- Email: [email protected]
The effects of felbamate on appetitive and aversive instrumental learning in adult rats
Epilepsy & Behavior 78 (2018) 14–19
Contents lists available at ScienceDirect
Epilepsy & Behavior journal homepage: www.elsevier.com/locate/yebeh
The effects of felbamate on appetitive and aversive instrumental learning in adult rats John J. Orczyk a, Preston E. Garraghty a,b,⁎ a b
Department of Psychological and Brain Sciences, Indiana University, Bloomington, IN, USA Program in Neuroscience, Indiana University, Bloomington, IN, USA
a r t i c l e
i n f o
Article history: Received 24 July 2017 Revised 19 October 2017 Accepted 19 October 2017 Available online xxxx Keywords: Antiepileptic Anticonvulsant Felbamate Instrumental learning Cognition
a b s t r a c t Antiepileptic medications are the frontline treatment for seizure conditions but are not without cognitive side effects. Previously, our laboratory reported learning deficits in phenytoin-, carbamazepine-, and valproatetreated rats. In the present experiment, the effects of felbamate (FBM) have been compared to water-treated controls (controls) using the same instrumental training tasks employed here. Rats treated with FBM displayed a deficit in acquiring a tone-signaled avoidance response, relative to controls, but this was true only if they had no prior appetitive experience. Terminal avoidance behavior was equivalent to healthy controls. In contrast, the FBM-treated rats showed enhanced acquisition of the avoidance response relative to controls when given the benefit of prior experience in the appetitive condition. Relative to animals treated with phenytoin, carbamazepine, or valproate, FBM-treated rats showed the lowest overall pattern of deficits using these instrumental learning tasks. While FBM treatment has been severely restricted because of rather low risks of serious medical side effects, we suggest that the risks are not substantially higher than those shown to exist for phenytoin, carbamazepine, or valproate. As psychologists, we further suggest that negative cognitive deficits associated with these various drugs, along with their quality-of-life costs, are of relevance in the design of treatment strategies for individuals with seizure disorders. © 2017 Elsevier Inc. All rights reserved.
1. Introduction Antiepileptic drugs (AEDs) are the frontline treatment for epilepsy but are not without cognitive side effects [e.g., 1,2]. Many studies have reported cognitive deficits in human patients treated with AEDs [e.g., 3–9]; for reviews, see [10–12]. Performance deficits have also been observed in animals treated with AEDs [e.g., 13–15], though other studies performed in animals have failed to detect cognitive deficits associated with AED treatment [16,17]. Previously, our laboratory has reported different types of behavioral deficits in adult rats treated with the AEDs phenytoin [18], carbamazepine [19], and valproate [20]. It was shown that phenytoin, and to a lesser degree, carbamazepine, both frontline AEDs, blocked the acquisition of an avoidance response in the second part of an instrumental appetitive-to-aversive transfer conditioning task [18,19]. More recently, we have shown that valproate produces a different pattern of deficits, impairing the acquisition of an avoidance response in the absence of prior appetitive training, but having no effect on avoidance learning in rats transferred from appetitive training [20].
⁎ Corresponding author at: Department of Psychological and Brain Sciences, Indiana University, 1101 E 10th Street, Bloomington, IN 47405, USA. E-mail address: [email protected] (P.E. Garraghty).
https://doi.org/10.1016/j.yebeh.2017.10.022 1525-5050/© 2017 Elsevier Inc. All rights reserved.
In the present study, we have employed this same withinsubject, tone-signaled bar press task in which rats are tested in both appetitive and aversive contexts. The task employed is complex and multicontextual. There are multiple rules in the aversive context that must be learned through conditioning. In the aversive context, rats must learn to both press the lever after the tone and not to press the lever during the intertrial period. This paradigm was developed to study appetitive and aversive learning in the same subjects and has been used in past work to evaluate learning, memory, and impairments that accompany cerebellar, hippocampal, cingulate, and prefrontal cortex lesions [21,22]. We have used this behavioral paradigm to evaluate the effects of phenytoin, carbamazepine, and valproate in adult rats [18–20,23], rats exposed to phenytoin in utero [24], rats with lesions of the basal nucleus of Meynart [25], ovariectomized female rats with or without estradiol replacement [26], and rats undergoing chronic restraint [27,28]. In the current study, we have extended our assessment of the cognitive side effects of AEDs to felbamate (FBM). Like phenytoin and carbamazepine, FBM and valproate have broad spectrum anticonvulsant activity [29], but unlike those compounds, FBM and valproate are also prescribed to be effective in treating absence and myoclonic seizures [e.g., 30,31]. Thus, the present study continues our effort to systematically evaluate the effects of various antiseizure medications using the same instrumental learning procedures, affording the
J.J. Orczyk, P.E. Garraghty / Epilepsy & Behavior 78 (2018) 14–19
opportunity for direct side-to-side comparisons of the various medications [see 12]. 2. Materials and methods 2.1. Subjects Data are reported for 55 adult Sprague–Dawley rats that were bred in the care facility in the Psychology Building at Indiana University. All experimental procedures were approved by the Indiana University Animal Care and Use Committee with policies derived from the National Institutes of Health guide for the care and use of Laboratory animals (NIH Publications No. 8023, revised 1978). Because previous publications have employed only female rats [18–20], only females were tested for the present report so that outcomes could be compared across experiments. Thus, we emphasize that generalizability to males for the present and previous results may be limited. Because subjects in the present experiments were tested over 25–56 days involving multiple estrus cycles, these cycles were not monitored, as any within-subject variability that might arise from hormonal fluctuations would be expected to be averaged out over the extended training period. The animals were placed on food restriction beginning 10 days prior to initial training. When target weights were achieved, they were maintained at 85% free-feeding body weight throughout the study. If the animals were to undergo appetitive training, the 45-mg food pellets used as reinforcers (Bio-Serve, Frenchtown, NJ) were introduced to the animals in their home cages at least 2 days prior to the beginning of training. Training was conducted in an operant testing box (Lafayette) placed in a lighted (10-W utility bulb) sound-attenuating chamber equipped with a center-mounted speaker to deliver a 2 kHz (at 90 dB sound pressure level (SPL)) tone. 2.2. Back wire implantation A simple surgical procedure was used to implant two back wires that served as the connection points for the active lead for the delivery of electric shock during aversive training. For this procedure, anesthesia was induced with a mixture of ketamine and xylazine (60 and 6 mg/kg, respectively, IM). As the procedure could be accomplished in a matter of minutes, supplemental doses were rarely required, but given as needed. Two double-loop 30 gauge surgical wires separated by approximately 1 cm apart were implanted subcutaneously between the scapulae of each animal. Animals also received 0.2 cm3 Dopram (IM) and antibiotic ointment on the area of the wires. The entire procedure took approximately 15 min per subject. 2.3. Drug administration Felbamate or water was administered as stated in the detailed methods for each experimental condition. Felbamate-treated animals received two daily doses totaling 3000 mg of the drug per day. One of the dosages was delivered 2 h before training. We have shown this regimen to produce plasma levels within or slightly below the human therapeutic range during the training period [32; and unpublished plasma assay data]. 2.4. Appetitive training All training sessions were separated by 24 h. For appetitive training, the animals were first shaped to bar-press for food reinforcement. When the animal achieved 100 reinforced responses within 30 min on a continuous reinforcement schedule, they were shifted to a fixedratio reinforcement schedule, with one reinforce delivered after four
15
responses. This served to render the behavior more resistant to extinction. Under these conditions, two consecutive days with the animal performing 400 bar presses (i.e., 100 reinforcements) within 30 min were required before tone training was initiated. During the tone-signaled sessions, the tone served as a positive discriminative stimulus. Reinforcement was delivered only for bar presses during the 3-s tone period. One session consisted of 100 tones, each lasting 3 s or until the food pellet was delivered. A reinforced response was followed by a 15-s intertrial interval (ITI) and a randomly determined 1–8-s pretone period. Bar presses during the pretone period restarted the period, and the trial was delayed until no bar presses occurred during the randomly determined pretone period. Appetitive tone training continued for a total of 31 days. 2.5. Aversive training At the conclusion of appetitive training, animals were transferred to the active avoidance task. Aversive training began with the animals receiving a shock that could be terminated by a bar press. The shock intensity was generally maintained at 0.7 mA, and never exceeded 1.0 mA. For the single session of aversive shaping, shock pulses were presented continuously until the bar was pressed. Animals were required to press the bar prior to the onset of the fifth shock pulse at least 15–20 times consecutively. Tone training, using the same 2 kHz tone used for appetitive training, commenced on the following day. For aversive training, the tone served as a discriminative stimulus for an impending foot shock. A bar press during the first 3 s of tone presentation permitted the animal to avoid the shock. If an avoidance response was not produced, the tone and the shock pulsed continued for another 3 s. A bar press during this latter 3 s interval terminated the shock and the tone (i.e., an escape response). The shocks were delivered as a series of four 250-millisecond pulses separated by 500-millisecond periods. Continuous shock pulses were delivered if the animal maintained a bar press for 5 s. This punished the animals for adopting a strategy of holding the bar down for excessive amounts of time (thereby avoiding the shock). Tone trials were separated by 8–12-s ITIs and a variable 2–6-s pretone period. A bar press during the ITI or pretone period reset the pretone period and delayed the initiation of the next trial. One session consisted of 300 tone trials. Aversive training continued for 25 days. 2.6. Experimental conditions 2.6.1. Effects of FBM on appetitive-to-aversive transfer Eight rats began receiving FBM at the conclusion of the 21st day of appetitive training. For 6 additional animals, water treatment was initiated. Appetitive tone training continued for 10 days to assess any possible effect of the drug and/or gavage procedure on the acquired response. Behavioral testing began 2 h after drug or water administration which continued daily throughout the remaining appetitive and total number of avoidance training sessions. A third cohort of rats served as untreated controls (N = 12), receiving no gavage treatment throughout the appetitive and avoidance training. 2.6.2. Effects of FBM on avoidance acquisition without prior appetitive experience Another group of animals began avoidance training 10 days after the initiation of treatment with FBM (N = 8) or water (N = 10). Again, drug or water treatment continued throughout the 25 days of avoidance training. Additional animals (N = 11) underwent aversive training with neither drug nor water delivered via gavage. As above, one session of aversive shaping was followed by the 25 days of tone-signaled avoidance training. All parameters in the aversive context remained as
16
J.J. Orczyk, P.E. Garraghty / Epilepsy & Behavior 78 (2018) 14–19
Fig. 1. Overview of experimental procedure. (A) One group of rats underwent both appetitive and aversive conditioning. Drug administration began 21 days into conditioning in the appetitive context. Both contexts were proceeded by a brief period in which responses were shaped (S). (B) A separate cohort of rats underwent conditioning in the aversive context only. Drug treatment was initiated 11 days before the start of conditioning in the aversive context.
described above, but without prior exposure to the appetitive context. An overview of the two experimental conditions appears in Fig. 1.
3.3. In the absence of prior appetitive experience, a quite different pattern emerges, with the FBM-treated rats showing an acquisition deficit that persists throughout much of the training (Fig. 4)
2.7. Calculations and statistics
Thus, during the first 5 days of training, FBM-treated rats had significantly lower CRs, 19.7 ± 4.53% compared with 42.3 ± 5.85% for controls [F(1,24) = 4.937, p b 0.05]. The average ERs for the FBM-treated animals (0.07 ± 0.029) also differed significantly from the controls
Correct response rates (CR) are simply the number of reinforced responses/100 trials for appetitive training and number of avoidance responses/300 trials for the aversive training. Efficiency ratios (ERs) for appetitive training are simply the number of reinforced responses/ total number of bar presses. For aversive training, ERs are the number of avoidance responses/total number of bar presses. To evaluate differences in performance between the drug-treated and control groups across training days, a mixed-design (split-plot) repeated-measure analysis of variance (ANOVA) model was applied. For the appetitive data, analyses have compared the last five training days prior to the initiation of drug treatment (days 16–20) with the last 5 days of appetitive training (days 27–31). Analyses of avoidance learning have examined acquisition (first 5 days of training) and terminal performance (last 5 days of training). Statistical decisions were based on a 0.05 significance level. The calculations were carried out using SPSS 22.0 software, on a computer running Microsoft Windows 7. Water-treated and untreated controls did not differ in terms of either acquisition (CRs [F(1,15) = 0.747, p N 0.40]; ERs [F(1,15) = 0.289, p N 0.55]) or terminal performance acquisition (CRs [F(1,15) = 0.038, p N 0.80]; ERs [F(1,15) = 0.010, p N 0.90]) on either of the dependent measures. Thus, their data have been collapsed for comparison with the FBM-treated animals. 3. Results 3.1. FBM treatment had no apparent impact on acquired appetitive performance (Fig. 2) A repeated measures ANOVA revealed no difference for either CRs [F(1,14) = 1.783, p N 0.05] or ERs [F(1,14) = 0.724, p N 0.05] between days 16–20, just before the initiation of drug treatment, and terminal performance during days 27–31. 3.2. FBM-treated animals showed enhanced acquisition in the aversive context with prior appetitive training relative to controls (Fig. 3) During the first 5 days of training, FBM-treated rats had significantly higher CRs, averaging 57.5 ± 5.75% compared with 37.8 ± 5.01% [F(1,23) = 4.873, p b 0.05], but not ERs [F(1,23) = 0.147, p N 0.70]. Terminal performance, on the other hand, was comparable to control rats in terms of both CRs [F(1,22) = 0.511 p N 0.40] and ERs [F(1,22) = 0.265, p N 0.60].
Fig. 2. (top) CRs averaged over 5 days across all 31 days of appetitive training. The arrow denotes the beginning of drug treatment on the 21st day of appetitive training. (bottom) ERs averaged over 5 days across all 31 days of appetitive training. The arrow denotes the beginning of drug treatment on the 21st day of appetitive training.
J.J. Orczyk, P.E. Garraghty / Epilepsy & Behavior 78 (2018) 14–19
Fig. 3. (top) CRs averaged over 5 days across all 25 days of aversive training with prior appetitive experience. Felbamate treatment improved acquisition while terminal performance is comparable between the two groups. (bottom) ERs averaged over 5 days across all 25 days of aversive training with prior appetitive experience. Felbamate treatment did not enhance ERs during acquisition as it did CRs.
(0.21 ± 0.035) [F(1,24) = 7.775, p b 0.05]. As is clear from the figure, these deficits in the FBM-treated animals appear to persist throughout much of the training period. When data across all 25 days of training are analyzed, FBM-treated rats differ from controls with respect to both CRs [F(1,24) = 4.433, p b 0.05] and ERs [F(1,24) = 7.371, p b 0.05]. By the final 5 days of training, however, these differences are diminished with respect to both CRs (FBM-treated rats: 55.0 ± 7.12%; Controls: 61.5 ± 6.26%; [F(1,22) = 0.344, p N 0.55]) and, to a lesser descriptive extent, ERs (FBM-treated rats: 0.27 ± 0.053; Controls: 00.45 ± 0.057; [F(1,24) = 3.811, p N 0.05]). 4. Discussion The present study extends our analysis of AEDs, using a variety of behavioral paradigms [18–20,23,32–34], to FBM. The same welldefined instrumental learning and memory paradigm used previously to investigate phenytoin, carbamazepine, and valproate was again employed in testing the effects of FBM. We find that FBM impaired acquisition of the avoidance response in naïve rats, but note that the terminal performance levels of the FBMtreated animals were equivalent to controls. In contrast, rats that had prior experience in the appetitive context showed faster acquisition of the avoidance response relative to controls, while again terminal avoidance performance was equivalent to that in healthy controls. Thus, one
17
Fig. 4. (top) CRs averaged over 5 days across all 25 days of aversive training without prior appetitive experience for control and FBM-treated rats. Felbamate-treated rats suffer from an acquisition deficit. (bottom) ERs averaged over 5 days across all 25 days of aversive training without prior appetitive experience. Felbamate treated rats also had lower ERs throughout training.
cannot conclude that FBM has no effects in this learning paradigm as any departure from the behavior of control animals is perforce abnormal. The abnormalities in the FBM-treated animals are, however, obviously not related simply to an aversive learning deficit. 4.1. Comparisons with other tested medications We have previously reported tests of the behavioral effects of phenytoin, carbamazepine, and valproate in both a Morris water maze task [32] and in the instrumental appetitive-to-aversive transfer task employed here [18–20]. With respect to the Morris water maze, phenytoin had pronounced effects on platform escape learning. The effects of carbamazepine were relatively mild, and valproate-treated animals were unaffected on all of the measures of performance collected in those experiments. Similarly, with respect to aversive learning after appetitive training, phenytoin and carbamazepine impaired the transfer of learning from the appetitive to aversive context, but the impairment was considerably more pronounced with phenytoin [18,19]. By comparison, any effects of valproate were quite modest, at best [20]. In contrast, valproate disrupts avoidance learning in behaviorally naïve animals, while phenytoin and carbamazepine do not. Thus, the overall pattern of results differs between the various drugs tested to date. Phenytoin and carbamazepine impair transfer learning to differing degrees but do not affect avoidance learning in naïve animals. Felbamate and valproate, on the other hand have no
18
J.J. Orczyk, P.E. Garraghty / Epilepsy & Behavior 78 (2018) 14–19
negative effects on transfer learning, while animals treated with either compound show deficits in avoidance learning with no prior appetitive experience. Again, it is noted that this latter deficit in the FBM-treated animals disappears by the end of the training period. Summarizing across all of the data, one can conclude that FBM is associated with the mildest degree of behavioral impairment. To the extent that the instrumental learning task employed here can serve as a proxy for the relative severity of learning/memory deficits associated with these various drugs, one would argue that FBM is the drug of choice for treating seizure disorders when this particular quality of life issue is paramount. Felbamate was originally marketed in early 1993 to treat seizure disorders in adults and children, and the drug was initially viewed as having great promise because of its effectiveness in seizure models, and its apparent neuroprotective effects [e.g., see 35]. However, it was subsequently discovered that FBM was associated with risks of aplastic anemia [e.g., 36], and hepatotoxicity [e.g., 37], with those risks estimated to be 1 in 4000–5000 and 1 in 26,000–34,000, respectively [29]. In both cases, fatalities occurred. For these reasons, FBM was largely withdrawn from use save for individuals with severe and uncontrolled seizures for whom these risks were deemed to be outweighed by the anticonvulsant benefits of the drug. Arguably, had phenytoin, carbamazepine, and valproate been introduced at the same time, they too could well have suffered the same fate as the toxicities associated with these compounds are well-documented [e.g., see 38–41]. The literature on the cognitive side effects of FBM are generally favorable in comparison with other antiepileptic medications (AEDs). Felbamate has been tested in rats and mice using a passive avoidance paradigm. Behavioral side effects were not observed for either species at any dose of FBM tested [42]. Moreover, contrary to the sedating effects of other AEDs, FBM produces stimulant-like side effects [43].
4.2. Avoidance learning in behaviorally naïve animals Naïve FBM-treated rats are impaired in acquiring the avoidance response. Thus, they are impaired in learning the simple aversive association. Modern theorists [e.g., see 44] propose that instrumental avoidance learning proceeds in three stages: Pavlovian fear (threat) conditioning, instrumental action-outcome learning, and habit formation. Here, the acquisition deficit could arise from either of the first two processes (though the avoidance acquisition data from the animals with prior appetitive training discussed below would seem to strongly support the latter). Empirically, this issue should be resolvable with investigations involving the amygdala and the nucleus accumbens (Nacc). Specifically, the Pavlovian fear conditioning appears to involve the convergence of the conditional and unconditional stimuli in the lateral nucleus of the amygdala (LA). The LA, in turn, sends projections to the central nucleus of the amygdala (CeA), the main output pathway of the amygdala with projections to structures that contribute, among other things, to the conditional stimulus (CS)-elicited freezing response [45,46]. While freezing is incompatible with an active avoidance response, we have no index of freezing in the FBM-treated animals. In any event, electrophysiological recording in LA during the avoidance training [47–50] would show whether there was retardation in the rate of Pavlovian fear conditioning. The action-outcome learning also involves LA, but is dependent on an alternate projection of LA to the basal amygdala (BA), and from BA to the shell region of NAcc [e.g., see 51]. Thus, monitoring activity in the LA–BA–NAcc shell pathway could resolve the extent to which the acquisition deficit in naïve FBM-treated rats is due to a disruption in this pathway. In any event, it is worth reemphasizing that this effect was transient such that normal levels of avoidance were achieved by the end of training. Thus, presumably, habit formation was not impaired in these FBM-treated animals.
4.3. Avoidance learning after prior appetitive training For the FBM-treated animals that underwent appetitive training prior to their transfer to aversive training, the avoidance acquisition abnormality is quite different. Here, the drug treated animals actually acquire the avoidance response more quickly than controls. Indeed, if asymptotic performance can serve as an index of habit formation, the FBM-treated animals progress through the Pavlovian fear conditioning and action-outcome stages of the avoidance learning very quickly. We have argued previously [18] that transferring from appetitive to aversive training may be inherently more challenging than learning the avoidance response without prior appetitive experience. Our argument was based on the notion that naïve animals learning the avoidance response were required to merely learn a relatively simple association whereas animals being transferred from appetitive training had also to reconfigure associations acquired during appetitive training. Thus, the tone came to signal an impending shock rather than food availability, and the lever press came to be associated with fundamentally different consequences. In this regard, when avoidance acquisition rates in the control animals reported here are examined closely, the avoidance rate is roughly 25% higher over the first five training days in the animals placed directly in the aversive learning context relative to those transferred from appetitive training, suggesting that the transfer is indeed more challenging. Thus, FBM appears to make the animals somewhat more flexible in the face of the contingency shifts that accompany transfer from appetitive to aversive training. Interestingly, since their avoidance acquisition is abnormal (i.e., differing from the healthy controls), any such increased behavioral flexibility would also have to be characterized as abnormal. We have also previously suggested that any deficits in transfer may be due to enhanced proactive interference [18]. As prefrontal cortex has been implicated in both proactive interference [e.g., 52] and active avoidance learning [e.g., 53], it is possible that a circuit including frontal cortex (possibly including amygdala and nucleus accumbens) is affected by treatment with FBM, contributing to the pattern of avoidance learning shown here. 4.4. Future directions and therapeutic considerations Future studies should extend the boundaries of our behavioral observations by investigating the effects of various antiepileptic drugs delivered directly to sites in the amygdala, nucleus accumbens, and prefrontal cortex. Furthermore, a number of cannabinoids are now emerging as potential new AEDs, particularly useful for the treatment of intractable and treatment-resistant epilepsy [54; see numerous references in 55]. These compounds need to be tested for cognitive impairment and compared with existing medications. Finally, as psychologists, and as others have suggested [e.g., 56], we would argue that negative cognitive side effects of any medication should be a substantial consideration in the selection of a treatment strategy. This is an issue that is presently receiving an increasing amount of attention with respect to the use of chemotherapeutic approaches to the treatment of cancer [e.g., 57]. Indeed, one can argue that consideration of cognitive quality-of-life issues should be paramount. Funding We thank the Indiana University Harlan Family Scholars Program for their generous support of our research. We thank the Harlan Family Behavioral Neuroscience Research Innovation Fund IUF account 0370004846 for support. Conflict of interest The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as potential conflicts of interest.
J.J. Orczyk, P.E. Garraghty / Epilepsy & Behavior 78 (2018) 14–19
References [1] Thompson PJ, Trimble MR. Anticonvulsant drugs and cognitive functions. Epilepsia 1982;23:531–44. [2] Pulliainen V, Jokelainen M. Effects of phenytoin and carbamazepine on cognitive functions in newly diagnosed epileptic patients. Acta Neurol Scand 1994;89:81-66. [3] Thompson PJ, Trimble MR. Sodium valproate and cognitive functioning in normal volunteers. Brit. J Clin Pharmacol 1981;12:819–24. [4] Butlin AT, Danta G, Cook ML. Anticonvulsant effects on the memory performance of epileptics. Clin Exp Neurol 1983;20:27–35. [5] Devinsky O. Cognitive and behavioral effects of antiepileptic drugs. Epilepsia 1995; 36(s2):S46–65. [6] Vermeulen J, Aldenkamp AP. Cognitive side-effects of chronic antiepileptic drug treatment: a review of 25 years of research. Epilepsy Res 1995;22:65–95. [7] Armon C, Shin C, Miller P, Carwile S, Brown E, Edinger JD, et al. Reversible parkinsonism and cognitive impairment with chronic valproate use. Neurology 1996;47: 626–35. [8] Meador KJ, Gilliam FG, Kanner AM, Pellock JM. Cognitive and behavioral effects of antiepileptic drugs. Epilepsy Behav 2001;2:SS1–SS17. [9] Drane DL, Meador KJ. Cognitive and behavioral effects of antiepileptic drugs. Epilepsy Behav 2002;3:49–53. [10] Gates JR. Side effect profiles and behavioral consequences of antiepileptic medications. Epilepsy Behav 2000;1:153–9. [11] Loring DW, Meador KJ. Cognitive and behavioral effects of epilepsy treatment. Epilepsia 2001;42(Suppl. 8):24–32. [12] Ortinski P, Meador KJ. Cognitive side effects of antiepileptic drugs. Epilepsy Behav 2004;5:S60–5. [13] Poling A, Blakely, White W, Picker M. Chronic effects of clonazepam, phenytoin, ethosuximide, and valproic acid on learning in pigeons as assayed by a repeated acquisition procedure. Pharmacol Biochem Behav 1986;24:1583–6. [14] Deacon RMJ. Pharmacological studies of a rat spatial delayed nonmatch-to-sample task as an animal model of dementia. Drug Dev Res 1991;24:67–79. [15] Umka J, Mustafa S, El Beltagy M, Thorpe A, Latif L, Bennett G, et al. Valproic acid reduces spatial working memory and cell proliferation in the hippocampus. Neuroscience 2010;166:15–22. [16] Continella G, Drago F, Amico-Roxas M, Scapagnini U. Effects of valproic acid derivatives on the acquisition of active avoidance behavior in the rat. Pharmacol Res Commun 1984;16:699–706. [17] Murphy KJ, Fox GB, Foley AG, Gallagher HC, O'Connell A, Griffin AM, et al. Pentyl-4yn-valproic acid enhances both spatial and avoidance learning and attenuates agerelated NCAM-mediated neuroplastic decline within the rat medial temporal lobe. J Neurochem 2001;78:704–14. [18] Banks MK, Mohr NL, Besheer J, Steinmetz JE, Garraghty PE. The effects of phenytoin on instrumental appetitive-to-aversive transfer in rats. Pharmacol Biochem Behav 1999;63:465–72. [19] Banks MK, Besheer J, Szypczak J, Goodpaster LL, Phipps EJ, Garraghty PE. The effects of carbamazepine on an appetitive-to-aversive transfer task: comparison to untreated and phenytoin. Prog Neuropsychopharmacol Biol Psychiatry 2001;25:551–72. [20] Orczyk JJ, Banks MK, Garraghty PE. The effects of valproic acid on appetitive and aversive instrumental learning in adult rats. Front Behav Neurosci 2014;8:1–9 Available from: http://journal.frontiersin.org/article/https://doi.org/10.3389/fnbeh. 2014.00113/full. [21] Steinmetz JE, Logue SF, Miller DP. Using signaled bar-pressing tasks to study the neural substrates of appetitive and aversive learning in rats: behavioral manipulations and cerebellar lesions. Behav Neurosci 1993;107:941–54. [22] Logue SF. An examination of the role of the hippocampus in the acquisition and timing of a signaled barpress in appetitive and aversive conditioning [dissertation]. Bloomington (IN): Indiana University; 1994. [23] McDowell AL, Samuelson DL, Dina BS, Garraghty PE. Discrete and contextual cue alterations eliminate the instrumental appetitive-to-aversive transfer impairment in phenytoin-treated rats. Integr Physiol Behav Sci 2004;39:307–17. [24] Mowery TM, McDowell AL, Garraghty PE. Chronic developmental exposure to phenytoin has long-term behavioral consequences. Int J Dev Neurosci 2008;26:401–7. [25] Butt AE, Schultz JA, Arnold LL, Garman EE, George CL, Garraghty PE. Lesions of the rat nucleus basalis magnocellularis disrupt appetitive-to-aversive transfer learning. Integr Physiol Behav Sci 2003;38:253–71. [26] Goodman MR, Garman, Arnold LL, Sengelaub DR, Garraghty PE. The effects of estradiol on avoidance learning in ovariectomized adult rats. Integr Physiol Behav Sci 2004;39:192–206. [27] McDowell AL, Heath KM, Garraghty PE. The effects of sex and chronic restraint on instrumental learning in rats. Neurosci J 2013:893126 (12 pages). Available from: https://doi.org/10.1155/2013/893126. [28] McDowell AL, Fransen KMH, Elliot KS, Elghouche A, Kostylev PV, O'Dea PK, et al. Sex differences and the impact of chronic stress and recovery from on instrumental
[29]
[30]
[31] [32]
[33]
[34]
[35] [36] [37] [38] [39] [40] [41]
[42] [43]
[44]
[45]
[46]
[47] [48]
[49]
[50]
[51] [52] [53]
[54] [55] [56] [57]
19
learning. Neurosci J 2015:697659 (12 pages). Available from: https://doi.org/10. 1155/2015/697659. Bialer M, Johannessen SI, Kupferberg HJ, Levy RH, Perucca E, Tomson T. Progress report on new antiepileptic drugs: a summary of the Eighth Eilat Conference (EILAT VIII). Epilepsy Res 2007;73:1–52. Jeavons PM, Clark JE, Maheshwari MC. Treatment of generalized epilepsies of childhood and adolescence with sodium valproate (‘epilim’). Dev Med Child Neurol 1977;19:9–25. Bourgeois BFD. New anti-epileptic drugs. Arch Neurol 1998;55:1181–3. Churchill JD, Fang PC, Voss SE, Besheer J, Herron AL, Garraghty PE. Some antiepileptic compounds impair learning by rats in a Morris water maze. Integr Physiol Behav Sci 2003;38:91–103. Churchill JD, Voss SE, Miller DP, Steinmetz JE, Garraghty PE. Phenytoin blocks the reversal of a classically conditioned discriminative eyeblink response in rabbits. Epilepsia 1998;39:584–9. Samuelson DL, Arnold LL, Mowery TM, Mesnard A, Garraghty PE. The effects of phenytoin on the performance of rats in a delayed match-to-place task. Integr Physiol Behav Sci 2005;40:17–27. Sofia RD. Felbamate: mechanisms of action. In: Levy RH, Mattson RH, Meldrum BS, editors. Antiepileptic drugs. 4th ed. New York: Raven Press; 1999. p. 791–7. Kaufman D, Kelly JP, Anderson T, Harmon DC, Shapiro S. Evaluation of case reports of aplastic anemia among patients treated with felbamate. Epilepsia 1977;38:1265–9. Pellock JM. Felbamate. Epilepsia 1999;40(s5):s57–62. Bruni J. Phenytoin: toxicity. In: Levy RH, Mattson RH, Meldrum BS, editors. Antiepileptic drugs. 4th ed. New York: Raven Press; 1995. p. 345–50. Dreifuss FE. Valproic acid: toxicity. In: Levy RH, Mattson RH, Meldrum BS, editors. Antiepileptic drugs. 4th ed. New York: Raven Press; 1995. p. 641–8. Holmes GL. Carbamazepine: toxicity. In: Levy RH, Mattson RH, Meldrum BS, editors. Antiepileptic drugs. 4th ed. New York: Raven Press; 1995. p. 567–79. McNamara JO. Drugs effective in the therapy of the epilepsies. In: Hardman JG, Limbird LE, Gilman AG, editors. Goodman and Gilman's the pharmalogical basis of therapeutics. 10th ed. New York: McGraw-Hill; 2001. p. 521–47. Smith RD, Grzelak ME, Coffin VL. Felbamate, a novel antiepileptic agent, does not affect cognition in rodents. Behav Pharmacol 1994;5:365–8. Ketter TA, Malow BA, Flamini R, Ko D, White SR, Post RM, et al. Felbamate monotherapy has stimulant-like effects in patients with epilepsy. Epilepsy Res 1996;23: 129–37. LeDoux JE, Moscarello J, Sears R, Campese V. The birth, death and resurrection of avoidance: a reconceptualization of a troubled paradigm. Mol Psychiatry 2017;22: 24–36. LeDoux JE, Iwata J, Cicchetti P, Reis DJ. Different projections of the central amydaloid nucleus mediate autonomic and behavioral correlates of conditioned fear. J Neurosci 1988;8:2517–29. Kim JJ, Rison RA, Fanselow MS. Effects of amygdala, hippocampus, and periaquductal gray lesions on short- and long-term contextual fear. Behav Neurosci 1993;107: 1093–8. Rogan MT, Staubli UV, LeDoux JE. Fear conditioning induces associative long-term potentiation in the amygdala. Nature 1997;390:604–7. Quirk GJ, Armony JL, LeDoux JE. Fear conditioning enhances different temporal components of tone-evoked spike trains in auditory cortex and lateral amygdala. Neuron 1997;19:613–24. Maren S. Auditory fear conditioning increases CS-elicited spike firing in lateral amygdala neurons even after extensive overtraining. Eur J Neurosci 2000;12: 4047–54. Repa JC, Muller J, Apergis J, Desrochers TM, Zhou Y, LeDoux JE. Two different lateral amygdala cell populations contribute to theinitiation and storage of memory. Nat Neurosci 2001;4:724–31. Ramirez F, Moscarello JM, LeDoux JE, Sears RM. Active avoidance requires a serial basal amygdala to nucleus accumbens shell circuit. J Neurosci 2015;25:3470–7. Moscovitch M. Multiple dissociations of function in amnesia. In: Cermak LS, editor. Human memory and amnesia. Hillsdale, NJ: Erlbaum; 1982. p. 337–70. Jiao X, Beck KD, Myers CE, Servatius RJ, Pang KC. Altered activity of the medial prefrontal cortex and amygdala during acquisition and extinction of an active avoidance task. Front Behav Neurosci 2015;9:00249. https://doi.org/10.3389/fnbeh2015. 00249. Cilio MR, Thiele EA, Devinsky O. The case for assessing cannabidiol in epilepsy. Epilepsia 2014;55:787–90. Szaflarsk J, Devinsky O, editors. Cannabinoids and epilepsy, Epilepsy Behav 2017; 70(Part B):277–370. Schmitz B. Effects of antiepileptic drugs on mood and behavior. Epilepsia 2006; 47(Suppl. 2):28–33. Janelsins MC, Kohli S, Mohile SG, Usuki K, Ahles TA, Morrow GR. An update on cancer- and chemotherapy-related cognitive dysfunction: current status. Semin Oncol 2011;38:431–8.
Contents lists available at ScienceDirect
Epilepsy & Behavior journal homepage: www.elsevier.com/locate/yebeh
The effects of felbamate on appetitive and aversive instrumental learning in adult rats John J. Orczyk a, Preston E. Garraghty a,b,⁎ a b
Department of Psychological and Brain Sciences, Indiana University, Bloomington, IN, USA Program in Neuroscience, Indiana University, Bloomington, IN, USA
a r t i c l e
i n f o
Article history: Received 24 July 2017 Revised 19 October 2017 Accepted 19 October 2017 Available online xxxx Keywords: Antiepileptic Anticonvulsant Felbamate Instrumental learning Cognition
a b s t r a c t Antiepileptic medications are the frontline treatment for seizure conditions but are not without cognitive side effects. Previously, our laboratory reported learning deficits in phenytoin-, carbamazepine-, and valproatetreated rats. In the present experiment, the effects of felbamate (FBM) have been compared to water-treated controls (controls) using the same instrumental training tasks employed here. Rats treated with FBM displayed a deficit in acquiring a tone-signaled avoidance response, relative to controls, but this was true only if they had no prior appetitive experience. Terminal avoidance behavior was equivalent to healthy controls. In contrast, the FBM-treated rats showed enhanced acquisition of the avoidance response relative to controls when given the benefit of prior experience in the appetitive condition. Relative to animals treated with phenytoin, carbamazepine, or valproate, FBM-treated rats showed the lowest overall pattern of deficits using these instrumental learning tasks. While FBM treatment has been severely restricted because of rather low risks of serious medical side effects, we suggest that the risks are not substantially higher than those shown to exist for phenytoin, carbamazepine, or valproate. As psychologists, we further suggest that negative cognitive deficits associated with these various drugs, along with their quality-of-life costs, are of relevance in the design of treatment strategies for individuals with seizure disorders. © 2017 Elsevier Inc. All rights reserved.
1. Introduction Antiepileptic drugs (AEDs) are the frontline treatment for epilepsy but are not without cognitive side effects [e.g., 1,2]. Many studies have reported cognitive deficits in human patients treated with AEDs [e.g., 3–9]; for reviews, see [10–12]. Performance deficits have also been observed in animals treated with AEDs [e.g., 13–15], though other studies performed in animals have failed to detect cognitive deficits associated with AED treatment [16,17]. Previously, our laboratory has reported different types of behavioral deficits in adult rats treated with the AEDs phenytoin [18], carbamazepine [19], and valproate [20]. It was shown that phenytoin, and to a lesser degree, carbamazepine, both frontline AEDs, blocked the acquisition of an avoidance response in the second part of an instrumental appetitive-to-aversive transfer conditioning task [18,19]. More recently, we have shown that valproate produces a different pattern of deficits, impairing the acquisition of an avoidance response in the absence of prior appetitive training, but having no effect on avoidance learning in rats transferred from appetitive training [20].
⁎ Corresponding author at: Department of Psychological and Brain Sciences, Indiana University, 1101 E 10th Street, Bloomington, IN 47405, USA. E-mail address: [email protected] (P.E. Garraghty).
https://doi.org/10.1016/j.yebeh.2017.10.022 1525-5050/© 2017 Elsevier Inc. All rights reserved.
In the present study, we have employed this same withinsubject, tone-signaled bar press task in which rats are tested in both appetitive and aversive contexts. The task employed is complex and multicontextual. There are multiple rules in the aversive context that must be learned through conditioning. In the aversive context, rats must learn to both press the lever after the tone and not to press the lever during the intertrial period. This paradigm was developed to study appetitive and aversive learning in the same subjects and has been used in past work to evaluate learning, memory, and impairments that accompany cerebellar, hippocampal, cingulate, and prefrontal cortex lesions [21,22]. We have used this behavioral paradigm to evaluate the effects of phenytoin, carbamazepine, and valproate in adult rats [18–20,23], rats exposed to phenytoin in utero [24], rats with lesions of the basal nucleus of Meynart [25], ovariectomized female rats with or without estradiol replacement [26], and rats undergoing chronic restraint [27,28]. In the current study, we have extended our assessment of the cognitive side effects of AEDs to felbamate (FBM). Like phenytoin and carbamazepine, FBM and valproate have broad spectrum anticonvulsant activity [29], but unlike those compounds, FBM and valproate are also prescribed to be effective in treating absence and myoclonic seizures [e.g., 30,31]. Thus, the present study continues our effort to systematically evaluate the effects of various antiseizure medications using the same instrumental learning procedures, affording the
J.J. Orczyk, P.E. Garraghty / Epilepsy & Behavior 78 (2018) 14–19
opportunity for direct side-to-side comparisons of the various medications [see 12]. 2. Materials and methods 2.1. Subjects Data are reported for 55 adult Sprague–Dawley rats that were bred in the care facility in the Psychology Building at Indiana University. All experimental procedures were approved by the Indiana University Animal Care and Use Committee with policies derived from the National Institutes of Health guide for the care and use of Laboratory animals (NIH Publications No. 8023, revised 1978). Because previous publications have employed only female rats [18–20], only females were tested for the present report so that outcomes could be compared across experiments. Thus, we emphasize that generalizability to males for the present and previous results may be limited. Because subjects in the present experiments were tested over 25–56 days involving multiple estrus cycles, these cycles were not monitored, as any within-subject variability that might arise from hormonal fluctuations would be expected to be averaged out over the extended training period. The animals were placed on food restriction beginning 10 days prior to initial training. When target weights were achieved, they were maintained at 85% free-feeding body weight throughout the study. If the animals were to undergo appetitive training, the 45-mg food pellets used as reinforcers (Bio-Serve, Frenchtown, NJ) were introduced to the animals in their home cages at least 2 days prior to the beginning of training. Training was conducted in an operant testing box (Lafayette) placed in a lighted (10-W utility bulb) sound-attenuating chamber equipped with a center-mounted speaker to deliver a 2 kHz (at 90 dB sound pressure level (SPL)) tone. 2.2. Back wire implantation A simple surgical procedure was used to implant two back wires that served as the connection points for the active lead for the delivery of electric shock during aversive training. For this procedure, anesthesia was induced with a mixture of ketamine and xylazine (60 and 6 mg/kg, respectively, IM). As the procedure could be accomplished in a matter of minutes, supplemental doses were rarely required, but given as needed. Two double-loop 30 gauge surgical wires separated by approximately 1 cm apart were implanted subcutaneously between the scapulae of each animal. Animals also received 0.2 cm3 Dopram (IM) and antibiotic ointment on the area of the wires. The entire procedure took approximately 15 min per subject. 2.3. Drug administration Felbamate or water was administered as stated in the detailed methods for each experimental condition. Felbamate-treated animals received two daily doses totaling 3000 mg of the drug per day. One of the dosages was delivered 2 h before training. We have shown this regimen to produce plasma levels within or slightly below the human therapeutic range during the training period [32; and unpublished plasma assay data]. 2.4. Appetitive training All training sessions were separated by 24 h. For appetitive training, the animals were first shaped to bar-press for food reinforcement. When the animal achieved 100 reinforced responses within 30 min on a continuous reinforcement schedule, they were shifted to a fixedratio reinforcement schedule, with one reinforce delivered after four
15
responses. This served to render the behavior more resistant to extinction. Under these conditions, two consecutive days with the animal performing 400 bar presses (i.e., 100 reinforcements) within 30 min were required before tone training was initiated. During the tone-signaled sessions, the tone served as a positive discriminative stimulus. Reinforcement was delivered only for bar presses during the 3-s tone period. One session consisted of 100 tones, each lasting 3 s or until the food pellet was delivered. A reinforced response was followed by a 15-s intertrial interval (ITI) and a randomly determined 1–8-s pretone period. Bar presses during the pretone period restarted the period, and the trial was delayed until no bar presses occurred during the randomly determined pretone period. Appetitive tone training continued for a total of 31 days. 2.5. Aversive training At the conclusion of appetitive training, animals were transferred to the active avoidance task. Aversive training began with the animals receiving a shock that could be terminated by a bar press. The shock intensity was generally maintained at 0.7 mA, and never exceeded 1.0 mA. For the single session of aversive shaping, shock pulses were presented continuously until the bar was pressed. Animals were required to press the bar prior to the onset of the fifth shock pulse at least 15–20 times consecutively. Tone training, using the same 2 kHz tone used for appetitive training, commenced on the following day. For aversive training, the tone served as a discriminative stimulus for an impending foot shock. A bar press during the first 3 s of tone presentation permitted the animal to avoid the shock. If an avoidance response was not produced, the tone and the shock pulsed continued for another 3 s. A bar press during this latter 3 s interval terminated the shock and the tone (i.e., an escape response). The shocks were delivered as a series of four 250-millisecond pulses separated by 500-millisecond periods. Continuous shock pulses were delivered if the animal maintained a bar press for 5 s. This punished the animals for adopting a strategy of holding the bar down for excessive amounts of time (thereby avoiding the shock). Tone trials were separated by 8–12-s ITIs and a variable 2–6-s pretone period. A bar press during the ITI or pretone period reset the pretone period and delayed the initiation of the next trial. One session consisted of 300 tone trials. Aversive training continued for 25 days. 2.6. Experimental conditions 2.6.1. Effects of FBM on appetitive-to-aversive transfer Eight rats began receiving FBM at the conclusion of the 21st day of appetitive training. For 6 additional animals, water treatment was initiated. Appetitive tone training continued for 10 days to assess any possible effect of the drug and/or gavage procedure on the acquired response. Behavioral testing began 2 h after drug or water administration which continued daily throughout the remaining appetitive and total number of avoidance training sessions. A third cohort of rats served as untreated controls (N = 12), receiving no gavage treatment throughout the appetitive and avoidance training. 2.6.2. Effects of FBM on avoidance acquisition without prior appetitive experience Another group of animals began avoidance training 10 days after the initiation of treatment with FBM (N = 8) or water (N = 10). Again, drug or water treatment continued throughout the 25 days of avoidance training. Additional animals (N = 11) underwent aversive training with neither drug nor water delivered via gavage. As above, one session of aversive shaping was followed by the 25 days of tone-signaled avoidance training. All parameters in the aversive context remained as
16
J.J. Orczyk, P.E. Garraghty / Epilepsy & Behavior 78 (2018) 14–19
Fig. 1. Overview of experimental procedure. (A) One group of rats underwent both appetitive and aversive conditioning. Drug administration began 21 days into conditioning in the appetitive context. Both contexts were proceeded by a brief period in which responses were shaped (S). (B) A separate cohort of rats underwent conditioning in the aversive context only. Drug treatment was initiated 11 days before the start of conditioning in the aversive context.
described above, but without prior exposure to the appetitive context. An overview of the two experimental conditions appears in Fig. 1.
3.3. In the absence of prior appetitive experience, a quite different pattern emerges, with the FBM-treated rats showing an acquisition deficit that persists throughout much of the training (Fig. 4)
2.7. Calculations and statistics
Thus, during the first 5 days of training, FBM-treated rats had significantly lower CRs, 19.7 ± 4.53% compared with 42.3 ± 5.85% for controls [F(1,24) = 4.937, p b 0.05]. The average ERs for the FBM-treated animals (0.07 ± 0.029) also differed significantly from the controls
Correct response rates (CR) are simply the number of reinforced responses/100 trials for appetitive training and number of avoidance responses/300 trials for the aversive training. Efficiency ratios (ERs) for appetitive training are simply the number of reinforced responses/ total number of bar presses. For aversive training, ERs are the number of avoidance responses/total number of bar presses. To evaluate differences in performance between the drug-treated and control groups across training days, a mixed-design (split-plot) repeated-measure analysis of variance (ANOVA) model was applied. For the appetitive data, analyses have compared the last five training days prior to the initiation of drug treatment (days 16–20) with the last 5 days of appetitive training (days 27–31). Analyses of avoidance learning have examined acquisition (first 5 days of training) and terminal performance (last 5 days of training). Statistical decisions were based on a 0.05 significance level. The calculations were carried out using SPSS 22.0 software, on a computer running Microsoft Windows 7. Water-treated and untreated controls did not differ in terms of either acquisition (CRs [F(1,15) = 0.747, p N 0.40]; ERs [F(1,15) = 0.289, p N 0.55]) or terminal performance acquisition (CRs [F(1,15) = 0.038, p N 0.80]; ERs [F(1,15) = 0.010, p N 0.90]) on either of the dependent measures. Thus, their data have been collapsed for comparison with the FBM-treated animals. 3. Results 3.1. FBM treatment had no apparent impact on acquired appetitive performance (Fig. 2) A repeated measures ANOVA revealed no difference for either CRs [F(1,14) = 1.783, p N 0.05] or ERs [F(1,14) = 0.724, p N 0.05] between days 16–20, just before the initiation of drug treatment, and terminal performance during days 27–31. 3.2. FBM-treated animals showed enhanced acquisition in the aversive context with prior appetitive training relative to controls (Fig. 3) During the first 5 days of training, FBM-treated rats had significantly higher CRs, averaging 57.5 ± 5.75% compared with 37.8 ± 5.01% [F(1,23) = 4.873, p b 0.05], but not ERs [F(1,23) = 0.147, p N 0.70]. Terminal performance, on the other hand, was comparable to control rats in terms of both CRs [F(1,22) = 0.511 p N 0.40] and ERs [F(1,22) = 0.265, p N 0.60].
Fig. 2. (top) CRs averaged over 5 days across all 31 days of appetitive training. The arrow denotes the beginning of drug treatment on the 21st day of appetitive training. (bottom) ERs averaged over 5 days across all 31 days of appetitive training. The arrow denotes the beginning of drug treatment on the 21st day of appetitive training.
J.J. Orczyk, P.E. Garraghty / Epilepsy & Behavior 78 (2018) 14–19
Fig. 3. (top) CRs averaged over 5 days across all 25 days of aversive training with prior appetitive experience. Felbamate treatment improved acquisition while terminal performance is comparable between the two groups. (bottom) ERs averaged over 5 days across all 25 days of aversive training with prior appetitive experience. Felbamate treatment did not enhance ERs during acquisition as it did CRs.
(0.21 ± 0.035) [F(1,24) = 7.775, p b 0.05]. As is clear from the figure, these deficits in the FBM-treated animals appear to persist throughout much of the training period. When data across all 25 days of training are analyzed, FBM-treated rats differ from controls with respect to both CRs [F(1,24) = 4.433, p b 0.05] and ERs [F(1,24) = 7.371, p b 0.05]. By the final 5 days of training, however, these differences are diminished with respect to both CRs (FBM-treated rats: 55.0 ± 7.12%; Controls: 61.5 ± 6.26%; [F(1,22) = 0.344, p N 0.55]) and, to a lesser descriptive extent, ERs (FBM-treated rats: 0.27 ± 0.053; Controls: 00.45 ± 0.057; [F(1,24) = 3.811, p N 0.05]). 4. Discussion The present study extends our analysis of AEDs, using a variety of behavioral paradigms [18–20,23,32–34], to FBM. The same welldefined instrumental learning and memory paradigm used previously to investigate phenytoin, carbamazepine, and valproate was again employed in testing the effects of FBM. We find that FBM impaired acquisition of the avoidance response in naïve rats, but note that the terminal performance levels of the FBMtreated animals were equivalent to controls. In contrast, rats that had prior experience in the appetitive context showed faster acquisition of the avoidance response relative to controls, while again terminal avoidance performance was equivalent to that in healthy controls. Thus, one
17
Fig. 4. (top) CRs averaged over 5 days across all 25 days of aversive training without prior appetitive experience for control and FBM-treated rats. Felbamate-treated rats suffer from an acquisition deficit. (bottom) ERs averaged over 5 days across all 25 days of aversive training without prior appetitive experience. Felbamate treated rats also had lower ERs throughout training.
cannot conclude that FBM has no effects in this learning paradigm as any departure from the behavior of control animals is perforce abnormal. The abnormalities in the FBM-treated animals are, however, obviously not related simply to an aversive learning deficit. 4.1. Comparisons with other tested medications We have previously reported tests of the behavioral effects of phenytoin, carbamazepine, and valproate in both a Morris water maze task [32] and in the instrumental appetitive-to-aversive transfer task employed here [18–20]. With respect to the Morris water maze, phenytoin had pronounced effects on platform escape learning. The effects of carbamazepine were relatively mild, and valproate-treated animals were unaffected on all of the measures of performance collected in those experiments. Similarly, with respect to aversive learning after appetitive training, phenytoin and carbamazepine impaired the transfer of learning from the appetitive to aversive context, but the impairment was considerably more pronounced with phenytoin [18,19]. By comparison, any effects of valproate were quite modest, at best [20]. In contrast, valproate disrupts avoidance learning in behaviorally naïve animals, while phenytoin and carbamazepine do not. Thus, the overall pattern of results differs between the various drugs tested to date. Phenytoin and carbamazepine impair transfer learning to differing degrees but do not affect avoidance learning in naïve animals. Felbamate and valproate, on the other hand have no
18
J.J. Orczyk, P.E. Garraghty / Epilepsy & Behavior 78 (2018) 14–19
negative effects on transfer learning, while animals treated with either compound show deficits in avoidance learning with no prior appetitive experience. Again, it is noted that this latter deficit in the FBM-treated animals disappears by the end of the training period. Summarizing across all of the data, one can conclude that FBM is associated with the mildest degree of behavioral impairment. To the extent that the instrumental learning task employed here can serve as a proxy for the relative severity of learning/memory deficits associated with these various drugs, one would argue that FBM is the drug of choice for treating seizure disorders when this particular quality of life issue is paramount. Felbamate was originally marketed in early 1993 to treat seizure disorders in adults and children, and the drug was initially viewed as having great promise because of its effectiveness in seizure models, and its apparent neuroprotective effects [e.g., see 35]. However, it was subsequently discovered that FBM was associated with risks of aplastic anemia [e.g., 36], and hepatotoxicity [e.g., 37], with those risks estimated to be 1 in 4000–5000 and 1 in 26,000–34,000, respectively [29]. In both cases, fatalities occurred. For these reasons, FBM was largely withdrawn from use save for individuals with severe and uncontrolled seizures for whom these risks were deemed to be outweighed by the anticonvulsant benefits of the drug. Arguably, had phenytoin, carbamazepine, and valproate been introduced at the same time, they too could well have suffered the same fate as the toxicities associated with these compounds are well-documented [e.g., see 38–41]. The literature on the cognitive side effects of FBM are generally favorable in comparison with other antiepileptic medications (AEDs). Felbamate has been tested in rats and mice using a passive avoidance paradigm. Behavioral side effects were not observed for either species at any dose of FBM tested [42]. Moreover, contrary to the sedating effects of other AEDs, FBM produces stimulant-like side effects [43].
4.2. Avoidance learning in behaviorally naïve animals Naïve FBM-treated rats are impaired in acquiring the avoidance response. Thus, they are impaired in learning the simple aversive association. Modern theorists [e.g., see 44] propose that instrumental avoidance learning proceeds in three stages: Pavlovian fear (threat) conditioning, instrumental action-outcome learning, and habit formation. Here, the acquisition deficit could arise from either of the first two processes (though the avoidance acquisition data from the animals with prior appetitive training discussed below would seem to strongly support the latter). Empirically, this issue should be resolvable with investigations involving the amygdala and the nucleus accumbens (Nacc). Specifically, the Pavlovian fear conditioning appears to involve the convergence of the conditional and unconditional stimuli in the lateral nucleus of the amygdala (LA). The LA, in turn, sends projections to the central nucleus of the amygdala (CeA), the main output pathway of the amygdala with projections to structures that contribute, among other things, to the conditional stimulus (CS)-elicited freezing response [45,46]. While freezing is incompatible with an active avoidance response, we have no index of freezing in the FBM-treated animals. In any event, electrophysiological recording in LA during the avoidance training [47–50] would show whether there was retardation in the rate of Pavlovian fear conditioning. The action-outcome learning also involves LA, but is dependent on an alternate projection of LA to the basal amygdala (BA), and from BA to the shell region of NAcc [e.g., see 51]. Thus, monitoring activity in the LA–BA–NAcc shell pathway could resolve the extent to which the acquisition deficit in naïve FBM-treated rats is due to a disruption in this pathway. In any event, it is worth reemphasizing that this effect was transient such that normal levels of avoidance were achieved by the end of training. Thus, presumably, habit formation was not impaired in these FBM-treated animals.
4.3. Avoidance learning after prior appetitive training For the FBM-treated animals that underwent appetitive training prior to their transfer to aversive training, the avoidance acquisition abnormality is quite different. Here, the drug treated animals actually acquire the avoidance response more quickly than controls. Indeed, if asymptotic performance can serve as an index of habit formation, the FBM-treated animals progress through the Pavlovian fear conditioning and action-outcome stages of the avoidance learning very quickly. We have argued previously [18] that transferring from appetitive to aversive training may be inherently more challenging than learning the avoidance response without prior appetitive experience. Our argument was based on the notion that naïve animals learning the avoidance response were required to merely learn a relatively simple association whereas animals being transferred from appetitive training had also to reconfigure associations acquired during appetitive training. Thus, the tone came to signal an impending shock rather than food availability, and the lever press came to be associated with fundamentally different consequences. In this regard, when avoidance acquisition rates in the control animals reported here are examined closely, the avoidance rate is roughly 25% higher over the first five training days in the animals placed directly in the aversive learning context relative to those transferred from appetitive training, suggesting that the transfer is indeed more challenging. Thus, FBM appears to make the animals somewhat more flexible in the face of the contingency shifts that accompany transfer from appetitive to aversive training. Interestingly, since their avoidance acquisition is abnormal (i.e., differing from the healthy controls), any such increased behavioral flexibility would also have to be characterized as abnormal. We have also previously suggested that any deficits in transfer may be due to enhanced proactive interference [18]. As prefrontal cortex has been implicated in both proactive interference [e.g., 52] and active avoidance learning [e.g., 53], it is possible that a circuit including frontal cortex (possibly including amygdala and nucleus accumbens) is affected by treatment with FBM, contributing to the pattern of avoidance learning shown here. 4.4. Future directions and therapeutic considerations Future studies should extend the boundaries of our behavioral observations by investigating the effects of various antiepileptic drugs delivered directly to sites in the amygdala, nucleus accumbens, and prefrontal cortex. Furthermore, a number of cannabinoids are now emerging as potential new AEDs, particularly useful for the treatment of intractable and treatment-resistant epilepsy [54; see numerous references in 55]. These compounds need to be tested for cognitive impairment and compared with existing medications. Finally, as psychologists, and as others have suggested [e.g., 56], we would argue that negative cognitive side effects of any medication should be a substantial consideration in the selection of a treatment strategy. This is an issue that is presently receiving an increasing amount of attention with respect to the use of chemotherapeutic approaches to the treatment of cancer [e.g., 57]. Indeed, one can argue that consideration of cognitive quality-of-life issues should be paramount. Funding We thank the Indiana University Harlan Family Scholars Program for their generous support of our research. We thank the Harlan Family Behavioral Neuroscience Research Innovation Fund IUF account 0370004846 for support. Conflict of interest The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as potential conflicts of interest.
J.J. Orczyk, P.E. Garraghty / Epilepsy & Behavior 78 (2018) 14–19
References [1] Thompson PJ, Trimble MR. Anticonvulsant drugs and cognitive functions. Epilepsia 1982;23:531–44. [2] Pulliainen V, Jokelainen M. Effects of phenytoin and carbamazepine on cognitive functions in newly diagnosed epileptic patients. Acta Neurol Scand 1994;89:81-66. [3] Thompson PJ, Trimble MR. Sodium valproate and cognitive functioning in normal volunteers. Brit. J Clin Pharmacol 1981;12:819–24. [4] Butlin AT, Danta G, Cook ML. Anticonvulsant effects on the memory performance of epileptics. Clin Exp Neurol 1983;20:27–35. [5] Devinsky O. Cognitive and behavioral effects of antiepileptic drugs. Epilepsia 1995; 36(s2):S46–65. [6] Vermeulen J, Aldenkamp AP. Cognitive side-effects of chronic antiepileptic drug treatment: a review of 25 years of research. Epilepsy Res 1995;22:65–95. [7] Armon C, Shin C, Miller P, Carwile S, Brown E, Edinger JD, et al. Reversible parkinsonism and cognitive impairment with chronic valproate use. Neurology 1996;47: 626–35. [8] Meador KJ, Gilliam FG, Kanner AM, Pellock JM. Cognitive and behavioral effects of antiepileptic drugs. Epilepsy Behav 2001;2:SS1–SS17. [9] Drane DL, Meador KJ. Cognitive and behavioral effects of antiepileptic drugs. Epilepsy Behav 2002;3:49–53. [10] Gates JR. Side effect profiles and behavioral consequences of antiepileptic medications. Epilepsy Behav 2000;1:153–9. [11] Loring DW, Meador KJ. Cognitive and behavioral effects of epilepsy treatment. Epilepsia 2001;42(Suppl. 8):24–32. [12] Ortinski P, Meador KJ. Cognitive side effects of antiepileptic drugs. Epilepsy Behav 2004;5:S60–5. [13] Poling A, Blakely, White W, Picker M. Chronic effects of clonazepam, phenytoin, ethosuximide, and valproic acid on learning in pigeons as assayed by a repeated acquisition procedure. Pharmacol Biochem Behav 1986;24:1583–6. [14] Deacon RMJ. Pharmacological studies of a rat spatial delayed nonmatch-to-sample task as an animal model of dementia. Drug Dev Res 1991;24:67–79. [15] Umka J, Mustafa S, El Beltagy M, Thorpe A, Latif L, Bennett G, et al. Valproic acid reduces spatial working memory and cell proliferation in the hippocampus. Neuroscience 2010;166:15–22. [16] Continella G, Drago F, Amico-Roxas M, Scapagnini U. Effects of valproic acid derivatives on the acquisition of active avoidance behavior in the rat. Pharmacol Res Commun 1984;16:699–706. [17] Murphy KJ, Fox GB, Foley AG, Gallagher HC, O'Connell A, Griffin AM, et al. Pentyl-4yn-valproic acid enhances both spatial and avoidance learning and attenuates agerelated NCAM-mediated neuroplastic decline within the rat medial temporal lobe. J Neurochem 2001;78:704–14. [18] Banks MK, Mohr NL, Besheer J, Steinmetz JE, Garraghty PE. The effects of phenytoin on instrumental appetitive-to-aversive transfer in rats. Pharmacol Biochem Behav 1999;63:465–72. [19] Banks MK, Besheer J, Szypczak J, Goodpaster LL, Phipps EJ, Garraghty PE. The effects of carbamazepine on an appetitive-to-aversive transfer task: comparison to untreated and phenytoin. Prog Neuropsychopharmacol Biol Psychiatry 2001;25:551–72. [20] Orczyk JJ, Banks MK, Garraghty PE. The effects of valproic acid on appetitive and aversive instrumental learning in adult rats. Front Behav Neurosci 2014;8:1–9 Available from: http://journal.frontiersin.org/article/https://doi.org/10.3389/fnbeh. 2014.00113/full. [21] Steinmetz JE, Logue SF, Miller DP. Using signaled bar-pressing tasks to study the neural substrates of appetitive and aversive learning in rats: behavioral manipulations and cerebellar lesions. Behav Neurosci 1993;107:941–54. [22] Logue SF. An examination of the role of the hippocampus in the acquisition and timing of a signaled barpress in appetitive and aversive conditioning [dissertation]. Bloomington (IN): Indiana University; 1994. [23] McDowell AL, Samuelson DL, Dina BS, Garraghty PE. Discrete and contextual cue alterations eliminate the instrumental appetitive-to-aversive transfer impairment in phenytoin-treated rats. Integr Physiol Behav Sci 2004;39:307–17. [24] Mowery TM, McDowell AL, Garraghty PE. Chronic developmental exposure to phenytoin has long-term behavioral consequences. Int J Dev Neurosci 2008;26:401–7. [25] Butt AE, Schultz JA, Arnold LL, Garman EE, George CL, Garraghty PE. Lesions of the rat nucleus basalis magnocellularis disrupt appetitive-to-aversive transfer learning. Integr Physiol Behav Sci 2003;38:253–71. [26] Goodman MR, Garman, Arnold LL, Sengelaub DR, Garraghty PE. The effects of estradiol on avoidance learning in ovariectomized adult rats. Integr Physiol Behav Sci 2004;39:192–206. [27] McDowell AL, Heath KM, Garraghty PE. The effects of sex and chronic restraint on instrumental learning in rats. Neurosci J 2013:893126 (12 pages). Available from: https://doi.org/10.1155/2013/893126. [28] McDowell AL, Fransen KMH, Elliot KS, Elghouche A, Kostylev PV, O'Dea PK, et al. Sex differences and the impact of chronic stress and recovery from on instrumental
[29]
[30]
[31] [32]
[33]
[34]
[35] [36] [37] [38] [39] [40] [41]
[42] [43]
[44]
[45]
[46]
[47] [48]
[49]
[50]
[51] [52] [53]
[54] [55] [56] [57]
19
learning. Neurosci J 2015:697659 (12 pages). Available from: https://doi.org/10. 1155/2015/697659. Bialer M, Johannessen SI, Kupferberg HJ, Levy RH, Perucca E, Tomson T. Progress report on new antiepileptic drugs: a summary of the Eighth Eilat Conference (EILAT VIII). Epilepsy Res 2007;73:1–52. Jeavons PM, Clark JE, Maheshwari MC. Treatment of generalized epilepsies of childhood and adolescence with sodium valproate (‘epilim’). Dev Med Child Neurol 1977;19:9–25. Bourgeois BFD. New anti-epileptic drugs. Arch Neurol 1998;55:1181–3. Churchill JD, Fang PC, Voss SE, Besheer J, Herron AL, Garraghty PE. Some antiepileptic compounds impair learning by rats in a Morris water maze. Integr Physiol Behav Sci 2003;38:91–103. Churchill JD, Voss SE, Miller DP, Steinmetz JE, Garraghty PE. Phenytoin blocks the reversal of a classically conditioned discriminative eyeblink response in rabbits. Epilepsia 1998;39:584–9. Samuelson DL, Arnold LL, Mowery TM, Mesnard A, Garraghty PE. The effects of phenytoin on the performance of rats in a delayed match-to-place task. Integr Physiol Behav Sci 2005;40:17–27. Sofia RD. Felbamate: mechanisms of action. In: Levy RH, Mattson RH, Meldrum BS, editors. Antiepileptic drugs. 4th ed. New York: Raven Press; 1999. p. 791–7. Kaufman D, Kelly JP, Anderson T, Harmon DC, Shapiro S. Evaluation of case reports of aplastic anemia among patients treated with felbamate. Epilepsia 1977;38:1265–9. Pellock JM. Felbamate. Epilepsia 1999;40(s5):s57–62. Bruni J. Phenytoin: toxicity. In: Levy RH, Mattson RH, Meldrum BS, editors. Antiepileptic drugs. 4th ed. New York: Raven Press; 1995. p. 345–50. Dreifuss FE. Valproic acid: toxicity. In: Levy RH, Mattson RH, Meldrum BS, editors. Antiepileptic drugs. 4th ed. New York: Raven Press; 1995. p. 641–8. Holmes GL. Carbamazepine: toxicity. In: Levy RH, Mattson RH, Meldrum BS, editors. Antiepileptic drugs. 4th ed. New York: Raven Press; 1995. p. 567–79. McNamara JO. Drugs effective in the therapy of the epilepsies. In: Hardman JG, Limbird LE, Gilman AG, editors. Goodman and Gilman's the pharmalogical basis of therapeutics. 10th ed. New York: McGraw-Hill; 2001. p. 521–47. Smith RD, Grzelak ME, Coffin VL. Felbamate, a novel antiepileptic agent, does not affect cognition in rodents. Behav Pharmacol 1994;5:365–8. Ketter TA, Malow BA, Flamini R, Ko D, White SR, Post RM, et al. Felbamate monotherapy has stimulant-like effects in patients with epilepsy. Epilepsy Res 1996;23: 129–37. LeDoux JE, Moscarello J, Sears R, Campese V. The birth, death and resurrection of avoidance: a reconceptualization of a troubled paradigm. Mol Psychiatry 2017;22: 24–36. LeDoux JE, Iwata J, Cicchetti P, Reis DJ. Different projections of the central amydaloid nucleus mediate autonomic and behavioral correlates of conditioned fear. J Neurosci 1988;8:2517–29. Kim JJ, Rison RA, Fanselow MS. Effects of amygdala, hippocampus, and periaquductal gray lesions on short- and long-term contextual fear. Behav Neurosci 1993;107: 1093–8. Rogan MT, Staubli UV, LeDoux JE. Fear conditioning induces associative long-term potentiation in the amygdala. Nature 1997;390:604–7. Quirk GJ, Armony JL, LeDoux JE. Fear conditioning enhances different temporal components of tone-evoked spike trains in auditory cortex and lateral amygdala. Neuron 1997;19:613–24. Maren S. Auditory fear conditioning increases CS-elicited spike firing in lateral amygdala neurons even after extensive overtraining. Eur J Neurosci 2000;12: 4047–54. Repa JC, Muller J, Apergis J, Desrochers TM, Zhou Y, LeDoux JE. Two different lateral amygdala cell populations contribute to theinitiation and storage of memory. Nat Neurosci 2001;4:724–31. Ramirez F, Moscarello JM, LeDoux JE, Sears RM. Active avoidance requires a serial basal amygdala to nucleus accumbens shell circuit. J Neurosci 2015;25:3470–7. Moscovitch M. Multiple dissociations of function in amnesia. In: Cermak LS, editor. Human memory and amnesia. Hillsdale, NJ: Erlbaum; 1982. p. 337–70. Jiao X, Beck KD, Myers CE, Servatius RJ, Pang KC. Altered activity of the medial prefrontal cortex and amygdala during acquisition and extinction of an active avoidance task. Front Behav Neurosci 2015;9:00249. https://doi.org/10.3389/fnbeh2015. 00249. Cilio MR, Thiele EA, Devinsky O. The case for assessing cannabidiol in epilepsy. Epilepsia 2014;55:787–90. Szaflarsk J, Devinsky O, editors. Cannabinoids and epilepsy, Epilepsy Behav 2017; 70(Part B):277–370. Schmitz B. Effects of antiepileptic drugs on mood and behavior. Epilepsia 2006; 47(Suppl. 2):28–33. Janelsins MC, Kohli S, Mohile SG, Usuki K, Ahles TA, Morrow GR. An update on cancer- and chemotherapy-related cognitive dysfunction: current status. Semin Oncol 2011;38:431–8.