Introduction

Sex steroids are steroid hormones, synthesized from cholesterol, that share the distinctive characteristic of being produced in three separate sites in the body: the gonads, the adrenals, and the brain. These sex hormones can be categorized into subtypes with distinct molecular functions: androgens (e.g., testosterone, dehydroepiandrosterone [DHEA]), estrogens (e.g., 17-alpha and 17-beta estradiol, estrone, estriol), and progestogens (e.g., progesterone, allopregnanolone, pregnenolone).

Multiple lines of research based on experimental animal models suggest that these hormones regulate brain and behavioral trajectories in a sexually dimorphic manner [1–4]. For example, females may benefit disproportionately from the neuroprotective effects of estrogens compared to males [5]. Androgens, in turn, could play an important role in the recovery from demyelinating processes, particularly in males [6]. Progestogens have been shown to dampen abnormal rates of neuronal proliferation and apoptosis after a traumatic brain injury or a stroke, with males appearing to benefit more from these protective effects than females [7, 8].

While there is ample evidence that sex hormones carry critical, sexually dimorphic, functions during neurodevelopment and in neural as well as glial repair during adult life, the existing evidence relies overwhelmingly on animal data [9–12]. Therefore, it remains to be clarified to which extent and through which mechanisms sex steroids affect the human brain. Numerous studies have tried to elucidate the link between sex steroids and sex differences in brain structure, but a plethora of contradictory findings have emerged at every level of development [10, 12]. For example, patients suffering from schizophrenia and depression have been observed to exhibit both increases and decreases in sex steroids, compared to controls [13–20]. These conflicting findings suggest that other factors modulate CNS effects of sex steroids in humans. For instance, these central effects may be related to the complex interactions between developmental timing, individual genotypes, and other hormones (e.g., cortisol, oxytocin, and vasopressin) [21].

Recent progress in the fields of neuroendocrinology, neurogenetics, and neuroimaging may hold the key to identifying which of these genetic, hormonal, and developmental factors significantly contribute to brain and behavioral changes. Here we will discuss the interplay between different sex steroid hormones, the critical timing of CNS sex steroid effects during the prenatal period, and the interactions between sex steroids and sex steroid receptor genes, neurotransmitters, and other hormonal systems (Fig. 1).

Fig. 1
figure 1

Effects of sex steroids: a multimodal perspective: Figure 1 outlines the different sections of this review and how they relate to the examination of sex steroid levels, neural and glial changes related to sex steroids, sexual differentiation of the brain and sex-specific vulnerability to psychiatric illness. All images labeled for non-commercial reuse with modification; courtesy of en.wikipedia.org; commons.wikimedia.org; pixabay.com; suttonhoo.blogspot.com; and clipartse.com

Full size image

Metabolomics of the Sex Steroid System

Metabolomics is the study of small-molecule metabolite profiles [22]. Metabolic profiles contain elements specific to each individual, which constitute a metabolic “signature” akin to a genetic fingerprint [23]. The sex steroid metabolome can give a comprehensive picture of the overall balance of androgens, estrogens, and progestogens pertaining to a specific individual [23]. In turn, the net effect of each sex hormone may depend on the overall balance of steroid hormones [22–25]. The sex steroid metabolome can be measured in two different ways through the collection of: (1) peripheral fluids/tissues (blood, urine, etc.) or of (2) central fluids/tissues (cerebrospinal fluid (CSF), brain tissue).

Sex Steroid Metabolome: Peripheral Nervous System

Studies comparing CSF sex steroid levels to serum levels suggest reasonable correlations between peripheral and central metabolomes [26, 27]. Androgens such as DHEA, androstenedione, and testosterone exhibit moderate to high CSF-to-serum correlations (r = 0.3–0.9) [26, 27], and active metabolites of progesterone such as allopregnanolone show high correlations (r = 0.5–0.8), supporting the notion that there is significant transport of peripheral steroid hormones across the blood brain barrier [26, 27].

The fetal metabolome (sampled from the umbilical vein, artery, and amniotic fluid) was also shown to be closely related to the maternal metabolome (sampled from the antecubital vein) [28, 29], suggesting that maternal sampling may be an adequate method to assess the fetal hormonal milieu during the critical period of prenatal brain development. Another study of treatment-naïve adolescents and young adults with schizophrenia showed lower non-sulfated progesterone metabolites (positive allosteric modulators of GABAA) and 5-beta-reduced metabolites (negative NMDA-R modulators) in patients vs. controls [25], supporting the existence of an overall state of neuronal hyperexcitability in patients with schizophrenia [25]. All steroids show some degree of transfer from the PNS to the CNS through the blood-brain barrier (BBB) [30]. The free portion of circulating steroid hormones and most of the portion (80–100 %) loosely bound to albumin diffuses through the BBB, while the portion tightly bound to sex hormone-binding globulin (SHBG) does not appear to be transported to any significant extent [30]. Therefore, the permeability of the BBB becomes an inverse function of the SHBG affinity of each steroid hormone, with progesterone showing the lowest SHBG affinity (and hence, the greatest diffusion across the BBB) followed by estradiol and testosterone [30]. These preliminary studies thus highlight the relevance of measuring the peripheral sex steroid metabolome during specific developmental windows of increased mental health vulnerability and hormonal instability, such as the prenatal and pubertal stages.

Sex Steroid Metabolome: Central Nervous System

Existing studies of the CNS metabolome have, for the most part, used animal samples. For example, in one study of the central metabolome, investigators showed that steroid receptor coactivator-3 (SRC-3) knockout mice had lower levels of brain amino acids, highlighting the importance of this molecule in maintaining brain protein metabolism [31]. Steroid receptor coactivators are important transcriptional regulators for all sex steroid hormones, and their binding is required to activate the process of gene expression [32]. Thus, results from this study support the notion that an alteration in sex steroid-related processes also correlates with a significant change in the brain’s metabolic signature. Still, several questions remain unanswered regarding the brain sex steroid metabolome and how changes in this metabolome may be related to brain structure and function in humans.

Sex Steroids and the Fetal CNS

The central nervous system is thought to be particularly sensitive to the effects of sex steroids during development [33–35]. Of all developmental periods, the prenatal period is viewed as especially critical, given the brain’s sensitivity to sex steroids during fetal life [36]. While there have previously been technical obstacles limiting the direct examination of sex steroid-related effects in the fetal brain, it is now possible to examine the fetal hormonal brain environment, e.g., by combining prenatal amniocentesis to fetal neuroimaging.

Prenatal Amniocentesis In Vivo

Prenatal amniocentesis is now done in an unprecedented high number of women due to technical advances in obstetrical care as well as socio-economic factors that have led to a steady increase in the age of primiparous women [37]. This test is routinely offered in the second trimester to mothers at increased risk for a fetal anomaly [38]. The second trimester of pregnancy has been shown to be critical for the development of the central nervous system and to coincide with a major surge in one of the main sex steroids, testosterone [39, 40]. Recent studies have shown that fetal testosterone, obtained through amniocentesis, is positively correlated to a variety of cognitive and behavioral measures linked to autistic spectrum disorders, such as decreased eye contact, impaired theory-of-mind and quality of social relationships [41, 42]. Therefore, the determination of sex steroid levels from amniotic fluid may be a viable avenue to characterize the fetal brain’s hormonal environment during a critical period of development. On the other hand, prenatal amniocentesis remains limited by the small but non-negligible risk of adverse fetal outcomes, the inherent bias in the population sampled (at higher risk for chromosomal anomalies), and the stress inherent to the amniocentesis procedure, which may be associated with its own associated hormonal changes [42].

Fetal Neuroimaging In Vivo

In the last decade, fetal neuroimaging in vivo has progressively evolved to allow the characterization of increasingly minute changes in fetal brain structures [43]. For example, Zhang et al. have recently compiled a high-resolution spatio-temporal atlas of fetal brain development showing the rapid growth of the anterior temporal and insular regions and the progressive changes in cortical layers occurring during the second trimester (publicly available at http://www.loni.ucla.edu/Atlases/) [44]. Concurrent with this progress, fetal functional neuroimaging has also benefited from computational advances, with investigators reporting changes in resting-state connectivity in the fetal brain as early as 24 gestational weeks [45] and responses from the auditory and visual systems as early as 28–29 weeks of age, a finding confirmed by other neural functional modalities such as electroencephalography and magnetoencephalography [46]. Finally, proton MR spectroscopy detects and quantifies signals from an array of biochemical metabolites linked to fetal brain development, including markers of neuronal density (N-acetyl-aspartate), membrane and myelin changes (choline), and cellular metabolism (creatine and phosphocreatine) [47–51]. MR spectroscopy may reveal changes in the fetal brain not yet detectable on structural and functional MRI, starting as early as 19 weeks of gestation [47–51]. In addition, advances in motion correction, water signal suppression, and sequence duration now enable the detection of reliable signal peaks in the fetal brain [47–51]. Hence, sampling of the hormonal environment during the late second trimester can now be combined with early third-trimester fetal neuroimaging, with the expectation that the hormonal milieu would serve as a predictor of the dynamic changes in brain chemistry, function, and structure occurring during the prenatal period. While current studies remain limited by relatively small samples sizes and the lack of consensus on data processing and analysis procedures [46, 51], it is hoped that with further technical advances and larger sample sizes, prenatal amniocentesis and fetal neuroimaging could be used simultaneously to illuminate aspects of sex steroid-related brain changes during the second and third trimesters.

Sex Steroids and Sex Steroid Receptor Genes

In addition to metabolomics and developmental timing, localized brain changes related to fluctuations in sex steroid levels are also likely to account for a non-negligible portion of sex steroid-related effects [52–55]. Of the many possible regional brain factors, the ones most directly linked to the effects of sex steroids are the distribution, density, and activity of sex steroid receptors. Indeed, there is a tight relationship between sex steroid levels and sex steroid receptor density and activity, such that upregulation or downregulation of receptors occurs in the presence of variable hormone levels [56]. In turn, sex steroid receptor activity is likely to be modulated by differences in DNA and DNA expression, as well as messenger RNA (mRNA) and protein levels. Because of complex regulatory mechanisms at every step of the genomic process (e.g., targeting of mRNAs for degradation, suppression of translational activity by microRNAs), neither mRNA nor protein levels show a perfect correlation with each other or with the level of transcriptional or translational activity. The measurement of genetic and epigenetic characteristics of genes coding for sex steroid receptors may therefore allow us to draw some inference with regard to their activity in vivo. For example, differences in (1) genetic polymorphisms, (2) pre-transcriptional epigenetics (i.e., reversible modifications to the DNA before mRNA transcription), and (3) post-transcriptional modifications (differential mRNA processing before translation into a protein) may all be involved in the regulation of sex steroid receptor function.

Sex Steroid Receptor Genetic Polymorphisms

Two studies [57, 58] looking at estrogen receptor α polymorphisms (rs9340799, rs2234693) in postmenopausal women have also found effects on gray matter volume of cortical and subcortical structures. In addition, the same estrogen receptor α polymorphisms have been linked to Alzheimer’s disease and to mood/anxiety disorders [59]. These findings support the notion that changes in brain structure associated with estrogen receptor polymorphisms may also be associated with changes in brain function. Still, while polymorphic variation does lead to significant changes in gene expression, they are not the only determinant of transcriptional changes. Additional pre-transcriptional and post-transcriptional modifications further regulate sex steroid receptor gene expression and may amplify or restrict the effect of genotype on expression and, ultimately, on receptor levels and activity.

Pre-Transcriptional Modifications of Sex Steroid Receptor Genes

Pre-transcriptional changes can modulate the rate and extent of transcription from DNA to mRNA. Two main categories of pre-transcriptional changes exist: DNA methylation and histone modifications. DNA expression can be modulated by the addition of methyl residues at the 5′ position of the cytosine rings of a cytosine-phosphate-guanine (CpG) sequence, a process mediated by the activity of DNA methyltransferases [60]. Regions in the DNA rich in CpG sequences (CpG islands) are more susceptible to the actions of DNA methyltransferases and end up being methylated to a variable extent, depending on the cell type and the activity of the DNA methyltransferases within that cell [60]. In turn, the degree of methylation confers to the cell a specific identity and transcriptional sensitivity for some genes [60]. Active chromatin, linked to higher transcriptional rates, is usually associated with unmethylated DNA, while chromatin linked to lower transcriptional rates tends to be associated with methylated DNA [60].

During the perinatal period, methylation changes in the estrogen receptor α gene promoter were shown to be associated with changes in co-activator protein binding and in receptor expression in female rat offsprings exposed to high degree of maternal care [61], supporting the importance of postnatal DNA methylation in the regulation of sex steroid receptor activity [60]. Furthermore, estradiol appears to regulate the activity of DNA methyltransferases, particularly in females [62], though the net effects on estrogen receptor expression have not been fully investigated. Thus, both changes in the methylation of the estrogen receptor promoter and estradiol-related changes in the methylation process of other neurodevelopmental genes may act together to mediate the enduring effects of estrogens on brain structure and function.

Once sex steroid receptor proteins bind their respective steroid hormone, they translocate to the nucleus, where they form large complexes with histone-modifying enzymes, such as histone acetyltransferases (HATs), histone deacetylases (HDAC), and histone demethylases [62]. This large sex steroid-bound complex then has the potential to alter the chemical structure of the bonds between DNA and histones. Histone methylation, on one hand, tends to decrease transcriptional rates, while histone acetylation leads to a loosening of the bonds between the positively charged histone and negatively charged DNA, allowing greater transcriptional rates. Notably, the androgen receptor that forms a complex with a histone acetyltransferase, CBP/p300 [63], and with a histone demethylase LSD1, which decreases methylation and presumably, may lead to a decrease in transcriptional rates [64]. This receptor-enzyme complex then moves into the nucleus, binding to specific genes with an androgen receptor response element, thereby increasing DNA transcriptional rates of these genes through an increase in histone acetylation and a decrease in histone methylation [63]. Further supporting the importance of androgens in modulating the rate and extent of histone modifications, another study found that: (1) neonatal brains of male mice showed increased histone acetylation and methylation compared to female brains and (2) disappearance of this sex difference with testosterone treatment in females [65].

Thus, when looking at the central effects of sex steroids, the important role of epigenetic modifications in early life cannot be ignored. A repertoire of DNA methylation across the entire human genome, using postmortem brain tissue, has recently become available (BrainCloud Methyl, http://braincloud.jhmi.edu/downloads.htm) [66] and represents an opportunity to further investigate the effects of methylation on sex steroid receptor expression. Still, the examination of sex steroids-receptor interactions remains limited by the complex and sometimes unexpected, effects of epigenetic changes in gene expression: for example, there is emerging evidence that DNA methylation may, in some cases, be associated with increases in gene transcription, while the cross talk between different histone modifications may lead to unforeseen effects on gene transcription.

Post-Transcriptional Modifications of Sex Steroid Receptor Genes

Post-transcriptional modifications have emerged as major regulators of receptor abundance and stability. Even at this very early stage in the field of proteomics, mRNA and protein measurements do not show a perfect 1:1 correlation (Pearson’s r from 0.36 to 0.76 depending on the organism, with an average of 0.63 for human studies) [67, 68]. Therefore, other, as yet unaccounted for, factors intervene at the post-transcriptional level to modify sex steroid receptor function and activity. Sex steroid receptors have been implicated in such post-transcriptional modifications through their regulation of microRNAs. Regulatory microRNAs can bind to regular-sized mRNAs and target them for degradation, therefore affecting mRNA stability and half-life [67]. Estrogen receptor α binds to the promoter of specific microRNAs, repressing their expression [69], and also inhibits one of the main pre-processing microRNA enzymes (DROSHA) [70]. In addition, estrogen receptor α appears to self-regulate its own transcriptional response through a subset of microRNA changes [71]. The androgen and progesterone nuclear receptors also appear to regulate different subsets of microRNAs [72]. More than 16,000 microRNAs have now been identified, of which several families are regulated by sex steroid receptors and sex steroid levels (www.mirbase.org [72]). Though the characterization of the entire microRNA transcriptome has yet to be completed, it has become evident that the microRNA signatures of sex steroid receptors, and their interactions with sex steroid levels, will be crucial in determining their net effects on brain structure.

Sex Steroids and Sex Chromosomes

The interaction between sex steroids and sex chromosomes (XX, XY) represents another relatively unexplored area of investigation that combines research into both genetic and hormonal processes. The sex chromosomes contain regions which, unlike the other chromosome pairs, do not undergo recombination and therefore have been under selective pressure to adopt separate, sex-specific evolutionary paths [73]. In keeping with this, the sex chromosomes have been recently recognized as playing an important role in sexual differentiation of the brain. For example, specific types of knockout animal models have allowed investigators to observe the persistence of at least some sexually dimorphic behavior related to XX and XY genotype, even in the near-absence of gonadal hormones [74–76]. Intriguingly, the sex steroid system and the sex chromosome complement appear to be involved in both synergistic and antagonist interactions in the process of sexual differentiation.

Synergistic Effects of Sex Steroids and Sex Chromosomes

Because the research targeting primitive, “hard-wired” behaviors such as sexual function, aggression, and parenting in animals have shown sex steroids and sex chromosomes to be both essential in the expression of the full male or female phenotype [74–76], sex hormones and sex chromosomes have been viewed as synergistic determinants of sexual differentiation. However, it remains unclear how these synergistic effects are mediated. The only direct association between sex steroids and sex chromosomes that has been documented thus far involves the SRY gene (or sex-determining region Y or testis-determining factor), critically involved in the formation of a functional male reproductive system and active gonads, leading to the subsequent testosterone surge that occurs during the second trimester in male fetuses [77, 78]. This prenatal surge, as well as postnatal increases in testosterone during the perinatal and pubertal period, is thought to be responsible for brain masculinization and defeminization in animals, i.e., the establishment of male-specific behavior and the absence of female-specific behavior [79]. Yet, the female phenotype is not only a “default” mode of brain development, as specific levels of estradiol are necessary for the full female phenotype to be expressed in XX animals [80]. The specific genes on the X chromosome contributing to these effects have not, however, been identified [80].

Antagonistic Effects of Sex Steroids and Sex Chromosomes

There may also be specific, antagonistic effects between sex steroids and sex chromosomes. For instance, by removing the Y-linked SRY gene and hence creating genetically male animals (XY genotype) with or without testes, investigators have demonstrated that testosterone protects myelin in the context of detrimental autoimmune responses but that the XY genotype alone seems to exacerbate autoimmune responses [81]. Although this remains to be confirmed, one could speculate that such antagonistic interactions between sex steroids and sex chromosomes also may exist for brain development.

Sex Steroids and Neurotransmitter Systems

Estrogens appear to increase serotonergic tone [82], to increase the release of acetylcholine [82], to increase norepinephrine turnover [83, 84], to potentiate D1 receptors [84, 85], to antagonize D2 receptors [84, 85], to increase glutamatergic activity function [86], and to decrease GABA-ergic activity [87, 88]. These effects of estrogens are thought to underlie, at least in part, their beneficial effects on mood, cognition (serotonin, acetylcholine) [82], and attention (norepinephrine) [83, 84]. In addition, estrogens may decrease the severity of psychotic symptoms through their antagonistic effects on D2 receptors) [84, 85]. Conversely, they may increase addictive behaviors due to their agonistic actions on D1 receptors [84, 85] and increase seizure frequency (glutamate+/GABA−) [87, 88].

Progestogens, in turn, interact with estrogens and have various effects on neurotransmitter systems. While some of the effects of progestogens are similar to those of estrogens, several of their biochemical actions directly oppose those of estrogens. For example, when combined with estrogens, progestogens seem to increase serotonergic activity. On the other hand, allopregnanolone, one of the most potent progesterone metabolites, inhibits the activity of serotonergic neurons of the raphe nucleus, and long-term progesterone treatment downregulates serotonin receptors [89–91]. Also, in contrast to estrogens, progestogens decrease sigma-1 receptor activation, involved in the release of norepinephrine, and increase the activity of monoamine oxidase [89, 92]. Further, progestogens are associated with a decrease in dopaminergic tone in the nucleus accumbens [91, 93]. Finally, progestogens, in particular allopregnanolone, are potent agonists of the GABA-receptor, similar to benzodiazepines and alcohol [94]. Adding another layer of complexity, synthetic and bioidentical progestogens do not seem to have similar effects on the CNS. For example, synthetic progestogens (e.g., medroxyprogesterone acetate) appear to have a detrimental effect on mood, cognition, and attention, though these effects may be mitigated in the presence of high levels of estrogens [95]. In contrast, bioidentical progesterone has been shown to be a potent neurotrophic agent which may prevent the motor, cognitive, and sensory impairments associated with aging [95].

Although there are fewer studies examining the relationship between androgens and neurotransmitters, it seems that these hormones, in particular testosterone, may increase serotonergic tone, though this may be related to its conversion to estradiol and action on estrogen receptors [96, 97]. Testosterone may also boost the effect of noradrenergic anti-depressants agents through its interaction with androgen receptors [97, 98]. DHEA (and its sulfated derivative DHEAS), on the other hand, appears to protect against NMDA-induced neurotoxicity and to act on a variety of receptors including GABAA, sigma-1, and other receptors, as well as direct activation of second-messenger signaling systems [99]. Although much remains to be clarified given the complexity of its actions, DHEA appears to have neuroprotective, anti-oxidant, and anti-glucocorticoid effects, and as such, may decrease the severity of depressive, anxious, psychotic symptoms, as well as the extent of cognitive deficits [99]. Thus, based on their interactions with neurotransmitter systems, androgens may have certain beneficial effects in the CNS, particularly in men [100–102].

Sex Steroids and Other Hormonal Systems

Further adding to the complexity of the sex steroid system is the presence of cross talk between sex steroids and other hormonal systems. Of these, two interactions are particularly relevant to human behavior: (1) the interaction between the hypothalamo-pituitary-gonadal (HPG) axis and hypothalamo-pituitary-adrenal (HPA) axis, involved in stress reactivity, and (2) the interaction between sex steroids and the oxytocin-vasopressin system, involved in the regulation of social and affiliative behaviors.

Sex Steroids and the HPA Axis

Changes in cortisol reactivity are thought to be a measure of the HPA axis reactivity to physical and emotional stress [103]. Females show greater HPA axis reactivity than males, as well as a fluctuation in HPA reactivity across the menstrual cycle, with higher HPA reactivity during ovulation and the luteal phase as opposed to the follicular phase [103–105]. In addition, depressive and anxiety disorders—both more prevalent in women—have been associated with HPA axis dysregulation [106]. Therefore, interactions between sex steroids (i.e., HPG axis) and HPA axis may be particularly relevant given their potential impact on sexually dimorphic psychiatric disorders. Estrogens and progestogens appear to increase HPA axis reactivity by inhibiting the retroactive feedback related to increasing glucocorticoid levels, while androgens appear to be associated with lower HPA reactivity and are thought to be necessary to the maintenance of the HPA axis [104].

In addition, HPA-HPG interactions appear to vary according to the level, timing, and duration of the specific stressors involved. Indeed, under acute, low-stress conditions, both HPA and HPG axes upregulate each other [104, 107]. In contrast, under conditions of high and chronic stress, the HPA/HPG axes have been shown to downregulate each other in a reciprocal fashion at the hypothalamic, pituitary, adrenal, gonadal, and brain levels [104, 107–110].

The clinical significance of HPA-HPG interactions may also vary according to the specific brain region involved. For example, rodent studies suggest that limbic areas show a high density of glucocorticoid and sex steroid receptors [111], which may lead to increased HPA-HPG sensitivity in the limbic system compared to other brain regions. Yet, primate studies have also documented the presence of glucocorticoid and sex steroid receptors in the neocortex and cerebellum [112]. Therefore, HPA-HPG interactions may carry widespread effects over the whole brain, though limbic regions may be particularly sensitive to these interactions. Given the complex relationship between the HPA and HPG axes, and their potential to directly modulate a range of behavioral parameters, such as anxiety, aggression, and depression [103, 106], it is important to take these interactions into consideration when examining the effects of sex steroids on the human brain.

Sex Steroids and the Oxytocin-Vasopressin System

The neuropeptides oxytocin and vasopressin have been identified as major regulators of socio-affiliative behaviors in animals [113], with oxytocin playing a more prominent role in bonding and maternal care in females [114], while vasopressin plays a more prominent role in attachment and pair-bonding in males [113, 115, 116]. Intriguingly, the oxytocin-vasopressin and sex steroid endocrine systems share several characteristics, such as (1) sex differences in receptor expression and hormonal trajectories across the lifespan and (2) the co-localization of sex steroid, oxytocin, and vasopressin receptors in limbic regions, particularly the hypothalamus [117]. In addition, there is evidence that the sex steroid system regulates the activity of the oxytocin-vasopressin system. For example, (1) estrogen and testosterone treatments have been shown to increase vasopressin and oxytocin gene expression as well as oxytocin receptor binding in specific brain regions [118–120]; (2) oxytocin and vasopressin genes both contain an estrogen-response element where estrogen receptors can bind and activate gene expression; and (3) the oxytocin receptor gene contains a half-palindromic element, or DNA sequence, analogous to the estrogen response element, suggesting that estrogen receptors bind to these neuropeptide genes to regulate their transcriptional rates [119, 121]. In sum, the sex steroid system and oxytocin-vasopressin system share sexually dimorphic expression, co-localization in limbic regions, and evidence of co-regulated gene expression, supporting the notion that both systems may act together in regulating socio-affiliative behaviors.

Conclusions

In sum, several challenges have limited our understanding of the effects of sex steroids in the human brain: the ethical and practical natures of these investigations; the lack of systematic, comprehensive assessment of the human sex steroid metabolome; the different developmental trajectories of specific sex steroids; the impact of genetic variation and epigenetic changes; and the plethora of interactions between different sex steroids, sex chromosomes, neurotransmitters, and other hormonal systems. Fortunately, exciting avenues of research lie ahead in these very areas of complexity, and there are some new tools to delve deeper into the multilayered sex steroid system in order to investigate the crucial role it may play in shaping sexually dimorphic human cognition and behavior.