The term neurosteroid was introduced in 1981 (). Evidence that the brain can be a steroidogenic tissue was derived from the finding that substantial quantities of pregnenolone, dehydroepiandrosterone (DHEA), their sulfate esters, and their fatty acid esters were found in the central nervous system (CNS) of mice, rats, pigs, guinea pigs, monkeys, and humans (). Concentration of these steroids in the brain exceeded their plasma concentration and were estimated to be up to 10 times higher, in the order of 10-100 nmol/L (). Furthermore, these steroids appeared to be independent of gonadal and adrenal synthesis because they were present after adrenalectomy and gonadectomy (). In addition, incubation of primary cultures of rat forebrain glial cultures with a precursor of cholesterol led to the formation of cholesterol, pregnenolone, 20-OH-pregnenolone, and progesterone (). These findings from animal studies are supplemented by results from postmortem human brains, in which high concentrations of neurosteroids far exceeding plasma concentrations have also been detected in all CNS regions (). Of special interest from a neuropsychobiological perspective are neuroactive steroids, which include those steroids that act via neuronal cell surface receptors ().
Biochemistry and pharmacology of neurosteroidocenesis
In adrenal tissue and other steroidogenic tissues including glial cells — in particular, oligodendrocytes () — conversion of cholesterol into pregnenolone is catalyzed by cytochrome P450scc, located in the inner mito-chondrial membrane (). Further synthesis of neurosteroids probably proceeds through different pathways than those used in adrenals, gonads, and placenta (). The brain contains enzymes that metabolize pregnenolone, progesterone, and 11-deoxycorticosterone (DOC) into a variety of neuroactive compounds. The major metabolites of progesterone include allopregna-nolone (3α-hydroxy-5α-pregnan-20-one, 3α-5α-tetrahydroprogesterone; THP). DOC is metabolized to 3α,21-dihydroxy-5α-pregnan-20-one, allotetrahydro-DOC (allo-THDOC), indicating that these neural tissues contain 5α-reductase and 3α-hydroxysteroid oxidoreductase (). Pregnenolone (3α-hydroxy-5β-pregnan-20-one, 3α-5β-tetrahydro-progesterone, 3α-5β-THP) and THDOC (3α,21-dihydroxy-5β-pregnan-20-one, 3α-5β-tetrahydroDOC) are synthesized via 5β-reductase instead of 5α-reductase activity and further by 3α-hydroxysteroid oxidoreductase.
Whereas 5α-reductase shows a significantly higher function in neurons than in glial cells, 3α-hydroxysteroid dehydrogenase, which converts pregnenolone to progesterone, is present mainly in type-1 astrocytes (). All enzymes mentioned are found in highest concentrations in the pituitary and the hypothalamus, as well as in the cerebellum, thalamus, midbrain, pineal, medulla, white matter, and peripheral nerves (). Because the adult rat brain does not have 17a-hydroxylase activity nor contains 17a-hydroxylase (P450cl7) messenger ribonucleic acid (mRNA), the origin of brain DHEA is unknown because it cannot be synthesized from pregnenolone via P450cl7, but its concentrations in brain persists long after removal of gonads and adrenals ().
Classic and nonclassic effects of steroids
Regardless of their source of synthesis, glucocorticoids, mineralocorticoids, androgens, estrogens, and progestins exert their effects on the brain through high-affinity intracellular steroid hormone receptors, thereby modifying the expression of specific genes (). Allopregnanolone and allo-THDOC have been shown to regulate neuronal function through effects on gene expression by activating the intracellular progesterone receptor (). However, the major body of work to date has investigated nongenomic effects of neurosteroids on transmitter-gated ion channels. A first description of rapidly mediated steroid effects was given by Selye (1941). He reported on the anesthetic and sedative properties of progesterone and DOC in the rat; this discovery eventually resulted in the development of steroid-based anesthetics, for example, alfaxalone and hydroxydione (). Electrophysiological and ligand binding studies have shown that pregnenolone and THDOC exert these properties by interacting with the γ-aminobutyric acid type A (GABAA) receptor (). These neurosteroids act as allosteric agonists of the GABAA receptor by increasing the frequency and duration of chloride channel openings and further potentiate the inhibitory action of GABA (). It is important to note that neurosteroid action at the GABAA receptor is coupled to the presence of a 3a-hydroxyl group (). Sulfated neurosteroids, for example, pregnenolone sulfate and DHEA sulfate (DHEAS), act as noncompetitive antagonist of the GABAA receptor and inhibit GABA-induced chloride transport by a reduction in GABA chloride channel opening frequency (). They are less selective and have a lower affinity for GABAA receptors than do allopregnanolone and allo-THDOC (). DHEA also acts as noncompetitive antagonist at the GABAA receptor, although three to four times less potently than DHEAS (). Neither progesterone nor pregnenolone are active as modulators of the GABAA receptor and probably display their in vivo effects via formation of their reduced metabolites ().
Whereas benzodiazepines, barbiturates, convulsant channel agonists, and GABA itself bind at the GABAA receptor ectodomain, neurosteroids most likely bind to a site distinct from the benzodiazepine site of the GABAA receptor (). Although the allosteric modulatory efficacy of benzodiazepines acting in the receptor ectodomain differs according to the structural diversity of GABAA receptors, neurosteroid modulation is much less dependent on GABAA receptor structure diversity (). Moreover, neurosteroids can operate GABAA receptors in the absence of GABA, whereas benzodiazepines and other anxiolytics not only require GABA for their action, but their maximal modulatory efficacy never surpasses the maximal response elicited by GABA in the same receptor channel (). This explains why neurosteroids do and benzodiazepines do not possess anesthetic properties surpassing their anxiolytic effect.
Another receptor system with which neurosteroids interact is the N-methyl-D-aspartate (NMDA) receptor. Pregnenolone sulfate acts as a positive allosteric modulator of the NMDA receptor, in analogy to GABA-ergic effects, by increasing the frequency and duration of NMDA-activated channel opening ().
The use of benzodiazepines in anxiety disorders represents a difficult issue because of the well-known addictive properties of this substance class; furthermore, the depressogenic lability of benzodiazepines represents a major problem in the treatment of generalized anxiety disorders, in which concomitant depressive symptoms occur in up to 50% of patients (). Therefore, a compound mimicking endogenous anxiolytic effects, and that might further possess potential antidepressive properties, is warranted. Neuroactive steroids could represent such a substance class. However, the following issues need to be addressed in this respect:
• Although theoretically lipophilic (nonsulfated) neurosteroids penetrate the blood-brain barrier easily, with plasma and CSF concentrations of these compounds being almost identical (), Corpechot et al. () reported that brain uptake of intravenously injected allopregnanolone is about 100-fold lower than that of progesterone. In this respect, it is noteworthy that allopregnanolone has failed to display anxiolytic or sedative effects following systemic administration in one earlier study (). Compared with allo-THDOC though, CNS uptake of the more hydrophobic allopregnanolone is significantly greater; therefore, potential difficulty in brain availability of these compounds resulting from drug penetrance or rapid systemic metabolization requires further clarification ().
• With regard to the administration of pregnenolone or DHEA, it will be important to take potential sulfation/desulfation of these substances () into consideration to avoid possible GABA agonistic/antagonistic effects. Indirect evidence for this possibility can be derived from first systemic administrations of neurosteroids in human sleep research, with pregnenolone inducing sleep patterns compatible with inverse agonistic GABAA receptor modulation () and DHEA exerting mixed GABAA-agonistic/antagonistic effects ().
• A further issue is raised with regard to dose-dependent effects, suggesting dual or U-shaped psychotropic responses, because recent studies have indicated that neurosteroids may induce anxiogenic or anxiolytic responses in relation to the dosage used and subsequent metabolizing steps involved ().
• Gender differences in response to GABA-ergic neurosteroids have also to be taken into consideration. Although many studies cited used male animals (), which were sensitive to the anxiolytic effects of the neurosteroids tested, gender or hormonal status, for example, estrous cycle phase, may influence the metabolization and/or CNS response to neuroactive steroids ().
• Finally, neuroanatomical distribution of steroid-sensitive GABAA receptors in the CNS needs to be further correlated with CNS areas involved in anxiety responses as well as their interaction with other transmitter systems, for example, CRH or dopamine (). Such characterization may result in the development of more specific compounds, possibly also involving concomitant antidepressive effects.
In conclusion, neurosteroid-based anxiolytics with low intrinsic toxicity may represent an exciting new pharmacological development with potential advantages over existing classes of anxiolytics with regard to tolerance, dependence, and abuse liability.
Selections from the book: “Pharmacotherapy for Mood, Anxiety, and Cognitive Disorders”, 2000.