Prolactin-releasing peptide (PrRP) reduces food intake and body weight and modifies body temperature when administered centrally in rats, suggesting a role in energy homeostasis. However, the mediators of PrRP's actions are unknown. The present study, therefore, first examined the possible involvement of the anorectic neuropeptides corticotropin-releasing hormone (CRH) and the melanocortins (e.g., α-melanocyte-stimulating hormone) in PrRP's effects on food intake and core body temperature and, second, determined if PrRP affects energy expenditure by measuring oxygen consumption (V̇o2). Intracerebroventricular injection of PrRP (4 nmol) to 24-h-fasted male Sprague-Dawley rats decreased food intake and modified body temperature. Blockade of central CRH receptors by intracerebroventricular coadministration of the CRH receptor antagonist astressin (20 μg) reversed the PrRP-induced reduction in feeding. However, astressin's effect on PrRP-induced changes in body temperature was complicated because the antagonist itself caused a slight rise in body temperature. In contrast, intracerebroventricular coadministration of the melanocortin receptor-3/4 antagonist SHU-9119 (0.1 nmol) had no effect on any of PrRP's actions. Finally, intracerebroventricular injection of PrRP (4 nmol) caused a significantly greater V̇o2 over a 3-h test period compared with vehicle-treated rats. These results show that the anorectic actions of PrRP are mediated by central CRH receptors but not by melanocortin receptors-3/4 and that PrRP can modify V̇o2.
- food intake
- oxygen consumption
- body temperature
prolactin-releasing peptide (PrRP) was first identified to be the endogenous ligand for the human orphan receptor GPR10/hGR3 (or UHR-1 in the rat) and was reported to display specific prolactin-releasing properties (15). PrRP mRNA and immunoreactive cell bodies are located exclusively in three brain regions: the dorsomedial hypothalamus (DMH) and in the nucleus tractus solitarius (NTS) and the ventrolateral medulla of the brain stem (2, 7, 18, 26, 30, 33, 39, 60). UHR-1 receptor (PrRP-R) mRNA is also found in the rat brain and is expressed most highly in the reticular nucleus of the thalamus, the periventricular hypothalamic nucleus, paraventricular hypothalamic nucleus (PVN), DMH, NTS, and the area postrema (7, 19, 22, 39). PrRP-immunoreactive fibers are detected in several brain regions that express PrRP-R mRNA, including the PVN and the periventricular hypothalamic nucleus (18, 30, 60).
We have proposed a role for PrRP in energy homeostasis because mRNA for the peptide is reduced, typically for an anorectic peptide, in states of negative energy balance (fasting and lactation) and in the obese Zucker rat (5, 22). Intracerebroventricular administration of PrRP reduces both fasting-induced and spontaneous feeding, and body weight gain in rats (5, 22, 23, 47). Furthermore, PrRP does not appear to affect feeding via nonspecific or nonhomeostatic actions, as it does not support conditioned taste aversion or disrupt the behavioral satiety sequence (23). Pair-feeding experiments suggest that the reduction in body weight after intracerebroventricular injection of PrRP is not due entirely to its anorectic actions (22), implying that PrRP may have additional effects, possibly on energy expenditure. In support of this idea, PrRP affects core body temperature, and hyperthermia is observed typically 2-8 h after intracerebroventricular injection (5, 22). However, an increase in body temperature is only an indicator that PrRP is increasing energy expenditure, and more direct evidence based on oxygen consumption (V̇o2) needs to be determined.
The mechanisms of PrRP's action on appetite and body temperature are unknown. It is unlikely that prolactin itself is a mediator as several groups have demonstrated that PrRP does not, or can only weakly, cause prolactin release both in vivo and in vitro (19, 32, 42-44, 48, 54). One possible downstream mediator is corticotropin-releasing hormone (CRH). Like PrRP, central administration of CRH reduces food intake and body weight (see Ref. 13) and modifies energy expenditure in rodents (see Refs. 3, 12, 41). PrRP-immunoreactive fibers appear to make synaptic contact with CRH-containing cell bodies in the rat PVN, providing morphological evidence for a relationship between the two neuropeptides (31). Furthermore, intracerebroventricular injection of PrRP increases neuronal activation (measured by induction of c-Fos expression) in the PVN of the rat (23), including in ∼80% of CRH-containing neurons (31), and in vitro PrRP increases hypothalamic CRH release (46). In addition, central administration of PrRP in rats stimulates plasma levels of ACTH, an effect that is dependent on CRH receptor activation (31, 46), and corticosterone (42). In view of the above, we hypothesized that the actions of PrRP on energy balance are mediated via CRH receptors.
Therefore, the main aim of this study was to test the hypothesis that central CRH receptors mediate the effects of PrRP on food intake and core body temperature. Another anorectic pathway, the melanocortin system, was considered because it partially mediates the effects of modulators such as leptin (11, 16, 45, 49) and because the acute actions of leptin and PrRP are very similar (5). Additionally, the effect of PrRP on energy metabolism was tested more directly by measuring V̇o2 using closed-circuit calorimetry.
Materials. Rat PrRP (Peptide Institute, Osaka, Japan), the melanocortin receptor-3/4 antagonist SHU-9119 [acetyl-(Nle4, Asp5, D-2-Nal7, Lys10)-cyclo-α-MSH(4-10)amide; Bachem, Saffron Walden, UK], and the CRH receptor antagonist astressin (Bachem) were all dissolved in 0.9% sterile saline.
Animals and surgery. Male Sprague-Dawley rats (body wt 250-300 g, Charles River, Sandwich, UK) were used in all experiments and were housed at a constant ambient temperature of 21 ± 1°C on a 12:12-h light-dark cycle (lights on from 0800-2000). Standard pelleted rat chow (Beekay International, Hull, UK) and tap water were provided ad libitum unless stated otherwise. All procedures conformed to the requirements of the United Kingdom Animals Scientific Procedures Act of 1986. Five to seven days before intracerebroventricular injections, animals were anesthetized with 2.5-3% halothane (Fluothane; AstraZeneca, Macclesfield, UK) in oxygen, and a guide cannula was inserted stereotaxically into the lateral ventricle [0.8 mm posterior, 1.5 mm lateral and 3.5 mm ventral (1 mm above injection site) to bregma], according to the atlas of Paxinos and Watson (35), and the placement was checked histologically postexperiment.
Measurement of food intake and core body temperature. To allow remote measurement of core body temperature in undisturbed animals, radiotransmitters (TA10TA-F40, Data Sciences, Minneapolis, MN) were implanted into the peritoneum at the same time as lateral ventricle cannulations.
A week after surgery and 24 h before the start of the experiment, animals were housed individually, without food, in cages over receiver pads that monitored output frequency (temperature-dependent radio signals) emitted from the transmitters. The signals were then converted (to °C) via a peripheral processor (BCM 100; Data Sciences).
Intracerebroventricular injections were performed in the conscious, unrestrained, 24-h-fasted animals, 2 h after the beginning of the light phase (i.e., 1000). Immediately after injections, a preweighed amount of food was presented to the animals. Food consumption was measured 1 and 2 h later, and core body temperature was monitored continuously over an 8-h period.
Measurement of V̇o2. V̇o2 was determined in closed-circuit respirometers maintained at the thermoneutral temperature for rats (29°C). The system allows eight rats to be tested individually at one time (52). All animals were accustomed to the respirometers and procedures on two occasions the week before the experiments. One week after lateral ventricle cannulation, animals were housed individually, without food, in the respirometers, and V̇o2 was then recorded every 5 min. After a 90-min measurement of baseline V̇o2, animals were taken out and lightly restrained, and intracerebroventricular injections were performed ∼2 h after the beginning of the light phase (i.e., 1000). After injections, rats were placed back into the respirometers, and measurement of V̇o2 was continued for a further 3 h.
Experiment 1: effect of a CRH receptor antagonist on PrRP actions on food intake and core body temperature. Groups of rats (n = 10-12) were infused intracerebroventricularly with or without PrRP (4 nmol in 2 μl) and astressin (20 μg in 2 μl). The total volume injected was always made up to 4 μl with saline vehicle. The dose of PrRP was selected on the basis a previous studies demonstrating a hyperthermic and anorectic action in ad libitum-fed or fasted rats (5, 22). In addition, intracerebroventricular injections of astressin, at doses between 10 and 100 μg, have been shown to inhibit CRH-induced actions, including effects on food intake (1, 20).
Experiment 2: effect of a melanocortin receptor-3/4 antagonist on PrRP actions on food intake and core body temperature. To assess the effect of a melanocortin receptor antagonist on PrRP's actions, separate groups of rats (n = 15-18) were treated as in experiment 1 but with the astressin being replaced by SHU-9119 (0.1 nmol in 2 μl). This dose of SHU-9119 has been shown previously to inhibit reductions in feeding induced by intracerebroventricular injection of the melanocortin receptor agonist MTII in rats but does not effect feeding alone (24).
Experiment 3: effect of a PrRP on core body temperature in satiated rats. To investigate the actions of PrRP on core body temperature in the absence of possible confounding effects of feeding and activity (seen in fasted rats; experiments 1 and 2), PrRP was given to nonfasted satiated rats. Intracerebroventricular injection of PrRP (4 nmol in 2 μl, n = 8) or vehicle (2 μl saline, n = 8) was administered to satiated rats at 1000. Core body temperature was monitored continuously for 8 h, and food intake was measured at 6 h.
Experiment 4: effect of PrRP on V̇O2. To test the thermogenic effect of PrRP, rats were injected intracerebroventricularly with either vehicle (2 μl sterile saline; n = 13) or PrRP (4 nmol in 2 μl; n = 13). On another occasion a separate group of rats (n = 5 per group) was administered with PrRP (4 nmol in 2 μl) with or without astressin (20 μg in 2 μl).
Data and statistical analyses. All data are presented as means ± SE. Body temperatures were plotted as the mean change (Δ) from the point of injection (time 0) and were analyzed by calculating the integrated temperature response for 2-8 h after injection [area under the curve (AUC), °C·h] for each animal by the trapezoidal method. Average AUC values were then determined for each treatment group. V̇o2 is expressed as milliliters oxygen per minute per kilogram metabolic body size (i.e., ml·min-1·kg-0.75). Thirty-minute averages of V̇o2 were calculated, and the time course of the change (Δ) inV̇o2 was determined by comparing the mean of the last 30 min of the baseline readings with the mean values obtained for the postinjection period. The mean change in V̇o2 from baseline over the 3-h period was calculated for statistical analysis.
Statistical comparisons were performed using an unpaired Student's t-test for two groups (experiments 3 and 4; see Fig. 5B), and all other experiments were analyzed using a one-way ANOVA, followed by a Tukey-Kramer comparisons post hoc test. Statistical significance assumed a two-tailed probability with a P < 0.05.
Experiment 1: effect of a CRH receptor antagonist on PrRP actions on food intake and core body temperature. Central injection of the CRH receptor antagonist astressin (20 μg) alone had no effect on food intake at the 0- to 1-h and 1- to 2-h time points compared with vehicle-treated animals (Fig. 1). Administration of PrRP (4 nmol) suppressed food intake at 1 h (67%) after injection when compared with vehicle-injected animals, but there was no difference in the second hour. Astressin reversed the effects of PrRP on food intake at 1 h, and values for the PrRP plus astressin coinjected group were not significantly different from control animals.
Remote radiotelemetry revealed that fasted animals, injected intracerebroventricularly with vehicle, reliably displayed an initial hyperthermia (over ∼2 h) probably due to increased feeding activity when food was returned. Astressin alone had no significant effect on core body temperature during the experimental period, although the graphs suggest that there was a steady, but nonsignificant, rise (over 2-8 h) in body temperature in this group of animals (AUC for 2-8 h: vehicle, 2.82 ± 0.36°C·h vs. astressin, 4.06 ± 0.83°C·h, P > 0.05, Fig. 2A). Injection of PrRP caused an initial hypothermic response followed by a significant increase in core body temperature over 2-8 h compared with vehicle injection (AUC for 2-8 h: vehicle, 2.82 ± 0.36 °C·h vs. PrRP, 5.58 ± 0.88 °C·h, P < 0.05, Fig. 2B). Coinjection of astressin prevented the initial hypothermic response induced by PrRP (Fig. 2C). However, core body temperature observed during the 2-8 h after PrRP plus astressin was also significantly greater compared with vehicle injection (AUC for 2-8 h: vehicle, 2.82 ± 0.36 °C·h vs. PrRP + astressin, 5.84 ± 0.99 °C·h, P < 0.05, Fig. 2C). There was no significant difference between the AUC for the PrRP alone and PrRP plus astressin groups (P > 0.05).
Experiment 2: effect of a melanocortin receptor-3/4 antagonist on PrRP actions on food intake and core body temperature. Intracerebroventricular injection of SHU-9119 alone had no effect on 0- to 1-h or 1- to 2-h fast-induced feeding (Fig. 3). PrRP caused a significant reduction in 1-h (56%) food intake compared with vehicle-treated rats, and as reported for experiment 1, there was no effect on feeding during 1-2 h after injection (Fig. 3). However, coinjection of SHU-9119 had no effect on the decrease in food intake observed after injection of PrRP. Food intake in this group of animals was reduced by 51% at 1 h vs. vehicle-treated groups and was not significantly different when compared with PrRP injection alone (Fig. 3). Food intake was also measured for up to 8 h after injections, and that confirmed that SHU-9119 did not have any delayed effect on the PrRP response (data not shown).
SHU-9119 alone also had no significant effect on the change in body temperature compared with vehicle-treated animals (AUC for 2-8 h: vehicle, 3.04 ± 0.27 °C·h vs. SHU-9119, 4.41 ± 0.49 °C·h, P > 0.05, Fig. 4A). As expected, PrRP administration caused a characteristic drop in core body temperature (Fig. 4B). This hypothermic phase was followed by a period where body temperature settled above that of vehicle-injected animals but failed to reach statistical significance in this experiment using post hoc analysis (AUC for 2-8 h: vehicle, 3.04 ± 0.27 °C·h vs. PrRP, 4.64 ± 0.56 °C·h, P > 0.05, Fig. 4B). Consistent with the effects on feeding, SHU-9119 had no effect on the temperature response induced by PrRP, as the hypothermic phase was still observed. In addition, core body temperature for the 2- to 8-h period was not significantly different compared with PrRP alone (AUC for 2-8 h: PrRP, 4.64 ± 0.56 °C·h vs. PrRP + SHU-9119, 5.44 ± 0.87 °C·h, P > 0.05) but was significantly increased compared with vehicle-treated animals (AUC for 2-8 h: vehicle, 3.04 ± 0.27 °C·h vs. PrRP + SHU-9119, 5.44 ± 0.87 °C·h, P < 0.05, Fig. 4C).
Experiment 3: effect of a PrRP on core body temperature in satiated rats. Normally fed rats did not eat significant amounts of food during the light phase (0800-2000), and there was no difference in food intake between vehicle and PrRP-treated rats at 6 h postinjection (0- to 6-h food intake: vehicle, 1.4 ± 0.5 g vs. PrRP, 1.7 ± 0.6 g, P > 0.05). PrRP injection into satiated (nonfasted) rats caused the typical effect on core body temperature as reported previously. An initial rapid hypothermia was followed by a prolonged period of hyperthermia (AUC for 2-8 h: vehicle, 2.01 ± 0.33 °C·h vs. PrRP, 4.53 ± 0.81 °C·h, P < 0.05).
Experiment 4: effect of PrRP on V̇O2. The effect of a single intracerebroventricular dose of PrRP (4 nmol) on V̇o2 is shown in Fig. 5A as the change (Δ) in V̇o2 from baseline. The V̇o2 in vehicle-treated animals declined slowly over the 3-h test period. This decline is normal and probably reflects a decrease in the diet-induced thermogenesis resulting from the last few nocturnal meals consumed just before the start of the light phase (0800). This decline, however, was less in animals that received PrRP, and V̇o2 was significantly greater than in vehicle-treated animals for the 3-h measurement period. Figure 5B shows the mean change from baseline over 3 h, with the vehicle-treated animals showing an average 14% decrease of 1.87 ± 0.18 ml·min-1·kg-0.75, which was significantly greater (P < 0.001) than the 4% decrease (0.63 ± 0.16 ml·min-1·kg-0.75) observed after PrRP treatment.
In a separate experiment, 4 nmol PrRP caused a similar effect on the 3-h change in V̇o2 compared with controls (vehicle, 10% decrease of 1.39 ± 0.12 ml·min-1·kg-0.75; vs. PrRP, 4% decrease of 0.61 ± 0.19 ml·min-1·kg-0.75; P < 0.01, Fig. 5C). However, 20 μg astressin also affected 3-h V̇o2 when administered alone (5% decrease of 0.63 ± 0.15 ml·min-1·kg-0.75; P < 0.01 compared with control group), and although coadministration of astressin partially inhibited the effects of PrRP (7% decrease of 0.93 ± 0.10 ml·min-1·kg-0.75; P > 0.05 compared with other groups), this was not significantly different from the PrRP alone-treated animals.
We have shown previously that intracerebroventricular injection of PrRP reduces food intake and body weight and modifies core body temperature in fasted and ad libitum-fed rats (5, 22, 23). The acute effects of PrRP on food intake, which occur within 1 h of injection, have been confirmed by Seal and coworkers (47) and here in the present study. The CRH receptor antagonist astressin but not the melanocortin-3/4 receptor antagonist SHU-9119 blocked the effect of PrRP on food intake. Agonists of melanocortin receptor-3/4 (e.g., α-melanocyte stimulating hormone, MTII) suppress food intake in rodents (6, 55). The dose of the antagonist SHU-9119 used in the present experiment (0.1 nmol) is effective as it blocks the anorectic actions of an equimolar injection of the melanocortin receptor agonist MTII (24). Furthermore, it has its effect when it is coadministered at the same time with either MTII, leptin, or d-fenfluramine (14, 24, 45). Because higher doses of SHU-9119 can initiate food intake at later time points (9, 40), we confirmed also that it did not reverse the effect of PrRP on cumulative food intake over an 8-h period.
The effect of PrRP on core body temperature is twofold, as after an initial, brief hypothermic period a hyperthermic phase is evident. Data presented here and previously (4, 5, 22) demonstrate that the temperature profile observed after PrRP is the same regardless of when injections are performed (light vs. dark phase) and whether the animals are eating or not. Because astressin reversed both the initial drop in body temperature and the reduction in food intake observed after PrRP injection, it is possible that the hypothermic response is responsible for the ensuing effects on feeding. However, we observe similar acute hypothermic phases after intracerebroventricular injection of several other neuropeptides, which are conversely orexigenic, that is, they stimulate feeding (e.g., growth hormone releasing peptide-6 and ghrelin; Ref. 25). Thus the acute hypothermia is neither sufficient to cause a reduction in feeding nor is it a result of reduced feeding. The subsequent increase in core body temperature was not significantly inhibited by astressin, although these results are complicated by the fact that astressin itself has a thermogenic effect.
Increases in core body temperature may be indicative of heightened energy expenditure, and we now provide more direct evidence for this, as PrRP increased V̇o2, at least acutely. These results therefore demonstrate that PrRP's effects on body weight are probably brought about by dual actions on food intake and energy expenditure. Astressin itself increased V̇o2 significantly postinjection, and this correlated with the increase in core body temperature. The actions of astressin on these parameters highlight a need for caution in interpretation of experiments involving this compound and the role of CRH receptors in energy balance. Thus the involvement of CRH receptors in mediating the metabolic response to PrRP may need further investigation to establish whether the PrRP-induced rise in energy expenditure is mediated by the same mechanism as the anorectic response.
There are several CRH receptor subtypes described to date: CRH receptor type 1 and CRH receptor type 2, which can exist as several splice variants (see Refs. 37, 56). The CRH receptor subtype involved in PrRP's anorectic effects cannot be ascertained from this study, as the antagonist astressin does not distinguish between the two (10, 36). Likewise, the actions of PrRP cannot be conclusively ascribed to CRH because there are several other endogenous ligands that act at CRH receptors. These include urocortin (58) and the recently identified urocortin II (17, 38) and urocortin III (17, 27). Each of these neuropeptides affect feeding in rodents (17, 38, 51) and, thus, could potentially be involved in PrRP's anorectic actions. This stated, CRH's involvement is supported by the observation that 80% of CRH-containing neurons in the PVN are activated by intracerebroventricular injection of PrRP (31), and in vitro PrRP causes hypothalamic CRH release (46). However, actions of PrRP in the PVN may not be direct, because although PrRP-R mRNA is located in this area (19, 22, 39) and PrRP-immunoreactive fibers make synaptic contact with CRH cell bodies in the PVN (31), a recent study suggests that only a small percentage of PrRP-R-expressing cells in the PVN contain CRH (28). Thus the described effects of PrRP in the PVN may be indirect due to activation of neuronal pathways that connect with the PVN.
Both CRH and urocortin induce a range of behaviors, such as anxiety and hyperactivity, when administered centrally (21, 50). Thus the significance of CRH and the urocortins in the physiological regulation of food intake is still debatable. Our own data demonstrate that, in addition to the PVN, increases in c-Fos after intracerebroventricular PrRP injection are observed in the amygdala (23), which is also a major site of CRH synthesis (53). Although we propose a role for CRH receptors in the actions of PrRP on feeding, the caveat noted above might indicate that PrRP can affect appetite, possibly by causing anxiety. However, we have demonstrated that PrRP, at doses that decrease feeding, enhances the behavioral satiety sequence and does not cause conditioned taste aversion (23) or affect water intake (22, 43). In addition, Seal and colleagues (47) have demonstrated that PrRP, at comparable doses, does not affect overt behaviors significantly. We have demonstrated recently that repeated doses of PrRP lead to tolerance to the anorectic action (4). However, we do note that one study failed to reproduce the acute effects of PrRP on feeding (59) at doses used in our own and the experiments of Seal and coworkers (47). These workers did report an effect with repeated central injections of PrRP at higher dose.
The acute effects of PrRP on feeding and body temperature are similar to those reported after central administration of leptin (5, 29), and leptin actions on food intake are partially blocked by either CRH receptor antagonists (8, 57) or an anti-CRH antibody (34). However, in contrast to PrRP, leptin also functions through melanocortin receptors (11, 16, 45, 49). We have shown previously that leptin and PrRP have an additive interaction when coadministered (5), and, thus, CRH receptors may be a point of convergence for these two catabolic factors.
This work was supported by the Biotechnology and Biological Sciences Research Council and AstraZeneca.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
↵† Deceased 26 March 2001.
- Copyright © 2004 the American Physiological Society