Statistics from Altmetric.com
Our knowledge of the potential genetic link between fat metabolism and the reproductive axis dates back to the 1950s when the obese (ob/ob) mouse strain was first described. These mice were not only distinctly hyperphagic and rapidly developed obesity associated with hyperglycaemia and insulin resistance, but they were also infertile.1 It was not until 1994 when the leptin (ob) gene was postionally cloned and a mutation was identified in the coding sequence of murine leptin in ob/ob mice that the cause of their obesity was recognised.2 As predicted from the phenotype of ob/ob mice, leptin, which is principally expressed in adipocytes, had potent actions to suppress appetite and stimulate energy expenditure. It rapidly became clear that leptin could also influence the reproductive system. The sterility of male and femaleob/ob mice could be reversed when recombinant leptin was administered.3 4
In humans, initial studies focused on the possible role that leptin may play in obesity. The hypothesis that leptin deficiency may contribute to common human obesity was soon rejected. An exponential relation between serum leptin concentration and body mass index or percentage body fat was described,5 implying that, as a person became fatter, so insensitivity to the anorexigenic action of leptin developed. However, the relation between body fat and reproductive ability in humans has long been recognised. Both anorexia nervosa and intense physical training are associated with reduced gonadotrophin levels,6 while Frisch had proposed that a certain amount of body fat must be accrued to achieve regular menstruation.7 There is therefore a clear link between peripheral energy stores (in fat) and central regulation of physical development and reproductive capacity. Studies in animals and humans over the last four years have provided compelling evidence that leptin may be a neurohumoral mediator capable of signalling between the extent of nutritional intake and body fat store to the central nervous system.
Examination of the link between leptin and hypothalamic-pituitary-gonadal function has been undertaken mainly in the murine model. The relation in higher species is less well defined. In addition, the role that leptin may play in the onset of puberty, when the gonadotrophin releasing hormone (GnRH) pulse generator is being re-activated, has not been clarified. A review of data from the mouse, rat, monkey, and man and discussion of the possible neural networks involved will be presented.
The murine model
In mice, the relation between leptin and gonadal function has been tested in two situations: fasting and by direct administration of leptin. Normal mice when fasted not only develop low leptin levels as expected but also hypogonadotrophic hypogonadism, reduced thyroxine levels, and elevated adrenocorticotrophin (ACTH) and corticosterone.8 These pituitary abnormalities could be partially reversed by leptin administered to fasted mice at a dose that returned their low leptin levels back to normal. Leptin can therefore be implicated in adaptation to starvation. Low leptin levels in times of food deprivation would result in an appetite drive, but a reduced reproductive capacity, protecting the female from conceiving and hence the energy demands of pregnancy. In addition, leptin administered to male and female ob/ob mice increased gonadotrophins (luteinising hormone in the female, follicle stimulating hormone in the male) and reproductive organ weights (ovary and uterus, testes and seminal vesicles).9
Intraperitoneal injections of leptin into normal mice in a dose that reduced appetite and hence body weight or in a smaller dose that had no effect on weight have been reported to bring forward the timing of normal puberty as defined by vaginal opening.10 11However, injections of leptin into normal prepubertal rats did not alter pubertal timing. In this model, the animals needed to be starved for an effect of leptin on puberty to be observed.12 13The time of first vaginal opening in leptin treated rats, with food intake reduced to 80% of normal, was similar to that in control ad libitum fed rats, with both having vaginal opening earlier than rats pair fed to the leptin group.12 However, if the food intake in the leptin treated and pair fed control groups was restricted further down to 70% of normal, then leptin only partially reversed the delay in vaginal opening. This suggested that leptin plays a permissive rather than initiating role, allowing puberty to proceed only in favourable circumstances.
The rhesus macaque monkey
In the monkey, investigation of the role of leptin in puberty has been observational rather than interventional. In the male monkey, serum leptin levels have been measured throughout the juvenile period.14 Leptin levels paralleled changes in testosterone, with high levels in infancy, a prepubertal nadir, then elevation through puberty. In a separate study in normal and castrated male monkeys, leptin levels were frequently monitored around the time of puberty.15 No change in leptin concentration was found either before or during the time when luteinising hormone levels were increasing, indicating re-activation of the GnRH pulse generator and the initiation of puberty. This would imply that leptin was certainly not a trigger to puberty. Experiments in which leptin is administered to monkeys in order to assess its effect on the timing of puberty have, however, not been reported.
Data in childhood have been by necessity observational. Most studies have been cross sectional in design.16-19 Many investigators have now reported that leptin increases gradually in both sexes over the prepubertal years. At each age, girls tend to have higher levels than boys. The leptin peak is reached at Tanner genital stage (G) 2–3 in boys, but in girls leptin continues to rise through puberty with a particular increase after menarche. In boys, leptin decreases back to early childhood levels by G5. Therefore, from late puberty and thereafter, there are strikingly discordant leptin levels between the sexes. Measures of body fatness (body mass index (BMI), BMI SD score, percentage body fat) are the most significant determinants of leptin through childhood. However, in both sexes before and during puberty (Tanner stages (TS) 1 and 2), age is a further independent determinant of leptin, implying that there is a maturational influence on leptin independent of body composition.16 In the later stages of puberty (TS 3–5), age remains a significant positive influence in girls, but, in boys, age is replaced by a negative effect related to increasing testicular volume. The latter is likely to reflect the inhibitory effect that testosterone has on leptin secretion. High affinity leptin binding activity, as measured by specific binding of 125I-leptin in serum, varies considerably from birth through childhood. It is relatively low, at 5%, in cord blood of normal neonates, has risen to 18% at age 5, then decreases to 6% in both sexes20 by completion of puberty. Leptin binding activity remains at the slightly higher level of 7.5% in normal adults, in whom its level does not fluctuate with age. This would suggest that leptin may become progressively more available to bind long form leptin receptors over childhood. It could then exert enhanced biological action over the period that a child is progressing towards and entering puberty.
All these data have provided further evidence that leptin has a permissive role in puberty rather than acting as a trigger. However, in one report in which leptin levels were assessed longitudinally in boys as they entered puberty, leptin appeared to show pronounced individual elevation just before the rise in testosterone.21 This may imply a triggering role. However, other reports have not confirmed this observation.22
The most compelling evidence for a role for leptin in human puberty comes from those very rare families with deleterious mutations in either leptin or the leptin receptor.23 24 In adulthood, homozygous subjects with either condition remain substantively, although not completely, hypogonadal. In a peripubertal child with leptin deficiency, treatment with leptin has led not only to pronounced effects on satiety and fat loss but also acute increases in nocturnal gonadotrophin secretion.25 In addition, in boys with constitutional delay in growth and puberty (CDGP), a common disorder of the tempo of puberty, leptin levels at pubertal onset were lower than predicted for age and BMI.26 In normal boys, an increase in leptin between G1 and G2 occurred as indicated above. However, leptin levels in prepubertal boys with constitutional delay in growth were not different from those in early puberty with CDGP. This suggested that the increase in leptin over the prepubertal years was not necessary to achieve puberty, but its absence was associated with a delay in entering puberty.
Leptin is clearly required for appropriate pubertal development and maintenance of secondary sexual characteristics. The combined murine and human data would infer that leptin has a permissive rather than triggering role in puberty.
Leptin and hypothalamic control of GnRH
Circulating leptin is transported into the central nervous system to signal through long form receptors, located on cell bodies, such as neuropeptide Y (NPY) neurones, in the lateral hypothalamus. Many neurotransmitters and neural pathways in the hypothalamus are being linked to the control of appetite and hence body weight (table 1). Likewise, many neurotransmitters have been implicated in the control of GnRH neurones, and potentially in the control of the onset of puberty (table 1). Some of these factors affect both processes. In particular, NPY, a potent orexigenic factor and regulator of GnRH secretion, is thought to be a mediator of the central actions of leptin on appetite.8 NPY has differing effects on GnRH secretion.27 Both activation and inhibition have been described, depending on the model (fasted versus fed, acute versus chronic administration) and the age of the animal. In the fed and nourished state, NPY increases GnRH pulses, but in the undernourished state it inhibits GnRH neurones. In this situation, leptin levels will be low, reducing inhibition on NPY neurones. NPY levels will rise inhibiting GnRH secretion. It is not clear how these opposing actions of NPY on GnRH are mediated. Nevertheless, NPY can clearly act as a central link between nutrition and reproductive function, just as leptin fulfils this role as a peripheral factor.
Leptin also acts on cell bodies, which express cocaine and amphetamine regulated transcript (CART), another anorectic peptide. In in vitro experiments using retrochiasmatic hypothalamic explants of GnRH neurones from prepubertal female rats, leptin can stimulate CART expression, which in turn reduces the interval between pulses of GnRH secretion.28 However, effects of NPY on GnRH pulse interval were not affected by antibodies to CART, implying that the leptin-CART pathway was independent of NPY. It is likely that a number of pathways that can link leptin through to GnRH neurones will be found.
There has been disappointment that leptin deficiency was not the answer to common obesity and that leptin treatment was unlikely to make a significant contribution to improving the health burden resulting from obesity. It appeared that serum leptin in adults was principally a marker of fat mass. In children, however, evidence is mounting that leptin has an important permissive role in the progression into puberty and the maintenance of normal hypothalamic-pituitary-gonadal function thereafter. The central networks in the hypothalamus that mediate this relation are complex and as yet not fully defined. Nevertheless there are neurotransmitters that impact on appetite and GnRH neurones. It will be important to understand these networks as the pharmaceutical drive to develop specific anorectic agents may have repercussions for pubertal and reproductive function.
At present the measurement of leptin in relation to puberty does not have a clinical application. However, further investigation of the exact relation between nutritional intake, body composition, growth, and development may be key to characterising mechanisms that control the tempo of growth.