Main

It has been shown recently that some steroid hormones, deriving from progesterone, have a specific role in the CNS. For this reason, they have been called neurosteroids or, better, neuroactive steroids(1). Numerous sites of production for these steroid hormones are distributed throughout the brain (cortex, hypothalamus, pituitary, and hippocampus)(2–6). These hormones may, however, also derive from precursors originating in endocrine glands (gonads and adrenal cortex), and are, therefore, detectable in peripheral blood.

Neurosteroids have a large spectrum of biologic effects: they participate in modulating estrous cycle, pregnancy, stress, depression anxiety, cognitive functions, sexual behavior, mood, memory, development, and aging processes(7–12). This neuromodulatory role is mediated by their capacity to bind GABA receptor sites(5,6).

In particularly, one of the best characterized is 5α-pregnan-3α-ol-20-one, named allopregnanolone. Little information is available on circulating allopregnanolone in human subjects. In adult women, the highest levels have been found during the luteal phase. Moreover, a GnRH or CRF test prompts an increase in serum allopregnanolone levels. These data are indicative of an ovarian/adrenal cortex origin(13). The inverse relationship between serum levels of allopregnanolone and the severity of distressing symptoms that characterize the premenstrual syndrome suggest that it may be a causative factor(14).

Neurosteroids may also play a role in modulating reproductive function. Female rats (when ovulation occurs) manifest higher hippocampal levels of allopregnanolone on proestrus morning and afternoon than on diestrus or on estrus(8), whereas hypothalamic allopregnanolone content is lowest on proestrus afternoon(9).

In addition, the suppression of GnRH-stimulated FSH release in rat cultured pituitary cells by 3α-hydroxy-4-pregnen-20-one(11,12), and its inhibitory effect on ovulation(9), indicates that allopregnanolone may be involved in modulating reproductive function.

The aim of the present study was to evaluate whether serum allopregnanolone levels vary during pubertal maturation in both sexes and whether changes of this neurosteroid correlate with those of other steroid hormones, i.e. progesterone, cortisol, and DHEA.

METHODS

Subjects. Two groups of healthy subjects were included in the study (none were taking any sort of medication): group 1 consisted of toddlers between 0 and 2 y of age (n = 72, 34 boys and 38 girls), coming to the Pediatric Clinic of the University of Modena for growth screening; and group 2 consisted of children between 6 and 18 y of age (n = 82, 44 boys and 38 girls) coming in for annual check-up at the Pediatric Clinic of the Universities of Modena and Pisa. The different developmental pubertal phases were staged according to Tanner's stages (Tanner I = 20; Tanner II = 14; Tanner III = 27; Tanner IV = 14; and Tanner V = 7). Girls with a menstrual cycle were studied during the early follicular phase. This study was approved by the Local Ethical Committees of the Universities of Modena and Pisa.

In each subject, a blood sample was drawn between 0800 and 0900 h. Blood samples were centrifuged and serum was stored at -20°C until assayed.

Assay for allopregnanolone, progesterone, DHEA, and cortisol. Analytical-grade solvents were purchased from Merck (Darmstadt, Germany); C-18 Sep-Pak cartridges were obtained from Waters (Milford, MA). Standard allopregnanolone was purchased from Sigma Chemical (St. Louis, MO) and pregnan-3α-ol-20-one, 5α-[9,11,12,-3H(N)] (45 Ci/nmol) from Amersham (London, UK). The polyclonal antisera, raised in sheep against allopregnanolone carboxymethyl ether and coupled to BSA, were kindly provided by Dr. R.H. Purdy(13). Serum samples (1 mL) were first frozen and then thawed. The assay was performed as previously described(14). The recovery of labeled allopregnanolone was 96.3 ± 4.8% (mean ± SD); after extraction and chromatography, it was 87.5 ± 7.9%; and for entire procedure, the recovery of unlabeled standard allopregnanolone was 77.5 ± 9.8%. The sensitivity of the assay, expressed as a minimal amount of allopregnanolone distinguishable from the zero sample with 95% probability, was 15-20 pg/tube (0.047-0.063 nM) and the intra- and interassay coefficients of variation were 7.2% and 9.1%, respectively. The cross-reactivities of DHEA and progesterone in the allopregnanolone assay are <0.001% and <0.70% respectively. A parallelism test was also performed: a sample containing a high concentration of standard unlabeled allopregnanolone (≅ 2000 pg/mL) was diluted with zero standard.

Serum progesterone and cortisol were determined, after ether extraction and chromatographic partition on C18 Sep-Pak cartridges, by RIA using commercially available kits (Radim SpA, Pomezia, Italy); the sensitivity of the assay was 50 pg/tube (0.16 nM) and the intra- and interassay coefficients of variation were 6.5% and 8.7%, respectively, for progesterone. The cortisol sensitivity was 900 pg/tube (2.5 nM) and intra- and interassay coefficients of variation were 3.7% and 5.8%, respectively.

Serum samples for the determination of DHEA were extracted with ether, purified through a C18 Sep-Pak cartridge, and then assayed by RIA using a trade kit (DSL, Webster, TX); the sensitivity was 15 pg/mL (0.052 nM) and the intra- and interassay coefficients of variation were 3.1% and 6.9%, respectively. The extraction and chromatography step was also used for these steroids to increase the reliability criteria of the specific methods.

Statistical analysis. The statistical analysis of the results was performed using a one-way ANOVA and linear regression analysis. We also performed a Mann-Whitney test and a Duncan test for multiple comparison to confirm statistically significant differences.

RESULTS

Circulating allopregnanolone levels were detectable in all blood venous samples. During the first 2 y of life, serum allopregnanolone levels did not show any significant age-related variations (1 versus 24 mo, p > 0.05) (Fig. 1A). We observed a significant decrease of serum DHEA levels throughout the first 2 y of life, with a significant inverse age-related correlation (p < 0.001) (Fig. 1B). Serum cortisol and progesterone levels showed a significant age-related increase (p < 0.001 and p < 0.05, respectively) (Fig. 1, C and D). Cortisol levels correlated positively with serum allopregnanolone levels (p < 0.01). There were no sex-related differences with regard to these data.

Figure 1
figure 1

Serum allopregnanolone (A), DHEA (B), cortisol (C), and progesterone (D) levels in children in the first 24 mo of life.

Full size image

Serum allopregnanolone levels progressively increased throughout puberty and were correlated to the ages of both girls and boys (p < 0.01 and p < 0.006, respectively) (Fig. 2). Allopregnanolone levels (Tanner I and II) were significantly higher in the first stages of puberty than during the first 2 y of life (0.33 ± 0.11 nM versus 0.23 ± 0.09 nM in boys, p < 0.01; 0.48 ± 0.30 nM versus 0.23 ± 0.13 in girls, p < 0.001). Serum allopregnanolone levels increased in both sexes according with Tanner's stage (0.32 ± 0.02 nM in stages I-II versus 0.7 ± 0.01 nM in stages IV-V, p < 0.01) (Fig. 3).

Figure 2
figure 2

Serum allopregnanolone levels throughout puberty in girls (left) and boys (right).

Full size image
Figure 3
figure 3

Serum allopregnanolone levels at the different Tanner stages (mean ± SEM).

Full size image

During pubertal maturation, both girls and boys showed a significant age-related increase of progesterone and DHEA levels (p < 0.001) (Fig. 4), whereas there were no significant changes in cortisol (Fig. 5). We observed a positive correlation between allopregnanolone and DHEA levels both in girls and in boys (p < 0.008 and p < 0.05, respectively).

Figure 4
figure 4

Serum progesterone (top) and DHEA (bottom) levels throughout puberty in girls (left) and boys (right).

Full size image
Figure 5
figure 5

Serum cortisol levels throughout puberty in girls (left) and boys (right).

Full size image

DISCUSSION

The present study demonstrated that serum allopregnanolone levels do not change in the first 2 y of life, whereas they increase throughout puberty according to age and to the Tanner's stage.

The age-related increase in circulating allopregnanolone during the pubertal period in both girls and boys, suggests a possible role for this hormone in the maturational process of hypothalamic-pituitary-gonadal and hypothalamic-pituitary-adrenal axes. Moreover, it correlated with the age-related increase in progesterone and DHEA levels, thus indicating in ovarian/adrenal cortex origin. The changes in serum allopregnanolone levels throughout the menstrual cycle, with high levels observed during the luteal phase(14,15), also suggest that this steroid is synthesized and released in peripheral blood by the ovary. However, inasmuch as allopregnanolone levels do not significantly decrease after menopause, the adrenal cortex probably contributes to its synthesis as well. In fact, CRF administration prompts an increase in serum allopregnanolone levels(14).

Circulating progesterone, DHEA, and allopregnanolone levels rise around 7 y of age and progressively increase, correlating positively with Tanner's stage, in both boys and girls(16). Therefore, the rise in allopregnanolone levels seems to reflect an increased adrenal steroidogenesis in the development of sexual characteristics during the early stages of puberty. The role of this neuroactive steroid is still undefined. The effect of DHEA on pubertal maturational processes is still debated, but this steroid seems to have a neurobiologic activity(5,6). In fact, the age-related increase in allopregnanolone, a GABA-agonist, and in DHEA, a GABA-antagonist, suggest that neurosteroids may exert a neuroendocrine modulation on pubertal processes. This trend was not observed in infants, inasmuch as there were no variations in allopregnanolone levels, although progesterone and DHEA levels decreased.

One study demonstrated that, after oral administration of micronized progesterone in women, plasma levels of progesterone and its anxiolytic metabolites were related to changes in mood, cognition, and motor performance(17). In fact, significant changes in fatigue and delayed verbal recall and symbol copying were found 1 h after progesterone administration in women whose levels of allopregnanolone and its epimeric steroid were higher. These results suggest that these anxiolytic steroids mediate the behavioral effects of progesterone in women. In physiologic studies, serum allopregnanolone was measured by RIA, and levels were found to fluctuate during the menstrual cycle, with the highest levels in the luteal phase(14). The fact that allopregnanolone levels do not significantly change after menopause suggests that both the ovary and the adrenal cortex may be sources of circulating allopregnanolone. This hypothesis is supported by the observation that allopregnanolone levels increase after short-term GnRH or CRF administration in adult subjects(14). A possible role of allopregnanolone on neuroendocrine modifications throughout puberty, going from prepuberty to adult age, may by hypothesized. A child's psychological maturation may be influenced by neurosteroids, which act on mood, anxiety, libido, and other behavioral functions. The present study suggests that allopregnanolone may be involved in the neuroendocrine mechanisms related to the onset of puberty, which is characterized by major behavioral changes. Children of both sexes show critical variations in mood during puberty. The interconnection between pubertal transition and behavioral problems is well established, and puberty is a time when children are more vulnerable to emotional and behavioral problems(18). Pubertal transition affects psychological distress of adolescents: the onset of puberty itself is a novel event that calls for psychological adaptation, physical maturation, and fertility(19). Increased levels in adrenal androgens seem to have a role in the sexual development during the early stages of puberty(16). Adrenarche is an event of pubertal development. DHEA and dehydroepiandrosterone sulfate (DHEAS) levels induce pubic and axillary hair growth during early puberty. The mechanism for the enhanced androgen secretion during the early phase of puberty remains obscure, but it is clear that this event is also correlated with behavioral changes(20,21).

The possible role of allopregnanolone in these neurobehavioral adaptive phenomena is suggested by the changes in allopregnanolone levels during stress and by its modulatory effects on stress mechanisms(5,8).