The neuroendocrine physiology of kisspeptin in the human
- First Online:
- Cite this article as:
- Dhillo, W.S., Murphy, K.G. & Bloom, S.R. Rev Endocr Metab Disord (2007) 8: 41. doi:10.1007/s11154-007-9029-1
Kisspeptin is a 54-amino acid peptide, encoded by the KiSS-1 gene, which activates the G protein-coupled receptor GPR54. Recent evidence suggests the kisspeptin/GPR54 system is a key regulator of reproduction. GPR54-deficient mice have abnormal sexual development. Central or peripheral administration of kisspeptin stimulates the hypothalamic-pituitary-gonadal (HPG) axis in animal models. This review discusses the evidence that kisspeptin also plays a key role in human reproduction. Inactivating GPR54 mutations cause normosmic hypogonadotrophic hypogonadism in humans. Mutations which increase GPR54 signaling are associated with gonadotrophin-dependant premature puberty. Acute intravenous administration of kisspeptin to healthy human male volunteers potently increased plasma LH levels and significantly increased plasma FSH and testosterone without side effects. Plasma kisspeptin is found at low concentrations in the circulation of men and non-pregnant women, but is markedly increased in pregnancy. The placenta is believed to be the source of these high levels of circulating kisspeptin. The kisspeptin-GPR54 system is also implicated in tumour biology. Consistent with this role, plasma kisspeptin concentrations are elevated in patients with abnormal proliferation of placental tissue (gestational trophoblastic neoplasia or GTN) at presentation and fall after treatment with chemotherapy. The kisspeptin/GPR54 system therefore appears to play an important role in the regulation of reproduction in humans. Kisspeptin represents a novel tool for the manipulation of the HPG axis in humans and plasma kisspeptin may be a novel tumour marker in patients with GTN.
Kisspeptin is a 54 amino acid peptide which is encoded by the KiSS-1 gene and activates the previously orphan G protein coupled receptor GPR54 . Initial studies demonstrated that KiSS-1 was a metastasis suppressor gene, producing a number of peptides named kisspeptins (kisspeptin -54, 14, 13, 10), which inhibited tumour progression (hence kisspeptin-54 is also known as metastin). The KiSS-1 gene was named by combining laboratory nomenclature for putative suppressor sequences with its geographical providence; researchers at the Pennsylvania State College of Medicine in Hershey christened it in honour of the famous Hershey chocolate kisses .
In 2003 the importance of the GPR54 receptor as a key regulator required for normal pubertal development was revealed when mice and humans with GPR54 null mutations were found to suffer from hypogonadotropic hypogonadism (absence of sexual maturation with low sex hormones and gonadotrophins) [3–5]. Male GPR54 −/− mice had significantly lower plasma testosterone concentrations compared to age-matched controls. The 17β-oestradiol concentrations in female GPR54 −/− mice were similar to those in wild type female mice in the non-oestrous phase of the reproductive cycle. The failure to produce sperm in male GPR54 −/− mice and lack of an oestrous cycle in female GPR54 −/− mice seems likely to be due to low serum FSH and LH rather than an inability of gonadal tissue to respond to gonadotrophins. The low gonadotrophins levels in GPR54 −/− mice appear not to be due to a problem at the level of the pituitary, as the pituitary gonadotrophs of GPR54 −/− mice synthesize LH and FSH normally. GPR54 deficient mice also have normal hypothalamic gonadotrophin releasing hormone (GnRH) content and the hypogonadotropic hypogonadism and pubertal delay in GPR54 −/− mice can be corrected with exogenous administration of GnRH. This suggests that the role of GPR54 signaling may be to regulate the processing or secretion of GnRH (4). Subsequent studies have confirmed that kisspeptin is a potent stimulator of gonadotrophins when administered peripherally or centrally to rodents, mice, sheep and monkeys (reviewed in ) and this effect is likely to be mediated via the release of GnRH.
2 Human mutations of the GPR54 gene
In humans, GPR54 mutations are associated with a phenotype of hypogonadotropic hypogonadism. Human hypogonadotropic hypogonadism can result from deficiencies in hypothalamic GnRH production, in GnRH receptor function at the pituitary, or in LH and FSH production by the pituitary gland. Hypogonadotropic hypogonadism in the absence of a hypothalamic or a pituitary lesion and with normal pituitary function is diagnosed as Kallman’s syndrome if associated with anosmia. In the absence of anosmia it is classified as isolated or idiopathic hypogonadotropic hypogonadism (IHH). IHH is likely to be due to a defect in GnRH synthesis, secretion or activity and can be caused by a number of different genetic defects, including GnRH receptor mutations. However, in many cases the genes involved are currently unidentified.
A number of inactivating mutations in the GPR54 gene have recently been shown to cause IHH [3, 4, 7–9]. Affected male subjects were able to undergo spermatogenesis and affected females to carry successful pregnancies following exogenous gonadotrophin or GnRH therapy. The phenotype of the patients in these studies therefore suggests that GPR54 signaling is essential for normal gonadotrophin release, but is not critical for spermatogenesis or gametogenesis at the level of the gonad, or for placental function, uterine contraction or lactation.
Since inactivating mutations of the GPR54 gene cause IHH, it might therefore be expected that GPR54 activating mutations cause premature puberty by increasing GnRH, and thus gonadotrophin, release. Consistent with this mutations resulting in enhanced GPR54 signaling have been found in children with gonadotrophin-dependant precocious puberty .
The phenotypes of patients with mutations in the GPR54 gene therefore suggest that the kisspeptin system plays an important role in the regulation of the hypothalamo-pituitary-gonadal axis in humans. Kisspeptin might therefore represent a novel tool for the manipulation of this axis.
3 Administration of kisspeptin-54 to humans
In vivo administration of kisspeptin has been shown to increase plasma LH levels in a number of species, including mice, rats, monkeys and sheep. Kisspeptin can stimulate LH secretion when administered by intracerebroventricular, intravenous, intraperitoneal or subcutaneous routes . Our group investigated the effect of intravenous administration of kisspeptin-54 to human males on reproductive hormone release . Since kisspeptin had not previously been exogenously administered to humans, we chose to administer it by intravenous infusion. An advantage of this mode of administration for a novel peptide is that the infusion can be immediately terminated in response to unexpected side effects, for example, nausea or changes in cardiovascular parameters. Most similar sized peptides are rapidly broken down in the circulation. Once the infusion of peptide is terminated the plasma levels of the peptide will therefore rapidly fall to normal physiological levels. Another advantage of intravenous administration is that the peptide is delivered directly into the circulation, thus avoiding any potential variability in absorption between individuals that can occur, for example, following subcutaneous administration. This also allows more accurate study of the pharmacokinetic characteristics of the peptide in the circulation, for example, half life and volume of distribution. Normal male volunteers were used rather than females in order to minimize the baseline variation in circulating gonadotrophins and sex steroids. Kisspeptin-54 was used because although kisspeptin-10, -14 and -54 have similar efficacy in vitro  there is evidence that kisspeptin-54 is more efficacious than shorter kisspeptin fragments in vivo [12, 13]. Kisspeptin-54 was initially infused at rates from 0.125 to 40 pmol/kg/min to construct a dose-response curve for its effects on LH. There was a dose dependant increase in LH over time of the infusion from an infusion rate of 0.25 pmol/kg/min up to 12 pmol/kg/min. Higher infusion rates than this did not further increase mean LH concentration, and it may be that this rate of infusion maximally stimulates kisspeptin-responsive GnRH neurons. The LH rise at higher infusion rates did appear to be sustained for a longer time period. In human circulation, the half life of kisspeptin-54 is approximately 28 min and therefore at high infusion rates plasma kisspeptin levels would remain elevated for several hours after the infusion was terminated.
Plasma FSH was also raised following kisspeptin-54 infusion, although the effect was not as dramatic as that on LH release. This is in accord with animal models which suggest that the ED50 for kisspeptin on LH release in vivo is approximately 100 fold lower than that for FSH release . There was also a trend towards an increase in mean testosterone levels over time of the infusion. Inhibin B is the major circulating inhibin in man and inhibits secretion of FSH via a negative feedback mechanism. Although the FSH levels were increased by kisspeptin-54 infusion, there was no change in inhibin B levels during the relatively short duration of the infusion.
Blood pressure and pulse rate were measured every 15 min during the infusion and every 30 min after discontinuation of the infusion until completion of the study. We noted no significant changes in these parameters following any of the doses of kisspeptin administered. It has recently been shown that GPR54 mRNA is expressed in smooth muscle cells of the aorta, coronary artery and umbilical vein and that kisspeptins are potent vasoconstrictors in isolated rings of coronary artery and umbilical vein . However, our results suggest that acutely raised circulating kisspeptin levels do not significantly influence systemic blood pressure. It would be interesting to investigate whether intravenous kisspeptin infusion affects, for example, local blood flow in specific tissues.
In the second set of infusions, a double blind crossover study was carried out in normal male volunteers to determine the time course of the effects of kisspeptin and its pharmacokinetics. An infusion rate of 4 pmol kisspeptin-54/kg/min was used as it resulted in a near maximal LH release in the dose response studies. Each volunteer received an intravenous infusion of kisspeptin-54 and a control infusion of saline (0.9%) at least 3 days apart and in random order. Kisspeptin-54 infusion resulted in a significant increase in LH, FSH and testosterone release compared with saline infusion. There was no effect of the kisspeptin infusion on sexual arousal as measured using visual analogue scales. The pharmacokinetics of circulating kisspeptin-54 were also assessed. The plasma half life of kisspeptin-54 was found to be 27.6 min and the volume of distribution suggested that kisspeptin is not extensively tissue-bound. Chromatographic analysis of plasma from volunteers infused with kisspeptin suggested that kisspeptin-54 is not degraded to smaller active kisspeptin fragments in vivo.
These studies clearly show that intravenous infusion of kisspeptin-54 potently stimulates LH secretion in men. These results are similar to the reported effects of intravenous infusion kisspeptin-10 in agonadal juvenile male rhesus monkeys, in which pituitary responsiveness to GnRH had been previously heightened by pulsatile GnRH treatment . Kisspeptin-10 was administered as a 10 μg (approximately 9 nmol) bolus injection, followed by an intravenous infusion of kisspeptin-10 at a rate of 100 μg/h (approximately 135 nmol infused over a 90 min time period). This doubled LH levels at 90 min compared to baseline LH levels. In our studies human males were infused with 4 pmol kisspeptin-54/kg/min for the first 30 min followed by half this dose for the remaining 60 min of the infusion (approximately 17 nmol over 90 min) resulting in a doubling of LH levels. The difference in doses of kisspeptin required to achieve a similar magnitude of LH secretion in these studies may be due to the putative longer half life of kisspeptin-54 compared to kisspeptin-10 in vivo, or alternatively due to species differences.
4 Comparison of the effects of kisspeptin-54 on LH release to other known LH secretagogues in humans
A number of studies have previously reported the effects of intravenous GnRH administration on LH release in men [17–19]. However, the majority of these studies administered GnRH either in a bolus of 100 μg (approximately 90 nmol) of GnRH or a 4 h infusion of GnRH. It is thus difficult to compare the effects of kisspeptin on the HPG axis in our study with the previously reported effects of GnRH. Comparison of LH levels between studies is also made more difficult by differences in the LH assays used. Bearing these cautions in mind, it has been reported that an IV bolus of 100 μg (approximately 90 nmol) of GnRH to normal fertile men resulted in a mean peak LH response 623% elevated from the individual’s preinjection LH concentration. The mean peak FSH response to the same stimulus was 221% elevated from control . A 4 h infusion of GnRH (at a dose of approximately 1 nmol per min i.e. 240 nmol in total over 4 h) to normal fertile men resulted in a mean peak LH response 400% higher than control, and a mean peak FSH response of 150% [18, 19]. Kisspeptin-54 infusion at a dose of 12 pmol/kg/min for the first 30 min followed by half this dose for the remaining 60 min of the infusion (approximately 50 nmol total peptide administered over the 90 min infusion) resulted in a mean LH response 504% higher than the individual preinjection gonadotrophin concentration control. The mean FSH response was elevated 146% compared to the same control . This data suggests that kisspeptin-54 administration is at least as potent, if not more potent than GnRH on stimulation of the HPG axis in healthy men. However, further studies directly comparing the effects of kisspeptin-54 and GnRH in humans are required. It will also be of great interest to determine whether kisspeptin can also stimulate the HPG axis in human females and whether it is effective when administered via alternative routes.
Leptin, the adipocyte hormone signaling the status of the body’s energy stores has recently been shown to affect the hypothalamic kisspeptin system. Kisspeptin neurones express leptin receptors and leptin administration is able to normalize defective KiSS-1 gene expression in rodent models with impaired gonadotrophin secretion caused by hypoleptinaemia [20, 21]. These findings suggest that peripheral leptin levels may mediate their permissive effects on the reproductive axis at least in part via the hypothalamic KiSS-1 system. Leptin is thought to be the critical link between sufficient energy stores and integrity of the HPG axis as exogenous administration of leptin prevents the fasting-induced inhibition of the HPG axis in mice and humans [22–24]. In healthy human males fasted for 72 h, administration of leptin subcutaneously four times per day for 3 days fully prevented the fasting-induced decrease in testosterone levels and LH pulsatility .
A number of hypothalamic neurotransmitters have been suggested to link metabolic regulation and the HPG axis. Neuropeptide Y (NPY) is one of the most potent hypothalamic orexigenic factors known and also stimulates LH secretion in rodents . In healthy men an IV bolus injection of 100 μg (24 nmol) of NPY alone did not affect LH secretion. However, administration of the same dose of NPY potentiated the effect of GnRH (100 μg or 90 nmol) on the HPG axis, increasing plasma LH by 150% compared to administration of GnRH alone. Thus NPY administration alone has no effect on LH release, but can augment the GnRH-induced LH secretion in healthy men .
Glutamate is an excitatory amino acid which increases LH secretion via N-methyl-D-aspartate (NMDA) and non-NMDA receptors. N-methyl-D-aspartate (NMDA) also increases LH secretion in a number of species including rats  and monkeys . Glutamate and NMDA stimulate LH release via their stimulatory effects on hypothalamic GnRH release. Oral administration of 500 mg of D-cycloserine, a partial agonist of the NMDA receptor, resulted in a small increase in plasma LH which was approximately 25% higher in subjects receiving D-cycloserine compared to saline controls at 240 min following administration .
Animal studies have also shown that substance P can influence gonadotrophin secretion . Intravenous administration of substance P at 1.5 pmol/kg/min for 60 min (approximately a total infusion of 7 nmol of substance P) resulted in an approximate doubling of plasma LH levels during the infusion in healthy men . Twenty minutes after the end of the infusion of substance P, plasma LH levels had returned to baseline values. There was no effect of the infusion of substance P on plasma FSH or testosterone. In our studies in human males 4 pmol kisspeptin-54/kg/min for the first 30 min followed by half this dose for the remaining 60 min of the infusion (approximately 17 nmol over 90 min) caused a doubling in LH levels. This suggests that peripherally administered kisspeptin has a similar potency to substance P on LH release. However, further studies would be required to directly compare the effects with substance P infused at higher doses to determine its maximal effect on LH release compared to kisspeptin and GnRH.
5 Regulation of plasma kisspeptin levels in humans
Plasma kisspeptin is found at low concentrations in the circulation of men and non-pregnant women, but is markedly increased in pregnancy . Chromatographic analysis of the kisspeptin-immunoreactivity (-IR) in plasma from both non-pregnant and pregnant women suggests that the major form of circulating kisspeptin is kisspeptin-54.
The mean plasma concentration of kisspeptin-IR in both men and non pregnant women is approximately 1.3 pmol/l. The source of this kisspeptin is currently unknown, though it seems possible that it may be the vascular endothelial cells. The mean concentrations of kisspeptin-IR in maternal plasma are 1,230 pmol/l in the first trimester, 4,590 pmol/l in the second trimester and 9,590 pmol/l in the third trimester. The placenta is believed to be the source of these high levels of circulating kisspeptin. Human trophoblast cells secrete kisspeptin-54, -14 and -10 in vitro. By 5 days post delivery, the plasma concentrations of kisspeptin-IR have almost returned to non pregnant levels (7.6 pmol/l) . The role of kisspeptin in human pregnancy is currently unknown. It is possible that kisspeptin mediates the down-regulation of the HPG axis in pregnancy. There is no data on GPR54 desensitization in humans. However, following continuous kisspeptin stimulation, accumulation of inositol phosphate (IP) in GPR54-transfected COS-7 cells reached a peak at 4 h and gradually decreased thereafter. By 18 h IP levels had returned to baseline despite the continued presence of kisspeptin, suggesting a desensitization effect . The effect of continuous administration of kisspeptin on LH secretion has also been examined in vivo in monkeys and rats. Seminara et al.  studied agonadal, juvenile male monkeys. Kisspeptin-10 was infused continuously and elicited a brisk LH response for approximately 3 h. This rise was then followed by a precipitous drop in LH despite continuous exposure of GPR54 to kisspeptin-10. On the fourth day, during the final 3 h of the kisspeptin-10 infusion, single boluses of kisspeptin-10 were unable to elicit LH pulses, suggesting desensitization of GPR54. Thompson et al.  investigated the effects of continuous peripheral kisspeptin-54 administration in male rats. Subcutaneous administration of kisspeptin-54 for 1 day increased plasma LH and testosterone. This effect was lost after 2 days of administration, suggesting a down regulation of the HPG axis response to kisspeptin following continuous administration. It might be expected therefore that the high circulating kisspeptin levels in pregnancy would cause desensitization of the GPR54, and thus perhaps prevent hypothalamic kisspeptin from stimulating the HPG axis. However, it is interesting that women with GPR54 mutations can carry successful pregnancies following exogenous gonadotrophin or GnRH therapy. This suggests that the GPR54 receptor is not critical in maintaining a normal pregnancy.
The polycystic ovary syndrome (PCOS) is common amongst women of reproductive age and causes hyperandrogenism and oligomenorrhoea. The pathogenesis of PCOS is unclear, but involves hypothalamic-pituitary disturbances in gonadotrophin release. Women with PCOS demonstrate higher LH levels compared to ovulatory women without the syndrome. It is possible that there are links between the kisspeptin system and PCOS. However, the results of a recent study suggest that plasma kisspeptin-IR concentrations are not significantly different in women with PCOS (0.28 ± 0.01 pmol/l) compared to controls (0.36 ± 0.05 pmol/l) . It is interesting that the plasma kisspeptin-IR in this study in normal controls was almost four fold lower than that previously reported . Further work is required to confirm this finding which suggests that kisspeptin is not directly associated with PCOS-associated LH hypersecretion.
The KiSS-1 gene was first discovered as an anti-metastasis gene. KiSS-1 suppresses metastasis in human breast carcinomas and KiSS-1 expression inversely correlates with increased metastasis in many tumours including cancer progression in gestational trophoblastic neoplasia (GTN). Therefore KiSS-1 represents a potential marker to distinguish metastatic from non-metastatic forms of specific cancers.
GTN compromises a number of disorders characterized by an abnormal proliferation of placental tissue. Complete (CHM) and partial hydatidiform moles (PHM) are the most common form of the disease and are regarded as premalignant. Sixteen percent of CHM and 0.5% of PHM transform in time into the malignant conditions of invasive mole, choriocarcinoma or placental site trophoblastic tumour. The latter three diseases are referred to as malignant GTN. All forms of GTN secrete human chorionic gonadotrophin (hCG) and serum measurement of this hormone is extremely useful in the diagnosis, staging and subsequent assessment of the therapeutic response of malignant GTN. However, many existing commercial assays for hCG detection are troubled by false positives, leading to inappropriate chemotherapy, or false negatives, leading to a delay in diagnosis. The development of new markers for GTN may provide better predictive models of trophoblastic behavior. Placental KiSS-1 gene expression is increased in normal and molar pregnancies and circulating kisspeptin-IR is increased in normal pregnancy. We therefore hypothesized that plasma kisspeptin-IR would be altered in patients with GTN. Plasma kisspeptin concentrations were elevated in patients with invasive mole at presentation, and fell similarly to hCG levels in each patient during and after treatment with chemotherapy . Plasma kisspeptin may thus act as a novel tumour marker in patients with malignant GTN. Placental KiSS-1 expression in molar pregnancy is elevated to a similar degree to that seen in normal pregnancy but is undetectable in choriocarcinoma. Plasma kisspeptin-IR may be elevated in patients with an invasive mole to inhibit further tumour metastasis and the development of choriocarcinoma. Additional studies are required to determine whether plasma kisspeptin-IR is significantly lower in patients with choriocarcinoma compared with plasma kisspeptin-IR in patients with molar pregnancy. In addition, it would be of great interest to determine whether molar pregnancy patients who later develop choriocarcinoma have different circulating kisspeptin levels from molar pregnancy patients who do not.
6 Possible therapeutic applications of the kisspeptin/GPR54 system
Since 2003, data from human and animal studies have confirmed a fundamental role for the kisspeptin-GPR54 system in reproduction. Administration of exogenous kisspeptin-54 to human males without side effects demonstrates that kisspeptin can be used to manipulate the hypothalamic-pituitary gonadal axis in humans. GPR54 agonists have therapeutic potential in initiating puberty in cases of pubertal delay, and in induction of ovulation in those with anovulation. GPR54 antagonists may be useful in treating children with GDPP. Further research is necessary to explore whether kisspeptin might also have utility in the detection and treatment of specific cancers, for example, prostate cancer and choriocarcinoma. High throughput screening approaches to identify small molecule GPR54 agonists and antagonists are already underway , and small peptidic GPR54 agonists have been used to determine the structure-activity relationship of the kisspeptin molecule [37, 38]. Were such studies to lead to the development of orally active agents that can agonise or antagonize the GPR54, the therapeutic potential of the kisspeptin system might then be fully realized.
W.S.D. is funded by a Department of Health Clinician Scientist Fellowship. K.G.M. is supported by a Biotechnology and Biological Sciences Research Council New Investigator Award.