| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Reproductive Endocrine Unit Massachusetts General Hospital Boston, Massachusetts 02114
Address all correspondence and requests for reprints to: Stephanie B. Seminara, Massachusetts General Hospital, Reproductive Endocrinology Unit, Bartlett Hall Extension 505, 55 Fruit Street, Boston, Massachusetts 02114. E-mail: seminara.stephanie{at}mgh.harvard.edu.
GnRH neurons, although few in absolute number, constitute a hypothalamic pulse generator, releasing their decapeptide hormone in discrete discharges (1). How these diffusely distributed neurons of the parvocellular system achieve coordinated pulsatile activity is not understood. How these neurons engage in characteristic patterns of activity throughout the reproductive life cycle is even more mysterious. In the human, the GnRH pulse generator is active in fetal life, driving the fetal pituitary by the second trimester (2, 3). Although its activity continues for several months after birth, the pulse generator diminishes in the first year, entering a lengthy period of relative quiescence modified by decreasing the amplitude but not frequency of the pulse generator (4) until the system once again amplifies, heralding puberty. Understanding the many signals that modulate the GnRH neuron is key to understanding the ontogeny of the GnRH pulse generator, and by extension, the timing of human puberty.
The quest for such factors has lead to the study of several molecules, some of which have been classified as "stimulatory" to GnRH neurons (i.e. glutamate) (5); others, inhibitory (i.e. 
-aminobutyric acid) (6). Peripheral metabolic signals, as communicated by leptin, are another category of neuronal input to the GnRH neuron (7). However, decades of investigation into these maturational triggers have not diminished the intensity of the search for additional modulators of GnRH secretion; the years have merely changed the tools for asking these questions. Motivating these searches is the fact that several reproductive disorders are characterized by abnormalities in the pattern of GnRH secretion, including hypogonadotropic hypogonadism, a condition characterized by an absence of normal pubertal development. Indeed, recent years have attested to the critical nature of genetic tools in patients with hypogonadotropic hypogonadism as a prism to understand GnRH physiology (8, 9, 10, 11).
Despite the successes of novel gene discovery in recent years [i.e. FGFR1 (fibroblast growth factor receptor-1), GPR54 (G protein-coupled receptor 54), PROK2 (prokineticin 2), PROKR2 (prokineticin receptor 2)] (8, 9, 10, 11), a tendency persists for many investigative groups to look at any one gene as a single factor, when in reality there are networks of interactions between multiple molecules, in multiple pathways. Today, high throughput technologies are transforming biological research by providing extraordinary amounts of data in several animal species simultaneously. Accordingly, scientists must be able to interpret this information in some integrative way. Systems biology is a field of study that aims to provide this comprehensive view. Rather than focusing on any single gene or protein, systems approaches offer the unique ability to integrate the complex interactions between genes and proteins.
In this issue of Endocrinology, Roth et al. (12) make this conceptual leap using a systems biology approach to develop a hierarchically arranged gene network underlying female puberty in the monkey. These authors have cleverly used a combination of DNA microarrays, "guilt by association"/"retrospective" approaches, and computational methodologies in both rodents and nonhuman primates. Using RNA extracted from the hypothalami of female nonhuman primates, the authors examined gene expression profiles on human cDNA microarrays. Their analyses identified a subset of genes previously involved in the regulation of tumor progression (tumor-related genes = TRGs) that were shown to increase during sexual maturation. Changes in gene expression in the rodent hypothalamus at the time of female puberty were examined in parallel with similar findings. Some of the positive results from the microarray analyses were then validated by real-time PCR and localization of the sites of expression of these selected genes was performed by in situ hybridization. Additional candidate genes were added based on retrospective examination of the literature and guilt-by-association analyses. With this information in hand, the authors built a model of a gene network using in silico analyses (containing five major hubs—CUTL1, USF2, YY1, MAF, and p53) to generate new hypotheses to explain the neuroendocrine systems controlling the onset of puberty in the mammalian brain.
These authors demonstrated about 10% of hypothalamic genes that increase their expression at the time of puberty are genes previously implicated in tumorogenesis or suppression. This biological motif has been observed in a singular fashion with other genes validated to be important for human puberty. For example the proteolytic processing of kisspeptin, a potent regulator of GnRH release, gives rise to metastin, a peptide known to inhibit metastases in vivo in melanoma and breast carcinoma models (13). However, the idea that tumor-related genes might play a more over-arching role in the pubertal process takes this notion much farther and hence is truly novel. Of particular interest in this regard is the well-studied tumor suppressor, p53, which has traditionally been thought of as quiescent until cells become stressed or damaged. The data presented in Roth et al. (12) suggest that, in fact, this tumor suppressor may also play a role in normal reproductive physiology. It is not known whether the activation of certain genes is specific to puberty and independent from their tumor suppressor activities and how the effects of sex steroids might modulate these roles.
While the power of such high throughput approaches and the advantages of the construction of a gene network are both obvious, there are also some limitations to this approach, as with any model. Gene profiling using mRNA by its very definition depends on changes in patterns of gene expression. Not surprisingly, GnRH, the prime mover in the onset of puberty, was not fished out by this approach since levels of its message and protein do not change dramatically during the pubertal transition (14). However, somewhat more concerning is that kisspeptin was also not captured by the arrays despite the fact that levels of this mRNA have been shown to increase across sexual maturation in multiple mammalian species (15, 16, 17). Therefore, readers have to be alert to the limitation that gene discovery using arrays may miss key players that are expressed only at certain windows of development or whose expression across the entire hypothalamus falls below the predetermined sensitivity thresholds for data analysis. Most importantly, although the authors did perform some tissue confirmation of their array data, their construction of a gene network is only a theoretical model until other biological validation can be performed.
Having acknowledged those limitations, the work of Roth et al. (12) is a leap forward. It challenges all of us to master new computational tools, step outside our comfort zone, and place the genes that we study into broader physiological context.
 |
Footnotes |
|---|
Received July 12, 2007.
Accepted for publication July 18, 2007.
|
References |
|---|
 Top References |
|---|
-Aminobutyric acid is an inhibitory neurotransmitter restricting the release of luteinizing hormone-releasing hormone before the onset of puberty. Proc Natl Acad Sci USA 91:395–399
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
