Role of central kisspeptin and RFRP‐3 in energy metabolism in the male Wistar rat

Abstract Kisspeptin (Kp) and (Arg)(Phe) related peptide 3 (RFRP‐3) are two RF‐amides acting in the hypothalamus to control reproduction. In the past 10 years, it has become clear that, apart from their role in reproductive physiology, both neuropeptides are also involved in the control of food intake, as well as glucose and energy metabolism. To investigate further the neural mechanisms responsible for these metabolic actions, we assessed the effect of acute i.c.v. administration of Kp or RFRP‐3 in ad lib. fed male Wistar rats on feeding behaviour, glucose and energy metabolism, circulating hormones (luteinising hormone, testosterone, insulin and corticosterone) and hypothalamic neuronal activity. Kp increased plasma testosterone levels, had an anorexigenic effect and increased lipid catabolism, as attested by a decreased respiratory exchange ratio (RER). RFRP‐3 also increased plasma testosterone levels but did not modify food intake or energy metabolism. Both RF‐amides increased endogenous glucose production, yet with no change in plasma glucose levels, suggesting that these peptides provoke not only a release of hepatic glucose, but also a change in glucose utilisation. Finally, plasma insulin and corticosterone levels did not change after the RF‐amide treatment. The Kp effects were associated with an increased c‐Fos expression in the median preoptic area and a reduction in pro‐opiomelanocortin immunostaining in the arcuate nucleus. No effects on neuronal activation were found for RFRP‐3. Our results provide further evidence that Kp is not only a very potent hypothalamic activator of reproduction, but also part of the hypothalamic circuit controlling energy metabolism.

of RFRP-3 is still under debate because stimulatory, inhibitory or absent effects have been reported according to species, sex and seasons. 3,4 Reproduction is a very expensive process in terms of energetic needs, which makes it essential for mammals to match the timing of reproduction with an optimal energetic and metabolic status. Thus, it is not that surprising that, recently, Kp and RFRP-3 have also been linked to the control of food intake, body weight regulation and glucose homeostasis. [5][6][7] The scarce and scattered data so far point towards RFRP-3 having an orexigenic effect in different mammalian species [7][8][9][10][11] and Kp having an anorexigenic effect. 10,12 Regarding glucose homeostasis, it has been shown that female mice with a knockout (KO) for the Kp receptor Kiss1r are glucose intolerant, 5 whereas i.p. administration of RFRP-3 changed circulating glucose concentrations and insulin receptor and glucose transporter expression in testis and adipose tissue. 13 Interestingly, it has been found that one in three men with type 2 diabetes present detrimental effects on gonadal activity (hypogonadism) 14 and testosterone replacement has positive effects on metabolic syndrome survival rates. 15,16 Within the hypothalamus, the arcuate nucleus (ARC), a brain region well known to receive and integrate many metabolic signals from the periphery, shows a high expression of both Kp 17 and RFRP-3 18,19 receptors. The two main populations of neurones within the ARC that are responsible for the control of energy metabolism and glucose homeostasis are the orexigenic neuropeptide Y (NPY)/ agouti-related peptide (AGRP)-expressing neurones and the anorexigenic pro-opiomelanocortin (POMC)/cocaine-and amphetamineregulated transcript (CART)-expressing neurones. 20 In our previous studies in the seasonal Djungarian hamster we showed that central administration of Kp increased body weight as well as NPY-and POMC expression, whereas RFRP-3 increased food intake, body weight and circulating levels of leptin and insulin, without changing NPY-and POMC expression in the ARC. 7,11 In the present study, we tested the hypothesis that Kp and RFRP-3 would also affect energy metabolism in male Wistar rats. Therefore, we assessed the central effects of Kp and RFRP-3 on feeding behaviour, energy metabolism and glucose homeostasis in this species and revealed possible hypothalamic pathways involved in the reported metabolic effects.

| Animals
Adult male Wistar rats (Charles River, Sulzfeld, Germany) weighing 250-280 g at the start of the experiment were used in all experiments.
Animals were housed in individual cages in an enriched environment with a wooden stick under a 12:12 hour light/dark photocycle (lights on 7.00 am; =ZT0). Food (24% protein, 58% carbohydrate and 18% fat) (Teklad global diet 2918; Envigo, Indianapolis, IN, USA) and water were provided ad lib. After arrival, rats could adapt to the animal facility with constant temperature (21 ± 2°C) and humidity (50 ± 5%) for at least 1 week. All experimental procedures performed were approved by the Animal Ethics Committee of the Royal Dutch Academy of Arts and Sciences (KNAW, Amsterdam, The Netherlands) and were in accordance with the guidelines on animal experimentation of the Netherlands Institute for Neuroscience (NIN).

| Surgery
To infuse either RFRP-3 or Kp in the i.c.v. space of the central nervous system, a unilateral brain cannula (Plastic One , Dusseldorf, Germany) reaching the lateral ventricle was implanted. Surgery was conducted under anaesthesia consisting of an i.m. injection of a mix of ketamine (80 mg kg -1 ; Eurovet Animal Health, Bladel, The Netherlands) and xylazine (8 mg kg -1 ; Bayer Health Care, Mijdrecht, The Netherlands). The coordinates were defined using the rat brain atlas 21 as a reference: −0.8 mm anteroposterior, +2.0 mm lateral from bregma and −3.2 mm ventral from the dura. In some of the animals, silicon catheters were surgically implanted into the right jugular vein and the left carotid artery for i.v. infusion and blood sampling, respectively. 22 Brain cannula and catheters were fixed to the skull using dental cement. A cannula dummy was used to seal the guide cannula maintaining it open until the infusion. A metallic connector that could be attached to a chain swivel was added to the dental cement, which allowed us to execute the experiment without handling of the animals during the experiment. The animals received carprofen as a postoperative analgesic (2.5 mg kg -1 ; Zoetis, Parsippany-Troy Hills, NJ, USA). The rats were allowed to recover for at least 10 days after the surgery, with experiments only being started after they had reached their initial pre-operative body weight again.

| Experimental set-up for indirect calorimetry
Seven days after surgery, animals were single housed in metabolic cages (TSE, Bad Homburg, Germany) for three consecutive days.
Day 1 was aimed for habituation, day 2 for a baseline measurement and then on the morning of day 3 animals were i.c.v. injected (~ZT5) and the automatised measurements continued for 24 hours. Animals had ad lib. access to water and food from hanging bottles and baskets, respectively. Food and water intake, respiratory exchange ratio (RER), energy expenditure and locomotor activity were recorded continuously during these 3 days. On the afternoon of day 4, animals were moved back to their regular housing conditions.

| Intracerebroventricular peptide infusion
Every animal received a cross-treatment with vehicle (sterile NaCl 0.9%) and Kp (3 nmol/5 µL; Rat Kp10 sequence; ToCris Bioscience, Bristol, UK; n = 15) or RFRP-3 (50 or 250 pmol/5 µL; Rat RFRP-3 sequence; Caslo Laboratory, Lyngby, Denmark; 50 pmol, n = 8; 250 pmol, n = 7). The initial dose of 50 pmol RFRP-3 was based on the 100 ng dose of Johnson et al 8 ; when we found no effects with 50 pmol, we increased the dose to 250 pmol. Brain injections were performed at a rate of 1 µL min -1 and patency was corroborated by tracking the movement of a small air bubble. All animals were injected between ZT4.5 and ZT5.5. Animals were handled for 3-5 min day -1 for at least 4 days before each i.c.v. injection to habituate them to the procedure. Animals were allowed to recover for 7 days after each i.c.v. injection.

| Perfusion and peripheral tissue sampling
At the end of the experiment, rats were given a third i.c.v. injection with either vehicle or one of the RF-amides and killed 1 hour after under an overdose of i.p. injected pentobarbital. A blood sample was taken by heart puncture and then animals were perfused intracardially with 150 mL of NaCl 0.9%. Next, animals were perfused with 100 mL of formaldehyde 4%, and then brains were removed and post-fixed overnight. Brains were then transferred to 30% su- were performed using a cross-over design with at least 1 week of recovery in between. On the evening prior to the EGP evaluation, animals were attached to a counterbalanced swivel that allowed blood sampling without handling the animal. On the following morning, food was removed at ZT0 and the arterial and venous catheters were connected to a tubing line filled with heparinised (1%) saline ( Figure 1). At ZT3, a blood sample was taken for basal measurements and the tubing for the i.c.v. infusion was filled up, connected to the cannula injector and sealed to avoid leaking into the ventricular space prior to the start of the brain infusion. At ZT4, the D2 glucose infusion was started using a primed-continuous administration protocol, starting first with a 5-minute infusion at a rate of 3000 µL h -1 and then continued with a rate of 500 µL h -1 until the end of the experiment. Ninety minutes later at ZT5.5, when a steady-state was reached, 200-µL blood samples were taken every 10 minutes for a total of three samples to calculate the basal level of EGP before the start of the brain infusion. Finally, at ZT6, a primed i.c.v. infusion was started. Either Kp (0.6 nmol µL -1 ), RFRP-3 (50 pmol µL -1 ) or vehicle (NaCl 0.9%) were infused at a rate of 1 µL min -1 for the first 5 minutes, which was then decreased to 5 µL h -1 for the remainder of the experiment. Blood samples were taken every 20 minutes over 2 hours to calculate the change in EGP during the brain infusion.
After 2 hours, the i.c.v. and D2-glucose infusions and blood sampling were stopped and all external tubing was removed. Animals were immediately returned to ad lib. water and food access. At the end of the experiment, animals were killed, then fresh brains were collected, frozen and sliced with a cryostat to verify the cannula placement. and testosterone levels were measured using isotope dilution-liquid chromatography-tandem mass spectrometry. 25

| Immunostaining
Brain slices containing hypothalamic nuclei were selected using bregma as reference according to the rat brain atlas from Paxinos and Watson (Table 1). 21  For all experiments, the specificity of the first antibody was assessed by verifying that removal of the primary antibody resulted in an absence of immunostaining. In addition, the specificity of the anti-POMC was verified by pre-absorption controls on ARC brain sections containing POMC neurones, where staining was abol- ished. 26,27 The specificity of the Santa Cruz rabbit c-Fos antibody

| Statistical analysis
Only data from animals with correct cannula placements and injections corroborated for patency were included in the data analysis.
All data showed normal distribution and homogeneous variance ac-

| Experiment 1: Indirect calorimetry
Only the animals that successfully received both i.c

| Central Kp injection decreases food intake
Rats injected with 3 nmol Kp exhibited a decrease in 24-hour food intake compared to the previous baseline day, as well as compared to their vehicle treatment ( Figure 3A-C). Both the comparison "Kp i.c.v.
vs vehicle i.c.v." and "Kp i.c.v. vs Kp baseline" revealed a significant effect of Treatment, respectively P = 0.0006 and P = 0.0059 (Table 2).
Water intake was not changed after Kp injection ( Figure 3D-F and Table 2). By contrast, neither the 50 pmol (Table S1; see also Supporting information, Figure S1), nor 250 pmol (Table S2; see also Supporting information, Figure S2) i.c.v. injections of RFRP-3 significantly changed food or water intake. In addition, when the two RFRP-3 experiments were combined (n = 9), no significant effects on food or water intake were found (see Supporting information, Figure   S3 and Table S3).

| Central Kp injection decreases the RER
The RER value indicates the main fuel source utilised by the body for energy production. This ratio fluctuates over the daily cycle and is close to 1.0 during the dark (ie, feeding) phase when mainly  (Table 2). In addition, the 24-hour mean RER level was decreased in the Kp-i.c.v. group ( Figure 4B). By contrast, energy expenditure was not significantly affected by Kp ( Figure 4C). Locomotor activity showed a significant Treatment effect in the "Kp i.c.v. vs Kp baseline" comparison (P = 0.0069) ( Figure 4A and Table 2), but no differences in 24-hour total activity were detected ( Figure 4A). None of the doses of RFRP-3, either separately or combined, significantly changed RER, locomotor activity or energy expenditure (see Supporting information, Figures   S4 and S5 and Tables S1-S3).  (Table 3).

| Effects of RF-amide infusion on LH, testosterone, insulin and corticosterone secretion
The i.c.v. infusion of Kp increased plasma LH and testosterone concentrations, showing significant Treatment and Interaction effects ( Figure 5A,C and Table 4). Post-hoc analysis revealed that mean plasma levels of LH had already increased at 20 minutes (ie, in the first sample after the start of the i.c.v. infusion), but a statistically significant difference was only reached at 80 minutes ( Figure 5A).
Plasma testosterone levels showed statistically significant differences compared to the vehicle group 60, 100 and 120 minutes after the start of the i.c.v. infusion ( Figure 5C). The i.c.v. RFRP-3 infusions did not show any statistically significant difference in plasma LH or testosterone levels ( Figure 5B,D and Table 4). Both RF-amides did not have any significant effects on either on plasma corticosterone or insulin levels (Table 4 and Figure 6).

| Glycaemia and EGP after RF-amide i.c.v. infusion
Both RF-amides did not result in any significant changes in blood glucose levels, but the i.c.v. administration of Kp resulted in a significant increase of EGP as attested by a significant Interaction effect (P = 0.0385). Post-hoc analysis showed that this increase started 20 minutes after the start of the i.c.v. infusion ( Figure 7C and TA B L E 2 Effects of i.c.v. kisspeptin injection (3 nmol) on metabolic outcomes EGP as indicated by the borderline significant Interaction effect (P = 0.050) ( Figure 7D and Table 4).

| Central targets of i.c.v. Kp and RFRP-3
At the end of the Experiment 1, rats were given a third i.c.v. injection with either vehicle or one of the RF-amides and, 1 hour later, animals were perfused and perfusion fixed brains processed for immunostaining.

| Activation of the median preoptic nucleus (MnPO) in response to Kp
From all the brain regions analysed for c-Fos immunoreactivity

| POMC neurones respond to Kp
POMC expression was analysed in the ARC of Kp-and RFRP-3treated rats (Figure 9) because most of the POMC neurones have been reported to express RF-amide receptors. 17 Kp induced a significant decrease in the number of POMC-immunoreactive cells in the posterior part of ARC (P = 0.024) ( Figure 9C) . Also, total POMC immunoreactivity in the ARC was reduced, which was mainly a result of a decrease in its posterior part (P = 0.004) ( Figure 9E). The number of POMC cells expressing c-Fos was not modified ( Figure 9G).

RFRP-3 injections had no significant effect on ARC c-Fos or POMC
expression ( Figure 9B,D,F,H). Although the currently reported anorexic effect of central Kp in ad libitum fed rats confirms earlier observations in mice, rats and Jerboas, 10,12,36 in the very first studies, no significant effects on feeding behaviour were found, 31 ΔEndogenous glucose production (µmol kg·min -1 ) To elucidate the neuronal targets that might be involved in the anorexigenic effect of Kp, we investigated the activity of POMC neurones because they are the main neuronal population in the ARC expressing the Kiss1R 17 and are very well characterised for their inhibitory role on food intake and body weight. 40 We found that  53 In addition, Kp is also known to change the expression of other (an)orexigenic peptides, such as brain-derived neurotrophic factor, melanin-concentrating hormone , nesfatin-1 and oxytocin. 36,54,55

ACK N OWLED G EM ENTS
We thank Rainier Epping for husbandry of the experimental animals. This study was funded by the Agence National de la Recherche (ANR-13-BSV1-001) and the NeuroTime Erasmus+ program of the European Commission.

CO N FLI C T O F I NTE R E S T S
The authors declare that they have no conflicts of interest.

PE E R R E V I E W
The peer review history for this article is available at https://publo ns.com/publo n/10.1111/jne.12973.

DATA AVA I L A B I L I T Y
The data that support the findings of this study are available from the corresponding author upon reasonable request.