Abstract
The anxiolytic potential of the selective ς2 ligand 1-(4-fluorophenyl)-3-[4-[4-(4-fluorophenyl)-1-piperidinyl]-1-butyl]-1H-indole (Lu 28-179) was assessed in various animal models of anxiety in rodents. Lu 28-179 facilitated the exploratory behavior of mice and rats in the black and white two-compartment box over a large dose range. In the rat, the minimal effective dose (MED) was 0.18 nmol/kg (0.1 μg/kg), and in the mouse, the MED was 0.00018 nmol/kg (0.1 ng/kg). The anxiolytic-like effect was maintained after treatment with 1 μg/kg/day for up to 14 days, and no anxiogenic-like effects were seen upon withdrawal from repeated treatment. Lu 28-179 increased the time that pairs of rats spent in active social interaction (unfamiliar high-light conditions), MED = 0.1 ng/kg. Daily treatment with Lu 28-179 (1.8 nmol/kg = 1 μg/kg/day) for up to 4 weeks increased the social interaction time significantly compared with controls, and no anxiogenic-like effects were seen upon withdrawal. Furthermore, Lu 28-179 reversed shock-induced suppression of drinking in the rat (MED = 18,000 nmol/kg = 10 mg/kg). Lu 28-179 did not inhibit footshock-induced ultrasonic vocalization in the rat or isolation-induced aggressive behavior in the mouse. Lu 28-179 was over 100 times more potent than diazepam in the rat and mouse black and white test box and the rat social interaction test, whereas the potency of Lu 28-179 was comparable to that of lorazepam in reversal of shock-induced suppression of drinking. Lu 28-179 neither induced sedation nor impaired motor coordination, even at high doses (70,000 nmol/kg = 40 mg/kg). In conclusion, Lu 28-179 exerts potent and long-lasting anxiolytic-like effects in rodents without inducing sedation and withdrawal anxiogenesis.
The benzodiazepines remain the predominant pharmacotherapy for anxiety disorders. However, much effort is still directed at the development of new non-benzodiazepine anxiolytics (e.g., review byPerregaard et al., 1993). The reason for this is mainly the side effects related to treatment with benzodiazepines. There is a potential risk that benzodiazepine treatment leads to physical dependence, and withdrawal symptoms are described in relation to discontinuation of prolonged treatment (review by Woods et al., 1992). Sedation, especially during the first period of treatment, is another drawback of benzodiazepine treatment.
Lu 28-179 (fig. 1) is a novel ς ligand with subnanomolar affinity and specificity for the ς2binding site (Perregaard et al., 1995). Two subtypes of ς binding sites, the ς1 and the ς2 site, have been characterized, but the functions of these sites still remain to be clarified (Quirion et al., 1992). The ς binding sites have mainly been hypothesized to play a role in schizophrenia. A number of neuroleptics (e.g., haloperidol and remoxipride) have affinity for ς binding sites, and ς ligands are described as interacting with midbrain dopamine neurons (Walker et al., 1990; Monnet, 1993; Zhang et al., 1993). A number of ς-selective compounds have been suggested as antipsychotics on the basis of preclinical data; examples include DuP 734 and XJ 448 (Gilligan et al., 1992), NPC 16377 (Clissold et al., 1993; Karbon et al., 1993) and NE-100 (Okuyamaet al., 1993; Chaki et al., 1994). These compounds either are nonselective or show preference for the ς1 binding site. The ς2 binding site has been discussed in relation to dystonic effects in rats (Walker et al., 1993).
Limited evidence suggests involvement of ς binding sites in anxiety. The two ς ligands DTG and (+)-pentazocine show an anxiogenic-like profile in rats tested in a modified version of the Vogel conflict test (Lai et al., 1989). DTG is non-selective with respect to ς1 and ς2 binding sites, whereas (+)-pentazocine has preference for the ς1 binding site (Quirion et al., 1992).
In the present study, we assess the anxiolytic potential of a ς2 ligand (Lu 28-179) after single and repeated treatment in various models of anxiety in rodents. Preliminary data on some of the anxiolytic-like effects of Lu 28-179 were presented at the 24th Annual Meeting of the Society for Neuroscience (Sánchez et al., 1994). In addition, the in vitro binding profile and ex vivo binding data for Lu 28-179 are presented.
Materials and Methods
General animal housing conditions.
All rodents were housed at a temperature of 21 ± 2°C and had free access to water and standard rodent laboratory chow.
Receptor binding studies and amine reuptake inhibition in vitro.
ς2 binding site: Inhibition of 3H-DTG binding to ς2 binding sites in homogenates from rat brain minus cerebellum was determined as described by Meier et al. (1997). ς1 binding site: Inhibition of binding of 3H(+)-pentazocine to ς1 receptors in homogenates from rat brain minus cerebellum was determined as described by Meier et al.(1997). DA D1 receptors: Inhibition of3H-SCH 23390 binding to DA D1 receptors in rat striatal membranes was determined as described by Hyttel (1982).DA D2 receptors: Inhibition of3H-spiperone binding to DA D2 receptors in rat striatal membranes was determined as described by Hyttel (1987).Serotonin1A (5-HT) receptor: Inhibition of3H-8-OH-DPAT binding to 5-HT1A receptors in membranes from rat brain minus cerebellum was determined as described by Hyttel et al. (1988). 5-HT2Areceptors: Inhibition of 3H-ketanserin binding to 5-HT2A receptors in membranes from rat cortex was determined as described by Hyttel et al. (1988).5-HT2C receptors: Inhibition of3H-mesulergine (0.5 nM) binding to a cloned rat 5-HT2C receptor expressed in 3T3 cells was determinedin vitro as described by Bøgesø et al. (1995).Alpha-1 adrenoceptors: Inhibition of3H-prazosin binding to alpha-1 adrenoceptors in membranes from rat whole brain was determined as described by Hyttel and Larsen (1985). Alpha-2 adrenoceptors: Inhibition of 3H-idazoxan binding to alpha-2 adrenoceptors in membranes from rat brain cortex was determined as described by Megens et al. (1986). Beta adrenoceptors: Inhibition of 3H-dihydroalprenolol binding to beta adrenoceptors in membranes from rat cortex was determined as described by Hyttel et al. (1984).Histamine H1 receptors: Inhibition of3H-mepyramine binding to histamine H1 receptors in membranes from rat brain minus cerebellum was determined as described by Hall and Ögren (1984). Muscarine cholinergic receptors: Inhibition of 3H-QNB binding to muscarinic cholinergic receptors in membranes from rat whole brain was determined as described by Meier et al. (1997). Inhibition of3H-DA uptake in rat striatal synaptosomes was determined as described by Hyttel (1982). Inhibition of3H-noradrenaline (NA) uptake in rat cortical synaptosomes was determined as described by Hyttel (1982).Inhibition of 3H-5-HT uptake in synaptosomes from rat brain minus cerebellum was determined as described by Hyttel (1982).
Estimation of ς2 binding site activity in the CNS using ex vivo binding.
Ex vivo binding studies were performed essentially as described by Freedman et al. (1989). Briefly, mice (NMRI, Bomholtgård, Denmark) were given test compound either s.c. or p.o., and were sacrificed after 3 h (Lu 28-179) or after 0.5 h (DTG or haloperidol). The forebrain was quickly removed and homogenized in 100 vol of assay buffer (50 mM Tris-buffer, pH 7.4). Binding activity (3H-DTG) was estimated on the forebrain homogenate, and ς2 binding site activity of the compound was quantified by the relative inhibition of binding compared with saline-treated control animals. For absorption and elimination experiments, mice were given 3.3 μmol/kg Lu 28-179 s.c. At different time intervals after drug administration, mice were decapitated, and ex vivo binding (3H-DTG) was determined in brain homogenates and expressed as 100 minus percent binding compared with saline-treated controls. Absorption and elimination data were fitted to the equation F(t) = F0 exp(−ka t) + exp(−ket), where F0 is the total binding activity, ex vivo, extrapolated to time 0,ka and ke are the absorption respective elimination constants and t is the time in days.
Black and white test box, rats.
We used male Wistar WU rats (Charles River, Hannover, Germany) housed in groups of four in macrolon type III cages and kept on a reversed light/dark cycle (dark period 7.00–19.00 h). The two-compartment test model is formerly described bySánchez et al. (1995). A white, open-topped compartment (80 cm × 65 cm × 33 cm) with the floor divided into nine squares was connected to a closed black box (25 cm × 21 cm) by an opening (10 cm × 10 cm) located in the center of one of the short sides of the white compartment. The white compartment was illuminated by bright light (2000 Lux). The test system was fully automated by 2 rows of 11 infrared light sources and photocells in the transverse direction and 1 row of 18 in the longitudinal direction (lower row). The lower row of photocells (5.5 cm above the cage floor) detected horizontal locomotor activity, whereas the upper row of photocells (13 cm above the cage floor) detected rearing activity.
The rats were taken from the dark holding room in a dark container to a dimly illuminated (red light) room. Treatment with test substance was performed after a 1 to 2-h period of adaptation. During the test, individual animals were placed in the center of the white compartment and observed for 7 min for number of rearings and number of line crossings between squares in the two compartments, number of entries into the black compartment and time spent in the white compartment. Each treatment group consisted of 8 to 24 rats. The dose-response relationship for single doses of Lu 28-179 was assessed, along with the duration of action of a single dose (1 μg/kg) of Lu 28-179.
Black and white test box, mice.
We used male BKW mice (Bradford strain, University of Bradford, UK) weighing 30 to 36 g, housed in groups of 10 in polypropylene cages and kept on a reversed light/dark cycle (dark period 7.00–19.00 h).
The test box is described in detail by Costall et al.(1989). Briefly, the box (45 cm × 27 cm × 27 cm) was open-topped, and the base was lined into 9-cm squares. Two-fifths of the box was painted black, illuminated under a dim red light and partitioned from the remainder of the box, which was painted white and brightly illuminated. The compartments were connected by an opening 7.5 cm × 7.5 cm located at floor level in the center of the partition.
The mice were taken from the dark holding room in a dark container to a dimly illuminated (red light) room. Treatment with test substance was performed after a 1 to 2-h period of adaptation. During the test, individual animals were placed in the center of the white compartment, and their behavior was recorded by remote video recording for 5 min. The recordings were subsequently evaluate, by a person unaware of the drug treatments, for time spent in each compartment, number of exploratory rearings in each compartment, number of crossings between squares in each compartment, and latency of the initial movement from the white to the black compartment. Each treatment group consisted of 5 to 10 mice. We assessed the dose-response relationship for single doses of Lu 28-179, as well as the effects of daily treatment with 0.0018 μmol/kg (1 μg/kg i.p.) for 3, 7, 10 or 14 days. Furthermore, separate groups of animals treated with Lu 28-179 daily for 14 days were assessed 8 h, 12 h and 1, 2, 4 and 10 days after withdrawal of treatment.
Social interaction test.
We used male Hooded-Lister rats (Bradford strain, University of Bradford, UK) weighing 225 to 275 g, housed in groups of five in polypropylene cages and kept on a 12-h light/dark cycle with lights on at 7.00 h.
The apparatus was an open-topped perspex box (51 cm × 51 cm × 20 cm) with 17 × 17 cm areas marked on the floor. The box was illuminated with bright white light. Tests were conducted in an illuminated room using a methodology based on the model of File (1980). Two naive rats, from separate housing cages, were both treated with test drug and placed into the brightly illuminated test box. Their behavior was recorded by remote video recording over a 10-min period. Social interaction between animals (sniffing of partner, crawling under or climbing over partner, genital investigation of partner and following partner) was determined by timing in seconds, and exploratory locomotion was measured as the number of crossings of the lines marked on the floor of the test box. The video recordings were evaluated by a person unaware of the drug treatments. A total of 6 to 12 pairs of rats were used per dose.
We assessed the dose-response relationship for single doses of Lu 28-179, as well as the effects of daily treatment with 0.0018 μmol/kg (1 μg/kg) for 1, 2, 3 and 4 weeks, respectively. Furthermore, separate groups of animals treated with Lu 28-179 for 4 weeks were assessed 24 h, 48 h and 7, 10 and 14 days after withdrawal of treatment.
Inhibition of shock-induced suppression of drinking.
We used male Wistar rats (Mol:Wist, Møllegård, Denmark) that weighed 150 to 175 g at the beginning of the study. The rats were housed in groups of four in macrolon type III cages and were kept on a 12-h light/dark cycle with lights on at 7.00.
The test chambers were stainless steel boxes (13 cm × 21 cm × 14 cm) with wire mesh floor and walls. A water bottle with a metal tube was mounted on one side of the chamber. The chambers were contained in a sound-attenuating cabinet. Shocks were provided by a two-pole shocker connected to a metal drinking tube and the floor of the box. A shock was delivered every 20 licks, or, at constant licking, a shock was delivered every 7 sec (equivalent to approximately 20 licks). The numbers of shocks and licks were recorded by means of a computer program.
Rats were deprived of water for 24 h before the first training session. Then the drinking behavior was assessed in a nonpunished 6-min session. Only rats that licked during this session progressed to a further 24-h deprivation of water. The rats were tested again 24 h later, this time receiving a shock for every 20 licks. The trial session time was 6 min, starting automatically when the rat completed 20 licks and received the first shock. If a rat had not licked after the first 3 min, the recording started automatically, thus resulting in a maximum duration of 9 min for a test session. The shock intensity was 0.6 mA, and the duration of the shock was 0.1 sec. Immediately after the second session, drugs were administered and the punished session was repeated 1 h or 2 h later. Each treatment group consisted of 8 to 24 animals.
Isolation-induced aggression.
We used male NMRI mice (BOM, SPF, Møllegård, Denmark) with starting weight 18 to 20 g. The isolated mice were single-housed in Macrolon type II cages, and the socially housed intruder mice were kept in groups of 10 in opaque plastic cages (35 cm × 30 cm × 12 cm). Lights were on 7.00 to 19.00 h.
The test was conducted as described by Sánchez et al.(1993). Briefly, the mice were kept isolated for about 21 days. After the isolation period the mice were trained to attack, initially by introducing a single housed intruder into the test cage, and subsequently by means of a nonaggressive intruder mouse. An attack was defined as biting or attempting to bite the intruder mouse. The training and the testing sessions took place in the home cage of the isolated mouse. Only mice with attack latencies of less than 10 sec were included in the studies. In the test sessions, the mice were tested immediately before drug treatment and 30 min or 2 h later. The attack latency was measured with a maximum observation time of 180 sec. Each group consisted of 8 to 16 aggressive mice and 16 to 32 nonaggressive mice (for testing before and after drug).
Inhibition of footshock-induced ultrasonic vocalization.
We used male Wistar WU rats (Charles River, Hannover, Germany) that weighed 150 to 175 g at the beginning of the study. The rats were housed in groups of four in macrolon type III cages and lights were on 7.00 to 19.00 h.
The test cages and procedure are described in detail by Sánchez (1993). Briefly, the test cages (22 cm × 22 cm × 22 cm) were made of grey perspex and equipped with a metal grid floor. Footshocks were delivered from a two-pole shocker. A microphone sensitive to ultrasounds in the range of 20 to 30 kHz was placed in the center of the lid of the test cages. The ultrasounds were sent from the microphone to a preamplifier and converted from AC signals to DC signals in a signal rectifier. The accumulated time when the voltage of the rectified signal was larger than the voltage of a previously determined threshold level was recorded.
Test procedure: Briefly, 24 h before the first test session, the animals were primed. The rat was placed in the test cage and received immediately thereafter four inescapable 1.0-mA footshocks, each lasting 10 sec with an intershock interval of 5 sec. The animals were left in the test cages for a total of 6 min after the last shock. On the test day, rats received the same shock regimen, and recording of ultrasonic vocalization started 1 min after the last shock and lasted for 5 min. The total time spent vocalizing was recorded. Each treatment group consisted of 8 to 16 animals.
Motor side effects.
We used male NMRI mice (BOM, SPF, Møllegård, Denmark) that weighed 18 to 25 g (locomotor activity) or 25 to 35 g (horizontal wire test) and were kept in groups of 10 to 15 in opaque plastic cages (35 cm × 30 cm × 12 cm). Lights were on 7.00 to 19.00 h.
Spontaneous locomotor activity was measured in automated activity cages. Each cage consisted of an open transparent perspex arena (20 cm × 32 cm × 20 cm) placed on a black screen with 40 photocells. A light source was placed above the cage, and the number of light-beam interruptions was recorded. Groups of three mice were treated with test substance or saline and, 1 h later, placed groupwise in the activity cages. The number of light-beam interruptions was recorded for 15 min. Each dose level was assessed in 2 to 4 activity cages, and the locomotor activity was expressed as percent activity relative to the activity of animals from parallel saline-treated control groups.
Performance on the horizontal wire test was assessed in the following way: The mouse was lifted by the tail and allowed to grasp with its forepaws a horizontally strung wire (35 cm long and 0.55 mm in diameter) 30 cm above the table level; it was then released. Animals that were unable to grasp the wire with the hindpaws within 15 sec were scored as responders (all-or-none criterion). A total of 3 to 6 animals were used per treatment group.
Statistics.
In the in vitro studies, IC50 values were estimated on the basis of two full concentration-response curves, each containing five concentrations of drug covering 3 to 5 log units (each data point in triplicate). Estimation of IC50 values was performed by means of the receptor program Ligand and/or Multicalc from Wallac. If the log ratio (logR) between the two determinations was greater than corresponding to 3 × S.D. (99% confidence interval), then extra determinations were performed and outliers discarded. The S.D. values were calculated from a series of n determinations of logR between double determinations. Antilog (S.D.) values applied for the individual binding assays are shown in table 1.
In vitro receptor binding. IC50 values were estimated on the basis of two full concentration-response curves, each containing five concentrations of drug covering 3 to 5 log units (each data point in triplicate).
In the in vivo studies, one-way analysis of variance was used, followed by post-hoc comparisons of means (Dunnett’s test) when the outcome of the analysis of variance was significant (P < .05). Inhibitory potencies on footshock-induced ultrasonic vocalization, isolation-induced aggression, locomotor activity and performance on the horizontal wire test were expressed as ED50 values, calculated by means of log-probit analysis.
Drugs.
Lu 28-179 hemifumarate or oxalate, mw 571 or 545, respectively (H. Lundbeck A/S, Copenhagen, Denmark), 10 mg, was dissolved in a mixture of 1 ml propyleneglycol and 0.2 ml 0.1 M methanesulphonic acid by heating. The solution was diluted by slowly adding distilled water (40–50 parts). Diazepam, mw 285 (Roche, Basel, Switzerland) was dissolved in a minimum amount of polyethylene glycol and diluted into saline or provided as ampules (Apozepam, Apothekernes Laboratorium A/S, Oslo, Norway) and diluted with saline. Lorazepam, mw 320 (Ferrosan, Copenhagen, Denmark) was suspended in 0.5% methylcellulose saline dissolution. Haloprendol (Tanson, Belgium) was dissolved in minimum amounts of tartaric acid and diluted with saline. DTG (Tocris Cookson, UK) was dissolved in 96% alcohol and propylene glycol and diluted with water.
The injection volumes were 10 ml/kg in mice, and 5 ml/kg (black and white box test, footshock-induced ultrasonic vocalization and conflict test) or 1 ml/kg (social interaction) in rats. Treatment times, doses and routes of administration are given with the results of the individual studies.
Results
In vitro receptor binding and amine reuptake inhibitory potency.
Lu 28-179 had very high affinity for ς2sites labeled by 3H-DTG (table 1). In contrast, the affinities for the various receptor sites, including the site labeled by 3H-(+)-pentazocine, were much lower (table 1). The reference compounds DTG, haloperidol and (+)-3-PPP had considerably lower (90 to several thousand-fold) affinities for the ς2binding site.
Lu 28-179 had negligible effects on the uptake of labeled amines into synaptosomes compared with their affinity for ς sites (table 1).
Ex vivo receptor binding.
Lu 28-179 easily penetrated the blood-brain barrier and reached its maximum value within 1 to 2 h after administration (fig.2). It was slowly eliminated from the brain. The elimination constant was estimated toke = 0.838 ± 0.0423 day−1, from which T1/2 = 0.827 days, or approximately 20 h, was calculated. Maximum inhibition of binding was seen 2 to 3 h after administration, but 60 to 70% inhibition was seen at 15 to 60 min after administration.
Absorption and elimination of Lu 28-179 (3.3 μmol/kg) from murine brain after s.c., p.o. and i.v. administration as estimated by 3H-DTG ex vivo binding. Values represent the mean ± S.E.M. of five determinations.
Lu 28-179 was effective after s.c. and p.o. administration, haloperidol displaced 3H-DTG at high doses and DTG was ineffective (table 2).
3H-DTG ex vivo binding studies in mice. For details, see “Materials and Methods”
Black and white test box, rats and mice.
The effects of Lu 28-179 and diazepam are shown in figures 3 and 4 for the rat and for the mouse, respectively.
The effect of Lu 28-179 (s.c. 2 h before test) and diazepam (0.035 μmol/kg, s.c. 30 min before test; hatched bars) on rat rearing behavior, line crossings, time spent in the white compartment and number of entries into the black part of a two-compartment black and white test box. C indicates the response of vehicle-treated control animals. Values represent the mean ± S.E.M. The numbers of animals per group were 24 controls; 8, 16, 16, 16, 8 and 8 Lu 28-179-treated rats; 16 diazepam-treated rats. Significant increases or decreases in responding compared with control values are indicated by * (P < .05; one-way analysis of variance followed by Dunnett’s t test).
The effect of Lu 28-179 (i.p. 40 min before test; double-hatched bars) and diazepam 4.6 μmol/kg = 1.3 mg/kg (i.p. 40 min before test; hatched bars) on mouse rearing behavior, line crossings, the percentage of time spent in the white compartment and latency of first movement from the white to the black part of a two-compartment black and white test box. C (open bars) indicates the response of vehicle-treated control animals. Values represent the mean ± S.E.M. The numbers of animals were 10 controls,n = 5 in each of the Lu 28-179-treated groups, and 10 diazepam-treated. Significant increases or decreases in responding compared with control values are indicated by * (P < .05; one-way analysis of variance followed by Dunnett’s t test).
In the rat, treatment with Lu 28-179 significant increased exploratory rearing in the white compartment without affecting the rearing activity in the black compartment (fig. 3). Similarly, the number of entries into the black compartment and the time spent in the white compartment increased significantly. The potency of Lu 28-179 was high, with a MED of 0.00018 μmol/kg (0.0001 mg/kg s.c., 2 h before test), and Lu 28-179 was active over a wide dose range, i.e., 0.00018 to 1.8 μmol/kg (0.0001–1.0 mg/kg). Diazepam 0.035 μmol/kg (0.01 mg/kg s.c., 30 min before test) induced a significant increase in number of rearings, line crossings and time spent in the white compartment.
In the mouse, treatment with Lu 28-179 (i.p., 40 min before test) caused a significant increase in rearings and line crossings in the white compartment and a corresponding significant reduction in the black compartment (fig. 4). Similarly, time spent in the black compartment decreased significantly. The potency of Lu 28-179 was extremely high, with a MED of 0.00018 nmol/kg (0.10 ng/kg), and Lu 28-179 was active over a very large dose range,i.e., 0.00018 nmol/kg to 1800 nmol/kg (0.1 ng/kg–1.0 mg/kg). Diazepam 4.6 μmol/kg (1.3 mg/kg i.p., 40 min before test) induced a significant increase in number of rearings and line crossings and time spent in the white compartment (fig. 4).
Lu 28-179 (0.0018 μmol/kg i.p.) administered daily for 1, 3, 7, 10 or 14 days to separate groups of mice, significantly increased number of rearings and line crossings and time spent in the white compartment (fig. 5, left half of each graph). A similar anxiolytic-like effect was obtained after 10 days of treatment with diazepam 4.6 μmol/kg/day (1.3 mg/kg/day i.p., test 40 min after the last dose). Groups of mice treated with Lu 28-179 (0.0018 μmol/kg = 1 μg/kg/day i.p.) for 14 days maintained an increased level of rearings and line crossings and time spent in the white compartment, 1 to 2 days after withdrawal from treatment (fig. 5, right half of each graph). At 4 and 10 days after withdrawal, the behavior was indistinguishable from that of control animals. A group treated daily with diazepam for 14 days and assessed 24 h after the last dose showed a significant decrease in number of rearings and line crossings and in time spent in the white compartment.
Left half of graphs) The effects of treatment with Lu 28-179 (0.0018 μmol/kg = 1 μg/kg/day i.p.; double-hatched bars) for 1, 3, 7, 10 or 14 days or of treatment with diazepam 4.6 μmol/kg (1.3 mg/kg/day i.p.) for 10 days (hatched bars) on mouse exploratory behavior in a two-compartment black and white test box. C (open bars) indicates the response of control animals treated with vehicle for 14 days. The mice were tested 40 min after the last injection. Right half of graphs) The effects of Lu 28-179 0.0018 μmol/kg/day (1 μg/kg/day i.p.) (double-hatched bars) or of diazepam 18 μmol/kg (5 mg/kg/day) for 14 days (hatched bars) on mouse exploratory behavior in a two-compartment black and white test box 8 h, 12 h or 1, 2, 4 or 10 days and 24 h after withdrawal from treatment with Lu 28-179 and diazepam, respectively. Each value is the mean ± S.E.M. of five determinations. Significant increases or decreases in responding compared with control values are indicated by * (P < .05; one-way analysis of variance followed by Dunnett’s t test).
Social interaction in the rat.
Lu 28-179 increased the time that pairs of rats spent in active social interaction over a wide dose range without modifying locomotor activity (fig.6). Lu 28-179 was extremely potent; MED = 0.00018 nmol/kg (0.1 ng/kg i.p., 40 min before test). The number of line crossings was not affected by Lu 28-179 at the highest dose tested, 1.8 μmol/kg (1.0 mg/kg). Diazepam 4.6 μmol/kg (1.3 mg/kg i.p., 40 min before test) induced a maximum increase in social interaction of the same magnitude as that induced by treatment with Lu 28-179.
The effect of Lu 28-179 (i.p. 40 min before test; double-hatched bars) and diazepam 4.6 μmol/kg = 1.3 mg/kg (i.p. 40 min before test; hatched bars) to increase rat social interaction under high-light unfamiliar conditions. C (open bars) indicates the response of vehicle-treated control animals. Pairs of rats received acute drug or vehicle administrations. Each value is the mean ± S.E.M. The numbers of animals were 12 pairs of control rats,n = 6 pairs in each of the Lu 28-179-treated groups, and 12 diazepam-treated pairs. Significant increases or decreases in responding compared with control values are indicated by * (P < .05; one-way analysis of variance followed by Dunnett’s ttest).
Daily treatment for 7, 14, 21 or 28 days with 0.0018 μmol/kg (1 μg/kg) significantly increased the time spent in active social interaction compared with the behavior of vehicle-treated animals (fig.7, left half). Diazepam treatment (35 μmol/kg = 10 mg/kg) for 7 days increased the time that pairs of rats spent in active social interaction. However, locomotor activity was significantly decreased.
Left) The effect of treatment with Lu 28-179 (0.0018 μmol/kg/day = 1 μg/kg/day i.p.) for 1, 2, 3 or 4 weeks (double-hatched bars) or of treatment with diazepam (35 μmol/kg/day = 10 mg/kg/day i.p.) for 7 days (hatched bar) on the time that pairs of rats spend in social interaction under high-light unfamiliar conditions. C (open bar) indicates the response of control animals treated with vehicle for 4 weeks. Pairs of rats were tested 40 min after the last injection. Right) The effect of Lu 28-179 (0.0018 μmol/kg/day = 1 μg/kg/day i.p. × 1) on rat social interaction 24 h or 2, 7, 10 or 14 days after withdrawal of a 4-week treatment with Lu 28-179 (double-hatched bars). The effect of diazepam (35 μmol/kg/day = 10 mg/kg/day i.p.) on rat social interaction 24 h after withdrawal of a 1-week treatment with diazepam is also shown (hatched bar). Each value is the mean ± S.E.M. of six determinations. Significant increases or decreases in responding compared to control values are indicated by * (P < .05; one-way analysis of variance followed by Dunnett’s t test).+Significantly different from day 7 (t test).
Rats treated for 4 weeks with Lu 28-179 0.0018 μmol/kg/day (1 μg/kg/day) maintained a significantly increased response 24 h after withdrawal from treatment (fig. 7, right half). The response recorded at 2, 7, 10 or 14 days after withdrawal was comparable to vehicle responses. A group of rats treated with diazepam 35 μmol/kg (10 mg/kg) for 7 days showed a significant decrease in time spent in social interaction, without change in number of line crossings, 24 h after withdrawal (fig. 7).
Inhibition of shock-induced suppression of drinking.
Lu 28-179 significantly increased the number of shocks received during a 6-min test period (fig. 8). The effective doses were much higher than those necessary to achieve anxiolytic-like effects with Lu 28-179 in the aforementioned models (MED = 18 μmol/kg = 10 mg/kg s.c., 2 h before test), but the doses were comparable to the effective doses of the benzodiazepine lorazepam (fig. 8). Furthermore, the maximum effects obtained with Lu 28-179 and lorazepam were of similar magnitude.
The effect of Lu 28-179 and lorazepam on shock-induced suppression of drinking behavior in rats. Rats were water deprived for 48 h. Subsequently, drinking behavior was assessed while rats received a shock (0.6 mA for 0.1 sec) for every 20 licks on the metal drinking tube. Immediately after this session, Lu 28-179 (s.c.) or lorazepam (p.o.) was administered, and the punished session was repeated 2 h or 1 h later, respectively. Each value is the mean ± S.E.M. The numbers of animals per group were 24 controls; 8, 8, 24, 16 and 8 Lu 28-179-treated rats; 8, 16, 16, 8 and 8 lorazepam-treated rats. Significant increases in responding compared with control values are indicated by * (P < .05; one-way analysis of variance followed by Dunnett’s t test).
Inhibition of footshock-induced ultrasonic vocalization in the rat and inhibition of isolation-induced aggression in the mouse.
Lu 28-179 did not inhibit footshock-induced ultrasonic vocalization in the rat, and Lu 28-179 did not inhibit isolation-induced aggression in male mice (0.01–20 mg/kg s.c., 2 h before test; table3). The effects of diazepam are included in the table for comparison.
Effects of Lu 28-179 and diazepam on footshock-induced ultrasonic vocalization in the rat and isolation-induced aggressive behavior of male mice. For details, see “Materials and Methods”
Motor effects.
In mice, Lu 28-179 neither inhibited spontaneous locomotor activity nor affected motor performance on a horizontal wire test, even at very high doses (table4). Diazepam significantly decreased spontaneous locomotor activity and significantly impaired performance on the horizontal wire test (table 4).
Effects of Lu 28-179 and diazepam on spontaneous locomotor activity and performance on a horizontal wire in mice. For details, see “Materials and Methods”
Discussion
The in vitro binding studies demonstrated that Lu 28-179 is a highly selective and very potent ς2 ligand compared with other ligands that have affinities for the ς2 binding site. Furthermore, the ex vivobinding studies suggested that Lu 28-179 penetrates the blood-brain barrier easily, that it is slowly eliminated with a half-life of about 20 h and that the compound is effective after s.c. as well as p.o. administration. The lack of effect of DTG in the ex vivobinding experiments may be due to poor CNS penetration of DTG. However, it should be emphasized that the tissue is diluted in 100 vol of buffer. Thus the ex vivo binding potency may underestimate the actual binding affinity as a result of dissociation of the test compound from its binding sites during incubation. For this reason, comparisons of potency between different compounds should be made with caution.
The present data demonstrate an anxiolytic-like profile of Lu 28-179 in a number of rodent models of anxiety. Lu 28-179 had an extremely high potency in the rat social interaction test and in the mouse black and white box test, and it also had a very potent effect in the rat black and white box test. The MED ranged from 0.01 to 100 ng/kg in these tests. It should be stressed that these high potencies are not reflected in the ex vivo binding studies. Methods for measuring such low drug concentration in tissue are presently not available. Furthermore, the functional correlate to stimulation of thesigma binding site is unknown. The anxiolytic-like effect had also a very long duration of action, as shown by the anxiolytic-like effects in the mouse black and white test box and the rat social interaction test persisting for 1 to 2 days upon withdrawal after 2 and 4 weeks of treatment, respectively. Furthermore, a single injection (1 μg/kg s.c.) of Lu 28-179 produces significantly anxiolytic-like effects in the rat black and white test box 24 h after administration (C. Sánchez, unpublished observations). The high potency might suggest that the anxiolytic-like effects are mediated by ς binding sites, because affinities for a variety of other receptors are at least 1000 times weaker. Lu 28-179 has recently been described as a ς ligand with selectivity for the ς2 binding site (Perregaard et al., 1995). Thein vitro affinities for the ς1 binding site (3H-(+)-pentazocine) and for the ς2 binding site (3H-DTG) are IC50 = 17 nM and IC50 = 0.19 nM, respectively, which corresponds to a ratio of about 90. This might suggest a role for the ς2 binding site in anxiety. However, we cannot exclude the possibility that the ς1 binding site is involved in the anxiolytic-like effects. Thus the high ς1/ς2 selectivity ratio is due to an extremely high affinity for the ς2binding site in the subnanomolar range, whereas the affinity for the ς1 binding site is in the nanomolar range. The ς1 binding site has formerly been associated with an anxiogenic-like effect; the ς1 ligand (+)-pentazocine induced an anxiogenic-like effect in a conflict model in rats (Laiet al., 1989). However, in the rat black and white test box (+)-pentazocine is found to induce an anxiolytic-like effect [effective doses from 0.035 to 3.5 μmol/kg (0.01–1 mg/kg)] without affecting locomotor activity (C. Sánchez, unpublished observations). Because the functional characterization of ς ligands with respect to intrinsic activity is hampered by the lack of appropriate models, additional information about the ς subtypes involved in the action of Lu 28-179 remains to be provided.
Lu 28-179 was about 100 times more potent than diazepam in the rat black and white test box and several thousand-fold more potent than diazepam in the mouse black and white test box and the rat social interaction test. Diazepam was effective at doses from 0.035 μmol/kg (0.010 mg/kg) in the former and from 0.46 μmol/kg (0.13 mg/kg) in the latter (Sánchez et al., 1995). Lu 28-179 does not have affinity for benzodiazepine receptors (E. Meier unpublished observation). Also, 5-HT1A receptor agonists (e.g., buspirone) show potent effects in this model (C. Sánchez, unpublished observations). However, the effect of Lu 28-179 can be explained by effects on neither 5-HT1A nor 5-HT3 receptors, because Lu 28-179 is devoid of affinity for these receptors. CCK receptors are also involved in mediating anxiety (review by Harro et al., 1993), and CCK antagonists facilitate the exploratory behavior in the mouse black and white test box (Costall et al., 1991; Hendrie et al., 1993). However, Lu 28-179 does not have affinity for either CCKAor CCKB receptors (E. Meier, unpublished observation).
The neurobiological mechanisms involved in mediation of the anxiolytic-like activity of Lu 28-179 are not known. Both ς1 and ς2 binding sites are widely distributed in the brain, particularly in the limbic system (e.g., hippocampus and amygdala) and in brain stem motor areas (Gundlach et al., 1986). The ς1 and ς2 binding sites differ in distribution; the former is at its highest level in the brain stem, the latter in hippocampal membranes (McCann et al., 1994). This might suggest that the ς2 binding sites are involved in modulation of emotional responses. However, a recent study suggests that ς1 sites are involved in mediation of conditioned fear stress (Kamei et al., 1996).
Lu 28-179 was also active against shock-induced suppression of drinking, which in the present form is a modification of the classical Vogel conflict test of anxiety (Petersen and Lassen, 1981). The effective doses of Lu 28-179 were much higher than those necessary to achieve anxiolytic-like effects in the black and white test box and the social interaction test, but the doses were comparable to the effective doses of the benzodiazepine lorazepam (fig. 9). Furthermore, the maximum effects obtained with Lu 28-179 and lorazepam were of similar magnitude. This is quite remarkable, because the responses of non-benzodiazepine anxiolytics in conflict models are generally variable and smaller than those of benzodiazepines (Perregaard et al., 1993). The model-dependent and very large differences among the MEDs of Lu 28-179 are not readily explained, although some differences in sensitivity are observed with diazepam, too. For example, reports from different laboratories on the MEDs in the mouse black and white test box range from 0.01 mg/kg (Onaivi and Martin, 1989) to 1.0 mg/kg (Young and Johnson, 1991). The intensity or type of stressor in anxiogenic stimuli may interfere with the MEDs needed. A recent study of the mouse black and white test box showed that tail suspension for 5 min immediately before test enhanced the anxiolytic-like response diazepam significantly (Sánchez, in press).
On the other hand, Lu 28-179 did not inhibit footshock-induced ultrasonic vocalization in adult rats, another paradigm suggested as a model of anxiety (Cuomo et al., 1988; Sánchez, 1993;De Vry et al., 1993). Like the conflict model, this model applied a relatively stressful secondary aversive stimulus. But no effect was observed with Lu 28-179, even at very high doses. Anxiolytics such as the 5-HT1A receptor agonist buspirone are much more effective inhibitors of footshock-induced ultrasonic vocalization in the rat than diazepam (Sánchez, 1993), whereas 5-HT1A receptor agonists have no or rather weak effects in the present version of the conflict test (C. Sánchez, unpublished observation). This model dependence of anxiolytic activity is a rather common finding for non-benzodiazepine anxiolytics (Lister, 1990; File, 1992; Perregaard et al., 1993). The type of stressor in the anxiogenic stimuli may interfere with the type of response in the different test models. Different neuronal pathways have been shown to be involved in mediating the behavioral responses to different anxiogenic stimuli. For example, injection of a 5-HT1Areceptor agonist in the basolateral part of the amygdala enhanced the anxiety response in the social interaction test but not in the elevated plus maze, whereas midazolam induced an anxiolytic response in the social interaction test but not in the elevated plus maze (Gonzalezet al., 1996). Another study demonstrated an anxiolytic response in the elevated plus maze after midazolam injection into basolateral but not central amygdala, whereas the opposite result was observed in the shock-probe burying test (Pesold and Treit, 1995). Furthermore, a recent study of different types of fear in the elevated T-maze suggested that stimulation of the ascending serotonergic pathway facilitated learned fear whereas it inhibited unconditioned fear (Graeff et al., 1996). Perhaps Lu 28-179 is effective only against a specific subtype of anxiety. However, this is pure speculation, because no clear relation between the different clinical diagnoses of anxiety and the different types of animal models has been established so far.
The anxiolytic-like responses of Lu 28-179 were observed from doses far below nmol/kg up to 1.8 μmol/kg, and no sedation was induced by any of these doses. In contrast, diazepam induced sedation and motor impairment. In the mouse two-compartment black and white test box, the anxiolytic-like profile was indistinguishable for mice treated daily for up to 2 weeks with Lu 28-179 and mice treated with a single dose. Furthermore, unlike the results with diazepam-treated mice, no anxiogenic-like effects were observed upon withdrawal after 2 weeks of treatment with Lu 28-179. In the rat social interaction test, a significant anxiolytic-like effect was seen after up to 4 weeks of daily treatment with Lu 28-179. The effect after 4 weeks of treatment was significantly lower than after 3 weeks but was similar to the effect after 1 week. The possibility that tolerance starts to develop after 4 weeks cannot be excluded, but this is unlikely because no signs of withdrawal-induced anxiogenesis were detectable after 4 weeks of treatment with Lu 28-179. The lack of sedative effects and withdrawal anxiogenesis after treatment with Lu 28-179 may indicate a more favorable side effect profile than that of the benzodiazepines.
In conclusion, Lu 28-179 exerts potent and very long-lasting anxiolytic-like effects in animal studies using mice and rats. Future clinical studies will show whether this new class of compounds is effective in relieving anxiety without inducing the unwanted effects of the benzodiazepines.
Acknowledgments
We thank Karin Larsen, Lone Hanne Petersen, Dorit Skov (H. Lundbeck A/S), Deborah Murphy and Ben Grayson (Postgraduate Studies in Pharmacology, University of Bradford) for excellent technical help, and Hanne Albertsen for typing the manuscript.
Footnotes
-
Send reprint requests to: C. Sánchez, Pharmacological Research, H. Lundbeck A/S, Ottiliavej 9, DK-2500 Copenhagen-Valby, Denmark.
-
↵1 Present address: C.S., J.A. and E.M.: Pharmacological and J.P.: Medicinal Chemistry Research, H. Lundbeck A/S, Ottiliavej 9, DK-2500 Copenhagen-Valby, Denmark.
-
↵2 B.C., M.E.K. and R.J.N.: Postgraduate Studies in Pharmacology, University of Bradford, Bradford West Yorkshire BD7 1DP, England.
- Abbreviations:
- Lu 28-179
- 1-(4-fluorophenyl)-3-[4-[4-(4-fluorophenyl)-1-piperidinyl]-1]butyl]-1H-indole
- MED
- minimal effective dose
- DTG
- 1,3-di-o-tolylguanidine
- DA
- dopamine
- 5-HT
- serotonin, 8-OH-DPAT, 8-hydroxy-2-(di-n-propylamino)tetraline
- QNB
- quinyclidinyl benzilate
- NA
- noradrenaline
- mw
- molecular weight
- 3-PPP
- N-n-propyl-3-(3-hydroxyphenyl)-piperidine
- CCK
- cholecystokinin
- Received March 19, 1996.
- Accepted August 4, 1997.
- The American Society for Pharmacology and Experimental Therapeutics
