GABAergic pathway in a rat model of chronic neuropathic pain: Modulation after intrathecal transplantation of a human neuronal cell line

Neuroscience Research 69 (2011) 111–120

Contents lists available at ScienceDirect

Neuroscience Research journal homepage: www.elsevier.com/locate/neures

GABAergic pathway in a rat model of chronic neuropathic pain: Modulation after intrathecal transplantation of a human neuronal cell line L. Vaysse a , J.C. Sol a , Y. Lazorthes a , M. Courtade-Saidi b , M.J. Eaton c , S. Jozan a,b,∗ a b c

Laboratory of «Approches expérimentales et Thérapeutiques des Douleurs Neuropathiques», 133 route de Narbonne, 31062 Toulouse, France Histology-Embryology, Medical School, Rangueil, 133 route de Narbonne, 31062, Toulouse, France VA Medical Research Center, Miami, Fl, USA

a r t i c l e

i n f o

Article history: Received 14 June 2010 Received in revised form 9 September 2010 Accepted 14 October 2010 Available online 21 October 2010 Keywords: Neuropathic pain hNT2 cell Cell transplant GABA GAD67 GAD65

a b s t r a c t Current understanding of chronic pain points a decrease in level of the inhibitory neurotransmitter GABA, in the spinal dorsal horn, leading to an imbalance between excitatory and inhibitory pathways. A subcloned derivative of the human NT2 cell line (hNT2.17) which, after neuronal differentiation, secretes different inhibitory neurotransmitters such as GABA and glycine has been recently isolated. In this study, we have investigated the effect of this new cell line on peripheral nerve injury induced by chronic constriction (CCI) and notably the effect on the cellular GABAergic pathway. Our data show that the decrease in GABA expression in the spinal dorsal horn of injured animals is concomitant with a decline of its synthetic enzyme GAD67-Ir and mRNA but not GAD65. Interestingly, in transplanted animals we observed a strong induction of GAD67 mRNA with one week after graft, which is followed by a recovery of GAD67 and GABA Ir. This effect paralleled a reduction of hindpaw hypersensitivity and thermal hyperalgesia induced by CCI. These results suggest that hNT2.17 GABA cells can modulate neuropathic pain after CCI certainly by minimizing the imbalance and restoring the cellular GABAergic pathway. © 2010 Elsevier Ireland Ltd and the Japan Neuroscience Society. All rights reserved.

1. Introduction Chronic neuropathic pain after nerve injury is a common symptom in clinical practice. Patients with chronic pain are subject to a greatly impaired quality of life and unfortunately, generally fail to respond to conventional pharmaceutical treatments. Neuropathic pain results from changes to both peripheral and central components of the pain transduction system that usually follow an initial injury leading to a malfunction of the network and an exaggerated sensation of pain (Campbell and Meyer, 2006; Kennedy, 2007). A possible mechanism for the maintenance of long term hyperexcitability and exaggerated sensory processing is impairment in spinal inhibitory function. -aminobutyric acid (GABA) is the main inhibitory neurotransmitteur in the nervous system, including laminae II of spinal dorsal horn where nociceptive nerve fibers terminate. Following spinal cord or peripheral nerve injury, levels of GABA are reduced along with modification in the level of its synthesizing enzyme glutamate decarboxylase (GAD) (CastroLopes et al., 1993; Ibuki et al., 1997; Eaton et al., 1998) and its

∗ Corresponding author at: Rangueil Medical School, Histology-Embryology Department and Laboratory of «Approches expérimentales et Thérapeutiques des Douleurs Neuropathiques», 133 route de Narbonne, 31062 Toulouse, France. Tel.: +33 562889025; fax: +33 562889027. E-mail address: [email protected] (S. Jozan).

receptors (Bhisitkul et al., 1990; Castro-Lopes et al., 1995; Moore et al., 2002). Moreover, pharmacological studies have shown that GABA receptor antagonists produce behavioral signs of tactile allodynia in rats whereas intrathecal injections of agonists tend to reduce pain behavior in pain injury models (Hwang and Yaksh, 1997; Malan et al., 2002; Gwak et al., 2006). Taken together theses studies indicate an important role for GABA in the expression of neuropathic pain, even if the exact mechanism remains unclear (Schoffnegger et al., 2006; Polgar and Todd, 2008). To restore the imbalance in inhibitory neurotransmitters, intrathecal grafting of cells that serve as biological pumps to provide various antinociceptive molecules shows promise in the treatment of chronic pain. Over the past 20 years, transplants of various preparations of primary adrenal chromaffin cells (Sagen et al., 1986, 1990; Siegan et al., 1996; Hentall et al., 2001; Kim et al., 2004; Sol et al., 2005), serotonergic (Eaton et al., 1997; Hains et al., 2001, 2003), GABAergic cells (Eaton et al., 1999b; Becerra et al., 2007; Mukhida et al., 2007), or recently human neural stem/progenitor cells (Mukhida et al., 2007) have been reported to provide a robust and enduring analgesic effect in various animal pain models. However, a serious limitation in the widespread application of this new therapy is still the lack of a safe, homogeneous, expandable cell source to supply the antinociceptive agents, although bioengineered cells may represent another promising source. (Eaton et al., 1999a, 2000; Cejas et al., 2000; Gajavelli et al., 2008; Cobacho et al., 2009).

0168-0102/$ – see front matter © 2010 Elsevier Ireland Ltd and the Japan Neuroscience Society. All rights reserved. doi:10.1016/j.neures.2010.10.006

112

L. Vaysse et al. / Neuroscience Research 69 (2011) 111–120

A subcloned derivative of the human NT2 cell line, the hNT2.17 cell line, which differentiates only to a GABAergic neuronal phenotype that secretes inhibitory neurotransmitters such as GABA and glycine has been recently isolated and has produced promising results after transplant in a central neuropathic pain model (Eaton et al., 2007; Wolfe et al., 2007). These cells could represent a new, homogenous and safe alternative for pain treatment. Once differentiated by retinoic acid treatment, the hNT2-derived cell lines do not proliferate and tumors or neoplastic masses have never been observed after transplantation of differentiated hNT neurons in the rat brain (Newman et al., 2005; Eaton et al., 2007). Moreover, preliminary human transplantation studies in the brain of stroke patients also demonstrated a lack of tumorigenicity and adverse effects of the hNT2 mother cell line (Nelson et al., 2002; Kondziolka et al., 2005). In the present study, we evaluated the effects of hNT2.17 GABA cell transplant in a rat model of neuropathic pain induced by unilateral chronic constriction injury (CCI) of the sciatic nerve. This model is associated with a loss of GABAergic inhibition in the superficial dorsal horn of the spinal cord and induced spontaneous and evoked behaviors associated with neuropathic pain (Bennett and Xie, 1988). The behavioral response to heat stimulus and hind limb weight bearing were observed and recorded during four weeks post-transplantation of pre-differentiated hNT2.17 cells (6 weeks post-surgery). To better understand the hNT2.17 GABA cell mechanism of action, we have also investigated the kinetics of GABA expression and of its two synthetic enzymes, GAD65 and GAD67 after CCI and cell transplant. We noted in injured animals, in parallel with a decline of GABA expression, a decrease in GAD67 expression but not GAD65 by immunohistochemistry and in situ hybridation. Moreover, we show that hNT2.17 grafts reversed the inhibition of GABA and GAD67 expression and this effect is concomitant with a reduction of pain behaviours. 2. Materials and methods 2.1. Animal experiments 2.1.1. Cell preparation for transplants The hNT2.17 cell line, subcloned from the human neuronal cell line NTera2/D1 (Eaton et al., 2007), was grown in DMEM/F12 medium (Gibco BRL) supplemented with 10% fetal bovine serum (FBS), 2 mM l-glutamine (Glu) and penicillin/streptomycin (PS, Gibco BRL). For neuronal cell differentiation, we used the cell aggregation method previously described by Cheung et al. (1999). hNT2.17 cells, near confluence, were seeded in 100 mm bacteriological grade Petri dishes in DMEM-HG (high glucose, Gibco BRL) supplemented with 10%FBS/Glu/PS. The day after, 10 ␮M retinoic acid (RA, Sigma, France) was added and the medium was changed every two days, for two weeks. After RA treatment, cells were removed and plated onto 100 mm tissue culture dishes, pre-coated with poly-d-lysine/laminin in supplemented DMEM-HG for inhibitors treatment (1 ␮M cytosine d-arabinofuranoside and 10 ␮M uridine) (Eaton et al., 2007). After 7 days, to separate differentiated hNT2.17 cells from non-neural cells, cells were briefly exposed to trysin/EDTA and neuronal cells were dislodged by gently striking the plate. Purified neuronal cells were replated on tissue culture dishes, precoated with poly-d-lysine and mouse laminin, and maintained for 2 weeks in DMEM-HG with 10%FBS/Glu/PS. 2.1.2. Animals and surgical procedures Experiments were performed in adult male Sprague–Dawley rats (Elevage Janvier, Le Genest-St-Isle, France), weighing 225–250 g at the beginning of the experiment. All animals

were maintained and treated according to the guidelines of the International Association for the Study of pain (Zimmermann, 1983). All efforts were made to minimize the suffering and the number of animals used. The surgery to produce CCI was first described by Bennett and Xie (1988). Briefly, animals were deeply anaesthetized with pentobarbital (50 mg/kg). The right common sciatic nerve was exposed at the level of the middle thigh by blunt dissection through the biceps femoris. Proximal to the nerve’s trifurcation, 5–7 mm of nerve was freed of adhering tissue and four ligatures (4–0 chromic gut) were tied loosely around with about 1 mm spacing. Care was taken to tie the ligatures so that the diameter of the nerve was barely constricted. The incision was closed in layers and rats were allowed to recover in their normal environment. Two weeks after injury, only animals that demonstrated a vigorous mechanical and thermal hypersensitivity were used for the rest of the experiment and were randomly assigned to the different groups. 2.1.3. Cell transplants Transplanted animals received, two weeks after injury a lumbar intrathecal cell graft with either hNT2.17 cells, differentiated for two weeks in vitro, or non-viable hNT2.17 cells. A third group of animals did not receive any cell transplant after CCI and a fourth group received neither CCI nor transplants and served as the control group. Cells were prepared immediately before each transplant to have optimal cell viability (more than 98%) as described by Eaton et al. (2007). Briefly, cells were gently dissociated from tissues culture plate with sterile 0.5 mM EDTA (Gibco BRL). Viability and cell counts were assessed by trypan blue exclusion and cells pellet was suspended in Ca2+ –Mg2+ free Hank’s buffered saline solution (CMF-HBSS, Gibco BRL) to make a cell suspension at 105 cells/␮l. Following a partial laminectomy (T12-L1 vertebral level) and a small puncture of the dura, cells, 106 cells were injected into the rat spinal subarachnoid space via a polyethylene tubing (PE-10) carefully advanced rostrally to 1 cm above the dural incision (lumbar spinal segment L4–L6) as described elsewhere (Sol et al., 2005). Following all intraspinal surgical procedures, the overlying muscle and the skin were sutured. All animals from each group were immunosuppressed by cyclosporin A, injected daily intraperitoneally, from 1 day before transplantation until 2 weeks after graft as previously described (Sol et al., 2005). Non-viable hNT2.17 cells were prepared by incubating 1 million cells in sterile water. Cell debris were then centrifugated, resuspended in 10 ␮l CMF-HBSS and injected into the subarachnoid space as described above. 2.1.4. Behavioral testing Animals were acclimated and trained on all behavioral tests before the study. Behavioral testing to detect signs of hindpaw hypersensitivity and thermal hyperalgesia was then performed on all animal (n = 59) on several occasions: one week and one day before CCI (baseline), two weeks post lesion to validate the CCI model and weekly during four weeks after cell transplantation. The animals were assigned to the different experiments: 8 animals from each group (control, CCI, CCI + hNT2.17 cells and CCI + non-viable cells) were keep until the end of the experiment for the kinetic of pain behavior, 24 others animals from the different groups (control n = 3, CCI n = 12, CCI + hNT2.17 cells n = 9) were sacrificed at different time points and used for the study of GABA and GAD expression. Three other animals from the CCI + hNT2.17 cell groups were used for cell graft survival study. 2.1.4.1. Hindpaw hypersensitivity. Hind limb weight bearing to evaluate hindpaw hypersensitivity was assessed using an incapacitance tester (Bioseb, France) that independently measures the

L. Vaysse et al. / Neuroscience Research 69 (2011) 111–120

weight in grams that the animal distributes to each hind paw. Control animals distribute their weight equally on both hind paws, while neuropathic pain causes the animals to favor the normal limb. The equipment was set to average the weight over 5 s, five measurements were taken and the result was expressed as the mean of weight ratio between injured and non-injured paw. 2.1.4.2. Thermal hyperalgesia. The response to a heat stimulus was tested with the Plantar test based on the thermal stimulus according to the method of Hargreaves et al. (1988) (Bioseb, France). Animals were placed in a clear plexiglass box on an elevated plexiglass floor. Animals were allowed to acclimate for approximately 5 min. A constant intensity, radiant heat source was located at the midplantar area of the hind paw and the time in seconds, taken by the animal to withdraw the leg, was automatically recorded. For each animal, both hind limb paws (intact and operated) were tested alternately at 5 min intervals. The median value of the withdrawal latency of five tests was calculated for each paw and the result was expressed as the withdrawal latency ratio between injured and non-injured paw. 2.2. In vitro molecular experiments 2.2.1. Tissue processing and immunohistochemistry In order to visualize, GABA and GAD expression in the spinal cord, after behavioral testing, animals from each experimental group were deeply anesthetized (pentobarbital, 60 mg/kg i.p) at different times post surgery. Rats were then immediately perfused intracardially with 300 ml heparizined (5000 U/L) PBS (pH 7.4), followed by 500 ml of 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4). The spinal cords from thoracic to sacral region were carefully removed with intact dura and postfixed overnight in Formalin solution 10%, neutral buffered (Sigma, St Louis). Each sample was then dehydrated and paraffin embedded. To study GABA and GAD expression, spinal cords were cut in serial cross sections (5 ␮m thick) and hematoxylin–eosin-stained sections were prepared every twelve sections to ensure that they were taken from the same level in the cord for each animal studied. Only sections in the region from L4 to L6 were collected on slides (Capillary gap microscope slides, Dakocytomation, Trappes, France). Before processing for immunostaining, sections were deparaffinized with xylene and rehydrated through progressive dilutions of ethanol. Sections were then allowed to react with different primary antibodies for the night at 4 ◦ C using optimal dilutions: mouse monoclonal anti-GABA, (clone GB69, Sigma, 1/1750), anti-glutamic acid decarboxylase 67 (1/1500, Chemicon), anti-glutamic acid decarboxylase 65 (1/3000, chemicon). The staining was revealed with the LSAB 2 kit (Dakocytomation) using DAB for the peroxidase system. The sections were counterstained with Hemalun. Spinal cord sections of control and treated animals at various times after injury were always run concurrently. In particular, the development of the chromogen was held constant. For each group, three different animals and six sections per animal were analyzed at each time point and for each staining. To study cell graft survival, the lumbar–sacral spinal cords of three other animals from the CCI + hNT2.17 cells groups were removed four weeks after injection of cells and were longitudinally sectioned. Immunohistochemistry was performed as described above using an anti-human nuclei antibody (clone 235-1, Chemicon, 1/50) to stain only the transplanted human neuronal cells. 2.3. In situ hybridization analysis (ISH) 2.3.1. GAD67 riboprobe synthesis In order to obtain riboprobes for GAD67, we have isolated total RNA from rat hippocampus which has been described

113

to express this enzyme (Stone et al., 1999). Reverse transcriptase and PCR experiments were conducted as previously described (Jozan et al., 2007), except that PCR amplification was conducted at 58 ◦ C using complementary oligonucleotides to GAD67 sequence. Oligonucleotide sequences to amplify GAD67 gene was synthesized by Sigma Proligo (Evry, France): forward primer: 5 -AACCTCCTCGAACGCGGGA-3 ; reverse primer: 5 -GTCAACCAGGATCTGCTCCA-3 . These primers have been chosen to amplify a sequence specific of GAD67 that not cross react with the isoenzyme GAD65 sequence. The 484 bp GAD67 cDNA obtained by RT-PCR from adult rat hippocampal tissue was ligated into the pGEM-T easy vector (Promega, Charbonnieres, France) after purification. Antisense or sense templates were generated respectively by SpeI and SAC II linearization of this vector. The labeled riboprobes were then synthetized using DIG RNA labeling kit (Roche) with respectively T7 and SP6 polymerase as previously described (Jozan et al., 2007). 2.3.1.1. In situ hybridization. For hybridization experiment, the different slide treatments and all the steps were previously described (Jozan et al., 2007) except that the hybridization temperature was 58 ◦ C and the probe concentration 1.2 ␮g/ml. For the visualization step, if a good signal for GAD67 mRNA could be obtained in various specific regions of the brain using visualization with the NBT/BCIP substrata (Roche Diagnostic, Meylan, France); in the dorsal horn of the spinal cord staining was too weak for quantification. In order to obtain optimal results, we used the HNPP fluorescent /Fast red TR (Roche Diagnostic) system as an amplifier with observation by fluorescence microscopy at 560 nm. For each group, three different animals and six sections per animal were analyzed at each time point. 2.3.2. Microscopic analysis Microscopic images were acquired with an Olympus microscope BX60 equipped with a C-5050200M digital camera. Images to be compared were acquired at the same illumination intensity and regions of interest (laminae I–III) were then outlined on the saved images. The GABA immuno-positive neurons or the GAD67 mRNA positive neurons in laminae I–III were scored by three operators in blinded procedure and the mean values were considered. 2.4. Statistical analysis Statistical analysis was performed for behavior tests, GABA immunohistochemistry and GAD67 hybridization in situ by a oneway ANOVA followed by a Bonferroni Multiple’s Comparison test and the data were expressed as the mean ± standard error of the mean (SEM). Results with p-value <0.05 were considered as significant. 3. Results 3.1. Effect of hNT2.17 cell transplantation on pain behaviour To study the effect of hNT2.17 cells after differentiation, the cells have been transplanted into the lumbar subarachnoid space two weeks after CCI. Just before transplantation, they were always analyzed by immuno-cytochemistry to confirm their differentiation into a neural phenotype (Eaton et al., 2007). After transplantation, to evaluate the effect of these cells on neuropathic pain, the behavioural response to hind limb weight bearing (Fig. 1A) and to heat stimulus (Fig. 1B), was observed and recorded during 6 weeks post CCI and 4 weeks post-transplantation.

114

L. Vaysse et al. / Neuroscience Research 69 (2011) 111–120

0.65 ± 0.04 whereas in control animals without surgery (control) no difference was observed in the weight bearing between hind limbs (weight ratio 0.98 ± 0.05). However, 7 days after hNT2.17 cells were transplanted near the lumbar spinal cord of CCI animals, the sensitivity of the ligated hind paw decreased considerably (weight ratio 0.90 ± 0.09) and the effect remained for the rest of the experiment (weight ratio 1.02 ± 0.04 four weeks after transplant). Transplant of differentiated hNT2.17 cells after CCI significantly reduced hindpaw hypersensitivity, whereas grafts of non-viable cells did not affect the development of neuropathic pain (p < 0.01, CCI + hNT2.17 cells group versus CCI + non-viable hNT2.17 cells group). 3.3. Thermal hyperalgesia after cell transplant

Fig. 1. Sensory behaviors after transplantation of hNT2.17 GABA cells in CCI model. Animals were either left unoperated (control) or underwent CCI (CCI). Two weeks after CCI, some rats were transplanted with hNT2.17 GABA cells (CCI + hNT2.17 cells) or with non-viable cells as a control (CCI + non-viable hNT2.17 cells). Animals were tested before and after CCI, and once a week after cell transplantation for hindpaw hypersensitivity (A) and thermal hyperalgesia (B). Data represent the mean value ± SEM of the ratio between injured and non-injured paw (n = 8 animals in each group). Values were analyzed by a one way ANOVA test followed by a Bonferroni’s multiple comparison test. Asterisks indicate when the group with differentiated hNT2.17 differed significantly from the non-viable cells (p < 0.01). Note that CCI signs start spontaneously to disappear after 5 weeks.

3.2. Hindpaw sensitivity after cell transplants CCI of the sciatic nerve resulted in vigorous hindpaw hypersensitivity which peaked between 2 and 4 weeks after injury and did not recover by 6 weeks (Fig. 1A, CCI group). The mean of weight ratio between injured and non-injured paw at this time was

The measure of sensitivity to noxious heat in animals after nerve injury and cell transplant is shown in Fig. 1B. In the animals with CCI, withdrawal latencies of the ipsilateral hindpaw to the radiant heat stimulus were considerably reduced compared to those of the contralateral hindpaw during all the experiment. Two weeks after CCI, the mean withdrawal latency ratio between injured and non-injured paw is 0.65 ± 0.1, whereas control animal displayed a similar sensitivity in each hind paw (withdrawal latency ratio of 0.98 ± 0.1). In the transplanted group, the thermal hypersensitivity in the ipsilateral hind paw completely disappeared one week after transplant of the differentiated hNT2.17 cells (withdrawal latency ratio of 0.97 ± 0.04), and no significant difference could be observed with the control group (p > 0.05). The high beneficial effect of the cell graft (p < 0.01, CCI + hNT2.17 cells group versus CCI) remained for the rest of the experiment (4 weeks after cell transplant), whereas transplantation of non-viable cells had no effect on the hypersensitivity induced by CCI (CCI + non-viable hNT2.17 cells group). The evolution of this last group remained identical with this of the CCI group (p > 0.05) for all the experiment. 3.4. hNT2.17 cells survival and phenotype in vivo Following behavioural testing, spinal cords were examined 4 weeks after transplant (6 weeks after CCI) to study the viability and the phenotype of the implanted cells (Fig. 2). Histological examination of spinal cord longitudinal section (from the beginning of the lumbar cord to the sacral level) of hNT2.17 transplanted animals demonstrated the presence of many viable cells surviving mainly near the injection site (L4–L6) in all animals. Grafted cells were observed as dense clusters around blood vessels in the subarachnoid space (Fig. 2, HE). All transplanted cells (hNu positive cells), still expressed the neurotransmitter GABA four weeks after transplant, whereas the bordering endothelial cells were not stained (Fig. 2, GABA and hNu). GABA cells were not detected in any of the control animals injected with non-viable cells (data not shown).

Fig. 2. Characterization of hNT2.17 GABA cells in the subarachnoid space 4 weeks after transplantation. Differentiated hNT2.17 GABA cells were injected into the subarachnoid space 2 weeks after CCI lesion. About 4 weeks after cell transplant, animals (n = 3) were sacrificed and serial longitudinal spinal cord sections (6 per animal) were stained with hematoxylin–eosin (HE) or examined by immunohistochemistry (DAB/peroxidase) for the human cell nuclei specific marker (hNu) or the neurotransmitter GABA (GABA). Scale bar: 50 ␮m.

L. Vaysse et al. / Neuroscience Research 69 (2011) 111–120

115

Fig. 3. GABA expression in the dorsal horn of the lumbar spinal cord after CCI and cell transplants. Immunohistochemistry for GABA reactivity was conducted, in parallel with behavioural testing, on serial cross sections of the lumbar spinal cord. (A) Representative photomicrographs of GABA expression respectively in the ipsilateral and contralateral side of the dorsal horns of control animals (A1 and A4), of CCI animals after three weeks injury (A2 and A5) and of animals three weeks after injury and one week after hNT2.17 cell transplants (A3 and A6). Laminae I–III are shown by dotted lines. Insets presented for ipsilateral side show a high magnification (×100) of a representative area in laminae I–III for each treatment. Scale bar: 200 ␮m. (B) Quantification of GABA expression variations in laminae I–III of the ipsilateral and contralateral side. For each group (control, CCI and CCI + hNT2.17 cells), three different animals were analyzed at each time point (n = 3) and the number of GABA stained neurons were evaluated. The data represent the mean value ± SEM. One-way ANOVA test followed by a Bonferroni’s multiple comparison tests were performed for the statistical analysis. Various symbols indicate when GABA-Ir cell number is significantly different: between control animals and various treatments (§ p < 0.001), between CCI and CCI + hNT2.17 cells for the ipsilateral side ( p < 0.005).

4. Temporal study of GABA pathway in CCI and CCI transplanted animals In parallel with behavioural assessment, we investigated the cellular changes of the GABAergic pathway in the dorsal horns of the spinal cord after CCI and the effect of the hNT2.17 cell transplant on these alterations. 4.1. GABA profile In order to observe the variation in GABA expression, transversal lumbar sections of the spinal cords were periodically analyzed by immunohistochemistry and the number of GABA-Ir cells was quantified (Fig. 3). In intact animals, GABA-Ir staining was observed outside the dorsal horn but a strong staining was only obtained in Rexed laminae I–III (Fig. 3A1 and A4). In this region, the GABA-Ir cells were small and rounded, and the neuropil were densely stained (high magnification in Fig. 3A1). For quantification, only rounded nucleated immunoreactive cells from layers I to III were counted. The number of GABA-Ir cells in control animals was evaluated at

186 ± 14 (Fig. 3B) with no significant difference between the two sides. In CCI animals, we observed a marked decrease of GABA expression in the both side of the dorsal horn (Fig. 3A2 and A5). This decrease is simultaneously observed in the soma and the neuropil (high magnification in Fig. 3A2). The maximum loss of GABA-Ir cells was observed at 3 and 4 weeks after injury especially in ipsilateral side (respectively 77 ± 13 and 66 ± 14 GABA-Ir cells, p < 0.001 versus control). Then a spontaneous slow recovery occurred at six weeks. In transplanted animals, at the same time point (3 and 4 weeks after injury), we observed a higher number of GABA-Ir cells, respectively 93 ± 11 and 103 ± 16 for the ipsilateral side, compared to the CCI group (p < 0.05 at four weeks) (Fig. 3A3 and B). At six weeks, as CCI animals started to recover, the difference between the various groups was less pronounced. 4.2. GAD67 and GAD65 profiles To better understand the change in GABA expression in neuropathic pain and the influence of implanting hNT2.17 cells, we stud-

116

L. Vaysse et al. / Neuroscience Research 69 (2011) 111–120

Fig. 4. Representative photomicrographs of GAD67 and GAD65 expressions in the dorsal horn of the lumbar spinal cord after CCI and cell transplants. Immunohistochemistry for GAD67 or GAD65 reactivity was conducted on serial cross sections (7 ␮m) of the dorsal horn of the lumbar spinal cord. (A) GAD 67 expression in the ipsilateral side of the dorsal horns of control animals (A1), of CCI animals after 3 weeks and 6 weeks of injury (respectively A2 and A4), and of animals three weeks after injury and one week after hNT2.17 cell transplant (A3) or 6 weeks after injury and 4 weeks after cell transplant (A5). (B) GAD 65 expression in the ipsilateral side of the dorsal horns of control animals (B1), of CCI animals after 3 weeks of injury (B2), and of animals three weeks after injury and one week after hNT2.17 cell transplant (B3). Six sections of three different animals were analyzed by group; laminae I–III were shown by dotted lines in all the figures. Scale bar: 200 ␮m. Insets show a high magnification (×100) of a representative area in laminae I–III.

ied the expression profiles of the two enzymes involved in GABA secretion. Profile staining of the spinal cord for the isoenzymes GAD67 and GAD65 is presented in Fig. 4. GAD67 staining for the control group was located solely in the dorsal horns, with a strong distinct labelled band in laminae II and III (Fig. 4A1). This staining is located in the neuropil and the soma is only counterstained (high magnification in Fig. 4A1). After CCI, we noted a decrease in this GAD67 labelling in the ipsilateral and contralateral dorsal horns. This decrease appeared two weeks after CCI and was maximal around 3–4 weeks. At this time, only the external and internal sides of laminae II and III are still stained in the ipsilateral dorsal horn (Fig. 4A2). After six weeks, a spontaneous recovery of GAD67 expression is observed in these injured animals (Fig. 4A4). When animals had been transplanted with hNT2.17 cells, a stronger staining appeared in the neuropil in the dorsal horn in laminae II and III within one week (Fig. 4A3) which persists until six weeks (Fig. 4A5). For GAD65, a strong staining was observed in the whole dorsal and ventral horns. However, this staining, mainly in the neuropil (high magnification, Fig. 4B1), had the same intensity and distribution irrespective of the treatment and remained stable throughout the experiment (Fig. 4B). To examine more in detail the variation of GAD67 isoform, we analyzed GAD67 mRNA expression by in situ hybridization. Fig. 5A shows the staining in ipsilateral dorsal horn and Fig. 5B summarizes the number variation of positive GAD67 mRNA neurons in the ipsi and the contralateral side for each experimental condition. In control animals, we observed the same number of positive neurons (24 ± 3) in the two dorsal horns with a perinuclear staining (Fig. 5A1). Few neurons outside the laminae I–III were stained for

GAD67, but only those inside the manual drawing were counted (Fig. 5B). The fine analysis of GAD67 mRNA by in situ hybridization in CCI group showed no change in GAD67 mRNA level at two weeks, then a complete disappearance of this mRNA three weeks after injury especially in ipsilateral side (Fig. 5A2 and B, 1 ± 1 positive neurons, p < 0.001 versus the control group). Four weeks after injury, a spontaneous recovery has already started for the ipsilateral side (11 ± 4, Fig. 5B) and this increase reached the control level by six weeks for the ipsilateral side (21 ± 5, p > 0.05 versus control group, Fig. 5B). In the contralateral side, we observe also at 3 and 4 weeks a GAD67 mRNA decrease but slighter than in the ipsilateral side and a delay in its recovery is also observed compared to the ipsilateral side. In the same injured animals, but injected with hNT2.17 cells, the GAD67 mRNA was strongly induced notably in ipsilateral horn (Fig. 5A3 and B, p < 0.001 versus CCI animal). After one week transplant, the level is in fact higher than in the control group (38 ± 5, p < 0.001). Then, it decreased at four weeks to reach that of the control level and it maintained until the end of experiment (Fig. 5B). In the contralateral side, the same evolution could be observed but the magnitude is slighter and the kinetic slower than for the ipsilateral side. 5. Discussion In this study, we have shown that intrathecal grafting of differentiated hNT2.17 GABA cells, after CCI, reduced both the hindpaw hypersensitivity and the thermal hyperalgesia induced by this injury. The effect persisted throughout the duration of the

L. Vaysse et al. / Neuroscience Research 69 (2011) 111–120

117

Fig. 5. GAD67 mRNA expressions in dorsal horn of the lumbar spinal cord after CCI and cell transplants. In situ hybridization was conducted in serial cross-sections (10 ␮m) of lumbar spinal cord of rats. (A) Representative photomicrographs of GAD67 mRNA expression in ipsilateral dorsal horns of the different rat groups: control group (A1), CCI group after three weeks injury and (A2) CCI group after three week injury and one week hNT2.17 cell transplant (A3). Cerebellar cortex was taken as a positive control (antisense riboprobe) and shows an intense staining only in the two indicated layers, as expected for GAD67. In order to not overload the figure, the negative control (sens riboprobe) was not presented. Scale bar: 200 ␮m. (B) Quantification of GAD67 mRNA expression in the superficial dorsal horn of animals at various times after injury and cell transplant. For each group, three different animals were analyzed at each time point (n = 3) and the number of positive fluorescent neurons was evaluated for the ipsilateral and contralateral side. The data represent the mean value ± SEM. One-way ANOVA test followed by a Bonferroni’s multiple comparison tests were performed for the statistical analysis. Symbols indicate significant difference between control and indicated treatments (§ p < 0.001), between CCI and CCI + hNT2.17 cells groups for the ipsilateral side(*p < 0.001) and for the contralateral side (× p < 0.05).

experiment, whereas grafts of non-viable cells did not affect the development of neuropathic pain, showing that the alleviation of pain was linked to the presence of grafted hNT2.17 cells. The GABAergic transplanted cells survived close to the lumbar cord for more than four weeks after implantation. These results, after peripheral nerve injury, support findings of an other study showing the therapeutic efficiency of this GABAergic cells in recovery of sensorimotor function after central nerve injury (Eaton et al., 2007). However, the mechanism of action of GABAergic cell transplant in the treatment of neuropathic pain is still largely discussed. Current understanding of central and supraspinal mechanisms for

the induction and maintenance of chronic pain suggest a major role for an impairment of GABAergic inhibitory pathways even if it is not clear whether is due to a loss of neurons in the spinal cord (Polgar et al., 2003) or to a down regulation of GABA synthesis (Ibuki et al., 1997). It has been proposed that intraspinally transplanted cells that secrete GABA near the lumbar dorsal may decrease pain behaviors by restoring this imbalance (Eaton et al., 2007; Mukhida et al., 2007). To better understand the mechanism of action of the hNT2.17 cell grafts, we investigated molecular changes in GABA expression and its synthetic enzymes after CCI and the effect of the hNT2.17 cell

118

L. Vaysse et al. / Neuroscience Research 69 (2011) 111–120

transplant on these alterations. In parallel with chronic pain syndromes, we observed, in accordance with others studies, a marked decrease in GABA expression in the dorsal horn of injured animals and a rapid recovery after cell transplant (Ibuki et al., 1997; Eaton et al., 1999a). Interestingly, we show in this study that the inhibition of GABA expression is concomitant with a decrease in GAD67 expression and mRNA especially in ipsilateral dorsal horn of the spinal cord. The same effects with a lower intensity and a slower kinetic were also noted in the contralateral side of injured animals. The results observed for the contralateral side are certainly due to “the mirror-image pain” often observed in human clinic and on several animal models (Seltzer et al., 1990; Aloisi et al., 1993; Chacur et al., 2001). This mechanism seems to involve central neural and humoral mechanisms different from the ones of the ipsilateral side (Dongyue and Buwei, 2010). Many evidences support the claim that the mirror image of pain could be produced only when the extent of the original pain is too important. In accordance with our results, it has been described as less intense and with a later occurrence than the ipsilateral pain. The strong decrease of GAD67 mRNA, 3 weeks after injury, is not observed in injured animals injected with hNT2.17 cells. In these animals, GAD67 mRNA were strongly induced one week after graft to levels well above controls and was followed by a recovery of GAD67 Ir and GABA expression showing that hNT2.17 cells could act directly on the GABAergic pathway of the dorsal horn. We hypothesize that hNT2 cells secrete GABA neurotransmitter or other active molecules in the LCS of the subarachnoid space which go through the blood–nervous system barrier and can act directly on the inter-neurons in the ipsilateral dorsal horn of the spinal cord. The slight quantities, released are apparently sufficient to act on various receptors producing certainly synaptic inhibition of primary afferent terminals and inhibiting the release of excitatory neurotransmitters to restore the imbalance. The rapid and strong reversion of GAD67 and GABA expression after hNT2.17 cell transplant and the natural increase observed 6 weeks after injury are in favour of a down regulation of the GABA inhibitory pathways after CCI more than a real loss of neurons by apoptosis as proposed (Whiteside and Munglani, 2001). In this work, in parallel with changes in GAD67 expression, we did not observe any change in GAD65-Ir which could explain that there is still a large number of GABA Ir neurons 3 weeks after injury, even if GAD67 level is very low. GABA synthesized by GAD65 and GAD67 does not seem to have the same function in the cell and our results suggest that it is the GAD67 pathway which is involved in neuropathic pain. The two isoforms: GAD65 and GAD67 are encoded by two separate genes in the adult nervous system (Bu et al., 1992) and differ in their intracellular distributions, their interactions with the cofactor pyridoxal phosphate and their pattern of expression in the central nervous system (Erlander et al., 1991). Both isoforms are present in most GABA containing neurons, but GAD65 appears to be targeted to membrane or nerve ending and synthesize GABA for vesicular release, whereas GAD67 is more widely distributed in cells and might preferentially synthesize cytoplasmic GABA (Soghomonian and Martin, 1998). The role of the two GADs to support the simple process of GABA synthesis is still not really explained, but seems critical as altered GABA pathways seems to play a key role in different human central nervous pathologies (Fatemi et al., 2002; Blatt, 2005). In spinal cord, few authors have looked at the involvement of each isoform in alterations of GABA synthesis in chronic neuropathic pain and results have been rather contradictory. In a cat model of complete thoracic spinal cord transection, an increase in GAD67, but no change in GAD65 protein and mRNA in the lumbar spinal cord were observed three months after injury (Tillakaratne et al., 2000). In a rat model of CCI and spared nerve injury, Moore et al. (2002) obtained a reduction in GAD65 levels measured by

Western blotting or immunohistochemical in the ipsilateral dorsal horn of animals two weeks after CCI, whereas GAD67 was little altered. In the same CCI model, Eaton et al. (1998) observed at three days after the ligation, a decrease in GAD67-Ir and then an increase which peaked by eight weeks to a significantly higher level than in controls. Unfortunately, these authors did not examine changes in GAD65. These studies underline the complexity of neuropathic pain mechanisms and that the time chosen following nerve injury to study GADs expression is crucial. It will be necessary to make a longer kinetic of the two enzymes to really conclude on the involvement of each one after nerve injury. However, GAD67 involvement in neuropathic pain is also demonstrated by the findings that transfer of GAD67 genes by viral technology reduces neuropathic pain (Liu et al., 2008). HNT2.17 cell transplant could reverse, in parallel of pain behavior, cellular GAD67 expression which seems to accelerate secretion of GABA and permits to repair and to recover functional GABAergic neurons. However, there is not a strict correlation between GABA and GAD level and the behavior. After 2 weeks of injury, the pain behavior induced by CCI is already maximal whereas the loss of GABA and GAD Ir cells are really significant 3 weeks after injury. Moreover, after 6 weeks, GABA and GAD levels are nearly the same in CCI and cell transplant group but behavior do not appear as robustly reversed in CCI group. Neuropathic pains are the result of different complex mechanisms leading to a dysfunction of the pain transduction system. Peripheral nerve injury is accompanied by transient local inflammation, which contributes to the initiation of neuropathic pain. At this stage, damaged nerves and activated glia cells release a variety of chemical mediators (interleukines, chemokines, TNF␣, excitatory amino acids, and ATP) which contribute to central sensitization by different mechanisms including increased primary afferent excitability and neurotransmitter release; potentiation of glutamate- mediated depolarization; and impaired local GABAergic inhibition. GABA level by itself can certainly not totally explain the pain behaviour and seems to be an important parameter in the maintenance of neuropathic pain more than in the instigation. This could explain the different kinetics between pain behaviours and GABA/GAD level in this study. Moreover, we cannot exclude that the beneficial effect of hNT2.17 cell graft is also due to various effects, as neuroprotective actions, since it has been shown that these cells express many factors after differentiation (5HT, neuropeptide Y, etc.) (Guillemain et al., 2000) and intrathecal injection may facilitate diffusion of these neuromediators. The reduction of pain behaviour obtained after cells graft seems in fact the result of a small number of cells. The percentage of cells survival observed 4 weeks after transplant in longitudinal cord section are very low compared to the number injected (106 ). Eaton et al. (2007) have estimated this percentage at only 0.3% after transplantation in the lumbar subarachnoid space. As they largely discussed in their paper, we think also that this percentage is underestimated, due to the difficulties to conserve and see grafted cells after the various technical step used in histology and to evaluate the number of clumped cells. However, as the GABA is delivered locally and during a long period by the cell graft, we believe that a tiny quantity of GABA or others substances is sufficient to induce a post-synaptic inhibition at the interneuron level in the spinal cord of injured animals. At least, in this study, we did not observed any beneficial effect with non-viable hNT2.17 cells, showing that the alleviation of pain during four weeks after cell transplant was inherent to the presence of hNT2.17 cells and not the result of unspecific factors. Furthers studies should be carried out in order to evaluate more precisely the number of cell survival and the molecular mechanisms involved in the alleviation of pain after hNT2.17 cell graft. If only a small number of cells are needed, future directions in

L. Vaysse et al. / Neuroscience Research 69 (2011) 111–120

cell therapy strategies will need to increase cell survival after graft in order to minimize the number of injected cells and reduce the possible side effect in human due to the injection of a large amount of cells. In conclusion, these results suggest that hNT2.17 cells may modulate pain generation after sciatic nerve constriction by acting on the GABAergic pathways and are in agreement with recent studies showing the therapeutic efficacy of these cells in another rodent neuropathic pain model (Eaton et al., 2007). Molecular mechanisms involved should be more precisely defined but intrathecal transplant of hNT2.17 GABA cells, which can be readily obtained in large quantities, may represent a promising strategy in the treatment of human neuropathic pain. Acknowledgement This study was supported by the Institut Pierre Fabre Médicament and by a grant from the ESSFN (European Society for Stereotactic and Functional Neurosurgery). We also want to thank Viviane Blanco and Cecile Pujol for their technical assistance. The authors declare that they have no financial or otherwise conflict of interest. References Aloisi, A.M., Porro, C.A., Cavazzuti, M., Baraldi, P., Carli, G., 1993. ’Mirror pain’ in the formalin test: behavioral and 2-deoxyglucose studies. Pain 55, 267–273. Becerra, G.D., Tatko, L.M., Pak, E.S., Murashov, A.K., Hoane, M.R., 2007. Transplantation of GABAergic neurons but not astrocytes induces recovery of sensorimotor function in the traumatically injured brain. Behav. Brain Res. 179, 118–125. Bennett, G.J., Xie, Y.K., 1988. A peripheral mononeuropathy in rat that produces disorders of pain sensation like those seen in man. Pain 33, 87–107. Bhisitkul, R.B., Kocsis, J.D., Gordon, T.R., Waxman, S.G., 1990. Trophic influence of the distal nerve segment on GABAA receptor expression in axotomized adult sensory neurons. Exp. Neurol. 109, 273–278. Blatt, G.J., 2005. GABAergic cerebellar system in autism: a neuropathological and developmental perspective. Int. Rev. Neurobiol. 71, 167–178. Bu, D.F., Erlander, M.G., Hitz, B.C., Tillakaratne, N.J., Kaufman, D.L., WagnerMcPherson, C.B., Evans, G.A., Tobin, A.J., 1992. Two human glutamate decarboxylases, 65-kDa GAD and 67-kDa GAD, are each encoded by a single gene. Proc. Natl. Acad. Sci. U.S.A. 89, 2115–2119. Campbell, J.N., Meyer, R.A., 2006. Mechanisms of neuropathic pain. Neuron 52, 77–92. Castro-Lopes, J.M., Malcangio, M., Pan, B.H., Bowery, N.G., 1995. Complex changes of GABAA and GABAB receptor binding in the spinal cord dorsal horn following peripheral inflammation or neurectomy. Brain Res. 679, 289–297. Castro-Lopes, J.M., Tavares, I., Coimbra, A., 1993. GABA decreases in the spinal cord dorsal horn after peripheral neurectomy. Brain Res. 620, 287–291. Cejas, P.J., Martinez, M., Karmally, S., McKillop, M., McKillop, J., Plunkett, J.A., Oudega, M., Eaton, M.J., 2000. Lumbar transplant of neurons genetically modified to secrete brain-derived neurotrophic factor attenuates allodynia and hyperalgesia after sciatic nerve constriction. Pain 86, 195–210. Chacur, M., Milligan, E.D., Gazda, L.S., Armstrong, C., Wang, H., Tracey, K.J., Maier, S.F., Watkins, L.R., 2001. A new model of sciatic inflammatory neuritis (SIN): induction of unilateral and bilateral mechanical allodynia following acute unilateral peri-sciatic immune activation in rats. Pain 94, 231–244. Cheung, W.M., Fu, W.Y., Hui, W.S., Ip, N.Y., 1999. Production of human CNS neurons from embryonal carcinoma cells using a cell aggregation method. Biotechniques 26, 946–948, 950–942, 954. Cobacho, N., Serrano, A.B., Casarejos, M.J., Mena, M.A., Paino, C.L., 2009. Use of transduced adipose tissue stromal cells as biologic minipumps to deliver levodopa for the treatment of neuropathic pain: possibilities and limitations. Cell Transplant. 18, 1341–1358. Dongyue, H., Buwei, Y., 2010. The mirror-image pain: an unclered phenomenon and its possible mechanism. Neurosc. Biobehav. Rev. 34, 528–532. Eaton, M.J., Frydel, B.R., Lopez, T.L., Nie, X.T., Huang, J., McKillop, J., Sagen, J., 2000. Generation and initial characterization of conditionally immortalized chromaffin cells. J. Cell Biochem. 79, 38–57. Eaton, M.J., Martinez, M.A., Karmally, S., 1999a. A single intrathecal injection of GABA permanently reverses neuropathic pain after nerve injury. Brain Res. 835, 334–339. Eaton, M.J., Plunkett, J.A., Karmally, S., Martinez, M.A., Montanez, K., 1998. Changes in GAD- and GABA- immunoreactivity in the spinal dorsal horn after peripheral nerve injury and promotion of recovery by lumbar transplant of immortalized serotonergic precursors. J. Chem. Neuroanat. 16, 57–72. Eaton, M.J., Plunkett, J.A., Martinez, M.A., Lopez, T., Karmally, S., Cejas, P., Whittemore, S.R., 1999b. Transplants of neuronal cells bioengineered to synthesize GABA alleviate chronic neuropathic pain. Cell Transplant. 8, 87–101.

119

Eaton, M.J., Santiago, D.I., Dancausse, H.A., Whittemore, S.R., 1997. Lumbar transplants of immortalized serotonergic neurons alleviate chronic neuropathic pain. Pain 72, 59–69. Eaton, M.J., Wolfe, S.Q., Martinez, M., Hernandez, M., Furst, C., Huang, J., Frydel, B.R., Gomez-Marin, O., 2007. Subarachnoid transplant of a human neuronal cell line attenuates chronic allodynia and hyperalgesia after excitotoxic spinal cord injury in the rat. J. Pain 8, 33–50. Erlander, M.G., Tillakaratne, N.J., Feldblum, S., Patel, N., Tobin, A.J., 1991. Two genes encode distinct glutamate decarboxylases. Neuron 7, 91–100. Fatemi, S.H., Halt, A.R., Stary, J.M., Kanodia, R., Schulz, S.C., Realmuto, G.R., 2002. Glutamic acid decarboxylase 65 and 67 kDa proteins are reduced in autistic parietal and cerebellar cortices. Biol. Psychiatry 52, 805–810. Gajavelli, S., Castellanos, D.A., Furmanski, O., Schiller, P.C., Sagen, J., 2008. Sustained analgesic peptide secretion and cell labeling using a novel genetic modification. Cell Transplant. 17, 445–455. Guillemain, I., Alonso, G., Patey, G., Privat, A., Chaudieu, I., 2000. Human NT2 neurons express a large variety of neurotransmission phenotypes in vitro. J. Comp. Neurol. 422, 380–395. Gwak, Y.S., Tan, H.Y., Nam, T.S., Paik, K.S., Hulsebosch, C.E., Leem, J.W., 2006. Activation of spinal GABA receptors attenuates chronic central neuropathic pain after spinal cord injury. J. Neurotrauma 23, 1111–1124. Hains, B.C., Johnson, K.M., Eaton, M.J., Willis, W.D., Hulsebosch, C.E., 2003. Serotonergic neural precursor cell grafts attenuate bilateral hyperexcitability of dorsal horn neurons after spinal hemisection in rat. Neuroscience 116, 1097–1110. Hains, B.C., Johnson, K.M., McAdoo, D.J., Eaton, M.J., Hulsebosch, C.E., 2001. Engraftment of serotonergic precursors enhances locomotor function and attenuates chronic central pain behavior following spinal hemisection injury in the rat. Exp. Neurol. 171, 361–378. Hargreaves, K., Dubner, R., Brown, F., Flores, C., Joris, J., 1988. A new and sensitive method for measuring thermal nociception in cutaneous hyperalgesia. Pain 32, 77–88. Hentall, I.D., Noga, B.R., Sagen, J., 2001. Spinal allografts of adrenal medulla block nociceptive facilitation in the dorsal horn. J. Neurophysiol. 85, 1788–1792. Hwang, J.H., Yaksh, T.L., 1997. The effect of spinal GABA receptor agonists on tactile allodynia in a surgically-induced neuropathic pain model in the rat. Pain 70, 15–22. Ibuki, T., Hama, A.T., Wang, X.T., Pappas, G.D., Sagen, J., 1997. Loss of GABAimmunoreactivity in the spinal dorsal horn of rats with peripheral nerve injury and promotion of recovery by adrenal medullary grafts. Neuroscience 76, 845–858. Jozan, S., Aziza, J., Chatelin, S., Evra, C., Courtade-Saidi, M., Parant, O., Sol, J.C., Zhou, H., Lazorthes, Y., 2007. Human fetal chromaffin cells: a potential tool for cell pain therapy. Exp. Neurol. 205, 525–535. Kennedy, J.D., 2007. Neuropathic pain: molecular complexity underlies continuing unmet medical need. J. Med. Chem. 50, 2547–2556. Kim, Y.M., Jeon, Y.H., Jin, G.C., Lim, J.O., Baek, W.Y., 2004. Immunoisolated chromaffin cells implanted into the subarachnoid space of rats reduce cold allodynia in a model of neuropathic pain: a novel application of microencapsulation technology. Artif. Organs 28, 1059–1066. Kondziolka, D., Steinberg, G.K., Wechsler, L., Meltzer, C.C., Elder, E., Gebel, J., Decesare, S., Jovin, T., Zafonte, R., Lebowitz, J., Flickinger, J.C., Tong, D., Marks, M.P., Jamieson, C., Luu, D., Bell-Stephens, T., Teraoka, J., 2005. Neurotransplantation for patients with subcortical motor stroke: a phase 2 randomized trial. J. Neurosurg. 103, 38–45. Liu, W., Liu, Z., Liu, L., Xiao, Z., Cao, X., Cao, Z., Xue, L., Miao, L., He, X., Li, W., 2008. A novel human foamy virus mediated gene transfer of GAD67 reduces neuropathic pain following spinal cord injury. Neurosci. Lett. 432, 13–18. Malan, T.P., Mata, H.P., Porreca, F., 2002. Spinal GABA(A) and GABA(B) receptor pharmacology in a rat model of neuropathic pain. Anesthesiology 96, 1161–1167. Moore, K.A., Kohno, T., Karchewski, L.A., Scholz, J., Baba, H., Woolf, C.J., 2002. Partial peripheral nerve injury promotes a selective loss of GABAergic inhibition in the superficial dorsal horn of the spinal cord. J. Neurosci. 22, 6724–6731. Mukhida, K., Mendez, I., McLeod, M., Kobayashi, N., Haughn, C., Milne, B., Baghbaderani, B., Sen, A., Behie, L.A., Hong, M., 2007. Spinal GABAergic transplants attenuate mechanical allodynia in a rat model of neuropathic pain. Stem Cells 25, 2874–2885. Nelson, P.T., Kondziolka, D., Wechsler, L., Goldstein, S., Gebel, J., DeCesare, S., Elder, E.M., Zhang, P.J., Jacobs, A., McGrogan, M., Lee, V.M., Trojanowski, J.Q., 2002. Clonal human (hNT) neuron grafts for stroke therapy: neuropathology in a patient 27 months after implantation. Am. J. Pathol. 160, 1201–1206. Newman, M.B., Misiuta, I., Willing, A.E., Zigova, T., Karl, R.C., Borlongan, C.V., Sanberg, P.R., 2005. Tumorigenicity issues of embryonic carcinoma-derived stem cells: relevance to surgical trials using NT2 and hNT neural cells. Stem Cells Dev. 14, 29–43. Polgar, E., Hughes, D.I., Riddell, J.S., Maxwell, D.J., Puskar, Z., Todd, A.J., 2003. Selective loss of spinal GABAergic or glycinergic neurons is not necessary for development of thermal hyperalgesia in the chronic constriction injury model of neuropathic pain. Pain 104, 229–239. Polgar, E., Todd, A.J., 2008. Tactile allodynia can occur in the spared nerve injury model in the rat without selective loss of GABA or GABA(A) receptors from synapses in laminae I-II of the ipsilateral spinal dorsal horn. Neuroscience 156, 193–202. Sagen, J., Pappas, G.D., Perlow, M.J., 1986. Adrenal medullary tissue transplants in the rat spinal cord reduce pain sensitivity. Brain Res. 384, 189–194. Sagen, J., Wang, H., Pappas, G.D., 1990. Adrenal medullary implants in the rat spinal cord reduce nociception in a chronic pain model. Pain 42, 69–79.

120

L. Vaysse et al. / Neuroscience Research 69 (2011) 111–120

Schoffnegger, D., Heinke, B., Sommer, C., Sandkuhler, J., 2006. Physiological properties of spinal lamina II GABAergic neurons in mice following peripheral nerve injury. J. Physiol. 577, 869–878. Seltzer, Z., Dubner, R., Shir, Y., 1990. A novel behavioral model of neuropathic pain disorders produced in rats by partial sciatic nerve injury. Pain 43, 205– 218. Siegan, J.B., Hama, A.T., Sagen, J., 1996. Alterations in rat spinal cord cGMP by peripheral nerve injury and adrenal medullary transplantation. Neurosci. Lett. 215, 49–52. Soghomonian, J.J., Martin, D.L., 1998. Two isoforms of glutamate decarboxylase: why? Trends Pharmacol. Sci. 19, 500–505. Sol, J.C., Li, R.Y., Sallerin, B., Jozan, S., Zhou, H., Lauwers-Cances, V., Tortosa, F., Chaynes, P., Mascott, C., Bastide, R., Lazorthes, Y., 2005. Intrathecal grafting of porcine chromaffin cells reduces formalin-evoked c-Fos expression in the rat spinal cord. Cell Transplant. 14, 353–365.

Stone, D.J., Walsh, J., Benes, F.M., 1999. Localization of cells preferentially expressing GAD(67) with negligible GAD(65) transcripts in the rat hippocampus. A double in situ hybridization study. Brain Res. Mol. Brain Res. 71, 201–209. Tillakaratne, N.J., Mouria, M., Ziv, N.B., Roy, R.R., Edgerton, V.R., Tobin, A.J., 2000. Increased expression of glutamate decarboxylase (GAD(67)) in feline lumbar spinal cord after complete thoracic spinal cord transection. J. Neurosci. Res. 60, 219–230. Whiteside, G.T., Munglani, R., 2001. Cell death in the superficial dorsal horn in a model of neuropathic pain. J. Neurosci. Res. 64, 168–173. Wolfe, S.Q., Garg, M., Cumberbatch, N.M., Furst, C., Martinez, M., Hernandez, M., Reimers, R., Berrocal, Y., Gomez-Marin, O., Eaton, M.J., 2007. Optimizing the transplant dose of a human neuronal cell line graft to treat SCI pain in the rat. Neurosci. Lett. 414, 121–125. Zimmermann, M., 1983. Ethical guidelines for investigations of experimental pain in conscious animals. Pain 16, 109–110.