Nongenomic And Genomic Properties Of Neurosteroids

In 1986, it was shown for the first time that the neurosteroids allopregnanolone (3a,5a-THPROG) and tetrahydrodeoxycorticosterone (3a,5a-THDOC) may modulate neuronal excitability via their interaction with the GABAa receptor complex (2). These steroids are capable of displacing t-butylbicyclophosphorothionate (TBPS) from the choride channel with an affinity superior to that of barbiturates and of enhancing the GABA-evoked chloride current (2). Thus, these neurosteroids may be considered as potent positive allosteric modulators of the GABAA receptor by increasing the frequency and duration of channel openings of the GABA-gated chloride channel (3,7). Studies concerning the structure-activity relationship of neurosteroids at the GABAA receptor have revealed that the presence of a 3a-hydroxy group within the A-ring of these molecules is the critical determinant for a positive allosteric activity at the GABAA receptor (8,9). Steroids lacking these particular properties, e.g., progesterone and the 5a-pregnane steroids 5a-dihydroprogesterone (5a-DHPROG) and 5a-dihydrodeoxycorti-costerone (5a-DHDOC), are devoid of any GABA-enhancing potential (8,10) (Fig. 1). Whereas steroids like 3a,5a-THPROG, 3a,5a-THDOC, or the synthetic compound alphaxalone are positive allosteric modulators of the GABAA receptor, dehydro-epiandrosterone sulfate (DHEAS) and PREG sulfate (PREGS) display functional antagonistic properties (11,12).

Although the sites of action of neurosteroids at the molecular level have been attributed to unique binding sites on neurotransmitter receptors in the cell membrane (2,3,13), a "neurosteroid binding site" has not yet been determined. Because of the distinct chemical properties of these steroids (14) and their lack of binding to the rat progesterone receptor (PR) (8), up until now these steroids were believed not to possess regulatory properties at the genomic level (3,8). The aim of our studies was to elucidate further molecular mechanisms of neurosteroid action that might be important for the future development of such compounds for potential use in neuropsychopharmacology. Therefore, we addressed the question whether these steroids may be able to regulate gene expression via intracellular steroid receptors (10).

A human neuroblastoma cell line (SK-N-MC) was transiently transfected with a reporter plasmid encoding the mouse mammary tumor virus (MMTV) promoter upstream of the luciferase gene (MMTV-LUC) (15). Expression vectors for the human glucocorticosteroid receptor (hGR) (16), the human mineralocorticosteroid receptor (hMR) (17), and the full-length isoforms of the chicken (cPRB) (18) and the human (hPRB) progesterone receptor (19) were cotransfected. Both the cPRB and hPRB were strongly activated by 3a,5a-THPROG and 3a,5a-THDOC. The cPRB responded to the neurosteroids in a progestin-like fashion whereas concentrations in the upper nanomolar range were required to activate the hPRB (Fig. 2) Moreover, the neurosteroids 3a,5a-THPROG and 3a,5a-THDOC were able to induce a complete nuclear translocation of the hPRB, indistinguishable from that achieved by progesterone (PROG) when the hPRB was expressed in COS-1 cells in immunofluorescence studies. As in previous reports in the rat (8), 3a,5a-THPROG and 3 a,5a-THDOC did not bind to the PR of either species when the respective steroid receptors were expressed in COS-1 cells. Because the cPR may be activated in a ligand-independent fashion via cyclic AMP or dopamine (20,21), we questioned whether the genomic effects of neurosteroids are mediated via

Fig. 1. Modulatory properties of steroids at the GABA receptor. (A) Positive allosteric modulation of the GABA-evoked chloride current by neurosteroids. The bar indicates the presence of 1 ^M GABA. (B) Modulatory properties of neurosteroids and of 5a-pregnane steroids at the GABA receptor in form of a representative experiment, and (C) as the mean ± SD of several independent experiments. (Adapted with permission from ref. 10.) (THP, 3a,5a-THPROG; THDOC, 3a,5a-THDOC; DHP, 5a-DHPROG; DHDOC, 5a-DHDOC.)

Fig. 1. Modulatory properties of steroids at the GABA receptor. (A) Positive allosteric modulation of the GABA-evoked chloride current by neurosteroids. The bar indicates the presence of 1 ^M GABA. (B) Modulatory properties of neurosteroids and of 5a-pregnane steroids at the GABA receptor in form of a representative experiment, and (C) as the mean ± SD of several independent experiments. (Adapted with permission from ref. 10.) (THP, 3a,5a-THPROG; THDOC, 3a,5a-THDOC; DHP, 5a-DHPROG; DHDOC, 5a-DHDOC.)

0.01 0.1 1 10 100 0.01 0.1 1 10 100 0.1 1 10 100 100C

Steroid [nM] Steroid [nM] Steroid [nil]

Fig. 2. Progesterone-receptor mediated gene expression by progestins and 5a-reduced neurosteroids. Induction of the MMTV-promoter after cotransfection of cPR or hPR expression vectors into SK-N-MC cells and incubation with steroids at the indicated concentrations. (A) progesterone (closed diamonds) and R 5020 (closed triangles) after transfection of cPR ; progesterone (open diamonds) and R 5020 (open triangles) after transfection of hPR. (B and CC) 3a,5a-THPROG (open circles) and 3a,5a-THDOC (closed circles), 5a-DHPROG (open squares) and 5a-DHDOC (closed squares) after transfection of cPR (B) and hPR (C). The baseline activity of the MMTV-promoter without addition of steroid is set as 1. (Adapted with permission from ref. 10.)

0.01 0.1 1 10 100 0.01 0.1 1 10 100 0.1 1 10 100 100C

Steroid [nM] Steroid [nM] Steroid [nil]

Fig. 2. Progesterone-receptor mediated gene expression by progestins and 5a-reduced neurosteroids. Induction of the MMTV-promoter after cotransfection of cPR or hPR expression vectors into SK-N-MC cells and incubation with steroids at the indicated concentrations. (A) progesterone (closed diamonds) and R 5020 (closed triangles) after transfection of cPR ; progesterone (open diamonds) and R 5020 (open triangles) after transfection of hPR. (B and CC) 3a,5a-THPROG (open circles) and 3a,5a-THDOC (closed circles), 5a-DHPROG (open squares) and 5a-DHDOC (closed squares) after transfection of cPR (B) and hPR (C). The baseline activity of the MMTV-promoter without addition of steroid is set as 1. (Adapted with permission from ref. 10.)

intracellular kinases. However, this effect was not demonstrable with the hPR. Therefore, a kinase-predominated mechanism seems to be rather unlikely to explain gene expression induced by neurosteroids. Site-directed mutagenesis revealed that the carboxyterminus of the PR is required to confer the genomic effects of neurosteroids. Gel-shift analysis was employed to characterize the DNA-binding properties of the PR induced by neurosteroids. These experiments revealed that, in contrast to PROG, the neurosteroids were able to induce a conformational change of the PR only when added to living cells, but not when added to the band-shift incubation mixture, suggesting that the neurosteroids act indirectly via metabolism in the host cells. In the neuroblastoma cells used for analysis of transactivation, radioactively-labeled 3a,5a-THPROG was converted rapidly into the 5a-pregnane steroid 5a-DHPROG by the 3a-hydroxysteroid oxidoreductase (Fig. 3). The 5a-pregnane steroids 5a-DHPROG and 5a-DHDOC bound to the cPR with considerable affinity, whereas binding to the hPR was less pronounced. Moreover, they were potent inducers of gene expression via cPR and hPR. We could show that the neurosteroids 3a,5a-THPROG and 3a,5a-THDOC not only act through membrane-bound receptors, but may also regulate gene expression via the PR. Intracellular oxidation into 5a-DHPROG and 5a-DHDOC appears to be essential to confer PROG-like induction of DNA binding and regulation of gene transcription. This dual action of brain-derived steroids seems to be of physiological relevance, as the concentrations required for activation of the PR (10) and the GABAa receptor (3,22) are in the nanomolar range. Nanomolar concentrations of 3a,5a-THPROG and 3a,5a-THDOC and conversions of these steroids into 5a-DHPROG and 5a-DHDOC were shown to occur in vivo in the rat brain (14,23). Whether genomic or nongenomic effects of these neurosteroids predominate may thus depend on the relative expression of PR (24,25), GABAA receptors (26), and of metabolizing enzymes (27,28) in the target tissues.

In further studies we were able to show that an array of naturally occurring and synthetic neurosteroids may regulate gene expression via the cPR, whereas the hPR displays a more selective transactivation pattern (29) (Fig. 4). Of particular interest was the

MEVALONOLACTONE —CHOLESTEROL—" f^Y)

DEOXYCORTICOSTERONE 5a-REDUCTASE

Was this article helpful?

0 0

Post a comment