlled temperature (2171 1C) and relative humidity (55 710%). Food and water were available ad libitum except during behavioral evaluation. 2.2. Surgery We used the incisional postoperative pain mouse model previously described (Brennan et al., 1996) and validated in our laboratory (Ce´le´rier et al., 2006). Animals were anesthetized with sevoflurane delivered for 30 min via a nose mask (induction, 3.5% v/v; surgery, 3.0% v/v) in a sterile operating room. A 0.7 cm longitudinal incision was made with a number 20 blade through the skin and fascia of the plantar surface of the right hind paw, starting 0.3 cm from the proximal edge of the heel extending toward the toes. The underlying plantaris muscle was exposed and incised longitudinally, keeping the muscle origin and insertion intact. After hemostasis with slight pressure, the skin was closed with two 6–0 silk sutures and the wound covered with povidone–iodine antiseptic ointment. After surgery, the animals were allowed to recover under a heat source in cages with sterile bedding. 2.3. Behavioral testing Punctate mechanical stimulus (referred as hyperalgesia throughout the text) served as a measure of nociception. Before the experiments, animals were habituated to the equipment for 2–3 days (without nociceptive stimulation). All behavioral experiments were performed between 9:00 AM and 4:00 PM. Hyperalgesia was

127

measured by the hind paw withdraw response to von Frey filament stimulation. Animals were placed in methacrylate cylinders (30 cm high, 9 cm diameter; acquired from Servei Estacio´, Barcelona, Spain) with a wire grid bottom through which the von Frey filaments were applied (bending force range from 0.008 to 2 g; North Coast Medical, Inc., San Jose, CA). Animals were allowed to habituate for 2 h before testing, to achieve immobility. The filament force was increased or decreased according to the response. The upper limit value (2 g) was assigned when there was no response and the threshold of response was calculated using the up–down method (Chaplan et al., 1994). Paw shaking or licking were considered nociceptive-like responses. Both hind paws were alternately tested. 2.4. Perfusion, tissue processing and immunohistochemistry For each time point, 4–5 mice were deeply anesthetized with sodium pentobarbital (60 mg/kg, intraperitoneal) and perfused intracardially with phosphate-buffered saline (PBS; 0.1 M, pH 7.4) followed by a fixative containing 10% formalin. The spinal cord and the dorsal root ganglia (ipsilateral and contralateral) from the L4–L6 segments were removed and post-fixed for 2 h. Tissues were removed and then cryoprotected for 24 h at 4 1C in 20% sucrose in 0.1 M PBS. After that, tissues were embedded in TissueTek O.C.T. compound (Sakura Finetek, Zoeterwoude, The Netherlands), and transverse sections (25 or 13 mm thickness, respectively) were cut using a cryostat (Leica, Madrid, Spain) at  21 1C. Then, six or seven sections of each tissue were serially cut, placed on gelatinized slides, and processed for immunohistochemistry. Every fifth section was picked from a series of consecutive tissue sections, preventing significant overlap of neuronal profiles and minimizing repeated counting of the same neuronal profile across multiple sections. Non-specific binding was blocked with 5% normal goat serum (Vector Laboratories, Burlingame, CA) for 1 h at room temperature. Tissue sections were processed for 36 h at 41C, for ionized calcium binding adaptor molecule 1 (Iba1), glial fibrillary acidic protein (GFAP) immunohistochemistry, according to the following primary antibody dilutions: Iba1 (1:2000, Wako Chemical, Richmond, VA), GFAP (1:500, DakoCytomation, Denmark). Sections were then incubated in secondary antibody AlexaFluor 488 (1:2000; Molecular Probes Europe BV, The Netherlands) or Cy3 (1:2000, Chemicon International, Temecula, CA) goat antirabbit IgG, for 1 h at room temperature in goat serum in PBS (5% goat serum and 1% Triton-X-100, Sigma). For control samples, the primary antibody was omitted. All sections were then mounted with Vectashields þDapi (Vector Laboratories, Burlingame, CA) and cover-slipped. 2.5. Quantification To avoid variability in the staining procedure, all the sections to be compared were processed together, and images were acquired under the same exposure conditions. Fluorescence images were acquired with a microscope (Leica DMR, Madrid, Spain) outfitted with a filter set for Alexa 488 and Cy3, and equipped with a digital camera (Leica DFC 300 FX, Madrid, Spain). Spinal cord images were captured and recorded under 4x or 20x objectives, and the dorsal root ganglia under a 10x objective. Immunoreactive images were counted in a blinded manner on randomly selected spinal cord and DRG from L4–L5 (n¼ 4–5 animals per experimental group and time point). For qualitative analysis, we used samples from each group and time point from at least 4 slices per animal; values were averaged and compared with the control group. All images were analyzed with Image J software (version 1.44, NIH, Bethesda, MD). Quantification of Iba1 and GFAP was performed by determining the immunofluorescence intensity of cells within a fixed area in both

128

A. Romero et al. / European Journal of Pharmacology 702 (2013) 126–134

tissues (laminae I-III of the spinal cord, and DRG); the number of activated cells for Iba1 or the percentage of stained occupied area for GFAP over these specific regions of interest were assessed. 2.6. Experiments All animals received the same inhaled concentration of sevoflurane (3.0–3.5% (v/v%)) plus a s.c. infusion of saline or remifentanil administered during a period of 30 min. Special care was taken to use the smallest number of animals per group. The experiments were performed in mice receiving one of the following treatments:

i) Sevofluraneþs.c. saline (control group, CTL), a treatment that does not alter nociceptive thresholds (Ce´le´rier et al., 2006). ii) Plantar incision (surgery) performed under sevoflurane anesthesiaþs.c. saline (incision group, INC). iii) Sevofluraneþs.c. remifentanil 80 mg/kg (remifentanil group, REMI). iv) Plantar incision (surgery) performed under sevoflurane anesthesiaþs.c. remifentanil 80 mg/kg (incisionþremifentanil group, INCþREMI) The following experiments were performed in order to determine: 2.6.1. The pronociceptive effects of incision, remifentanil, and incision þremifentanil on mechanical thresholds, using von Frey filaments These experiments were performed to corroborate our previous work in this model. After habituation, the average mechanical thresholds from two to three consecutive days were registered in order to calculate baseline values. After manipulation, animals were tested at 4 h and on days 2, 4, 7, 10, 14, and 21, using 8–10 animals per group. We did not assess nociceptive behavior at time 30 min, to avoid possible residual effects of the anesthetics (sevoflurane and/or remifentanil). Based on our previous results with the same model, the highest degree of hyperalgesia is observed in the INC þREMI group 4 h to 2 days after manipulation; afterwards thresholds return to baseline ˜ ero et al., approximately 10 days after manipulation (Caban 2009; Campillo et al., 2010; Ce´le´rier et al., 2006). 2.6.2. The effects of saline, ( ) naloxone or ( þ) naloxone on mechanical thresholds after complete recovery of nociceptive thresholds On day 21 after manipulation, after all mice had completely recovered the baseline mechanical thresholds and in blinded conditions, they received subcutaneously (s.c.), one of the following: saline, ( ) naloxone (1 mg/kg), or (þ) naloxone (1 mg/kg). Twenty minutes later, mechanical thresholds were re-assessed using Von Frey filaments in all groups of study. 2.6.3. Morphological changes of microglia/macrophages and astrocytes/satellite glial cells in the spinal cord and the dorsal root ganglia, using immunohistochemistry After sacrifice, samples were obtained from non-stimulated mice at 30 min, 4 h, and 2, 7, 14, and 21 days after manipulation (4–5 animals per group and time point). The effects of s.c. saline, (  ) naloxone (1 mg/kg) or ( þ) naloxone (1 mg/kg) on glial immunoreactivity were also assessed in all groups of study, both in the spinal cord and the dorsal root ganglia. In these experiments, we evaluated only the ipsilateral spinal cord and dorsal root ganglia in the different groups of animals.

2.7. Drugs Sevoflurane (Sevoranes; Abbot Laboratories SA, Madrid, Spain), remifentanil (Ultivas; GlaxoSmithKline, Madrid, Spain), and naloxone hydrochloride (Naloxone KERN PARMAs, Kern Pharma, Barcelona, Spain) were supplied by the Department of Anesthesiology of the Hospital del Mar (Barcelona, Spain). ( þ) Naloxone hydrochloride was a gift from the National Institute of Drug Abuse (Bethesda, MD). Drug doses were selected on the ˜ ero et al., basis of previous studies: remifentanil (80 mg/kg; Caban 2009; Campillo et al., 2010) was dissolved in saline (NaCl 0.9%) and infused s.c. over a period of 30 min (rate, 0.8 ml/h) using a KD Scientific pump (KD Scientific Inc., Holliston, MA). (  ) Naloxone (1 mg/kg, a non receptor-specific dose; Ce´le´rier et al., 2001; Richebe´ et al., 2005), ( þ)-naloxone hydrochloride (1 mg/kg) (at the same doses of the active isomer). Control animals (sham-operated) underwent a sham procedure that consisted on the administration of sevoflurane plus s.c. saline in identical conditions.

2.8. Statistical analysis 2.8.1. Behavioral testing For each mouse and time point, the nociceptive responses in grams (von Frey) were also expressed as percent changes (%)7standard error of the mean (S.E.M.), with respect to baseline values. Negative values indicate net pronociceptive effects (shown in Table 1). To analyze changes in mechanical thresholds in the different groups of treatment, two-way ANOVA models for repeated measures (time and treatment) were used. For the post-hoc analyses comparing the values at each time with the baseline values, Dunnett’s many to-one test in the framework of these models was applied. The time-effect Areas Above Curve (AACs) (0–21days before injecting saline, ( ) or ( þ) naloxone), corresponding to each experimental condition, were also calculated and comparisons were accomplished with one-way ANOVA, followed by a post-hoc Tukey’s test.

2.8.2. Immunohistohemistry The number of Iba1-immunoreactive cells and the percent stained area for GFAP is shown as mean values 7S.E.M. Quantification of immunohistochemical data was carried out using sections of samples from at least 4 different animals per time point and group. Data were analyzed with a one-way ANOVA, followed by a post-hoc Games–Howell. All statistical analysis was performed with SPSS (version 18, Chicago, IL) and AACs with GraphPad Prism software. In all instances, a P value o0.05 was considered statistically significant.

3. Results In the incisional pain model in mice, we have previously reported that surgery performed under remifentanil and/or sevoflurane anesthesia induces postoperative pain hyperalgesia lasting up to 10 days. In order to reduce the length of the manuscript, in the Figs. 1–5 of the present report we show the results obtained in the INCþREMI group, while the complete data (all groups) is included in the Tables 1–3.

A. Romero et al. / European Journal of Pharmacology 702 (2013) 126–134

129

Table 1 Time course of mechanical hyperalgesia in the four groups of study. Time

Mechanical thresholds (g) and percent change over baseline (%) CTL

INC

REMI

INC þ REMI

Baseline 4h 2 days 4 days 7 days 10 days 14 days 21 days

1.16 7 0.07 1.077 0.05 1.037 0.07 1.17 7 0.07 1.15 7 0.09 1.067 0.08 1.12 7 0.06 1.077 0.01

(  6.447 4.54) (  7.527 10.33) (2.067 5.50) (4.217 3.90) (  0.677 3.99) (  5.687 6.83) (  9.107 8.37)

1.19 7 0.06 0.53 7 0.07 0.59 7 0.03 0.72 7 0.06 1.02 7 0.06 1.09 7 0.08 1.18 7 0.06 1.23 7 0.06

( 54.56 7 6.71)c ( 49.83 7 3.53)c ( 39.66 7 5.08)c ( 12.71 7 7.55) (  10.557 2.80) (  0.417 2.77) (3.667 2.45)

0.98 7 0.06 0.46 7 0.02 0.53 7 0.04 0.63 7 0.04 0.86 7 0.11 0.88 7 0.07 0.94 7 0.08 1.01 7 0.08

(  51.97 74.41)c (  46.08 73.82)c (  39.44 76.16)c (  11.76 710.21)a (  8.08 710.16) (  3.99 76.14) (4.11 78.44)

1.07 70.06 0.32 70.05 0.36 70.05 0.63 70.11 0.75 70.08 0.78 70.06 1.00 70.08 1.10 70.08

( 71.65 7 4.28)c ( 65.62 7 6.45)c ( 42.91 7 9.72)c ( 28.31 7 8.91)b ( 26.81 7 5.77)b (  7.487 7.80) (2.167 8.51)

21 days þSS 21 days þ(  )Nx 21 days þ(þ )Nx

1.16 7 0.04 1.067 0.03 1.077 0.05

(1.357 4.56) (0.387 4.59) (1.887 5.64)

1.03 7 0.08 0.607 0.02 1.07 7 0.05

(  5.607 5.44) ( 38.67 7 3.87)c (  0.417 3.51)

1.04 7 0.06 0.607 0.03 1.05 7 0.04

(2.80 76.37) (  43.60 75.96)c (0.36 75.07)

1.05 70.07 0.36 70.03 1.03 70.08

(  4.897 4.15) ( 66.69 7 2.89)c (  2.957 3.79)

SS, saline; Nx, naloxone (1 mg/kg). Two-way ANOVA for repeated measures test followed by Dunnett test. a b c

P o 0.05. Po 0.01. Po 0.001 vs. baseline.

hyperalgesia in all groups of study (Table 1). In all groups, the extent of naloxone-induced hyperalgesia was similar to that observed on day 2, although it was greater in the INCþREMI group (Table 1), confirming recent results from our group (Campillo et al., 2011). The hyperalgesia observed in all groups after a challenge with () naloxone reveals the development of latent pain sensitization in our model (Campillo et al., 2011; Romero et al., 2011). 3.2. Immunoreactive expression of Iba1 and GFAP (0–21 days) in the dorsal horn of the spinal cord

Fig. 1. Time course of mechanical hyperalgesia induced by a plantar incision performed under remifentanil anesthesia. Effects of naloxone (Nx). Each point (n¼ 8–10 mice) shows mean mechanical thresholds (g) 7 S.E.M., from ipsilateral hind paws (right), at 4 h (h) and on days (d) 2, 4, 7, 10, 14, and 21, after manipulation. On day 21 animals received a s.c. injection of saline (SS), (  ) naloxone (1 mg/kg) or ( þ) naloxone (1 mg/kg). Baseline (b) values are represented by the horizontal broken line. Black dots, incision þremifentanil group (INC þREMI), and gray circles, control group (CTL). Two-way ANOVA for repeated measures followed by Dunnett’s test. nPo 0.05, nnP o 0.01, nnnP o0.001 vs. baseline.

3.1. The pronociceptive effects of incision, remifentanil, and incision þremifentanil on mechanical thresholds, using von Frey filaments. Effects of naloxone Fig. 1 illustrates the time course of postoperative mechanical hyperalgesia in the INCþREMI group, although similar time-effect curves were obtained for each group of study. Mean mechanical thresholds (in grams) and the percent changes (%) with respect to baseline in all experimental groups are shown in Table 1. The areas above curve (AACs) for the 0–21 day period were also calculated, revealing that the INCþ REMI group induced a greater pronociceptive effect (749.0768.5) than the INC (350.0790.2) or REMI (480.0768.0) groups (Po0.001), corroborating our previous ˜ero et al., 2009; Campillo et al., 2010). results (Caban On day 21 after manipulation, when the surgical wound was completely healed, and nociceptive thresholds had returned to baseline values in all groups (Table 1) and the s.c. administration of ( ) but not (þ) naloxone or saline, induced significant mechanical

In the CTL group, the number of resting microglia cells (Iba1) and the % astrocyte stained area (GFAP) were 72.00 72.68 and 6.0570.37%, respectively (Table 2). In all groups receiving active treatment, the immunoreactive expression of Iba1 and GFAP was significantly increased at 30 min, 4 h and 2 days after manipulation (Po0.001 versus CTL). Glial cell immunoreactivity returned to control levels (CTL group) on day 7, and remained unchanged up to day 21. Fig. 2 shows the time course of microglia (panel A) and astrocyte (panel B) activation in the INC þREMI group, and Table 2 summarizes the data obtained in all experimental groups, at the different times of evaluation. The administration of either ( ) or (þ) naloxone on day 21, did not alter microglia immunoreactivity but induced significant astrocyte re-activation in all treatment groups, that was of a greater magnitude in the INCþ REMI group (Po0.001, Fig. 2). These results suggest that surgery and/or remifentanil induce latent, long-lasting changes in spinal cord astrocytes involving opioid and non-opioid mechanisms. Representative photomicrographs of Iba1 and GFAP immunoreactivity in the INCþREMI group (Fig. 3A and B, respectively), show the time points 30 min (immediate postoperative period), and the effects of naloxone on day 21. 3.3. Immunoreactive expression of Iba1 and GFAP (0–21 days) in the dorsal root ganglia. Effects of naloxone A significant increase in the number of activated cells expressing Iba1 (macrophages) was observed in all treatment groups at 30 min and 4 h (Po0.001 vs. CTL group), that remained activated until day 2 in the INC and INC þREMI groups (Table 3 and Fig. 4A). The three groups recovered the basal immunoreactivity values on day 7, that remained unchanged up to day 21 after manipulation.

130

A. Romero et al. / European Journal of Pharmacology 702 (2013) 126–134

Fig. 2. Microglia and astrocytes activation in the spinal cord, after surgery (incision) performed under remifentanil anesthesia. Effects of naloxone. The histograms show the mean values of the number of Iba-1 positive cells (microglia, panel A), and percent GFAP stained area (astrocytes, panel B), in the ipsilateral spinal cord, in the control (CTL), and in the incisionþ remifentanil (INC þREMI) groups, at different times after manipulation; the vertical bars indicate the S.E.M. On day 21 mice received a s.c. injection of saline (SS), (  ) naloxone (1 mg/kg) or (þ) naloxone (1 mg/kg). One-way ANOVA, followed by Games–Howell test. n¼ 4–5 Animals per group and time of evaluation. nP o 0.001 vs. control (CTL).

Fig. 3. Iba1 and GFAP expression in the spinal cord by immunofluorescence in the control and incisionþremifentanil groups. Representative photomicrographs showing Iba-1 (panel A) and GFAP (panel B) immunoreactivity, in control (CTL), and INCþ REMI groups at 30 min, and on day 21 after manipulation, in the ipsilateral dorsal horn of the spinal cord (C). On day 21 mice received a s.c. injection of saline (SS), ( ) naloxone (1 mg/kg) or (þ) naloxone (1 mg/kg). Scale bars: panels A and B, 50 mm; panel C, 200 mm.

Satellite glial cells (SGCs) showed increased GFAP immunoreactivity expression in all groups of study at 30 min and 4 h after manipulation (Po0.001 vs. CTL group). In the INC and REMI groups, immunoreactivity returned to control levels on day 2 (Table 3) and remained unaltered thereafter; however, in the INCþREMI group, a significant GFAP up-regulation was observed

at all times of evaluation up to day 21 (Po0.001 vs. CTL) (Table 3 and Fig. 4B). These results show that in our model, the activation of SGCs persists after the initial stimulus, in the absence of mechanical hypersensitivity. The administration of either ( ) or ( þ) naloxone on day 21, did not induce changes on macrophage resting state in any of the

A. Romero et al. / European Journal of Pharmacology 702 (2013) 126–134

131

Fig. 4. Macrophages and satellite glial cell activation in the dorsal root ganglia in mice that underwent surgery under remifentanil anesthesia. Effects of naloxone. The histograms show the mean number of Iba-1 positive cells (macrophages, panel A) and percent GFAP stained area (satellite glial cells, panel B), in the ipsilateral dorsal root ganglia, in the control (CTL), and in the incision þremifentanil (INCþ REMI) groups, at different times after manipulation; the vertical bars indicate the S.E.M. On day 21 mice received a s.c. injection of saline (SS), (  ) naloxone (1 mg/kg) or (þ ) naloxone (1 mg/kg). One-way ANOVA, followed by Games–Howell test. n ¼4–5 Animals per group and time of evaluation.nP o 0.001 vs. CTL.

Fig. 5. Iba-1 and GFAP immunoreactivity in the dorsal root ganglia, in the control and incision þ remifentanil groups. The photomicrographs show Iba1 (panel A) and GFAP (panel B) immunoreactivity of the ipsilateral dorsal root ganglia, in the control (CTL) and INC þ REMI groups at 30 min, and 21 days after manipulation. On day 21 mice received a s.c. injection of saline (SS), ( ) naloxone (1 mg/kg) or (þ) naloxone (1 mg/kg). Scale bars: 100 mm.

groups (Table 3). However, significant increases in SGCs immunoreactivity were observed in all treatment groups after ( ) but not after (þ ) naloxone, that was of a greater magnitude in the

INCþREMI group (Po0.001, Fig. 4B). In the same group, the administration of saline did not alter the enhanced levels of GFAP observed on day 21, while ( þ) naloxone returned

132

A. Romero et al. / European Journal of Pharmacology 702 (2013) 126–134

Table 2 Time course of the morphological changes in spinal cord glial cells induced by surgery and/or remifentanil administration. Immunoreactive expression of Iba-1 (microglia) and GFAP (astrocytes). Effect of naloxone. Time

Microglia (number of cells activated)

Astrocytes (% stained area)

CTL

INC

REMI

INC þREMI

CTL

INC

REMI

INC þ REMI

30 min 4h 2 days 7 days 14 days 21 days

72.007 2.68 – – – – – –

150.64 710.41a 112.22710.27a,b 111.50 77.15a 66.09 73.07 64.20 77.15 64.47 75.49

140.14 7 4.50a 131.06 7 7.27a,b 96.68 7 6.78a 67.007 3.46 66.807 4.31 69.65 7 5.37

163.04 7 7.90a 178.90 7 9.25a 117.397 5.85a 67.80 7 2.87 65.88 7 2.52 69.20 7 2.83

6.057 0.37 – – – – – –

12.067 0.76a 12.12 7 0.58a 11.407 0.44a 6.647 0.32 6.627 0.21 6.477 0.43

13.82 7 0.62a 13.75 7 0.69a 11.29 7 0.96a 6.78 7 0.56 6.93 7 0.43 6.45 7 0.44

12.13 7 0.80a 14.17 7 0.67a 10.877 0.57a 6.807 0.22 6.527 0.17 6.587 0.38

21 days þSS 21 days þ(  )Nx 21 days þ(þ )Nx

71.60 75.79 69.69 76.47 70.88 76.10

71.64 7 7.89 74.65 7 3.20 75.207 7.21

73.02 7 6.91 69.44 7 5.74 68.84 7 4.20

6.14 71.42 5.707 1.25 6.68 71.72

7.217 1.34 12.72 7 0.75a,b 9.357 0.62a

6.83 7 0.69 10.65 7 0.89a,b 9.37 7 0.48a

7.037 1.18 15.32 7 1.14a 11.34 7 1.18a

75.52 76.04 73.24 75.93 72.01 75.02

Data expressed as mean values7 S.E.M. SS, saline; Nx, naloxone (1 mg/kg ). CTL¼ control group; INC ¼ incision; REMI ¼remifentanil. One-way ANOVA followed by Games–Howell test. a b

P o0.001 vs. CTL. Po 0.001 vs. INC þ REMI.

Table 3 Time course of the morphological changes in glial cells of the dorsal root ganglia induced by surgery and/or remifentanil administration. Immunoreactive expression of Iba-1 (macrophages) and GFAP (satellite glial cells). Time

30 min 4h 2 days 7 days 14 days 21 days 21 days þSS 21 days þ(  )Nx 21 days þ(þ )Nx

Macrophages (number of cell activated)

Satellite glial cells (% stained area)

CTL

INC

REMI

INC þREMI

CTL

22.007 1.57 – – – – – –

33.307 2.42c 41.67 7 2.76a 28.55 7 1.79a,d 26.46 7 1.29 18.37 7 1.18 20.917 1.00

30.85 7 3.10c 33.88 7 3.27c 19.89 7 2.37e 22.47 7 1.60 18.24 7 2.17 20.64 7 1.87

57.00 7 8.38c 44.00 7 5.41c 38.40 7 4.61a 27.86 7 5.26 18.00 7 4.62 21.00 7 1.83

6.507 0.34 – – – – – –

19.67 7 1.56 21.54 7 1.23 20.567 2.45

21.34 7 0.98 20.64 7 2.17 19.24 7 1.87

20.02 7 1.04 23.00 7 0.79 20.06 7 1.03

5.47 70.82 6.47 71.27 6.62 71.17

22.27 7 6.26 19.92 7 3.57 18.38 7 3.53

INC

11.1 70.53c 11.5 70.64c,d 8.3 71.00f 8.10 70.19g 8.3 70.25 7.6 70.39 6.28 70.45 11.46 70.34c 6.30 70.85

REMI

INC þREMI

11.17 70.49c 10.80 70.90c,e 6.7 70.18f 6.90 70.50d 5.6 70.25d 5.8 70.31d

12.9 7 0.44c 13.9 7 0.37c 13.5 7 0.66c 11.47 7 0.54b 11.2 7 0.61b 10.1 70.28a

5.1 70.82 12.32 70.56b 6.2 70.25

11.16 7 0.31c 13.89 7 0.28c 5.98 7 0.17

Results are expressed as mean values 7 S.E.M. CTL¼ control group; INC ¼ incision; REMI ¼remifentanil. One-way ANOVA followed by Games–Howell test. a

P o0.05. Po 0.01. Po 0.001 vs. CTL. d Po 0.05. e P o0.01. f Po 0.001 vs. INC þ REMI group. g P o0.01 vs. REMI group. b c

GFAP immunoreactivity levels to control values (Table 3 and Figs. 4 and 5B). We show representative photomicrographs of Iba1 (Fig. 5A) and GFAP (Fig. 5B) immunoreactivity in the INCþREMI group.

4. Discussion The present investigation shows for the first time that astrocyte but not microglia participates in latent pain sensitization after surgery in mice. We also show distinct patterns of activation of glial cells in the SC (astrocytes) and DRG (SGCs), the latter lasting for at least three weeks. The persistent activation of SGCs could contribute to maintain a state of hyper-excitability of the peripheral nervous systems after surgery. Our work also shows for the first time that in the SC, astrocyte re-activation occurs after the administration of both steroisomers of naloxone, involving opioid and non-opioid mechanisms.

Our results demonstrate that tissue injury (surgery) and/or opioid administration (remifentanil) induce a transient activation of astrocytes in the SC that is no longer present 7 days after treatment. The same stimulus induces latent (silent) and persistent changes in the neuronal/glial complex compatible with the development of latent pain sensitization, a poorly characterized phenomenon (Campillo et al., 2011; Le Roy et al., 2011; Richebe´ et al., 2005; Rivat et al., 2007, 2002), proposed as the first step in the development of sub-acute/chronic post-surgical pain. In our work, the latent morphological changes in astrocyte were substantiated by the administration of ( ) naloxone 21 days after manipulation, that induced significant astrocyte-reactivation in the SC. It has been proposed that in the nervous system, injury could stimulate a simultaneous and long term enhancement of the inhibitory (i.e. endogenous opioids) and excitatory systems involved in nociceptive transmission (NMDA, other) that would be regulated at abnormal levels (Ce´le´rier et al., 2001) partially explaining latent pain sensitization. The systemic administration of (  ) naloxone would block the opioid receptors in neurons and

A. Romero et al. / European Journal of Pharmacology 702 (2013) 126–134

astrocytes, antagonizing the inhibitory modulation of endogenous opioids in both cells; as a consequence, hyperalgesia and astrocyte activation would occur. The transient astrocyte activation in the early postoperative period, and its re-activation after a challenge with (  ) naloxone, closely correlate with nociceptive behavior in our model, where we observe mechanical hyperalgesia during the first week after surgery that re-appears after (  ) naloxone administration on day 21. Further times were not evaluated in the present investigation because the experiments became too long and the animals overweight, although we have previously reported in the same model the presence naloxone-induced hyperalgesia up to 150 days after surgery (Campillo et al., 2011). Interestingly, the administration of ( þ) naloxone on day 21, also induced astrocyte re-activation in the SC, but did not produce hyperalgesia when assessing nociceptive behavior. This could be related to the fact that neuronal and glial cells both express opioid receptors in cell membranes (Bokhari et al., 2009; Burbassi et al., 2010; Kao et al., 2012) that are able to bind the (  ) but not the ( þ) isomers of endogenous and exogenous opioids (agonists or antagonists). ( þ) Naloxone cannot bind to neuronal opioid receptors, and does not induce hyperalgesia on day 21, but induced significant astrocyte re-activation in the spinal cord (Fig. 2). The positive correlation between hyperalgesia and astrocyte re-activation induced by ( ) naloxone suggests the implication of opioid receptors expressed on neuronal or glial cells in latent pain sensitization. Astrocyte re-activation after ( þ) naloxone, in the absence of hyperalgesia would probably reflect binding to non-opioid receptors located in astrocytes. However, based on the present investigation we cannot exclude that (  ) naloxone may induce hyperalgesia binding to both, neuronal opioid receptors and glial Toll Like Receptors (TLR). The effects of TLR antagonists or those of astrocyte inhibitors were not tested in the present investigation. Astrocytes and other glial cells express TLR on the cell surface. TLR4 sense endogenous ligands released after tissue injury as danger signals, inducing inflammation in infectious and noninfectious situations (Miyake, 2007). In animal models of neuropathic pain, genetic manipulation of cell surface receptor TLR4 expression, significantly attenuated behavioral hyperalgesia and the expression of spinal microglial and astrocytic markers, supporting the role of these receptors in the development of hyperalgesia after nerve injury (Tanga et al., 2005). Several studies also show that the effects of the (þ)isomer of opioid agonists and antagonists on glial cells are mediated through mechanisms distinct from binding to conventional opioid receptors (Wu et al., 2006). Thus, chronic infusion of either (  ) or ( þ) naloxone suppressed neuropathic pain and microglial activation in a chronic constriction injury model in rats (Hutchinson et al., 2006), suggesting the involvement of TLR4 signaling. Hutchinson et al. (2010b) using in vitro, vivo, and in silico approaches also showed that opioids may affect non-stereoselectively TLR4 signaling, and that the opioid-inactive ( þ) naloxone and the opioid-active ( ) naloxone both blocked TLR4 signaling by opioids and lipopolysaccharide (Hutchinson et al., 2010a). Other groups have investigated glial cell activation in the SC in rodents, using similar models of incisional pain (Ito et al., 2009; Obata et al., 2006; Romero-Sandoval et al., 2008; RomeroSandoval and Eisenach, 2007; Wen et al., 2011); these studies consistently report changes in microglia and astrocyte in the first days of the postoperative period. In the Obata study, microglia was immunoreactive several days after astrocyte activation, while we observe a simultaneous activation of both glial cells (microglia and astrocytes), a finding that could be related to the different animal species (rat, mice) and/or markers of glial activity.

133

On the basis of our results, we would like to propose that astrocyte activation participates in the plastic adaptative changes induced by surgery and/or remifentanil in our model, contributing to the development of latent pain sensitization in mice. However, we do not have a realistic explanation for the lack of correlation between astrocytes re-activation after ( þ) naloxone and hyperalgesia. The DRG is the first peripheral structure where modulation of sensory information occurs. The relevance of the DRG in sensory abnormalities (including pain), was first described by Devor and Wall (1990), that observed electrical cross-excitation between DRG neurons, facilitating central sensitization in the SC (Amir and Devor, 2000; Pogatzki et al., 2012). Recent reports in the literature show that after injury, the DRG undergoes significant plastic changes, involving glial activation, playing a relevant role in the development of chronic pain (Dublin and Hanini, 2007; Ohara et al., 2009; Takeda et al., 2009). In our experiments, we assessed glial cell activation in the SC and the DRG in the same animals, determining the expression of GFAP in SGCs by immunohistochemistry. Surgery and remifentanil, each one individually, induced significant increases in SGCs immunoreactivity immediately after surgery (30 min and 4 h), that could not be observed at later times of evaluation. However, when both detrimental stimulus were applied simultaneously (INCþREMI group), as routinely occurs in clinical practice when surgery is performed under remifentanil anesthesia, SGC activation was slightly enhanced in magnitude, but remarkably prolonged over time. Increased immunostaining of SGCs was observed as early as 30 min after treatment, then stabilized on day 7 at about 50% of the initial increase, and remained significantly increased up to day 21 (three weeks after surgery). At present, we do not have an explanation for this finding other than a potential predominance of the excitatory, in detriment of the inhibitory modulation in response to injury/opioids, in this peripheral tissue. Similarly, we do not have an explanation for the effect of ( þ) naloxone on day 21 that consistently returned SGC immunoreactivity to baseline control levels. Thus, SGC activation does not seem to correlate with pain hyperalgesia in our model, although due to its persistence over time, could be a reasonable marker of latent pain sensitization in mice. Spinal microglia and DRG macrophages were activated early after manipulation (30 min to approximately 2d) probably in relation to the immunological response to injury. In both tissues, the administration of remifentanil induced a comparable degree of immune-cell activation that the surgical incision, supporting the similarity of the mechanisms involved in opioid-induced hyperalgesia and peripheral tissue injury (Angst and Clark, 2006; Nickel et al., 2012). Macrophages and microglia cells were not reactivated after naloxone administration, suggesting that they do not take part in the plastic changes involved in latent pain sensitization in our model.

5. Conclusion Surgery under remifentanil anesthesia induces morphological changes in the spinal cord and the dorsal root ganglia in the early postoperative period in mice. A challenge with (  ) naloxone 21d after manipulation, stereospecifically blocked opioid receptors and triggered hyperalgesia and astrocyte and SGCs reactivation. The administration of ( þ) naloxone did not produce hyperalgesia, but induced astrocyte-reactivation, also reversing the sustained satellite glial activation in the periphery; the results suggest the implication of opioid and non-opioid receptors in postoperative latent pain sensitization in mice.

134

A. Romero et al. / European Journal of Pharmacology 702 (2013) 126–134

Acknowledgments The authors thank Klaus Langohr, Ph.D. for his assistance in the statistical analysis. ( þ) Naloxone hydrochloride was kindly provided from the National Institute on Drug Abuse (Bethesda, MD, USA). This work was supported by Grants from Fondo de Investigacio´n Sanitaria, Instituto de Salud Carlos III, Madrid, Spain (PS09/01270), Marato´ de Televisio´ de Catalunya TV3, Barcelona, Spain (071110), and the Endowed Chair in Pain Management  Universitat Autonoma de Barcelona-Parc de Salut Mar-Menarini (MMP). References Amir, R., Devor, M., 2000. Functional cross-excitation between afferent A- and C-neurons in dorsal root ganglia. Neuroscience 95, 189–195. Angst, M.S., Clark, J.D., 2006. Opioid-induced hyperalgesia: a qualitative systematic review. Anesthesiology 104, 570–587. Asiedu, M.N., Tillu, D.V., Melemedjian, O.K., Shy, A., Sanoja, R., Bodell, B., Ghosh, S., Porreca, F., Price, T.J., 2011. Spinal protein kinase M z underlies the maintenance mechanism of persistent nociceptive sensitization. J. Neurosci. 31, 6646–6653. Berta, T., Liu, T., Liu, Y.C., Xu, Z.Z., Ji, R.R., 2012. Acute morphine activates satellite glial cells and up-regulates IL-1beta in dorsal root ganglia in mice via matrix metalloprotease-9. Mol. Pain 22, 8–18. Bokhari, S.M., Yao, H., Bethel-Brown, C., Fuwang, P., Williams, R., Dhillon, N.K., Hegde, R., Kumar, A., Buch, S.J., 2009. Morphine enhances Tat-induced activation in murine microglia. J. Neurovirol. 15, 219–228. Brennan, T.J., Vandermeulen, E.P., Gebhart, G.F., 1996. Characterization of a rat model of incisional pain. Pain 64, 493–501. Burbassi, S., Sengupta, R., Meucci, O., 2010. Alterations of CXCR4 function in m-opioid receptor-deficient glia. Eur. J. Neurosci. 32, 1278–1288. ˜ ero, D., Campillo, A., Ce´le´rier, E., Romero, A., Puig, M.M., 2009. Pronociceptive Caban effects of remifentanil in a mouse model of postsurgical pain: effect of a second surgery. Anesthesiology 111, 1334–1345. ˜ ero, D., Romero, A., Garcı´a-Nogales, P., Puig, M.M., 2011. Campillo, A., Caban Delayed postoperative latent pain sensitization revealed by the systemic administration of opioid antagonists in mice. Eur. J. Pharmacol. 25, 89–96. ˜ ero, D., Garcı´a-Nogales, P., Romero, A., Campillo, A., Gonza´lez-Cuello, A., Caban Milane´s, M.V., Laorden, M.L., Puig, M.M., 2010. Increased spinal dynorphin levels and phospho-extracellular signal-regulated kinases 1 and 2 and c-Fos immunoreactivity after surgery under remifentanil anesthesia in mice. Mol. Pharmacol. 77, 185–194. ˜ ero, D., Puig, M.M., 2006. OpioidCe´le´rier, E., Gonza´lez, J.R., Maldonado, R., Caban induced hyperalgesia in a murine model of postoperative pain: role of nitric oxide generated from the inducible nitric oxide synthase. Anesthesiology 104, 546–555. Ce´le´rier, E., Laulin, J.P., Corcuff, J.B., Le Moal, M., Simonnet, G., 2001. Progressive enhancement of delayed hyperalgesia induced by repeated heroin administration: a sensitization process. J. Neurosci. 2, 4074–4080. Chaplan, S.R., Bach, F.W., Pogrel, J.W., Chung, J.M., Yaksh, T.L., 1994. Quantitative assessment of tactile allodynia in the rat paw. J. Neurosci. Methods 53, 55–63. Devor, M., Wall, P.D., 1990. Cross-excitation in dorsal root ganglia of nerve-injured and intact rats. J. Neurophysiol. 64, 1733–1746. Dublin, P., Hanani, M., 2007. Satellite glial cells in sensory ganglia: their possible contribution to inflammatory pain. Brain Behav. Immun. 21, 592–598. Garrido, E., Pe´rez-Garcı´a, C., Alguacil, L.F., Dı´ez-Ferna´ndez, C., 2005. The alpha2adrenoceptor antagonist yohimbine reduces glial fibrillary acidic protein upregulation induced by chronic morphine administration. Neurosci. Lett. 29, 141–144. Hay, J.L., White, J.M., Bochner, F., Somogyi, A.A., Semple, T.J., Rounsefell, B., 2009. Hyperalgesia in opioid-managed chronic pain and opioid-dependent patients. J. Pain 10, 316–322. Horvath, R.J., Landry, R.P., Romero-Sandoval, E.A., DeLeo, J.A., 2010. Morphine tolerance attenuates the resolution of postoperative pain and enhances spinal microglial p38 and extracellular receptor kinase phosphorylation. Neuroscience 25, 843–854. Hutchinson, M.R., Lewis, S.S., Coats, B.D., Rezvani, N., Zhang, Y., Wieseler, J.L., Somogyi, A.A., Yin, H., Maier, S.F., Rice, K.C., Watkins, L.R., 2010a. Possible involvement of toll-like receptor 4/myeloid differentiation factor-2 activity of opioid inactive isomers causes spinal proinflammation and related behavioral consequences. Neuroscience 167, 880–893. Hutchinson, M.R., Zhang, Y., Shridhar, M., Evans, J.H., Buchanan, M.M., Zhao, T.X., Slivka, P.F., Coats, B.D., Rezvani, N., Wieseler, J., Hughes, T.S., Landgraf, K.E., Chan, S., Fong, S., Phipps, S., Falke, J.J., Leinwand, L.A., Maier, S.F., Yin, H., Rice,

K.C., Watkins, L.R., 2010. Evidence that opioids may have toll-like receptor 4 and MD-2 effects. Brain Behav. Immun. 24, 83–95. Hutchinson, M.R., Zhang, Y., Brown, K., Coats, B.D., Shridhar, M., Sholar, P.W., Patel, S.J., Crysdale, N.Y., Harrison, J.A., Maier, S.F., Rice, K.C., Watkins, L.R., 2006. Non-stereoselective reversal of neuropathic pain by naloxone and naltrexone: involvement of toll-like receptor 4 (TLR4). Eur. J. Neurosci. 28, 20–29. Ito, N., Obata, H., Saito, S., 2009. Spinal microglial expression and mechanical hypersensitivity in a postoperative pain model: comparison with a neuropathic pain model. Anesthesiology 111, 640–648. Kao, S.C., Zhao, X., Lee, C.Y., Atianjoh, F.E., Gauda, E.B., Yaster, M., Tao, Y.X., 2012. Absence of m opioid receptor mRNA expression in astrocytes and microglia of rat spinal cord. Neuroreport 23, 378–384. Kehlet, H., Jensen, T.J., Woolf, C.J., 2006. Persistent post surgical pain: risk factors and prevention. Lancet 367, 1618–1625. Laboureyras, E., Chateauraynaud, J., Richebe´, P., Simonnet, G., 2009. Long-term pain vulnerability after surgery in rats: prevention by nefopam, an analgesic with antihyperalgesic properties. Anesth. Analg. 109, 623–631. Le Roy, C., Laboureyras, E., Gavello-Baudy, S., Chateauraynaud, J., Laulin, J.P., Simonnet, G., 2011. Endogenous opioids released during non-nociceptive environmental stress induce latent pain sensitization Via a NMDAdependent process. J. Pain 12, 1069–1079. Macrae, W.A., 2008. Chronic post surgical pain: ten years on. Br. J. Anaesth. 101, 77–86. Miyake, K., 2007. Innate immune sensing of pathogens and danger signals by cell surface Toll-like receptors. Semin. Immunol. 19, 3–10. ¨ Nickel, F.T., Seifert, F., Lanz, S., Maihofner, C., 2012. Mechanisms of neuropathic pain. Eur. Neuropsychopharmacol. 22, 81–91. Obata, H., Eisenach, J.C., Hussain, H., Bynum, T., Vincler, M., 2006. Spinal glial activation contributes to postoperative mechanical hypersensitivity in the rat. J. Pain 7, 816–822. Ohara, P.T., Vit, J.P., Bhargava, A., Romero, M., Sundberg, C., Charles, A.C., Jasmin, L., 2009. Gliopathic pain: when satellite glial cells go bad. Neuroscientist 15, 450–463. Peters, C.M., Eisenach, J.C., 2010. Contribution of the chemokine (C–C motif) ligand 2 (CCL2) to mechanical hypersensitivity after surgical incision in rats. Anesthesiology 112, 1250–1258. Pogatzki, E.M., Vandermeulen, E.P., Brennan, T.J., 2012. Effect of plantar local anesthetic injection on dorsal horn neuron activity and pain behaviors caused by incision. Pain 97, 151–161. Raghavendra, V., Tanga, F.Y., DeLeo, J.A., 2004. Complete freunds adjuvant-induced peripheral inflammation evokes glial activation and proinflammatory cytokine expression in the CNS. Eur. J. Neurosci. 20, 467–473. Richebe´, P., Rivat, C., Laulin, J.P., Maurette, P., Simonnet, G., 2005. Ketamine improves the management of exaggerated postoperative pain observed in perioperative fentanyl-treated rats. Anesthesiology 102, 421–428. Rivat, C., Laboureyras, E., Laulin, J.P., Le Roy, C., Richebe´, P., Simonnet, G., 2007. Non-nociceptive environmental stress induces hyperalgesia, not analgesia, in pain and opioid-experienced rats. Neuropsychopharmacology 32, 2217–2228. Rivat, C., Laulin, J.P., Corcuff, J.B., Ce´le rier, E., Pain, L., Simonnet, G., 2002. Fentanyl enhancement of carrageenan-induced long-lasting hyperalgesia in rats: prevention by the N-methyl-D-aspartate receptor antagonist ketamine. Anesthesiology 96, 381–391. ˜ ero, D., Gispert, J.D., Herance, J.R., Campillo, A., Puig, Romero, A., Rojas, S., Caban M.M., 2011. A 18F-fluorodeoxyglucose MicroPET imaging study to assess changes in brain glucose metabolism in a rat model of surgery-induced latent pain sensitization. Anesthesiology 115, 1072–1083. Romero-Sandoval, A., Chai, N., Nutile-McMenemy, N., Deleo, J.A., 2008. A comparison of spinal Iba1 and GFAP expression in rodent models of acute and chronic pain. Brain Res. 1219, 116–126. Romero-Sandoval, A., Eisenach, J.C., 2007. Spinal cannabinoid receptor type 2 activation reduces hypersensitivity and spinal cord glial activation after paw incision. Anesthesiology 106, 787–794. Silverman, S.M., 2009. Opioid induced hyperalgesia: clinical implications for the pain practitioner. Pain Physician 12, 679–684. Takeda, M., Takahashi, M., Matsumoto, S., 2009. Contribution of the activation of satellite glia in sensory ganglia to pathological pain. Neurosci. Biobehav. Rev. 33, 784–792. Tanga, F.Y., Nutile-McMenemy, N., DeLeo, J.A., 2005. The CNS role of Toll-like receptor 4 in innate neuroimmunity and painful neuropathy. Proc. Natl. Acad. Sci. USA 102, 5856–5861. Wen, Y.R., Tan, P.H., Cheng, J.K., Liu, Y.C., Ji, R.R., 2011. Microglia: a promising target for treating neuropathic and postoperative pain, and morphine tolerance. Formos. Med. Assoc. 110, 487–494. Wu, H.E., Sun, H.S., Cheng, C.W., Terashvili, M., Tseng, L.F., 2006. dextro-Naloxone or levo-naloxone reverses the attenuation of morphine antinociception induced by lipopolysaccharide in the mouse spinal cord via a non-opioid mechanism. Eur. J. Neurosci. 24, 2575–2580. Xie, W., Strong, J.A., Zhang, J.M., 2009. Early blockade of injured primary sensory afferents reduces glial cell activation in two rat neuropathic pain models. Neuroscience 160, 847–857.