PAINÒ 152 (2011) 2131–2137

www.elsevier.com/locate/pain

Nociceptive thresholds are controlled through spinal b2-subunit-containing nicotinic receptors

Ipek Yalcin a, Alexandre Charlet a,b, Matilde Cordero-Erausquin a, Luc-Henri Tessier a, Marina R. Picciotto c, ⇑ Rémy Schlichter a,b, Pierrick Poisbeau a,b, Marie-José Freund-Mercier a,b, Michel Barrot a, a Institut des Neurosciences Cellulaires et Intégratives, Centre National de la Recherche Scientifique, Strasbourg, France b Faculté des Sciences de la Vie, Université de Strasbourg, Strasbourg, France c Division of Molecular Psychiatry, Abraham Ribicoff Research Facilities, Connecticut Mental Health Center, Yale University School of Medicine, New Haven, CT, USA

Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article. article info abstract

Article history: Although drugs are known to modulate nociception, the role of endogenous acetylcholine in Received 13 October 2010 nociceptive processing remains unclear. In the current study, we evaluated the role of cholinergic trans- Received in revised form 8 April 2011 mission through spinal b2-subunit-containing nicotinic acetylcholine receptors in the control of nocicep- Accepted 8 May 2011 tive thresholds. We show that mechanical and thermal nociceptive thresholds are significantly lowered ⁄ ⁄ in b2 -knockout (KO) mice. Using nicotinic antagonists in these mice, we demonstrate that b2 -nAChRs are responsible for tonic inhibitory control of mechanical thresholds at the spinal level. We further Keywords: hypothesized that tonic b ⁄-nAChR control of mechanical nociceptive thresholds might implicate GAB- Nicotinic 2 Aergic transmission since spinal nAChR stimulation can enhance inhibitory transmission. Indeed, the b -nAChRs 2 ⁄ Acetylcholine GABAA antagonist bicuculline decreased the mechanical threshold in wild-type but not b2 -KO ⁄ ⁄ GABA mice, and the restored basal mechanical threshold in b2 -KO mice. Thus, b2 -nAChRs Nociception appeared to be necessary for GABAergic control of nociceptive information. As a consequence of this ⁄ defective inhibitory control, b2 -KO mice were also hyperresponsive to capsaicin-induced C-fiber stimu- ⁄ lation. Our results indicate that b2 -nAChRs are implicated in the recruitment of inhibitory control of nociception, as shown by delayed recovery from capsaicin-induced allodynia, potentiated nociceptive response to inflammation and neuropathy, and by the loss of high-frequency transcutaneous electrical ⁄ nerve stimulation (TENS)–induced analgesia in b2 -KO mice. As high-frequency TENS induces analgesia ⁄ through Ab-fiber recruitment, these data suggest that b2 -nAChRs may be critical for the gate control of nociceptive information by non-nociceptive sensory inputs. In conclusion, acetylcholine signaling ⁄ through b2 -nAChRs seems to be essential for setting nociceptive thresholds by controlling GABAergic inhibition in the spinal cord. Ó 2011 Published by Elsevier B.V. on behalf of International Association for the Study of Pain.

1. Introduction horn, with a predominance of the a4, b2, and a7 subunits [8]. These nAChR subunits are found on primary afferent terminals [17] and Neuronal nicotinic acetylcholine receptors (nAChRs) are ligand- intrinsic spinal interneurons [7]. Application of nicotinic gated channels formed of five subunits surrounding a central cat- results in enhanced neurotransmission in the dorsal horn of the ionic pore [9]. These receptors are widely expressed in the central spinal cord [13,36,30]. and peripheral nervous system, where they contribute to many Although the effect of exogenous cholinomimetic drugs, such as neuronal functions, including development and degeneration or , on nociceptive transmission in the spinal [44], cognitive functions [19], emotional and motivational pro- cord has been studied [6,8,31], the role of endogenous acetylcho- cesses [27,35], as well as nociception and pain processes [7,37]. line in nociceptive processing remains unclear. It has already been ⁄ The dorsal horn of the spinal cord plays an important role in the reported that nAChRs containing the b2-subunit (b2 -nAChR) con- reception, integration, and transmission of peripheral nociceptive tribute to nicotine-induced analgesia, as the ability of nicotine to ⁄ information. Most nAChR subunits are expressed in the dorsal attenuate thermal nociception is reduced in b2 -knockout (KO) mice [20]. However, acute nociceptive responses to heat in the

⇑ Corresponding author. Tel.: +33 3 68 85 14 50; fax: +33 3 88 61 33 47. tail-flick and hot-plate tests were not altered in these KO mice ⁄ E-mail address: [email protected] (M. Barrot). [20], suggesting that the endogenous stimulation of b2 -nAChRs

0304-3959/$36.00 Ó 2011 Published by Elsevier B.V. on behalf of International Association for the Study of Pain. doi:10.1016/j.pain.2011.05.022 2132 I. Yalcin et al. / PAINÒ 152 (2011) 2131–2137 might not be critical for setting these responses under basal condi- The latency to the first hind paw licking or withdrawal was taken tions. However, more recent pharmacological studies using nAC- as an index of nociceptive response. A cut-off time was set at 15 sec- ⁄ hRs antagonists suggest that acute blockade of b2 -nAChR activity onds to avoid damage to the paw. For the dynamic hot-plate (DHP) might induce thermal hyperalgesia [30]. test, mice were placed on the DHP (Bioseb, France) at 30 ± 0.1°C, In the current study, we used pharmacological and genetic ap- and the plate temperature was increased to 43°C with a computer- ⁄ proaches in KO mice to evaluate the role of spinal b2 -nAChR sub- controlled rate of 1°C/min as previously described [40]. During each units in the control of acute nociceptive sensitivity. Our results degree interval, we scored escape behavior (jumps). ⁄ reveal that b2 -nAChRs are critical for establishing mechanical and thermal nociceptive thresholds through endogenous choliner- 2.4. Sustained pain models gic and GABAergic networks. 2.4.1. Inflammatory pain 2. Methods Under isoflurane anesthesia, a volume of 10 ll of Complete Fre- und Adjuvant (CFA, Sigma-Aldrich, Saint Quentin Fallavier, France) 2.1. Animals was injected subcutaneously into the plantar surface of the right hindpaw using 50 ll Hamilton syringe with a 27-gauge needle. ⁄ Baseline threshold to mechanical stimuli was determined for each As a genetic approach to assess b2 -nAChR function, we used age- ⁄ animal before CFA injection and then measured 6, 24, and 48 hours matched, adult, wild-type or b2 -KO mice backcrossed for more than 10 generations onto the C57BL/6J background [28]. All mice were and 4 days after the injection using von Frey tests. male, except for the neuropathic allodynia study, for which one- third of the mice in each group were female. These experimental 2.4.2. Neuropathic pain mice were littermates from heterozygous breeding pairs and were Neuropathic pain was induced by placing a cuff around the right genotyped at weaning. The Pasteur Institute kindly provided breed- common sciatic nerve in mice as described previously [41]. Surgery ers through the Charles River mouse repository (St Germain sur was performed under /xylazine anesthesia (ketamine L’Arbresle France). Mice were group-housed four to five animals 17 mg/ml and xylazine 2.5 mg/ml; intraperitoneal [i.p.] adminis- per cage and maintained on a 12-hour light/dark cycle (lights on at tration, 4 ml/kg) (Centravet, Taden, France). The common branch 6 AM) with food and water available ad libitum. The dose–response of the right sciatic nerve was exposed and a 2-mm section of split studies were performed in C57BL/6J male mice (Charles River, St PE-20 polyethylene tubing (Harvard Apparatus, Les Ulis, France) Germain sur L’Arbresle France). The animal facilities are legally reg- was placed around it (Cuff group) [1]. Sham-operated mice under- istered for animal experimentation under Animal House Agreement went the same surgical procedure without cuff implantation B67-482-1/C67-482-1. All procedures were performed in accor- (Sham group). dance with the guidelines for animal experimentation of the Inter- national Association for the Study of Pain and the European 2.5. Transcutaneous electrical nerve stimulation Communities Council Directive 86/6609/EEC. Animals were lightly anesthetized by inhalation and 2.2. Nociceptive tests: mechanical sensitivity the right hindlimb was prepared for electrode placement by shav- ing and cleaning the area with ethanol. Two silver electrodes cov- We evaluated the mechanical threshold by using von Frey hairs ered with conducting gel were placed at the surface of the right (Bioseb, Chaville, France) as previously described [3,41,42]. Mice hindlimb. Pulse duration was set at 250 microseconds. The inten- were placed in clear Plexiglas boxes (7 cm  9cm 7 cm) on an ele- sity was set at a sensory level, just below muscle contraction level vated mesh screen, and were allowed to habituate for 15 minutes (6–12 mA). Transcutaneous electrical nerve stimulation (TENS) before testing. Filaments were applied to the plantar surface of each was delivered at very high frequency (2000 Hz) for 20 minutes. hindpaw in a series of ascending forces. We approached the filament This procedure has been shown previously to recruit sensory Ab fi- toward the plantar surface slowly until it slightly bent at contact. At bers selectively [18,21]. After either the TENS procedure or light that point the pressure was immediately removed [24]. Each fila- anesthesia only (controls, No TENS), mice were placed in clear ment was tested five times per paw and the threshold was defined Plexiglas boxes on the elevated mesh screen and allowed to re- as three or more withdrawals observed among the five consecutive cover for approximately 30 minutes before mechanical sensitivity trials. The results were expressed in grams.The filaments (Bioseb) testing with von Frey hairs. used in the studies were labeled as follows: 0.008, 0.02, 0.04, 0.07, 0.16, 0.4, 0.6, 1, 1.4, 2, 4, 6, 8, 10, and 15 grams. 2.6. Drugs

2.3. Nociceptive tests: thermal sensitivity chloride, N-n-decylnicotinium iodide (NDNI), methyllycaconitine from brownii seeds, muscimol 2.3.1. Hargreaves test hydrobromide, bicuculline methiodide, and capsaicin were ob- Mice were placed in clear Plexiglas boxes (7 cm  9cm 7 cm) tained from Sigma-Aldrich (St. Quentin Fallavier, France). a-Cono- on a glass surface and allowed to habituate for 15 minutes. The MII was purchased from Tocris. All drugs were dissolved in infrared beam of the radiant heat source (7370 Plantar Test; Ugo 0.9% sodium chloride (NaCl). Each drug and dose was tested in Basile, Comerio, Italy) was applied to the plantar surface of each an independent set of animals. hindpaw. A 15-second cut-off time was used. Three measures of Intraperitoneal (i.p.) injections were given in a volume of 5 ml/ paw withdrawal latency were taken and averaged for each hind- kg. Intrathecal (i.t.) injections in a volume of 10 ll were performed paw. To control for the reproducibility of results, mice were tested under gaseous anesthesia (halothane 3%) as previously described. on three consecutive days. Briefly, a 27-gauge needle connected to a 50-ll Hamilton syringe

was inserted between the L5 and L6 vertebrae, into the sub-arach- 2.3.2. Hot plate tests noidal space [16]. Placement of the needle was verified by the elic- For the conventional hot plate test, mice were placed on the hot itation of a tail flick movement. Mice were then tested for plate (Bioseb, France) with the temperature adjusted to 55 ± 0.1°C. mechanical sensitivity 15 minutes after the injection. I. Yalcin et al. / PAINÒ 152 (2011) 2131–2137 2133

Capsaicin (0.1 mmol/L or 10 mmol/L, dissolved in 50% ethanol [20], we found no difference in noxious heat sensitivity between ⁄ with 0.9% NaCl) was applied topically on the plantar surface of WT and b2 -KO mice in the hot plate test and the Hargreaves test, ⁄ both hindpaws with cotton-tipped applicators [22,40] under light confirming the lack of heat hyperalgesia in b2 -KO mice. However, gaseous anesthesia (halothane 3%). Mice were tested 15 minutes using the DHP test [40], we noticed a significant decrease in noci- ⁄ ⁄ after this application. ceptive threshold in b2 -KO mice (Fig. 1D). Indeed, b2 -KO mice showed an increased number of escape responses at 38°C, whereas ⁄ 2.7. Data analysis wild-type mice started to respond at 40°C (for b2 -KO mice: 2 v 10 = 87.68, P < .001, post hoc: P < .05 from 38° to 40°C when com- 2 When data met criteria for parametric tests (homogeneity of pared with 30°C; for wild-type mice: v 13 = 72.3, P < .001, post hoc: variances and normality of samples), significance was measured P < .05 from 40° to 43°C when compared with 30°C). It is important using multi-factor analysis of variance (ANOVA) followed by Dun- to note that we used a slow heating rate (1°C minÀ1) in the DHP can post hoc tests. If data did not fit conditions for parametric test- test. It has been previously shown by our laboratory [40] and in ing, we used the Kruskal–Wallis test or the multiple-factor analysis other studies [43] that slow heat ramps preferentially activate C-fi- of variance (ANOVA) Friedman test to compare multiple indepen- bers. The different results obtained with the Hargreaves and the dent or dependent group, respectively. The Wilcoxon test was used DHP tests might therefore be due to the involvement of different for paired group comparisons. The significance level was set at heat thermonociceptors (i.e., Ad vs C nociceptors). ⁄ P < .05, and data were expressed as mean ± SEM for graphs. Taken together, these findings indicate that b2 -nAChRs are crit- ical for establishing both mechanical and thermal nociceptive 3. Results thresholds under physiological conditions.

⁄ ⁄ 3.1. b2 -nAChR deficiency induced mechanical and thermal allodynia 3.2. Peripheral non–b2 -nAChRs are involved in the setting of mechanical nociceptive thresholds The mechanical nociceptive threshold has not been evaluated ⁄ ⁄ previously in b2 -KO mice. We therefore used von Frey hairs and To analyze the role of b2 -nAChRs in nociceptive thresholds fur- found that the mechanical threshold was significantly lower in ther, we focused our studies on mechanical sensitivity. We ob- ⁄ b2 -KO mice compared with their wild-type littermates (Fig. 1A, served that i.p. injection of the nonselective, peripherally active F1,22 = 71.43, P < .001). We also tested thermal sensitivity of the hexamethonium (5 mg/kg) resulted in mice using the conventional hot plate test at 55°C and the Har- mechanical allodynia. This effect was observed in both wild-type ⁄ greaves test (Fig. 1B and C). In agreement with previous report (P < .05) and b2 -KO mice (P < .05) (Fig. 2A, left panel). This result was confirmed when presenting the data as a percentage of the threshold measured before i.t. injection (Fig. 2A, right panel). As A Von-Frey hexamethonium crosses the blood brain barrier poorly, these data 8 suggest that peripheral nAChRs influence nociceptive thresholds, 6 WT KO 4 * * * L L A 8 100 2 R

Pressure (g) R 6 80 0 60 * 12 3 4 Time (days) 40 * 2 *

Pressure (g) Pressure 20 * threshold B Hot-plate C Hargreaves 0 %of nociceptive 15 6 WT KO WT KO NaCl (0.9%) Pretest 10 4 HX (5mg/kg, i.p) HX (5mg/kg,i.p) WT 5 2 KO B 8 8 150 Latency (s) Latency Latency (s) Latency % 0 0 6 6 100 123 50 * Time (days) 4 * 4 0

Dynamic hot-plate 2 (g) Pressure 2 *

D (g) Pressure * 25 + * 0 0 20 + + 0 0.5 2.5 5 WT KO 15 Dose (μg/10μL) * WT 10 Pretest Pretest * + KO HX (i.t) HX (5 μg/10μL, i.t) 5 Responses (nb) 0 Fig. 2. Peripheral and spinal influence of endogenous nAChRs. (A) Peripheral non- 30 32 34 36 38 40 42 44 ⁄ b2 -nAChRs control nociceptive transmission. Effect of the nonselective nAChR Degree (C) antagonist hexamethonium (HX, 5 mg/kg i.p.) on the mechanical threshold in wild- ⁄ type and b2 -KO mice (left panel). Data are also presented as percentage of the ⁄ ⁄ ⁄ Fig. 1. Mechanical and thermal allodynia in b2 - KO mice. In wild-type and b2 -KO threshold measured before i.t. injection (right panel). (B, left panel) Spinal b2 - littermate mice we measured (A) mechanical sensitivity using von Frey hairs nAChRs control nociceptive transmission. Dose–response of i.t. HX on the (L = left hindpaw, R = right hindpaw), (B) thermal hyperalgesia using the conven- mechanical threshold in C57BL/6J mice. (B, right panel) The highest effective doses ⁄ tional hot-plate at 55 ± 0.1°C and (C) the Hargreaves test, and (D) thermal allodynia of HX (5 lg/10 ll) were used to test the mechanical sensitivity of wild-type and b2 - using the dynamic hot-plate test. ⁄P < .05 when compared with the response at KO mice. (B, right inset panel) To account for differences in basal nociceptive starting temperatures in WT mice and +P < .05 when compared with the response at threshold, data are also presented as a percentage of the threshold measured before starting temperatures in KO mice. n = 11–13 per group. i.t. injection. n = 5–6 per group. 2134 I. Yalcin et al. / PAINÒ 152 (2011) 2131–2137

⁄ but that b2 -nAChRs do not play a major role in this peripheral le- i.t. delivery resulted in reduced mechanical threshold (a- vel of control. MII: P < .05 at 34.2 and 171 pg/10 ll; NDNI: P < .05 at 0.05, 0.5, and 5.58 lg/10 ll). Similar to hexamethonium, these selective antago- ⁄ 3.3. Spinal b2 -nAChRs are critical to establish mechanical nociceptive nists induced mechanical allodynia in wild-type mice (Fig. 3A, thresholds P < .05; Fig. 3B, P < .05), without affecting the mechanical thresh- ⁄ olds in b2 -KO mice. In contrast, i.t. administration of the selective ⁄ To assess the involvement of spinal nAChRs in mechanical sen- a7 -nAChR antagonist methyllycaconitine citrate (MLA) [2] sitivity, we delivered hexamethonium i.t. A dose-response study in (8.74 lg/10 ll) had no significant effect in wild-type mice or in ⁄ C57BL/6J mice (Fig. 2B) revealed that i.t. hexamethonium at doses b2 -KO mice (Fig. 3C). These results confirmed that, in the spinal ⁄ ⁄ of 2.5 or 5 lg/10 ll induced mechanical allodynia, indicating a role cord, b2 -nAChRs but not a7 -nAChRs are critical for the tonic nic- for spinal nAChRs in a tonic control of nociceptive responses to otinic control of mechanical nociceptive thresholds under basal mechanical stimulation (H = 20.32, P < .001). Using the effective conditions. dose inducing significant allodynia in C57BL/6J mice (5 lg/10 ll), ⁄ we observed mechanical allodynia in wild-type (P < .05) but not 3.4. Spinal b2 -nAChR control of mechanical nociceptive thresholds is ⁄ ⁄ b2 -KO mice (Fig. 2B). These data suggest that spinal b2 -nAChRs mediated through a GABAA-dependent mechanism are critical for the tonic nicotinic control of mechanical nociceptive thresholds. Inhibitory mechanisms in the dorsal horn of the spinal cord are ⁄ To test the involvement of spinal b2 -nAChRs, we also used the crucial for the processing of nociceptive information [34]. As spinal antagonists a-conotoxin MII and NDNI which selectively target nAChR stimulation by agonists can result in enhanced inhibitory ⁄ a3b2/a6b2-nAChRs [33] and a4b2-nAChRs [39], respectively. For transmission [29], we hypothesized that tonic b2 -nAChR control both drugs, dose–response studies in C57BL/6J mice showed that of mechanical nociceptive thresholds might involve GABAergic

transmission. To test this possibility, we used the GABAA bicuculline. A dose–response study in C57BL/6J mice showed that i.t. delivery of bicuculline resulted in mechanical allo- A 8 8 150 dynia (Fig. 4A, P < .05 at 0.0018, 0.018, and 0.037 lg/10 ll). Using

% the highest dose of bicuculline that is devoid of convulsive effects 6 6 100 (0.037 lg/10 ll), we observed a decrease in the mechanical thresh- 50 ⁄ 4 4 * old in wild-type mice (Fig. 4A, P < .05) but not in b2 -KO mice. * 0 This finding suggested that the tonic GABAergic control of noci- Pressure (g) Pressure (g) 2 * 2 ⁄ * ceptive thresholds was deficient in mice lacking b2 -nAChRs. To test whether GABA receptors were still functional in KO mice, 0 0 A 0 3.42 34.2 171 WT KO we used the agonist muscimol. A dose–response study in C57BL/ Dose (pg/10μL) 6J mice showed that i.t. delivery of muscimol induced analgesia Pretest Pretest α-con (i.t) α-con (34.2pg/10μL, i.t)

8 8 B 8 8 150 A 150 % 6 6 100 6 6 100 %

50 * 50 4 * 4 4 4 * * * 0 * Pressure (g) Pressure (g) Pressure (g) 2 * Pressure (g) 2 * 2 * * 2 *

0 0 0 0 0.05 0.5 5.58 WT KO 0 0.0018 0.018 0.037 WT KO Dose (μg/10μL) Dose (μg/10μL) Pretest Pretest Pretest Pretest Bicuculline NDNI (i.t) NDNI (5.58μg/10μL, i.t) Bicuculline (i.t) (0.037μμ g/10 L, i.t)

8 B * C 150 15 * 8 6 6 * 100 10 4 4 50 2 threshold 5 Pressure (g) Pressure Pressure (g) Pressure (g) 2 % of nociceptive 0 0 0 WT KO WT KO 0 0.058 0.116 0.28 0.58 WT KO Pretest Dose (μg/10μl) MLA (8.74μg/10μL, i.t) Pretest Pretest Muscimol (i.t) Muscimol ⁄ Fig. 3. Spinal b2 -nAChRs tonically control nociceptive transmission. (A and B, left (0.116μg/10μL, i.t) panels) Dose–response of i.t. a-conotoxin MII (a-con, antagonist of a3b2/a6b2- ⁄ nAChRs) (A, left panel) and N-n-decylnicotinium iodide (NDNI, antagonist of a4b2- Fig. 4. Spinal b2 -nAChRs are permissive for the tonic inhibitory control of nAChRs) (B, left panel) on the mechanical threshold in C57BL/6J mice. (A and B, nociceptive thresholds by GABAergic transmission. Dose–response of i.t. (A, left right panels) The highest effective doses of a-con (34.2 pg/10 ll) (A, right panel) panel) bicuculline and (B, left panel) muscimol in C57BL/6J mice. (A, right panel) and NDNI (5.58 lg/10 ll) (B, right panel) were used to test the mechanical The highest effective dose of the GABAA receptor antagonist bicuculline (0.037 lg/ ⁄ ⁄ sensitivity of wild-type and b2 -KO mice. (A and B, right insets panels) To account 10 ll, i.t.) was tested in wild-type and b2 -KO mice. To account for differences in for differences in basal nociceptive threshold, data are also presented as a basal nociceptive thresholds, the data are also presented as a percentage of the percentage of the threshold measured before i.t. injection. (C) Effect of the selective threshold measured before i.t. injection (A, right inset panel). (B, right panel) Effect ⁄ a7 -nAChR antagonist methyllycaconitine citrate (MLA, 8.74 lg/10 ll, i.t) in wild- of a low dose of the GABAA receptor agonist muscimol (0.116 lg/10 ll, i.t.; ⁄ type and b2 -KO mice (left panel). The data are also presented as a percentage of the ineffective in C57BL/6J mice) on mechanical nociceptive thresholds in wild-type ⁄ threshold measured before i.t. injection (right panel). n = 5–6 per group. and b2 -KO mice. n = 5-6 per group. I. Yalcin et al. / PAINÒ 152 (2011) 2131–2137 2135

(Fig. 4B, P < .05 at 0.28 and 0.58 lg/10 ll). We then used a sub- We then aimed to determine whether the nociceptive response ⁄ threshold dose, ineffective in wild-type mice (0.116 lg/10 ll), of b2 -KO mice would also be affected in sustained inflammatory or and observed that this low dose of muscimol fully restored the ba- neuropathic pain paradigms. In the test of CFA-induced inflamma- ⁄ sal mechanical nociceptive thresholds in b2 -KO mice (Fig. 4B, tion, decrease of mechanical thresholds was observed in both wild- ⁄ P < .05). Loss of tonic GABAergic control in b2 -KO mice was thus type (F4,32 = 10.51, P < .001, P < .001 for 6, 24, and 48 hours) and KO not related to deficient GABAA receptors but rather to a loss in their mice (F4,32 = 9.02, P < .001, P < 0.05 for 6, 24, and 48 hours) (Fig. 5C, ⁄ tonic stimulation. These results show that b2 -nAChRs are permis- left panel). Interestingly, 24 hours after the injection, the relative sive and necessary for the GABAergic control of nociceptive decrement in mechanical thresholds was even higher in KO mice information. (À95% ± 0.9%) than in wild-type mice (À77 ± 1.6%) (genotype x

treatment interaction: F1,8 = 33.21, P < .001; post hoc: CFA in wild- 3.5. Enhanced nociceptive transmission and decreased TENS-induced type or KO vs their baseline, P < .05; wild-type vs KO after CFA, ⁄ analgesia in b2 -nAChR deficient mice P < .05, Fig. 5C, right panel). For cuff-induced neuropathy, we ob- served no difference between male and female mice for the ⁄ If inhibitory control is lacking in the spinal cord of b2 -KO mice, mechanical threshold at baseline and after the induction of the these mice should be more sensitive to a mild recruitment of noci- neuropathy. Cuff implantation induced mechanical allodynia in ceptive fibers and less sensitive to the recruitment of inhibitory both wild-type (F4,48 = 16.16, P < .001, P < .001 at 4, 7, 11, and 14 control. Using a sub-threshold dose of capsaicin with no effect in days after the surgery) and KO mice F4,28 = 3.36, P < .001, P < .001 wild-type mice (application of 0.1 mmol/l of solution under the at 4, 7, 11, and 14 days after the surgery) (Fig. 5D, left panel). Seven hindpaw), we observed an increase in mechanical allodynia in days after the surgery, mechanical allodynia was potentiated in ⁄ ⁄ b2 -KO mice (Fig. 5A). Indeed, application of 0.1 mmol/l capsaicin b2 -KO mice when compared with wild-type mice (geno- ⁄ resulted in a 90% reduction in the nociceptive threshold in b2 - type  treatment interaction: F1,10 = 49.35, P < .001; post hoc: cuff KO mice, whereas capsaicin had no effect in wild-type mice at in wild-type or KO vs their baseline, P < .001; wild-type vs KO in the same concentration, suggesting that mild stimulation of cuff-implanted mice, P < .05, Fig. 5D, right panel) These results ob- peripheral nociceptors was no longer filtered at spinal level in tained from inflammatory and neuropathic pain models also sup- ⁄ b2 -KO mice. port the hypothesis that nociceptive transmission is enhanced in ⁄ Inhibitory control can be recruited by sustained nociceptive b2 -KO mice. stimulation [34]. We thus used a concentration of capsaicin (appli- The stimulation of Ab fibers can have an analgesic action cation of a 10 mmol/l solution on the hindpaw) that induced allo- through the recruitment of inhibitory control. We used TENS at dynia in wild-type mice, and studied the time-course of recovery. high frequency to stimulate the Ab fibers [18,21], and showed that ⁄ The allodynic effect of capsaicin lasted longer in b2 -KO mice it induced analgesia in C57BL/6J and wild-type mice but had no ef- ⁄ ⁄ (P < .05 until 60 hours when compared with pretest) than in their fect in b2 -KO mice (Fig. 6). The lack of b2 -nAChRs thus resulted in wild-type siblings (P < .05 until 31 hours when compared with pre- a deficit in the recruitment of inhibitory control of nociceptive test) (Fig. 5B). information.

AB8 100 8 80 6 6 60 4 Pretest WT 40 Capsaicin 4 KO Pressure (g) 2 *

%of nociceptive threshold 20 (0.1mM) * Pressure (g) 2 0 0 WT KO WT KO 0 0 60 120 180 240 300 24 36 48 60 Time (min) Time (hours) + +

C 8 100 D 8 100 80 6 80 6 60 Pretest 60 Pretest 4 4 40 * CFA 40 Cuff 2 threshold 2 + +++ * + ++ 20 Pressure (g) 20 Pressure (g) * %of nociceptive * * * 0 * * * 0 0 ** 0 WT KO

0 6 24 48 96 0 5 10 15 %of nociceptive threshold Time (hours) Time (days)

WT Saline WT CFA WT Sham WT Cuff KO Saline KO CFA KO Sham KO Cuff

⁄ Fig. 5. b2 -KO mice are hypersensitive to capsaicin, CFA-induced inflammation and cuff-induced neuropathy. (A, left panel) Effect of a sub-threshold concentration of ⁄ capsaicin (0.1 mM applied on both hindpaws) on the mechanical threshold in wild-type and b2 -KO mice. (A, right panel) To account for differences in basal nociceptive threshold, the data are also presented as a percentage of the threshold measured before capsaicin application. (B) Time-course of capsaicin effect at high dose (10 mmol/l ⁄ ⁄ applied on both hindpaws) in wild-type and b2 -KO mice. (C, left panel) Ipsilateral mechanical threshold in wild-type and b2 -KO mice following intraplantar CFA inflammation. ⁄KO saline versus KO CFA, +wild-type Saline versus wild-type CFA. (C, right panel) To account for differences in basal nociceptive threshold, the data obtained 24 hours after CFA injection are also presented as a percentage of the baseline threshold. ⁄Pretest versus CFA, + wild-type CFA versus KO CFA. (D, left panel) Ipsilateral ⁄ ⁄ + mechanical threshold in wild-type and b2 -KO mice after cuff-induced neuropathy. KO Sham versus KO Cuff, wild-type Sham versus wild-type Cuff. (D, right panel) To account for differences in basal nociceptive threshold, the data obtained 7 days after the cuff implantation are also presented as a percentage of the baseline threshold. ⁄Pretest versus Cuff, + wild-type Cuff versus KO Cuff. n = 5–7 per group. 2136 I. Yalcin et al. / PAINÒ 152 (2011) 2131–2137

15 * 15 * administration of hexamethanonium induces mechanical allodynia ⁄ in both wild-type and b2 -KO mice, showing that, at the peripheral

) ⁄

g level, non–b2 -nAChRs contribute to the control of mechanical 10 ( 10

e

r nociceptive thresholds. Together, these data indicate that spinal

u

s ⁄

s Pretest b2 -nAChRs are of fundamental importance in controlling the neu-

e 5 r 5

P After TENS Pressure (g) ronal network that sets mechanical thresholds associated with acute nociceptive stimulation. 0 0 In the dorsal horn of the spinal cord, b ⁄-nAChRs are expressed No TENS TENS WT KO 2 by both inhibitory and excitatory neurons [8]; however, it has been ⁄ Fig. 6. b2 -KO mice are insensitive to high frequency transcutaneous electrical proposed that cholinergic antinociception might be due to facilita- nerve stimulation (TENS). The effects of TENS at very high frequency (2000 Hz) tion of inhibitory synaptic transmission. Indeed, electrophysiolog- were tested on the mechanical threshold in C57BL/6J mice (left panel), and in wild- ical and pharmacological studies have shown that nAChR agonists ⁄ type and b2 -KO littermate mice (right panel). n = 5–6 per group. enhance inhibitory GABAergic transmission in the dorsal horn of 4. Discussion the spinal cord [13,30,36]. Spinal nAChRs have been identified on local inhibitory neurons where they are, or are not, under tonic Our results indicate a critical role for cholinergic transmission cholinergic control depending on the cell type [6]. nAChRs have ⁄ also been identified on serotonergic terminals from descending through b2 -nAChRs in the control of nociceptive thresholds. Indeed, both mechanical and thermal nociceptive thresholds measured by pathways which are tonically activated by acetylcholine [6]. Thus, ⁄ stimulation by nicotinic agonists, as well as by endogenous acetyl- DHP are significantly lowered in b2 -KO mice. Combining this genetic approach with pharmacological manipulation of nicotinic receptors, , may also modulate these descending pathways. However, ⁄ a4 or b2 subunits are not necessary to the nicotine induced seroto- we show that spinal, but not peripheral, b2 -nAChRs are critical for nicotinic control of mechanical thresholds. Using i.t. manipulation nin release as confirmed in KO mice and the lack of antagonism by ⁄ of GABA receptor, we also show that the influence of b ⁄-nAChRs MLA suggests that a7 -nAChRs are not involved either in this effect A 2 ⁄ on nociceptive thresholds is mediated through spinal GABAergic [6]. Here, we show a clear loss of tonic GABAergic inhibition in b2 - transmission. Moreover spinal b ⁄-nAChRs are also implicated in KO mice. Indeed, bicuculline induces mechanical allodynia in wild- 2 ⁄ the recruitment of inhibitory control of nociception, as shown by de- type mice but had no effect in b2 -KO mice whereas muscimol in- layed recovery from capsaicin-induced allodynia, exaggerated re- creases mechanical nociceptive thresholds in KO mice to a level sponse to inflammation- and neuropathy-induced nociception, and comparable to that of WT mice. These results suggest that the ⁄ ⁄ low mechanical threshold detected in b2 -KO mice might be be- by the loss of TENS-induced analgesia in b2 -KO mice. The spinal cholinergic system is known to have an antinocicep- cause of a decrease in tonic GABAA receptor-mediated inhibition tive role. Acetylcholine participates in the analgesic action of opi- in the dorsal horn of the spinal cord. oids and of the -adrenoceptor agonist clonidine, and it has We thus propose that the lack of GABAergic control renders the a2 ⁄ intrinsic analgesic properties [4,15,25]. Although many studies b2 -KO mice more sensitive to the nociceptive inputs, and/or that emphasize the participation of muscarinic receptors [5,10,11, these mice display a reduced inhibitory control mediated by non- 25,26], there is also evidence for an involvement of nicotinic recep- nociceptive primary afferents on nociceptive transmission. Indeed, tors in the cholinergic control of nociception [30,31]. Indeed, i.t capsaicin, at a dose ineffective in wild-type mice, potentiates injection of nicotinic antagonists in mice induces thermal hyperal- mechanical allodynia in KO mice. In addition, the action of an effec- ⁄ gesia [30] and mechanical allodynia (present work) suggesting the tive dose of capsaicin lasts longer in b2 -KO mice compared with presence of tonic inhibitory control of nociception through nico- wild-type mice. The inhibitory control exerted on C-fiber-mediated ⁄ tinic receptors in the spinal cord. nociception (triggered by capsaicin) appears to be deficient in b2 - ⁄ Although several subunits of the nAChR are expressed in the KO mice. The exaggerated mechanical allodynia observed in b2 -KO dorsal horn of the spinal cord [8,14], the nAChRs locally responsi- mice after the induction of inflammation or neuropathy also sup- ble for the tonic inhibition of nociception under physiological con- ports the hypothesis that nociceptive transmission is enhanced in ditions appear to be the b ⁄-nAChRs. Indeed, i.t. delivery of these mice. Moreover, stimulation of Ab fibers by TENS has no ef- 2 ⁄ fect in b2 -KO mice whereas it induces analgesia in wild-type mice. nicotinic antagonists selective for nAChRs containing the b2 sub- unit induce mechanical allodynia in wild-type mice without affect- The alleviation of pain observed after TENS stimulation can be ⁄ interpreted within the context of the gate control theory, which ing mechanical nociceptive threshold in b2 -KO mice. In contrast, ⁄ postulates that the stimulation of large non-nociceptive Ab sensory the i.t. blockade of a7 -nAChRs did not affect mechanical nocicep- tive thresholds in either wild-type or b ⁄-KO mice. These behav- fibers inhibits the transmission of nociceptive messages conveyed 2 ⁄ ioral findings are in agreement with the results showing the lack by peripheral nociceptive fibres [23]. Our results indicate that b2 - ⁄ nAChRs might play a fundamental role in the initiation of inhibi- of involvement of spinal a7 -nAChRs in the setting of thermal and mechanical nociceptive hypersensivity in mice under basal tory interaction between non-nociceptive and nociceptive physiological conditions [30] and in the analgesic action of nicotine afferents. and epibatidine [31]. It suggests that even though ⁄-nAChRs may In conclusion, pharmacological and genetic approaches show a7 ⁄ influence nociceptive transmission in spinal cord slices [12], these that b2 -nAChRs are essential for spinal nicotinic tone setting noci- spinal receptors are not involved in the tonic control of nociceptive ceptive thresholds via the control of GABAergic inhibition. This ⁄ thresholds. Nevertheless, we have to bear in mind that while spinal participation of b2 -nAChRs in the modulation of nociceptive trans- ⁄ mission suggest that these receptors might be targets of interest a7 -nAChRs are not implicated in the tonic control of nociceptive threshold under physiological conditions, they play an important for developing selective pharmacological compounds in the con- role in the modulation of peripheral inflammatory pain since cho- text of pain treatment. ⁄ line, an agonist of a7 -nAChRs, displays antinociceptive properties in models of inflammation [32,38]. The nonselective nicotinic Conflict of interest statement antagonist hexamethonium has no allodynic effect after i.t. admin- ⁄ ⁄ istration in b2 -KO mice, suggesting that spinal non-b2 -nAChRs are The authors state that they have no disclosures or conflict of not critical for establishing nociceptive thresholds. In contrast, i.p. interest. I. Yalcin et al. / PAINÒ 152 (2011) 2131–2137 2137

Acknowledgments Reduced antinociception in mice lacking neuronal nicotinic receptor subunits. Nature 1999;398:805–10. [21] Matsumoto M, Xie W, Inoue M, Ueda H. Evidence for the tonic inhibition of This work was supported by the Centre National de la Recher- spinal pain by nicotinic cholinergic transmission through primary afferents. ⁄ che Scientifique and the Université de Strasbourg. b2 -nAChRs- Mol Pain 2007;3:41. deficient mice were kindly provided by the Institut Pasteur. The [22] McMullan S, Simpson DA, Lumb BM. A reliable method for the preferential activation of C- or A-fibre heat nociceptors. J Neurosci Methods authors thank Stéphane Doridot for animal care. 2004;138:133–9. [23] Melzack R, Wall PD. Pain mechanisms: a new theory. Science 1965;150:971–9. References [24] Osikowicz M, Mika J, Makuch W, Przewlocka B. Glutamate receptor ligands attenuate allodynia and hyperalgesia and potentiate morphine effects in a mouse model of neuropathic pain. Pain 2008;139:117–26. [1] Benbouzid M, Pallage V, Rajalu M, Waltisperger E, Doridot S, Poisbeau P, [25] Pan HL, Chen SR, Eisenach JC. Intrathecal clonidine alleviates allodynia in Freund-Mercier MJ, Barrot M. Sciatic nerve cuffing in mice: a model of neuropathic rats: interaction with spinal muscarinic and nicotinic receptors. sustained neuropathic pain. Eur J Pain 2008;12:591–9. Anesthesiology 1999;90:509–14. [2] Blondel A, Sanger DJ, Moser PC. Characterisation of the effects of nicotine in the [26] Pan HL, Wu ZZ, Zhou HY, Chen SR, Zhang HM, Li DP. Modulation of pain five-choice serial reaction time task in rats: antagonist studies. transmission by G-protein-coupled receptors. Pharmacol Ther Psychopharmacology (Berl) 2000;149:293–305. 2008;117:141–61. [3] Bohren Y, Karavelic D, Tessier LH, Yalcin I, Gaveriaux-Ruff C, Kieffer BL, Freund- [27] Picciotto MR, Brunzell DH, Caldarone BJ. Effect of nicotine and nicotinic Mercier MJ, Barrot M. Mu-opioid receptors are not necessary for nortriptyline receptors on anxiety and depression. Neuroreport 2002;13:1097–106. treatment of neuropathic allodynia. Eur J Pain 2010;14:700–4. [28] Picciotto MR, Zoli M, Lena C, Bessis A, Lallemand Y, Le Novere N, Vincent P, Pich [4] Chen SR, Pan HL. Spinal endogenous acetylcholine contributes to the analgesic EM, Brulet P, Changeux JP. Abnormal avoidance learning in mice lacking effect of systemic morphine in rats. Anesthesiology 2001;95:525–30. functional high-affinity nicotine receptor in the brain. Nature 1995;374:65–7. [5] Chen SR, Pan HL. Up-regulation of spinal muscarinic receptors and increased [29] Rashid HI, Ajeel A, Tyrrell MR. Persistent popliteal fossa reflux following antinociceptive effect of intrathecal muscarine in diabetic rats. J Pharmacol saphenopopliteal disconnection. Br J Surg 2002;89:748–51. Exp Ther 2003;307:676–81. [30] Rashid MH, Furue H, Yoshimura M, Ueda H. Tonic inhibitory role of [6] Cordero-Erausquin M, Changeux JP. Tonic nicotinic modulation of alpha4beta2 subtype of nicotinic acetylcholine receptors on nociceptive serotoninergic transmission in the spinal cord. Proc Natl Acad Sci USA transmission in the spinal cord in mice. Pain 2006;125:125–35. 2001;98:2803–7. [31] Rashid MH, Ueda H. Neuropathy-specific analgesic action of intrathecal [7] Cordero-Erausquin M, Marubio LM, Klink R, Changeux JP. Nicotinic receptor nicotinic agonists and its spinal GABA-mediated mechanism. Brain Res function: new perspectives from knockout mice. Trends Pharmacol Sci 2002;953:53–62. 2000;21:211–7. [32] Rowley TJ, McKinstry A, Greenidge E, Smith W, Flood P. Antinociceptive and [8] Cordero-Erausquin M, Pons S, Faure P, Changeux JP. Nicotine differentially anti-inflammatory effects of choline in a mouse model of postoperative pain. activates inhibitory and excitatory neurons in the dorsal spinal cord. Pain Br J Anaesth 2010;105:201–7. 2004;109:308–18. [33] Salminen O, Whiteaker P, Grady SR, Collins AC, McIntosh JM, Marks MJ. The [9] Corringer PJ, Le Novere N, Changeux JP. Nicotinic receptors at the amino acid subunit composition and pharmacology of alpha-conotoxin MII-binding level. Annu Rev Pharmacol Toxicol 2000;40:431–58. nicotinic acetylcholine receptors studied by a novel membrane-binding [10] Duttaroy A, Gomeza J, Gan JW, Siddiqui N, Basile AS, Harman WD, Smith PL, assay. Neuropharmacology 2005;48:696–705. Felder CC, Levey AI, Wess J. Evaluation of -induced [34] Sandkuhler J. Models and mechanisms of hyperalgesia and allodynia. Physiol analgesia in muscarinic knockout mice. Mol Rev 2009;89:707–58. Pharmacol 2002;62:1084–93. [35] Shytle RD, Silver AA, Lukas RJ, Newman MB, Sheehan DV, Sanberg PR. Nicotinic [11] Ellis JL, Harman D, Gonzalez J, Spera ML, Liu R, Shen TY, Wypij DM, Zuo F. acetylcholine receptors as targets for antidepressants. Mol Psychiatry Development of muscarinic analgesics derived from epibatidine: role of the 2002;7:525–35. M4 receptor subtype. J Pharmacol Exp Ther 1999;288:1143–50. [36] Takeda D, Nakatsuka T, Papke R, Gu JG. Modulation of inhibitory synaptic [12] Genzen JR, McGehee DS. Short- and long-term enhancement of excitatory activity by a non-alpha4beta2, non-alpha7 subtype of nicotinic receptors in transmission in the spinal cord dorsal horn by nicotinic acetylcholine the substantia gelatinosa of adult rat spinal cord. Pain 2003;101:13–23. receptors. Proc Natl Acad Sci USA 2003;100:6807–12. [37] Vincler M. Neuronal nicotinic receptors as targets for novel analgesics. Expert [13] Genzen JR, McGehee DS. Nicotinic modulation of GABAergic synaptic Opin Investig Drugs 2005;14:1191–8. transmission in the spinal cord dorsal horn. Brain Res 2005;1031:229–37. [38] Wang Y, Su DM, Wang RH, Liu Y, Wang H. Antinociceptive effects of choline [14] Genzen JR, Van Cleve W, McGehee DS. Dorsal root ganglion neurons express against acute and inflammatory pain. Neuroscience 2005;132:49–56. multiple nicotinic acetylcholine receptor subtypes. J Neurophysiol [39] Wilkins Jr LH, Grinevich VP, Ayers JT, Crooks PA, Dwoskin LP. N-n- 2001;86:1773–82. alkylnicotinium analogs, a novel class of nicotinic receptor antagonists: [15] Hood DD, Mallak KA, James RL, Tuttle R, Eisenach JC. Enhancement of analgesia interaction with alpha4beta2⁄ and alpha7⁄ neuronal nicotinic receptors. J from systemic opioid in humans by spinal inhibition. J Pharmacol Exp Ther 2003;304:400–10. Pharmacol Exp Ther 1997;282:86–92. [40] Yalcin I, Charlet A, Freund-Mercier MJ, Barrot M, Poisbeau P. Differentiating [16] Inoue M, Rashid MH, Fujita R, Contos JJ, Chun J, Ueda H. Initiation of thermal allodynia and hyperalgesia using dynamic hot and cold plate in neuropathic pain requires lysophosphatidic acid receptor signaling. Nat Med rodents. J Pain 2009;10:767–73. 2004;10:712–8. [41] Yalcin I, Choucair-Jaafar N, Benbouzid M, Tessier LH, Muller A, Hein L, Freund- [17] Khan IM, Taylor P, Yaksh TL. Stimulatory pathways and sites of action of Mercier MJ, Barrot M. Beta(2)-Adrenoceptors are critical for antidepressant intrathecally administered nicotinic agents. J Pharmacol Exp Ther treatment of neuropathic pain. Ann Neurol 2009;65:218–25. 1994;271:1550–7. [42] Yalcin I, Tessier LH, Petit-Demouliere N, Waltisperger E, Hein L, Freund- [18] Kurucsai G, Joo I, Fejes R, Szekely A, Szekely I, Tihanyi Z, Altorjay A, Funch- Mercier MJ, Barrot M. Chronic treatment with agonists of beta(2)-adrenergic Jensen P, Varkonyi T, Madacsy L. Somatosensory hypersensitivity in the receptors in neuropathic pain. Exp Neurol 2010;221:115–21. referred pain area in patients with chronic biliary pain and a sphincter of Oddi [43] Yeomans DC, Proudfit HK. Nociceptive responses to high and low rates of dysfunction: new aspects of an almost forgotten pathogenetic mechanism. Am noxious cutaneous heating are mediated by different nociceptors in the rat: J Gastroenterol 2008;103:2717–25. electrophysiological evidence. Pain 1996;68:141–50. [19] Levin ED. Nicotinic systems and cognitive function. Psychopharmacology [44] Zoli M, Picciotto MR, Ferrari R, Cocchi D, Changeux JP. Increased (Berl) 1992;108:417–31. neurodegeneration during ageing in mice lacking high-affinity nicotine [20] Marubio LM, del Mar Arroyo-Jimenez M, Cordero-Erausquin M, Lena C, Le receptors. EMBO J 1999;18:1235–44. Novere N, de Kerchove d’Exaerde A, Huchet M, Damaj MI, Changeux JP.