The kisspeptin-GnRH pathway in human reproductive health and disease.

Since its discovery, hypothalamic secretion of GnRH has been robustly established as the key pathway that initiates and controls reproductive function. Whilst the pivotal central role played by GnRH remains undisputed, a number of functional limitations of the GnRH neuronal network have been identified. For example, in rats, GnRH neurones lack estrogen receptor (ER)-alpha ( Herbison and Theodosis, 1992 ), the principal ER, thus suggesting the need for an intermediary signalling pathway mediating gonadal feedback. It was only a decade ago that the discovery of the obligate role of kisspeptin in human puberty revolutionized current understanding of the neuroendocrine regulation of human reproduction ( de Roux et al. , 2003 ; Seminara et al. , 2003 ). Kisspeptin, a hypothalamic peptide coded by the KiSS1 gene, is a novel neuromodulator that acts upstream of GnRH, and is sensitive to sex steroid feedback and metabolic cues. Kisspeptin is now recognized as a crucial regulator of the onset of puberty, the regulation of sex hormone-mediated secretion of gonadotrophins, and the control of fertility ( Pinilla et al. , 2012 ). The related discovery of a reproductive role for neurokinin B has stimulated further interest in the field. The same functional neuronal network secretes kisspeptin and neurokinin B—now called kisspeptin-neurokinin B-dynorphin (KNDy) neurones as they also co-secrete dynorphin, a well-established opioid inhibitor ( Goodman et al. , 2007 ). Exogenous kisspeptin has been administered to healthy volunteers and a limited number of patients, with a view to restoring reproductive function in certain conditions. In this review, we summarize current understanding of the physiological regulation of GnRH pulse frequency by kisspeptin, and appraise the clinical relevance of the discoveries of kisspeptin and neurokinin B. The focus will predominantly be on human findings, using animal data where human studies are lacking but where there is direct translational potential.

The authors' studies in this field are supported by the Medical Research Council ( G0701682 ), the Novo Nordisk UK Research Foundation and Sanofi Excellence for Diabetes Research Awards. K.S.'s current position as a clinical research fellow is funded by the Wellcome Trust through the Scottish Translational Medicine and Therapeutics Initiative (STMTI). J.T.G.'s current position is part-funded by the University of Oxford Diabetes Trials Unit and the NIHR through the Oxford Biomedical Research Centre. Funding to pay the Open Access publication charges for this article was provided by the Wellcome Trust through the Scottish Translational Medicine and Therapeutics Initiative.

A systematic literature search was performed using PUBMED for all English language articles published up to January 2014 using the terms ‘kisspeptin’ and ‘reproduction’. The search was performed without limitations by species although subsequent priority was given to human studies, where available. The initial search identified 390 manuscripts, which were used as background material for the review. In addition, the reference lists of all relevant original research articles and reviews were examined and selected if judged to be relevant. Relevant abstracts from recent scientific meetings were included in the review.

K.S., J.T.G. and R.A.A. contributed equally to determining the scope of the review. K.S. and J.T.G. undertook the literature review. K.S. drafted the manuscript, which was edited by J.T.G. and R.A.A. All authors have approved the final manuscript for submission.

GnRH analogues are extensively used in clinical practice in the treatment of hormone-dependent diseases and infertility. Current therapies manipulate the HPG axis at the level of GnRH receptors on pituitary gonadotrophs, largely to suppress gonadal function, e.g. in the treatment of prostate and breast cancer, endometriosis and uterine fibroids. As reproductive endocrine conditions can be broadly categorized into those with pathologically diminished (delayed puberty, hypothalamic amenorrhoea, hypogonadism in diabetes) and pathologically enhanced (polycystic ovary syndrome (PCOS), menopause, precocious puberty) GnRH and associated gonadotrophin pulsatility, the newly discovered hypothalamic peptides kisspeptin and neurokinin B offer a novel therapeutic approach with potential advantages over the existing therapies in several clinical contexts (Fig.  4 ). Figure 4 Potential clinical applications of novel kisspeptin-based modulation of LH secretion. Schematic presentation of LH pulses in health and in reproductive endocrine disorders. In health, an LH pulse occurs about every 90 min. The frequency of LH pulses is diminished in patients with hypothalamic amenorrhoea, male hypogonadism and pubertal delay, whereas LH pulse frequency is enhanced in women with polycystic ovary syndrome, menopause and precocious puberty. Therapeutic opportunities to correct abnormal LH pulse frequency by manipulating KNDy neurones with relevant agonists and antagonists are emerging. PCOS, polycystic ovary syndrome. Potential clinical applications of novel kisspeptin-based modulation of LH secretion. Schematic presentation of LH pulses in health and in reproductive endocrine disorders. In health, an LH pulse occurs about every 90 min. The frequency of LH pulses is diminished in patients with hypothalamic amenorrhoea, male hypogonadism and pubertal delay, whereas LH pulse frequency is enhanced in women with polycystic ovary syndrome, menopause and precocious puberty. Therapeutic opportunities to correct abnormal LH pulse frequency by manipulating KNDy neurones with relevant agonists and antagonists are emerging. PCOS, polycystic ovary syndrome. Enhancing the stimulatory tone of kisspeptin and neurokinin B by appropriate agonists and suppressing the inhibitory tone of dynorphin by its antagonists, may have therapeutic potential for diseases with decreased gonadotrophin secretion. Exogenous kisspeptin enhances diminished LH pulsatility in hypogonadal men with diabetes and stimulates LH secretion in women with hypothalamic amenorrhoea ( Jayasena et al. , 2010 ; George et al. , 2013 ). Kisspeptin initiates puberty in monkeys and rodents, but this has not been tested in children with delayed puberty ( Navarro et al. , 2004b , Plant et al. , 2006 ). Pulsatile gonadotrophin secretion is restored by kisspeptin administered to patients with hypogonadotropic hypogonadism secondary to mutations in neurokinin B and/or its receptor ( Young et al. , 2013 ). The role of dynorphin antagonists, such as naloxone, in patients with abnormally low LH secretion remains to be elucidated. Kisspeptin therapy has the potential to ‘fine tune’ IVF techniques. Kisspeptin triggered the LH surge during following ovulation induction for assisted reproduction ( Abbara et al. , 2013 ) with successful achievement of a live birth. Kisspeptin might stimulate a more physiological pattern of gonadotrophin secretion, avoiding the risk of ovarian hyperstimulation syndrome associated with currently used hCG injections although clearly much remains to be discovered regarding potential advantages and disadvantages over current approaches. Suppressing the stimulatory role of kisspeptin and neurokinin B by specific receptor antagonists and enhancing the inhibitory action of dynorphin by its receptor agonist is desirable in scenarios of increased GnRH pulsatility where a reduction rather than complete suppression of GnRH is required. Increased frequency of GnRH and therefore LH pulsatile secretion (with little effect on FSH secretion) is central to the pathophysiology of PCOS, the most common endocrinopathy in women. As GnRH pulse frequency primarily determines LH but not FSH secretion ( McNeilly et al. , 2003 ), slowing GnRH might normalize the relative LH hypersecretion often seen in PCOS. Normalization of LH secretion (and perhaps the consequent hyperandrogenism) in PCOS may promote folliculogenesis and ovulation. Studies using a neurokinin B antagonist are currently underway to reduce high LH secretion in PCOS. The ability of kisspeptin antagonists to limit follicular development and inhibit ovulation offers potential for a novel female contraceptive, perhaps being specifically advantageous in the scenarios where exogenous estrogen is contraindicated. However it might be limited by the resulting lack of progesterone exposure and adverse effects on the endometrium. Given the preferential stimulation of LH secretion in response to kisspeptin in humans ( Dhillo et al. , 2005 , 2007 ; George et al. , 2011 , 2012 ), kisspeptin antagonists might potentially result in relative sparing of FSH compared with LH, which might reduce or prevent the unwanted side effects of estrogen deficiency, such as vasomotor symptoms and risk of osteoporosis, associated with GnRH analogue administration. A similar therapeutic approach might support the use of kisspeptin and neurokinin B suppressive therapies in the treatment of precocious puberty. Similarly, it may alleviate menopausal hot flushes since KNDy neurones project to preoptic thermoregulatory areas that express neurokinin B receptor in rats and KNDy neurone ablation reduces cutaneous vasodilation ( Burke et al. , 2006 ; Rance, 2009 ; Hrabovszky et al. , 2010 ; Krajewski et al. , 2010 ; Mittelman-Smith et al. , 2012 ; Rance et al. , 2013 ). Although the inhibitory role of opioids on GnRH and LH pulsatility is well known, manipulation of this system does not have the apparent specificity of the kisspeptin or neurokinin B pathways. The potential more subtle effects of kisspeptin antagonists reducing LH pulsatility contrast with the profound suppression resulting from GnRH analogue administration, decreasing gonadotrophin and sex steroid secretion to castration levels with consequent side effects, including hot flushes, loss of libido and decreased bone mineral density ( Roseweir et al. , 2009 ). Complete suppression of gonadotrophins and sex steroids is necessary is some conditions, such as prostate cancer, but partial suppression is more appropriate in benign prostatic hyperplasia, endometriosis and uterine fibroids. Clinical effectiveness in the management of endometriosis and uterine fibroids with GnRH suppression with add back, and with selective progesterone receptor modulators ( Chabbert-Buffet et al. , 2005 ), suggests that approaches not based on complete suppression of the HPG axis have clear clinical value. Targeted partial gonadotrophin suppression, such as that afforded via kisspeptin and/or neurokinin B inhibition, has the potential to overcome the existing drawbacks of GnRH analogues although the emerging data on kisspeptin analogues ( MacLean et al. , 2013 ; Scott et al. , 2013 ) demonstrate the potential for profound suppression as well.

J.T.G. serves as the International Co-ordinating Investigator for an AstraZeneca sponsored clinical trial in PCOS, as a consultant for AstraZeneca and Takeda Pharmaceuticals; and has received educational grants, speaker fees or advisory board fees from most leading pharmaceutical companies active in the field of diabetes. R.A.A. has undertaken consultancy work for AstraZeneca and Takeda Pharmaceuticals.

KISS1 , the gene encoding kisspeptins, was originally identified in 1996 as a suppressor of metastasis in human malignant melanoma ( Lee et al. , 1996 ). As it was discovered in Hershey (PA, USA), the gene was named after the famous chocolate ‘Kisses’ produced in the town. The SS in KiSS1 acknowledges that it is a ‘suppressor sequence’. The KISS1 gene is localized to chromosome 1q32 and has four exons, the first two of which are not translated ( West et al. , 1998 ). The gene encodes the precursor 145 amino acid peptide, which is cleaved to a 54 amino acid protein ( West et al. , 1998 ). To acknowledge its metastasis inhibitory properties, the 54 amino acid transcript was named ‘metastin’ ( Ohtaki et al. , 2001 ). This can be further cleaved to 14, 13 and 10 amino acid peptides. The 54 amino acid and the shorter peptides belong to the RF amide group of peptides, sharing the C-terminal sequence of Arg-Phe-NH 2 , and are now collectively referred to as kisspeptins ( Kotani et al. , 2001 ). In 2001, kisspeptin was identified as a ligand for the orphan G-protein coupled receptor 54 (GPR54), which was first described in the rat brain and subsequently in human (then named AXOR12 and hOT7T175) ( Lee et al. , 1999 ; Muir et al. , 2001 ; Ohtaki et al. , 2001 ), now termed KISS1R ( Gottsch et al. , 2009 ). KISS1R maps to chromosome 19p13.3 and includes five exons, encoding a 398 amino acid protein with seven hydrophobic trans-membrane domains ( Muir et al. , 2001 ). It has an amino acid sequence close to that of the galanin receptor family (40% identity), although it does not bind either galanin or galanin-like peptide ( Lee et al. , 1999 ). Upon binding by kisspeptin, KISS1R activates phospholipase C and recruits secondary intracellular messengers, inositol triphosphate and diacylglycerol, which in turn mediate calcium release and protein kinase C activation to mediate kisspeptin's function ( Muir et al. , 2001 ; Liu et al. , 2008 ; Constantin et al. , 2009 ). Activation of Kiss1r results in a biphasic increase in intracellular calcium, with a rapid increase followed by a more sustained second phase ( Min et al. , 2014 ). To maintain this second phase and therefore sustain signalling, kisspeptin receptor trafficking involving internalization, recycling and recruitment from an intracellular pool, is required ( Min et al. , 2014 ). Without receptor trafficking, the kisspeptin receptor undergoes desensitization following an initial acute phase ( Min et al. , 2014 ). Since the discovery of kisspeptin-KISS1R signalling, many different terms have been used to describe its components. The nomenclature used in this review for kisspeptin and its receptor will be that recently recommended by Gottsch et al. (2009) . GnRH neurones extend from the preoptic area through to the infundibular nucleus (homologue to the arcuate nucleus in other species) of the hypothalamus in humans, whereas in rodents GnRH neurones reside predominantly in the preoptic area ( Lehman et al. , 1986 ; Schwanzel-Fukuda and Pfaff, 1989 ; Clifton and Steiner, 2009 ) (Fig.  1 ). GnRH axons project from these nuclei to the median eminence, where GnRH is secreted into the portal circulation in a coordinated and pulsatile manner. Similarly, kisspeptin neurones are located in the rostral preoptic area and the infundibular nucleus in the human hypothalamus ( Rometo et al. , 2007 ; Hrabovszky et al. , 2010 ). The anatomical distribution of kisspeptin neurones and their appositions with other hypothalamic endocrine networks are described below. Areas of incongruity between data from human studies and those carried out in other species are also highlighted. Figure 1 Schematic diagram showing the neuroanatomy of the kisspeptin-GnRH pathway and the relationship between KNDy neurones and GnRH neurones in humans and rodents. Kisspeptin signals directly to the GnRH neurones, which express kisspeptin receptor. The location of kisspeptin neurone populations within the hypothalamus is species specific, residing within the anteroventral periventricular nucleus (AVPV) and the arcuate nucleus in rodents, and within the preoptic area (POA) and the infundibular nucleus in humans. Kisspeptin neurones in the infundibular (humans)/arcuate (rodents) nucleus co-express neurokinin B and dynorphin (KNDy neurones), which via neurokinin B receptor and kappa opioid peptide receptor autosynaptically regulate pulsatile kisspeptin secretion, with neurokinin B being stimulatory and dynorphin inhibitory. Negative (red) and positive (green) sex steroid feedback is mediated via distinct kisspeptin populations in rodents, via the AVPV and the arcuate nucleus, respectively. In humans KNDy neurones in the infundibular nucleus relay both negative (red) and positive (green) feedback. The role of the POA kisspeptin population in mediating sex steroid feedback in humans is incompletely explored. ME, median eminence; +, stimulatory; −, inhibitory; ERα, estrogen receptor alpha; PR, progesterone receptor; Kiss1/KiSS1, kisspeptin; NKB, neurokinin B; Dyn, dynorphin. Schematic diagram showing the neuroanatomy of the kisspeptin-GnRH pathway and the relationship between KNDy neurones and GnRH neurones in humans and rodents. Kisspeptin signals directly to the GnRH neurones, which express kisspeptin receptor. The location of kisspeptin neurone populations within the hypothalamus is species specific, residing within the anteroventral periventricular nucleus (AVPV) and the arcuate nucleus in rodents, and within the preoptic area (POA) and the infundibular nucleus in humans. Kisspeptin neurones in the infundibular (humans)/arcuate (rodents) nucleus co-express neurokinin B and dynorphin (KNDy neurones), which via neurokinin B receptor and kappa opioid peptide receptor autosynaptically regulate pulsatile kisspeptin secretion, with neurokinin B being stimulatory and dynorphin inhibitory. Negative (red) and positive (green) sex steroid feedback is mediated via distinct kisspeptin populations in rodents, via the AVPV and the arcuate nucleus, respectively. In humans KNDy neurones in the infundibular nucleus relay both negative (red) and positive (green) feedback. The role of the POA kisspeptin population in mediating sex steroid feedback in humans is incompletely explored. ME, median eminence; +, stimulatory; −, inhibitory; ERα, estrogen receptor alpha; PR, progesterone receptor; Kiss1/KiSS1, kisspeptin; NKB, neurokinin B; Dyn, dynorphin. Initial studies of the neuroanatomical distribution of kisspeptin neurones in the human brain carried out in autopsy samples from premenopausal and post-menopausal women localized KISS1 expression to the infundibular nucleus only ( Rometo et al. , 2007 ). A more recent study, using both male and female autopsy samples, has confirmed the localization of the majority of kisspeptin cell bodies in the infundibular nucleus, but identified a second dense population of kisspeptin cells in the rostral preoptic area ( Hrabovszky et al. , 2010 ). Whilst kisspeptin neurones are located in the infundibular/arcuate nucleus across all species, including humans, the rostral population is species specific ( Clarkson and Herbison, 2006 ; Pompolo et al. , 2006 ; Ramaswamy et al. , 2008 ; Clarkson et al. , 2009 ; Hrabovszky et al. , 2010 ). In rodents, the rostral population is located in the anteroventral periventricular nucleus (AVPV) and the periventricular nucleus (PeN), the continuum of this region known as the rostral periventricular region of the third ventricle (RP3V) ( Clarkson and Herbison, 2006 ; Clarkson et al. , 2009 ) (Fig.  1 ). In contrast, humans and ruminants lack this well-defined RP3V population and have more scattered kisspeptin cell bodies within the preoptic region ( Pompolo et al. , 2006 ; Oakley et al. , 2009 ; Hrabovszky et al. , 2010 ). Unlike in humans and ruminants, Kiss1 mRNA was not detectable in the preoptic area in the adult rhesus monkey ( Ramaswamy et al. , 2008 ). Kisspeptin axons form dense pericapillary plexuses in the human infundibular stalk, the site of neurosecretion of GnRH ( Hrabovszky et al. , 2010 ). Axo-somatic, axo-dendritic and axo-axonal contacts between kisspeptin and GnRH axons were also demonstrated in the infundibular stalk, in keeping with data from rodents, sheep and monkeys, where kisspeptin and GnRH neuronal networks are in close proximity ( Clarkson and Herbison, 2006 ; Ramaswamy et al. , 2008 ; Smith et al. , 2008a ; Hrabovszky et al. , 2010 ; Uenoyama et al. , 2011 ). GnRH neurones express Kiss1r mRNA ( Irwig et al. , 2004 ; Han et al. , 2005 ; Messager et al. , 2005 ). These findings indicate direct involvement of kisspeptin in the neurosecretion of GnRH. However, in humans as well as other species studied to date, not all GnRH neurones receive kisspeptin neurone contacts and the incidence of these contacts seems low ( Clarkson and Herbison, 2006 ; Ramaswamy et al. , 2008 ; Smith et al. , 2008a ; Hrabovszky et al. , 2010 ), suggesting a subtle regulation of GnRH secretion by kisspeptin and other neuropeptides. There is considerable overlap in the distribution of kisspeptin, neurokinin B and dynorphin in the hypothalamus, with frequent colocalization. Mapping of kisspeptin and neurokinin B neurones was similar in the infundibular nucleus of post-menopausal women, prompting the identification of a subpopulation of kisspeptin neurones which express neurokinin B and dynorphin in the human infundibular nucleus ( Rometo et al. , 2007 ; Hrabovszky et al. , 2010 ). This unique region expressing kisspeptin, neurokinin B and dynorphin is conserved across species and resides in the hypothalamic arcuate nucleus in sheep and rodents ( Burke et al. , 2006 ; Goodman et al. , 2007 ; Navarro et al. , 2009 ). Neurokinin B and dynorphin are, however, absent from the kisspeptin population in the preoptic area/RP3V. These infundibular (human)E/arcuate (rodent and ruminant) nucleus neurones which co-express all three neuropeptides are referred to as KNDy neurones ( Cheng et al. , 2010 ) (Fig.  1 ). KNDy neurones in rats and sheep also co-localize with the glutamate transporter-2, and glutamate has been implicated in mediating estrogen positive feedback resulting in the pre-ovulatory GnRH surge. Whether KNDy cells express glutamate receptors remains to be determined ( Pompolo et al. , 2003 ; Ciofi et al. , 2006 ). Kisspeptin neurones in the preoptic area/RP3V are not KNDy neurones, but in the mouse AVPV they co-express tyrosine hydroxylase (the key regulatory step in dopamine synthesis) ( Oakley et al. , 2009 ). This differential expression of neuropeptides reflects complex signalling within the hypothalamus and distinct functions of the two kisspeptin populations ( Ojeda et al. , 2010 ; Tello et al. , 2010 ). KNDy neurones of the infundibular/arcuate nucleus influence the activity of GnRH by acting on both GnRH cell bodies and neurosecretory terminals ( Krajewski et al. , 2005 ; Ciofi et al. , 2006 ; Ramaswamy et al. , 2008 ) (Fig.  1 ). KNDy neurones make direct contact with GnRH cell bodies and dendrites in humans and project to the median eminence in rodents, sheep and monkeys ( Krajewski et al. , 2005 ; Ciofi et al. , 2006 ; Clarkson and Herbison, 2006 ; Ramaswamy et al. , 2008 ; Dahl et al. , 2009 ). KNDy cells act synergistically to produce coordinated and pulsatile GnRH secretion by controlling the neuroactivity of other KNDy cells, as inferred from a reciprocally interconnected KNDy cell network within the arcuate nucleus in the sheep and rat ( Foradori et al. , 2002 ; Burke et al. , 2006 ; Lehman et al. , 2010 ). This is supported by the expression of neurokinin B receptors and the kappa opioid peptide receptor (the receptor for dynorphin) within the KNDy cells, but not the kisspeptin receptor, which predominantly co-localizes with GnRH neurones ( Krajewski et al. , 2005 ; Navarro et al. , 2009 ; Herbison et al. , 2010 ) (Fig.  1 ). This implies that the stimulatory role of neurokinin B and the inhibitory action of dynorphin autosynaptically coordinate the pulsatile release of kisspeptin, which in turn drives the pulsatile secretion of GnRH and LH ( Navarro et al. , 2009 ). Kisspeptin-mediated GnRH stimulation is sex steroid dependent. Estrogen and progesterone modulate kisspeptin activity at both the AVPV nucleus and the arcuate/infundibular nucleus through sex steroid receptors ( Ciofi et al. , 1994 ; Goubillon et al. , 2000 ; Foradori et al. , 2002 ; Smith et al. , 2005 ; Franceschini et al. , 2006 ) (Fig.  1 ). It is becoming clear that not only do kisspeptin neurones mediate both negative and positive sex steroid feedback, but also that distinct subgroups, which are species specific, are involved in these two critical regulatory functions, described more fully below (sections: Kisspeptin mediates negative sex steroid feedback and Kisspeptin may also mediate estrogenic positive feedback). In rodents, the AVPV and the arcuate nucleus respond to positive and negative sex steroid feedback, respectively ( Smith et al. , 2005 , 2006b , Herbison, 2008 ), whereas in humans, the infundibular nucleus alone relays sex steroid signalling ( Rometo et al. , 2007 ; Oakley et al. , 2009 ) (Fig.  1 ). Thus while there is less marked anatomical differentiation of the two feedback pathways in humans, it remains possible (and perhaps likely) that the two functions are mediated by different neurones. There is evidence for sexual dimorphism in kisspeptin pathways in the human, probably reflecting these functional differences discussed above. Female hypothalami have significantly more kisspeptin fibres in the infundibular nucleus and ventral periventricular zone than are seen in men ( Hrabovszky et al. , 2010 ). There is also a striking sex difference in the number and expression of kisspeptin cell bodies, which are observed in the rostral periventricular zone of the female only ( Hrabovszky et al. , 2010 ). Likewise only a few kisspeptin cell bodies are present in the infundibular nucleus in males in contrast to the abundant kisspeptin cell bodies in females ( Hrabovszky et al. , 2010 ). Similarly, sex differences have been reported in the arcuate nucleus of the sheep ( Cheng et al. , 2010 ). Pre-ovulatory positive sex steroid feedback is unique to the female, and the adult female mouse and rat hypothalamus contain 10-fold more kisspeptin neurones than males in the RP3V region, whereas the arcuate nucleus responsible for negative sex steroid feedback does not display such dimorphism ( Clarkson and Herbison, 2006 ; Kauffman et al. , 2007 ).

Human reproductive function is influenced by both extremes of nutrition—undernutrition and obesity. Kisspeptin may provide a link between nutritional/metabolic status and reproduction by sensing energy stores and translating this information into the pulsatile secretion of GnRH. The expression of Kiss1 mRNA and gonadotrophin secretion is reduced in mice, pubertal rats and monkeys subject to fasting ( Castellano et al. , 2005 ; Cota et al. , 2006 ; Roa et al. , 2009 ; Wahab et al. , 2011 ). Kisspeptin is able to restore delayed vaginal opening and increases low gonadotrophin and estrogen levels associated with chronic undernutrition in pre-pubertal rats ( Navarro et al. , 2004b , Castellano et al. , 2005 ). Humans with mutations in leptin or leptin receptor show hypogonadism ( Farooqi and O'Rahilly, 2009 ). The leptin receptor (Ob-Rb) is not present on GnRH neurones, but 40% of kisspeptin neurones in the mouse arcuate nucleus express the leptin receptor ( Smith et al. , 2006a ), suggesting a role for kisspeptin in mediating the metabolic signals of leptin on the HPG axis. Leptin-deficient mice show decreased expression of Kiss1 mRNA, which is partially up-regulated by leptin ( Smith et al. , 2006a ). Incomplete restoration of Kiss1 mRNA expression suggests that other mediators are involved in inhibiting kisspeptin signalling in leptin deficiency. Furthermore, mice with selective deletion of leptin receptor from kisspeptin neurones display normal pubertal development, sexual maturation and fertility, demonstrating that leptin action on kisspeptin neurones is not obligatory for these processes ( Donato et al. , 2011 ). Low levels of testosterone have also been observed in men with obesity and type 2 diabetes, where decreased secretion of GnRH is thought to be the causative factor ( Dandona et al. , 2008 ). A rat model of diabetes (streptozocin treated) has reduced levels of hypothalamic Kiss1 mRNA with subsequently low levels of circulating gonadotrophins and sex steroids, which are corrected by kisspeptin ( Castellano et al. , 2006 , 2009 ). This raises the possibility that diminished kisspeptin secretion is a potential mechanism for hypogonadotropic hypogonadism in patients with obesity and diabetes ( George et al. , 2010 ). Indeed, as described above, kisspeptin-10 increased LH pulse frequency and LH secretion in hypogonadal men with type 2 diabetes ( George et al. , 2013 ). The likely pathways for down-regulation of kisspeptin signalling include negative feedback by estrogen, which is markedly elevated in obesity ( Schneider et al. , 1979 ), resistance to leptin, also seen in human obesity ( Finn et al. , 1998 ), insulin resistance and hyperglycaemia ( Castellano et al. , 2006 , 2009 ), and inflammation, which is up-regulated in hypogonadal men with diabetes ( Dandona et al. , 2008 ) and is associated with decreased kisspeptin expression in rats ( Iwasa et al. , 2008 ). Current data indicate that kisspeptin acts downstream to metabolic signals and conveys information about energy stores to GnRH neurones, thereby regulating reproduction. This gives promise for a potential novel therapeutic role of kisspeptin to restore the reproductive axis in conditions of negative energy balance, such as anorexia nervosa, and in diabetes.

The discovery of kisspeptin has transformed our understanding of the neuroendocrine signals controlling the reproductive axis. Kisspeptin coordinates GnRH secretion, mediates gonadal steroid negative and positive feedback, controls the onset of puberty, and relays information regarding the body's energy stores. The last decade has thus seen a huge resurgence in interest in neuroendocrinology, and the potential for translational application is already being explored in human studies. However, much remains to be learnt before kisspeptin can replace or be used in conjunction with GnRH and gonadotrophin analogues, the current mainstay of infertility and reproductive endocrine disorder treatments. The mode of kisspeptin administration, as with most peptides, remains a challenge and there is thus the need for novel approaches and the development of non-peptide analogues, which is already well underway. These will also allow refinement of experimental approaches to explore physiological pathways (such as elaboration of the importance of the sex steroid environment) as well as novel treatment strategies across a wide range of conditions requiring manipulation of gonadal function. Co-administration of kisspeptin, opioid and neurokinin B modifying agents will allow fine modulation of the HPG axis that may open new therapeutic avenues.