Role of leptin in the regulation of food intake in fasted mice.

Abstract Leptin is well acknowledged as an anorexigenic hormone that plays an important role in feeding control. Hypothalamic GABA system plays a significant role in leptin regulation on feeding and metabolism control. However, the pharmacological relationship of leptin and GABA receptor is still obscure. Therefore, we investigated the effect of leptin or combined with baclofen on the food intake in fasted mice. We detected the changes in hypothalamic c‐Fos expression, hypothalamic TH, POMC and GAD67 expression, plasma insulin, POMC and GABA levels to demonstrate the mechanisms. We found that leptin inhibit fasting‐induced increased food intake and activated hypothalamic neurons. The inhibitory effect on food intake induced by leptin in fasted mice can be reversed by pretreatment with baclofen. Baclofen reversed leptin's inhibition on c‐Fos expression of PAMM in fasted mice. Therefore, these results indicate that leptin might inhibit fasting‐triggered activation of PVN neurons via presynaptic GABA synaptic functions which might be partially blocked by pharmacological activating GABA‐B. Our findings identify the role of leptin in the regulation of food intake.


| INTRODUC TI ON
Feeding behaviour is one of the most basic physiological activities of humans and animals and is closely related to energy expenditure and metabolism. It is well acknowledged that acute or chronic fasting can trigger overeating and hyperphagia. 1 In the past decades, leptin, a hormone encoded by ob gene and secreted by adipose cells, is well studied in energy balance and glucose metabolism. 2,3 A number of animal studies have shown that leptin can suppress appetite and reduce the food consumption via acting on specific neurons in hypothalamus. 4 Hypothalamic gamma-aminobutyric acid (GABA)ergic neurons play significant role in the control of feeding and metabolism. It has been shown that leptin-responsive neurons in hypothalamus are mainly GABAergic and largely located in the dorsomedial hypothalamus (DM), the arcuate nucleus of hypothalamus (ARC) and the lateral hypothalamus (LH). 5 It was reported that leptin can selectively activate anorexigenic pro-opiomelanocortin (POMC) neurons via influencing presynaptic release of GABA. 6  GABA-B, both of which have been reported to have important effect in the regulation of feeding behaviour. 7 It was found that administration of GABA-A receptor agonist, muscimol, can enhance feeding behaviour. [8][9][10][11] Systematic administration of benzodiazepine partial receptor agonist provoked a hyperphagic response. 9,12 Conversely, some evidence reported that administration of GABA-A antagonist into the LH promote eating, while the agonist of GABA-A receptor suppressed feeding. 13 Intracerebroventricular administration of antagonist of GABA-A receptor bicuculline increased eating and drinking behaviours. 14 Pharmacological studies showed that peripheral or intracranial administration of the GABA-B agonist baclofen increased food intake. [15][16][17][18][19] Conversely, administration of GABA-B receptor antagonist 3-aminopropyl-diethoxy-methyl-phosphinic acid (CGP 35348) reduced food consumption in rats after 22 hours of fasting. 20 Chronic systematic administration of 4 mg/kg baclofen did not affect daily cumulative food intake, but reducing the bodyweight gain in comparison with control rats. 21 These results indicate that GABA receptors exert potential anti-anorexia effect via different molecular mechanisms. A recent research found that leptin-mediated inhibition of orexin neuron activities was independent of GABA-A receptor. 22 Currently, still few studies investigated the pharmacological effect of GABA-B receptor implicated in the anorectic effects of leptin. Therefore, in the present study, we investigated the role of GABA-B receptor played in the feeding regulation of leptin in fasted mice and non-fasted mice. We further demonstrate the potential molecular mechanisms via investigating hypothalamic neuronal activation, glutamate decarboxylase 67 (GAD67), tyrosine hydroxylase (TH) and POMC protein levels of hypothalamus, serum insulin, POMC and GABA levels alterations after systematic administration of leptin or combined with baclofen.

| Animals
All animals we used were 5-week-old, 20 ± 2 g weight imprinting control region (ICR) strain male mice which were purchased from Jilin University. All mice were single-housed in a cage (25.5 × 15 × 14 cm) for at least 1 week before all experimental procedures. This study was carried out under the approval from the ethics committee of Second Hospital of Jilin University.

| Drugs
Leptin (Tocris) was dissolved in saline in dose of 1 mg/kg. About 1 mg leptin was dissolved in 1 mL saline, and the compound was further diluted into final concentration of 0.1 mg/mL for injection. Baclofen (Sigma) was dissolved in saline in dose of 8 mg/kg. About 80 mg baclofen was dissolved in 10 mL saline, and the compound was further diluted into final concentration of 0.8 mg/mL for injection. All drugs and vehicle (saline) were intraperitoneally (i.p.) injected in a volume calculated based on the bodyweight (0.01 mL/g).

| Food intake measurements
Before food intake measurements, each animal was single-housed and fasted for 16 hours (overnight) from 5:00 pm to the next day 9:00 am. Animals were injected with leptin (1 mg/kg), baclofen (8 mg/kg) and saline according to the experimental groups. Baclofen was injected 15 minutes before leptin injection. After all drug treatment, each animal was placed in a new cage and refed with 4-6 food pellets. The food intake was measured at 0.5, 1 and 1.5 hours after the final injection as we described previously. 23 For calculation of the cumulative food intake, the latter weight of feed was subtracted from the former value. After recording the food intake, all mice were killed by decapitation. The brain tissue and blood sample were saved for subsequent neurochemical test.

| Western blot analysis for GAD67, TH and POMC
Hypothalamus were extracted from saved brain tissue and ho- with Tween 20 buffer, membranes were then incubated in horseradish peroxidase (HRP)-conjugated secondary antibodies (Proteintech, SA00001-2, 1:1000), and then membranes were exposed to chemiluminescent detection reagents and imaged. The optical density of all sample bands was detected by ImageJ software.

| c-Fos immunohistochemistry
c-Fos immunohistochemistry analysis method was presented as our previous studies. 23 All mice were intraperitoneal injected with 400 mg/kg chloral hydrate for anesthetization. The perfused fixed brain tissue was cut into 30 μm slices on a cryostat. After washing three times in phosphate buffer saline (PBS) containing 0.1% Triton X-100 (PBST), brain slices were incubated 15 minutes with 0.6% hydrogen peroxide in PBS. After three times washing with PBS, sections were incubated with anti-c-Fos (Santa Cruz, CA; #sc-52, 1:1000) primary diluents which is diluted with PBS containing 0.3% Triton X-100, 0.05% sodium azide and 2% normal goat serum for 72 hours at 4°C. The sections were next successively incubated with biotinylated secondary antibody diluent (ZSGB-Bio, ZB2305, 1:400) and the diluent of avidin-biotinylated horseradish peroxidase complex for 75 minutes at room temperature, and then using the glucose oxidase-diaminobenzidine-nickel method to visualize the reaction product and subsequently terminating the reaction by washing in sodium acetate buffer. Sections were attached onto chrome alum/ gelatin-coated slides and air-dried at room temperature. The c-Fospositive cells were identified by dense brown nuclear staining and counted under ×200 magnification from different nucleus of hypothalamus. The c-Fos-positive neurons of per bilateral regions were pictured under microscopy (Eclipse 50i Microscope, Nikon) at least from 5-8 sections as described previously. 24 The counting was based on average of counting c-Fos numbers by experimenters who were blind to the experimental conditions.

| ELISA analysis for plasma insulin, GABA and POMC
Collected blood was left at 4°C overnight and then centrifuged at 4°C for 15 minutes at 1509 g. And then, we transferred the serum to a new eppendorf tubes until use. We use Mouse GABA, insulin and POMC (Jiancheng) ELISA Kits to detect the serum insulin, GABA and POMC levels according to the manufacturers' instructions.

| Statistical analysis
All data are presented as mean ± SEM Statistical analysis was performed with SPSS statistics for Windows (version 25.0). All data we presented were using one-way analysis of variance (ANOVA) followed by a post hoc Dunnett's test. Significant differences were considered exist when the P values were less than .05.

| Effect of leptin or combined with baclofen on food intake in fasted mice
Firstly, we detected the cumulative food intake in the 0-1.5 hours period after all drug treatments ( Figure 1). We found that 16 hours food deprivation significantly increased the food intake during 0-0.5 hours (one-way ANOVA, F (4,35) = 15,755, P < .001). We found that pretreatment with 8 mg/kg baclofen significantly reduce the food intake during 0-0.5 hours period (P < .001) and 0.5-1 hours (P < .05). During the 0.5-1 hours period after drug treatment (oneway ANOVA, F (4,35) = 5.457), we found that leptin treatment significantly reduced food intake in comparison with the fasting group (P < .05), and baclofen pretreatment reversed this effect (P < .05).
No significant difference on cumulative food intake was detected in 1-1.5 hours period after drug treatment (one-way ANOVA,

| Effect of leptin or combined with baclofen on c-Fos expression in hypothalamus of fasted mice
We next detected the neuronal activation of hypothalamus in fasted mice after leptin or combined with baclofen treatment. We found that 16 hours food deprivation significantly increased c-Fos expression in DM ( Figure 2G, F (4,19)  We also found that single baclofen injection after fasting significantly reduced the c-Fos number in DM (P < .01), VM (P < .05), PAMM (P < .01) and AHP (P < .01) of hypothalamus (Bac group versus Fast group. Figure 5). alone also decreased GAD67 expression (P < .01, Bac group versus Fast group). We observed increased TH expression in fasted mice (P < .05, Fast group versus Cont group). There was no significant difference in POMC expression between groups.

| Effect of leptin or combined with baclofen on insulin POMC and GABA levels in fasted mice
We finally detected serum insulin ( Figure 7A, F (4,29)

| D ISCUSS I ON
Our result found that 16-hour food deprivation significantly increased the cumulative food intake and the increase was gradually weakened over time as previously reported. 25 Systematic injection of 1 mg/kg leptin inhibited fasting-induced increase of food consumption in mice. This action of leptin was the most significant in the 0.5-1 hours period after drug treatments. We also observed that pretreatment with baclofen reversed leptin-induced reduction of food intake in fasted mice, indicating that GABA-B receptor might be implicated in the regulation of food intake by leptin. We have previously reported that 5-HT 3 receptor agonist SR-57227 significantly inhibited the food intake of fasted mice which can be blocked by 5-HT 3 receptor antagonist ondansetron. 26 Some evidence have shown that systematic baclofen administration can increase the food intake of animals in a dose-dependent manner. 16,27,28 Chronic 2 mg/kg baclofen treatment enhanced shortterm food intake while had no significant effect on daily cumulative food intake and the weight gain of rats. 19 Ebenezer et al 27 reported that 1-8 mg/kg systematic administration of baclofen caused a dose-related increase in cumulative food intake and the increase was the most significant in CFLP mice treated with 8 mg/kg baclofen. They subsequently reported that the promotion of baclofen on food intake was related to hunger or satiety conditions and 1-4 mg/kg baclofen treatment had no significant effect on feeding behaviours of rats treated with 22-hour food deprivation, 28 which is consistent with an earlier study. 29  Leptin has been shown to modulate mesolimbic DA system via activating LepR-containing LH neurons. 40 Earlier research reported fasting can promote expression of TH, a catecholamine biosynthesis-mediated enzyme, and DA release. 41,42 However, in present study, we observed a reduction of TH expression of hypothalamus in fasted mice, indicating that the role of catecholaminergic neurons played in fasting-induced overeating needs to be further investigated. It was reported that LepR-expressed GABA and POMC neurons were implicated in increased glucose utilization upon insulin deficiency. 30 However, we did not observe statistical difference of insulin levels after leptin or combined with baclofen injection in fasted mice. We found that 8 mg/kg baclofen injection significantly increased serum insulin levels, indicating that GABA system might be synergic with insulin regulation in dietary metabolic.
In addition, some pharmacological evidence showed the role of GABA-A receptor played in feeding control. Infusion of a nonspecific GABA-A receptor agonist muscimol to PVN elicit feeding behaviour. 11 However, some findings showed that administration of GABA-A antagonist into the LH promoted eating, while the agonist of GABA-A receptor suppressed feeding. 13 Intracerebroventricular administration of antagonist of GABA-A receptor bicuculline increased eating and drinking behaviours. 14 Systematic administration of benzodiazepine partial receptor agonist provoked a hyperphagic response. 9,12 Loss of GABA-A receptor function of the PVN disrupted neuropeptide Y (NPY)-induced hyperphagia, which depend on GABAergic projections from non-POMC non-AgRP neurons. 43 Similarly, deletion of GABA-A receptor in the PVN reduced feeding. 39 However, few studies demonstrated the role of GABA-B played in hypothalamus feeding control.
Some limitations still exist in the present article. Firstly, do not reconcile with previous research showing significant alterations of POMC expression induced by leptin treatment or food deprivation, and we did not find significant changes of POMC expression in Western blot measurement and ELISA test. We propose that refeeding process after drug administration before killing might cause compensate expression of anorectic peptides. Secondly, our strategy still lacks in comparative experiments and analysis on GABA-A and GABA-B receptor and other antagonists or agonists of GABA-B receptor. Finally, recent findings have been shown that leptin signalling of astrocytes in hypothalamus also implicates in feeding regulation. 27 Thus, comprehensive consideration of the interactions of F I G U R E 7 Effect of leptin or combined with baclofen on plasma insulin, GABA and POMC levels. A, Effect of leptin or combined with baclofen on plasma insulin levels. B, A, Effect of leptin or combined with baclofen on plasma GABA levels. C, Effect of leptin or combined with baclofen on plasma POMC levels. Data are represented as the mean ± SEM; (*P < .05, **P < .01, ***P < .001. n = 5-6 mice per group) hormones, synaptic connections with different type cells in hypothalamus is important to understand leptin action in feeding control.
These issues need to be further investigated.
In summary, these findings indicate that leptin inhibit fasting-promoted food intake and activation of hypothalamic neurons. The inhibition of leptin on food intake of fasted mice can be prevented by pretreatment with GABA-B receptor agonist baclofen. We speculate that leptin might inhibit fasting-triggered activation of PAMM neurons via presynaptic GABA synaptic functions and this effect might be partially blocked by pharmacological activating GABA-B. The specific neurophysiological mechanisms of GABA implicated in the role of leptin played in feeding regulation of fasted mice need to be further investigated.

CO N FLI C T O F I NTE R E S T
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

AUTH O R CO NTR I B UTI O N S
TTG, FLZ, XXY and XHZ conceived the idea and performed experiments. WY, RJC and BJL wrote the manuscript. BJL revised the manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data are available on request.