Prokineticin 2 Is a Hypothalamic Neuropeptide That Potently Inhibits Food Intake

PK2 was purchased from Peprotech (London, U.K.), PK2 antibody was purchased from Santa Cruz Biotechnology (Santa Cruz, CA), and control IgG was generated from a random sequence peptide epitope (21).

Adult male Wistar rats weighing 200–250 g and adult male C57BL/6 mice weighing 20–25 g (Harlan, Bicester, U.K.) were maintained in individual cages (width 24.5 cm, length 41.5 cm, and depth 18.5 cm) at 21–23°C with a 12-h light/dark cycle with ad libitum access to food (RM1 diet; SDS, Witham, U.K.) and water unless specified in procedure protocol.

Studies were performed in C57BL/6 mice when they had developed diet-induced obesity and their body weight was stable (see supplementary Methods 1, available in an online appendix at http://diabetes.diabetesjournals.org/cgi/content/full/db09-1198/DC1). Animal studies were approved under British Home Office Scientific Procedures Act 1986.

Three groups of rats (ad libitum fed, fasted for 12 or 24 h, n = 24) were killed at the beginning of the light phase. Whole hypothalami were dissected and total RNA was extracted using an RNAqueous-4PCR kit (Applied Biosystems, Austin, TX) according to the manufacturer's protocol. cDNA was synthesized from 0.5 μg of RNA using a High Capacity cDNA Reverse Transcription Kit (Applied Biosystems) according to the manufacturer's protocol. Quantitative real-time PCR was performed in triplicate using the following primer sets: 18S RNA primer assay ID:4310893E, PK2 primer assay ID:Rn00593837_m1.

ICV cannulation was performed as previously described (22). Male rats (n = 10–12/group) were injected intracerebroventricularly with saline or PK2 at doses of 0.005, 0.015, 0.05, and 0.15 nmol/rat at the beginning of the dark phase. All rats were injected within a single 20-min period with individual times of injection noted. The subsequent measurements of food intake are relative to the recorded injection time. Food intake was measured 1, 2, 4, 8, and 24 h after injection for these studies.

In a second study, rats (n = 10–12/group) were injected intracerebroventricularly with saline or PK2 at doses of 0.15, 0.5, and 1.5 nmol/rat at the beginning of the dark phase. To determine whether PK2 has anorectic effects in animals refeeding after a fast, rats (n = 10–12/group) fasted for 24 h were injected intracerebroventricularly in the early light phase with saline or PK2 at doses of 0.15, 1.5, or 4.5 nmol/rat. Food intake was measured 1, 2, 4, 8, and 24 h after injection for these studies.

Rats (n = 10–12/group) were injected intracerebroventricularly with saline or PK2 1.5 nmol/rat at the beginning of the dark phase. This dose of PK2 was used because it potently reduced food intake. The ambulatory activity of each animal was measured simultaneously using the optical beam technique (Opto M3; Columbus Instruments, Columbus, OH) (supplementary Methods 2).

Adult male Wistar rats weighing 200–250 g (n = 10–12/group) were injected intracerebroventricularly with saline or PK2 1.5 nmol/rat and behavioral patterns monitored continuously for 120 min after injection. Behavior was classified into eight categories: feeding, drinking, grooming, burrowing, rearing, locomotion, head down, and sleeping, as previously described (23). Abnormal behavior was defined by a significant increase in locomotor activity, rearing, head down or burrowing, or reduced sleeping or grooming, as previously described (23,24).

Rats (n = 10–12/group) were injected intracerebroventricularly with saline or PK2 1.5 nmol/rat at the beginning of the dark phase and food was removed after injection but water was available ad libitum. Oxygen consumption was measured by indirect calorimetry using an open-circuit Oxymax system of the Comprehensive Lab Animal Monitoring System (Columbus Instruments) (25).

Rats (n = 10–12/group) were injected intracerebroventricularly with either control IgG or anti-PK2 antibody (10 or 30 pmol) in the early light phase. Food intake was measured 1, 2, 4, 8, and 24 h after injection.

Rats were cannulated into the lateral ventricle as previously described (26). PK2 (1.5 nmol/rat) or saline (n = 4/group) was injected into the lateral ventricle of ad libitum–fed rats over 1 min in the early light phase. Immunocytochemistry (ICC) for c-fos was performed on brain sections from animals as previously described (27). Total numbers of c-fos–positive cells were counted bilaterally in matched sections from hypothalamic nuclei.

Rats (200–250 g) had permanent indwelling, unilateral, 26-gauge stainless steel guide cannulae (Plastics One, Roanoke, VA) stereotactically implanted into the supraoptic nucleus (SON), arcuate nucleus (ARC), paraventricular nucleus (PVN), anterior hypothalamic area (AHA), ventromedial hypothalamus (VMH), dorsomedial nucleus (DMN), SCN, and lateral hypothalamic area (LHA) of the hypothalamus (coordinates listed in supplementary Table 1) (n = 16/nucleus), as previously described (28). The study was of a randomized crossover design. Each animal received both injections (saline or 0.025 nmol PK2) in a random order 4 days apart. Food intake was measured 1, 2, 4, 8, and 24 h after injection. The dose of PK2 was chosen based on previous studies that show that 10% of the effective ICV dose results in significant effects when directly administered into responsive hypothalamic nuclei and minimizes nonspecific activation (28). Cannula placement was determined at the end of the study by injection of Indian ink (28).

Adult male Wistar rats weighing 200–250 g were killed by decapitation and hypothalamic explants prepared as previously described (29). The hypothalami were incubated for 45 min in 600 μl artificial cerebrospinal fluid (aCSF; basal period). The tissues were then exposed to PK2 (10, 100, or 1,000 nmol/l) in 600 μl aCSF for 45 min (n = 9–12/treatment). Finally, the viability of the tissue was verified by a 45-min exposure to 56 mmol/l KCl. α-Melanocyte–stimulating hormone (α-MSH), cocaine and amphetamine–regulated transcript, thyrotropin-releasing hormone, corticotrophin-releasing hormone, neuropeptide Y, and agouti-related peptide (AgRP) in the aCSF were measured using established radioimmunoassay (22,30,,,–34).

Rats (n = 10–12/group) were intracerebroventricularly injected with either 1) saline, 2) AgRP (1 nmol/rat), 3) α-MSH (1 nmol/rat), 4) α-MSH (1 nmol/rat) and AgRP (1 nmol/rat), 5) PK2 (0.15 nmol/rat), or 6) PK2 (0.15 nmol/rat) and AgRP (1 nmol/rat) at the beginning of the dark phase. The doses of α-MSH and AgRP were chosen based on previous studies (35). Food intake was measured 1, 2, 4, 8, and 24 h after injection.

Animals (n = 5 per group) were injected into the lateral ventricle with 1.5 nmol PK2. In situ hybridization (ISH) for proopiomelanocortin (POMC) and ICC for c-fos were performed on sections including the ARC, as previously described (36,37). A riboprobe corresponding to nucleotides 307–795 of the POMC rat sequence (accession no. NM_139326) was used for ISH. Total numbers of positive cells per animal were counted from matched sections of the ARC, and colocalized cells were expressed as a percentage of the total number of POMC and c-fos neurons.

Rats (n = 10–12/group) were injected intraperitoneally with saline or PK2 at doses of 2.3, 7, or 20 nmol/kg at the beginning of the dark phase. Food intake was measured 1, 2, 4, 8, and 24 h after injection. Due to the limited availability of recombinant PK2, the effects of peripheral administration of PK2 on food intake were further characterized in mice. A similar study was conducted in groups of C57BL/6 mice (n = 10–12/group) injected with saline or PK2 at doses of 7, 20, 60, 180, or 540 nmol/kg.

Adult male C57BL/6 mice weighing 20–25 g (n = 10/group) were given twice daily intraperitoneal injections (early light phase and just prior to the dark phase) of either saline or PK2 180 nmol/kg for 5 days. Food intake was measured 4 h after the injection at the beginning of the light phase and 1 h after the injection just prior to the dark phase. Daily food intake and body weight were also measured.

A similar study was carried out in C57BL/6 diet-induced obese (DIO) mice that were randomized to 1) saline treatment with ad libitum access to food, 2) PK2 treatment (540 nmol/kg per injection) with ad libitum access to food, or 3) pair-fed group: saline treatment but food restricted to the daily median food consumed by the PK2-treated mice over the previous 24-h period (n = 10/group).

Data are shown as mean values ± SEM except c-fos and behavioral analysis data, which are presented as median and interquartile range. The studies of food intake and hypothalamic PK2 expression were analyzed using a one-way ANOVA, followed by post hoc Dunnett test except for the intranuclear food intake study, which was analyzed using the Holm Bonferroni test. Food intake data expressed as a change compared with saline-treated animals were analyzed using a one-way ANOVA, followed by post hoc least significant difference test. For the c-fos immunocytochemistry and dual ISH/ICC studies, the Mann-Whitney U test was used. Data from behavioral studies were compared using Kruskal-Wallis one-way ANOVA on ranks. For the CLAMS (Comprehensive Laboratory Animal Monitoring System) studies, the generalized estimating equation and the Mann-Whitney U test were used. In all cases, P < 0.05 was considered statistically significant.

The expression levels of hypothalamic neuropeptides that reduce food intake (e.g., α-MSH) are often elevated in states of positive energy balance and reduced in states of negative energy balance (38). Hypothalamic PK2 expression was significantly reduced by 45% in rats fasted for 12 or 24 h (Fig. 1). This is consistent with the hypothesis that PK2 is an endogenous anorectic hypothalamic neuropeptide.

ICV administration of PK2 caused a dose-dependent significant reduction in food intake that, at doses greater than 0.15 nmol/rat, produced an 85% reduction in food consumed in the first hour after injection (Fig. 2,A and B, supplementary Fig. 1, and supplementary Table 2A and B). When administered intracerebroventricularly in the early light phase to fasted rats, PK2 caused a similarly potent inhibition of food intake (Fig. 2,C and supplementary Table 2C). In addition, rats injected with 1.5 and 4.5 nmol PK2 had a 30% reduction in 24-h food intake compared with saline-injected rats (Fig. 2 D and supplementary Table 2C).

A reduction in food intake can be due to an indirect effect, for example, changes in locomotor activity or behavior (39). ICV administration of PK2 to ad libitum–fed rats did not significantly alter locomotor activity or result in abnormal behavior compared with saline-injected animals (Fig. 2,E and F and supplementary Table 3). However, consistent with an increase in satiety, ICV PK2 significantly reduced the number of feeding episodes (supplementary Table 3). Many regulators of food intake also regulate energy expenditure. However, this does not appear to be the case for PK2, because ICV administration of PK2 did not alter oxygen consumption, a surrogate for energy expenditure (Fig. 2 G). These data suggest that ICV administration of PK2 specifically reduces food intake for up to 24 h.

Rats injected with anti-PK2 antibody ate significantly more than rats injected with control IgG antibody, suggesting that endogenous PK2 may restrain appetite (Fig. 3 and supplementary Table 4).

ICV administration of PK2 resulted in a significant increase in c-fos immunoreactivity in the SON, ARC, PVN, and AHA (Fig. 4,A–F). No significant changes in c-fos expression were observed in the VMH, DMN, SCN, or LHA (Fig. 4 A and supplementary Fig. 2).

To establish which hypothalamic nuclei mediate PK2's anorectic effects, PK2 was administered into the hypothalamic nuclei showing c-fos activation after ICV administration of PK2. In addition, other nuclei expressing PKR2 were also injected with PK2 as negative controls. PK2 significantly reduced 0–1 h food intake in rats after administration into the SON, ARC, PVN, and AHA (Fig. 4,G), but there was no significant effect of PK2 after injection into the VMH, DMN, SCN, or LHA (Fig. 4 G).

PK2 significantly stimulated the release of α-MSH (Fig. 5 A) but did not alter the release of the other hypothalamic neuropeptides measured (supplementary Table 5).

We therefore hypothesized that ICV PK2 may mediate part of its anorectic effect via the melanocortin system. To further investigate the relationship between PK2 and the melanocortin system in vivo, we coadministered AgRP with PK2. In this paradigm, AgRP attenuated the effect of PK2, suggesting that PK2 may reduce food intake in part via the melanocortin system (Fig. 5,B). To confirm that the dose of AgRP used was antagonizing α-MSH, we demonstrated that ICV administration of α-MSH reduced 0–2 h food intake in rats, whereas coadministration of AgRP with the same dose of α-MSH abolished the anorectic effect of α-MSH (Fig. 5 B).

To determine whether ICV administration of PK2 resulted in activation of ARC POMC neurons (which produce α-MSH), we performed colocalization studies of c-fos and POMC after ICV administration of PK2. ICV administration of PK2 significantly increased the number of arcuate POMC-expressing neurons exhibiting c-fos immunoreactivity (Fig. 5,C and Table 1). Together these data suggest that PK2 may mediate part of its anorectic effect via the ARC melanocortin system.

A single intraperitoneal injection of PK2 (20 nmol/kg) reduced food intake by 45% in the first hour after injection but did not affect food intake at the other time points studied (Fig. 6 A and supplementary Table 6A).

Intraperitoneal injection of PK2 in mice reduced 0–1 h food intake at similar doses to those in rats (Fig. 6,B). Higher doses of PK2 in mice resulted in a further dose-dependent reduction in food intake for up to 24 h after injection (Fig. 6,B and supplementary Table 6B). The highest dose of PK2 (540 nmol/kg) administered produced a 20% reduction in 24-h food intake (Fig. 6 C and supplementary Table 6B).

Twice daily intraperitoneal injection of PK2 for 5 days in lean mice significantly decreased cumulative food intake (Fig. 7,A) and body weight compared with saline-injected controls (Fig. 7 B). The reduction in food intake after each single injection of PK2 was similar in magnitude throughout the study period, suggesting that PK2 remained equally potent after recurrent administration (supplementary Table 7).

DIO mice are commonly used as a rodent model of human obesity (40). Some anorectic factors, for example leptin, are ineffective in obese mice (41). To investigate whether DIO mice were sensitive to PK2, we administered PK2 twice daily by intraperitoneal injection for 5 days to DIO mice. PK2 significantly reduced cumulative food intake (Fig. 7,C) and body weight (Fig. 7,D) compared with saline-injected controls. DIO mice pair fed to the PK2-treated group lost a similar amount of body weight as the mice given PK2 (Fig. 7 D), suggesting the effect of PK2 on body weight was mediated predominantly via a reduction in food intake. As in lean mice, injection of PK2 was equally potent at reducing food intake on each of the 5 days of the study (supplementary Table 8). These data suggest that repeated administration of PK2 results in reduced food intake and body weight without tachyphylaxis in lean and obese mice.

The hypothalamus is a key center of the brain involved in appetite regulation. PK2 and its receptors are expressed in hypothalamic nuclei known to affect food intake, but the effect of PK2 on food intake has not previously been reported. Our data suggest that PK2 is a novel hypothalamic anorectic neuropeptide. It was important to determine whether the anorectic effects of PK2 were due to behavioral changes because PK2- or PKR2-null mice display abnormal circadian rhythms and locomotor activity (16,–18). ICV administration of PK2 potently inhibited food intake in rats up to 24 h after injection, but did not alter locomotor activity or cause abnormal behaviors. These data suggest the anorectic effects of PK2 are not secondary to effects on locomotor activity or other behavioral abnormalities.

To investigate the possibility that endogenous PK2 may affect appetite, we determined the effect of immunoblockade of endogenous hypothalamic PK2 in rats. ICV administration of PK2 antibody in the early light phase (when endogenous CNS PK2 levels are highest [14]) increased food intake in rats, suggesting that elevated CNS PK2 signaling may inhibit food intake in the light phase. However, this finding needs to be interpreted with caution because the increase in food intake after ICV administration of PK2 antibody was small. Conversely, CNS PK2 expression is at its lowest in the dark phase, and this reduction in PK2 inhibitory tone may therefore contribute to the nocturnal increase in food intake. Our results also show that hypothalamic PK2 mRNA expression was lower in fasted rats compared with ad libitum–fed rats, supporting the hypothesis that PK2 may act as an endogenous anorectic signal.

If PK2 is an endogenous inhibitor of food intake, one might expect PK2-null mice to be obese. Mice lacking PK2 have not been reported to have increased body weight compared with their wild-type littermates (8,16). However, PK2 is critical in CNS development and these mice have reduced voluntary and spontaneous locomotor activities, show a reduction in the time spent sleeping compared with wild-type mice (16), and have reduced fertility (11). These factors are likely to confound the effects of the lack of PK2 signaling on energy homeostasis in this mouse model. In addition, it is known that developmental compensation in embryonic knockout models of appetite-regulating factors can mask roles in energy homeostasis, for example, as has been suggested to occur with neuropeptide Y and AgRP (42). A postembryonic or conditional knockout model of PK2, or animal studies in which local hypothalamic PK2 expression is reduced, may help to further determine the role of endogenous PK2 in appetite regulation.

To investigate the hypothalamic sites that may mediate the anorectic effects of PK2, we used the induction of c-fos as a marker of neuronal activation (43). ICV administration of PK2 resulted in c-fos induction in the SON, ARC, PVN, and AHA, and direct injection of PK2 into each of these nuclei reduced food intake, suggesting that the anorectic effects of PK2 may be mediated directly via these nuclei. After ICV administration of PK2 and direct injection of PK2 into the VMH, DMN, SCN, or LHA, which all express prokineticin receptors, there was no induction of c-fos and PK2 did not significantly affect food intake. These data suggest that the anorectic effects of PK2 may be mediated via specific hypothalamic pathways.

ICV administration of PK2 induced c-fos expression in the ARC and direct injection of PK2 into the ARC reduced food intake, suggesting that ICV PK2 may mediate part of its anorectic effect via the ARC. PK2 may therefore mediate its anorectic effects via alteration in hypothalamic ARC neuropeptides. Our results show that PK2 increased the release of α-MSH from hypothalamic explants ex vivo and ICV coadministration of AgRP with PK2 attenuated the anorectic effects of PK2. Consistent with this, ICV administration of PK2 resulted in c-fos activation in ARC POMC neurons that produce α-MSH. Together these data suggest that PK2 may mediate part of its anorectic effect via the hypothalamic ARC melanocortin system.

To investigate whether peripheral administration of PK2 had anorectic effects, we investigated the effects of intraperitoneal administration of PK2 on food intake and body weight in rodents. Peripheral administration of PK2 acutely reduced food intake with similar efficacy in rats and mice. This led us to investigate whether repeated PK2 administration would result in a sustained reduction in food intake and body weight, since repeated administration of anorectic agents can cause tachyphylaxis (44), resulting in an attenuated effect after repeated administration. PK2 administration to lean or obese mice caused a similar reduction in food intake after each injection and resulted in a significant reduction in body weight, suggesting that tachyphylaxis to the anorectic effects of PK2 does not occur using this administration protocol.

Leptin reduces food intake and body weight in lean animals but is ineffective in obese animals (41). This may be due to differences between the appetite circuits of lean and obese animals (45) and/or the development of resistance to leptin in obese animals (46). We therefore investigated the effect of repeated intraperitoneal PK2 administration in DIO mice, which are commonly used as a model of human obesity (40). PK2 resulted in a similar reduction in food intake after each injection, and reduced cumulative food intake and body weight in DIO mice. Higher doses of PK2 were required to reduce food intake in DIO mice than in lean mice, but this resulted in greater weight loss in DIO mice compared with lean mice.

In both lean and DIO mice, the effect of PK2 on body weight is likely to be mediated predominantly via a reduction in food intake because control mice pair fed to PK2-injected mice lost a similar amount of weight as the PK2-injected group. This is consistent with our studies in rats in which ICV administration of PK2 reduced food intake without affecting energy expenditure. In addition, it is unlikely that this effect is due to an acute reduction in water intake because the prokineticin receptor agonist Bv8 has actually been shown to slightly increase, rather than reduce, water intake (19). This phenomenon of a rapid weight loss on the first day is widely observed in investigations of anorectic agents. The rapid decrease is thought to occur as a result of the rapid utilization of glycogen stores, and is especially true in rodents. The decrease is followed by a period in which fat is lost, coinciding with the less rapid weight loss (47,48).

In summary, our data identify a novel role for PK2 in appetite regulation. The anorectic effects of PK2 appear to be mediated in part via the ARC melanocortin system. Repeated peripheral administration of PK2 for 5 days reduced food intake and body weight in obese mice. Further studies investigating the effects of longer term administration of PK2 on food intake and body weight will determine the potential of PK2 as a target for the development of antiobesity agents.

W.S.D. is funded by a National Institute for Health Research (NIHR) Clinician Scientist Award, a Wellcome Trust Value in People Award, and a Higher Education Funding Council for England Clinical Senior Lecturer Award. G.A.B. is supported by a Diabetes Research and Wellness Nonclinical Research Fellowship. K.G.M. is supported by a Biotechnology and Biological Sciences Research Council New Investigator Award. S.H. is funded by a Medical Research Council Clinical Research Training Fellowship. The department is funded by an Integrative Mammalian Biology (IMB) Capacity Building Award and funding from the NIHR Biomedical Research Centre Funding Scheme.

S.C. and S.P.V. are employees of RenaSci Consultancy, a fee-for-service company that tests candidate molecules in CNS models and models of obesity and diabetes. No financial arrangement exists between RenaSci Consultancy and any of the other authors. All work undertaken by S.C. and S.P.V. in the preparation of this article was an extension of existing ongoing academic collaborations with Imperial College, University of London. No other potential conflicts of interest relevant to this article were reported.

The data in this article were presented in abstract form at the 28th annual meeting of the British Endocrine Society, Harrogate, U.K., 16–19 March 2009 and the 91st annual meeting of the American Endocrine Society, Washington, D.C., 10–13 June 2009.