Evidence That a Sex Difference in Neonatal DNA Methylation Organizes Two Distinct Phenotypic Characteristics of Neurons in the Murine Forebrain
Year: 2017
Type: article
Abstract: Perinatal manipulations of the availability of testosterone (or of its aromatized metabolite, estradiol) have long been known to modify adult capacity for expressing female-typical and male-typical courtship behaviors in a wide variety of mammalian species (1). In their seminal paper, Phoenix et al. (2) speculated that the ability of exogenous fetal testosterone to organize male-typical adult courtship capacity (e.g., less lordosis, increased mounting behavior) in genetic female guinea pigs likely resulted from some permanent, testosterone-dependent change in aspects of brain structure and/or function. Beginning more than a decade later, numerous studies using a variety of mammalian species identified sex differences in the number and morphologic and/or neurochemical phenotypes of neurons in specific regions of the forebrain and spinal cord (3,–5). These studies also showed that perinatal administration of testosterone to females organized male-typical features, whereas perinatal deprivation of testosterone (or blocking aromatization of testosterone) in males prevented the masculinization of these neuronal features, resulting in the female-typical phenotype. Over the last decade, considerable experimental effort has been dedicated to identifying the signaling mechanisms whereby perinatal exposure to testosterone (or estradiol formed via neural aromatization of testosterone) organizes male-typical neuronal features. In some instances, sex differences in the size/neuronal number in the forebrain (6) and spinal cord (7) have been linked to testosterone-dependent differences in the expression of genes controlling programmed cell death. More recently, sex differences in perinatal hormone-induced epigenetic modulation of gene expression have been linked to the production of male-typical features of brain morphology and in some cases to correlated behavioral functions (8). Epigenetic regulation involves the modulation of ongoing gene expression through changes in chromatin structure (8, 9). The most commonly studied examples of epigenetic modulation include the acetylation of histone proteins that typically enhances gene expression and DNA cytosine methylation at CpG sites, which typically (although not always) suppresses gene expression. DNA cytosine methylation is promoted by any of three different DNA methyltransferase (DNMT) enzymes. A recent example of the latter type of epigenetic regulation includes the demonstration (10) that global DNA methylation in the medial preoptic area (mPOA) at postnatal day 1 is significantly greater in female than male rats (and in females treated neonatally with estradiol), together with the observation that neonatal intracerebroventricular (icv) administration of zebularine, an inhibitor of DNMT activity, augmented dendritic spine formation in mPOA neurons of female rats up to the level seen in males. This masculinization of mPOA dendritic morphology was correlated with a substantial increase in the ability of adult testosterone treatment to stimulate mounting behavior by these females after ovariectomy. Nugent et al. (10) concluded that the methylation of CpG sites on any of several identified candidate genes that are expressed neonatally in the mPOA reflects the female default status. They speculated that testosterone (or estradiol) acts in males during a critical neonatal period to demethylate these genes, leading to their later expression, which augments dendritic spine formation and the capacity to show male-typical mating behaviors. A study by Mosley et al. (11), published in this issue of Endocrinology, significantly expands our understanding of the role of neonatal DNA methylation as a mediator of the actions of testosterone in masculinizing the phenotype of neurons in the mouse forebrain. Mosley et al. (11) asked whether neonatal blockade of DNMT activity in mice of both sexes would disrupt the differentiation of two established sex differences in the phenotypes of neurons in the mPOA and in the ventrolateral portion of the ventromedial nucleus of the hypothalamus (VMHvl). In the case of the sexually dimorphic nucleus (SDN) of the murine mPOA, more neurons are immunoreactive for the calcium binding protein, calbindin-D28k (CALB), in males than in females (12). In addition, in both the VMHvl and mPOA, more neurons are immunoreactive for estradiol receptor (ER)α protein in females than in males (13). Prior research has shown that the sex differences in both of these phenotypic differences of neurons depend on perinatal testosterone action in males and are not altered by adult sex hormone manipulations (14, 15). In the case of the sex difference in the CALB-SDN, downregulation of the Bax gene, which eliminates programmed cell death, failed to disrupt the observed sex difference (male > female) in the number of mPOA neurons that expressed CALB (16). This outcome implies that a sex difference in programmed cell death cannot account for the observed sex difference in the phenotype of CALB neurons. In their initial experiment, Mosley et al. (11) found that neonatal icv administration of zebularine (which blocks all DNMT enzymatic activity) to female mice augmented the number of CALB-immunoreactive (ir) neurons in the SDN-mPOA up to the level seen in control males when subjects were euthanized at the prepubertal age of postnatal day 26. This outcome in the mouse mPOA resembles that obtained by Nugent et al. (10) in the rat mPOA: in both studies, blocking DNMT activity in the neonatal female brain with zebularine (the same drug dosage was used in the two studies) led to a male-typical mPOA neural phenotype (more dendritic spines in female rats; more CALB-ir neurons in female mice) many weeks later. This result implies that in the mPOA of the female, the Calb1 gene is hypermethylated, leading to the suppression of its expression. This result also implies that testosterone (or its neural metabolite, estradiol) acts neonatally in the male mPOA to block DNMT activity, leading to the heightened expression of Calb1 and the eventual appearance of more CALB-ir neurons in the male vs female SDN mPOA. Interestingly, in the study by Mosley et al. (11), neonatal zebularine further augmented CALB expression in male subjects above the level seen in control males. This result implies that some substantial degree of methylation of the Calb1 gene persists in normal testes-intact males after perinatal exposure to male-typical levels of circulating testosterone. More research is needed to identify the relevant regulatory mechanism. Another important result of the Mosley et al. (11), study is the observation that neonatal blockade of DNMT activity with zebularine elevated the number of ERα-ir neurons in both the VMHvl and mPOA of male mice up to the level seen in control females, thereby eliminating the sex difference in Esr1 expression. Note that in contrast with the effects of zebularine on Calb1 expression in males, this treatment in females failed to augment Esr1 expression above the level seen in control females. These results imply that hypermethylation of the Esr1 gene (leading to suppression of ΕRα in both the mPOA and VMHvl of males) is the normal default state in the male, despite perinatal exposure to appreciable testosterone. This outcome contrasts with the Calb1 gene in one of the same brain regions (mPOA) in which hypermethylation, occurring only in the absence of male-typical perinatal testosterone action, is the default condition in females. It is important to note that Mosley et al. (11) also demonstrated that neonatal treatment with zebularine failed to modulate cell density in the mPOA in mice of either sex euthanized at postnatal day 26. Likewise, there was no effect of neonatal zebularine on the expression in the mPOA of a marker of programmed cell death. These control experiments strengthen the authors’ argument that zebularine’s effects on the number of CALB-ir and ERα-ir forebrain cells, visualized several weeks after drug treatment, were not an artifact of an undetected change in the cellular composition of these brain regions, but instead reflected a change in the expression of the relevant genes in response to drug-induced reduction in their methylation. Additional studies are needed to determine the molecular mechanisms that enable perinatal testosterone (or estradiol formed via neural aromatization) to differentially modify the methylation status of genes that encode proteins that are differentially expressed in the forebrain over a lifetime. A final important result reported by Mosley et al. (11) was that icv treatment with zebularine on postnatal day 0 and again on postnatal day 1 failed to affect labeling for ERα in mice of both sexes that were euthanized only 6 hours after the second injection of the drug. Furthermore, ERα cell density was quite high (no sex difference) in both the mPOA and the VMHvl at this early neonatal age. Other workers (17) have shown that a sex difference in ERα cell density emerges several weeks later, presumably because of a downregulation of Esr1 expression in the male forebrain. Neonatal administration of testosterone to female mice was reported to modify the DNA methylation status of relatively few genes at postnatal day 4, whereas many more genes were found to be methylated in other mice that were euthanized several weeks later (18). McCarthy (19) coined the term epigenetic echo to describe this delayed effect of the earlier, neonatal presence of testosterone. The article by Mosley et al. (11) establishes a role for the same type of epigenetic echo in the differentiation of two distinct sex differences (male more than female for CALB-ir neurons; female more than male for ERα-ir neurons) in the phenotypes of neurons in the mouse forebrain. More research is needed to determine how DNA methylation along with other types of epigenetic modulation of gene expression establish both morphologic and functional differences in neuronal circuitry that cause sex differences in behavior and neuroendocrine function. calbindin-D28k DNA methyltransferase estradiol receptor intracerebroventricular immunoreactive medial preoptic area sexually dimorphic nucleus ventromedial nucleus of the hypothalamus. I thank Stuart Tobet for providing useful suggestions on this manuscript prior to its publication. Disclosure Summary: The authors have nothing to disclose.
Source: Endocrinology
Author: Michael J. Baum
Institution: Boston University
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: bronze
OpenAlex ID: W2620697552