Increased Opioid Dependence in a Mouse Model of Panic Disorder

Panic disorder is a highly prevalent neuropsychiatric disorder that shows co-occurrence with substance abuse. Here, we demonstrate that TrkC, the high-affinity receptor for neurotrophin-3, is a key molecule involved in panic disorder and opiate dependence, using a transgenic mouse model (TgNTRK3). Constitutive TrkC overexpression in TgNTRK3 mice dramatically alters spontaneous firing rates of locus coeruleus (LC) neurons and the response of the noradrenergic system to chronic opiate exposure, possibly related to the altered regulation of neurotrophic peptides observed. Notably, TgNTRK3 LC neurons showed an increased firing rate in saline-treated conditions and profound abnormalities in their response to met5-enkephalin. Behaviorally, chronic morphine administration induced a significantly increased withdrawal syndrome in TgNTRK3 mice. In conclusion, we show here that the NT-3/TrkC system is an important regulator of neuronal firing in LC and could contribute to the adaptations of the noradrenergic system in response to chronic opiate exposure. Moreover, our results indicate that TrkC is involved in the molecular and cellular changes in noradrenergic neurons underlying both panic attacks and opiate dependence and support a functional endogenous opioid deficit in panic disorder patients.


INTRODUCTION
There is a high rate of co-occurrence between substance abuse and other psychiatric disorders (Regier et al., 1990;Kessler et al., 1994;Kandel et al., 2001) such as anxiety and panic disorder (Cottler et al., 1992;Kushner et al., 2000;Breslau et al., 2003;Zvolensky et al., 2005Zvolensky et al., , 2006. Based on clinical data, it has been proposed that deregulation of the opioid system at the brainstem level may potentially underlie critical aspects of panic disorder's neurobiology including respiratory abnormalities and sensitivity to suffocation. The locus coeruleus (LC) underlies the physiologic and affective changes seen in opiate withdrawal and in panic attacks (Aghajanian, 1978;Aghajanian and Wang, 1986;Gear et al., 1995;El-Kadi and Sharif, 1997;Nakai et al., 2002) and could be a common structure for both processes so that withdrawal syndrome and panic attacks may be different symptomatic expressions of similar neurobiological mechanisms. So, in subjects with overactivity of endogenous opioid system, naltrexone might induce panic attacks by removing this inhibitory effect (Aghajanian, 1978;Arntz et al., 1993;Maremmani et al., 1998;Sinha et al., 2007) on noradrenergic neurons in the LC. Moreover, studies focused on the treatment of pain, the most frequent fear of dying patients have used morphine to reduce states of panic (Roller, 2000), as well in studies with rats in which morphine administration inhibited the dorsal periaqueductal grey (dPAG), a brain region responsible for the escape behavior that is thought to be related to fear and panic (Nicolas et al., 2007). Sklair-Tavron and Nestler, 1995). Noradrenergic LC neurons of BDNF-defi cient mice respond to chronic morphine treatment with a paradoxical down regulation of cAMP-mediated excitation and lack of dynamic regulation of TH, and these mice also have reduced opiate withdrawal symptoms despite normal antinociceptive tolerance (Akbarian et al., 2002).
In the present study, we have used transgenic mice overexpressing TrkC (TgNTRK3), a model of panic disorder (Dierssen et al., 2006;Sahún et al., 2007b), to test the hypothesis that the LC opioid system might be altered with an impact on opiate dependence as previously suggested in human patients.. We have characterized the behavioral and electrophysiological responses to acute and chronic morphine treatment and during opiate withdrawal along with the changes in expression of BDNF, NT-3 TrkB and TrkC mRNAs. Our work highlights the implication of TrkC in panic disorder and substance abuse.

ANIMALS
Same sex hybrid B6/SJL-F1J mice were housed in groups of 3-5 animals per cage in standard Macrolon® cages (40 cm long × 25 cm wide × 20 cm high) under a 12 h light/dark schedule (lights on 08:00 to 20:00) in controlled environmental conditions of humidity (50%-70%) and temperature (22 ± 2ºC) with food and water supplied ad libitum. The production TgNTRK3 mice has previously been described (Dierssen et al., 2006). Briefl y, TgNTRK3 mice were obtained as heterozygotes on a B6/SJL-F1J genetic background (Charles River, Barcelona, Spain). Hybrid founders were crossed using B6/SJL-F1J females and all experiments were performed using male mice from the F16-F20 generation to attenuate littermate's genetic differences. The non-transgenic (WT) littermates obtained from crosses of males TgNTRK3 mice and B6/SJL-F1J females served as controls. All experimental procedures were approved by the local ethics committee (CEEA-IMIM and CEEA-PRBB), and met the guidelines of the local (Catalan law 5/1995 andDecrees 214/97, 32/2007) and European regulations (EU directives 86/609 and 2001-486) and the Standards for Use of Laboratory Animals nº A5388-01 (NIH). The CRG is authorized to work with genetically modifi ed organisms (A/ES/05/I-13 and A/ES/05/14).

DRUGS
Morphine was provided by the Ministerio de Sanidad y Consumo (Spain). Naloxone and Met 5 -enkephalin (ME) was purchased from Sigma Chemical Co. (Madrid, Spain). Both compounds were dissolved in physiological saline (0.9% NaCl).

CHARACTERIZATION OF THE TgNTRK3 BEHAVIORAL RESPONSE TO OPIATES
For the pharmacological in vivo study of opiate effects in TgNTRK3 mice we used 15-20 male mice of 2-4 months of age per group of treatment and genotype.

Acute effects of morphine
We evaluated the acute effects of morphine (10 mg/kg, s.c.) on nociception using the tail immersion and hot plate tests and on locomotion in WT and TgNTRK3 mice. For the hot plate test we used different doses of morphine in order to avoid a ceiling effect (2.5, 5, and 10 mg/kg, s.c.). For the tail immersion test, mice were gently placed in a restrainer cylinder and the nociceptive threshold was assessed as described previously Ruiz et al., 1993;Simonin et al., 1998) by measuring the time to withdraw the tail immersed in a thermostatically controlled water bath (50 ± 0.1ºC; Clifton, Scientifi c Instruments, England), with a cutoff latency of 15 s to prevent tissue damage. We also used a hot plate analgesic meter (Columbus, OH, USA) that consisted of a glass cylinder (19-cm high, 19-cm diameter) to keep the mice on the heated surface of the plate (52 ± 0.1ºC). Two nociceptive indexes, licking of the paws and jumping, were registered. The cutoff latency was of 30 s for the licking response and of 240 s for the jumping response, respectively. Locomotor responses were evaluated using activity boxes (9 cm long × 20 cm wide × 11 cm high; Imetronic, Lyon, France); with a line of photocells 2 cm above the fl oor to measure horizontal movements and another line located 6 cm above the fl oor to measure vertical activity (rearing). On the testing day, the mice were individually placed in the boxes and the total locomotor activity (ambulatory movements plus horizontal movements) was recorded for 30 min in a low luminosity environment (5-15 Lux).
In order to habituate the animals to the test environment and to obtain a stable baseline, tail immersion response and locomotor activity were measured for 2 days before the experiment. Saline injections were given on these days. On the experimental day, locomotor activity was evaluated immediately after morphine injection (10 mg/kg, s.c.) during 30 min. Then, tail immersion test was performed, followed by the hot plate test.

Tolerance and sensitization to morphine effects
Tolerance refers to drug-induced adaptations that lead to diminishing effects of a constant, repeated drug dose. Sensitization, or reverse tolerance, refers to drug-induced adaptations that enhance drug responsiveness with repeated drug exposure . Morphine causes both tolerance and sensitization, with antinociceptive effects decreasing over time while locomotor activity increases. As for the previous experiment, basal tail immersion response and locomotor activity were evaluated for 2 days before the initiation of the chronic morphine treatment on day 1. Morphine (10 mg/kg, s.c.) was administered chronically twice a day (12 h inter-injection interval) for 7 days. Locomotor activity was recorded immediately after the morning injection of morphine, as described above. Changes in nociceptive threshold were assessed 30 min later using the tail immersion test (days 1-7).

Naloxone-precipitated morphine withdrawal
The naloxone-precipitated withdrawal syndrome in morphinedependent mice was evaluated as previously described (Maldonado, 1997). Briefl y, behavioral ratings were calculated over a 15 min period by a trained observer blind to the genotype. The animals were placed individually into test chambers consisting of transparent plastic cylinders (50 cm high, 30 cm diameter) with a white fl oor, to evaluate the behavioral signs of withdrawal. Behavioral responses were recorded in two sessions: 15 min before naloxone injection and the 30 min after injection. Wet dog shakes, jumping, paw tremor and sniffi ng were quantifi ed. Ptosis, diarrhea, teeth chattering, body tremor and piloerection were evaluated over February 2010 | Volume 3 | Article 60 | 3 Gallego et al.
TrkC increases opiate dependence 5 min periods with the presence of each sign during each period being scored as 1. The numbers of periods showing each sign were then counted (maximum score = 6). A global withdrawal score, determined by all the checked signs was calculated individually for each mouse. To obtain this global value and to assign a proportional weight to all the signs, the score obtained from each sign was multiplied by a constant as previously reported (Maldonado et al., 1996). Naloxone-precipitated withdrawal was evaluated after two different treatment regimes of morphine as follows in order to avoid ceiling effects on withdrawal symptoms that would blind possible differences between genotypes.

Naloxone-precipitated morphine withdrawal after increasing doses of morphine.
Morphine was injected i.p. twice a day (12-h inter-injection interval) for 6 days. The morphine dose was progressively increased as follows: 1st day, 20 mg/kg/injection??; 2nd day, 40 mg/kg; 3rd day, 60 mg/kg; 4th day, 80 mg/kg; 5th day, 100 mg/kg; 6th day (only morning injection), 100 mg/kg. Control mice of both genotypes were treated with saline using the same injection schedule. Withdrawal was precipitated in each animal by injecting naloxone (1 mg/kg, s.c.) 2 h after the last morphine administration.

Naloxone-precipitated morphine withdrawal after constant dose of morphine.
Morphine was injected i.p. twice a day (12-h interinjection interval) for 6 days. In this experimental group, the dose of morphine was constant (20 mg/kg/injection?) during the 6 days. Control mice of both genotypes were treated with saline using the same injection schedule. Withdrawal was precipitated in each animal by injecting naloxone (1 mg/kg, s.c.), 2 h after the last morphine administration.
Anxiety-like behavior during morphine withdrawal. The anxiogenic effects of morphine withdrawal were studied using the zeromaze test immediately after evaluation of withdrawal symptoms, in animals receiving constant doses of morphine (20 mg/kg; see Naloxone-precipitated morphine withdrawal after constant dose of morphine). The zero-maze consisted of a circular path (runway width 5.5 cm, 46 cm diameter) with two open and two closed segments (walls 8 cm high) and was elevated 50 cm. Animals were placed into the closed segments and their movements analyzed for 5 min. The experiment was carried out in a dimly lit room to minimize any effects of stress/anxiety and data were scored with the aid of a video camera. The latency to enter the open segment, the number of crossings and the time spent in the open segments were recorded. An open segment visit was scored when the mouse moved all four paws into the segment.

IN SITU HYBRIDIZATION
Three TgNTRK3 and 3 WT littermates were used for each treatment condition following the increasing dose morphine injection schedule detailed in Section "Naloxone-precipitated morphine withdrawal after increasing doses of morphine". Animals were killed 2 h after naloxone injection to study morphine withdrawal or 4 h after the last morphine injection to study chronic morphine effects. The same schedule was used in the saline-treated group. Then, brains were dissected out and frozen in dry ice and serial coronal cryostat sections (14 µm) were processed for TrkB, TrkC, NT-3, and BDNF riboprobes as follows. After rinsing in PBS, tissue was deproteinated in 0.2 M HCl for 10 min, acetylated with 0.25% acetic anhydride in 0.1 M ethanolamine for 10 min, post-fi xed in 4% paraformaldehyde, and dehydrated with increasing concentrations of ethanol. Slices were incubated for 16 h in a humidifi ed chamber at 53°C with 3 × 105 c.p.m. of antisense probes in 150 µL of hybridization cocktail (50% formamide, 20 mM Tris-HCl, pH 7.6, 1 mM EDTA, pH 8.0, 0.3 M NaCl, 0.1 M dithiothreitol, 0.5 mg/ mL yeast tRNA, 0.1 mg/mL poly (A+) RNA, 1 × Denhardt's solution, and 10% dextran sulfate). Antisense cRNA probes to detect the transcript were prepared by in vitro transcription using T7 RNA polymerase (Promega, WI, USA) and [35S] UTP (Amersham Pharmacia Biotech, Uppsala, Sweden). The TrkB probe recognized both full-length and truncated isoforms (Lindefors et al., 1995) and the TrkC probe detected all TrkC receptor forms (Laurenzi et al., 1994). For control experiments, sense cRNA probe was obtained by in vitro transcription using T3 RNA polymerase (Promega, WI, USA). After hybridization, slices were fi rst washed at room temperature in 1 × SSC, followed by two washes in 1 × SSC at 37°C for 15 min each. Single-stranded RNA was digested by RNase treatment (40 µg/mL) for 30 min at 37°C in 0.5 M NaCl, 20 mM Tris-HCl, pH 7.6, and 2 mM EDTA. Tissue was washed twice in 1 × SSC in 65°C for 10 min, dehydrated in ethanol, and air-dried. Slices were exposed to β-max x-ray fi lm (Amersham Pharmacia Biotech, Uppsala, Sweden) for 20 days and dipped in LM-1 photoemulsion (Amersham Pharmacia Biotech, Uppsala, Sweden), exposed at 4°C for 2 months, developed with D19 (Eastman Kodak), fi xed, and counterstained with cresyl violet staining. Consecutive slices were stained against TH in order to delimit the LC and thus allowing an accurately quantifi cation of this region.

QUANTITATIVE-PCR
To determine changes in endogenous TrkC receptors [full length (TrkC FL) and truncated (TrkC TF)], quantitative-PCR was carried out in duplicate in brainstem from adult (2-4 month of age) TgNTRK3 (n = 4) and WT mice (n = 3) with the cDNA solution using the LightCycler® 480 SYBR Green I Master kit (Roche Diagnostics GmbH, Mannheim, Germany) in a LightCycler® 480 Real-Time PCR system (Roche Diagnostics GmbH, Mannheim, Germany) according to the manufacturer's recommendations, except that the fi nal volume of the reaction was 20 µL in order to avoid errors due to high volumes. The cDNA solution was submitted to 45 cycles of PCR amplifi cation by using specifi c primers from the endogenous TrkC: TrkC FL -F 5′-ACCGGAAGTTCACCACAGAG -3′ and TrkC FL -R 5′-GCAGACTCTGGGTCTCTCCA -3′; TrkC TF -F 5′-CCAGAGTGGGGATGTGTCTT -3′ and TrkC TF -R 5′-CCATGGTTAAGAGGCTTGGA -3′. The Pgk1 gene was used as the reference gene. Final volume was 20 µL, with 1 µL of one-tenth H2O-diluted cDNA from reverse transcription reaction, 0.3 µmol/L of each specifi c primer, and 10 µL 2 × SYBR Green I Master solutions. Quantitative-PCR cycles were set at 10 min denaturation followed by 45 cycles of 95ºC for 10 s, 61ºC for 15 s, and 72ºC for 10 s and 95ºC. Dissociation curves were inspected for each pair of primers, and only dissociation peak must be present for each run of reaction before the results were considered to be valid. A 4-fold serial dilution of a reference sample was used for construction of standard curves with respect to each pair of primers. All samples were amplifi ed in duplicate and the CT mean was obtained for further calculations. Duplicates with SD ≥ 0.38 were excluded, and the corresponding samples were re-amplifi ed. The ΔΔCt method [Relative Expression Software Tool (REST)] was used to quantify the relative amount of mRNA in each sample in comparison with the control samples. Next, 2−ΔΔCt of each gene was transformed into logarithmic scale (log 2 ratio).

ELECTROPHYSIOLOGICAL STUDIES
Single-unit extracellular recordings of LC neurons were performed in slices from saline treated (WT n = 16 and TgNTRK3 n = 12) and morphine-treated (WT n = 16; TgNTRK3 n = 16) mice, as previously described (Pineda et al., 1996) and following the increasing dose morphine injection schedule of Section "Naloxone-precipitated morphine withdrawal after increasing doses of morphine". Briefl y, mice were anesthetized with chloral hydrate (400 mg/kg, i.p.) and decapitated 1-4 h after the last injection of morphine. Coronal slices (400 µm) of the brainstem including the LC were prepared with a vibratome and incubated in a Haas-type gas-liquid interface chamber at 33ºC, which was continuously perfused with artifi cial cerebrospinal fl uid (aCSF) at a fl ow rate of 1-1.5 mL/min (saturated with 95% O 2 /5% CO 2 ). The aCSF was composed of: 129 mM NaCl, 3 mM KCl, 1.25 mM NaH 2 PO 4 , 2 mM MgSO 4 , 2 mM CaCl 2 , 22 mM NaHCO 3 and 10 mM D-glucose. The extracellular signal was recorded by blunt-tipped Omegadot glass micropipettes fi lled with a 50 mM NaCl solution (3-5 MΩ), and then monitored through a highinput impedance amplifi er on an oscilloscope and also with an audio unit. Individual neuronal (single-unit) spikes were isolated from the background noise with a window discriminator and the fi ring rate was analyzed by means of a PC-based custom-made program, which generated rate bar histograms in consecutive 10-s bins (HFCP®, Cibertec S.A., Madrid, Spain). The LC was identifi ed visually in the rostral pons as a dark oval area on the lateral borders of the central gray and the 4th ventricle, immediately anterior to the genu of the facial nerve. LC cells displayed spontaneous fi ring activity at regular rhythms and long-lasting positive-negative waveforms.
Opioid withdrawal was spontaneously precipitated in vitro on the slice preparation in chronic morphine-treated mice of both genotypes by allowing morphine to wash out for 90 min before recordings (Kogan et al., 1992). Next, consecutive neurons were sampled during in vitro opiate withdrawal for about 90 min by multiple electrode tracks randomly positioned within the LC. Complete washout of residual morphine was confi rmed at the end of each experiment by testing the lack of naloxone (1 µM) effect. To evaluate opiate tolerance in the LC, concentration-effects curves were constructed for the inhibitory effect of the µ-opioid receptor (MOR) agonist ME on the LC in saline and chronic morphine-treated mice of both genotypes following the increasing dose morphine injection schedule of Section "Naloxone-precipitated morphine withdrawal after increasing doses of morphine". Increasing concentrations of ME (0.1-100 µM) were added at 5 min intervals until a maximal effect was reached. ME was chosen because its washout is very rapid even at high concentrations and its action is mediated almost exclusively by MOR in the LC (Williams and North, 1984).

DATA ANALYSIS AND STATISTICAL EVALUATION
All values are given as mean ± SEM. For the behavioral studies, data obtained from the tail immersion and the hot plate tests were expressed as the percentage of maximal possible effect (% MPE). The calculation of the % MPE was performed using the following formula: [(test latency -control mean latency)/(cutoff -control mean latency)] × 100. For data obtained from acute morphine administration, a two-way analysis of variance (ANOVA) was used to determine the interaction between genotype and treatment. Posterior Student's t test was used to determine differences between TgNTRK3 and WT mice. The development of tolerance to morphine antinoception in the tail immersion test and the development of locomotor sensitization in activity cages were analyzed by using a General Linear Model -Repeated Measures analysis in order to detect differences between curves across days. Two-way ANOVA was used to determine the interaction between treatment and genotype. Data related to the total score obtained from withdrawal signs were analyzed by using Student's t test. Levels of NT-3, BDNF and their receptors, TrkC and TrkB mRNAs were analyzed by using one-way ANOVA for the treatment and Student's t test for the genotype. Bonferroni post hoc was used when necessary. Results from quantitative RT-PCR analysis were obtained in the Relative Expression Software Tool -384 (REST-384© v.2), and analyzed using Pair Wise Re-allocation Randomisation Test©.
For the electrophysiological studies, the basal fi ring rate was obtained in each cell by averaging the values from the rate bars recorded for at least 60 s in the absence of any drug, so that the stability of the discharge was ensured. To evaluate opiate withdrawal in the LC, the mean fi ring rate of all cells sampled in each slice was considered as an individual data point and used for the statistical analysis. The inhibitory effect of each ME concentration was calculated as the percentage of the reduction of fi ring rate from the baseline. In these assays, the fi ring rates after each ME application was the average value obtained from the rate bars in 9 consecutive bins after the onset of the perfusion. Curve fi tting analysis was performed by the computer program GraphPad Prism (version 3.0 for Windows, San Diego, USA) to obtain the best simple nonlinear fi t to the following three-parameter logistic equation: , where E is the effect induced by each concentration of ME (A), Emax is the maximal effect, EC 50 is the concentration of the agonist needed to elicit a 50% of the maximal effect and n is the slope factor of the concentration-effect curve. Emax was constrained to be 100% since the maximal inhibition was always complete. For statistical analysis, the EC 50 values were transformed to the corresponding logarithmic data to convert them to a Gaussian distribution. Data among groups were compared by one-way ANOVA followed, when signifi cant, by a Student's t test (for concentration-effect curve parameter evaluations) or a Tukey's multiple comparison t test (for fi ring rate evaluations). Probability values lower than 0.05 were accepted as signifi cantly different.

THE ACUTE ANTINOCICEPTIVE EFFECTS OF MORPHINE WERE SLIGHTLY ATTENUATED IN TgNTRK3
The acute effects of morphine were evaluated by the changes induced in nociception (Figure 1) and locomotor activity (Figure 2A). In the hot plate test, two-way ANOVA revealed no interaction between Nociceptive threshold was determined during the 7 days of morphine administration after measuring locomotor activity, by means of the tail immersion test. Data are expressed as means ± SEM of the maximum possible effect (MPE %) for the antinociceptive effect or the distance traveled (cm) for locomotor activity. genotype and treatment in any of the responses (F (2,89) = 0.04, n.s., jumping, and F (2,89) = 1.27, n.s., licking). However, the licking response ( Figure 1A), which is highly sensitive to analgesic agents and is considered an accurate measure of antinociceptive potency in rodents, showed a signifi cantly reduced antinociceptive effect of acute administration of 10 mg/kg morphine in TgNTRK3 mice (t (1,47) = 2.02, p < 0.05, Student's t test). Besides, no differences between genotypes were observed when using neither 2.5 nor 5 mg/kg of morphine (t (1,20) = 0.04, n.s., and t (1,18) = -0.59, n.s., respectively, Student's t test) in independent groups of mice. Regarding the jumping response, the higher doses of morphine (5 and 10 mg/kg) showed a ceiling effect ( Figure 1B). However, when using a dose of 2.5 mg/kg of morphine no signifi cant differences were observed between genotypes (t (1,20) = 0.18, n.s., Student's t test). In the tail immersion test (Figure 1C) no signifi cant differences in the acute antinociceptive effects of morphine were observed between genotypes as refl ected by a similar tail withdrawal latency (t (1,48) = 0.66, n.s., Student's t test). The analysis of locomotor activity revealed that the acute effect of morphine on locomotion was similar between genotypes (F (1,107) = 0.81, n.s., two-way ANOVA, Figure 2A). An independent experiment measuring locomotor activity during longer time periods (during 60 min after morphine administration) also did not show differences between genotypes ( Figure S1 in Supplementary Material)

MORPHINE-INDUCED LOCOMOTOR SENSITIZATION AND DEVELOPMENT OF TOLERANCE TO MORPHINE ANTINOCICEPTION WERE NOT MODIFIED BY THE OVEREXPRESSION OF NTRK3
We monitored the development of tolerance and sensitization over the course of 7 days, induced by daily injections of morphine (10 mg/kg twice-a-day). Locomotor activity was measured after saline or morphine injection during the fi rst hours of the light phase (Figure 2A). Animals of either genotype were habituated to the activity cages during 2 days showing a reduction in total locomotor activity between the fi rst and second day (F (1,106) = 76.91, p < 0.001, ANOVA repeated measures). The habituation to the apparatus was similar in both genotypes (F (1,106) = 0.13, n.s., one-way ANOVA). Morphine administration along 7 days after habituation elicited sensitization to total locomotor activity in WT and TgNTRK3 mice when compared to saline groups (F (7,97) = 3.02, p < 0.01; ANOVA repeated measures) with no differences between genotypes (F (7,97) = 0.04, n.s. two-way ANOVA). Moreover, no differences between genotypes were observed along the 7 days of treatment in neither morphine groups (F (7,42) = 0.43, n.s., one-way ANOVA) or saline groups (F (7,49) = 1.01, n.s., one-way ANOVA). Locomotor sensitization also was studied changing the administration protocol. In this experiments, when morphine (10 mg/kg) was administered only once every 2 days again no differences were observed between genotypes ( Figure S1 in Supplementary Material).
Development of tolerance was studied measuring antinociception with the tail immersion test (Figure 2B). During the fi rst 2 days mice of either genotype did not show differences in the habituation to the apparatus (F (1,120) = 0.23, n.s., one-way ANOVA). Both TgNTRK3 and WT mice exhibited a progressive loss of morphine antinociceptive effect along sessions in the tail immersion test as refl ected by no differences in the reduction of the %MPE (F (7,41) = 0.31, n.s., ANOVA repeated measures). Saline-treated mice also showed no genotype differences during the 8 days (F (7,62) = 0.04, n.s., one-way ANOVA). Two-way ANOVA repeated measures revealed no interaction between treatment and genotype (F (7,109) = 0.42, n.s.) along days.

NALOXONE-PRECIPITATED WITHDRAWAL SYNDROME AFTER INCREASING DOSES OF MORPHINE WAS SIMILAR IN TgNTRK3 AND WT MICE
In the fi rst experiment, withdrawal was induced after 6 days of progressively increasing doses of morphine with a subcutaneous injection of naloxone 2 h after the last morphine injection (Figure 3A).
Both WT and TgNTRK3 mice showed the characteristic signs of opiate withdrawal, including autonomic hyperactivity, wet dog shakes, paw tremor, piloerection, jumps, and other withdrawalrelated behaviors. No differences between genotypes were observed in the score obtained for each behavioral sign checked. The total withdrawal score was similar for the TgNTRK3 and WT mice (F (1,79) = 0.82, n.s., two-way ANOVA).

NALOXONE-PRECIPITATED WITHDRAWAL SYNDROME WAS INCREASED AFTER BY CONSTANT DOSES OF MORPHINE IN TgNTRK3 MICE
In order to discard a possible ceiling effect on withdrawal symptoms after dependence induced by increasing dose of morphine, we examined withdrawal symptoms after dependence induced by a lower constant dose of morphine (20 mg/kg) ( Figure 3B). Two-way ANOVA revealed interaction between genotype and treatment (F (1,41) = 5.99, p < 0.05). Student's t test analysis revealed a signifi cant increase in global symptoms score in TgNTRK3 as compared to WT mice (t (1,16) = −2.46, p < 0.05). No differences were observed in animals treated with saline plus naloxone (t (1,21) = −0.99, n.s., Student's t test).

DECREASED ANXIETY BEHAVIOR OF TgNTRK3 MICE DURING WITHDRAWAL SYNDROME
Anxiety was evaluated in the zero-maze test during morphine withdrawal (Figure 4). No interaction between genotype and treatment was revealed by two-way ANOVA in any of the quantifi ed behaviors (F (1,38) = 1.43, n.s., number of crossings;  (Figure 4B) was lower during morphine withdrawal in both TgNTRK3 (t (1,19) = 2.83, p = 0.01, Student's t test) and WT mice, although the latter did not reach signifi cant differences (t (1,19) = 1.13, n.s., Student's t test).

NT-3, BDNF AND THEIR RECEPTORS TrkC AND TrkB MRNAS LEVELS ARE DIFFERENTIALLY REGULATED AFTER CHRONIC MORPHINE ADMINISTRATION AND PRECIPITATED WITHDRAWAL IN TgNTRK3 MICE
To test whether opiate-induced regulation of neurotrophin expression was modifi ed in TgNTRK3 mice, coronal sections throughout the LC of saline-treated and morphine-treated mice were processed for in situ hybridization of NT-3, BDNF and TrkC, TrkB mRNAs ( Figure 5; Table 1). Densitometric analysis of fi lm autoradiograms confi rmed the increased levels of TrkC mRNA that were previously reported in the LC (Dierssen et al., 2006) of TgNTRK3 after saline plus naloxone (t (1,95) = −2.66, p < 0.01, Student's t test). These increased levels of TrkC mRNA were accompanied by an increase in NT-3 mRNA levels (t (1,55) = −3.53, p = 0.001, Student's t test), thus confi rming previous data observed at the protein level (Dierssen et al., 2006).
In contrast to the increase of TrkC mRNA levels observed in TgNTRK3 mice under saline plus naloxone conditions, TrkB mRNA levels were reduced as compared to WT mice (t (1,53) = 4.70, p < 0.001, Student's t test), but no differences in BDNF mRNA levels were observed between genotypes (t (1,37) = 1.53, n.s., Student's t test). TrkB mRNA expression levels were similarly reduced after chronic morphine administration showing no differences between genotypes (t (1,39) = 1.05, n.s., Student's t test) but a signifi cantly greater decrease was observed in TgNTRK3 mice after precipitated withdrawal (t (1,65) = 2.05, p < 0.05, Student's t test). However, whereas the reduction in TrkB levels with respect to saline-treated mice was signifi cant in WT mice (F (2,80) = 11.00, p < 0.001; oneway ANOVA), TgNTRK3 mice showed no differences in mRNA TrkB expression among treatment groups (F (2,81) = 1.02, n.s. oneway ANOVA). Bonferroni post hoc analysis revealed signifi cant differences between saline and chronic morphine (p < 0.001) and also between saline and morphine withdrawal (p < 0.001) TgNTRK3 mice. The expression levels of BDNF changed after chronic morphine treatment and precipitated withdrawal in both WT (F (2,73) = 5.51, p < 0.001, one-way ANOVA) and TgNTRK3 mice (F (2,68) = 9.59, P < 0.001, one-way ANOVA), but a Bonferroni post hoc test showed signifi cant differences only in TgNTRK3 but not in WT mice when comparing chronic morphine treatment vs. saline (p < 0.001). However, signifi cant differences were observed in both genotypes when comparing precipitated withdrawal vs. saline (p < 0.01 in WT and p < 0.001 in TgNTRK3 mice).
To test whether the mRNA expression observed in the in situ hybridization was due to altered mRNA expression of endogenous TrkC in LC, we used Lightcycler qRT-PCR analysis. Data showed no differences in mRNA expression in brainstem of TgNTRK3 mice ( Table 2) using probes to discriminate between mouse and human receptors. The statistical analysis showed no signifi cant differences in endogenous TrkC FL (1.01 ± 0.33, p = 0.93, n.s.) and Trkc TF (1.00 ± 0.25, p = 0.91, n.s.) expression of TgNTRK3 mice with respect to WT.

NTRK3 OVEREXPRESSION ALTERS NORADRENERGIC NEURONAL ACTIVITY AND OCCLUDES MORPHINE-INDUCED FUNCTIONAL ADAPTATIONS IN THE LC
An elevated activity of LC neurons has been postulated to contribute to the expression of opiate withdrawal in morphine-dependent rats (Kogan et al., 1992). To examine whether the TrkC receptor regulates this withdrawal-induced hyperactivity, single-unit extracellular recordings of LC neurons were obtained in brain slices from WT and TgNTRK3 mice. Spontaneous fi ring rates were recorded during in vitro opioid withdrawal in a total of 192 LC cells (on average, 6 cells/slice) from mice chronically treated with morphine or saline. After saline treatment, fi ring rate values of LC neurons in WT animals (0.75 ± 0.1, n = 8) were within the range previously reported in rat brain slices (Kogan et al., 1992). In vitro opiate withdrawal in morphine-treated WT mice increased the fi ring rates of LC cells (+73%, n = 8, p < 0.05) compared to the values in the corresponding saline-treated group (Table 3; Figure 6A). The magnitude of this increase was similar to that previously reported in slices from morphine-dependent mice (Martin et al., 2003). Paradoxically, although fi ring rates in TgNTRK3 after saline treatment were signifi cantly elevated with respect to WT (1.28 ± 0.1 Hz, n = 8; p < 0.05) (Table 3; Figure 6A), they were not signifi cantly elevated during opiate withdrawal (n = 8) with respect to WT or saline-treated groups (Table 3; Figure 6A), suggesting that TrkC receptor overexpression is able to mimic and thereby occlude morphine-induced functional adaptations in the LC. The possibility of residual morphine present in the slice from dependent animals was explored by the administration of naloxone immediately after the cell sampling: naloxone (1 µM) failed to change the fi ring rate of LC cells in either group (data not shown).
To further characterize the changes found in the LC from mice overexpressing TrkC, we tested the functionality of µ-opioid receptors by constructing concentration-effect curves for ME in chronically morphine-and saline-treated mice. In WT mice, treatment with increasing doses of morphine induced a moderate degree of opioid tolerance in the LC, as revealed by a rightward shift in the concentration-effect curves ( Figure 6B) and a 4-fold increase in the EC 50 for ME (n = 9) compared to the saline group (n = 8; One-way ANOVA, F(3,25) = 3.16; post hoc, p < 0.05; Table 3). Interestingly, the EC 50 value obtained from concentration-effect curves for ME in saline-treated TgNTRK3 mice was also signifi cantly shifted to the right (n = 4, p < 0.05) compared to that in the corresponding WT group (n = 8; Table 3). However, after chronic treatment with morphine in TgNTRK3 mice no further change was observed in the EC 50 value for ME (n = 8) as compared to the corresponding curve in saline-treated mice ( Table 3). In fact, concentration-effect curves in morphine-treated TgNTRK3 mice were not different from those in morphine-treated WT mice ( Figure 6B; Table 3), confi rming that TrkC receptor overexpression occludes morphine-induced functional adaptations in the LC.

DISCUSSION
In the present work, constitutive TrkC overexpression in a model of panic disorder, TgNTRK3 mice (Dierssen et al., 2006), dramatically alters spontaneous fi ring rates of LC neurons and the   response of the noradrenergic system to chronic opiate exposure, possibly related to the paradoxical regulation of neurotrophic peptides observed in this model. Notably, TgNTRK3 LC neurons showed an increased fi ring rate in saline-treated conditions and profound abnormalities in their response to ME. In contrast, the behavioral withdrawal pattern was not modifi ed, nor were either the development of tolerance or the locomotor sensitization to morphine when it was administered at increasing doses. Conversely, chronic constant doses of morphine induced a significantly increased behavioral withdrawal syndrome in TgNTRK3. LC is involved in the pathogenesis of panic disorder (Protopopescu et al., 2006) and has served as a model system for the pharmacology and cellular basis of opiate tolerance, dependence, and withdrawal . Several studies have suggested that panic attacks might be the consequence of hyperactivity of LC noradrenergic cells (David Johnson, 2003;Valenca et al., 2004) and changes in the expression of neurotrophins and their receptors are essential for opiate-induced adaptations (Numan et al., 1998;Akbarian et al., 2001Akbarian et al., , 2002Hatami et al., 2007). In the present work we observed profound abnormalities in LC neuronal responses in basal conditions and after chronic opiate exposure in TgNTRK3 mice. Spontaneous LC neurons basal fi ring rates of TgNTRK3 mice were abnormally elevated with respect to WTs. These neurons possess µ-opioid receptors and are innervated by enkephalin and endorphin fi bers (Strahlendorf et al., 1980;Drolet et al., 1992). Similarly, exogenous opioid agonists, acting at µ receptors, produce inhibition of LC cells, a pharmacological inhibition reversed by opioid antagonists such as naloxone, which in turn induces hyperactivity of noradrenergic cells. The nucleus paragigantocellularis (Ennis and Aston-Jones, 1988), which has been implicated in the hyperactivity of LC neurons during withdrawal, is an important source of NT-3 and BDNF (Rasmussen and Aghajanian, 1989) to the LC.
In fact, deregulation of the brainstem opioid system may underlie critical aspects of panic disorder neurobiology, since clinical studies show that normal subjects experience strong physiological reactivity to naloxone-lactate (Sinha et al., 2007) although other studies (Liebowitz et al., 1984) found that panic patients' response to lactate infusion was not modulated by naloxone pretreatment. This lack of effect of naloxone in panic patients has been interpreted as a tonic or phasic opioid hypofunction, although in some cases panic attacks can be precipitated by the administration of naltrexone in a subgroup of panic disorder patients (Maremmani et al., 1998). We tested the functionality of µ-opioid receptors in the LC of TgNTRK3 mice, by constructing concentration-effect curves for ME in chronically morphine-and saline-treated mice. Interestingly, the EC 50 value for ME in saline-treated TgNTRK3 mice was signifi cantly increased compared to the corresponding WT group. These results suggest that overexpression of TrkC leads to hypofunction of opioid receptors, and to a reduced endogenous inhibitory control of LC neurons. Consistently with these results, TgNTRK3 mice present a signifi cantly reduced acute antinociceptive effect of morphine.
It could be argued that an increased excitatory afferent input to the LC could contribute to disinhibition of noradrenergic neurons (Williams and North, 1984;Akaoka and Aston-Jones, 1991;Aghajanian et al., 1994). However, in our experiments, given the fact that the recordings were conducted on coronal slice preparations that disrupt afferent inputs of the LC, it is unlikely that the observed increase in LC neuronal fi ring in the mutants is caused by an imbalance between GABAergic and glutamatergic excitatory signal transmission (Van Bockstaele et al., 2000) to the LC. In fact, in WT mice, treatment with morphine induced a rightward shift in the concentration-effect curves for ME that revealed a moderate degree of opioid tolerance, but no change was observed in the ME EC 50 in TgNTRK3 mice. Thus, we monitored the development of tolerance and sensitization over the course of several days, and we observed that both TgNTRK3 and control mice developed a similar change in morphine responsiveness, indicating that neither tolerance nor sensitization were affected by TrkC overexpression. It is tempting to suggest that TrkC receptor overexpression is able to mimic and thereby occlude morphine-induced functional adaptations in the LC, suggesting that overexpression of TrkC could be a mechanism in the noradrenergic plasticity in LC (Akbarian et al., 2001(Akbarian et al., , 2002. TrkC increases opiate dependence increase in anxiety in humans and mouse models (Schulteis et al., 1998), we tested the possible anxiogenic effects of withdrawal in our panic disorder mouse model. Surprisingly, in our experiments using the zero-maze test we could not detect an increase in anxiety-like behaviors in either WT or TgNTRK3. Conversely both genotypes spent more time in the aversive zones but this reduced anxiety levels were more clear in TgNTRK3 mice. This supports the possibility that zero-maze test behaviors might not be driven exclusively by anxiety levels but rather by other withdrawal-induced behaviors, as has been reported using the elevated plus maze (Hodgson et al., 2008;Buckman et al., 2009). Increased TH expression and catalytic activity (Guitart et al., 1990;Lane-Ladd et al., 1997;Boundy et al., 1998) in the LC is considered a key element in the neuroadaptations produced by chronic morphine administration has been shown to. This upregulation of TH would be expected to increase the capacity of noradrenergic neurons to synthesize norepinephrine, which then could contribute to the dramatic increase in norepinephrine release and to behavioral changes described during opiate withdrawal (Koob, 1992;Grasing et al., 1997). In TgNTRK3 mice, we previously described an up-regulation of TH that may be due to the increased cell numbers in LC (Dierssen et al., 2006). Therefore, one would predict that the TH upregulation in the LC of TgNTRK3 mice would contribute to the increased basal LC activity, observed in the present work.
In our model, it was also relevant to establish the possible modifi cations in neurotrophin expression, since the NT-3/TrkC system has been suggested to interfere with the biological effects of opiates through direct effects on noradrenergic neurons. NT-3 opposes, in vitro, the decline in embryonic A6 neuron numbers after chronic morphine exposure (Sklair-Tavron and Nestler, 1995) and mice with a CNS-wide deletion of NT-3 had attenuated withdrawal reactions that were restored by transgene-derived overexpression of NT-3 in noradrenergic neurons (Akbarian et al., 2001). In TgNTRK3 mice the LC phenotype could thus emanate from the alteration of both BDNF and NT-3 systems. Interestingly, neurotrophins can recover a noradrenergic phenotype in 'resting noradrenergic' neurons in the adult LC (Bezin et al., 2000), a mechanism that could be involved in the pathological alteration of LC in psychiatric conditions such as panic disorder or substance abuse. Chronic opiate treatment and Regarding opioid withdrawal, electrophysiological studies have shown that after chronic morphine treatment, naloxone-induced withdrawal triggers intense hyperactivity of LC neurons that is prevented by clonidine treatment (Taylor et al., 1988;Kogan et al., 1992). In our experiments, LC cells from WT mice chronically treated with morphine showed the typical hyperactivation during in vitro withdrawal. However, this elevation of fi ring rates during opioid withdrawal was not observed in TgNTRK3 mice, thus suggesting that the neuronal adaptations that take place during chronic morphine treatment had not been produced in TgNTRK3 mice. Since an increase in LC activity has been argued to be both necessary and suffi cient to induce most opiate withdrawal behaviors Maldonado and Koob, 1993;Nestler and Aghajanian, 1997), the fact that the LC fi ring rate was not further increased by morphine withdrawal in TgNTRK3 mice would be in accordance with previous studies showing that lesions affecting the LC do not abolish opiate withdrawal reactions (Christie et al., 1997;Caille et al., 1999;Delfs et al., 2000). Thus, we decided to investigate the behavioral withdrawal syndrome. Using increasing doses of morphine, we did not observe qualitative or quantitative differences in somatic or behavioral withdrawal signs between WT and TgNTRK3 animals. In fact, the spontaneous fi ring rates of LC neurons of morphine-treated mice using the same dosage regime were not different between both genotypes. However, it has been argued that decreasing the dose or duration of morphine administration can reduce drug's withdrawal behavioral symptoms thus increasing the sensitivity of the assay (Haghparast et al., 2008). In fact, in TgNTRK3 mice constant low doses of morphine induced a signifi cantly more intense behavioral withdrawal syndrome. This suggests that a ceiling effect may have been reached in the previous experiment, and that the appearance of adverse effects may have hidden the real withdrawal signs. This ceiling effect may be explained either by an opposing effect of down-regulated TrkC in TgNTRK3 mice treated with increasing doses of morphine (see Figure 5A) or by the fact that LC neurons from TgNTRK3 mice are more resistant to opioid inhibition (see Figure 6), which would attenuate induction of spontaneous hyperactivity during morphine withdrawal as shown by Cruz et al. (2008). Alternatively, other structures than the LC that are regulated by morphine may participate in opiate withdrawal. Also, since withdrawal has been related to an February 2010 | Volume 3 | Article 60 | 12 Gallego et al.
TrkC increases opiate dependence noradrenergic system in response to opiate exposure (Numan et al., 1998;Akbarian et al., 2001Akbarian et al., , 2002. It has been shown (Numan et al., 1998), that chronic morphine exposure resulted in modest increases in BDNF and NT-3 but unchanged TrkB or TrkC mRNA expression in LC, while precipitated withdrawal leads to a marked, rapid, and prolonged increase in BDNF mRNA, a delayed decrease in NT-3 mRNA, and to an elevation of both TrkB and TrkC mRNAs at 2 and 6 h of withdrawal in the rat. In our experiments, chronic morphine treatment led to small changes in NT-3 and BDNF mRNA levels and precipitated opiate withdrawal resulted in a robust increase in BDNF along with a slight decrease in NT-3 mRNA levels in WT mice. However, in contrast to previous observations, TrkC was not modifi ed after chronic morphine treatment, but a robust decrease in TrkB expression was observed in WT mice. In TgNTRK3 mice we observed a different regulation of NT-3, BDNF and their receptors TrkC and TrkB mRNA after chronic morphine administration and precipitated withdrawal. These data suggest that variable combinations and intensities of neurotrophin-signaling pathways in neuronal circuits that are interconnected with noradrenergic neurons could adjust and fi ne-tune opiate-related adaptations of the noradrenergic system. In conclusion, we show here that the NT-3/TrkC is an important regulator of neuronal fi ring in LC and would contribute to the adaptations of the noradrenergic system in response to chronic opiate exposure. These results may explain the perturbation of naloxone response in panic disorder patients and naltrexoneinduced panic attacks (Arntz et al., 1993;Maremmani et al., 1998;Sinha et al., 2007). It remains to be elucidated whether the molecular mechanisms that mediate NT-3/TrkC neuronal plasticity in the noradrenergic system beyond these opiate-induced adaptations resemble those that operate in the hippocampus, where the overexpression of TrkC markedly increases synaptic activity and LTP (Sahún et al., 2007a). We propose that TrkC may act as a putative genetic interface for the co-occurrence of panic disorder and substance abuse.

SUPPLEMENTARY MATERIAL
The Supplementary Material for this article can be found online at http://www.frontiersin.org/behavioralneuroscience/paper/10.3389. neuro.08/060.2009/  To assess whether physical withdrawal signs of TgNTRK3 mice were paralleled at the cellular level by increased activities of LC neurons, we recorded the fi ring rates of these neurons in slices from WT (black bars) and TgNTRK3 animals (white bars) treated with saline or morphine. Single-unit extracellular recordings were performed 2 h after slices preparation to allow morphine to wash out. Bars represent the mean ± SEM. *p < 0.05 when compared with the corresponding WT group; †p < 0.05 when compared with the corresponding saline group (one-way ANOVA followed by a Tukey's multiple comparison t test). (B) Concentrationeffect curves were constructed for the inhibitory effect of the µ-opioid receptor agonist ME on the LC, by adding increasing concentrations of the drug (0.1-100 µM, 3.3×) at 5-min intervals until a maximal effect was reached. The horizontal axis represents the logarithmic value of ME concentrations. The vertical axis expresses the percentage of reduction in the fi ring rate from the basal rate. Data are mean ± SEM. at each ME concentration. The lines are the theoretical curves in each group constructed from the mean of the individual concentration-effect curve parameters, as estimated by nonlinear regressions (see Materials and Methods for details and Table 3 for mean values). In all groups, the maximal effect of ME was 100% of baseline.