Neurosteroids differ from nearly all the other transmitters and mediators in that they are lipid-soluble and can easily cross the blood-brain barrier. Thus it is necessary to distinguish those steroids that are produced in the brain from those that find their way there from the circulation after being released from the adrenal cortex or gonads. There are many natural and synthetic steroids that have some effect on neuronal function and can be considered neuroactive but few are actually produced in the brain to act on neurons, i.e. the true neurosteroids.
Steroids which are found in the brain may be grouped as follows. The list is not exhaustive.
(a) Those formed in peripheral glands: Corticosteroids
Corticosterone (a glucocorticoid) Aldosterone (a minerolocorticoid) Reproductive steroids Oestradiol (an oestrogen) Testosterone (an androgen)
All these steroids disappear from the brain in animals after removal of the adrenals or gonads (ovary and testis). This also applies to tetrahydrodeoxycorticosterone for although it is formed by reduction of deoxycorticosterone within the brain, its synthesis depends on that steroid coming from the blood.
(b) Those formed in both the periphery and CNS: Progesterone (PROG) (a progestogen)
Tetrahydroprogesterone (3a5aThPROG) or allopregnenolone, a reduced metabolite of progesterone
Pregnenolone (PREG) and its reduced (20a dihydropregnenolone) and sulphated (pregnenolone sulphate, PREGS) metabolites
The levels of these steroids are reduced, especially that of progesterone (70%) by adrenalectomy and/or gonadectomy but sufficient remains to indicate some central synthesis.
(c) Those formed within the CNS:
Dehydroepiandrosterone (DHEA) and dehydroepiandrosterone sulphate, its sulphated derivative DHEA levels are unaffected by adrenolectomy or gonadectomy
The neurosteroids of most interest are DHEA, PREG, PREGS, PROG and 3a5aThPROG.
The chemical structures and interrelationship of the neurosteroids listed in (b) and (c) above are shown in Fig. 13.5 but the synthetic pathways are not well established. PREG and DHEA are known as 3^-hydroxy-A5-steroids and are also found in peripheral glands as intermediaries between cholesterol and active steroids such as PROG and testosterone. It seems that cholesterol may be the starting point of neurosteroid synthesis in the brain. Cytochrome P450scc, the specific hydroxylase needed for cleavage of the cholesterol side-chain to give PREG, has been found widely in white matter and in myelinating oligodendrocytes, but not in neurons. Enzymes are present for the conversion of PREG to PROG and both are reduced in glia and neurons. This does not occur with DHEA and very little is known of either its synthesis or metabolism.
Neurosteroids are widely and fairly evenly distributed in the brain with few noteworthy localisations but the concentration of the conjugated forms (sulphated and reduced) of PREG and DHEA can exceed that of the parent compounds. Values given by Baulieu (1997) are PREG 8.9ng/g, with 14.2 and 9.2 for its sulphate and hydroxy metabolites, DHEA 0.24 ng/g (1.7 and 0.45) and PROG 2.2ng/g. By contrast, although the concentration of progesterone rises some twelvefold in plasma and eight times in hippocampus of animals and humans as they pass from the follicular to luteal phase of the ovarian cycle, it increases by 300 in the cortex, suggesting a considerable variation in the ability of different brain regions to concentrate it. In considering the neurosteroids as possible NTs it should be remembered that neither specific steroid neurons nor an evoked neuronal release of steroids have been demonstrated. There are, however, receptors for them in the CNS and no shortage of actions attributed to them.
There is no doubt that steroids have behavioural effects. Given clinically in the therapy of inflammatory conditions, the glucocorticoids are considered to produce euphoria, followed after prolonged or higher dosing by depression and, of course, when we are stressed the corticocosteroids pumped out by the adrenal cortex easily enter the brain and may initiate some of the behavioural response. In women the premenstrual syndrome (irritability, depression and anxiety) is thought to be associated in some way with progesterone since not only does its concentration rise then fall during that time but in post-menopausal women the use of sequential oestrogen and progestogen hormone replacement therapy (HRT) shows that similar mood changes accompany only the addition of progestogen. More specifically, in women with epilepsy while the incidence of seizures decreases when plasma progesterone is high, it increases during the immediate premenstrual period as progesterone levels fall, rather as with withdrawing
Figure 13.5 Structure and interrelationship of the major neurosteroids. Pregnenolone (PREG) (1) is synthesised from cholesterol and is then either metabolised, reduced to 20a dihydropregnenolone (7) or sulphated (6), or converted by 3^-hydroxysteroid dehydrogenase to progesterone (PROG) (2) or to dehydroepiandrosterone (DHEA) (3). The former (PROG) can then be reduced to allopregnanolone (3a5aThPROG) (4) and DHEA sulphated to dehydroepiandrosterone sulphate (5). It is PREG, DHEA and 3a5aThPROG which appear to be important centrally. The structures of alphaxalone, a steroid anaesthetic and tetrahydrodeoxy-corticosterone, which is formed in the brain from deoxycorticosterone of peripheral origin, are also shown
Figure 13.5 Structure and interrelationship of the major neurosteroids. Pregnenolone (PREG) (1) is synthesised from cholesterol and is then either metabolised, reduced to 20a dihydropregnenolone (7) or sulphated (6), or converted by 3^-hydroxysteroid dehydrogenase to progesterone (PROG) (2) or to dehydroepiandrosterone (DHEA) (3). The former (PROG) can then be reduced to allopregnanolone (3a5aThPROG) (4) and DHEA sulphated to dehydroepiandrosterone sulphate (5). It is PREG, DHEA and 3a5aThPROG which appear to be important centrally. The structures of alphaxalone, a steroid anaesthetic and tetrahydrodeoxy-corticosterone, which is formed in the brain from deoxycorticosterone of peripheral origin, are also shown an antiepileptic drug. Of course, it cannot be assumed that plasma levels are reflected in the brain but in rats stressed by insertion in the Morris water maze, so that they have to swim to a safe platform (see Chapter 18), there is still some increase in the concentration of brain PROG and in particular its reduced metabolic (ThPROG), even after adrenalectomy (Purdy et al. 1991). The aggressiveness of castrated male mice exposed to lactating females in a cage can also be reduced by DHEA administration.
These observations, while implicating steroids in brain function and behaviour, cannot be taken as a reliable indicator of their actual effect on neuronal function. Nevertheless, some neurosteroids produce CNS depression with a rapid inhibition of neuronal excitability and one progesterone derivative, alphaxalone (3a-hydroxy-5a pregnane-11, 20 dione, see Fig. 13.5) has been used effectively as an intravenous anaesthetic in humans.
Intracellular steroid receptors, which alter gene expression, exist for corticosteroids, oestrogens and progesterone in the brain, as in the periphery but they cannot account for the relatively rapid depression of CNS function induced by some steroids. This was explained when Harrison and Simmonds (1984) discovered that alphaxalone (the steroid anaesthetic) potentiated the duration of GABA-induced currents at the GABAa receptor in slices of rat cuneate nucleus just like the barbiturates (Fig. 13.6). Of the
Figure 13.6 Potentiation of GABA action at the GABAA receptor by the steroid anaesthetic, alphaxalone. Depolarisations recorded extracellularly from dorsal funiculus fibres and terminals in the rat cuneate brain slice after superfusion for 2 min with muscimol 2.5 p,M (m), GABA 50 p,M (G) and glycine 2mM (g). In the presence of alphaxalone (1 pM), responses to GABA and the GABAA agonist muscimol, but not those to glycine, were substantially enhanced. The effect was reversible with responses slowly returning to normal after 3 h. (From Harrison and Simmonds 1984 and reproduced by permission of Elsevier Science)
naturally occurring neurosteroids ThPROG also increases GABAa receptor currents increasing both the duration and probability of Cl_ channel opening. The sulphated metabolite of PREG is similarly active at low (nM) concentrations. These allosteric effects are still seen after maximal barbiturate potentiation and are not affected by benzodiazepine antagonists suggesting a specific and separate modulating site for the steroids (see Paul and Purdy 1992) although it has not been found. Also while their activity changes with the subunit composition of recombinent GABAA receptors no specific configuration has been established for their effectiveness but expression of a2 with a1 + B1 or a2 + B1 gives a more responsive receptor than the inclusion of a3 (Shingai, Sutherland and Barnard 1991).
Not all neurosteroids are inhibitory. The GABA potentiation seen with low concentrations of PREG sulphate changes to antagonism at higher strengths and both this compound and sulphated DHEA, which also inhibits GABAA receptors, are proconvulsant. There is also evidence that the sulphates of PREG and DHEA potentiate NMDA receptors while glucocorticoids reduce seizure threshold without affecting GABA receptors. Thus even without considering reports of effects on glycine sigma and ACh nicotinic receptors, the electrophysiology of the steroids is complex. In practice, although steriods modulate GABAA receptors at realistic nM concentrations, unphysiological ^M amounts are required at other ligand-gated ion channels (see Rupprecht and Holsboer 1999).
Two unrelated steroid effects warrant some mention. The discovery that PREGS levels are significantly lower in aged than young male rats prompted an interest in its possible role in memory function. In rats, subjected to spatial memory tests in water and Y mazes, impairments in performance were mirrored by reductions in hippocampal PREGS levels (see Baulieu 1997), while aged rats with established memory impairment showed improvement after PREGS administration. Whether this depends on the known ability of PREGS to increase NMDA activity and the accepted role of that receptor in LTP maintenance and memory functions remains to be seen.
In the periphery PROG and PREG may well have an important trophic action since their production in Schwann cells has been shown to result in increased myelin synthesis in regenerating rat sciatic nerve and cultured dorsal root ganglia (see Koenig et al. 1995).
With so many different neurosteroids with differing and even opposing neuronal effects, much will depend on their relative concentrations at any time and any evaluation of their function must take this into consideration. Hopefully the synthesis and use of appropriate antagonists will throw more light on the physiological role of steroids in the CNS and facilitate the development and clinical use of new neuroactive steroids (see Gasior et al. 1999).
Continue reading here: Adrenaline
Was this article helpful?