- Email: [email protected]
Anticonvulsant effects of the selective melatonin receptor agonist ramelteon
Epilepsy & Behavior 16 (2009) 52–57
Contents lists available at ScienceDirect
Epilepsy & Behavior journal homepage: www.elsevier.com/locate/yebeh
Anticonvulsant effects of the selective melatonin receptor agonist ramelteon Kristina Fenoglio-Simeone a, Andrey Mazarati b, Sepideh Sefidvash-Hockley c, Don Shin b, Julianne Wilke c, Heather Milligan c, Raman Sankar b, Jong M. Rho c, Rama Maganti c,* a
Creighton University School of Medicine, Omaha, NE, USA Department of Pediatrics, David Geffen School of Medicine at the University of California, Los Angeles, CA, USA c Barrow Neurological Institute, St Joseph’s Hospital and Medical Center, Phoenix, AZ, USA b
a r t i c l e
i n f o
Article history: Received 27 May 2009 Revised 13 July 2009 Accepted 14 July 2009 Available online 13 August 2009 Keywords: Ramelteon Melatonin Melatonin receptor Epilepsy Circadian rhythms
a b s t r a c t Objective: The endogenous hormone melatonin has previously been shown to exert anticonvulsant effects in a variety of experimental models. Accordingly, we asked whether ramelteon, a synthetic and selective melatonin receptor agonist, might also possess anticonvulsant and/or antiepileptogenic properties. Methods: The effects of ramelteon (30 or 100 mg/kg intraperitoneally twice daily for 5 days) were evaluated in two animal models of epilepsy. In the rat rapid kindling model, baseline hippocampal afterdischarge properties, kindling progression, and hippocampal excitability in kindled animals were measured. Anti-ictogenic efficacy was assessed after acute administration in untreated kindled rats. In the spontaneously epileptic Kcna1-null mouse model, we determined seizure frequency and periodicity using continuous video/EEG monitoring over 72 hours. Further, circadian rest–activity rhythms in ramelteontreated animals were studied with actigraphy. Results: In kindled animals, ramelteon reversed kindling-induced hippocampal excitability; however, it did not modify baseline afterdischarge properties, the progression and establishment of the kindled state in the rapid kindling model. However, in Kcna1-null mice, ramelteon (200 mg/kg/day) significantly attenuated seizure periodicity and frequency and improved circadian rest–activity rhythms compared with control animals. Conclusions: The selective melatonin receptor agonist ramelteon possesses anticonvulsant properties in a chronic epilepsy model. Our findings provide further support for melatonin receptors being potential novel targets for anticonvulsant drug development. Ó 2009 Elsevier Inc. All rights reserved.
1. Introduction Epilepsy remains a continuing health concern despite the increasing availability of newer anticonvulsant drugs. Although seizures in two-thirds of patients can be successfully controlled with these medications, seizures in the remaining third remain refractory to medical therapy [1]. Most anticonvulsant drugs target neuronal membrane-bound ion channels, such as voltage-gated sodium and calcium channels, N-Methyl-D-aspartate (NMDA) receptors, and c-aminobutyric acid type A (GABAA) receptors, and there are many other drugs—both currently available and in various stages of development—that have unknown mechanisms of action. In this context, there is a strong need to develop antiepileptic drugs with novel mechanisms of action. Melatonin is an endogenous neuroactive compound that plays a critical role in regulating circadian rhythms and sleep–wake cycles * Corresponding author. Address: 500 West Thomas Road, Suite 300, Phoenix, AZ 85013, USA. Fax: +1 602 406 6299. E-mail address: [email protected] (R. Maganti). 1525-5050/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.yebeh.2009.07.022
[2]. Intriguingly, melatonin has been shown to possess anticonvulsant properties in several experimental models of epilepsy. Specifically, melatonin demonstrates anticonvulsant efficacy against pentylenetetrazol (PTZ)-induced seizures [3] and retards amygdala kindling [4]. Further, exogenous melatonin administration prevents the reduction in seizure threshold in the PTZ model caused by alterations in light–dark cycles [5]. Finally, pinealectomized animals have an increased susceptibility to lithium–pilocarpine-induced seizures compared with controls [6], suggesting that melatonin deficiency lowers seizure threshold. Despite these laboratory observations, there is presently a lack of compelling human data supporting the anticonvulsant properties of melatonin. Ramelteon is a selective melatonin MT-1 and MT-2 receptor agonist approved in the United States for treatment of insomnia. As such, and given the broad experimental literature on the anticonvulsant effects of melatonin, we asked whether ramelteon might also possess anticonvulsant efficacy. Specifically, we hypothesized that ramelteon might be efficacious in either induced and/or spontaneous models of epilepsy. For our studies, we chose the rapid kindling model of epileptogenesis [7–9], wherein both
K. Fenoglio-Simeone et al. / Epilepsy & Behavior 16 (2009) 52–57
antiepileptogenic and anti-ictogenic effects can be accurately assessed [10,11], and also the Kcna1-null mouse model of spontaneous recurrent seizures, which closely models a genetic form of refractory partial-onset epilepsy in humans [12–15].
53
2.1.5. Analysis Data were analyzed using Prism 4 Software (GraphPad, San Diego, CA, USA), with repeated-measures ANOVA + Bonferroni post hoc test. 2.2. Kcna1-null mouse model
2. Methods 2.1. Rapid kindling model 2.1.1. Animals Experiments were performed in male Wistar rats (Charles River, Wilmington, MA, USA), aged Postnatal Days 50 to 60 (P50–P60). All procedures were conducted in accordance with the policies of the National Institutes of Health and the UCLA Office for the Protection of Research Subjects.
2.1.2. Surgery Animals were anesthetized with isoflurane and stereotaxically implanted with a twisted bipolar stimulating electrode (PlasticsOne, Inc., Roanoke, VA, USA) in the left ventral hippocampus (4.8 mm posterior, 5.3 mm lateral from bregma, 6.5 mm deep from the brain surface) and a tripolar recording electrode (PlasticsOne, diameter 0.23 mm) in the neocortex [9,10].
2.1.3. Studies of antiepileptogenic effects Three to five days after surgery, the animals were connected to a DS8000 electrical stimulator (World Precision Instruments, Sarasota, FL, USA) and to an MP100/EEG100B acquisition system (BIOPAC, Santa Barbara, CA, USA). Afterdischarge threshold (ADT) and afterdischarge duration (ADD) were measured by applying electrical stimuli to and recording from the hippocampal electrode. Stimulation parameters were: 10-s train, 50-ms peak interval, 1-ms pulse duration, square-wave biphasic waveform, starting at 0.1 mA with 0.05-mA increments delivered every 10 minutes [9,10]. On detection of ADT, animals were injected intraperitoneally with either ramelteon (30 or 100 mg/kg dissolved in dimethylsulfoxide [DMSO]) or DMSO as control treatment. Twenty minutes later, afterdischarge properties were studied again, and the animals were subjected to rapid kindling: 60 stimulations delivered every 5 minutes at the parameters described above, but at 0.05 mA above ADT, as determined after ramelteon or DMSO injection [9,10]. During kindling, electrical activity was recorded from the cortical electrode; animal behavior was recorded using a digital video camera. Twenty-four hours after the last kindling stimulation, hippocampal afterdischarge properties were examined again. ADT and ADD were analyzed before kindling; the stimulations required to reach the first stage 1 seizure [16], the stimulations required to reach the first stage 4 seizure [16], and all stage 4 seizures were counted during kindling; and ADT and ADD were measured after kindling.
2.1.4. Studies of anti-ictogenic effects These experiments were performed using previously described methods [9,10]. Animals were subjected to rapid kindling as described above, but in the absence of ramelteon treatment; 24 hours after the last kindling stimulation, afterdischarge properties were examined and the animals were injected with ramelteon (30 or 100 mg/kg IP) or DMSO (control). Twenty to thirty minutes after the injections, afterdischarge properties were recorded again. The following parameters were analyzed: ADT, ADD, and behavioral seizures in response to the threshold stimulation.
2.2.1. Animals Kcna1-null mice were bred at the Barrow Neurological Institute (BNI) vivarium, reared in a quiet, temperature-controlled room, and entrained to a 12-hour light/dark cycle, with lights on at Zeitgeber Time (ZT) 00:00. Tail clips were taken by P7 and sent to Transnetyx Inc. (Cordova, TN, USA) for genotyping. All protocols were conducted in accordance with NIH guidelines and approved by Barrow Institutional Animal Care and Use Committee. 2.2.2. Monitoring and treatment Mice were individually placed in an 8 8 16-in. transparent Plexiglas arena (with bedding; food and water provided ad libitum) and allowed to habituate for 3–4 hours. Electroclinical seizures (in Kcna1-null mice) and rest–activity patterns (of Kcna1-null and wild-type [WT] mice) were monitored for 3–5 days prior to treatment and for 5 days during treatment. During treatment, mice were injected twice daily with ramelteon (30 or 100 mg/kg in DMSO) or DMSO as the control treatment at approximately zeitgeber (ZT)00:00 and ZT10:00. 2.2.3. EEG electrode implantation surgery and seizure scoring Electrical and behavioral seizures were recorded using Stellate video/EEG technology and Harmonie software (Stellate, Quebec, Canada). Animals were anesthetized with isoflurane (5% induction, 2% maintenance) prior to transmitter implantation. A wireless, PhysioTel telemetry transmitter (Data Sciences International, St. Paul, MN, USA) was implanted in a subcutaneous pocket along the dorsal flank. The biopotential leads were implanted bilaterally on the dura, 2 mm lateral of the midsagittal suture and 1 mm caudal of bregma, with the ground implanted in the occipital bone (the transmitter contained an internal reference electrode). Animals were allowed to recover 3 days prior to seizure and rest–activity monitoring. Behavioral seizures were scored on a modified Racine scale [16] and correlated with EEG interictal and ictal activity. Generalized tonic–clonic seizures typically began with tonic arching and tail extension, followed by forelimb clonus, or rearing and forelimb clonus, then generalized synchronous forelimb and hindlimb clonus, after which there was postictal depression. Time of onset and severity were recorded for each seizure. Number of seizures observed during each hour was subsequently collapsed into 4-hour ZT time bins—ZT 00:00; ZT 04:00; ZT 08:00; ZT 12:00; ZT 16:00; ZT 20:00—for each animal. 2.2.4. Actigraphy Actigraphy is a noninvasive method of monitoring human sleep–wake and animal rest–activity cycles. Behavioral rest–activity cycles were assessed using the Vital View data acquisition system, which integrates radio telemetry technology and switchclosure activity monitoring (Mini Mitter Company, Inc; Bend, OR, USA). The activity was monitored in 3-minute epochs and scored on an activity scale of 0–50. Data were analyzed with ActiView Biological Rhythm Analysis software (Mini Mitter Co., Inc.). The time of peak activity was determined by the maximum value of a fitted cosine function. A v2 periodogram method was used to determine the length of the rest–activity (nocturnal) period (length of the average rest–activity cycle or oscillation in hours) in all groups of animals.
54
K. Fenoglio-Simeone et al. / Epilepsy & Behavior 16 (2009) 52–57
2.2.5. Statistical analyses Spontaneous seizures were examined as a 2 4 6 factor design: treatment (prior to treatment, during treatment) x drug (30 mg/kg, 100 mg/kg ramelteon) x time point (ZT00:00, 04:00, 08:00, 12:00, 16:00, 20:00). Rest–activity patterns were examined as a 2 2 4 6 factor design: genotype (WT, Kcna1-null) treatment (prior to treatment, during treatment) drug (naïve, DMSO control, 30 mg/kg ramelteon, 100 mg/kg ramelteon) time point (ZT00:00, 04:00, 08:00, 12:00, 16:00, 20:00). Statistical significance (set at P < 0.05) among experimental groups was determined using ANOVA with Bonferroni’s post hoc test (Prism, GraphPad).
3. Results Fig. 1. Effects of ramelteon in kindled animals. Data are presented as means ± SEM. Control, n = 6; RAM30, n = 9; RAM100, n = 8. *P < 0.05 versus control (t test). Seizure response was analyzed using the Mann–Whitney test (P > 0.05).
3.1. Rapid kindling model 3.1.1. Effects on rapid kindling Ramelteon administered at either 30 or 100 mg/kg daily failed to alter baseline afterdischarge properties (Table 1, rows 1–4) or rapid kindling progression (Table 1, rows 5–7). Twenty-four hours after completion of rapid kindling, all ramelteon-treated and control animals exhibited statistically similar decreases in ADT and increases in ADD, and developed behavioral seizures in response to threshold stimulation (Table 1, rows 8–11).
3.1.2. Anti-ictogenic effects In control (DMSO-treated) rats, rapid kindling resulted in decreased ADT, increased ADD, and the development of behavioral convulsions in response to threshold stimulation (Fig. 1). With the 30 mg/kg twice daily dosing regimen, ramelteon did not modify afterdischarge properties and behavioral seizures in kindled animals compared with controls. At the higher dose (100 mg/kg twice daily), ramelteon alleviated kindling-induced decreases in ADT and increases in ADD (Fig. 1); this treatment also resulted in a lessening of behavioral seizure severity in response to threshold stimulation, although this did not reach statistical significance (Fig. 1).
Table 1 Effects of ramelteon on rapid kindling epileptogenesis.
ADT before kindling (mA) Before RAM After RAM ADD before kindling (s) Before RAM After RAM Number of stimulations to first stage 1 seizure Number of stimulations to first stage 4 seizure Number of stage 4 seizures ADT after kindling (mA) ADD after kindling (s) Seizure score at AD after kindling Number of animals with stage 4 or 5/number of animals without stage 4 or 5 sizures at AD after kindling
Control (n = 6)
RAM, 30 mg/kg (n = 9)
RAM, 100 mg/kg (n = 10)
0.40 ± 0.03 0.43 ± 0.04
0.41 ± 0.03 0.45 ± 0.03
0.5 ± 0.04 0.49 ± 0.04
40.8 ± 5.4 43.3 ± 4.8 8.7 ± 1.1
39.4 ± 5.0 41.1 ± 4.8 9.8 ± 1.1
43.0 ± 4.4 45.0 ± 3.5 8.7 ± 1.1
17.2 ± 2.1
19.6 ± 1.8
17.6 ± 1.6
16.0 ± 1.8 0.22 ± 0.02a 90.8 ± 5.7a 3.7 ± 0.2 4/2
17.1 ± 2.1 0.24 ± 0.03a 87.8 ± 6.9a 3.7 ± 0.2 6/3
18.7 ± 1.5 0.26 ± 0.02a 92.2 ± 6.2a 3.5 ± 0.17 5/5
Note. Data are presented as means ± SEM. ADD, afterdischarge duration; ADT, afterdischarge threshold; RAM, ramelteon. a P < 0.05 versus respective parameter before kindling (both before and after ramelteon treatment, repeated-measures ANOVA + Bonferroni post hoc test).
3.2. Kcna1-null mouse model 3.2.1. Ramelteon reduces spontaneous seizures in Kcna1-null mice Timing, duration, and severity of electroclinical seizures of Kcna1-null mice were recorded continuously for 2 days prior to and during the 5-day ramelteon treatment eriod. There was a nonsignificant trend toward seizure reduction (in 75% of mice) treated with the low-dose (30 mg/kg twice daily) regimen (P = 0.19) (Fig. 2A). Kcna1-null mice treated with the higher (100 mg/kg twice daily) dose experienced significantly fewer seizures per day when compared with each animal’s baseline seizure frequency (P < 0.05) (Fig. 2B). Fig. 2A and B (lower panels) depict the number of seizures experienced by each animal before (baseline) and after ramelteon treatment. Seizure occurrence was segregated into 4-hour bins to determine whether they followed a diurnal periodicity (i.e., occurring with a 24-hour rhythm). As shown in Fig. 2C (filled squares), seizures were more frequent during periods of rest and were less frequent while mice were active (between ZT12:00 and 20:00). Interestingly, there was no significant difference in the timing of seizures in mice treated with the high-dose regimen. However, there was a trend indicating seizure reduction at five of the six time points (ZT00:00, 04:00, 08:00, 16:00, 20:00). The total numbers of seizures per day that occurred during the rest period (light phase) and during the active period (dark phase) were summed across animals (Fig. 2D). Ramelteon treatment (100 mg/ kg twice daily) reduced the total number of seizures by approximately 58%; there was a 53% reduction during the rest period and a 63% reduction during the active period. DMSO alone did not influence seizure frequency.
3.2.2. Ramelteon treatment restores rest–activity rhythms of Kcna1null mice Rest–activity patterns were monitored in Kcna1-null and WT mice 2 days prior to and for 5 days during ramelteon treatment. As shown in Fig. 3A, Kcna1-null mice exhibit intrinsically disrupted rest–activity cycles when compared with WT controls, with increased activity during the light phase (when mice are more restful) and reduced activity during the dark phase (when mice are more active). Ramelteon treatment (100 mg/kg twice daily) significantly reduced the activity of Kcna1-null mice at the end of the light phase (i.e., 08:00–12:00) (Fig. 3B). There was no effect of ramelteon on rest–activity cycles of WT or Kcna1-null mice when given at the lower doses (30 mg/kg twice daily). Rest–activity patterns of DMSO-treated mice did not differ from those of naïve controls.
K. Fenoglio-Simeone et al. / Epilepsy & Behavior 16 (2009) 52–57
55
Fig. 2. Ramelteon reduces seizures in Kcna1-null mice. (A) Number of seizures experienced per day prior to (baseline, filled bars) and during (open bars) ramelteon treatment. (B) The ‘‘before–after” graph depicts the number of seizures experienced by each animal prior to and during ramelteon treatment. (C) Seizures of Kcna1-null mice occurred in a diurnal rhythm and statistically differ from ZT04:00, P < 0.05. There was no difference in the timing of seizure occurrence after ramelteon treatment (100 mg/kg). (D) Total number of seizures per day during the rest phase (lights on) and during the active phase (lights off).
The rest–activity cycles of WT mice entrained to a 12-hour light/dark cycle followed a 24 ± 0.01-hour period (Fig. 4, left panel). In contrast, epileptic Kcna1-null mice entrained to the same 12-hour light/dark cycle had a significantly longer period (28.4 ± 0.8 hour). Ramelteon treatment (100 mg/kg twice daily) restored activity periods of Kcna1-null mice to 24.89 ± 1.04 hours, values similar to those of WT mice and significantly less than those of nontreated Kcna1-null mice. Rest–activity cycles were fit to a cosine function to determine the time of peak activity. For WT mice, peak activity occurred at ZT17.11 ± 0.16, whereas activity did not peak at a particular time point for epileptic mice and occurred throughout the 24-hour cycle, ranging from ZT00:05 to 23:45 (Fig. 4, right panel). Treatment with ramelteon (100 mg/kg twice daily) clustered the timing of peak activity to ZT20.79 ± 0.3 and reduced the range to 4 hours (ZT19:21–22:15). Ramelteon treatment did not influence the period or peak activity of WT mice.
4. Discussion To our knowledge, the present study is the first to report that the selective melatonin receptor agonist ramelteon possesses anticonvulsant properties in both induced and spontaneous models of epilepsy. In the rapid kindling model, although ramelteon did not prevent kindling progression (indicating a lack of antiepileptogenic effects), it did reverse kindling-induced hippocampal hyperexcitability and exhibited a trend toward decreased severity of kindled seizures at the higher 100 mg/kg twice daily dose (suggesting an anti-ictogenic potential of the medication). By contrast, in Kcna1null mice, ramelteon significantly decreased the frequency of spontaneous seizures when administered at 100 mg/kg twice daily. Moreover, ramelteon attenuated the circadian periodicity of seizures, and improved circadian rest–activity periods at the higher dose. It should be noted that in both models tested, ramelteon exerted dose-dependent effects.
Fig. 3. Ramelteon improves rest–activity rhythms of Kcna1-null mice. (A) Rest–activity cycles of WT and Kcna1-null mice have nocturnal rhythmicity. Rest–activity cycles are disrupted in Kcna1-null mice, with increased activity during the light phase and reduced activity during the dark phase. Two-factor ANOVA: *Statistical difference from WT ZT00:00 (n = 9). aDiffers significantly from Kcna1-null mice ZT00:00 (n = 21). **Kcna1-null mice differed significantly from WT, P < 0.05. (B) Treatment with 100 mg/kg ramelteon reduced activity of Kcna1-null mice during their rest period (n = 9). *Statistically differs from nontreated Kcna1-null mice, P < 0.05. (C) Wild type with and without ramelteon. (D) Treatment with 30 mg/kg ramelteon has no significant effect on rest–activity cycles in Kcna1-null mice.
56
K. Fenoglio-Simeone et al. / Epilepsy & Behavior 16 (2009) 52–57
Fig. 4. Ramelteon treatment restored the period and clustered the peak activity of Kcna1-null mice. (Left) Rest–activity cycles were analyzed using a v2 periodogram method to determine whether ramelteon treatment influenced the period of WT and Kcna1-null mice. Kcna1-null mice have a significantly extended period when compared with that of WT mice. The period of epileptic mice treated with 100 mg/kg ramelteon did not differ from that of WT controls. *Differs significantly from Kcna1-null mice. (Right) Rates of activity were fit to a cosine function to determine at what time point activity peaked. Ramelteon treatment clustered peak activity in a nonsignificant dose-dependent manner. *Statistically different from WT values. DMSO treatment did not alter period or peak activity in either group.
The rapid kindling model represents a model of ‘‘compressed” epileptogenesis, in that naïve animals are induced into an epileptic state within a matter of hours [8]. This contrasts with other wellestablished models, such as the status epilepticus models involving kainate or pilocarpine administration [17], or conventional kindling, wherein epileptogenesis ensues over many days to weeks [18]. As such, rapid kindling affords investigators a practical and expedient approach toward screening of investigational agents with potential anticonvulsant (and antiepileptogenic) properties. Our earlier studies [9,11,19] using this rapid kindling paradigm helped establish specific antiepileptogenic and/or disease-modifying effects of a number of drugs that are currently available for clinical use or that are in various stages of development (i.e., topiramate, retigabine, bumetanide, ganaxolone). In further support of the utility of rapid kindling, other investigators have reported antiepileptogenic effects of yet another anticonvulsant, levetiracetam [20]. Accordingly, we chose to study the potential anticonvulsant effects of ramelteon using a similar protocol. When ramelteon was compared with other anticonvulsant compounds tested to date against rapid kindling, ramelteon appeared to exert only a moderate anti-ictogenic effect, evidenced by its mitigation of kindling-induced hippocampal hyperexcitability. Ramelteon, even at a total daily dose of 200 mg/kg, does not possess antiepileptogenic properties, at least in the rapid kindling model employed in the current study. Whether ramelteon might be effective in other models of epileptogenesis remains unknown. Although not directly compared, ramelteon is much weaker compared with such anticonvulsants as levetiracetam, topiramate, and retigabine. In contrast to rapid kindling, which is an induced model, the Kcna1-null mouse is a spontaneous, genetic model of developmental epileptogenesis, where electroclinical seizures are noted as early as the third postnatal week of life. As treatment was instituted after the onset of seizures, our protocol mirrors more closely the human paradigm in which anticonvulsant medications are administered after a clear diagnosis of epilepsy is established. Thus, we propose that the Kcna1-null model not only offers the advantage of post hoc treatment as in humans with epilepsy, but is also a chronic model in which the natural ontogeny of seizure activity can be documented and appropriately timed interventions carefully studied. In our hands, ramelteon clearly demonstrated dose-dependent anticonvulsant effects in Kcna1-null mice.
The mechanisms underlying the anticonvulsant effects of ramelteon are unknown, as melatonin receptors are not established anticonvulsant targets. However, melatonin receptors are widely distributed throughout the brain in seizure-prone areas such as the cerebral cortex [21] and the hippocampus, specifically in CA1, CA3, dentate gyrus, and subiculum [22]. Additionally, in the hippocampus, MT-1 receptors are more prevalent than MT-2 receptors [22]. This fact, combined with the observation that MT1 receptor blockade with luzindole can abolish the anticonvulsant effects of melatonin [23], lends further support to the notion that MT-1 receptor agonism may help attenuate hippocampal hyperexcitability, and as such, agents that activate MT-1 receptors (e.g., ramelteon) may exert anticonvulsant activity. In a prior study [15], baseline seizure activity in Kcna1-null mice followed a diurnal periodicity, and diurnal rhythms were disturbed compared with those of wild-type controls. Ramelteon attenuated the diurnal periodicity of seizures in the present study. Moreover, we noted improvement in the circadian rhythm of rest–activity periods following administration of the drug. Specifically, the rest–activity patterns of Kcna1-null mice were similar to those of wild-type mice, especially when the higher dose of ramelteon was used. These findings are consistent with our previous study [15] demonstrating that circadian rhythms, seizure frequency, and periodicity in Kcna1-null mice all improved following treatment with an anticonvulsant ketogenic diet. Although actigraphy cannot examine sleep architecture directly, improvements in rest–activity (circadian) rhythms implicate a normalization of sleep–wake cycles in these epileptic animals that could profoundly influence seizure activity. Unfortunately, our findings cannot clearly distinguish whether the observed reduction in seizure activity results from an improvement in circadian function or is a direct anticonvulsant effect of ramelteon. There are other limitations to the present study. Although we noted improvement in circadian rhythms in Kcna1-null mice following ramelteon administration, we cannot explain the mechanism accounting for the observed improvement. Nevertheless, the rest–activity patterns of treated epileptic mice were similar to those of control mice, indicating that the drug did not induce sedative effects or make the animals inactive during the dark phase (when they are most active). Perhaps the most significant limitation is the fact that the doses of ramelteon employed (on a mg/ kg basis) are vastly higher than that approved for the treatment
K. Fenoglio-Simeone et al. / Epilepsy & Behavior 16 (2009) 52–57
of insomnia in humans [24], and are much higher than those used in prior animal studies of insomnia [25,26]. Finally, we cannot extrapolate the doses used in the present study to humans because we did not measure serum concentrations of the drug and, therefore, have limited pharmacokinetic information. The doses of ramelteon we used in this study are roughly four to eight times the dose used for treatment of insomnia (8 mg). In summary, we found that ramelteon, a specific MT-1 and MT-2 receptor agonist, exhibits anticonvulsant properties in both the induced and spontaneous animal models of epilepsy. Additional studies are needed to confirm these preliminary findings and to examine potential anticonvulsant mechanisms for this novel compound. Importantly, the effects of ramelteon on circadian rhythm function need to be investigated further. Despite the limitations of the present study, our data provide additional support for melatonin receptors as novel targets for anticonvulsant drug development.
References [1] Kwan P, Brodie MJ. Early identification of refractory epilepsy. N Engl J Med 2000;342:314–9. [2] Dubocovih ML. Melatonin receptors: role on sleep and circadian rhythm regulation. Sleep Med 2007;8(Suppl. 3):34–42. [3] Solmaz I, Gürkanlar D, Gökçil Z, Göksoy C, Ozkan M, Erdog˘an E. Antiepileptic activity of melatonin in guinea pigs with pentylenetetrazol-induced seizures. Neurol Res 2009 [Epub ahead of print]. [4] Mevissen M, Ebert U. Anticonvulsant effects of melatonin in amygdale-kindled rats. Neurosci Lett 1998;257:13–6. [5] Tahsili-Fahadan P, Yahyavi-Firouz-Abadi N, Riazi K, Ghahremani MH, Dehpour AR. Effect of acute and chronic photo-period modulation on pentylenetetrazolinduced clonic seizure threshold in mice. Epilepsy Res 2008;82:64–9. [6] Janjoppi L, Silva de Lacerda AF, Scorza FA, Amado D, Cavalheiro EA, Arida RM. Influence of pinealectomy on the amygdale kindling development in rats. Neurosci Lett 2006;392:150–3. [7] Lothman EW, Hatlelid JM, Zorumski CF, Conry JA, Moon PF, Perlin JB. Kindling with rapidly recurring hippocampal seizures. Brain Res 1985;360:83–91. [8] Lothman EW, Williamson JM. Rapid kindling with recurrent hippocampal seizures. Epilepsy Res 1993;14:209–20. [9] Mazarati A, Shin D, Auvin S, Sankar R. Age-dependent effects of topiramate on the acquisition and the retention of rapid kindling. Epilepsia 2007;48:765–73.
57
[10] Mazarati A, Wu J, Shin D, Kwon YS, Sankar R. Antiepileptogenic and antiictogenic effects of retigabine under conditions of rapid kindling: an ontogenic study. Epilepsia 2008;49:1777–86. [11] Mazarati A, Shin D, Sankar R. Bumetanide inhibits rapid kindling in neonatal rats. Epilepsia 2009 [Epub ahead of print]. doi: 10.1111/j.15281167.2009.02048.x. [12] Smart SL, Lopantsev V, Zhang CL, et al. Deletion of Kv1.1 potassium channel causes epilepsy in mice. Neuron 1998;20:809–19. [13] Zuberi SM, Eunson LH, Spauschus A, et al. A novel mutation in the human voltage-gated potassium channel gene (Kv1.1) associated with episodic ataxia type-1 and sometimes with partial epilepsy. Brain 1999;122:817–25. [14] Rho JM, Szot P, Tempel BL, Schwartzkroin PA. Developmental seizure susceptibility of kv1.1 potassium channel knock out mice. Dev Neurosci 1999;21:320–7. [15] Fenoglio-Simeone KA, Wilke JC, Milligan HL, Allen CN, Rho JM, Maganti RM. Ketogenic diet treatment abolishes seizure periodicity and improves diurnal rhythmicity in epileptic Knca1-null mice. Epilepsia 2009 [Epub ahead of print]. doi: 10.1111/j.1528-1167.2009.02163.x. [16] Racine RJ. Modification of seizure activity by electrical stimulation: II. Motor seizure. Electroencephalogr Clin Neurophysiol 1972;32:281–94. [17] Cavalheiro EA, Naffah-Mazzacoratti MG, Mello LE, Leite JP. The pilocarpine model of seizures. In: Pitkanen A, Schwartzkroin PA, Moshe SL, editors. Models of seizures and epilepsy. Amsterdam: Elsevier; 2006. p. 433–48. [18] McIntyre DC. The rapid kindling model of seizures. In: Pitkanen A, Schwartzkroin PA, Moshe SL, editors. Models of seizures and epilepsy. Amsterdam: Elsevier; 2006. p. 352–64. [19] Mazarati A, Shin D, Rogawski M, Sankar R. Modification of rapid kindling epileptogenesis by ganaxolone and bumetanide [abstract]. Epilepsia 2008;49(Suppl. 7):369. [20] De Smedt T, Vonck K, Raedt R, et al. Rapid kindling in preclinical antiepileptic drug development: the effect of levetiracetam. Epilepsy Res 2005;67:109–16. [21] Morgan PJ, Barrett P, Howell HE, Helliwell R. Melatonin receptors: localization, molecular pharmacology and physiological significance. Neurochem Int 1994;24:101–46. [22] Musshoff U, Riewenherm D, Berger E, Fauteck JD, Speckmann EJ. Melatonin receptors in rat hippocampus: molecular and functional investigations. Hippocampus 2002;12:165–73. [23] Ray M, Mediratta PK, Reeta Kh, Mahajan P, Sharma KK. Receptors mechanisms involved in the anticonvulsant effect of melatonin in maximal electroshock seizures. Methods Find Exp Clin Pharmacol 2004;26:177–81. [24] Karim A, Tolbert D, Cao C. Disposition kinetics and tolerance of escalating single doses of ramelteon, a high affinity MT1 and MT2 melatonin receptor agonist indicated for treatment of insomnia. J Clin Pharmacol 2006;46:140–8. [25] Miyamoto M, Nishikawa H, Doken Y, Hirai K, Uchikawa O, Ohkawa S. The sleep promoting action of ramelteon (TAK-375) in freely moving cats. Sleep 2004;27:1319–25. [26] Yukuhiro N, Kimura H, Nishikawa H, Ohkawa S, Yoshikubo S, Miyamoto M. Effects of ramelteon (TAK-375) on nocturnal sleep in freely moving monkeys. Brain Res 2004;1027:59–66.
Contents lists available at ScienceDirect
Epilepsy & Behavior journal homepage: www.elsevier.com/locate/yebeh
Anticonvulsant effects of the selective melatonin receptor agonist ramelteon Kristina Fenoglio-Simeone a, Andrey Mazarati b, Sepideh Sefidvash-Hockley c, Don Shin b, Julianne Wilke c, Heather Milligan c, Raman Sankar b, Jong M. Rho c, Rama Maganti c,* a
Creighton University School of Medicine, Omaha, NE, USA Department of Pediatrics, David Geffen School of Medicine at the University of California, Los Angeles, CA, USA c Barrow Neurological Institute, St Joseph’s Hospital and Medical Center, Phoenix, AZ, USA b
a r t i c l e
i n f o
Article history: Received 27 May 2009 Revised 13 July 2009 Accepted 14 July 2009 Available online 13 August 2009 Keywords: Ramelteon Melatonin Melatonin receptor Epilepsy Circadian rhythms
a b s t r a c t Objective: The endogenous hormone melatonin has previously been shown to exert anticonvulsant effects in a variety of experimental models. Accordingly, we asked whether ramelteon, a synthetic and selective melatonin receptor agonist, might also possess anticonvulsant and/or antiepileptogenic properties. Methods: The effects of ramelteon (30 or 100 mg/kg intraperitoneally twice daily for 5 days) were evaluated in two animal models of epilepsy. In the rat rapid kindling model, baseline hippocampal afterdischarge properties, kindling progression, and hippocampal excitability in kindled animals were measured. Anti-ictogenic efficacy was assessed after acute administration in untreated kindled rats. In the spontaneously epileptic Kcna1-null mouse model, we determined seizure frequency and periodicity using continuous video/EEG monitoring over 72 hours. Further, circadian rest–activity rhythms in ramelteontreated animals were studied with actigraphy. Results: In kindled animals, ramelteon reversed kindling-induced hippocampal excitability; however, it did not modify baseline afterdischarge properties, the progression and establishment of the kindled state in the rapid kindling model. However, in Kcna1-null mice, ramelteon (200 mg/kg/day) significantly attenuated seizure periodicity and frequency and improved circadian rest–activity rhythms compared with control animals. Conclusions: The selective melatonin receptor agonist ramelteon possesses anticonvulsant properties in a chronic epilepsy model. Our findings provide further support for melatonin receptors being potential novel targets for anticonvulsant drug development. Ó 2009 Elsevier Inc. All rights reserved.
1. Introduction Epilepsy remains a continuing health concern despite the increasing availability of newer anticonvulsant drugs. Although seizures in two-thirds of patients can be successfully controlled with these medications, seizures in the remaining third remain refractory to medical therapy [1]. Most anticonvulsant drugs target neuronal membrane-bound ion channels, such as voltage-gated sodium and calcium channels, N-Methyl-D-aspartate (NMDA) receptors, and c-aminobutyric acid type A (GABAA) receptors, and there are many other drugs—both currently available and in various stages of development—that have unknown mechanisms of action. In this context, there is a strong need to develop antiepileptic drugs with novel mechanisms of action. Melatonin is an endogenous neuroactive compound that plays a critical role in regulating circadian rhythms and sleep–wake cycles * Corresponding author. Address: 500 West Thomas Road, Suite 300, Phoenix, AZ 85013, USA. Fax: +1 602 406 6299. E-mail address: [email protected] (R. Maganti). 1525-5050/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.yebeh.2009.07.022
[2]. Intriguingly, melatonin has been shown to possess anticonvulsant properties in several experimental models of epilepsy. Specifically, melatonin demonstrates anticonvulsant efficacy against pentylenetetrazol (PTZ)-induced seizures [3] and retards amygdala kindling [4]. Further, exogenous melatonin administration prevents the reduction in seizure threshold in the PTZ model caused by alterations in light–dark cycles [5]. Finally, pinealectomized animals have an increased susceptibility to lithium–pilocarpine-induced seizures compared with controls [6], suggesting that melatonin deficiency lowers seizure threshold. Despite these laboratory observations, there is presently a lack of compelling human data supporting the anticonvulsant properties of melatonin. Ramelteon is a selective melatonin MT-1 and MT-2 receptor agonist approved in the United States for treatment of insomnia. As such, and given the broad experimental literature on the anticonvulsant effects of melatonin, we asked whether ramelteon might also possess anticonvulsant efficacy. Specifically, we hypothesized that ramelteon might be efficacious in either induced and/or spontaneous models of epilepsy. For our studies, we chose the rapid kindling model of epileptogenesis [7–9], wherein both
K. Fenoglio-Simeone et al. / Epilepsy & Behavior 16 (2009) 52–57
antiepileptogenic and anti-ictogenic effects can be accurately assessed [10,11], and also the Kcna1-null mouse model of spontaneous recurrent seizures, which closely models a genetic form of refractory partial-onset epilepsy in humans [12–15].
53
2.1.5. Analysis Data were analyzed using Prism 4 Software (GraphPad, San Diego, CA, USA), with repeated-measures ANOVA + Bonferroni post hoc test. 2.2. Kcna1-null mouse model
2. Methods 2.1. Rapid kindling model 2.1.1. Animals Experiments were performed in male Wistar rats (Charles River, Wilmington, MA, USA), aged Postnatal Days 50 to 60 (P50–P60). All procedures were conducted in accordance with the policies of the National Institutes of Health and the UCLA Office for the Protection of Research Subjects.
2.1.2. Surgery Animals were anesthetized with isoflurane and stereotaxically implanted with a twisted bipolar stimulating electrode (PlasticsOne, Inc., Roanoke, VA, USA) in the left ventral hippocampus (4.8 mm posterior, 5.3 mm lateral from bregma, 6.5 mm deep from the brain surface) and a tripolar recording electrode (PlasticsOne, diameter 0.23 mm) in the neocortex [9,10].
2.1.3. Studies of antiepileptogenic effects Three to five days after surgery, the animals were connected to a DS8000 electrical stimulator (World Precision Instruments, Sarasota, FL, USA) and to an MP100/EEG100B acquisition system (BIOPAC, Santa Barbara, CA, USA). Afterdischarge threshold (ADT) and afterdischarge duration (ADD) were measured by applying electrical stimuli to and recording from the hippocampal electrode. Stimulation parameters were: 10-s train, 50-ms peak interval, 1-ms pulse duration, square-wave biphasic waveform, starting at 0.1 mA with 0.05-mA increments delivered every 10 minutes [9,10]. On detection of ADT, animals were injected intraperitoneally with either ramelteon (30 or 100 mg/kg dissolved in dimethylsulfoxide [DMSO]) or DMSO as control treatment. Twenty minutes later, afterdischarge properties were studied again, and the animals were subjected to rapid kindling: 60 stimulations delivered every 5 minutes at the parameters described above, but at 0.05 mA above ADT, as determined after ramelteon or DMSO injection [9,10]. During kindling, electrical activity was recorded from the cortical electrode; animal behavior was recorded using a digital video camera. Twenty-four hours after the last kindling stimulation, hippocampal afterdischarge properties were examined again. ADT and ADD were analyzed before kindling; the stimulations required to reach the first stage 1 seizure [16], the stimulations required to reach the first stage 4 seizure [16], and all stage 4 seizures were counted during kindling; and ADT and ADD were measured after kindling.
2.1.4. Studies of anti-ictogenic effects These experiments were performed using previously described methods [9,10]. Animals were subjected to rapid kindling as described above, but in the absence of ramelteon treatment; 24 hours after the last kindling stimulation, afterdischarge properties were examined and the animals were injected with ramelteon (30 or 100 mg/kg IP) or DMSO (control). Twenty to thirty minutes after the injections, afterdischarge properties were recorded again. The following parameters were analyzed: ADT, ADD, and behavioral seizures in response to the threshold stimulation.
2.2.1. Animals Kcna1-null mice were bred at the Barrow Neurological Institute (BNI) vivarium, reared in a quiet, temperature-controlled room, and entrained to a 12-hour light/dark cycle, with lights on at Zeitgeber Time (ZT) 00:00. Tail clips were taken by P7 and sent to Transnetyx Inc. (Cordova, TN, USA) for genotyping. All protocols were conducted in accordance with NIH guidelines and approved by Barrow Institutional Animal Care and Use Committee. 2.2.2. Monitoring and treatment Mice were individually placed in an 8 8 16-in. transparent Plexiglas arena (with bedding; food and water provided ad libitum) and allowed to habituate for 3–4 hours. Electroclinical seizures (in Kcna1-null mice) and rest–activity patterns (of Kcna1-null and wild-type [WT] mice) were monitored for 3–5 days prior to treatment and for 5 days during treatment. During treatment, mice were injected twice daily with ramelteon (30 or 100 mg/kg in DMSO) or DMSO as the control treatment at approximately zeitgeber (ZT)00:00 and ZT10:00. 2.2.3. EEG electrode implantation surgery and seizure scoring Electrical and behavioral seizures were recorded using Stellate video/EEG technology and Harmonie software (Stellate, Quebec, Canada). Animals were anesthetized with isoflurane (5% induction, 2% maintenance) prior to transmitter implantation. A wireless, PhysioTel telemetry transmitter (Data Sciences International, St. Paul, MN, USA) was implanted in a subcutaneous pocket along the dorsal flank. The biopotential leads were implanted bilaterally on the dura, 2 mm lateral of the midsagittal suture and 1 mm caudal of bregma, with the ground implanted in the occipital bone (the transmitter contained an internal reference electrode). Animals were allowed to recover 3 days prior to seizure and rest–activity monitoring. Behavioral seizures were scored on a modified Racine scale [16] and correlated with EEG interictal and ictal activity. Generalized tonic–clonic seizures typically began with tonic arching and tail extension, followed by forelimb clonus, or rearing and forelimb clonus, then generalized synchronous forelimb and hindlimb clonus, after which there was postictal depression. Time of onset and severity were recorded for each seizure. Number of seizures observed during each hour was subsequently collapsed into 4-hour ZT time bins—ZT 00:00; ZT 04:00; ZT 08:00; ZT 12:00; ZT 16:00; ZT 20:00—for each animal. 2.2.4. Actigraphy Actigraphy is a noninvasive method of monitoring human sleep–wake and animal rest–activity cycles. Behavioral rest–activity cycles were assessed using the Vital View data acquisition system, which integrates radio telemetry technology and switchclosure activity monitoring (Mini Mitter Company, Inc; Bend, OR, USA). The activity was monitored in 3-minute epochs and scored on an activity scale of 0–50. Data were analyzed with ActiView Biological Rhythm Analysis software (Mini Mitter Co., Inc.). The time of peak activity was determined by the maximum value of a fitted cosine function. A v2 periodogram method was used to determine the length of the rest–activity (nocturnal) period (length of the average rest–activity cycle or oscillation in hours) in all groups of animals.
54
K. Fenoglio-Simeone et al. / Epilepsy & Behavior 16 (2009) 52–57
2.2.5. Statistical analyses Spontaneous seizures were examined as a 2 4 6 factor design: treatment (prior to treatment, during treatment) x drug (30 mg/kg, 100 mg/kg ramelteon) x time point (ZT00:00, 04:00, 08:00, 12:00, 16:00, 20:00). Rest–activity patterns were examined as a 2 2 4 6 factor design: genotype (WT, Kcna1-null) treatment (prior to treatment, during treatment) drug (naïve, DMSO control, 30 mg/kg ramelteon, 100 mg/kg ramelteon) time point (ZT00:00, 04:00, 08:00, 12:00, 16:00, 20:00). Statistical significance (set at P < 0.05) among experimental groups was determined using ANOVA with Bonferroni’s post hoc test (Prism, GraphPad).
3. Results Fig. 1. Effects of ramelteon in kindled animals. Data are presented as means ± SEM. Control, n = 6; RAM30, n = 9; RAM100, n = 8. *P < 0.05 versus control (t test). Seizure response was analyzed using the Mann–Whitney test (P > 0.05).
3.1. Rapid kindling model 3.1.1. Effects on rapid kindling Ramelteon administered at either 30 or 100 mg/kg daily failed to alter baseline afterdischarge properties (Table 1, rows 1–4) or rapid kindling progression (Table 1, rows 5–7). Twenty-four hours after completion of rapid kindling, all ramelteon-treated and control animals exhibited statistically similar decreases in ADT and increases in ADD, and developed behavioral seizures in response to threshold stimulation (Table 1, rows 8–11).
3.1.2. Anti-ictogenic effects In control (DMSO-treated) rats, rapid kindling resulted in decreased ADT, increased ADD, and the development of behavioral convulsions in response to threshold stimulation (Fig. 1). With the 30 mg/kg twice daily dosing regimen, ramelteon did not modify afterdischarge properties and behavioral seizures in kindled animals compared with controls. At the higher dose (100 mg/kg twice daily), ramelteon alleviated kindling-induced decreases in ADT and increases in ADD (Fig. 1); this treatment also resulted in a lessening of behavioral seizure severity in response to threshold stimulation, although this did not reach statistical significance (Fig. 1).
Table 1 Effects of ramelteon on rapid kindling epileptogenesis.
ADT before kindling (mA) Before RAM After RAM ADD before kindling (s) Before RAM After RAM Number of stimulations to first stage 1 seizure Number of stimulations to first stage 4 seizure Number of stage 4 seizures ADT after kindling (mA) ADD after kindling (s) Seizure score at AD after kindling Number of animals with stage 4 or 5/number of animals without stage 4 or 5 sizures at AD after kindling
Control (n = 6)
RAM, 30 mg/kg (n = 9)
RAM, 100 mg/kg (n = 10)
0.40 ± 0.03 0.43 ± 0.04
0.41 ± 0.03 0.45 ± 0.03
0.5 ± 0.04 0.49 ± 0.04
40.8 ± 5.4 43.3 ± 4.8 8.7 ± 1.1
39.4 ± 5.0 41.1 ± 4.8 9.8 ± 1.1
43.0 ± 4.4 45.0 ± 3.5 8.7 ± 1.1
17.2 ± 2.1
19.6 ± 1.8
17.6 ± 1.6
16.0 ± 1.8 0.22 ± 0.02a 90.8 ± 5.7a 3.7 ± 0.2 4/2
17.1 ± 2.1 0.24 ± 0.03a 87.8 ± 6.9a 3.7 ± 0.2 6/3
18.7 ± 1.5 0.26 ± 0.02a 92.2 ± 6.2a 3.5 ± 0.17 5/5
Note. Data are presented as means ± SEM. ADD, afterdischarge duration; ADT, afterdischarge threshold; RAM, ramelteon. a P < 0.05 versus respective parameter before kindling (both before and after ramelteon treatment, repeated-measures ANOVA + Bonferroni post hoc test).
3.2. Kcna1-null mouse model 3.2.1. Ramelteon reduces spontaneous seizures in Kcna1-null mice Timing, duration, and severity of electroclinical seizures of Kcna1-null mice were recorded continuously for 2 days prior to and during the 5-day ramelteon treatment eriod. There was a nonsignificant trend toward seizure reduction (in 75% of mice) treated with the low-dose (30 mg/kg twice daily) regimen (P = 0.19) (Fig. 2A). Kcna1-null mice treated with the higher (100 mg/kg twice daily) dose experienced significantly fewer seizures per day when compared with each animal’s baseline seizure frequency (P < 0.05) (Fig. 2B). Fig. 2A and B (lower panels) depict the number of seizures experienced by each animal before (baseline) and after ramelteon treatment. Seizure occurrence was segregated into 4-hour bins to determine whether they followed a diurnal periodicity (i.e., occurring with a 24-hour rhythm). As shown in Fig. 2C (filled squares), seizures were more frequent during periods of rest and were less frequent while mice were active (between ZT12:00 and 20:00). Interestingly, there was no significant difference in the timing of seizures in mice treated with the high-dose regimen. However, there was a trend indicating seizure reduction at five of the six time points (ZT00:00, 04:00, 08:00, 16:00, 20:00). The total numbers of seizures per day that occurred during the rest period (light phase) and during the active period (dark phase) were summed across animals (Fig. 2D). Ramelteon treatment (100 mg/ kg twice daily) reduced the total number of seizures by approximately 58%; there was a 53% reduction during the rest period and a 63% reduction during the active period. DMSO alone did not influence seizure frequency.
3.2.2. Ramelteon treatment restores rest–activity rhythms of Kcna1null mice Rest–activity patterns were monitored in Kcna1-null and WT mice 2 days prior to and for 5 days during ramelteon treatment. As shown in Fig. 3A, Kcna1-null mice exhibit intrinsically disrupted rest–activity cycles when compared with WT controls, with increased activity during the light phase (when mice are more restful) and reduced activity during the dark phase (when mice are more active). Ramelteon treatment (100 mg/kg twice daily) significantly reduced the activity of Kcna1-null mice at the end of the light phase (i.e., 08:00–12:00) (Fig. 3B). There was no effect of ramelteon on rest–activity cycles of WT or Kcna1-null mice when given at the lower doses (30 mg/kg twice daily). Rest–activity patterns of DMSO-treated mice did not differ from those of naïve controls.
K. Fenoglio-Simeone et al. / Epilepsy & Behavior 16 (2009) 52–57
55
Fig. 2. Ramelteon reduces seizures in Kcna1-null mice. (A) Number of seizures experienced per day prior to (baseline, filled bars) and during (open bars) ramelteon treatment. (B) The ‘‘before–after” graph depicts the number of seizures experienced by each animal prior to and during ramelteon treatment. (C) Seizures of Kcna1-null mice occurred in a diurnal rhythm and statistically differ from ZT04:00, P < 0.05. There was no difference in the timing of seizure occurrence after ramelteon treatment (100 mg/kg). (D) Total number of seizures per day during the rest phase (lights on) and during the active phase (lights off).
The rest–activity cycles of WT mice entrained to a 12-hour light/dark cycle followed a 24 ± 0.01-hour period (Fig. 4, left panel). In contrast, epileptic Kcna1-null mice entrained to the same 12-hour light/dark cycle had a significantly longer period (28.4 ± 0.8 hour). Ramelteon treatment (100 mg/kg twice daily) restored activity periods of Kcna1-null mice to 24.89 ± 1.04 hours, values similar to those of WT mice and significantly less than those of nontreated Kcna1-null mice. Rest–activity cycles were fit to a cosine function to determine the time of peak activity. For WT mice, peak activity occurred at ZT17.11 ± 0.16, whereas activity did not peak at a particular time point for epileptic mice and occurred throughout the 24-hour cycle, ranging from ZT00:05 to 23:45 (Fig. 4, right panel). Treatment with ramelteon (100 mg/kg twice daily) clustered the timing of peak activity to ZT20.79 ± 0.3 and reduced the range to 4 hours (ZT19:21–22:15). Ramelteon treatment did not influence the period or peak activity of WT mice.
4. Discussion To our knowledge, the present study is the first to report that the selective melatonin receptor agonist ramelteon possesses anticonvulsant properties in both induced and spontaneous models of epilepsy. In the rapid kindling model, although ramelteon did not prevent kindling progression (indicating a lack of antiepileptogenic effects), it did reverse kindling-induced hippocampal hyperexcitability and exhibited a trend toward decreased severity of kindled seizures at the higher 100 mg/kg twice daily dose (suggesting an anti-ictogenic potential of the medication). By contrast, in Kcna1null mice, ramelteon significantly decreased the frequency of spontaneous seizures when administered at 100 mg/kg twice daily. Moreover, ramelteon attenuated the circadian periodicity of seizures, and improved circadian rest–activity periods at the higher dose. It should be noted that in both models tested, ramelteon exerted dose-dependent effects.
Fig. 3. Ramelteon improves rest–activity rhythms of Kcna1-null mice. (A) Rest–activity cycles of WT and Kcna1-null mice have nocturnal rhythmicity. Rest–activity cycles are disrupted in Kcna1-null mice, with increased activity during the light phase and reduced activity during the dark phase. Two-factor ANOVA: *Statistical difference from WT ZT00:00 (n = 9). aDiffers significantly from Kcna1-null mice ZT00:00 (n = 21). **Kcna1-null mice differed significantly from WT, P < 0.05. (B) Treatment with 100 mg/kg ramelteon reduced activity of Kcna1-null mice during their rest period (n = 9). *Statistically differs from nontreated Kcna1-null mice, P < 0.05. (C) Wild type with and without ramelteon. (D) Treatment with 30 mg/kg ramelteon has no significant effect on rest–activity cycles in Kcna1-null mice.
56
K. Fenoglio-Simeone et al. / Epilepsy & Behavior 16 (2009) 52–57
Fig. 4. Ramelteon treatment restored the period and clustered the peak activity of Kcna1-null mice. (Left) Rest–activity cycles were analyzed using a v2 periodogram method to determine whether ramelteon treatment influenced the period of WT and Kcna1-null mice. Kcna1-null mice have a significantly extended period when compared with that of WT mice. The period of epileptic mice treated with 100 mg/kg ramelteon did not differ from that of WT controls. *Differs significantly from Kcna1-null mice. (Right) Rates of activity were fit to a cosine function to determine at what time point activity peaked. Ramelteon treatment clustered peak activity in a nonsignificant dose-dependent manner. *Statistically different from WT values. DMSO treatment did not alter period or peak activity in either group.
The rapid kindling model represents a model of ‘‘compressed” epileptogenesis, in that naïve animals are induced into an epileptic state within a matter of hours [8]. This contrasts with other wellestablished models, such as the status epilepticus models involving kainate or pilocarpine administration [17], or conventional kindling, wherein epileptogenesis ensues over many days to weeks [18]. As such, rapid kindling affords investigators a practical and expedient approach toward screening of investigational agents with potential anticonvulsant (and antiepileptogenic) properties. Our earlier studies [9,11,19] using this rapid kindling paradigm helped establish specific antiepileptogenic and/or disease-modifying effects of a number of drugs that are currently available for clinical use or that are in various stages of development (i.e., topiramate, retigabine, bumetanide, ganaxolone). In further support of the utility of rapid kindling, other investigators have reported antiepileptogenic effects of yet another anticonvulsant, levetiracetam [20]. Accordingly, we chose to study the potential anticonvulsant effects of ramelteon using a similar protocol. When ramelteon was compared with other anticonvulsant compounds tested to date against rapid kindling, ramelteon appeared to exert only a moderate anti-ictogenic effect, evidenced by its mitigation of kindling-induced hippocampal hyperexcitability. Ramelteon, even at a total daily dose of 200 mg/kg, does not possess antiepileptogenic properties, at least in the rapid kindling model employed in the current study. Whether ramelteon might be effective in other models of epileptogenesis remains unknown. Although not directly compared, ramelteon is much weaker compared with such anticonvulsants as levetiracetam, topiramate, and retigabine. In contrast to rapid kindling, which is an induced model, the Kcna1-null mouse is a spontaneous, genetic model of developmental epileptogenesis, where electroclinical seizures are noted as early as the third postnatal week of life. As treatment was instituted after the onset of seizures, our protocol mirrors more closely the human paradigm in which anticonvulsant medications are administered after a clear diagnosis of epilepsy is established. Thus, we propose that the Kcna1-null model not only offers the advantage of post hoc treatment as in humans with epilepsy, but is also a chronic model in which the natural ontogeny of seizure activity can be documented and appropriately timed interventions carefully studied. In our hands, ramelteon clearly demonstrated dose-dependent anticonvulsant effects in Kcna1-null mice.
The mechanisms underlying the anticonvulsant effects of ramelteon are unknown, as melatonin receptors are not established anticonvulsant targets. However, melatonin receptors are widely distributed throughout the brain in seizure-prone areas such as the cerebral cortex [21] and the hippocampus, specifically in CA1, CA3, dentate gyrus, and subiculum [22]. Additionally, in the hippocampus, MT-1 receptors are more prevalent than MT-2 receptors [22]. This fact, combined with the observation that MT1 receptor blockade with luzindole can abolish the anticonvulsant effects of melatonin [23], lends further support to the notion that MT-1 receptor agonism may help attenuate hippocampal hyperexcitability, and as such, agents that activate MT-1 receptors (e.g., ramelteon) may exert anticonvulsant activity. In a prior study [15], baseline seizure activity in Kcna1-null mice followed a diurnal periodicity, and diurnal rhythms were disturbed compared with those of wild-type controls. Ramelteon attenuated the diurnal periodicity of seizures in the present study. Moreover, we noted improvement in the circadian rhythm of rest–activity periods following administration of the drug. Specifically, the rest–activity patterns of Kcna1-null mice were similar to those of wild-type mice, especially when the higher dose of ramelteon was used. These findings are consistent with our previous study [15] demonstrating that circadian rhythms, seizure frequency, and periodicity in Kcna1-null mice all improved following treatment with an anticonvulsant ketogenic diet. Although actigraphy cannot examine sleep architecture directly, improvements in rest–activity (circadian) rhythms implicate a normalization of sleep–wake cycles in these epileptic animals that could profoundly influence seizure activity. Unfortunately, our findings cannot clearly distinguish whether the observed reduction in seizure activity results from an improvement in circadian function or is a direct anticonvulsant effect of ramelteon. There are other limitations to the present study. Although we noted improvement in circadian rhythms in Kcna1-null mice following ramelteon administration, we cannot explain the mechanism accounting for the observed improvement. Nevertheless, the rest–activity patterns of treated epileptic mice were similar to those of control mice, indicating that the drug did not induce sedative effects or make the animals inactive during the dark phase (when they are most active). Perhaps the most significant limitation is the fact that the doses of ramelteon employed (on a mg/ kg basis) are vastly higher than that approved for the treatment
K. Fenoglio-Simeone et al. / Epilepsy & Behavior 16 (2009) 52–57
of insomnia in humans [24], and are much higher than those used in prior animal studies of insomnia [25,26]. Finally, we cannot extrapolate the doses used in the present study to humans because we did not measure serum concentrations of the drug and, therefore, have limited pharmacokinetic information. The doses of ramelteon we used in this study are roughly four to eight times the dose used for treatment of insomnia (8 mg). In summary, we found that ramelteon, a specific MT-1 and MT-2 receptor agonist, exhibits anticonvulsant properties in both the induced and spontaneous animal models of epilepsy. Additional studies are needed to confirm these preliminary findings and to examine potential anticonvulsant mechanisms for this novel compound. Importantly, the effects of ramelteon on circadian rhythm function need to be investigated further. Despite the limitations of the present study, our data provide additional support for melatonin receptors as novel targets for anticonvulsant drug development.
References [1] Kwan P, Brodie MJ. Early identification of refractory epilepsy. N Engl J Med 2000;342:314–9. [2] Dubocovih ML. Melatonin receptors: role on sleep and circadian rhythm regulation. Sleep Med 2007;8(Suppl. 3):34–42. [3] Solmaz I, Gürkanlar D, Gökçil Z, Göksoy C, Ozkan M, Erdog˘an E. Antiepileptic activity of melatonin in guinea pigs with pentylenetetrazol-induced seizures. Neurol Res 2009 [Epub ahead of print]. [4] Mevissen M, Ebert U. Anticonvulsant effects of melatonin in amygdale-kindled rats. Neurosci Lett 1998;257:13–6. [5] Tahsili-Fahadan P, Yahyavi-Firouz-Abadi N, Riazi K, Ghahremani MH, Dehpour AR. Effect of acute and chronic photo-period modulation on pentylenetetrazolinduced clonic seizure threshold in mice. Epilepsy Res 2008;82:64–9. [6] Janjoppi L, Silva de Lacerda AF, Scorza FA, Amado D, Cavalheiro EA, Arida RM. Influence of pinealectomy on the amygdale kindling development in rats. Neurosci Lett 2006;392:150–3. [7] Lothman EW, Hatlelid JM, Zorumski CF, Conry JA, Moon PF, Perlin JB. Kindling with rapidly recurring hippocampal seizures. Brain Res 1985;360:83–91. [8] Lothman EW, Williamson JM. Rapid kindling with recurrent hippocampal seizures. Epilepsy Res 1993;14:209–20. [9] Mazarati A, Shin D, Auvin S, Sankar R. Age-dependent effects of topiramate on the acquisition and the retention of rapid kindling. Epilepsia 2007;48:765–73.
57
[10] Mazarati A, Wu J, Shin D, Kwon YS, Sankar R. Antiepileptogenic and antiictogenic effects of retigabine under conditions of rapid kindling: an ontogenic study. Epilepsia 2008;49:1777–86. [11] Mazarati A, Shin D, Sankar R. Bumetanide inhibits rapid kindling in neonatal rats. Epilepsia 2009 [Epub ahead of print]. doi: 10.1111/j.15281167.2009.02048.x. [12] Smart SL, Lopantsev V, Zhang CL, et al. Deletion of Kv1.1 potassium channel causes epilepsy in mice. Neuron 1998;20:809–19. [13] Zuberi SM, Eunson LH, Spauschus A, et al. A novel mutation in the human voltage-gated potassium channel gene (Kv1.1) associated with episodic ataxia type-1 and sometimes with partial epilepsy. Brain 1999;122:817–25. [14] Rho JM, Szot P, Tempel BL, Schwartzkroin PA. Developmental seizure susceptibility of kv1.1 potassium channel knock out mice. Dev Neurosci 1999;21:320–7. [15] Fenoglio-Simeone KA, Wilke JC, Milligan HL, Allen CN, Rho JM, Maganti RM. Ketogenic diet treatment abolishes seizure periodicity and improves diurnal rhythmicity in epileptic Knca1-null mice. Epilepsia 2009 [Epub ahead of print]. doi: 10.1111/j.1528-1167.2009.02163.x. [16] Racine RJ. Modification of seizure activity by electrical stimulation: II. Motor seizure. Electroencephalogr Clin Neurophysiol 1972;32:281–94. [17] Cavalheiro EA, Naffah-Mazzacoratti MG, Mello LE, Leite JP. The pilocarpine model of seizures. In: Pitkanen A, Schwartzkroin PA, Moshe SL, editors. Models of seizures and epilepsy. Amsterdam: Elsevier; 2006. p. 433–48. [18] McIntyre DC. The rapid kindling model of seizures. In: Pitkanen A, Schwartzkroin PA, Moshe SL, editors. Models of seizures and epilepsy. Amsterdam: Elsevier; 2006. p. 352–64. [19] Mazarati A, Shin D, Rogawski M, Sankar R. Modification of rapid kindling epileptogenesis by ganaxolone and bumetanide [abstract]. Epilepsia 2008;49(Suppl. 7):369. [20] De Smedt T, Vonck K, Raedt R, et al. Rapid kindling in preclinical antiepileptic drug development: the effect of levetiracetam. Epilepsy Res 2005;67:109–16. [21] Morgan PJ, Barrett P, Howell HE, Helliwell R. Melatonin receptors: localization, molecular pharmacology and physiological significance. Neurochem Int 1994;24:101–46. [22] Musshoff U, Riewenherm D, Berger E, Fauteck JD, Speckmann EJ. Melatonin receptors in rat hippocampus: molecular and functional investigations. Hippocampus 2002;12:165–73. [23] Ray M, Mediratta PK, Reeta Kh, Mahajan P, Sharma KK. Receptors mechanisms involved in the anticonvulsant effect of melatonin in maximal electroshock seizures. Methods Find Exp Clin Pharmacol 2004;26:177–81. [24] Karim A, Tolbert D, Cao C. Disposition kinetics and tolerance of escalating single doses of ramelteon, a high affinity MT1 and MT2 melatonin receptor agonist indicated for treatment of insomnia. J Clin Pharmacol 2006;46:140–8. [25] Miyamoto M, Nishikawa H, Doken Y, Hirai K, Uchikawa O, Ohkawa S. The sleep promoting action of ramelteon (TAK-375) in freely moving cats. Sleep 2004;27:1319–25. [26] Yukuhiro N, Kimura H, Nishikawa H, Ohkawa S, Yoshikubo S, Miyamoto M. Effects of ramelteon (TAK-375) on nocturnal sleep in freely moving monkeys. Brain Res 2004;1027:59–66.