- Email: [email protected]
Anxiolytic-like effects of leptin on fixed interval responding
Pharmacology, Biochemistry and Behavior 148 (2016) 15–20
Contents lists available at ScienceDirect
Pharmacology, Biochemistry and Behavior journal homepage: www.elsevier.com/locate/pharmbiochembeh
Anxiolytic-like effects of leptin on fixed interval responding Susan M. Tyree 1, Robert G.K. Munn 2, Neil McNaughton ⁎ Department of Psychology, University of Otago, Dunedin, New Zealand
a r t i c l e
i n f o
Article history: Received 23 February 2016 Received in revised form 9 May 2016 Accepted 10 May 2016 Available online 11 May 2016 Keywords: Leptin Anxiety Anxiolytic Fixed interval Behavioural inhibition 5-HT
a b s t r a c t Leptin has been shown to affect energy homeostasis, learning and memory, and some models of anxiolytic action. However, leptin has produced inconsistent results in previous non-operant behavioural tests of anxiety. Here, we test the anxiolytic potential of leptin in an operant paradigm that has produced positive results across all classes of anxiolytic so far tested. Rats were tested in the Fixed Interval 60 Seconds (FI60) task following administration of 0/0.5/1.0 mg/kg (i.p.) leptin or an active anxiolytic control of 5 mg/kg (i.p.) chlordiazepoxide (CDP). By the end of the 14 days of testing in the FI60 task, 0.5 mg/kg leptin released suppressed responding in a manner similar to CDP, and 1.0 mg/kg leptin produced a relative depression in responding, a similar outcome pattern to previously tested 5HT-agonist anxiolytics. This suggests that leptin behaves similarly to established serotonergic anxiolytics such as buspirone and fluoxetine; with the delay in development of effect during testing, and the inverted-U dose–response curve explaining the inconsistent behaviour of leptin in behavioural tests of anxiety, as this type of pattern is common to serotonergic anxiolytics. © 2016 Elsevier Inc. All rights reserved.
1. Introduction Leptin is released primarily from adipocytes in response to the ingestion of dietary fat (Havel, 2000). It has long been thought to be a feedback signal involved in satiety, but has since been shown to have diverse roles unrelated to energy regulation (Margetic et al., 2002). One of these novel roles for leptin is to act centrally on some aspects of cognition and anxiety — functions typically thought to be controlled by the hippocampus (Asakawa et al., 2003; Farr et al., 2006; Gisou et al., 2009; Paulus et al., 2005). There is substantial behavioural evidence linking endogenous leptin to anxiety. Genetically modified ob/ob mice, which fail to produce leptin, display more anxious behaviour (Finger et al., 2010) and the administration of exogenous leptin reverses this difference (Asakawa et al., 2003). They also, like animals treated with anxiolytic drugs (McNaughton and Morris, 1987, 1992), show impaired spatial memory (Farr et al., 2006), implicating structures involved in the regulation of anxiety, such as the hippocampus, in the action of leptin. Conversely, maternal hyperleptinaemia, via injection of leptin while pups are in utero, produces adult rats that display improved spatial memory and reduced anxious behaviour compared to animals whose dams were injected with vehicle (Fraga-Marques et al., 2010). ⁎ Corresponding author at: Department of Psychology, University of Otago, P.O. Box 56, Dunedin 9054, New Zealand. E-mail addresses: [email protected] (S.M. Tyree), [email protected] (R.G.K. Munn), [email protected] (N. McNaughton). 1 Present address: Department of Molecular Genetics, German Institute of Human Nutrition, Potsdam-Rehbrücke, Germany. 2 Present address: Department of Neurobiology, Stanford University, Stanford, CA, USA.
http://dx.doi.org/10.1016/j.pbb.2016.05.005 0091-3057/© 2016 Elsevier Inc. All rights reserved.
There is also evidence for interaction of exogenous leptin with the control of anxiety. The Hypothalamic–Pituitary–Adrenocortical (HPA) axis is a central structure in anxiety (Faravelli et al., 2012; Landgraf et al., 1999). Leptin has been shown to diminish hyperactivity of the HPA axis (Holmes, 2015; Perry et al., 2014) and activation of neurons expressing the leptin receptor (LepRb) in the lateral hypothalamic area results in the inhibition of HPA axis activity (Bonnavion et al., 2015). Leptin is so central to this system that it has been included in the HPA axis as the Hypothalamic–Pituitary–Adrenal–Leptin axis (Aschbacher et al., 2014). It is also possible that leptinergic modulation of this axis is through 5-HT (Guo and Lu, 2014; Haleem et al., 2015; Kurhe et al., 2015). Leptin can also behave like established anxiolytics in some behavioural tests of anxiety (Liu et al., 2010; Wang et al., 2015). The picture in relation to exogenous anxiolytic action is not homogeneous, however. Several experiments have produced non-positive results. Suomalainen and Mannisto (1998) failed to obtain changes in anxiety-related behaviour in the elevated plus maze or open field. Buyse et al. (2001) found that leptin significantly reduced the number of open-arm entries in the elevated plus maze, as did Thorsell et al. (2002). In contrast to these anxiogenic-like findings in the elevated plus maze, Liu et al. (2010) showed that leptin administration produced anxiolytic-like effects in the elevated plus maze, but not in the open field test. Both the elevated plus maze and open field test are standard behavioural tests of anxiety. Hogg (1996) points out that the variability in response to the elevated plus maze, particularly in response to serotonergic agonists, is highly contingent on a variety of factors such as the aversiveness of the testing environment and the prior handling experience of the animals in the tests. There was also considerable variation in the dose of leptin used in these tests; Suomalainen and
16
S.M. Tyree et al. / Pharmacology, Biochemistry and Behavior 148 (2016) 15–20
Mannisto (1998) used 10 and 20 mg/kg, Buyse et al. (2001) 0.4–1 mg/ kg, and Liu et al. (2010) 0.25 and 1 mg/kg. In contrast to the i.p. administration in the other studies, Thorsell et al. (2002) delivered 0.1 mg/kg intraventricularly. While thought to bind primarily to the LEPR-encoded leptin receptor family there is evidence that leptin modulates, and is modulated by, the activity of 5-HT (Garcia-Alcocer et al., 2010; Morrison, 2004), perhaps in an indirect manner. 5-HT is a key neurotransmitter system that is involved in the anxiolytic function of serotonergic anxiolytics such as buspirone (Riblet et al., 1982) and fluoxetine (Fuller et al., 1991). Administration of leptin increases brain serotonin (5-HT) metabolism in mice (Calapai et al., 1999). There is also evidence that administration of the 5-HT precursor 5-hydroxytryptophan significantly increases serum leptin levels (Yamada et al., 1999); and serotonergic neurons have been shown to be targets for leptin in the monkey (Finn et al., 2001), strongly implying a serotonergic modulatory role for leptin in anxiety. Other lines of evidence indirectly link 5-HT to leptin. The 5-HT transporter, which is targeted by Serotonin-Specific Reuptake Inhibitors, is linked to obesity (Üçeyler et al., 2010). There is also evidence that polymorphisms of the 5-HT receptor are linked to anorexia and bulimia (Collier et al., 1997; Nacmias et al., 1999). This suggests that disorders of eating behaviour which are modulated by leptin, amongst other things, are linked to the 5-HT system. With increasing evidence of a connection between leptin and the 5HT system, as well as studies supporting its anxiolytic potential, and the established role for 5-HT agonists as anxiolytics, the inconsistent performance of leptin in behavioural assays of anxiety is perplexing. The open field and elevated plus maze tests used to assess leptin also produce inconsistent results when testing well-established clinical anxiolytics, not just with regard to failure to detect anxiolytic potential in known anxiolytics, but with some non-anxiolytic compounds being incorrectly detected as anxiolytic (for review, see (Crawley, 1985; Lister, 1990). In contrast to these spatial exploratory paradigms, which can produce mixed results, the Fixed Interval 60 Seconds (FI60) is an operant behavioural task that involves non-spatial behavioural inhibition. It assesses response suppression in a similar manner to the Vogel conditioned suppression of licking, Geller-Seifter, and Conditioned Emotional Response tests, but has the advantage of suppressing responding with the aversive state of frustration and not using shock and so not involving a confound with any analgesic action a drug may have. Fear and frustration have similar eliciting properties and their suppression of responding is similarly sensitive to anxiolytic drugs (Gray, 1977). Control animals in the FI60 paradigm rapidly learn to inhibit the bulk of their responding until after the 60-second non-reward period has elapsed. However, following administration of an anxiolytic drug animals are typically less able to inhibit their behaviour, making a larger number of inappropriately early responses during the non-reward period. This is true for all anxiolytics so far tested, regardless of precise pharmacological mode of action (Munn and McNaughton, 2008; Panickar and McNaughton, 1991; Zhu and McNaughton, 1995). One further advantage of the FI60 paradigm is that it provides an indirect assay of pharmacological mode of action. GABA-ergic anxiolytics such as the barbiturates and benzodiazepines show a linear dose–response relationship in the FI60; larger doses typically produce a larger release of responding. In contrast, the serotonergic anxiolytics so far tested tend to show an inverted-U dose–response curve; low doses produce a release of responding, while higher doses suppress responding (Munn and McNaughton, 2008; Panickar and McNaughton, 1991). It should be noted that while the form of the dose–response curve is different for the different classes of drugs, the low dose effects all involve a similar release of inhibition in the initial period of responding and the differences between the classes disappear with long-term administration (Zhu and McNaughton, 1995). The aim of the present study is therefore twofold. First, given the inconsistent results produced by standard tests of anxiety, we first aim to
examine the anxiolytic potential of leptin in the more reliable FI60. Secondly, should leptin behave as an anxiolytic in the FI60, we predict that it will show a dose response curve consistent with 5HT (i.e. response depression at higher doses) rather than GABA agency. 2. Methods 2.1. Subjects The subjects were 24 naïve male Sprague–Dawley rats, obtained from the University of Otago Department of Laboratory Animal Sciences. They weighed between 243 and 323 g immediately prior to the experiments. The home room was maintained on a 12 h light, 12 h dark cycle; (lights on at 0600 and off at 1800). The temperature of the home room was maintained at 20–22 °C. Animals in all cages had access to water and Reliance stock food pellets ad libitum. The animals were initially housed in 49 cm × 31 cm × 26.5 cm cages, in groups of four. Twelve days before the pre-training sessions began the animals were taken off ad libitum access to food and placed on a restricted diet of Reliance Stock food pellets. Animals were weighed daily, and the amount of food they received each day was varied in order to maintain their weight at 80% of their free feeding weight. Animals had ad libitum access to water throughout the experiment. 2.2. Apparatus Six standard operant chambers (Campden Instruments, U.K.) were used throughout the training sessions as well as all the experimental trials. The dimensions of the operant chambers were 57.5 cm × 34.5 cm × 39 cm. The interior of the operant chambers contained a smaller interior chamber which was 24 cm × 24 cm × 26 cm. A food delivery system distributed individual 45 mg Dustless Precision Pellets (Campden Instruments Ltd., UK) to a food hopper in the interior chamber through a plastic tube. The front-facing wall of the operant chamber was hinged from the bottom in order to allow access to the interior. One of the long exterior walls of the chamber had a circular tinted window 16.6 cm in diameter, which allowed a view of the interior of the chamber. The interior chamber of the operant chamber had three metal walls, a metal ceiling, and a horizontal grid floor. The front-facing wall of the interior chamber was a transparent Perspex wall that could be unlatched from the top of the box and pulled down to allow access to the chamber. One of the walls had a recessed 5 cm by 6.5 cm food hopper with a Perspex door hinged from the top, the food hopper was set in the middle and at the bottom of the wall. Two retractable metal levers were set into the wall, one on either side of, and equidistant from, the hopper. Only the left hand lever was extended into the chamber in the current experiment. Directly above each lever there were two 2.8 W stimulus lights set into the wall. The other two metal walls of the housing chamber were bare. In the centre of the metal ceiling of the housing chamber there was a 2.8 W house light, which was on throughout the training and experimental trials. There was also an electric fan, which provided ventilation for the animals, and also produced a constant level of background noise. Three IBM compatible computers, which were running Visual Basic 6 software with LABJACK control components, each controlled two operant chambers. Custom Visual Basic 6 programs were used to deliver the three training schedules. This meant that the computers controlled the timing and delivery of the reinforcements, as well as recording the number of lever presses the animal made, and the number of nose pokes into the food hopper. The computers also recorded the time each lever press was made during the FI60 task. These responses were divided into one of twelve five-second bins, according to the interval since the last reward that they were made. Responses made between 0 and 5 s since the last reward were allocated to bin one; responses made between 5 and 10 s since the last reward were allocated to bin two and so on.
S.M. Tyree et al. / Pharmacology, Biochemistry and Behavior 148 (2016) 15–20
2.3. Procedure 2.3.1. Autoshaping Each of the animals participating in the FI60 behavioural experiment received preliminary training to ensure that the animals would reliably insert their head into the hopper in order to retrieve a food pellet. Each animal was run on a random time (i.e. not response contingent) 30 s (RT30) schedule once a day. Each of these RT30 training sessions lasted for 30 min and each of the animals were run once a day at the same time each day. In order to encourage the animals to investigate the food hopper and put its head inside, a small pellet of the animal's regular food (Reliance Stock Pellets) was placed in the hopper so that it wedged open the Perspex hopper door. Once each animal was reliably responding to the light and sound cues of reinforcement by putting its head in the hopper, they were moved onto the next stage of training. The animals all required five sessions of RT30 training before they were reliably responding to reinforcement cues. 2.3.2. Continuous reinforcement Having completed the RT30 training schedule, all animals started on a continuous reinforcement (CRF) schedule. Each CRF training schedule lasted for 30 min and each of the animals ran once a day, at the same time each day, just as described in the RT30 schedule. Once the animals were consistently pressing the lever and receiving a food pellet, another two drug-free CRF sessions were run. Four training days were required to establish reliable lever pressing responses in the animals. Prior to the final CRF training session, animals were assigned to the four separate drug groups based on their CRF responses to ensure that the average CRF response rates were consistent across the four groups. The mean CRF response rates of the resultant groups differed by not N 5% overall. The animals were then rehoused to ensure that there was not more than one animal from each drug group in any one cage. This rehousing was carried out in order to remain consistent with previous experiments carried out in the current lab (Munn and McNaughton, 2008). Note that all drug groups will have been similarly affected by rehousing. Animals then ran a final CRF training session after having been administered the drug corresponding to the group to which they were assigned. Drug administration was started on the last day of CRF to prevent drug state acting as a signal of schedule change. The animals began their Fixed Interval training the day following this final CRF training session. 2.3.3. Drug preparation and administration The animals were divided into four groups: one passive control group, which was given the phosphate buffered saline (PBS) vehicle solution; one active control group, which was given a previously confirmed positive anxiolytic dose (5 mg/kg) of the benzodiazepine chlordiazepoxide (CDP); and two leptin groups, which were given doses of 0.5 mg/kg and 1.0 mg/kg, respectively. These doses of leptin were chosen in order to facilitate comparison to the doses used in those studies that show an anxiolytic effect of leptin (Haleem et al., 2015; Liu et al., 2010; Wang et al., 2015). All PBS used was pH adjusted to 8.0. Recombinant mouse leptin was supplied by NIDDK's National Hormone and Peptide Program and A.F. Parlow. Chlordiazepoxide Hydrochloride (CDP; Sigma Chemical Co., Louis, USA) was dissolved in 0.9% saline, for a concentration of 5 mg/ml. The drug solutions were prepared freshly on the day of injection and were administered via intraperitoneal (i.p.) injection using a 1 ml syringe and a 26 g needle at a volume of 1 ml/kg of body weight. Animals in the leptin drug groups were administered their injections 35 min before they were scheduled to begin their FI60 training session. This was to ensure that the maximum effect of leptin would be occurring during their testing session, specifically to peak half way through the thirty-minute FI60 session. The time interval was chosen based on the results of previous experiments in the laboratory (data not shown), which suggest that 50 min is the time of maximum drug effect on hippocampal rhythmic slow activity caused by 1.0 mg/kg leptin.
17
Animals in the 5 mg/kg CDP drug group were administered CDP 10 min before they were scheduled to begin their FI60 training session to ensure that the maximum effect of CDP would be occurring during the session, this time interval was based on previous data from electrophysiological and FI60 experiments using CDP (Woodnorth and McNaughton, 2002). A specific time-matched control group was not used to limit the use of animals and because only qualitative rather than quantitative positive control results were required. Animals therefore received a drug administration for one day of CRF training and each day of FI60 testing, for a total of 15 administrations of drug. 2.3.4. Fixed interval Having completed the CRF training, the animals began the FI60 reinforcement schedule. Animals were run in the FI60 task once each day, at the same time each day (between 1200 and 1600). The FI60 training sessions ran every day for 14 consecutive days. Each FI60 session lasted 30 min. 2.4. Data collection and analysis For the purposes of analysis and illustration, the twelve recorded 5 s bins were compressed into six 10 s bins so that the responses from bins one and two were now placed into bin one, responses in bins three and four were now in bin two, etc. Raw data were transformed by a logarithmic [log10(X + 1.0)] function in order to normalise the error variance (Zar, 2009). The log transformed raw data were subjected to Analysis of Variance (ANOVA) using the Genstat package (VSN international). Within subject factors such as responding across bins and days were extracted, and the presence of linear, quadratic, and cubic polynomial components in the data were analysed to investigate the effect of each drug on the response rates in the beginning, middle, and end of the FI60 schedule as well as over experimental days. The data were first subjected to an analysis including all groups to assess omnibus effects overall. This was followed by post-hoc ANOVAs and t-tests to determine the source of the omnibus effects. 3. Results Fig. 1 shows the average number of bar press responses per bin for each drug group at the beginning, middle, and end of the 15 days of testing. At times immediately after reinforcement, animals displayed low responding. Responses increased as time to reinforcement approached (bins[lin], F(1,3300) = 9278.22, p b 0.001) reaching a maximum in close proximity to reinforcement. Averaged over days, responding tended to be suppressed in the initial bins, rise during intermediate bins, and then flatten off in the final bins (bins[cub], F(1, 3300) = 318.39. p b 0.001). Animals showed a progressive decrease in responding in bin 1 and an increase in responding in bin 6 as they received more days of training with the response curve showing a mild positive inflection (inverted U-shape assessed by the quadratic component of bins) initially that changed to a more negative (U-shape) as training progressed (compare Fig. 1A and C; days[lin] × bins[quad], F(1,3300) = 4.99. p = 0.002). There were also differences across the different drug groups. The difference in overall response rate (averaged across bins) between the different drug groups increased steadily over days of testing (drug × days[lin], F(3, 280) = 3.73, p = 0.012; Fig. 2) but with an apparent decrease in the differences as training reached asymptote that did not achieve significance (drug × days[quad], F(3, 280) = 2.15, p = 0.094). This drug difference was clearest, overall, in bins 2–4 (drug × bins [lin] F(3, 3300) = 4.50, p = 0.004; [quad] F(3, 3300) = 46.36, p b 0.001; [cub] F(3, 3300) = 12.38, p b 0.001) with the differences generally increasing over days (drug × days[lin] × bins [quad] F(3, 3300) = 4.99, p = 0.002). The response curve for the CDP group showed an inverted
18
S.M. Tyree et al. / Pharmacology, Biochemistry and Behavior 148 (2016) 15–20
Fig. 2. The mean number of responses made by the 0 mg/kg Leptin (open circles), 0.5 mg/ kg leptin (grey circles), and 1 mg/kg leptin (filled black circles) groups (n = 6 in each group) in each day of FI60 testing. All three groups begin testing with similar rates of response, but rapidly diverge such that the 0.5 mg/kg leptin group makes consistently more responses than control, and the 1.0 mg/kg leptin group consistently makes fewer responses. The fitted grey lines illustrate the significant linear and marginal quadratic components of the drug × days interaction found in the main analysis (0 mg/kg; solid line, 0.5 mg/kg; dotted and dashed line, 1.0 mg/kg; dashed line). The non-linear y-axis is a result of the logarithmic transform used to normalise the data.
F(1, 1650) = 5.01, p = 0.025) and more so for 1 mg/kg (leptin[1.0 versus 0.0] × bins[quad), F(1, 1650) = 53.32, p b 0.001). As can be seen in Fig. 1C, by the end of training, the behaviour of the 0.5 mg/kg group was very similar to that of the CDP group, with the greatest release of responding, relative to control, being in bin 2 when the controls show the greatest response suppression. The 1 mg/kg leptin group tended to make progressively fewer responses than the 0 mg/kg control group over days until towards the end of testing, when the responses in the 1 mg/kg group increased until the groups were approximately equal (leptin[1.0 versus 0.0] × days[quad], F(1140) = 4.76, p b 0.05; Fig. 2); but with a difference limited to intermediate bins (3 and 4) becoming most obvious at the end of training (leptin[1.0 versus 0.0] × days[lin] × bins[quad), F(1, 1650) = 6.99, p = 0.008; Fig. 1C). 4. Discussion
Fig. 1. The average number of responses per bin of the FI60 schedule by each of the 0 mg/ kg leptin, 0.5 mg/kg leptin, 1.0 mg/kg leptin, and 5.0 mg/kg CDP groups. Each bin-pair contains the average data from all six animals in each drug group averaged across A) Days 1–4, B) days 6–9, and C) days 11–15. The non-linear response axis is the result of the logarithmic transform used to normalise the data. Over days 1–4 responses in the saline and both leptin groups are equivalent over all bins, while the CDP group shows a general release of responding. Over days 6–9 the saline and 0.5 mg/kg leptin groups are equivalent over all bins, but there is a general depression of responding in the 1.0 mg/kg leptin group and a continued release of responding in the CDP group. By days 11–15 the responses of both the CDP and 0.5 mg/kg groups show an equivalent release of responding in comparison to saline control, while the responses in the 1.0 mg/kg leptin group remain depressed relative to saline control. The error bars are 2 × S.E.M. The nonlinear y-axis is a result of the logarithmic transform used to normalise the data.
U-shape, and that of the 1.0 mg/kg leptin group a U-shape, to the greatest extent in the middle of training compared to the beginning and the end of training (Fig. 1B; drug × days[quad] × bins[quad], F(1,3300) = 4.75. p = 0.003; drug × days[quad] × bins[lin], F(1,3300) = 2.74. p b 0.05). The significant drug effects in the overall analysis were explored with separate post-hoc comparisons of each leptin dose against 0 mg/kg control. The overall effect of drug on the intermediate bins was significant for the 0.5 mg/kg dose (leptin[0.5 versus 0.0] × bins[quad),
As expected, the active control drug CDP at a dose of 5.0 mg/kg caused a release of responding throughout training compared to control, most obviously in the early bins, as in previous experiments with this drug. This result is typical of a GABA agonist anxiolytic agent. In a similar manner, low dose (0.5 mg/kg) leptin released responding in the early bins of the FI60 schedule, particularly towards the end of testing; by the final days of FI60 testing low dose leptin and CDP were producing a comparable release of responding. This finding adds support to the notion that leptin may be clinically anxiolytic, at least at lower doses. Our high dose (1.0 mg/kg) of leptin produced a robust depression of responding in comparison with vehicle control. This effect is strikingly similar to that typically observed when a high dose of serotonergic agents such as buspirone or fluoxetine are tested; this effect is consistent in animals as diverse as monkeys, pigeons, and rats (Leander and Carter, 1984; McKearney, 1982; McNaughton et al., 1996; Munn and McNaughton, 2008; Panickar and McNaughton, 1991). The dose–differential effect of leptin on responding in the FI60 task is consistent with the growing body of evidence that leptin acts, at least for its anxiolytic action, through serotonergic modulation. Leptin typically produces dose-dependent reductions in food intake starting from doses of 0.5 mg/kg (Widdowson et al., 1997). In contrast, we observe an increase of responding compared to control at this dose, suggesting
S.M. Tyree et al. / Pharmacology, Biochemistry and Behavior 148 (2016) 15–20
that the satiety-inducing effect of leptin is not a factor in these experiments. Anxiolytics often increase food intake but effects on appetite would be expected to affect terminal rate responding (when food is anticipated) not initial rate responding (when behavioural inhibition is strongest). A similar argument rules out changes in impulsivity since, in the FI schedule, rat models of ADHD show increased terminal but not initial responding (Wickens et al., 2004). The FI60 has been shown to produce characteristic changes in behaviour depending on whether an anxiolytic functions via GABA-ergic or serotonergic action (Munn and McNaughton, 2008; Panickar and McNaughton, 1991). This difference in behaviour caused by these two families of drugs is believed to be mediated by their effect on the stress hormone corticosterone. The 5-HT1A agonist buspirone, unlike benzodiazepines such as CDP, does not block stress-induced corticosterone release (Urban et al., 1986), and in fact, at high doses (over 1 mg/kg), triggers the release of corticosterone without the presence of an external stressor (De Boer et al., 1990; Matheson et al., 1988). This effect shows tolerance following long-term treatment with high doses of buspirone (De Boer et al., 1990) or fluoxetine (Serra et al., 2001). In the FI60, long-term administration of serotonergics or pharmacologic blockade of corticosterone release render the action of serotonergics similar to that of the GABA-agonists (McNaughton et al., 1996; Zhu and McNaughton, 1995), suggesting that acute serotonin-driven corticosterone release, which can show tolerance, is likely the cause of the U-shaped dose–response relationship observed when testing these drugs. Interestingly, leptin has also previously been shown to modulate corticosterone levels: optogenetic activation of LepRb neurons reduces corticosterone levels in stressed mice (Bonnavion et al., 2015) and conversely intracerebroventricular administration of leptin resulted in increased basal levels of corticosterone in rats (Thorsell et al., 2002). Like the clinical response, the effects of serotonergic and GABA drugs are, as noted above, similar on the FI when they have been previously administered 3 times per day for 60 days (Zhu and McNaughton, 1995). However, neither the anxiolytic nor depressant effects of buspirone on the FI shows much change across days of testing when administered as in the current experiment (Panickar and McNaughton, 1991). The slow-development of the leptin effects in the current study may, then, partially clarify the inconsistent results seen in previous behavioural screenings of leptin for anxiolytic potential. The doses and time scales used in these experiments are very heterogeneous. For example, studies testing the anxiolytic potential of leptin in the elevated plus maze showed variations either in leptin dosage; 0/0.25/0.5 mg/kg (Liu et al., 2010), 0.4/1.0 mg/kg (Buyse et al., 2001), or 10/20 mg/kg (Suomalainen and Mannisto, 1998), or delay between drug administration and testing; 30 min (Liu et al., 2010), 1 h (Buyse et al., 2001), or 10– 20 min (Suomalainen and Mannisto, 1998). It is important to note that the positive anxiolytic-typical findings in these tasks occur at doses and times similar to those that are used in the present experiments that positively identify leptin as anxiolytic. The results from the present study show that high dose leptin produces a qualitatively similar depression in responding on the FI60 task to high doses of buspirone or fluoxetine. This finding, in addition to previous evidence showing leptin has been implicated in modulation of the serotonergic system (Calapai et al., 1999; Charnay et al., 2000; Yamada et al., 1999), and evidence that leptin may also release, or enhance the release of, corticosterone (Thorsell et al., 2002), suggests that leptin may act, directly or indirectly, as an agonist of serotonin, potentially at 1A receptors, as part of its anxiolytic mode of action. Author disclosures There were no external funds used in the production of this study. N.M. has a confidential disclosure and consulting agreement with Janssen Research & Development, LLC. R.M. and N.M. designed the study and wrote the protocol. S.T. ran the experimental procedures. S.T., R.M. and N.M. undertook the statistical analysis, and S.T. wrote
19
the first draft of the manuscript. All authors contributed to- and have approved the final manuscript. Acknowledgements Leptin was obtained from the NIH National Institute of Diabetes and Digestive and Kidney diseases (NIDDK) National Hormone and Peptide Program and A.F. Parlow. R. Munn was supported by a postdoctoral fellowship from the Neurological Foundation of New Zealand.
References Asakawa, A., Inui, A., Inui, T., Katsuura, G., Fujino, M.A., Kasuga, M., 2003. Leptin treatment ameliorates anxiety in ob/ob obese mice. J. Diabetes Complicat. 17 (2), 105–107. Aschbacher, K., Rodriguez-Fernandez, M., van Wietmarschen, H., Tomiyama, A.J., Jain, S., Epel, E., ... van der Greef, J., 2014. The hypothalamic–pituitary–adrenal–leptin axis and metabolic health: a systems approach to resilience, robustness and control. Interface Focus 4 (5). Bonnavion, P., Jackson, A.C., Carter, M.E., de Lecea, L., 2015. Antagonistic interplay between hypocretin and leptin in the lateral hypothalamus regulates stress responses. Nat. Commun. 6. Buyse, M., Bado, A., Daugé, V., 2001. Leptin decreases feeding and exploratory behaviour via interactions with CCK 1 receptors in the rat. Neuropharmacology 40 (6), 818–825. Calapai, G., Corica, F., Corsonello, A., Sautebin, L., Rosa, D., Massimo, C., Giuseppe, M., ... Caputi, A.P., 1999. Leptin increases serotonin turnover by inhibition of brain nitric oxide synthesis. J. Clin. Investig. 104 (7), 975. Charnay, Y., Cusin, I., Vallet, P.G., Muzzin, P., Rohner-Jeanrenaud, F., Bouras, C., 2000. Intracerebroventricular infusion of leptin decreases serotonin transporter binding sites in the frontal cortex of the rat. Neurosci. Lett. 283 (2), 89–92. Collier, D.A., Arranz, M.J., Li, T., Mupita, D., Brown, N., Treasure, J., 1997. Association between 5-HT2A gene promoter polymorphism and anorexia nervosa. Lancet 350 (9075), 412. Crawley, J.N., 1985. Exploratory behavior models of anxiety in mice. Neurosci. Biobehav. Rev. 9 (1), 37–44. De Boer, S.F., Slangen, J.L., Van der Gugten, J., 1990. Effects of chlordiazepoxide and buspirone on plasma catecholamine and corticosterone levels in rats under basal and stress conditions. Endocrinol. Exp. 24 (1–2), 229–239. Faravelli, C., Lo Sauro, C., Lelli, L., Pietrini, F., Lazzeretti, L., Godini, L., ... Ricca, V., 2012. The role of life events and HPA axis in anxiety disorders: a review. Curr. Pharm. Des. 18 (35), 5663–5674. http://dx.doi.org/10.2174/138161212803530907. Farr, S.A., Banks, W.A., Morley, J.E., 2006. Effects of leptin on memory processing. Peptides 27 (6), 1420–1425. http://dx.doi.org/10.1016/j.peptides.2005.10.006. Finger, B.C., Dinan, T.G., Cryan, J.F., 2010. Leptin-deficient mice retain normal appetitive spatial learning yet exhibit marked increases in anxiety-related behaviours. Psychopharmacology 210 (4), 559–568. http://dx.doi.org/10.1007/s00213-010-1858-z. Finn, P.D., Cunningham, M.J., Rickard, D.G., Clifton, D.K., Steiner, R.A., 2001. Serotonergic neurons are targets for leptin in the monkey. J. Clin. Endocrinol. Metab. 86 (1), 422–426. Fraga-Marques, M.C., Moura, E.G., Silva, J.O., Claudio-Neto, S., Pereira-Toste, F., Passos, M.C.F., ... Manhaes, A.C., 2010. Effects of maternal hyperleptinaemia during lactation on short-term memory/learning, anxiety-like and novelty-seeking behavioral traits of adult male rats. Behav. Brain Res. 206 (1), 147–150. http://dx.doi.org/10.1016/j. bbr.2009.08.029. Fuller, R.W., Wong, D.T., Robertson, D.W., 1991. Fluoxetine, a selective inhibitor of serotonin uptake. Med. Res. Rev. 11 (1), 17–34. Garcia-Alcocer, G., Rodriguez, A., Moreno-Layseca, P., Berumen, L.C., Escobar, J., Miledi, R., 2010. Serotonin receptor 5-HT(5A) in rat hippocampus decrease by leptin treatment. Neurosci. Lett. 486 (3), 171–173. Gisou, M., Soheila, R., Nasser, N., 2009. Evaluation of the effect of intrahippocampal injection of leptin on spatial memory. Afr. J. Pharm. Pharmacol 3 (9), 443–448. Gray, J.A., 1977. Drug effects on fear and frustration: possible limbic site of action of minor tranquilizers. In: Iversen, L.L., Iversen, S.D., Snyder, S.H. (Eds.), Handbook of PsychopharmacologyDrugs, Neurotransmitters and Behaviour vol. 8. Plenum Press, New York, pp. 433–529. Guo, M., Lu, X.Y., 2014. Leptin receptor deficiency confers resistance to behavioral effects of fluoxetine and desipramine via separable substrates. Transl. Psychiatry 4 (12), e486. Haleem, D.J., Haque, Z., Inam, Q.U., Ikram, H., Haleem, M.A., 2015. Behavioral, hormonal and central serotonin modulating effects of injected leptin. Peptides http://dx.doi. org/10.1016/j.peptides.2015.10.002. Havel, P.J., 2000. Role of adipose tissue in body-weight regulation: mechanisms regulating leptin production and energy balance. Proc Nutr Soc 59 (3), 359–371. Hogg, S., 1996. A review of the validity and variability of the elevated plus-maze as an animal model of anxiety. Pharmacol. Biochem. Behav. 54 (1), 21–30. http://dx.doi.org/ 10.1016/0091–3057(95)02126–4. Holmes, D., 2015. Neuroendocrinology: leptin attenuates HPA-axis activation and stress responses. Nat. Revs. Endocrinol. 11 (5), 255. http://dx.doi.org/10.1038/nrendo. 2015.34. Kurhe, Y., Mahesh, R., Devadoss, T., 2015. QCM-4, a 5-HT3 receptor antagonist ameliorates plasma HPA axis hyperactivity, leptin resistance and brain oxidative stress in
20
S.M. Tyree et al. / Pharmacology, Biochemistry and Behavior 148 (2016) 15–20
depression and anxiety-like behavior in obese mice. Biochem. Biophys. Res. Commun. 456 (1), 74–79. http://dx.doi.org/10.1016/j.bbrc.2014.11.036. Landgraf, R., Wigger, A., Holsboer, F., Neumann, I.D., 1999. Hyper-reactive hypothalamo– pituitary–adrenocortical axis in rats bred for high anxiety-related behaviour. J. Neuroendocrinol. 11 (6), 405–407. Leander, J.D., Carter, R.B., 1984. Effects of norepinephrine and serotonin uptake inhibitors on the schedule-controlled behavior of the pigeon. Pharmacol. Biochem. Behav. 20 (3), 391–395. http://dx.doi.org/10.1016/0091-3057(84)90277–6. Lister, R.G., 1990. Ethologically-based animal models of anxiety disorders. Pharmacol. Ther. 46 (3), 321–340. Liu, J., Garza, J.C., Bronner, J., Kim, C.S., Zhang, W., Lu, X.-Y., 2010. Acute administration of leptin produces anxiolytic-like effects: a comparison with fluoxetine. Psychopharmacology 207, 535–545. http://dx.doi.org/10.1007/s00213-009-1684-3. Margetic, S., Gazzola, C., Pegg, G.G., Hill, R.A., 2002. Leptin: a review of its peripheral actions and interactions. Int. J. Obes. Relat. Metab. Disord. 26 (11), 1407–1433. http:// dx.doi.org/10.1038/sj.ijo.0802142. Matheson, G.K., Gage, D., White, G., Dixon, V., Gipson, D., 1988. A comparison of the effects of buspirone and diazepam on plasma corticosterone levels in rat. Neuropharmacology 27 (8), 823–830. McKearney, J.W., 1982. Effects of tricyclic antidepressant and anticholinergic drugs on fixed-interval responding in the squirrel monkey. J. Pharmacol. Exp. Ther. 222 (1), 215–219. McNaughton, N., Morris, R.G.M., 1987. Chlordiazepoxide, an anxiolytic benzodiazepine, impairs place navigation in rats. Behav. Brain Res. 24, 39–46. McNaughton, N., Morris, R.G.M., 1992. Buspirone produces a dose-related impairment in spatial navigation. Pharmacol. Biochem. Behav. 43, 167–171. McNaughton, B.L., Barnes, C.A., Gerrard, J.L., Gothard, K., Jung, M.W., Knierim, J.J., ... Weaver, K.L., 1996. Deciphering the hippocampal polyglot: the hippocampus as a path integration system. J. Exp. Biol. 199, 173–185. Morrison, S.F., 2004. Activation of 5-HT1A receptors in raphe pallidus inhibits leptinevoked increases in brown adipose tissue thermogenesis. Am. J. Physiol. Regul. Integr. Comp. Physiol. 286 (5), R832–R837. Munn, R.G.K., McNaughton, N., 2008. Effects of fluoxetine on hippocampal rhythmic slow activity and behavioural inhibition. Behav. Pharmacol. 19 (3), 257–264. Nacmias, B., Ricca, V., Tedde, A., Mezzani, B., Rotella, C.M., Sorbi, S., 1999. 5-HT 2A receptor gene polymorphisms in anorexia nervosa and bulimia nervosa. Neurosci. Lett. 277 (2), 134–136. Panickar, K.S., McNaughton, N., 1991. Effects of buspirone on fixed interval responding in rats. J. Psychopharmacol. 5 (4), 410–417. http://dx.doi.org/10. 1177/026988119100500436. Paulus, K., Schulz, C., Lehnert, H., 2005. Central nervous effects of leptin and insulin on hippocampal leptin and insulin receptor expression following a learning task in
Wistar rats. Neuropsychobiology 51 (2), 100–106. http://dx.doi.org/10.1159/ 000084167. Perry, R.J., Zhang, X.M., Zhang, D., Kumashiro, N., Camporez, J.P., Cline, G.W., ... Shulman, G.I., 2014. Leptin reverses diabetes by suppression of the hypothalamic–pituitary–adrenal axis. Nat. Med. 20 (7), 759–763. http://dx.doi.org/10.1038/nm.3579. Riblet, L.A., Taylor, D.P., Eison, M.S., Stanton, H.C., 1982. Pharmacology and neurochemistry of buspirone. J Clin Psychiatry 43 (12 Pt 2), 11–18. Serra, M., Pisu, M.G., Muggironi, M., Parodo, V., Papi, G., Sari, R., ... Biggio, G., 2001. Opposite effects of short- versus long-term administration of fluoxetine on the concentrations of neuroactive steroids in rat plasma and brain. Psychopharmacology 158 (1), 48–54. http://dx.doi.org/10.1007/s002130100853. Suomalainen, M., Mannisto, P.T., 1998. Lack of effect of leptin on the behaviour of mice predicting the level of anxiety and depression. Pharmacol. Toxicol. 83 (3), 139–142. Thorsell, A., Caberlotto, L., Rimondini, R., Heilig, M., 2002. Leptin suppression of hypothalamic NPY expression and feeding, but not amygdala NPY expression and experimental anxiety. Pharmacol. Biochem. Behav. 71 (3), 425–430. Üçeyler, N., Schütt, M., Palm, F., Vogel, C., Meier, M., Schmitt, A., ... Sommer, C., 2010. Lack of the serotonin transporter in mice reduces locomotor activity and leads to genderdependent late onset obesity. Int. J. Obes. 34 (4), 701–711. Urban, J.H., Van de Kar, L.D., Lorens, S.A., Bethea, C.L., 1986. Effect of the anxiolytic drug buspirone on prolactin and corticosterone secretion in stressed and unstressed rats. Pharmacol. Biochem. Behav. 25 (2), 457–462. Wang, W., Liu, S.-L., Li, K., Chen, Y., Jiang, B., Li, Y.-K., ... Wang, F., 2015. Leptin: a potential anxiolytic by facilitation of fear extinction. CNS Neurosci. Ther. 21 (5), 425–434. http://dx.doi.org/10.1111/cns.12375. Wickens, J.R., Macfarlane, J., Booker, C., McNaughton, N., 2004. Dissociation of hypertension and fixed interval responding in two separate strains of genetically hypertensive rat. Behav. Brain Res. 152, 393–401. Widdowson, P.S., Upton, R., Buckingham, R., Arch, J., Williams, G., 1997. Inhibition of food response to intracerebroventricular injection of leptin is attenuated in rats with dietinduced obesity. Diabetes 46 (11), 1782–1785. Woodnorth, M.A., McNaughton, N., 2002. Similar effects of medial supramammillary or systemic injection of chlordiazepoxide on both theta frequency and fixed-interval responding. Cogn. Affect. Behav. Neurosci. 2 (1), 76–83. Yamada, J., Sugimoto, Y., Ujikawa, M., 1999. The serotonin precursor 5-hydroxytryptophan elevates serum leptin levels in mice. Eur. J. Pharmacol. 383 (1), 49–51. Zar, J.H., 2009. Biostatistical Analysis, 4th Impression. Dorling Kindersley (India) Pvt. Ltd., Delhi, 110, 92. Zhu, X.O., McNaughton, N., 1995. Similar effects of buspirone and chlordiazepoxide on a fixed interval schedule with long-term, low-dose administration. J. Psychopharmacol. 9, 326–330.
Contents lists available at ScienceDirect
Pharmacology, Biochemistry and Behavior journal homepage: www.elsevier.com/locate/pharmbiochembeh
Anxiolytic-like effects of leptin on fixed interval responding Susan M. Tyree 1, Robert G.K. Munn 2, Neil McNaughton ⁎ Department of Psychology, University of Otago, Dunedin, New Zealand
a r t i c l e
i n f o
Article history: Received 23 February 2016 Received in revised form 9 May 2016 Accepted 10 May 2016 Available online 11 May 2016 Keywords: Leptin Anxiety Anxiolytic Fixed interval Behavioural inhibition 5-HT
a b s t r a c t Leptin has been shown to affect energy homeostasis, learning and memory, and some models of anxiolytic action. However, leptin has produced inconsistent results in previous non-operant behavioural tests of anxiety. Here, we test the anxiolytic potential of leptin in an operant paradigm that has produced positive results across all classes of anxiolytic so far tested. Rats were tested in the Fixed Interval 60 Seconds (FI60) task following administration of 0/0.5/1.0 mg/kg (i.p.) leptin or an active anxiolytic control of 5 mg/kg (i.p.) chlordiazepoxide (CDP). By the end of the 14 days of testing in the FI60 task, 0.5 mg/kg leptin released suppressed responding in a manner similar to CDP, and 1.0 mg/kg leptin produced a relative depression in responding, a similar outcome pattern to previously tested 5HT-agonist anxiolytics. This suggests that leptin behaves similarly to established serotonergic anxiolytics such as buspirone and fluoxetine; with the delay in development of effect during testing, and the inverted-U dose–response curve explaining the inconsistent behaviour of leptin in behavioural tests of anxiety, as this type of pattern is common to serotonergic anxiolytics. © 2016 Elsevier Inc. All rights reserved.
1. Introduction Leptin is released primarily from adipocytes in response to the ingestion of dietary fat (Havel, 2000). It has long been thought to be a feedback signal involved in satiety, but has since been shown to have diverse roles unrelated to energy regulation (Margetic et al., 2002). One of these novel roles for leptin is to act centrally on some aspects of cognition and anxiety — functions typically thought to be controlled by the hippocampus (Asakawa et al., 2003; Farr et al., 2006; Gisou et al., 2009; Paulus et al., 2005). There is substantial behavioural evidence linking endogenous leptin to anxiety. Genetically modified ob/ob mice, which fail to produce leptin, display more anxious behaviour (Finger et al., 2010) and the administration of exogenous leptin reverses this difference (Asakawa et al., 2003). They also, like animals treated with anxiolytic drugs (McNaughton and Morris, 1987, 1992), show impaired spatial memory (Farr et al., 2006), implicating structures involved in the regulation of anxiety, such as the hippocampus, in the action of leptin. Conversely, maternal hyperleptinaemia, via injection of leptin while pups are in utero, produces adult rats that display improved spatial memory and reduced anxious behaviour compared to animals whose dams were injected with vehicle (Fraga-Marques et al., 2010). ⁎ Corresponding author at: Department of Psychology, University of Otago, P.O. Box 56, Dunedin 9054, New Zealand. E-mail addresses: [email protected] (S.M. Tyree), [email protected] (R.G.K. Munn), [email protected] (N. McNaughton). 1 Present address: Department of Molecular Genetics, German Institute of Human Nutrition, Potsdam-Rehbrücke, Germany. 2 Present address: Department of Neurobiology, Stanford University, Stanford, CA, USA.
http://dx.doi.org/10.1016/j.pbb.2016.05.005 0091-3057/© 2016 Elsevier Inc. All rights reserved.
There is also evidence for interaction of exogenous leptin with the control of anxiety. The Hypothalamic–Pituitary–Adrenocortical (HPA) axis is a central structure in anxiety (Faravelli et al., 2012; Landgraf et al., 1999). Leptin has been shown to diminish hyperactivity of the HPA axis (Holmes, 2015; Perry et al., 2014) and activation of neurons expressing the leptin receptor (LepRb) in the lateral hypothalamic area results in the inhibition of HPA axis activity (Bonnavion et al., 2015). Leptin is so central to this system that it has been included in the HPA axis as the Hypothalamic–Pituitary–Adrenal–Leptin axis (Aschbacher et al., 2014). It is also possible that leptinergic modulation of this axis is through 5-HT (Guo and Lu, 2014; Haleem et al., 2015; Kurhe et al., 2015). Leptin can also behave like established anxiolytics in some behavioural tests of anxiety (Liu et al., 2010; Wang et al., 2015). The picture in relation to exogenous anxiolytic action is not homogeneous, however. Several experiments have produced non-positive results. Suomalainen and Mannisto (1998) failed to obtain changes in anxiety-related behaviour in the elevated plus maze or open field. Buyse et al. (2001) found that leptin significantly reduced the number of open-arm entries in the elevated plus maze, as did Thorsell et al. (2002). In contrast to these anxiogenic-like findings in the elevated plus maze, Liu et al. (2010) showed that leptin administration produced anxiolytic-like effects in the elevated plus maze, but not in the open field test. Both the elevated plus maze and open field test are standard behavioural tests of anxiety. Hogg (1996) points out that the variability in response to the elevated plus maze, particularly in response to serotonergic agonists, is highly contingent on a variety of factors such as the aversiveness of the testing environment and the prior handling experience of the animals in the tests. There was also considerable variation in the dose of leptin used in these tests; Suomalainen and
16
S.M. Tyree et al. / Pharmacology, Biochemistry and Behavior 148 (2016) 15–20
Mannisto (1998) used 10 and 20 mg/kg, Buyse et al. (2001) 0.4–1 mg/ kg, and Liu et al. (2010) 0.25 and 1 mg/kg. In contrast to the i.p. administration in the other studies, Thorsell et al. (2002) delivered 0.1 mg/kg intraventricularly. While thought to bind primarily to the LEPR-encoded leptin receptor family there is evidence that leptin modulates, and is modulated by, the activity of 5-HT (Garcia-Alcocer et al., 2010; Morrison, 2004), perhaps in an indirect manner. 5-HT is a key neurotransmitter system that is involved in the anxiolytic function of serotonergic anxiolytics such as buspirone (Riblet et al., 1982) and fluoxetine (Fuller et al., 1991). Administration of leptin increases brain serotonin (5-HT) metabolism in mice (Calapai et al., 1999). There is also evidence that administration of the 5-HT precursor 5-hydroxytryptophan significantly increases serum leptin levels (Yamada et al., 1999); and serotonergic neurons have been shown to be targets for leptin in the monkey (Finn et al., 2001), strongly implying a serotonergic modulatory role for leptin in anxiety. Other lines of evidence indirectly link 5-HT to leptin. The 5-HT transporter, which is targeted by Serotonin-Specific Reuptake Inhibitors, is linked to obesity (Üçeyler et al., 2010). There is also evidence that polymorphisms of the 5-HT receptor are linked to anorexia and bulimia (Collier et al., 1997; Nacmias et al., 1999). This suggests that disorders of eating behaviour which are modulated by leptin, amongst other things, are linked to the 5-HT system. With increasing evidence of a connection between leptin and the 5HT system, as well as studies supporting its anxiolytic potential, and the established role for 5-HT agonists as anxiolytics, the inconsistent performance of leptin in behavioural assays of anxiety is perplexing. The open field and elevated plus maze tests used to assess leptin also produce inconsistent results when testing well-established clinical anxiolytics, not just with regard to failure to detect anxiolytic potential in known anxiolytics, but with some non-anxiolytic compounds being incorrectly detected as anxiolytic (for review, see (Crawley, 1985; Lister, 1990). In contrast to these spatial exploratory paradigms, which can produce mixed results, the Fixed Interval 60 Seconds (FI60) is an operant behavioural task that involves non-spatial behavioural inhibition. It assesses response suppression in a similar manner to the Vogel conditioned suppression of licking, Geller-Seifter, and Conditioned Emotional Response tests, but has the advantage of suppressing responding with the aversive state of frustration and not using shock and so not involving a confound with any analgesic action a drug may have. Fear and frustration have similar eliciting properties and their suppression of responding is similarly sensitive to anxiolytic drugs (Gray, 1977). Control animals in the FI60 paradigm rapidly learn to inhibit the bulk of their responding until after the 60-second non-reward period has elapsed. However, following administration of an anxiolytic drug animals are typically less able to inhibit their behaviour, making a larger number of inappropriately early responses during the non-reward period. This is true for all anxiolytics so far tested, regardless of precise pharmacological mode of action (Munn and McNaughton, 2008; Panickar and McNaughton, 1991; Zhu and McNaughton, 1995). One further advantage of the FI60 paradigm is that it provides an indirect assay of pharmacological mode of action. GABA-ergic anxiolytics such as the barbiturates and benzodiazepines show a linear dose–response relationship in the FI60; larger doses typically produce a larger release of responding. In contrast, the serotonergic anxiolytics so far tested tend to show an inverted-U dose–response curve; low doses produce a release of responding, while higher doses suppress responding (Munn and McNaughton, 2008; Panickar and McNaughton, 1991). It should be noted that while the form of the dose–response curve is different for the different classes of drugs, the low dose effects all involve a similar release of inhibition in the initial period of responding and the differences between the classes disappear with long-term administration (Zhu and McNaughton, 1995). The aim of the present study is therefore twofold. First, given the inconsistent results produced by standard tests of anxiety, we first aim to
examine the anxiolytic potential of leptin in the more reliable FI60. Secondly, should leptin behave as an anxiolytic in the FI60, we predict that it will show a dose response curve consistent with 5HT (i.e. response depression at higher doses) rather than GABA agency. 2. Methods 2.1. Subjects The subjects were 24 naïve male Sprague–Dawley rats, obtained from the University of Otago Department of Laboratory Animal Sciences. They weighed between 243 and 323 g immediately prior to the experiments. The home room was maintained on a 12 h light, 12 h dark cycle; (lights on at 0600 and off at 1800). The temperature of the home room was maintained at 20–22 °C. Animals in all cages had access to water and Reliance stock food pellets ad libitum. The animals were initially housed in 49 cm × 31 cm × 26.5 cm cages, in groups of four. Twelve days before the pre-training sessions began the animals were taken off ad libitum access to food and placed on a restricted diet of Reliance Stock food pellets. Animals were weighed daily, and the amount of food they received each day was varied in order to maintain their weight at 80% of their free feeding weight. Animals had ad libitum access to water throughout the experiment. 2.2. Apparatus Six standard operant chambers (Campden Instruments, U.K.) were used throughout the training sessions as well as all the experimental trials. The dimensions of the operant chambers were 57.5 cm × 34.5 cm × 39 cm. The interior of the operant chambers contained a smaller interior chamber which was 24 cm × 24 cm × 26 cm. A food delivery system distributed individual 45 mg Dustless Precision Pellets (Campden Instruments Ltd., UK) to a food hopper in the interior chamber through a plastic tube. The front-facing wall of the operant chamber was hinged from the bottom in order to allow access to the interior. One of the long exterior walls of the chamber had a circular tinted window 16.6 cm in diameter, which allowed a view of the interior of the chamber. The interior chamber of the operant chamber had three metal walls, a metal ceiling, and a horizontal grid floor. The front-facing wall of the interior chamber was a transparent Perspex wall that could be unlatched from the top of the box and pulled down to allow access to the chamber. One of the walls had a recessed 5 cm by 6.5 cm food hopper with a Perspex door hinged from the top, the food hopper was set in the middle and at the bottom of the wall. Two retractable metal levers were set into the wall, one on either side of, and equidistant from, the hopper. Only the left hand lever was extended into the chamber in the current experiment. Directly above each lever there were two 2.8 W stimulus lights set into the wall. The other two metal walls of the housing chamber were bare. In the centre of the metal ceiling of the housing chamber there was a 2.8 W house light, which was on throughout the training and experimental trials. There was also an electric fan, which provided ventilation for the animals, and also produced a constant level of background noise. Three IBM compatible computers, which were running Visual Basic 6 software with LABJACK control components, each controlled two operant chambers. Custom Visual Basic 6 programs were used to deliver the three training schedules. This meant that the computers controlled the timing and delivery of the reinforcements, as well as recording the number of lever presses the animal made, and the number of nose pokes into the food hopper. The computers also recorded the time each lever press was made during the FI60 task. These responses were divided into one of twelve five-second bins, according to the interval since the last reward that they were made. Responses made between 0 and 5 s since the last reward were allocated to bin one; responses made between 5 and 10 s since the last reward were allocated to bin two and so on.
S.M. Tyree et al. / Pharmacology, Biochemistry and Behavior 148 (2016) 15–20
2.3. Procedure 2.3.1. Autoshaping Each of the animals participating in the FI60 behavioural experiment received preliminary training to ensure that the animals would reliably insert their head into the hopper in order to retrieve a food pellet. Each animal was run on a random time (i.e. not response contingent) 30 s (RT30) schedule once a day. Each of these RT30 training sessions lasted for 30 min and each of the animals were run once a day at the same time each day. In order to encourage the animals to investigate the food hopper and put its head inside, a small pellet of the animal's regular food (Reliance Stock Pellets) was placed in the hopper so that it wedged open the Perspex hopper door. Once each animal was reliably responding to the light and sound cues of reinforcement by putting its head in the hopper, they were moved onto the next stage of training. The animals all required five sessions of RT30 training before they were reliably responding to reinforcement cues. 2.3.2. Continuous reinforcement Having completed the RT30 training schedule, all animals started on a continuous reinforcement (CRF) schedule. Each CRF training schedule lasted for 30 min and each of the animals ran once a day, at the same time each day, just as described in the RT30 schedule. Once the animals were consistently pressing the lever and receiving a food pellet, another two drug-free CRF sessions were run. Four training days were required to establish reliable lever pressing responses in the animals. Prior to the final CRF training session, animals were assigned to the four separate drug groups based on their CRF responses to ensure that the average CRF response rates were consistent across the four groups. The mean CRF response rates of the resultant groups differed by not N 5% overall. The animals were then rehoused to ensure that there was not more than one animal from each drug group in any one cage. This rehousing was carried out in order to remain consistent with previous experiments carried out in the current lab (Munn and McNaughton, 2008). Note that all drug groups will have been similarly affected by rehousing. Animals then ran a final CRF training session after having been administered the drug corresponding to the group to which they were assigned. Drug administration was started on the last day of CRF to prevent drug state acting as a signal of schedule change. The animals began their Fixed Interval training the day following this final CRF training session. 2.3.3. Drug preparation and administration The animals were divided into four groups: one passive control group, which was given the phosphate buffered saline (PBS) vehicle solution; one active control group, which was given a previously confirmed positive anxiolytic dose (5 mg/kg) of the benzodiazepine chlordiazepoxide (CDP); and two leptin groups, which were given doses of 0.5 mg/kg and 1.0 mg/kg, respectively. These doses of leptin were chosen in order to facilitate comparison to the doses used in those studies that show an anxiolytic effect of leptin (Haleem et al., 2015; Liu et al., 2010; Wang et al., 2015). All PBS used was pH adjusted to 8.0. Recombinant mouse leptin was supplied by NIDDK's National Hormone and Peptide Program and A.F. Parlow. Chlordiazepoxide Hydrochloride (CDP; Sigma Chemical Co., Louis, USA) was dissolved in 0.9% saline, for a concentration of 5 mg/ml. The drug solutions were prepared freshly on the day of injection and were administered via intraperitoneal (i.p.) injection using a 1 ml syringe and a 26 g needle at a volume of 1 ml/kg of body weight. Animals in the leptin drug groups were administered their injections 35 min before they were scheduled to begin their FI60 training session. This was to ensure that the maximum effect of leptin would be occurring during their testing session, specifically to peak half way through the thirty-minute FI60 session. The time interval was chosen based on the results of previous experiments in the laboratory (data not shown), which suggest that 50 min is the time of maximum drug effect on hippocampal rhythmic slow activity caused by 1.0 mg/kg leptin.
17
Animals in the 5 mg/kg CDP drug group were administered CDP 10 min before they were scheduled to begin their FI60 training session to ensure that the maximum effect of CDP would be occurring during the session, this time interval was based on previous data from electrophysiological and FI60 experiments using CDP (Woodnorth and McNaughton, 2002). A specific time-matched control group was not used to limit the use of animals and because only qualitative rather than quantitative positive control results were required. Animals therefore received a drug administration for one day of CRF training and each day of FI60 testing, for a total of 15 administrations of drug. 2.3.4. Fixed interval Having completed the CRF training, the animals began the FI60 reinforcement schedule. Animals were run in the FI60 task once each day, at the same time each day (between 1200 and 1600). The FI60 training sessions ran every day for 14 consecutive days. Each FI60 session lasted 30 min. 2.4. Data collection and analysis For the purposes of analysis and illustration, the twelve recorded 5 s bins were compressed into six 10 s bins so that the responses from bins one and two were now placed into bin one, responses in bins three and four were now in bin two, etc. Raw data were transformed by a logarithmic [log10(X + 1.0)] function in order to normalise the error variance (Zar, 2009). The log transformed raw data were subjected to Analysis of Variance (ANOVA) using the Genstat package (VSN international). Within subject factors such as responding across bins and days were extracted, and the presence of linear, quadratic, and cubic polynomial components in the data were analysed to investigate the effect of each drug on the response rates in the beginning, middle, and end of the FI60 schedule as well as over experimental days. The data were first subjected to an analysis including all groups to assess omnibus effects overall. This was followed by post-hoc ANOVAs and t-tests to determine the source of the omnibus effects. 3. Results Fig. 1 shows the average number of bar press responses per bin for each drug group at the beginning, middle, and end of the 15 days of testing. At times immediately after reinforcement, animals displayed low responding. Responses increased as time to reinforcement approached (bins[lin], F(1,3300) = 9278.22, p b 0.001) reaching a maximum in close proximity to reinforcement. Averaged over days, responding tended to be suppressed in the initial bins, rise during intermediate bins, and then flatten off in the final bins (bins[cub], F(1, 3300) = 318.39. p b 0.001). Animals showed a progressive decrease in responding in bin 1 and an increase in responding in bin 6 as they received more days of training with the response curve showing a mild positive inflection (inverted U-shape assessed by the quadratic component of bins) initially that changed to a more negative (U-shape) as training progressed (compare Fig. 1A and C; days[lin] × bins[quad], F(1,3300) = 4.99. p = 0.002). There were also differences across the different drug groups. The difference in overall response rate (averaged across bins) between the different drug groups increased steadily over days of testing (drug × days[lin], F(3, 280) = 3.73, p = 0.012; Fig. 2) but with an apparent decrease in the differences as training reached asymptote that did not achieve significance (drug × days[quad], F(3, 280) = 2.15, p = 0.094). This drug difference was clearest, overall, in bins 2–4 (drug × bins [lin] F(3, 3300) = 4.50, p = 0.004; [quad] F(3, 3300) = 46.36, p b 0.001; [cub] F(3, 3300) = 12.38, p b 0.001) with the differences generally increasing over days (drug × days[lin] × bins [quad] F(3, 3300) = 4.99, p = 0.002). The response curve for the CDP group showed an inverted
18
S.M. Tyree et al. / Pharmacology, Biochemistry and Behavior 148 (2016) 15–20
Fig. 2. The mean number of responses made by the 0 mg/kg Leptin (open circles), 0.5 mg/ kg leptin (grey circles), and 1 mg/kg leptin (filled black circles) groups (n = 6 in each group) in each day of FI60 testing. All three groups begin testing with similar rates of response, but rapidly diverge such that the 0.5 mg/kg leptin group makes consistently more responses than control, and the 1.0 mg/kg leptin group consistently makes fewer responses. The fitted grey lines illustrate the significant linear and marginal quadratic components of the drug × days interaction found in the main analysis (0 mg/kg; solid line, 0.5 mg/kg; dotted and dashed line, 1.0 mg/kg; dashed line). The non-linear y-axis is a result of the logarithmic transform used to normalise the data.
F(1, 1650) = 5.01, p = 0.025) and more so for 1 mg/kg (leptin[1.0 versus 0.0] × bins[quad), F(1, 1650) = 53.32, p b 0.001). As can be seen in Fig. 1C, by the end of training, the behaviour of the 0.5 mg/kg group was very similar to that of the CDP group, with the greatest release of responding, relative to control, being in bin 2 when the controls show the greatest response suppression. The 1 mg/kg leptin group tended to make progressively fewer responses than the 0 mg/kg control group over days until towards the end of testing, when the responses in the 1 mg/kg group increased until the groups were approximately equal (leptin[1.0 versus 0.0] × days[quad], F(1140) = 4.76, p b 0.05; Fig. 2); but with a difference limited to intermediate bins (3 and 4) becoming most obvious at the end of training (leptin[1.0 versus 0.0] × days[lin] × bins[quad), F(1, 1650) = 6.99, p = 0.008; Fig. 1C). 4. Discussion
Fig. 1. The average number of responses per bin of the FI60 schedule by each of the 0 mg/ kg leptin, 0.5 mg/kg leptin, 1.0 mg/kg leptin, and 5.0 mg/kg CDP groups. Each bin-pair contains the average data from all six animals in each drug group averaged across A) Days 1–4, B) days 6–9, and C) days 11–15. The non-linear response axis is the result of the logarithmic transform used to normalise the data. Over days 1–4 responses in the saline and both leptin groups are equivalent over all bins, while the CDP group shows a general release of responding. Over days 6–9 the saline and 0.5 mg/kg leptin groups are equivalent over all bins, but there is a general depression of responding in the 1.0 mg/kg leptin group and a continued release of responding in the CDP group. By days 11–15 the responses of both the CDP and 0.5 mg/kg groups show an equivalent release of responding in comparison to saline control, while the responses in the 1.0 mg/kg leptin group remain depressed relative to saline control. The error bars are 2 × S.E.M. The nonlinear y-axis is a result of the logarithmic transform used to normalise the data.
U-shape, and that of the 1.0 mg/kg leptin group a U-shape, to the greatest extent in the middle of training compared to the beginning and the end of training (Fig. 1B; drug × days[quad] × bins[quad], F(1,3300) = 4.75. p = 0.003; drug × days[quad] × bins[lin], F(1,3300) = 2.74. p b 0.05). The significant drug effects in the overall analysis were explored with separate post-hoc comparisons of each leptin dose against 0 mg/kg control. The overall effect of drug on the intermediate bins was significant for the 0.5 mg/kg dose (leptin[0.5 versus 0.0] × bins[quad),
As expected, the active control drug CDP at a dose of 5.0 mg/kg caused a release of responding throughout training compared to control, most obviously in the early bins, as in previous experiments with this drug. This result is typical of a GABA agonist anxiolytic agent. In a similar manner, low dose (0.5 mg/kg) leptin released responding in the early bins of the FI60 schedule, particularly towards the end of testing; by the final days of FI60 testing low dose leptin and CDP were producing a comparable release of responding. This finding adds support to the notion that leptin may be clinically anxiolytic, at least at lower doses. Our high dose (1.0 mg/kg) of leptin produced a robust depression of responding in comparison with vehicle control. This effect is strikingly similar to that typically observed when a high dose of serotonergic agents such as buspirone or fluoxetine are tested; this effect is consistent in animals as diverse as monkeys, pigeons, and rats (Leander and Carter, 1984; McKearney, 1982; McNaughton et al., 1996; Munn and McNaughton, 2008; Panickar and McNaughton, 1991). The dose–differential effect of leptin on responding in the FI60 task is consistent with the growing body of evidence that leptin acts, at least for its anxiolytic action, through serotonergic modulation. Leptin typically produces dose-dependent reductions in food intake starting from doses of 0.5 mg/kg (Widdowson et al., 1997). In contrast, we observe an increase of responding compared to control at this dose, suggesting
S.M. Tyree et al. / Pharmacology, Biochemistry and Behavior 148 (2016) 15–20
that the satiety-inducing effect of leptin is not a factor in these experiments. Anxiolytics often increase food intake but effects on appetite would be expected to affect terminal rate responding (when food is anticipated) not initial rate responding (when behavioural inhibition is strongest). A similar argument rules out changes in impulsivity since, in the FI schedule, rat models of ADHD show increased terminal but not initial responding (Wickens et al., 2004). The FI60 has been shown to produce characteristic changes in behaviour depending on whether an anxiolytic functions via GABA-ergic or serotonergic action (Munn and McNaughton, 2008; Panickar and McNaughton, 1991). This difference in behaviour caused by these two families of drugs is believed to be mediated by their effect on the stress hormone corticosterone. The 5-HT1A agonist buspirone, unlike benzodiazepines such as CDP, does not block stress-induced corticosterone release (Urban et al., 1986), and in fact, at high doses (over 1 mg/kg), triggers the release of corticosterone without the presence of an external stressor (De Boer et al., 1990; Matheson et al., 1988). This effect shows tolerance following long-term treatment with high doses of buspirone (De Boer et al., 1990) or fluoxetine (Serra et al., 2001). In the FI60, long-term administration of serotonergics or pharmacologic blockade of corticosterone release render the action of serotonergics similar to that of the GABA-agonists (McNaughton et al., 1996; Zhu and McNaughton, 1995), suggesting that acute serotonin-driven corticosterone release, which can show tolerance, is likely the cause of the U-shaped dose–response relationship observed when testing these drugs. Interestingly, leptin has also previously been shown to modulate corticosterone levels: optogenetic activation of LepRb neurons reduces corticosterone levels in stressed mice (Bonnavion et al., 2015) and conversely intracerebroventricular administration of leptin resulted in increased basal levels of corticosterone in rats (Thorsell et al., 2002). Like the clinical response, the effects of serotonergic and GABA drugs are, as noted above, similar on the FI when they have been previously administered 3 times per day for 60 days (Zhu and McNaughton, 1995). However, neither the anxiolytic nor depressant effects of buspirone on the FI shows much change across days of testing when administered as in the current experiment (Panickar and McNaughton, 1991). The slow-development of the leptin effects in the current study may, then, partially clarify the inconsistent results seen in previous behavioural screenings of leptin for anxiolytic potential. The doses and time scales used in these experiments are very heterogeneous. For example, studies testing the anxiolytic potential of leptin in the elevated plus maze showed variations either in leptin dosage; 0/0.25/0.5 mg/kg (Liu et al., 2010), 0.4/1.0 mg/kg (Buyse et al., 2001), or 10/20 mg/kg (Suomalainen and Mannisto, 1998), or delay between drug administration and testing; 30 min (Liu et al., 2010), 1 h (Buyse et al., 2001), or 10– 20 min (Suomalainen and Mannisto, 1998). It is important to note that the positive anxiolytic-typical findings in these tasks occur at doses and times similar to those that are used in the present experiments that positively identify leptin as anxiolytic. The results from the present study show that high dose leptin produces a qualitatively similar depression in responding on the FI60 task to high doses of buspirone or fluoxetine. This finding, in addition to previous evidence showing leptin has been implicated in modulation of the serotonergic system (Calapai et al., 1999; Charnay et al., 2000; Yamada et al., 1999), and evidence that leptin may also release, or enhance the release of, corticosterone (Thorsell et al., 2002), suggests that leptin may act, directly or indirectly, as an agonist of serotonin, potentially at 1A receptors, as part of its anxiolytic mode of action. Author disclosures There were no external funds used in the production of this study. N.M. has a confidential disclosure and consulting agreement with Janssen Research & Development, LLC. R.M. and N.M. designed the study and wrote the protocol. S.T. ran the experimental procedures. S.T., R.M. and N.M. undertook the statistical analysis, and S.T. wrote
19
the first draft of the manuscript. All authors contributed to- and have approved the final manuscript. Acknowledgements Leptin was obtained from the NIH National Institute of Diabetes and Digestive and Kidney diseases (NIDDK) National Hormone and Peptide Program and A.F. Parlow. R. Munn was supported by a postdoctoral fellowship from the Neurological Foundation of New Zealand.
References Asakawa, A., Inui, A., Inui, T., Katsuura, G., Fujino, M.A., Kasuga, M., 2003. Leptin treatment ameliorates anxiety in ob/ob obese mice. J. Diabetes Complicat. 17 (2), 105–107. Aschbacher, K., Rodriguez-Fernandez, M., van Wietmarschen, H., Tomiyama, A.J., Jain, S., Epel, E., ... van der Greef, J., 2014. The hypothalamic–pituitary–adrenal–leptin axis and metabolic health: a systems approach to resilience, robustness and control. Interface Focus 4 (5). Bonnavion, P., Jackson, A.C., Carter, M.E., de Lecea, L., 2015. Antagonistic interplay between hypocretin and leptin in the lateral hypothalamus regulates stress responses. Nat. Commun. 6. Buyse, M., Bado, A., Daugé, V., 2001. Leptin decreases feeding and exploratory behaviour via interactions with CCK 1 receptors in the rat. Neuropharmacology 40 (6), 818–825. Calapai, G., Corica, F., Corsonello, A., Sautebin, L., Rosa, D., Massimo, C., Giuseppe, M., ... Caputi, A.P., 1999. Leptin increases serotonin turnover by inhibition of brain nitric oxide synthesis. J. Clin. Investig. 104 (7), 975. Charnay, Y., Cusin, I., Vallet, P.G., Muzzin, P., Rohner-Jeanrenaud, F., Bouras, C., 2000. Intracerebroventricular infusion of leptin decreases serotonin transporter binding sites in the frontal cortex of the rat. Neurosci. Lett. 283 (2), 89–92. Collier, D.A., Arranz, M.J., Li, T., Mupita, D., Brown, N., Treasure, J., 1997. Association between 5-HT2A gene promoter polymorphism and anorexia nervosa. Lancet 350 (9075), 412. Crawley, J.N., 1985. Exploratory behavior models of anxiety in mice. Neurosci. Biobehav. Rev. 9 (1), 37–44. De Boer, S.F., Slangen, J.L., Van der Gugten, J., 1990. Effects of chlordiazepoxide and buspirone on plasma catecholamine and corticosterone levels in rats under basal and stress conditions. Endocrinol. Exp. 24 (1–2), 229–239. Faravelli, C., Lo Sauro, C., Lelli, L., Pietrini, F., Lazzeretti, L., Godini, L., ... Ricca, V., 2012. The role of life events and HPA axis in anxiety disorders: a review. Curr. Pharm. Des. 18 (35), 5663–5674. http://dx.doi.org/10.2174/138161212803530907. Farr, S.A., Banks, W.A., Morley, J.E., 2006. Effects of leptin on memory processing. Peptides 27 (6), 1420–1425. http://dx.doi.org/10.1016/j.peptides.2005.10.006. Finger, B.C., Dinan, T.G., Cryan, J.F., 2010. Leptin-deficient mice retain normal appetitive spatial learning yet exhibit marked increases in anxiety-related behaviours. Psychopharmacology 210 (4), 559–568. http://dx.doi.org/10.1007/s00213-010-1858-z. Finn, P.D., Cunningham, M.J., Rickard, D.G., Clifton, D.K., Steiner, R.A., 2001. Serotonergic neurons are targets for leptin in the monkey. J. Clin. Endocrinol. Metab. 86 (1), 422–426. Fraga-Marques, M.C., Moura, E.G., Silva, J.O., Claudio-Neto, S., Pereira-Toste, F., Passos, M.C.F., ... Manhaes, A.C., 2010. Effects of maternal hyperleptinaemia during lactation on short-term memory/learning, anxiety-like and novelty-seeking behavioral traits of adult male rats. Behav. Brain Res. 206 (1), 147–150. http://dx.doi.org/10.1016/j. bbr.2009.08.029. Fuller, R.W., Wong, D.T., Robertson, D.W., 1991. Fluoxetine, a selective inhibitor of serotonin uptake. Med. Res. Rev. 11 (1), 17–34. Garcia-Alcocer, G., Rodriguez, A., Moreno-Layseca, P., Berumen, L.C., Escobar, J., Miledi, R., 2010. Serotonin receptor 5-HT(5A) in rat hippocampus decrease by leptin treatment. Neurosci. Lett. 486 (3), 171–173. Gisou, M., Soheila, R., Nasser, N., 2009. Evaluation of the effect of intrahippocampal injection of leptin on spatial memory. Afr. J. Pharm. Pharmacol 3 (9), 443–448. Gray, J.A., 1977. Drug effects on fear and frustration: possible limbic site of action of minor tranquilizers. In: Iversen, L.L., Iversen, S.D., Snyder, S.H. (Eds.), Handbook of PsychopharmacologyDrugs, Neurotransmitters and Behaviour vol. 8. Plenum Press, New York, pp. 433–529. Guo, M., Lu, X.Y., 2014. Leptin receptor deficiency confers resistance to behavioral effects of fluoxetine and desipramine via separable substrates. Transl. Psychiatry 4 (12), e486. Haleem, D.J., Haque, Z., Inam, Q.U., Ikram, H., Haleem, M.A., 2015. Behavioral, hormonal and central serotonin modulating effects of injected leptin. Peptides http://dx.doi. org/10.1016/j.peptides.2015.10.002. Havel, P.J., 2000. Role of adipose tissue in body-weight regulation: mechanisms regulating leptin production and energy balance. Proc Nutr Soc 59 (3), 359–371. Hogg, S., 1996. A review of the validity and variability of the elevated plus-maze as an animal model of anxiety. Pharmacol. Biochem. Behav. 54 (1), 21–30. http://dx.doi.org/ 10.1016/0091–3057(95)02126–4. Holmes, D., 2015. Neuroendocrinology: leptin attenuates HPA-axis activation and stress responses. Nat. Revs. Endocrinol. 11 (5), 255. http://dx.doi.org/10.1038/nrendo. 2015.34. Kurhe, Y., Mahesh, R., Devadoss, T., 2015. QCM-4, a 5-HT3 receptor antagonist ameliorates plasma HPA axis hyperactivity, leptin resistance and brain oxidative stress in
20
S.M. Tyree et al. / Pharmacology, Biochemistry and Behavior 148 (2016) 15–20
depression and anxiety-like behavior in obese mice. Biochem. Biophys. Res. Commun. 456 (1), 74–79. http://dx.doi.org/10.1016/j.bbrc.2014.11.036. Landgraf, R., Wigger, A., Holsboer, F., Neumann, I.D., 1999. Hyper-reactive hypothalamo– pituitary–adrenocortical axis in rats bred for high anxiety-related behaviour. J. Neuroendocrinol. 11 (6), 405–407. Leander, J.D., Carter, R.B., 1984. Effects of norepinephrine and serotonin uptake inhibitors on the schedule-controlled behavior of the pigeon. Pharmacol. Biochem. Behav. 20 (3), 391–395. http://dx.doi.org/10.1016/0091-3057(84)90277–6. Lister, R.G., 1990. Ethologically-based animal models of anxiety disorders. Pharmacol. Ther. 46 (3), 321–340. Liu, J., Garza, J.C., Bronner, J., Kim, C.S., Zhang, W., Lu, X.-Y., 2010. Acute administration of leptin produces anxiolytic-like effects: a comparison with fluoxetine. Psychopharmacology 207, 535–545. http://dx.doi.org/10.1007/s00213-009-1684-3. Margetic, S., Gazzola, C., Pegg, G.G., Hill, R.A., 2002. Leptin: a review of its peripheral actions and interactions. Int. J. Obes. Relat. Metab. Disord. 26 (11), 1407–1433. http:// dx.doi.org/10.1038/sj.ijo.0802142. Matheson, G.K., Gage, D., White, G., Dixon, V., Gipson, D., 1988. A comparison of the effects of buspirone and diazepam on plasma corticosterone levels in rat. Neuropharmacology 27 (8), 823–830. McKearney, J.W., 1982. Effects of tricyclic antidepressant and anticholinergic drugs on fixed-interval responding in the squirrel monkey. J. Pharmacol. Exp. Ther. 222 (1), 215–219. McNaughton, N., Morris, R.G.M., 1987. Chlordiazepoxide, an anxiolytic benzodiazepine, impairs place navigation in rats. Behav. Brain Res. 24, 39–46. McNaughton, N., Morris, R.G.M., 1992. Buspirone produces a dose-related impairment in spatial navigation. Pharmacol. Biochem. Behav. 43, 167–171. McNaughton, B.L., Barnes, C.A., Gerrard, J.L., Gothard, K., Jung, M.W., Knierim, J.J., ... Weaver, K.L., 1996. Deciphering the hippocampal polyglot: the hippocampus as a path integration system. J. Exp. Biol. 199, 173–185. Morrison, S.F., 2004. Activation of 5-HT1A receptors in raphe pallidus inhibits leptinevoked increases in brown adipose tissue thermogenesis. Am. J. Physiol. Regul. Integr. Comp. Physiol. 286 (5), R832–R837. Munn, R.G.K., McNaughton, N., 2008. Effects of fluoxetine on hippocampal rhythmic slow activity and behavioural inhibition. Behav. Pharmacol. 19 (3), 257–264. Nacmias, B., Ricca, V., Tedde, A., Mezzani, B., Rotella, C.M., Sorbi, S., 1999. 5-HT 2A receptor gene polymorphisms in anorexia nervosa and bulimia nervosa. Neurosci. Lett. 277 (2), 134–136. Panickar, K.S., McNaughton, N., 1991. Effects of buspirone on fixed interval responding in rats. J. Psychopharmacol. 5 (4), 410–417. http://dx.doi.org/10. 1177/026988119100500436. Paulus, K., Schulz, C., Lehnert, H., 2005. Central nervous effects of leptin and insulin on hippocampal leptin and insulin receptor expression following a learning task in
Wistar rats. Neuropsychobiology 51 (2), 100–106. http://dx.doi.org/10.1159/ 000084167. Perry, R.J., Zhang, X.M., Zhang, D., Kumashiro, N., Camporez, J.P., Cline, G.W., ... Shulman, G.I., 2014. Leptin reverses diabetes by suppression of the hypothalamic–pituitary–adrenal axis. Nat. Med. 20 (7), 759–763. http://dx.doi.org/10.1038/nm.3579. Riblet, L.A., Taylor, D.P., Eison, M.S., Stanton, H.C., 1982. Pharmacology and neurochemistry of buspirone. J Clin Psychiatry 43 (12 Pt 2), 11–18. Serra, M., Pisu, M.G., Muggironi, M., Parodo, V., Papi, G., Sari, R., ... Biggio, G., 2001. Opposite effects of short- versus long-term administration of fluoxetine on the concentrations of neuroactive steroids in rat plasma and brain. Psychopharmacology 158 (1), 48–54. http://dx.doi.org/10.1007/s002130100853. Suomalainen, M., Mannisto, P.T., 1998. Lack of effect of leptin on the behaviour of mice predicting the level of anxiety and depression. Pharmacol. Toxicol. 83 (3), 139–142. Thorsell, A., Caberlotto, L., Rimondini, R., Heilig, M., 2002. Leptin suppression of hypothalamic NPY expression and feeding, but not amygdala NPY expression and experimental anxiety. Pharmacol. Biochem. Behav. 71 (3), 425–430. Üçeyler, N., Schütt, M., Palm, F., Vogel, C., Meier, M., Schmitt, A., ... Sommer, C., 2010. Lack of the serotonin transporter in mice reduces locomotor activity and leads to genderdependent late onset obesity. Int. J. Obes. 34 (4), 701–711. Urban, J.H., Van de Kar, L.D., Lorens, S.A., Bethea, C.L., 1986. Effect of the anxiolytic drug buspirone on prolactin and corticosterone secretion in stressed and unstressed rats. Pharmacol. Biochem. Behav. 25 (2), 457–462. Wang, W., Liu, S.-L., Li, K., Chen, Y., Jiang, B., Li, Y.-K., ... Wang, F., 2015. Leptin: a potential anxiolytic by facilitation of fear extinction. CNS Neurosci. Ther. 21 (5), 425–434. http://dx.doi.org/10.1111/cns.12375. Wickens, J.R., Macfarlane, J., Booker, C., McNaughton, N., 2004. Dissociation of hypertension and fixed interval responding in two separate strains of genetically hypertensive rat. Behav. Brain Res. 152, 393–401. Widdowson, P.S., Upton, R., Buckingham, R., Arch, J., Williams, G., 1997. Inhibition of food response to intracerebroventricular injection of leptin is attenuated in rats with dietinduced obesity. Diabetes 46 (11), 1782–1785. Woodnorth, M.A., McNaughton, N., 2002. Similar effects of medial supramammillary or systemic injection of chlordiazepoxide on both theta frequency and fixed-interval responding. Cogn. Affect. Behav. Neurosci. 2 (1), 76–83. Yamada, J., Sugimoto, Y., Ujikawa, M., 1999. The serotonin precursor 5-hydroxytryptophan elevates serum leptin levels in mice. Eur. J. Pharmacol. 383 (1), 49–51. Zar, J.H., 2009. Biostatistical Analysis, 4th Impression. Dorling Kindersley (India) Pvt. Ltd., Delhi, 110, 92. Zhu, X.O., McNaughton, N., 1995. Similar effects of buspirone and chlordiazepoxide on a fixed interval schedule with long-term, low-dose administration. J. Psychopharmacol. 9, 326–330.