Address correspondence and reprint requests to Abdel Ghoumari, INSERM U488, 80 rue du Général Leclerc, 94276 Kremlin-Bicêtre, France, Tel. (33) 1 49 59 18 97, E-mail: firstname.lastname@example.org
We have previously shown that progesterone (PROG) is synthesized by Schwann cells and promotes myelin formation in the peripheral nervous system (PNS). We now report that this neurosteroid also stimulates myelination in organotypic slice cultures of 7-day-old (P7) rat and mouse cerebellum. Myelination was evaluated by immunofluorescence analysis of the myelin basic protein (MBP). After 7 days in culture (7DIV), we found that adding PROG (2–5 × 10−5 M) to the culture medium caused a fourfold increase in MBP expression when compared to control slices. The effect of PROG on MBP expression involves the classical intracellular PROG receptor (PR): the selective PR agonist R5020 significantly increased MBP expression and the PR antagonist mifepristone (RU486) completely abolished the effect of PROG on this MBP expression. Moreover, treatment of P7-cerebellar slice cultures from PR knockout (PRKO) mice with PROG had no significant effect on MBP expression. PROG was metabolized in the cerebellar slices to 5α-dihydroprogesterone (5α-DHP) and to the GABAA receptor-active metabolite 3α,5α-tetrahydroprogesterone (3α,5α-THP, allopregnanolone). The 5α-reductase inhibitor L685-273 partially inhibited the effect of PROG, and 3α,5α-THP (2–5 × 10−5 M) significantly stimulated the MBP expression, although to a lesser extent than PROG. The increase in MBP expression by 3α,5α-THP involved GABAA receptors, as it could be inhibited by the selective GABAA receptor antagonist bicuculline. These findings suggest that progestins stimulate MBP expression and consequently suggest an increase in CNS myelination via two signalling systems, the intracellular PR and membrane GABAA receptors, and they confirm a new role of GABAA receptors in myelination.
Gey's balanced salt solution containg 5 mg/mL glucose
gas chromatography/mass spectrometry
myelin basic protein
peripheral nervous system
The study of the synthesis and actions of progesterone (PROG) in the nervous system has become a very active field of investigation. Many new concepts have emerged over the past few years regarding the significance of this pleitropic hormone, which regulates many neuronal and glial functions in the brain, spinal cord and in peripheral nerves (Schumacher et al. 2000; Schumacher and Robert 2002). Thus, PROG has been shown to be neuroprotective and to promote peripheral nervous system (PNS) myelination (Schumacher et al. 2001; Stein 2001). Recent research has also revealed multiple mechanisms of PROG action, in particular the regulation of gene transcription after its binding to an intracellular receptor (PR), rapid non-genomic effects on cell-signalling pathways and direct actions on membrane receptors for neurotransmitters (Rupprecht and Holsboer 1999; Boonyaratanakornkit et al. 2001; Mani and O'Malley 2002). Particularly well characterized has been the positive modulation of neuronal GABAA receptors by the PROG metabolite 3α,5α-tetrahydroprogesterone (3α,5α-THP), also known as allopregnanolone (Lambert et al. 2001). Interactions of 3α,5α-THP with membrane GABAA receptors may explain the multiple psychopharmacological effects of progestins, including their anaesthetic, anxiolytic, anticonvulsant, analgesic actions and also their effects on sleep patterns and memory (Majewska 1992; Smith 1994).
Progesterone present in nervous tissues originates either from the endocrine glands, as steroids easily cross the blood–brain and blood–nerve barriers, or from local synthesis. Steroids synthesized within the nervous system have been named neurosteroids (Baulieu 1997; Robel et al. 1999). The presence of the cytochrome P450scc, which catalyses the conversion of cholesterol to pregnenolone, has been demonstrated in both glial cells and neurones of the vertebrate brain (Le Goascogne et al. 1987; Mensah-Nyagan et al. 1999; Tsutsui et al. 2000; Mellon et al. 2001). Pregnenolone, either locally synthesized or derived from the circulation, is converted to PROG in neurones and glial cells by the 3β-hydroxysteroid dehydrogenase (3β-HSD; Jung-Testas et al. 1989; Guennoun et al. 1995; Guennoun et al. 1997; Ukena et al. 1999). In Schwann cells, the myelinating glial cells of peripheral nerves, expression and activity of the 3β-HSD are dependent on a diffusible neuronal factor (Robert et al. 2001). Recently, the synthesis of PROG has also been studied in the oligodendroglial lineage, showing that only PSA-NCAM+ pre-progenitors and progenitors express the 3α-HSD mRNA and synthesize PROG from pregnenolone (Gago et al. 2001). PROG is further converted in the nervous system to 5α-DHP by the 5α-reductase and to the GABAA receptor-active neurosteroid 3α,5α-THP by the 3α-hydroxysteroid oxidoreductase (3α-HSOR; Melcangi et al. 1994; Guennoun et al. 1997). The presence of an important 5α-reductase activity in purified myelin membranes indicated an important role of this enzymatic step in myelination (Melcangi et al. 1988).
An important role of PROG in myelination has been demonstrated in the PNS, both in vivo after lesion of the sciatic nerve and in co-cultures of dorsal root ganglia (DRG) sensory neurones and Schwann cells (Koenig et al. 1995; Chan et al. 1998). Recently, the prolonged administration of PROG has also been shown to reverse age-related structural abnormalities of the peripheral myelin sheaths (Azcoitia et al. 2003). PROG may promote myelination and remyelination by acting directly on Schwann cells and stimulating the expression of myelin protein genes and of transcription factors such as Krox-20 (Désarnaud et al. 1998; Guennoun et al. 2001; Mercier et al. 2001). PROG and its reduced metabolites regulate myelination in peripheral nerves either through the classical intracellular PR or through a membrane GABAA receptor, which are both expressed in Schwann cells (Jung-Testas et al. 1996; Magnaghi et al. 2001). In addition, PROG may influence the myelination process indirectly by acting on DRG neurones, which also express the PR (Chan et al. 2000).
The role of PROG and its metabolites in CNS myelination by oligodendrocytes is not so well established, although it has been suggested by a few studies. Thus, in cultures of glial cells prepared from newborn rat brains, PROG increased the number of myelin basic protein (MBP)-immunoreactive oligodendrocytes (Jung-Testas et al. 1994). In aged rats, in which spontaneous remyelination is very much delayed after a demyelinating lesion, the prolonged administration of PROG has recently been shown to promote the formation of new myelin sheaths (Ibanez et al. 2003b).
To investigate the effect of PROG and its 5α-reduced metabolites on myelination, we have developed a model of organotypic slice cultures of rat cerebellum. We used P7 newborn rats because, as we have previously demonstrated, Purkinje cells die in explant cultures by apoptosis between P1 and P5 (Dusart et al. 1997; Ghoumari et al. 2000, 2002). Moreover, the second post-natal week corresponds to a period of high 3β-HSD mRNA expression, elevated levels of PROG (Ukena et al. 1999) and intense myelination (Notterpek et al. 1993) in the cerebellum. Our results demonstrate a marked effect of PROG on MBP-immunoreactivity in cerebellar slices, which is mediated by the classical intracellular PR. Its metabolite 3α,5α-THP also increased MBP expression by a bicuculline-sensitive mechanism involving GABAA receptors.
Materials and methods
New born animals of post-natal day 7 (P7) from Sprague–Dawley Rats (Janvier, Le Genest St Isle, France) and from PR-KO mice were used. P0 was the day of birth. For each experiment at least three animals and 18 slices were used. After decapitation, brains were dissected out into cold Gey's balanced salt solution containing 5 mg/mL glucose (GBSS-Glu) and meninges were removed. Cerebellar parasagittal slices (350-µm thick) were cut on a MacIlwain tissue chopper and transferred onto membranes of 30-mm Millipore culture inserts with 0.4 µm pore size (Millicell, Millipore, Bedford, MA, USA). Slices were maintained in culture in 6-well plates containing 1 mL of medium at 35°C in an atmosphere of humidified 5% CO2. The medium was composed of 50% basal medium with Earle's salts (Invitrogen, Gaithersburg, MD, USA), 25% Hanks' balanced salt solution (Life Technologies, Grand Island, NY, USA), 25% horse serum (Life Technologies), l-glutamine (1 mm) and 5 mg/mL glucose.
To investigate whether steroids might enhance the expression of MBP protein in the organotypic slice cultures, PROG and its 5α-reduced metabolites (5α-DHP and 3α,5α-THP) were applied to the medium of P7 rat or mice cerebellar slice cultures. An anti5α-reductase (L685-273) and an antiprogestin RU486 (Mifepristone) were used to block PROG metabolism and binding of PROG to its receptors (PR), respectively. R5020 (progesterone-receptor agonist), muscimol (GABAA receptor agonist) and bicuculline (GABAA receptor antagonist) were also applied in this study. PROG, muscimol and bicuculline were dissolved in dimethyl sulphoxide (DMSO); 5α-DHP, 3α,5α-THP, L685-273 and RU486 in ethanol. The final concentrations of vehicles were about 0.1%. Control media contained the same amounts of vehicles. Dose–response curves were determined by treating wild-type cerebellar slices with different concentrations of each compound and we further retained the doses with maximal efficiency: PROG, 5α-DHP and 3α,5α-THP (20–50 µm); L685-273 (20 µm), R5020 (20 µm) and RU486 (5–10 µm), Muscimol and Bicuculine (100 µm). Cerebellar slices were maintained in culture for 7 days (7DIV). Medium, with the respective steroids or drugs, was replaced once after 2 or 3 days.
Antibodies and staining procedures
The following primary antibodies were used: rabbit polyclonal antibody against Calbindin protein (CaBP, 1/10.000 dilution; Swant, Bellinzona, Switzerland) was used to visualize Purkinje cells. Monoclonal antibodies against MBP 1/1000 dilution; Chemicon International (Temecula, CA, USA) sulfatide O4 (clone 81, 1/500 dilution, Boehringer Mannheim, Mannheim, Germany) and galactocerebroside (GalC; clone MGalC, 1/1000 dilution, Chemicon International) were used to visualize myelin, oligodendrocyte progenitors and oligodendrocytes, respectively.
The organotypic cultures were fixed in 4% paraformaldehyde in phosphate buffer (0.1 m, pH 7.4) for 1 h at room temperature. After washing in phosphate-buffered saline (PBS), slices were taken off the Millicell and processed for immunocytochemistry. In all cases, slices were incubated for 1 h in PBS 0.12 m (pH 7.4) containing 0.25% Triton-X, 0.2% gelatin, 0.1% sodium azide (PBSGTA) and lysine (0.1 m), before the overnight incubation with the primary antibodies (diluted in PBSGTA). The following secondary antibodies were used: goat anti-rabbit CY3 (1/200 dilution; Jackson ImmunoResearch Laboratories, Inc, West Baltimore Pike, MD, USA), and goat anti-mouse Alexa Fluor488 (1/1000 dilution, Molecular Probes, Leiden, the Netherlands). After 2 h incubation, slices were washed several fold in PBS, mounted with mowiol (Calbiochem, San Diego, CA, USA) and analysed using a fluorescent microscope (Axiovert 135M Zeiss, Carl Zeiss Inc, Göttingen, Germany).
Quantification of myelination in cerebellar slice cultures
MBP staining in organotypic slice cultures of rat cerebella was measured using a confocal Zeiss LSM 410 (Carl Zeiss Inc) image analysing system. Images were acquired with a non-confocal configuration (488 nm and 543 nm excitations). For each measurement, a region of interest was located around the deep nuclei zone and the apical ends of each lobule (see Fig. 1b), according to our observation that myelination begins at the deep nuclei zone in the cerebellum and thereafter progresses in the white matter along the axons. The MBP staining intensity was measured in these areas, using NIH image software. This staining intensity is quantified on a continuous scale of 0–255 (darkest). In order to minimize differences among the respective measurement, we set as control an arbitrary level of staining 100. The MBP staining intensity was evaluated as %(light pixels/light + dark pixels). Oligodendrocyte density, studied in the present report, was also quantified using this equation.
Cerebellar slice cultures were washed with PBS and an equal volume of lysis buffer [Tris–HCL 1.5 m, pH 7,5; NaCl 100 mm; MgSO4 2 mm; Triton X-100 4%; PMSF 40 mm; leupeptine 0.5mg/mL; ditihiothreitol (DTT) 1 m] was added to the slices and then vortexed. The protein content for each extract was quantified and 25 µg of total protein from each sample was separated by electrophoresis using 12% polyacrylamide gel and transferred on polyvinylidene difluoride (PVDF) membranes. After blocking with 5% dry milk, the membranes were incubated overnight at 4°C with a primary monoclonal anti-MBP antibody (1/500 dilution; Chemicon International). After washing with Tween-20 PBS buffer, membranes were incubated for 1 h with the peroxidase-conjugated AffiniPure GAM (1/20000 dilution; Jackson Immunoreseasrch Laboratories, West Grove, PA, USA). The blots were developed with the Enhanced Chemiluminescence ECL + plus detection kit (Amersham, Little Chalfont, UK).
Progesterone metabolism in cerebellar slice cultures using thin-layer chromatography
The metabolism of PROG was investigated by incubating cerebellar slice cultures with 1 µCi 3H-PROG (Amersham Life Science, 86 Ci/mmole, diluted to 5 µm with cold PROG) for different incubation time (24, 48 and 72 h). Some incubations were carried out in the presence of 20 µm of L685-273, a 5α-reductase inhibitor. After the incubation period, steroids were extracted three times, from slice cultures and media separately, with 2 volumes of acetate/isooctane (1 : 1, vol : vol). The organic phases were collected and taken to dryness by heating at 60°C under air pressure. Steroids were separated and characterized by thin-layer chromatography (TLC) on silica gel 60F254 plates (Merck Eurolab, Fontenay-sous-Bois, France), developed in the solvent system chloroform/ethyl acetate (4 : 1, vol:vol). Reference steroids were run in separate lanes. Chromatograms were analysed after development with an automatic TLC-linear analyser (Berthold Trace Master 20; Berthold, Colombes, France). The relative mobilities of compounds (Rfs) were: 0.0 for polar metabolites; 0.45 for 3β,5α-THP; 0.53 for 3α,5α-THP; 0.66 for PROG; 0.79 for 5α-DHP and 0.9 for non-polar metabolites. The identity of these metabolites was verified by gas chromatography/mass spectrometry (GC/MS) which allows the identification and the measurement of small amounts of steroids with high specificity and sensitivity (Liere et al. 2000).
The non-polar fraction was analysed after saponification by heating with 1 mL of 95% ethanol, 40% KOH, 95 : 5 (v/v) for 1 h at 80°C and then overnight at room temperature under nitrogen atmosphere. The solution was diluted with 1 mL water, acidified with 1 N HCl and steroids were extracted repeatedly with ethyl acetate, separated from lipids by passage on a C18 reverse phase column (IST Isolute C18 500 mg) and characterized by TLC (as described above).
Data were expressed as mean of at least 18 cerebellar slices (n = 18) from three independent animals (n = 3) and in three independent experiments ± SEM. The Newman–Keuls tests after one-way anova was used to analyse the differences in MBP intensities or in oligodendrocytes densities between steroid-treated and-non-treated slices.
When organotypic cerebellar slice cultures are prepared from rats between P1 and P5, Purkinje cells rapidly die by apoptosis (Ghoumari et al. 2000). We used P7 rat cerebella in the present study because, at this developmental stage, neurones survive and because it corresponds to a period of active dendritic remodelling and neuronal synaptogenesis (Dusart et al. 1997; Ghoumari et al. 2000) and also to the beginning of myelin formation (Notterpek et al. 1993). At later stages, myelination has progressed (Muse et al. 2001) and the evaluation of hormonal effects becomes more difficult.
Progesterone stimulates MBP expression in a dose-dependent manner
Cerebella from female P7 rats were dissected and slices were cultured for 7DIV in the presence or absence of PROG (0–75 µm). MBP immunohistological staining was quantified by using the NIH image software. At 5 µm, PROG did not affect MBP-immunostaining, even when added every day to the culture medium (Fig. 1a). Between 10 and 50 µm, PROG increased MBP expression in a dose-dependent manner, reaching a maximum at 50 µm. Concentrations of PROG higher than 100 µm were less efficient in stimulating MBP expression, possibly reflecting a cytotoxic effect, as dying Purkinje cells were observed. In the presence of 20 µm PROG, the MBP staining intensity was increased threefold when compared to control slices. When 50 µm PROG were used, the expression of MBP increased fourfold. These results were corroborated by western blot analysis (Fig. 1b). Similar results were obtained when the experiment was performed by using slices from P7 male rats (data not shown). The increase in MBP expression by PROG was mainly observed around the deep nuclei area of cerebellum (Figs 1c and d'). This region, indicated by a dotted square in Fig. 1(c'), is represented in Figs 1(d and d') at a higher magnification, used for the quantification of MBP expression.
Progesterone accelerates MBP expression in female as well as in male
We then analysed MBP expression after 3, 7 and 15DIV in slice cultures of P7 male or female rat cerebella. Slices were cultured in the presence or absence of PROG (50 µm). Control cultures were treated with vehicle (DMSO) alone. At 7DIV, the intensity of MBP staining was significantly increased by PROG (fourfold) when compared to control slices (Fig. 2). Between 7DIV and 15DIV, MBP-immunostaining also rapidly increased in the control slices and at 15DIV, MBP expression reached approximately the same level in the control slices as in the PROG-treated ones. The increase in MBP expression in PROG-treated slices at 3 and 7DIV suggest that PROG accelerates the process of myelination. The rates of myelination in the presence or absence of PROG were similar in slices from female and male rats.
The action of progesterone involves the intracellular progesterone receptor
To investigate whether the effect of PROG on MBP expression is mediated through its classical intracellular receptor (PR), we treated P7 cerebellar slice cultures with the selective PR agonist R5020 and/or with the PR antagonist RU486. R5020 (20 µm) caused a threefold increase in MBP expression (Fig. 3a), which is similar to the effect obtained with 20 µm of PROG (Figs 3a and b). Moreover, RU486 (5–10 µm), when used simultaneously, completely blocked the effects of 20 µm of PROG (Fig. 3b) or 20 µm of R5020 (data not shown). In these cases, MBP staining did not differ from control. RU486 alone had no effect. These data demonstrate that the effect of PROG is mediated by the classical PR.
We also used mice lacking the PR (PRKO mice; Lydon et al. 1995) to confirm the role of the intracellular PR in mediating the effect of PROG on MBP expression. P7 cerebellar slices were prepared from heterozygous (Htz), homozygous (Hz) and from wild-type (WT) mice. Slices from one cerebellum were separated into two groups: one group served as control and the other group was treated with 20 µm of PROG. PROG did not increase MBP expression in slices from Hz mice, as there was no difference in MBP-immunostaining between treated and non-treated slices (Figs 4c and d). In contrast, in cerebellar slices of WT and Htz mice, PROG caused, respectively, a three- and 1.5-fold increase in the intensity of MBP staining. These experiments demonstrate that PROG does not increase MBP expression in cerebellar slices from mice lacking both isoforms of the PR (PRA and PRB) and provide further evidence that the intracellular PR is necessary for the effect of PROG on myelination.
Progesterone increases the density of O4+ and GalC+ oligodendrocytes in cerebellar slice cultures
Cerebellar slices from P7 rats were treated for 7 days with 20 µm PROG or with vehicle (DMSO). The density of O4+ and GalC+ oligodendrocytes was then determined by NIH image software analysis. Figure 4 shows that PROG cause a 4.5- to fivefold increase in the density of both O4+ and GalC+ cells when compared to controls.
In P7 rat and mouse cerebellar slice cultures, some of the Purkinje cells still die by apoptosis (Dusart et al. 1997; Ghoumari et al. 2000), and PROG may have increased the number of O4+ and GalC+ oligodendrocytes indirectly by promoting neuronal survival. This was not the case as we did not observe any significant difference in the number of calbindin-positive Purkinje cells between PROG-treated and non-treated slices (control slices: 214 ± 30; PROG-treated slices: 242 ± 42 Purkinje cells). These experiments demonstrate that PROG stimulates the expression of different oligodendrocyte markers, namely, MBP, O4 and GalC to the same extent.
Metabolism of progesterone in cerebellar slices
Rat cerebellar slice cultures were incubated with 5 µm of [3H]PROG for 48 h. About 40% of the substrate were metabolized. Radioactive metabolites were first separated and characterized by TLC. Figure 5(a) shows a representative radiochromatogram. Three principal metabolites were detected in organotypic slice cultures and in the culture medium. Their Rf values corresponded to those of 5α-DHP, 3α,5α-THP and 3β,5α-THP and the identity of these metabolites was verified by GC/MS. A small amount of [3H]PROG (2%) was converted to a non-polar metabolite. Its saponification released [3H] 3α,5α-THP, suggesting a fatty acid ester of this steroid. Higher levels of PROG metabolites were detected in the medium than in the slices, indicating their release (medium: 425 nm of 5α-DHP and 250,300 nm of 3α,5α-THP and 3β,5α-THP, respectively; slices: 100 nm of 5α-DHP and 50,115 nm of 3α,5α-THP and 3β,5α-THP, respectively; Fig. 5a). In the presence of 20 µm of the 5α-reductase inhibitor L685-273, the synthesis of 5α-reduced metabolites was almost completely abolished (Fig. 5b).
Role of the 5α-reduced metabolites of progesterone in enhancing MBP expression
A series of experiments was performed to determine whether the 5α-reduced metabolites of PROG, formed in the cerebellar slices, play a role in myelination. A role of 3α,5α-THP in myelination has indeed recently been demonstrated in peripheral nerves (Magnaghi et al. 2001). Cerebellar slices from P7 rats were incubated for 7 days in the presence of PROG, 5α-DHP or 3α,5α-THP. Dose–response curves were first established (data not shown), and 20 µm or 50 µm of each steroid were found to significantly increase MBP expression. At 50 µm, PROG, 5α-DHP and 3α,5α-THP, respectively, caused a four-, three- or 2.5-fold increase in MBP-immunoreactivity (Fig. 6a). The increase in MBP expression produced by the 5α-reduced metabolites was significantly lower than that of PROG. To ascertain the role of the 5α-reduction of PROG in MBP expression, cerebellar slices were incubated with 20 µm of PROG alone or together with the selective 5α-reductase inhibitor L685-273 (50 µm). Figure 6(b) shows that L685-273 reduced the PROG-stimulated MBP-immunostaining by 50%. The simultaneous addition of 3α,5α-THP (50 µm) to the culture medium reversed this inhibition.
Involvement of membrane GABAA receptors in stimulating MBP expression
5α-DHP binds to the intracellular PR and activates gene transcription (Rupprecht et al. 1993). In contrast, 3α,5α-THP does not bind to the PR, but it is a well known positive modulator of GABAA receptors (Lambert et al. 2001). To determine whether the stimulatory effect of 3α,5α-THP on MBP expression involved GABAA receptors, we incubated cerebellar slices from P7 rats for 7 days with 50 µm of the steroid in the presence or absence of the selective GABAA receptor antagonist bicuculline (100 µm). Results show that bicuculline significantly reduced the increase in MBP-immunostaining by 3α,5α-THP (Fig. 7a). In agreement with a role of GABAA receptors in stimulating MBP expression, the selective GABAA receptor agonist muscimol (100 µm) also caused a 1.7-fold increase in MBP-immunoreactivity, comparable to that observed with 3α,5α-THP. This effect could be completely blocked by bicuculline (Fig. 7b). Taken together, these results indicate a role of GABAA receptors in myelination and they show that the stimulatory effects of the PROG metabolite 3α,5α-THP on MBP expression are mediated through these receptors.
An important role for PROG in myelination has previously been documented in the PNS (Schumacher et al. 2001). The significance of progestins in CNS myelination is more difficult to evaluate. Previous studies have shown that PROG increases the number of MBP-immunoreactive oligodendrocytes in cultures of rat glial cells (Jung-Testas et al. 1994) and that the systemic administration of PROG slightly increases the remyelination of axons in the cerebellar peduncle of old rats after a demyelinating lesion (Ibanez et al. 2003b). In the present study, we have examined the effects of PROG and of its 5α-reduced metabolites on CNS myelination in cerebellar slice cultures. This in vitro system closely reproduces developmental in vivo events and thus provides an unique model for examining the process of myelin formation and its regulation in detail (Notterpek et al. 1993). It has already been used to study the effects of PROG on dendritic growth and synaptogenesis of developing Purkinje neurones (Sakamoto et al. 2001). We evaluated the rate of myelination by measuring MBP accumulation, as the quantification of MBP-immunostaining provides a reliable and sensitive method for assessing the progress of myelination in the brain (Hamano et al. 1996; Muse et al. 2001).
The results obtained demonstrate that PROG and its 5α-reduced metabolites accelerate the rate of MBP expression in the developing rat and mouse cerebellum. Immunostaining of two other myelin-specific markers, O4 and GalC, was also increased by PROG to the same extent as that of MBP, respectively, 4.5-fold and fivefold. Whereas the stimulatory effect of PROG on MBP expression is mediated by the classical intracellular PR, the increase in MBP expression by its metabolite 3α,5α-THP involves membrane GABAA receptors.
Progesterone stimulates MBP expression in organotypic cerebellar slice cultures: involvement of the classical progesterone receptor
The quantification of MBP-immunoreactivity showed that the effect of PROG on MBP expression was dose-dependent. MBP protein expression was already significantly increased by 10 µm of PROG and highest MBP-immunoreactivity was observed at 50 µm. As only 44% of PROG added to the culture medium was metabolized within 48 h and as the hormone was replaced every second day, the cerebellar slices were constantly exposed to high micromolar concentrations of PROG throughout the experiment. With respect to our culture system, the observation that only 44% of the [3H]PROG added to the culture medium were metabolized after 48 h suggests that the hormone does not penetrate the 350 µm thick slices easily, which were cultured on top of a microporous membrane and were not submerged by the culture medium (Ghoumari et al. 2000, 2002). Nevertheless, overall PROG could be introduced in slices, but only a half was metabolized.
A time-course analysis showed that PROG accelerated MBP expression rather than enhanced it. In both control and PROG-treated slices, MBP-immunostaining considerably increased between 3 and 15DIV. At 7DIV, it was markedly increased in slices cultured in the presence of PROG when compared to control slices. However, between 7 and 15DIV, the control group caught up and MBP expression became similar in both groups at 15DIV. Because cerebellar slices were prepared from P7 rats, myelination at 15DIV may correspond to maximal myelination observed in vivo at P21 (Hamano et al. 1996; Muse et al. 2001).
PROG accelerated the MBP expression to a similar extent in cerebellar slices from both males and females. This observation suggests that PROG may play an important role in myelination during the post-natal brain development in both sexes. That is, measurable amounts of PROG, mainly of adrenal origin, circulate in the blood of both male and female rats during the first post-natal weeks (Döhler and Wuttke 1974; Weisz and Ward 1980). In addition, a functional 3β-HSD is expressed in the post-natal rat brain, suggesting that PROG is also locally synthesized (Ibanez et al. 2003a). The PR system is also functional in the post-natal rat brain of both sexes and nuclear progestin binding increases during the period of intense myelination (MacLusky and McEwen 1980; Kato et al. 1984). In the cerebellum, PR mRNA becomes detectable at P0 and rapidly increases at P7 (Sakamoto et al. 2001).
Our results demonstrate that the effect of PROG on MBP expression in cerebellar slices requires the presence of the intracellular PR, as it could be mimicked by the selective PR agonist R5020 and inhibited by the PR antagonist RU486. Moreover, PROG did not stimulate MBP expression in cerebellar slices from PRKO mice. These mice lack both isoforms of the PR (PRA and PRB), which result from differential transcription of a single gene (Lydon et al. 1995).
3α,5α-tetrahydroprogesterone stimulates MBP expression in organotypic cerebellar slice cultures: involvement of GABAA receptors
[3H]PROG, when added to the culture medium, was converted by the cerebellar slices to 5α-DHP, 3α,5α-THP and 3β,5α-THP. These metabolites were detected in the slices and were also released into the medium. Their formation was completely blocked by adding the 5α-reductase inhibitor L685-273. In addition to these three 5α-reduced metabolites, a very low quantity of a non-polar metabolite was detected in the slices. This unidentified metabolite released 3α,5α-THP after its saponification and may thus correspond to a fatty acid ester of the steroid. The brain does concentrate fatty acid esters of 3-hydroxy steroids and brain microsomes contain steroid acyl-transferase activity (Robel et al. 1987; Smith and Watson 1997). Although their biological functions have yet to be established, they are generally considered as storage forms of steroids (Larner et al. 1985).
The conversion of PROG to 3α,5α-THP by the cerebellar slices played a role in myelination. Indeed, the 5α-reductase inhibitor L685-273 partially inhibited the stimulatory effect of PROG on MBP expression, and this inhibitory effect could be reversed by the simultaneous administration of 3α,5α-THP. The increase in MBP expression by 3α,5α-THP, which does not bind to the intracellular PR, involved GABAA receptors as it could be partially blocked by the GABAA receptor antagonist bicuculline. A role for GABAA receptors in MBP expression was further confirmed by the observation that the selective agonist muscimol was also active in a bicuculline-sensitive manner.
That progestins promote myelination through the classical intracellular PR and through membrane GABAA receptors has previously been demonstrated in the PNS (Magnaghi et al. 2001). These two receptor systems were found to activate the expression of distinct peripheral myelin proteins: PROG and 5α-dihydroprogesterone, which both bind with high affinity to the intracellular PR, increased P0 expression; 3α,5α-TH PROG, a positive allosteric modulator of GABAA receptors which does not bind to the PR, increased PMP22 expression (Melcangi et al. 1999; Magnaghi et al. 2001). The finding that progestins regulate myelination in a concerted manner through the intracellular PR and through membrane GABAA receptors is a new concept and is notable, because it shows that steroids can regulate slow processes such as myelination by acting on a membrane neurotransmitter receptor.
However, although the modulation of GABAA receptors by 3α,5α-THP obviously plays a role in CNS myelination, binding of PROG to its intracellular receptor seems to be required, as the effect of PROG was completely blocked by the PR antagonist RU486 and as PROG had no effect on MBP expression in cerebellar slices from PRKO mice. A key role of the PR in increasing MBP expression and consequently myelination was further supported by the following observations: (i) PROG more efficiently increased MBP expression than 3α,5α-THP; (ii) the 5α-reductase inhibitor L685-273 only partly inhibited the effect of PROG on MBP expression, although it completely blocked the formation of 3α,5α-THP; (iii) bicuculline only partly inhibited the effect of 3α,5α-THP. This finding could be explained by the fact that 3α,5α-THP can be converted back to 5α-DHP by the 3α-HSOR, and 5α-DHP binds with high affinity to the intracellular PR and activates gene transcription (Rupprecht et al. 1993). However, once PR signalling is activated, 3α,5α-THP may have additional effect via GABAA receptors. Thus, 3α,5α-THP may promote MBP expression via modulation of membrane GABAA receptors and via the intracellular PR after its metabolism to 5α-DHP. Obviously, the respective functions of the PR and of GABAA receptors in CNS myelination, and in particular the role of the latter, remain to be clarified.
Cellular targets for the actions of progestins
Progestins may promote myelination in the CNS by acting directly on the myelinating glial cells. Indeed, oligodendrocytes in culture have been shown to express a functional PR (Jung-Testas et al. 1991) and cells of the oligodendroglial lineage express different GABAA receptor subunits, compatible with the expression of steroid-sensitive membrane receptors (N. Gago and M. El-Etr, unpublished observation). Whether progestins stimulate the maturation of oligodendrocytes or whether they promote the elaboration of the myelin sheaths remains to be studied. In favour of the former mechanism of action are the observations that PROG accelerates the process of myelination and increases expression of the early oligodendrocyte marker O4. In addition, cells of the oligodendroglial lineage express 3α-HSD mRNA and synthesize PROG from pregnenolone only during very early stages of maturation, and PSA-NCAM+ oligodendroglial preprogenitors are the cells which convert the largest amount of PROG to 3α,5α-THP (Gago et al. 2001). However, PROG has also been reported to increase MBP synthesis by acting at a post-transcriptional step (Verdi and Campagnoni 1990).
In addition, PROG may influence the myelination process indirectly by acting on neurones. Indeed, it has been shown that myelination, and in particular MBP expression, are strongly influenced by neuronal activity (Zeller et al. 1985; Demerens et al. 1996; Pareek et al. 1997). There are several observations in favor of an important role of PROG in Purkinje cell functions: these neurones express the intracellular PR during post-natal life (Sakamoto et al. 2001), express a functional 3β-HSD (Ukena et al. 1999) and PROG increases their dendritic outgrowth and the density of Purkinje spine synapses (Sakamoto et al. 2001). Purkinje cells may also be a target for 3α,5α-THP, as progestins have been shown to augment their inhibitory responses to GABA (Smith 1989). A similar situation may prevail in the PNS, where myelination by Schwann cells is under the tight control of neuronal signals (Bolin and Shooter 1993; Reynolds and Woolf 1993). Dorsal root ganglia (DRG) sensory neurones express a functional 3β-HSD, contain an intracellular PR which translocates into the nucleus during myelin synthesis by co-cultured Schwann cells, express specific genes in response to PROG treatment and express GABAA receptors (Ma et al. 1993; Guennoun et al. 1997; Chan et al. 2000). Thus, in the CNS and PNS, the actions of progestins on both neurones and myelinating glial cells may be complementary and contribute to the co-ordinated maturation of neuronal processes and their myelin sheaths.
In summary, we report here that PROG increases MBP expression and consequently might suggest an increase in myelination in organotypic slice cultures of P7 rat and mouse cerebellum. PROG accelerates the rate of MBP, especially during the first week of culture. We found that the intracellular PR is necessary for mediating this effect. PROG is further converted to 5α-DHP and 3α,5α-THP. The latter also stimulates MBP expression by a mechanisms involving GABAA receptors. Both neurones and oligodendrocytes are potential targets for the actions of PROG, and their respective contribution in the myelination process needs to be defined. The present study thus casts a new light on the influence of steroids on myelination in the CNS, in particular during the post-natal development of the rat cerebellum.
This work was partly supported by the Myelin Project (USA), the Projet Myéline (France), by the Commission of the European Communities, specific RTD program ‘Quality of life and Management of Living Resources’, QLK6-CT- 2000–00179, ‘The role of neurosteroids in healthy aging: therapeutical perspectives’ and by Artemis-Fondation Nationale de Gerontologie (Grants to Etienne Emile Baulieu).