Neurosteroids for a successful pregnancy

Abstract

Steroid hormones play a critical role in the initiation and maintenance of pregnancy. In particular, the important role that the progesterone metabolite, and neurosteroid, allopregnanolone, may play in fetal and adolescent development is becoming increasingly evident. Unlike steroid hormones, neurosteroids act at nontraditional targets in the central and peripheral nervous systems, including GABA_A receptor complexes. This commentary discusses the three works in this issue that elucidate the important role of allopregnanolone in the mechanisms that regulate stress hypo-sensitivity of rodents in late pregnancy, neuroprotective effects in fetal sheep exposed to a hypoxic insult, and the continuing role that prefrontal cortex formation of allopregnanolone may play on the cognitive development of gestationally stressed rat offspring, grown to adolescence. The narrative that these works comprise was facilitated by the 5^th International Meeting on Steroids and the Nervous System (Torino, Italy), which is organized to update our knowledge on the relationships between steroid hormones synthesized in different organs and the nervous system. Topics covered in this most recent meeting included sex differences in, and hormonal influences on, cannabinoid-regulated biology; steroids and pain; the importance of co-regulatory factors for steroid receptor action in the brain; mechanism and role of estrogen-induced nonclassical signaling in the brain; vitamin D as the forgotten neurosteroid; neurosteroids and GABA_A receptors; and pathogenic mechanisms mediated by glucocorticoid receptors in psychiatric disorders. The 6^th International Meeting on Steroids and the Nervous System will be held in Torino, Italy in February 2011.

Keywords::

• Parity
• parturition
• delivery
• neuroactive steroids
• allopregnanolone

Steroid hormones play fundamental roles in the development and/or function of the central nervous system (CNS). Traditionally, hormones secreted by the gonads are considered to have both activating and organizing effects on the brain, neuroendocrine function, and behavior. Activational effects of hormones are those changes in brain and/or behavior due to hormones acting in the developed CNS. Throughout the lifespan, mature females experience greater variations in, and higher levels of, progestogens (progesterone and its natural metabolites), than do males. For example, during the follicular phase of the menstrual cycle, progestogen levels of women are low (similar to men); however, luteal phase circulating levels are two- to fourfold higher (2–4 nmol/l/plasma than follicular phase levels (∼1 nmol/l/plasma; Purdy et al. 1990; Wang et al. 1996; Genazzani et al. 1998; Sundström and Bäckström 1998a,b; Freeman et al. 2002; Gracia et al. 2003;). Organizing effects of hormones occur during critical periods of development (typically pre- or perinatal), which result in permanent changes in the CNS structure (i.e. sexually dimorphic areas of the hypothalamus) and/or function (e.g. hypothalamic-pituitary-gonadal or adrenal (HPA) axes; Melcangi and Panzica 2003). During pregnancy, serum levels of progestogens increase, peak in the third trimester (50–160 nmol/l), and reach nadir 1–24 h post-parturition (Sundström et al. 1999; Luisi et al. 2000; Herbison 2001;). The pattern of secretion of progestogens during pregnancy suggests that progestogens may have some organizing effects on fetal programing and neural development of offspring, as well as activating effects for the dam.

While progestogens can be synthesized in the peripheral nervous system (PNS), in gonads, adrenals, corpora lutea, or placenta, they are also produced de novo from sources within the CNS, including neurons and glia (Baulieu 1980, 1991). Centrally derived steroid hormones are often termed “neurosteroids.” Levels of neurosteroids are typically greater in the CNS and PNS than in the circulation. Enzymes involved in peripheral gland steroidogenesis have been identified in the CNS and PNS (Furukawa et al. 1998; Compagnone and Mellon 2000). As well, high CNS and PNS levels of neurosteroids persist after extirpation of peripheral glands (gonadectomy and/or adrenalectomy; Baulieu 1980, 1991; Majewska 1992; Paul and Purdy 1992; Mellon 1994). Thus, in addition to peripheral synthesis and secretion, progestogens can also be produced in the CNS.

Among the most well-studied pregnane neurosteroids is allopregnanolone, which is formed via metabolism of progesterone by the 5α-reductase enzyme, and the subsequent metabolism of dihydroprogesterone by the 3α-hydroxysteroid dehydrogenase enzyme. Unlike its prohormones, allopregnanolone is devoid of affinity for cognate, intracellular progestin receptors, but can effectively modulate membrane-bound neurotransmitter targets, such as the GABA_A receptor. As such, neurosteroids like allopregnanolone have important benzodiazepine-like actions that are associated with reductions in stress responding and restoration of HPA homeostasis (Patchev et al. 1996; Barbaccia et al. 2001).

Maintenance of pregnancy has long been known to be dependent upon elevated levels of “progestational” progestogens. Indeed, progestogens can have suppressive effects on labor and/or stress/HPA responding. For example, the administration of progestins (i.e. synthetic progestogens) reduces the incidence of preterm birth (PTB, which is delivery prior to 36-week gestation) among women (Keirse 1990; Doggrell 2003; Sanchez-Ramos 2005). In a multicenter, randomized placebo-controlled trial of pregnant women with a history of spontaneous PTBs, weekly administration of 17α-hydroxyprogesterone (17α-OHP) reduced PTB by 33% compared to placebo (Petrini et al. 2009). In another study, weekly 17α-OHP administration postponed labor by at least one week in 21 of 24 women (Kauppila et al. 1980). Recent studies have investigated the role of allopregnanolone to ascertain if it can have similar effects.

During pregnancy, plasma and hippocampal allopregnanolone levels are higher for females compared to typical ovarian cycling-induced changes (Holzbauer 1975). In pregnant women, plasma allopregnanolone levels rise throughout gestation, peak in the third trimester (∼50–100 nmol/l), and then decline to luteal levels within ∼1 h of delivery. Similar decline in circulating allopregnanolone level is observed among rodents (Concas et al. 1998). In rats, decline in allopregnanolone level also occurs in brain just prior to parturition (Concas et al. 1998). Notably, there is an inverse relationship between allopregnanolone and glucocorticoid secretion. Studies in sheep show that allopregnanolone concentrations in the fetal brain also increase during late gestation reaching maximal levels near term. Concentrations then fall dramatically after birth (Nguyen et al. 2003).

Aberrations that occur during labor, such as spontaneous PTB, are among the foremost causes of neonatal death in the United States (MacDorman et al. 2007) and are associated with stress (Hedegaard et al. 1993; Wadhwa et al. 1993; Paarlberg et al. 1995; Rini et al. 1999). Stressors, whether they are psychological, physical and/or immune, double the risk of PTB. For instance, 14% of a survivor cohort that were pregnant while residing in an area severely hit by Hurricane Katrina had PTBs compared to 6% of survivors pregnant in an area less severely damaged (Xiong et al. 2008). As well, a recent study finds that risk-taking and an immune cytokine response (serum macrophage migration inhibitory factor) predict PTB (Pearce et al. 2008). Thus, stressors are a significant contributor to PTB, and of interest is whether allopregnanolone may mitigate some of these effects.

Evidence to suggest that allopregnanolone may play a critical role in modulating stress effects related to PTB is as follows. Allopregnanolone is an endogenous modulator of the stress axis. Its secretion in brain can dampen (para)sympathetic responses to stressors. During pregnancy, unchecked stress responses can be pathological and/or teratogenic. Notably, stress responding is attenuated at the end of pregnancy, when allopregnanolone levels are most elevated (Brunton and Russell 2008; Paris and Frye 2008). Although placental 11β-hydroxysteroid dehydrogenase type-2 limits the passage of cortisol from maternal circulation to the fetus, this barrier is not complete and fetal cortisol levels rise when maternal levels are markedly raised (Williams et al. 1999). Thus, an elevated progestogen milieu occurs with the reduction of abortive stress responses and this may protects the fetus from elevated maternal glucocorticoid levels.

In a paper in this issue, Brunton and Russell describe novel mechanisms involved in attenuated neuroendocrine responses to stress in pregnancy. They use a model of immune challenge in rats to reveal a novel opioid-mediated mechanism of inhibiting stress responding during gestation. Excitatory noradrenergic neurons in the hindbrain normally drive the activity of corticotrophin-releasing hormone neurons in the paraventricular nucleus in response to physical stressors, including immune challenge, and oxytocin neurons are also stimulated, which might trigger PTB. However, this input is suppressed in late pregnancy by allopregnanolone, so the HPA axis is not stimulated, nor is oxytocin secretion (preventing contractile uterine responses to stress). Moreover, these effects are the result of central inhibitory actions of endogenous opioids in late pregnancy. This opioid inhibition is induced and maintained by allopregnanolone. These effects may represent a global mechanism through which HPA axis responses to other stressful stimuli are restrained in late pregnancy.

Another paper in this issue by Yawno and colleagues, describes the necessity for allopregnanolone formation to mitigate effects of potential negative birth outcomes. These studies were performed in sheep which, like humans, have a relatively long gestation. In these species, the placenta becomes the major source of progesterone and plasma concentrations rise markedly with advancing gestation leading to very high levels in the circulation by late gestation. Much of the progesterone that crosses to the fetus is rapidly metabolized on entering the fetal circulation leading to the formation of abundant precursors for neurosteroid production in the fetal brain. The human placenta expresses of 5α-reductases type-1 & -2 (Vu et al. 2009) which convert progesterone to 5α-dihydroprogesterone (Vu et al. 2009). This metabolite may be further converted to allopregnanolone by 3α-hydroxysteroid reductases in the brain contributing to the high gestational levels (Nguyen et al. 2003). Previous studies using finasteride to suppress conversion of progesterone to neuroactive metabolites showed that gestational neurosteroid have a major role in suppressing fetal CNS activity (Nicol et al. 1997). These studies showed that the suppression of neurosteroid synthesis with finasteride perturbed patterns of electrocorticographic activity. The incidence of sleep-like behavior that typifies the electrocorticographic patterns seen in the late gestation fetal sheep were reduced by finasteride and were replaced by an arousal-like pattern. The studies described by Yawno and colleagues in this issue highlight the clinical importance of these observations. These studies revealed that the inhibition of allopregnanolone formation promoted aberrant brain activity in response to asphyxiation and further showed that finasteride treatment increased the incidence of the sub-low voltage electrocorticographic activity and the incidence of electrocorticographic spiking activity. These patterns suggest the onset of potentially damaging seizure-like activity in the fetus. Allopregnanolone concentrations decline dramatically following the loss of the placenta at birth. The finasteride treatment in the studies of Yawno and colleagues was used to mimics the marked decline in allopregnanolone concentrations that occurs following PTB. This premature fall may occur before the fetus has experienced adequate exposure to normal late gestation neurosteroid levels and when the neonates own synthetic capacity may be limited (Vu et al. 2009). The studies of Yawno and colleagues, therefore, indicate the decline in neurosteroid levels may lower seizure threshold and predispose the preterm neonate to hyperactivity. Thus, neurosteroid actions at GABA_A receptor sites may underlie an important mechanism by which fetal brain quiescence is maintained to normal term. Preterm human neonates are more vulnerable to seizures compared to term neonates and this susceptibility is increased further after periods of hypoxia/ischemia at birth. Furthermore, seizures in very preterm neonates are difficult to treat using GABAergic agonists such as barbiturates, possibly due to immaturity in GABA_A receptor signaling mechanisms (Rivera et al. 1999; Dzhala et al. 2005). The action of neurosteroids in stimulating GABA_A receptors that are located largely extrasynaptically to modulate tonic activity suggests that these steroids may better elevate seizure threshold and protect against hyperexcitability (Belelli et al. 2005). The finding that allopregnanolone, alfaxalone, reduced aberrant electrocorticographic activity suggests neurosteroid derivatives may be a better approach for treating hyperactivity in preterm neonates compared to other GABA_A receptor agonists.

The report by Yawno and colleagues builds on previous work using finasteride to examine the effects of suppressing neurosteroid concentrations in the fetal brain (Yawno et al. 2007). The previous findings revealed that the inhibition of allopregnanolone formation markedly potentiated asphyxia (umbilical cord occlusion) induced cell death in the vulnerable hippocampal region. The paper by Yawno and colleagues indicates the hyperexcitability and seizure activity contribute to this increased cell death. These observations suggest that the preterm neonate is compromised by the lack of neurosteroids in the brain. Importantly, the studies described by Paris and Frye in this issue show adverse effects of reduced neurosteroid exposure continue in the offspring, on cognitive performance to adolescence. We speculate that the lower fetus exposure to neurosteroids due to very PTB may have similar ongoing detrimental effects. These observations suggest that the replacement therapy with appropriate neurosteroids should be evaluated for the treatment of preterm neonates.

The report by Paris and Frye in this issue shows that cognitive performance of adolescent rats is perturbed when they have been exposed to stress in late gestation, and why inadequate allopregnanolone formation in the stressed dam may be an important mechanism contributing to these outcomes. Indeed, reduced circulating allopregnanolone concentrations in stressed dams, compared with controls, predict reduced prefrontal cortex formation of allopregnanolone later in life among adolescent offspring, as well as the impaired cognitive performance of adolescent offspring. These findings provide further support for the concept that allopregnanolone formation in response to stress during pregnancy may underlie important mechanisms that have consequences, not only for the immediate support of the pregnancy, but also for lasting effects on offspring cognitive function.

Collectively, these papers indicate an important role for pregnane neurosteroid formation throughout pregnancy and development. Stressors elicit an HPA axis response, irrespective of whether they are immune stressors (as in Brunton and Russell 2010), hypoxic insults (as in Yawno et al. 2010), or physical/psychological stressors (as in Paris and Frye 2010). Allopregnanolone may serve as an important mediator during pregnancy to promote quiescence of stress-activated neurons in the dam brain. When a stressor, such as hypoxic insult, is experienced by the offspring before or during birth, allopregnanolone's inhibitory actions at GABA_A receptors may reduce neurodegeneration, in part, via its actions to regulate central activation in the fetal brain. Stress challenge at a critical time during gestation may result in perturbations of allopregnanolone formation in developing offspring, dysregulating these protective mechanisms. Such dysregulation in allopregnanolone formation may promote behavioral consequences that manifest later in life, such as poorer cognitive performance in adolescence. As such, allopregnanolone formation and actions at its endogenous targets may underlie a critical mechanism for optimal development that extends beyond the womb.

The papers in this issue of Stress, briefly reviewed above, were presented at The 5^th International Meeting on Steroids and the Nervous System in Torino, Italy, in 2009. This conference is organized to update our knowledge on the relationships between steroid hormones synthesized in different organs and the nervous system. This is a wide research field covering different areas from molecular biology to behavior. Topics covered at this most recent meeting included: sex differences in, and hormonal influences on, cannabinoid-regulated biology; steroids and pain; the importance of co-regulatory factors for steroid receptor action in the brain; estrogen-induced nonclassical signaling in the brain: mechanism and role; vitamin D and the forgotten neurosteroid; neurosteroids and GABA_A receptors; and pathogenic mechanisms mediated by glucocorticoid receptors in psychiatric disorders: from bench to clinics. An important and continuing goal of the International Meeting on Steroids and the Nervous System is to increase participation by trainees and junior investigators. As in past years, a Young Investigators Symposium was held, wherein six young investigators were chosen to present their research in 10-minute talks. At the 2009 meeting, there were also two round tables entitled, “Androgens and androgen receptors,” and “Neurosteroids and pregnancy: long-term effects on development.” In addition to the symposia and poster sessions, three highlight lectures were given: Dr L.M. Garcia-Segura (Spain) “Estrogen and neuroprotection,” Dr A.M. Etgen (USA) “Intracellular signal transduction cascades mediating the behavioral actions of estrogens,” and Dr J. Lambert (UK) “Neurosteroids: neuron specific allosteric modulators of the GABA_A receptor.” The next meeting will be in Torino, Italy in February 2011. Please see the web site for additional information: http://www.dafml.unito.it/anatomy/panzica/neurosteroids/SNS11/HOME.html

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.