Puberty, A Sensitive Window of Hypothalamic Development and Plasticity

Abstract

Puberty is a developmental period characterized by a broad range of physiologic changes necessary for the acquisition of adult sexual and reproductive maturity. These changes mirror complex modifications within the central nervous system, including within the hypothalamus. These modifications result in the maturation of a fully active hypothalamic–pituitary–gonadal (HPG) axis, the neuroendocrine cascade ensuring gonadal activation, sex steroid secretion, and gametogenesis. A complex and finely regulated neural network overseeing the HPG axis, particularly the pubertal reactivation of gonadotropin-releasing hormone (GnRH) secretion, has been progressively unveiled in the last 3 decades. This network includes kisspeptin, neurokinin B, GABAergic, and glutamatergic neurons as well as glial cells. In addition to substantial modifications in the expression of key targets, several changes in neuronal morphology, neural connections, and synapse organization occur to establish mature and coordinated neurohormonal secretion, leading to puberty initiation. The aim of this review is to outline the current knowledge of the major changes that neurons secreting GnRH and their neuronal and glial partners undergo before and after puberty. Emerging mediators upstream of GnRH, uncovered in recent years, are also addressed herein. In addition, the effects of sex steroids, particularly estradiol, on changes in hypothalamic neurodevelopment and plasticity are discussed.

Puberty is a critical period of development characterized by the acquisition of sexual maturity. In humans, the age of puberty onset varies according to multiple factors, such as genetic background, ethnicity, and gender. The first pubertal signs are generally observed between the ages of 8 and 12 years in girls, with the appearance of mammary buds (thelarche) and between 9 and 14 years in boys, with the onset of testicular enlargement (1). A number of hormonal, neural, and behavioral changes are observed during this period, highlighting the important contribution of neurodevelopment and brain plasticity (2). The hypothalamic–pituitary–gonadal (HPG) axis is the physiologic cascade ensuring gonadal activation and triggering the onset of puberty. The HPG axis, already active during the final months of prenatal life and shortly after birth, remains silent throughout childhood and resumes activity just prior to pubertal changes (3). At the level of the hypothalamus, this reactivation results from the awakening of a complex neuronal network that culminates in the release of the gonadotropin-releasing hormone (GnRH) neuropeptide (Fig. 1) (4, 5). An extensive body of knowledge has identified GnRH as the master regulator of the HPG axis (6, 7). GnRH neurons originate from the olfactory placode and follow a migratory route through the forebrain to settle in discrete hypothalamic regions. Once in place, these neurons project their axons to the median eminence, from which they release the decapeptide GnRH in a pulsatile manner. GnRH enters the hypophyseal portal circulation to stimulate the secretion of the pituitary gonadotropins, luteinizing hormone (LH) and follicle-stimulating hormone, that stimulate the gonads to synthesize and secrete sex steroids and produce gametes. Circulating gonadal sex steroids, responsible for the development of external genitalia along with other physical changes, in turn exert either negative or positive feedback at the level of the hypothalamus to regulate the release of GnRH. Over the past 3 decades, a complex and finely regulated neural network overseeing GnRH activity has been progressively unveiled. Neural elements producing kisspeptin, neurokinin B, glutamate, GABA, and other neuropeptides have been identified as major regulators of pulsatile GnRH secretion. These neuronal systems, together with glial and other non-neuronal partners, undergo a host of morphological, synaptic, and molecular changes during development, particularly as puberty approaches (2, 8). While many aspects of pubertal maturation of the brain are not yet fully understood, particularly within the hypothalamus, this review aims to present the current knowledge of pubertal hypothalamic development and plasticity and to discuss future challenges.

Figure 1.

Schematic depiction of levels of hypothalamic factors involved in the regulation of the GnRH network according to developmental stage. The upper panel (A) shows changes in protein levels of GABA and glutamate receptor components and mRNA levels of neuropeptides involved in regulation of the GnRH network. The central panel (B) shows changes in mRNA levels of Mkrn3 and Dlk1, 2 genes in which loss of function mutations in humans result in central precocious puberty. The lower panel (C) shows GnRH secretion over time during postnatal development.

Human Brain Development and Plasticity During Puberty

In humans, the brain develops from the third week of prenatal life until adolescence. Neurogenesis, gliogenesis, and cell migration are the first embryonic developmental processes. Under the effect of locally produced guidance cues, neurons migrate to their final destination, where they differentiate, grow dendrites and axons, and begin to form synapses (8). This neuronal maturation begins prenatally and continues throughout adolescence. It is well demonstrated, especially in the cortex, that there is developmental overproduction of neurons and synapses. A high level of neuronal cell death occurs prenatally, which is critical for the establishment of effective and functional neural circuits (8). In addition, there is a peak of synapse formation between the first and the second year of life, which is later eliminated in the course of postnatal and pubertal development. This process of refinement is referred to as synaptic pruning and has proven to be essential for efficient information processing in adulthood (9, 10). In parallel, glial cell differentiation and maturation (astrocytes, microglia, oligodendrocytes) take place primarily postnatally, lasting through adolescence (8). In adult brains, astrocytes outnumber neurons and provide several functions, going far beyond trophic support. Oligodendrocytes are particularly essential for neuronal maturation, as they form the myelin sheaths surrounding axons, referred to as myelination, which is a prerequisite for effective neurotransmission (11, 12). Finally, in addition to the intrinsic properties of both neurons and glial cells and neuron–glia interaction, extracellular matrix and cell adhesion molecules have been shown to play prominent roles in brain development by guiding cell migration, neurite outgrowth, and synaptogenesis (13, 14).

Throughout the lifespan, changes in the brain can occur in each of these processes, driven by both genetic and environmental influences. Brain plasticity (or neuroplasticity) refers to the ability of the brain to change its activity in response to intrinsic or extrinsic stimuli by reorganizing its structures, functions, or connections. Interestingly, structural and functional changes are particularly enhanced at the time of puberty (15). Magnetic resonance imaging (MRI) technologies have allowed the assessment of human brain changes across development and puberty. Apart from conventional T1- and T2-weighted images, the most common radiofrequency pulse sequences that highlight fat tissue (T1) or fat and water (T2), a variety of additional magnetic resonance–based techniques have helped to refine our understanding of brain structure and provided additional data relevant to plasticity. For instance, water diffusion, assessed by diffusion tensor imaging, has permitted indirect investigation of the microstructural components of white matter and quantitative myelination (16). Cortical gray matter density, indirectly measuring glia, vasculature, and neurons (including dendrites and synapses), can be described by an inverted U-shaped curve, peaking in late childhood and then declining throughout adolescence (17, 18). This decline temporally correlates with postmortem findings of increased synaptic pruning during adolescence (9). By contrast, cortical white matter, constituted essentially by myelinated axons, increases progressively from birth to adolescence across most brain regions (19, 20). This observation is consistent with postmortem histological evidence of ongoing myelination during this period (21).

While initial studies analyzed brain development in relation to chronological age, more recent studies have shown that the development of several cortical and subcortical regions correlates more specifically with pubertal stages than with chronological age (22, 23). Indeed, independently of age, larger white matter volumes and increased maturation have been observed in postpubertal girls compared with prepubertal girls (16, 24). Likewise, at equivalent ages, girls showing signs of pubertal maturation develop central nervous system (CNS) changes, such as smaller gray matter volumes in several regions of the brain, including the frontal cortex, the amygdala, and hippocampus, compared with girls who had not yet entered puberty (22). These findings strengthen the hypothesis of a direct influence of pubertal timing on brain development and plasticity (2).

Hypothalamic Development and Plasticity in the Regulation of Pubertal Timing

Human studies of neuroplasticity have focused primarily on cortical and subcortical regions, particularly the cortex, hippocampus, and amygdala. To our knowledge, no structural or functional studies have been reported on the human hypothalamus at puberty. The absence of data about hypothalamic structural changes during puberty could in part be explained by the lack of sufficient fine spatial resolution of current neuroimaging techniques, which do not allow precise delimitation of hypothalamic areas and nuclei. New advances in neuroimaging should permit a more detailed exploration of the human hypothalamus across puberty. In this regard, an atlas of the adult human hypothalamus has been described using MRI and optimized protocols using high-resolution functional MRI (fMRI) have been developed (25-28). Diffusion tensor imaging (DTI) followed by fiber tractography has also been used recently in adults to identify fibers passing through specific hypothalamic nuclei (29).

While human studies regarding the mechanisms underlying pubertal hypothalamic neuroplasticity are limited, the use of animal models has allowed a better understanding of the pubertal changes that occur at the level of the GnRH neuron and its neural network. Studies in sheep and primates have considerably enhanced our knowledge of the physiological regulation of pulsatile GnRH secretion and have been thoroughly reviewed elsewhere (30, 31). This review will focus on the regulation of the timing of puberty in rodents in which the majority of studies regarding hypothalamic development and plasticity have been conducted.

GnRH neurons

GnRH neurons have a unique characteristic within the CNS of being born outside the brain in the olfactory placode and migrating through the forebrain to the hypothalamus during embryonic development. After migration, GnRH cell bodies are diffusely scattered within the medial septum, the organum vasculosum of the lamina terminalis, and the rostral preoptic area of the hypothalamus. The number of GnRH cell bodies is relatively low, with approximately 1000 neurons in rodent and 1500 to 2000 in human hypothalami (32, 33). Remarkably, using whole-mount immunolabeling and 3DISCO technology, a recent study identified around 10 000 GnRH cells in the human brain during fetal development, with around 2000 GnRH neurons located in the hypothalamus and 8000 GnRH neurons in extrahypothalamic regions (34). These data corroborate previous observations using in situ hybridization (35). The role of this extrahypothalamic GnRH neuronal population is still unknown and needs further investigation.

Once discretely settled in the hypothalamus, GnRH neurons send extensive projections to the median eminence (5, 36). These projections may extend for long distances (up to 1000 μm) and feature both axonal and dendritic properties (37, 38). Indeed, GnRH projections have spike initiation sites and actively conduct action potentials, and at the same time receive and integrate synaptic inputs along their entire length, thereby regulating the excitability of the neuron. These unique projections have been termed “dendrons” (38).

Although this organization is established before birth, morphological analyses in rodents have revealed remodeling of the dendritic structure and spine density of GnRH neurons across postnatal and prepubertal development. Immature multipolar GnRH neurons located in the rostral preoptic area are submitted to synaptic pruning and a complex dendritic reorganization to achieve unipolar or bipolar arrangements typical of mature GnRH neurons (Fig. 2) (39, 40). Concomitantly, a postnatal increase in the number of spines has been observed at the level of both somal and proximal dendrites (39, 41). Electrophysiological analyses show that GnRH neurons of juvenile mice, although displaying the same passive membrane properties as adult neurons, exhibit an enhanced heterogeneity in their firing properties (42). Interestingly, a recent study evaluating the firing activity of GnRH neurons by extracellular recording of GnRH-green fluorescent protein (GFP) neurons revealed that GnRH neurons are more active during the prepubertal period than in adulthood. Indeed, there is an increase in GnRH neuronal activity during the first 3 weeks of development, followed by a decline to adult levels (43). It is worth mentioning that, if experimentally isolated from the rest of the brain, prepubertal GnRH neurons are capable of generating pulses (44). This suggests that a strong network upstream of the GnRH neuron controls the juvenile restraint of GnRH neuronal activity, which is subsequently counterbalanced by an increase in excitatory influences. How these inhibitory and stimulatory networks develop and mature during the postnatal/prepubertal period and the underlying neuroplasticity of these networks remain unclear. The current knowledge of the major GnRH regulators and their roles in neurodevelopment and neuroplasticity are presented in the following sections.

Figure 2.

Schematic representation of pubertal morphological changes in GnRH neurons and the upstream neural network. During pubertal development, the GnRH neuron undergoes synaptic pruning, dendritic reorganization and increased synaptogenesis. There is an increase in GABAergic and glutamatergic projections on GnRH neurons, as well as an increase in kisspeptin neuron projections. GnRH secretion is also regulated by glial cells including astrocytes and tanycytes. POA, preoptic area; ME, median eminence. RP3V, rostral periventricular area.

Kisspeptin neurons

The large postnatal/prepubertal increase in spine density of GnRH neurons suggests an important increase in excitatory inputs directly at the level of the GnRH neuron in order to activate the HPG axis. A large body of knowledge based on clinical, molecular, and pharmacological studies has clearly demonstrated that kisspeptin is 1 of the most robust activators of GnRH neurons. Kisspeptin neurons are indispensable in the initiation of puberty and in the maintenance of a fully functional HPG axis throughout adult life (45-48). Mutations in genes encoding kisspeptin (KISS1) or its receptor (KISS1R) in humans, as well as deletion of these genes in mice, led to severe GnRH deficiency and failure to enter puberty (49-51). In rodents, kisspeptin neurons are prevalent in 2 distinct hypothalamic regions, the rostral periventricular area (RP3V) and the arcuate nucleus (ARC), respectively involved in the generation of the female LH surge and the activation of the GnRH pulse generator in both sexes. The mutual role of these 2 distinct kisspeptin populations in the control of puberty onset is, however, still debated. Although variable according to the species, a postnatal increase in Kiss1 mRNA and kisspeptin protein expression has been detected in both regions, with a more marked increase observed in the female RP3V (Fig. 1) (52). The pubertal regulation of Kiss1 expression in the ARC is tightly regulated by both repressive and stimulatory epigenetic pathways, particularly involving the Polycomb group (PcG-repressor) and the Trithorax group (TrxG-facilitator) complexes (53). It is important to note that a third population of kisspeptin neurons has more recently been identified in the medial amygdala, and appears to contribute to the timing of puberty (54, 55). A pubertal increase of Kiss1 expression has also been reported in this region (56).

GnRH neurons receive kisspeptin projections and express Kiss1r mRNA and kisspeptin receptor protein as early as embryonic day (E)13.5 in mice (57, 58). The percentage of GnRH neurons expressing Kiss1r increases markedly across postnatal development, from ~40% at birth to ~70% by postnatal day 20 (PND20) (59). Kisspeptin projections to GnRH neurons also increase between PND25 and adulthood (Fig. 2) (52). In addition, electrophysiological analyses reveal a significant increase in the percentage of GnRH neurons responsive to kisspeptin stimulation with pubertal maturation, from ~25% in juvenile to >90% in adult mice (60). Together, these findings reveal the presence of important morphological and functional neuronal changes within the kisspeptin/GnRH system across pubertal maturation.

KNDy neurons

In the rodent ARC, kisspeptin, neurokinin B (NKB) and dynorphin are expressed in the same neurons, referred to as KNDy neurons. The KNDy neuron population is essential in order to coordinate GnRH pulses (61, 62). In adult mice, this process appears to be regulated through autocrine and paracrine regulatory loops involving either stimulatory (kisspeptin and NKB) or inhibitory (dynorphin) actions (61). Beyond kisspeptin, NKB has been shown to be another major stimulator of GnRH and the HPG axis. Its role is supported by evidence of GnRH deficiency and pubertal failure in patients with loss-of-function mutations in NKB (encoded by TAC3 gene) and its receptor (TAC3R). In contrast, Tac3 knockout in female mice results in pubertal delay but maintenance of fertility (63, 64). In rodents, Tac3 and Tacr3 expression increases in the ARC during pubertal maturation (Fig. 1) (65, 66). The pubertal control of Tac3 expression is regulated, like Kiss1, by an epigenetic switch from transcriptional repression to activation (53). Several studies have also revealed the involvement of neurokinin A and substance P, belonging to the family of tachykinins, in the modulation of pubertal timing in rodents (67, 68). How this tachykinin neuronal system develops during the postnatal/prepubertal period and its underlying synaptic plasticity remains to be elucidated.

GABAergic neurons

Apart from kisspeptin and KNDy neurons, GnRH neurons also receive significant synaptic inputs from both GABAergic and glutamatergic neurons, the principal inhibitory and excitatory systems in the adult brain (Fig. 2). These 2 systems have been shown to undergo significant modifications across postnatal and pubertal brain maturation (11, 69, 70). GnRH neurons express functional GABA (GABA_A and GABA_B) and glutamate (NMDA and AMPA) receptors (71-73). Little is known about the locations of glutamatergic and GABAergic cell bodies that project to GnRH neurons, but the RP3V seems to represent a primary site (74). In addition, kisspeptin neurons in the ARC appear to be mostly glutamatergic, while those in the RP3V are predominantly GABAergic and thereby account for some of the projections (75, 76).

GnRH neurons express functional GABA_A receptors from the earliest stages of embryonic development (Fig. 1) (77). GABA_A receptors are pentameric chloride channels composed of specific α, β, γ, and δ subunits. A postnatal increase in GABA_A receptor sensitivity is observed in conjunction with the maturation of the GnRH neuron. Indeed, electrophysiological analyses showed that juvenile animals required up to 10-fold higher GABA concentrations than adults to establish a concentration–response curve and displayed a reduction in the level of membrane depolarization in response to GABA compared with adults (78). This increase is accompanied by a prepubertal reorganization of GABA_A receptor signaling, with a change in receptor subunit composition and consequently a change in cell signaling properties (78, 79). A recent electrophysiological study revealed an increase in GABAergic transmission frequency in GnRH neurons during the first 4 weeks of postnatal/prepubertal development (80). Although heavily debated, the prevailing view supports an inhibitory action of GABA, through GABA_A receptors, on adult LH secretion. Nevertheless, surprisingly both excitatory and inhibitory actions of GABA_A receptor activation on adult GnRH neurons have been reported. This appears to be dependent on a variable concentration of chloride (Clˉ) in the intracellular milieu of the GnRH neuron (81). It is important to note that in most brain regions, GABA signaling typically shifts from being depolarizing (with excitatory effects) in early postnatal stages to hyperpolarizing (inhibitory) in adulthood. This effect is seemingly mediated by GABA_A receptors and is in part explained by a decline in intracellular Clˉ levels during postnatal life (82). Differential expression of the sodium potassium chloride cotransporter 1 (NKCC1) and the potassium chloride cotransporter 2 (KCC2) contribute to this decline. The excitatory property of early GABAergic projections seems to play an important role in neuronal differentiation and dendritic arborization (83). Within the hypothalamus, it is currently recognized that GABA_A receptor activation exerts a predominantly depolarizing and excitatory response on embryonic and prepubertal GnRH neurons (84, 85). However, the depolarizing/hyperpolarizing switch of GABA action, observed in most brain regions, still needs to be demonstrated in GnRH populations (86, 87).

Glutamatergic neurons

Glutamatergic stimulation of GnRH neurons is considered another essential factor in the reactivation of GnRH secretion at puberty. An increase in glutamatergic transmission is observed during postnatal development in many brain regions, including the hippocampus and the hypothalamus (11, 69, 70). In rodents, GnRH neurons express the three types of ionotropic glutamate receptors, the α-amino-3-hydroxyl-5-methyl-4-isoxazolepropionic acid (AMPA) receptor, the N-methyl-D-aspartic acid (NMDA) receptor, and the kainate receptor. Neuroanatomical studies showed that GnRH neurons express all 4 AMPA receptor subunits (GluR1-4), the NR1 and NR2 NMDA subunits, as well as both GluK1 and GluK5 kainate subunits (71). The specific combination of subunits determines the biophysical and pharmacological receptor properties. GnRH neurons appear to express NMDA and AMPA receptors at a later stage of postnatal development than GABAergic receptors, which are expressed since embryonic development. In GnRH neurons, a progressive increase in expression of the NMDA receptor NR1 subunit is observed during postnatal/prepubertal development in mice and rats (Fig. 1) (70, 88). One study in rat also showed a pubertal increase in expression of the kainate receptor GluK5 subunit (89). In addition, electrophysiological analyses revealed an increase in the percentage of GnRH neurons responding to AMPA in adult compared to prepubertal (PND18-25) mice. In contrast, the number of GnRH neurons responding to NMDA appeared relatively stable between these 2 groups (90). Altogether, these findings suggest that, similarly to other regions of the brain, a delicate interplay of GABAergic/glutamatergic transmission occurs at the hypothalamic level, and more specifically at the level of the GnRH network. During the prenatal period, there is formation of excitatory GABAergic synapses, followed by progressive postnatal development of glutamatergic synapses (5, 70). The progressive increase in excitatory transmission from these 2 neurotransmitters on GnRH neurons is likely to play a critical role in remodeling the synaptic network required for the proper control of the onset of puberty.

Glial cells

Beyond neuronal regulation of GnRH secretion, accumulating evidence exists for a critical role of glial cells in the regulation of brain plasticity during puberty. Within the hypothalamus, astrocytes and tanycytes have been shown to play a role in regulating not only dendritic plasticity but also cell adhesion, cell–cell communication, and axon guidance (91). Glial regulation has been demonstrated in both the preoptic area, where GnRH cell bodies reside, and the median eminence, where GnRH projections and terminals are located (Fig. 2) (91).

Several astrocyte signaling molecules have been identified to contribute to puberty initiation, including prostaglandin E2 (PGE2), tyrosine kinase receptors belonging to the erythroblastosis B (ERBB) family (ERBBs), and synaptic cell adhesion molecule 1 (SynCAM1) (91). In rodents, an estradiol-induced increase in hypothalamic PGE2 synthesis occurs during prepubertal development and triggers the first preovulatory GnRH surge (92). The PGE2 release requires the activation of ERBBs, known to be expressed in hypothalamic astrocytes that physically interact with GnRH cell bodies. The expression of ERBB increases during pubertal development (91, 93, 94). In astrocytes, selective genetic disruption of targets involved in ERBB1 and ERBB4 expression or in PGE2 synthesis is able to delay puberty in mice (95, 96). PGE2 has been shown to mediate the postnatal, estradiol-induced increase in dendritic spine plasticity in the preoptic area (97). Whether these changes specifically affect GnRH neurons remains to be determined.

Selective disruption of SynCAM1 in astrocytes has also been associated with delayed puberty (98). SynCAM1 is a synaptic cell adhesion molecule known for promoting synapse formation and enhancing excitatory synaptic transmission. Interestingly, SynCAM1 expression increases in the brain over the first 3 weeks of postnatal development, the major period of synaptogenesis in rodents (99). Within the hypothalamus, SynCAM1 is expressed by both astrocytes and GnRH neurons and has been shown to drive the adhesiveness of astrocytes to GnRH neuron cell bodies (100). These results highlight the importance of cell–cell adhesion and communication between astrocytes and GnRH neurons for proper GnRH neuron maturation.

In the median eminence, GnRH axon terminals are in close apposition to the endfeet of tanycytes, specialized unciliated cells lining the floor of the third ventricle (91, 101). In adult female mice, tanycytes demonstrate a high degree of structural plasticity across the estrous cycle. The retraction of tanycyte endfeet before ovulation allows direct access of GnRH neurons to the pericapillary space, thereby facilitating the release of GnRH into the pituitary portal circulation (102). This process involves PGE2 and ERBB signaling (103). Interestingly, recent studies revealed that this mechanism also involves semaphorins, a family of secreted and membrane proteins first identified as axon guidance cues. This class of molecules, together with their receptors neuropilins and plexins, regulate cell migration by exerting chemoattracting/repulsing properties. GnRH neuron development and migration have been shown to be regulated by semaphorin signaling during embryogenesis (104-107). Their contribution to activation of the HPG axis is underscored by the occurrence of GnRH deficiency in patients harboring deleterious mutations in SEMA3A, SEMA3E, and SEMA7A genes, as well as in their receptors PLXNA1, NRP1, and NRP2 (108-111). Together with PGE2 and ERBBs, endothelium-secreted SEMA3A and tanycyte-derived SEMA7A have been demonstrated to participate to tanycyte/GnRH terminal plasticity in the median eminence at adulthood (112, 113). In effect, in addition to cell attraction/repulsion processes, semaphorin signaling appears to be involved in neuroplasticity, axonal pruning, and synapse formation (114). Surprisingly, this morphological plasticity of GnRH projections at the level of the median eminence has only been studied in adulthood. Whether median eminence plasticity would promote the reactivation of GnRH secretion to contribute to the onset of puberty remains to be elucidated.

Additional partners: multigenic control of puberty initiation

Puberty has been demonstrated to be regulated by hundreds to thousands of genes (115). In humans, a recent genome-wide association study identified 389 independent regions associated with age at menarche (116). Interestingly, a substantial number of these regions encompassed genes linked to neuron development (116). Moreover, whole-exome sequencing studies of families with disorders of puberty such as central precocious puberty and delayed puberty have identified several new genes associated with pubertal timing with putative roles in neurodevelopment (117-120). These data underscore the importance of the early embryonic and prepubertal development periods, in complement to the pubertal period, in controlling the timing of puberty.

Makorin ring finger protein 3 (MKRN3) and Delta-like homolog 1 (DLK1) were the first genes to be identified as GnRH inhibitors. Loss-of-function mutations in both MKRN3 and DLK1 are associated with central precocious puberty (117, 118). MKRN3 belongs to the makorin family of E3 ubiquitin ligases (121, 122). In the rodent hypothalamus, Mkrn3 expression is high from E10.5 to the second week of postnatal development, then decreases sharply before puberty initiation (117, 123-125). Mkrn3 expression has been found in the mediobasal hypothalamus (containing the ARC) and in the preoptic area. Interestingly, Mkrn3 is expressed in kisspeptin neurons in the mediobasal hypothalamus. Moreover, in vitro analyses demonstrated that MKRN3 selectively inhibits KISS1 and TAC3 promoter activity, suggesting an action of MKRN3 upstream of GnRH secretion via regulation of kisspeptin and NKB (125). Increasing evidence shows that E3 ubiquitin ligases, and particularly the makorin protein family, play important roles in neurodevelopment and synaptic plasticity from embryonic stages through adolescence (126-129).

DLK1 is a transmembrane protein and member of the epidermal growth factor-like family, which is involved in cell differentiation and cell fate determination (130). In the mouse hypothalamus, Dlk1 expression has been shown as early as E10.5 and increases progressively during pubertal development (123, 131). Dlk1 acts as a negative regulator of the Notch signaling pathway (132). Interestingly, while the mechanism of action of Dlk1 in the regulation of puberty initiation is still unknown, the Notch signaling pathway has been shown to be critical for neurogenesis and development of the hypothalamus, specifically for the formation of kisspeptin neurons (133). Taken together, these observations suggest potential roles for MKRN3 and DLK1 in the neural development of the GnRH network before puberty initiation.

Whole-exome sequencing studies of families with delayed puberty have also identified additional targets involved in GnRH neuron development, such as immunoglobulin superfamily member 10 (IGSF10) and leucine-rich repeat containing G protein–coupled receptor 4 (LGR4) (119, 120). IGSF10 is a secreted protein involved in the control of early migration of GnRH neurons. Although the presence of specific receptors for IGSF10 on GnRH bodies or dendrites has not been formally demonstrated, lack of functional IGSF10 in zebrafish prevents GnRH neurons from extending projections to the hypothalamus (119). LGR4 has been shown to mediate Wnt/β-catenin signaling, which is critical to the development of GnRH neurons (120). These findings highlight the importance of early developmental processes in the regulation of pubertal timing.

Regulation of Pubertal Neuroplasticity by Sex Steroids

Within the HPG axis, gonadal sex steroids participate in the dynamic control of GnRH secretion through negative and positive feedback regulatory loops that are organized during postnatal and pubertal development. Indeed, sex steroid levels increase in both males and females during postnatal/prepubertal development and, in parallel with steroid-independent mechanisms, have been shown to contribute to the timing of puberty (134-136). Sex steroids have been associated with brain development and plasticity during puberty. Human MRI studies have shown changes in cortical gray and white matter volumes, thickness of the cortex, and hippocampus and amygdala morphology in association with rising levels of testosterone and estradiol (137, 138). There is also considerable evidence that gonadal hormones impact synaptogenesis, dendritic branching, synaptic pruning, and glial function (135).

In the rodent hypothalamus, GnRH neurons do not appear to express the androgen receptor (encoded by AR) or the estrogen receptor (ER) α (encoded by Esr1), the main receptors mediating the reproductive actions of testosterone and estradiol, respectively (139, 140). Nevertheless, while the mechanisms are still unclear, ERβ (encoding by Esr2) has been found in GnRH neurons, and both neural and GnRH neuron-specific ablation of Esr2 result in delayed puberty (141, 142). Remarkably, the dendritic morphological changes observed during postnatal development of GnRH neurons occur independently of gonadal steroid feedback (40). Thus, it would seem that sex steroid regulation of hypothalamic neuroplasticity occurs upstream of the GnRH neuron.

Unlike GnRH neurons, kisspeptin neurons strongly express Esr1 and AR in rodent RP3V and ARC. Moreover, Kiss1 expression in these regions is differentially regulated by sex steroids. Estradiol inhibits Kiss1 expression in the ARC neuronal population involved in negative feedback regulation of GnRH secretion. In contrast, estradiol stimulates Kiss1 expression in the RP3V, which controls the positive feedback regulation of GnRH secretion in females (143). Esr1 deletion in murine kisspeptin neurons produces early signs of pubertal progression, indicating that estradiol exerts a prepubertal restraint on GnRH secretion (144, 145). However, these mice fail to ovulate, probably because of defective kisspeptin neuron maturation in the RP3V. Similarly, ablation of Esr1 from Tac2-expressing cells advanced puberty, consistent with ERα restraint of GnRH activation via the ARC KNDy neurons (146). These findings have contributed to the development of a model, at least in rodents, in which interplay between the ARC and RP3V populations of kisspeptin neurons determines the timing of puberty and overall HPG integrity (147). It has been shown that ERβ also participates in the increase of postnatal Kiss1 mRNA and kisspeptin protein expression in the female RP3V (141). Taken together, these data show that estradiol controls postnatal/prepubertal expression of Kiss1 in the hypothalamus; however, its impact on the formation and plasticity of kisspeptin projections to GnRH neurons remains to be investigated.

While the involvement of glutamate and GABA in mediating estradiol feedback in adulthood has been known for decades, few studies have been devoted to understanding the impact on prepubertal development of GnRH neurons (148-150). A subset of GABAergic and glutamatergic cells in the RP3V and ARC have been shown to express Esr1 (151). Recent findings in mice showed that the specific deletion of Esr1 from Vgat-expressing (GABAergic) neurons did not alter puberty initiation. In contrast, Esr1 ablation in Vglut2-expressing (glutamatergic) neurons caused advanced puberty. Thus, Esr1-modulated glutamate transmission appears to be a critical component of the estradiol-induced prepubertal restraint on GnRH neuron activity (151). Further studies are needed to better understand the contribution of estradiol to the postnatal increase in glutamatergic synaptic transmission to GnRH neurons and to the GABAergic synaptic changes occurring postnatally.

In addition to neuronal cells, estradiol also regulates the function of glial cells, which also express estrogen receptors (152). Within the hypothalamus, an increase in the synthesis of astrocytic PGE2 correlates with the pubertal increase in estradiol levels (92). Similarly, the morphological plasticity of the tanycytes observed across the estrous cycle in female is estrogen dependent, and both SEMA3A and SEMA7A expression are regulated by ovarian steroids (91, 112, 113).

Additional studies are needed to understand whether gonadal steroids might also regulate other partners of the HPG axis. While no studies have yet been conducted on the newly identified regulators of puberty, DLK1, IGSF10, and LGR4, recent evidence shows that hypothalamic Mkrn3 expression is not regulated by sex steroids. Indeed, Mkrn3 mRNA levels in the ARC and RP3V were unchanged in the GnRH deficient hypogonadal hpg mouse, which carry a genomic rearrangement deleting the Gnrh1 locus. Moreover, treatment of PND11 female mice with estradiol did not modify hypothalamic Mkrn3 expression (125). Further investigation is needed to better understand the effects of sex steroids on hypothalamic neurodevelopment and neuroplasticity.

Conclusion

Puberty is a critical transitional developmental period during which extensive physical, hormonal, neural, and behavioral changes take place. These changes are essential for entering adulthood with reproductive maturity and with cognitive and emotional independence. It is a period of intense cerebral neuroplasticity, finely tuned by numerous factors working in concert to enable proper brain maturation. While our understanding of the neuroendocrine and physiological bases of puberty has improved tremendously in recent decades, the underlying processes impacting neuroplasticity are largely unknown. Both neurons and glial cells are known to contribute to neuroplasticity and regulation of GnRH neuron plasticity is no exception. In comparison with other CNS neurons, GnRH neurons have some distinct characteristics, such as their origin outside the brain, their diffuse distribution throughout the anterior hypothalamus, and their small population. Synchronization of GnRH neurons is critical for correct temporal pacing and for generating coordinated secretory pulses. These observations highlight the importance of high synaptic morphological and adaptive plasticity, to drive the pattern of GnRH secretion and efficiently stimulate pituitary gonadotropes. Interestingly, our current understanding underscores the remarkable morphological changes and neural plasticity that occur during the first few postnatal weeks in rodents, well before the onset of puberty. These and other findings suggest that pubertal development is not simply the result of a neural network awakening at a prespecified time from a previously period of quiescence. Rather, it appears that a broad set of plastic changes occur dynamically and continuously throughout a spectrum of developmental stages, beginning well before birth, to ensure the proper development and maturation of neural circuits to enable the juvenile–adult transition. Across these developmental stages, the fetal, perinatal, and prepubertal periods have been defined as critical windows of plasticity during which neural circuits are highly receptive to genetic, epigenetic and environmental factors. These periods are essential for the organization of hypothalamic circuits that underlie the regulation of the HPG. Thus, the development and maturation of the hypothalamus occurs in conjunction with developmental regulation of the activity of the HPG axis via both internal and external factors. For instance, the activation of GnRH secretion during the perinatal period demonstrates the ability of immature GnRH neurons to be activated. Hence, alongside the maturation of the GnRH neuron itself, there exists a fine regulation of the afferent network that controls the secretion of GnRH. In contrast, the postnatal/prepubertal period of development is a period of high level of plasticity within the hypothalamus, while the HPG axis is maintained in quiescence. These paradoxes highlight the existence of strong inhibitory influences of internal and external factors concomitantly with the development of the hypothalamus. While the mechanisms underlying this inhibitory effect are still unclear, the recent identification of the inhibitory factors Mkrn3 and Dlk1 open a new area of research. Furthermore, we can propose an important role of plasticity events in the modulation of this inhibitory action. In addition, external factors, particularly gonadal steroids, regulate GnRH secretion. The gonads are silent during early postnatal development, thereby participating in the quiescence of the HPG axis until puberty initiation. It is likely that a combination of both steroid-dependent and steroid-independent mechanisms govern puberty initiation. It is worth noting that the model proposing these changes in plasticity apply primarily to mice and rats, as most studies have focused on rodents. It is possible that these mechanisms may differ across species. For instance, while prepubertal negative feedback by sex steroids on GnRH neurons is prominent in rodents, steroid-independent mechanisms seem more important in primates (30). Studies of hypothalamic neuroplasticity during puberty are lacking in humans. Further investigations are needed to evaluate whether neurodevelopment and neuroplasticity processes may account in part for the sex-based differences in puberty initiation and more generally for the regulation of the HPG axis.

Glossary

Abbreviations

AMPA

α-amino-3-hydroxyl-5-methyl-4-isoxazolepropionic acid

ARC

arcuate nucleus

CNS

central nervous system

DTI

diffusion tensor imaging

ER

estrogen receptor

GnRH

gonadotropin-releasing hormone

HPG

hypothalamic–pituitary–gonadal

LH

luteinizing hormone

MRI

magnetic resonance imaging

NKB

neurokinin B

NMDA

N-methyl-D-aspartic acid

PGE2

prostaglandin E2

PND

postnatal day

RP3V

rostral periventricular area

SynCAM1

synaptic cell adhesion molecule 1

Additional Information

Disclosure Summary: The authors have nothing to disclose.

Data Availability

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.