Gaboxadol ��� a different hypnotic profile with no tolerance to sleep EEG and sedative effects after repeated daily dosing Bjarke Ebert a,���, Neil J. Anderson a, Thomas I. Cremers c, Stine Rasmussen d, Vanessa Vogel a, Jeanne M. Fahey d, Connie S��nchez b a H. Lundbeck A/S, Ottiliavej 9, DK-2500 Valby, Denmark b Lundbeck Research US, Paramaus, NJ, USA c Brains-on-Line, Antonius Deusinglaan 1, 9713 AV Groningen, The Netherlands d Department of Pharmacology & Experimental Therapeutics Tufts University School of Medicine, Boston MA, USA Available online 5 February 2008 Abstract Gaboxadol, a selective extra synaptic GABAA receptor agonist, has been in clinical development for the treatment of insomnia. Development of tolerance to therapeutic effects (e.g. hypnotic and anticonvulsant and sedative) and withdrawal symptoms (e.g. REM sleep rebound and reduced seizure threshold) upon treatment discontinuation is reported for GABAA receptor allosteric modulators acting via the benzodiazepine binding site, e.g. zolpidem and indiplon. We conducted a head to head comparison in rats of the hypnotic (sleep EEG after 21 daily doses and 24 and 48 h after the last dose) and seizure threshold modifying (bicuculline assay 24 h after 28 daily doses) effects of gaboxadol and benzodiazepine ligands. Furthermore, we investigated in further details a previously reported apparent rapid development of tolerance to gaboxadol's effects in a rat rotarod motor coordination assay and related this effect to CNS exposure levels and in vitro potency at extra synaptic GABAA receptors. Sleep EEG studies demonstrated lack of tolerance and withdrawal effects after 28 daily doses with gaboxadol, whereas zolpidem produced both tolerance and withdrawal effects under a similar dosing regimen. Daily dosing with gaboxadol, zolpidem or indiplon for 28 days and acute discontinuation of treatment left the threshold to bicuculline-induced seizures unchanged. The rapidly attenuated effect of repeated gaboxadol dosing was confirmed in the rotarod model. However, re-challenge of gaboxadol insensitive animals with gaboxadol produced a maximum response, ruling out that receptor desensitisation accounts for these effects. By comparing CNS exposure at rotarod responses and concentration response relation at cloned GABAA receptors expressed in Xenopus oocytes it appears that the decline in response in the rotarod model coincides with the steep part of the concentration response curve for gaboxadol at extra synaptic GABAA receptors. In conclusion, rat sleep EEG repeated dose studies of gaboxadol confirm a hypnotic-like profile and no withdrawal effects, whereas tolerance and withdrawal effects were shown with zolpidem. Withdrawal from gaboxadol, zolpidem and indiplon did not affect the seizure threshold to bicuculline. Gaboxadol's apparent rapid development of tolerance in the rotarod assay appears to be kinetically determined. �� 2008 Elsevier Inc. All rights reserved. Keywords: GABA Extra synaptic receptor Gaboxadol In vivo In vitro Tolerance Discontinuation 1. Introduction Even though hypnotics modulating GABAA receptors have been in clinical use for more than four decades, novel GABAA receptor active hypnotics are still in development. Most recently the GABAA receptor agonist gaboxadol was in clinical development, but discontinued very late in the development program. Publicly available clinical data (Walsh et al., 2007 Deacon et al., 2007 Anderson et al., 2007a Bodkin et al., 2007 Hedner et al., 2007a,b Mathias et al., 2005 Lancel et al., 2001) indicate that the hypnotic effects of gaboxadol are maintained over long term dosing and that withdrawal phenomena after acute discontinuation of treatment are very minor. Gaboxadol is a GABAA receptor agonist with functional selectivity for extra synaptic GABAA receptors. Experiments carried out in cell lines and Xenopus oocytes expressing Available online at www.sciencedirect.com Pharmacology, Biochemistry and Behavior 90 (2008) 113���122 www.elsevier.com/locate/pharmbiochembeh ��� Corresponding author. Department of Electrophysiology, H. Lundbeck A/S, Ottiliavej 9, DK-2500 Valby, Denmark. Tel.: +45 36433043. E-mail address: bjeb@lundbeck.com (B. Ebert). 0091-3057/$ - see front matter �� 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.pbb.2008.01.021

different GABAA receptor combinations have shown that gaboxadol has functional selectivity for ��4��3��, ��6��3�� and ��4��3 containing GABAA receptors over ��2 containing GABAA receptors (Storustovu and Ebert, 2006). Since extra synaptic receptors, with the exception of ��5��2/3��2 receptors in the hippocampus, are characterised by the absence of ��2, it has been proposed that gaboxadol should be characterised as a Selective Extra Synaptic GABAA receptor Agonist, abbreviated SEGA (Nutt, 2005). Several studies have demonstrated that the pharmacological activity of gaboxadol is dependent on both the ��4 and �� subunits. Chandra et al. (2006) demonstrated that gaboxadol in ��4 knockout mice was unable to induce tonic currents in thalamic slices and had very little effect on motor coordination impairment, whereas the inverse was the case in wild type mice. Boehm et al. (2006) showed that in delta knockout mice the anaesthetic effect of high doses of gaboxadol was relatively shorter lasting than in wild type mice, suggesting that �� containing receptors play an important role in the activity of gaboxadol. Recently, Winsky-Sommerer et al. (2007) were able to demonstrate that the hypnotic effects of gaboxadol were highly dependent on the presence of the �� subunit. Measuring the effects of gaboxadol on sleep architecture parameters during the normal wake period, the authors demonstrated that the massive increase in slow-wave activity seen during the waking EEG after dosing with gaboxadol in wild type mice was completely abolished in �� knockout mice. These data therefore tied the activity to �� containing GABAA receptors. As mentioned above, the clinical studies demonstrated hypnotic effects of gaboxadol maintained over months and very low incidence of withdrawal symptoms upon acute discontinua- tion, suggesting that the substrate for gaboxadol ��� probably the ��4��3�� containing receptors ��� is not markedly affected by long term dosing. In contrast, tolerance developed acutely to the motor coordination impairment effects in mice (Chandra et al., 2006). This discrepancy between motor coordination data in animal studies and sleep data in patients led us to speculate whether different tolerance mechanisms are involved in these effects. We therefore characterised gaboxadol in a series of animal models with the aim of identifying functional consequences of down regulation of GABAA receptors activated by gaboxadol. By combining these models with CNS exposure data, we selected a dose range of gaboxadol, which yielded peak concentrations up to 3 ��M (Cremers and Ebert, 2007), corresponding to doses up to 10 mg/kg/day in rats. 2. Methods 2.1. Animals Adult, male Sprague Dawley rats (200���250 g M&B, Denmark) pair-housed in macrolon cages (425��266��180 mm) were used in the sleep EEG studies. Male Wistar rats (150���200 g at study start M&B, Denmark) housed 2���4 per cage (2 per cage if weighing more than 200 g) were used in the rotarod studies. The rats were house under a 12:12 hour light:dark cycle (lights on at 06:00) and had free access to standard laboratory chow and water. Room temperature (21��2 ��C), relative humidity (55��5%), and air exchange (16 times per hour) were automatically controlled. Male Sprague���Dawley rats (125���150 g at study start Charles River Breeding Laboratories, Wilmington, MA, USA) housed 3 per cage on a 12 hour light:dark cycle and with food and water ad libitum were used in the bicuculline seizure threshold studies. 2.2. Ethics Ethical permissions for the sleep EEG and rotarod studies were granted by the animal welfare committee, appointed by the Danish Ministry of Justice. All animal procedures were carried out in compliance with the EC Directive 86/609/EEC and with the Danish law regulating experiments on animals. The protocols for the bicuculline seizure threshold studies were approved by the Tufts University School of Medicine and New England Medical Centre, Institutional Animal care and Use Committee, Boston, MA. 3. Sleep EEG studies 3.1. Surgery Using established methodology (Vogel et al., 2002), the transmitter (TL10M3-F50-EEE implant. Data Sciences Inter- national. USA) was implanted in the peritoneum (i.p.) of the anesthetized rat. EEG leads were then placed supradurally, 2 mm anterior to bregma and 2 mm on either side of the midline for the frontal electrodes and 2 mm anterior to lambda and 2 mm on either side of the midline for the parietal electrodes. The EMG leads were placed in either side of the musculus cervicoauricularis and were sutured in place. The animals were allowed to recover one-week post-surgery during which an antibiotic Baytril Vet��� (enrofloxacin 10 mg kg��� 1) and an analgesic (Rimadyl�� carprofen 0.1 ml per 100 g) were administered once daily subcutaneously (s.c.). 3.2. Sleep recording and scoring Dataquest A.R.T. Gold 2.2 was used to simultaneously record EEG and EMG for 5 h immediately after dosing. The EEG data were scored as wake (W), slow-wave sleep-1 (SWS- 1), slow-wave sleep-2 (SWS-2) or paradoxical sleep (PS or REM-like) according to visual analysis of EEG frequency as well as amplitude characteristics and EMG activity (Neck- elmann and Ursin, 1993). The EEG recordings for each animal were visually and manually scored on the computer screen in 10 s epochs over the 5-hour recording session (3000 epochs). Data were analysed using the sleep analysis software program, Somnologica 3 (Flagahf Medical Devices, Iceland). 3.3. Drug EEG studies Drugs were administered between 08:00 and 09:00 (lights on 06:00) and EEGs were recorded for the following 5 h. Dose��� response relationships for gaboxadol (1.25, 2.5 or 5 mg/kg s.c.) and zolpidem (2.5, 5.0 and 10 mg/kg s.c.) were established using acute administration in two groups of 8 rats. In a chronic dosing 114 B. Ebert et al. / Pharmacology, Biochemistry and Behavior 90 (2008) 113���122