Department of Medicine, Division of Endocrinology, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts 02215
Address correspondence and requests for reprints to: Dr. Jeffrey S. Flier, Beth Israel Deaconess Medical Center and Harvard Medical School, Department of Endocrinology, 325 RN, 99 Brookline Avenue, Boston, Massachusetts 02215. E-mail: jflier{at}caregroup.harvard.edu
![]() |
Introduction |
---|
![]() ![]() ![]() |
---|
Our understanding of the mechanistic links between nutrition and the reproductive axis is incomplete. An important early observation was that timing of puberty is predicted more precisely by body weight than by chronological age (2). This observation led to the hypothesis that attainment of a critical body weight (2), a critical fat mass (11), and/or critical levels of metabolites (12) must be achieved before puberty can occur. However, many issues that emerged from this general hypothesis remained unanswered, including the identity of the signals that inform the brain about nutritional status, the identity of the brain sites that receive and analyze this information, and the mechanism by which this information is transposed into a signal that alters GnRH release. About 20 yr ago, Frisch (11) suggested that the level of body fat could in some way trigger initiation of reproductive function. However, it was difficult to envision how the brain could sense the degree of adiposity. Potential signals that might link growth, adiposity, or metabolic status with reproductive development have included: the amount of calories (13); the level of oxidizable metabolic fuels (14); availability of glucose (15); the level of hormones, such as insulin (16); and specific amino acids (12). Although useful insights emerged from these studies, it was clear that the discovery of the adipose-derived hormone leptin (17) marked a major advance in the field. In the fed state, circulating levels of leptin and leptin messenger RNA levels are closely correlated with degree of adiposity (18), and with caloric restriction, leptin levels fall rapidly (19). Thus, leptin is a potential signal to the brain reflecting both energy stores and energy balance. Furthermore, messenger RNA encoding the predominant signaling isoform of the leptin receptor has been localized within the hypothalamus of all species studied so far. More importantly, these receptors colocalize with several neuropeptides thought to be important for controlling both food intake and reproduction, and leptin either activates or inhibits these neurons (20, 21). Thus, leptin is ideally situated to serve as a signal linking metabolic status and brain function.
One early indication that leptin might have an impact on reproduction came from the observation that ob/ob mice (which lack functional leptin) or db/db mice (which lack functional leptin receptor) are infertile and fail to undergo normal sexual maturation (22). Administration of leptin restores fertility to the ob/ob mouse (23, 24). Similar findings have also been reported for humans (25, 26, 27). Importantly, fertility of ob/ob mice is not restored by simply reducing body weight, indicating an effect of leptin per se on reproductive function. In keeping with its predominant role as a signal for starvation (19), leptin also seems to be important in mediating undernutrition-induced deficits in reproductive function. Evidence from several species demonstrates that leptin levels fall during fasting or food restriction in conjunction with decreased levels of LH and a delayed return to normal estrous cycles. In addition, several laboratories have shown that administration of leptin during food restriction, either intracerebroventricularly (ICV) or ip, restores LH release and normal estrous cycle periodicity (28, 29, 30, 31, 32, 33). Thus, it is established that leptin is critical for normal reproductive function and that the fall in leptin levels during food restriction is an important component of the impaired fertility that accompanies this state.
Whereas there is agreement on a critical role for leptin in regulating reproduction, at issue has been the specific role that leptin plays in regulating pubertal onset. That is, is leptin merely permissive (i.e. basal levels are required) or does it provide, through a rise in circulating levels at some key point, a "signal" for triggering pubertal onset. In this context, Foster and Nagatani (9) have defined a pubertal signal as one that changes between the sexually immature and sexually mature state in a way that is temporally meaningful to the initiation of puberty, and when administered at the appropriate time alters onset of puberty. A number of studies have attempted to address the first aspect of this definition by examining developmental changes in leptin. Although some studies in rodents (34) and humans (35) have reported rises in leptin levels before prepubertal increases in estradiol or testosterone release, other studies have been unable to correlate changes in leptin with puberty (29, 36, 37, 38, 39). It is important to realize, however, that although discrete events such as menarche or vaginal opening are often used to define puberty onset, in reality puberty is the culmination of a prolonged series of prepubertal events, making it inherently difficult to assess temporal relationships between changes in putative signals, such as leptin, and puberty. Prepubertal changes in LH secretion, which are necessary for onset of puberty in mammalian species, often occur well in advance of vaginal opening or menarche. Thus, it may be more appropriate to examine the relationship between developmental changes in leptin secretion and prepubertal changes in LH secretion. These experiments have not been performed in any species until now.
In this issue of the journal, Suter et al. (40) report that nocturnal levels of leptin increase just prior to the nocturnal, prepubertal increase in pulsatile LH release. In two earlier reports it was concluded that leptin secretion in primates changed little during the peripubertal period when LH, FSH, and testosterone begin their ascent to adult levels (38, 39). One possible explanation for the differing results is that in the previous studies, data were derived from a limited number of daytime samples. This is a critical difference given the fact that, in primates, prepubertal changes in nocturnal LH release can occur up to 5 months before increases in daytime LH release (41, 42). Another possible explanation is that the earlier studies used normal intact or castrated males, whereas Suter et al. (40) used a castrated male model wherein the pituitary was sensitized to endogenous GnRH stimulation by intermittent, exogenous GnRH administration. This model has the advantage that it allows for a more sensitive assessment of prepubertal increases in GnRH release (by measuring LH release) independent of any influence of differences in pituitary sensitivity (43, 44). Suter et al. (40) also observed changes in insulin-like growth factor I and GH levels before puberty. Although correlative in nature, these data nonetheless are consistent with the possibility that peripubertal changes in nocturnal leptin, along with other possible signals, act either independently or in concert to initiate events that ultimately culminate in puberty. The peripubertal appearance of a nocturnal leptin peak has recently been observed by Nagatani et al. (45) in the rat. Clearly, much additional data are needed to clarify this important issue. Furthermore, it is clear that in studies of this type, careful attention must be given to the sampling paradigm used. Also, regarding the role of leptin, differences in species may be operative, as may differences between males and females.
A number of studies have attempted to satisfy the second aspect of the definition by examining whether leptin administration could accelerate puberty onset. Results from these studies have been variable. Ahima et al. (46) reported that leptin administered after weaning advanced the age at vaginal opening, vaginal estrus, and onset of consistent estrous cycles, changes accompanied by increased uterine and ovarian weight. In contrast, studies in rats failed to show an ability of leptin to accelerate normal pubertal onset (28, 29), even though leptin did prevent the delay in vaginal opening induced by chronic food restriction in this species (29). Apart from the species difference, the explanations for these differences are not readily apparent. As mentioned above, it should be remembered that the search for a role of leptin as a "pubertal trigger" must take into consideration when the triggered event takes place (which may be different from onset of secondary sexual characteristics, menarche, or vaginal opening) the sampling paradigm used, and the time of day when samples are taken. For example, if a critical rise in leptin occurs at postnatal day 10 in the rodent, then administration of leptin postweaning (day 21) may be an insensitive means to assess the role of leptin as a signal of puberty. Development of a leptin receptor antagonist will be important for addressing the physiological relevance of leptin as a signal that initiates puberty as well as determining the relative importance of the early postnatal leptin rise in mice (34).
Although leptin clearly influences reproduction, where leptin acts to exert its effects is not yet resolved. Much data supports the notion that the reproductive actions of leptin involve actions in the brain and, more specifically, the hypothalamus. The long form of the leptin receptor that is responsible for signal transduction is heavily expressed in the arcuate nucleus and ventromedial hypothalamus (47), areas important for controlling GnRH release and sexual behavior, respectively. Leptin stimulated GnRH release from isolated hypothalamic explants in vitro (48), and ICV administration of leptin antibodies reduced pulsatile LH release (49). Also, ICV administration of leptin at doses that did not influence peripheral leptin concentrations restored LH secretion during fasting in rats (29). Collectively, these data suggest that leptin acts centrally to influence reproduction, but leave open the question of whether these actions are exerted directly on GnRH neurons or indirectly through interneural inputs. To date, leptin receptors have not been identified on GnRH neurons, favoring the idea of neuronal intermediaries of the actions of leptin on GnRH release. The identity of neural systems that may link leptin and GnRH release is unknown, but recent data from NPY knockout mice (50) and MC4 receptor knockout mice (51), both of which are fertile, would indicate that these two systems are not critical for mediating the reproductive effects of leptin.
It has been suggested that the influence of leptin on reproduction might involve actions outside the brain, as well. Presence of leptin receptor and actions of leptin on both the pituitary (48, 52, 53, 54) and gonads (52, 55, 56, 57, 58) have been described. In addition, according to one view, the effect of leptin on fertility may involve alterations in intracellular metabolism or metabolic fuel supply or sensing at extrahypothalamic sites. Combined administration of 2-deoxyglucose (2DG), a competitive inhibitor of glucose utilization, and methyl palmoxirate, an inhibitor of fatty acid oxidation, at doses which when delivered separately are without effect, disrupts estrous cyclicity in hamsters (4). Administration of higher doses of 2DG alone reduces pulsatile GnRH release in sheep (59), pulsatile LH release in rats (60) and decreases C-fos expression in hamster GnRH neurons (61). Not unexpectedly then, normal GnRH neuronal activity is dependent on sufficient energy availability. Because there are only a few papers that have attempted to examine the relationship between leptin and metabolic fuel availability, however, it is difficult to evaluate the degree to which these pathways interact. In fasted hamsters treated with 2DG, leptin given either ip (30) or ICV (31) only marginally improves the number of animals exhibiting estrous cycles, a finding interpreted by the authors to mean that leptin is unable to overcome blockade of peripheral metabolic fuel supply. However, in these studies fasting alone can reduce estrous cycle expression, and thus, addition of 2DG may result in an extreme state of metabolic distress, which leptin may not be able to overcome. In addition, the nonspecific effects of a general metabolic inhibitor like 2DG further complicate interpretation of the data. Additional study will be necessary to clarify actions of leptin in the periphery and to define the relative importance of these actions in regulating reproduction.
To summarize, the discovery of leptin has illuminated a fundamental link between nutrition and the biology of reproduction that was not clearly anticipated by most workers in the field. We can now state unequivocally that leptin is an adipocyte-derived hormone, sufficient levels of which are critical for normal sexual development and fertility, and falling levels of which play a necessary, and possibly sufficient, role in mediating fasting- or feed restriction-induced deficits in reproduction. Further work will be required to more fully elucidate the role that leptin plays in pubertal onset. Evaluation of leptin as a signal for pubertal onset should likely focus on the relationship between leptin and developmental events such as increased LH release. As shown by Suter et al. (40) in this issue, special attention should be given to the time during development when samples are collected as well as the sampling paradigm to be used. It will also be necessary to more fully define the relative importance of the central vs. peripheral effects of leptin on reproduction. Whatever the outcome, however, it is clear that future studies of reproductive regulation must take account of this hormone.
Received December 14, 1999.
Accepted December 16, 1999.
![]() |
References |
---|
![]() ![]() ![]() |
---|