| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Institut National de la Santé et de la Recherche Médicale Unité Mixte de Recherche (UMR) 488 (J.G., A.T., M.S., C.M.), Institut Fédératif de Recherche 93, 94276 Le Kremlin-Bicêtre Cedex, France; Faculté de Médecine Paris-Sud (J.G., A.T., M.S., C.M.) 94270 Le Kremlin-Bicêtre Cedex, France; Centre National de la Recherche Scientifique (CNRS) UMR 8125 (A.C.), Institut Gustave Roussy, 94805 Villejuif, France; and Biologie Cellulaire des Membranes (J.C.), Institut Jacques Monod, UMR 7592 CNRS/Universités Paris 6 et Paris 7, 75251 Paris Cedex 05, France
Address all correspondence and requests for reprints to: Charbel Massaad, Institut National de la Santé et de la Recherche Médicale Unité Mixte de Recherche 488, 80 rue du Général Leclerc, 94276, Le Kremlin-Bicêtre, Cedex. E-mail: massaad{at}kb.inserm.fr.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
GCs bind to and activate two major nuclear receptors, depending on their circulating levels: the high-affinity mineralocorticoid receptor and the lower affinity GR (13, 14, 15). The GR is a member of the superfamily of nuclear receptors (NRs) and contains several functional domains, among them the N-terminal region, including the ligand-independent activating function, the central DNA-binding domain, and the C-terminal hormone-binding domain, bearing the ligand-dependent activating function 2 (16). After hormonal stimulation, the GR binds as a homodimer to a glucocorticoid response element (GRE), generally located upstream of GC-sensitive genes, and acts as a transcriptional regulator. The transcriptional complex formed by the GR involves the transcription coactivators CREB-binding protein (CBP) or its close homolog p300 (17) and p160 family members (18). The latter comprise SRC-1 (steroid receptor coactivator 1; nuclear receptor coactivator 1) (19), SRC-2 [nuclear receptor coactivator 2/GR-interacting protein 1/transcriptional intermediary factor 2 (TIF-2) (20)], and SRC-3 [p300/cAMP response element binding protein (CREB)-binding protein (CBP) interacting protein/acetyltransferase/thyroid hormone receptor activator molecule 1/receptor-associated coactivator 3) (21). They are characterized by the presence of several conserved functional domains: a CBP-interacting domain (activating domain 1), a glutamine-rich region, a C-terminal activation domain (activating domain 2), and several NR boxes (LXXLL) involved in nuclear receptor binding.
Two functionally distinct isoforms of SRC-1 have been identified, SRC-1a and SRC-1e, issued from alternative splicing and differing in their C-terminal region (22). In the nervous system, SRC-1a and SRC-1e are highly expressed in the hypothalamus, hippocampus, cerebellum, thalamus, and amygdala, whereas some differences in their expression were observed between brain nuclei (23). Only a few studies have so far explored the functional significance of the different p160 coactivators. SRC-1 is thought to participate in estrogen receptor (ER) signaling, implicated in the defeminization of the male rat brain, but not in androgen receptor signaling leading to brain masculinization (24). Moreover, disruption of SRC-1 has been shown to delay Purkinje cell development and maturation, resulting in moderate motor dysfunction in adulthood (25).
Despite its close homology with SRC-1, SRC-2 seems to have specific functions: it is involved in lipid metabolism and energy balance, and appears to be essential for adhesion of Sertoli cells to germ cells. Its absence in SRC-2–/– females induces placental hypoplasia (26). In the brain, SRC-2 appears to be less expressed than SRC-1, and it is predominantly present in the anterior pituitary gland (23). SRC-2 is implicated in the hormone effects on sexual behavior, because acute administration of antisense oligodeoxynucleotides against SRC-2 into the hypothalamus compromises sexual behavior (27).
The third p160 family member, SRC-3, is mainly expressed in the hippocampus and olfactory bulbs. In contrast to SRC-1 and SRC-2, the inactivation of SRC-3 by antisense oligodeoxynucleotide did not inhibit hormone-induced reproductive behavior (27). SRC-3 is implicated in the regulatory pathway of GH and in the production of estrogen, and it is essential for both progesterone receptor (PR) and ER activities (28). Indeed, disruption of SRC-3 leads to decreased mammary gland development, even after combined treatment with estrogen and progesterone.
Given the pleiotropic effects of GCs within the nervous system, we investigated the role of the p160 family members in the GR signaling pathway within glial cells. In the central nervous system, astrocytes have been implicated in the maturation of neurons, in the regulation of their electrical activity and synaptic transmission, and in the development of neurological diseases (29, 30). In peripheral nerves, Schwann cells elaborate myelin sheaths and play an important role in regeneration processes. Moreover, perisynaptic astrocytes, as well as terminal Schwann cells, can be viewed as integral modulatory elements of synapses (31, 32).
In this study, we used primary cultures of astrocytes and an immortalized Schwann cell line (MSC80) (33). Our results show that the expression of p160 coactivators and their nucleo-cytoplasmic distribution differ between both types of glial cells, accounting for the cell-specific action of each SRC member. Moreover, by overexpression and RNA interference experiments, we showed a cell-specific recruitment of these coactivators by the GR: in Schwann cells, GR recruits SRC-1a, SRC-1e, or SRC-3, whereas in astrocytes, SRC-1e and SRC-2 are the only two active p160s toward the GR. Finally, we have shown that, in astrocytes, the NR2 domain of SRC-1a participates in its exclusion from the GR transcriptional complex, whereas SRC-1e was involved in GR transactivation.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
shows that the three p160s mRNAs were present in the two cell types. The presence of the corresponding SRC proteins was verified in astrocytes and Schwann cells by Western blotting using specific antibodies, selectively recognizing SRC-1, SRC-2, or SRC-3 proteins. As shown in Fig. 1B, all the three SRC proteins were detected in both cells, but their expression levels were uneven. Interestingly, we have observed two forms of SRC-3 in astrocytes whereas in MSC80 cells, only one form is present. The presence of one or two forms of SRC-3 was dependent on the use or not of serum depleted of steroids. The use of charcoal-treated serum provoked the presence of one form of SRC-3 (see Fig. 5B) whereas normal fetal calf serum resulted in the appearance of two forms of SRC-3 in astrocytes.
|
|
), SRC-1 and SRC-2 showed a nuclear staining in control as well as in Dex-treated cells, and no significant nucleo-cytoplasmic shuttling was observed under Dex treatment. Concerning SRC-3, we found an unusual localization. It was predominantly localized in the cytoplasm and appeared to be concentrated near the nucleus, in a layered cytoplasmic organite that could correspond to the Golgi apparatus. To confirm this hypothesis, we used an antibody recognizing giantin, a specific Golgi membrane protein (34). Double labeling revealed that giantin staining strikingly surrounded SRC-3 staining, indicating that SRC-3 may be present in the Golgi apparatus (Fig. 2B). This observation was then confirmed by analysis of tridimensional pictures (data not shown). Nevertheless, after 3 h and up to 24 h of Dex treatment, a little fraction of SRC-3 also localized in the nucleus (Fig. 2A). This unusual localization of SRC-3 seems to be specific to astrocytes, because SRC-3 localized exclusively in the nucleus of T47D cells (Fig. 2C). Finally, we have performed cell fractioning of astrocytes (Fig. 2D). We have prepared cellular extracts of astrocytes: membranes, cytoplasm, and Golgi apparatus. The Golgi apparatus was identified by means of an antibody against GM-130, a protein of the Golgi. No membrane localization of SRC-3 was detected, but we observed a faint slowly migrating band in the cytoplasmic extracts. Interestingly, we obtained two bands in the Golgi fraction: one faint slowly migrating band, and another major fast migrating band. These two bands could correspond to two forms of SRC-3. Their natures are currently under investigation.
|
), as we have previously shown (35). SRC-1 was predominantly cytoplasmic in control cells, whereas 3 h of Dex treatment led to a rapid nuclear concentration of SRC-1, as almost the entire immunoreactive protein was concentrated in the nucleus. After 24 h of Dex treatment, we observed a partial export of SRC-1 out of the nucleus and its concentration in cytoplasmic granules. SRC-2 staining was exclusively nuclear, and Dex treatment did not affect its localization, even after 24 h. Finally, SRC-3 appeared to be predominantly cytoplasmic in the absence of Dex or after 3 h of treatment. After 24 h of treatment with Dex, however, a small fraction of SRC-3 was present in the nucleus (Fig. 3A). The cytoplasmic staining appeared diffuse all around the nuclear apparatus, contrary to what was observed in astrocytes. A double staining of giantin and SRC-3 did not reveal any colocalization of these two proteins, indicating that, in contrast to astrocytes, SRC-3 was not localized in the Golgi apparatus in MSC80 cells (Fig. 3B).
|
, transfection of Rip140 inhibited chloramphenicol acetyltransferase (CAT) activity to the same extent in astrocytes and in MSC-80 cells; the maximal inhibition of GR transactivation reached 70%. This inhibition was specific to the GR signaling pathway because the basal transcriptional activity measured in the absence of Dex was not affected (data not shown), and no inhibitory effect on GRE-less promoters [simian virus 40 (SV40)-luciferase, cytomegalovirus-luciferase] was observed. These results highlighted the implication of p160 coactivator family members in GR signaling in astrocytes as well as in Schwann cells.
|
The inhibitory efficiency of each siRNA vector was assayed by immunocytochemistry and confocal analysis on astrocytes and MSC80 cells cotransfected with green fluorescent protein, as the indicator of transfected cells, and either short interfering SRC1 (siSRC1), siSRC2, or siSRC3 expression vectors. Transfected astrocytes (Fig. 5A
) and MSC80 cells (Fig. 5C) appeared green as they expressed green fluorescent protein, and endogenous coactivators were detected using specific antibodies coupled to a red fluorescent secondary antibody. Both astrocytes and MSC80 cells transfected with empty pSuper vector (green) showed staining for SRC-1, SRC-2 and SRC-3 (red), revealing that pSuper transfection did not inhibit p160s synthesis. When astrocytes or MSC80 were transfected with siSRC-1, green cells showed a lack of red SRC-1 staining, showing that siSRC-1 transfection inhibited SRC-1 synthesis in both types of glial cells. Similarly, siSRC-2 and siSRC-3, respectively, inhibited SRC-2 and SRC-3 staining in both astrocytes and Schwann cells. These inhibitions were specific because transfection with a given siSRC only inhibited red staining of the corresponding SRC in transfected cells (Fig. 5, A and C). We have further confirmed the efficiency of the siRNA inhibition of p160 expression by performing Western blots on whole-cell extracts of astrocytes (Fig. 5B) or MSC80 cells (Fig. 5D) transfected with siSRCs. Each siRNA knocked down specifically the expression of its corresponding p160 protein (Fig. 5, B and D). The extinction of the protein expression of each p160 was not total in astrocytes due to the difficulty to transfect these primary culture cells (40–50% efficiency of the transfection). Finally, the extinction of each p160 expression by siRNA did not affect the levels of the GR mRNA as suggested by real-time PCR (data not shown).
We tested the implication of SRC-1 in GR signaling in glial cells on the levels of (GRE)2-TATA promoter. Inhibition of SRC-1 expression by RNA interference provoked a 40% inhibition of GR activity in primary astrocytes and a 50% inhibition in Schwann cells, suggesting that SRC-1 may intervene in GR signaling in both cell types (Fig. 5E). SRC-2 knockdown in Schwann cells did not significantly affect Dex activation of the (GRE)2-TATA promoter, whereas it inhibited GR signaling in astrocytes (60% inhibition). These observations strongly suggest that SRC-2 may be specifically implicated in GR signaling in astrocytes. Finally, siSRC-3 knockdown inhibited GR transactivation in Schwann cells (70% inhibition), but not in astrocytes. Therefore, SRC-3 seems to play a potent role in MSC-80 cells, but not in astrocytes.
Effect of p160s Overexpression on GR Transactivation
The interesting results obtained using siRNA indicated differences between astrocytes and Schwann cells in p160s recruitment by the GR. We then studied the effect of p160s overexpression on (GRE)2-TATA-CAT promoter activity in the two cell types.
In the previous experiment, SRC-1 siRNA was not able to discriminate between SRC-1a and SRC-1e isoforms because they are very homologous, and their mRNAs differ only at their 3'-termini. Therefore, we first transfected increasing amounts of SRC-1a expression vector together with the reporter gene in astrocytes and MSC80 cells. Figure 6A revealed that in Schwann cells, SRC-1a potentiated GR signaling on (GRE)2-TATA-CAT in a dose-dependent manner whereas no significant increase in CAT activity was observed when SRC-1a was overexpressed in astrocytes. When SRC-1e expression vector was transfected, it significantly potentiated GR signaling in both astrocytes and MSC80 cells. The level of overexpressed SRC-1a and -e were assessed by Western blot on whole-cell extracts of astrocytes or MSC80 transfected with the coactivators expression vectors tagged with HA. As shown in Fig. 6B, both SRC-1a and SRC-1e were overexpressed in these cells, indicating that the lack of activity of SRC-1a in astrocytes was not due to a deficient expression.
|
) showing the enhancement of SRC-2 and SRC-3 expression levels in both cells. These results showed that the recruitment of p160 family members is cell type dependent.
Implication of the NR2 Domain of SRC-1a
Our findings suggested a cell-specific recruitment of SRC-1a in GR signaling, because overexpression of SRC-1a potentiated GR activity in Schwann cells, but not in astrocytes. To understand such a cell-specific difference, we performed a functional study based on transient transfection of a NR2 domain-deleted mutant of SRC-1a (SRC-1a 
NR2). Figure 7, A and B, showed that SRC-1a NR2 potentiated GR signaling in both cell types in a dose-dependent manner: up to 3-fold in astrocytes and 2.5-fold in Schwann cells (for 2 µg transfected vector). We have performed Western blot analysis to assess expression of overexpressed SRC-1a and SRC-1a NR2 in astrocytes and MSC80 cells. As shown in Fig. 7, C and D, SRC-1a and its deletion mutant were expressed in these cells, showing that the lack of activity of SRC-1a in astrocytes was not due to a deficiency in expression. These results indicated that in astrocytes, deletion of the C-terminal NR interacting domain of SRC-1a enabled the protein to participate in GR signaling. Thus, in this cell type, the presence of the C-terminal domain would probably exclude SRC-1a from the transcriptional complex of the GR. We assessed the expression of SRC-1 a and e isoforms in astrocytes and MSC80 by using a couple of primers that specifically recognize SRC-1e, which contains an insertion in its 3'-end, and another couple of primers able to amplify both SRC-1 (a and e) isoforms. At the same number of PCR cycles (35 cycles), we only detected the SRC-1e mRNA in astrocytes, whereas both SRC-1a and SRC-1e mRNA isoforms were detected in MSC80 cells (Fig. 7E). Thus, SRC-1a, which was the inactive isoform for GR signaling in astrocytes, has a clear difference in abundance compared with SRC-1e.
|
). As expected, we observed a Golgi localization of SRC-3 in control cells and a faint nuclear translocation in the nucleus after 24 and 48 h of treatment with Dex. Strikingly, a longer treatment with Dex (72 h) provoked the translocation of SRC-3 into the nucleus. These observations were completed by functional studies. Astrocytes were transiently transfected with (GRE)2-TATA-CAT construct and the siSRC-3 and were treated during 24, 48, and 72 h with Dex. As shown in Fig. 8B, siSRC-3 inhibition became significant after 72 h of treatment with Dex (30% inhibition). In contrast, at 24 h and 48 h, we only observed a very weak and not significant inhibition of Dex stimulation. In conclusion, the nuclear localization of SRC-3 after 72 h of treatment with Dex probably corresponds to its participation in the GR complex.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
We showed by functional overexpression experiments and siRNA-specific knockdown that in MSC80 cells the GR preferentially recruits both SRC-1 isoforms (a and e) and SRC-3, whereas in astrocytes, the GR recruits SRC-1e, SRC-2, and, to a lesser extent, SRC-3. In addition to these functional studies, we examined the cellular localization of each p160 in the two glial cells. Interestingly, we found strong differences in the subcellular localization of the p160s between astrocytes and Schwann cells.
Schwann cells exhibited a nucleo-cytoplasmic localization of the SRC-1 protein, depending on the time of Dex treatment, whereas in astrocytes this coactivator remained strictly localized in nuclear speckles independently of Dex treatment. The cytoplasmic localization of a coactivator could be unexpected. However, Kim et al. (36) showed by cellular fractionation and Western blotting that SRC-1 is present in the cytoplasm of HeLa cells. Moreover, similar observations have been reported by Amazit et al. (37), who found that the subcellular localization of SRC-1 in nonneural cells was dependent on PR signaling. Because the GR appears to be the only active steroid receptor in MSC80 cells (38), one could hypothesize that the cellular relocalization of SRC-1 that we observed may be dependent on GR activation in this cell type. In the central nervous system, Ogawa et al. (39) have demonstrated a nuclear distribution of SRC-1 in neurons and astrocytes of the hippocampus (23, 39), which is in accordance with our observation that SRC-1 is strictly nuclear in astrocytes. Other studies have reported a nuclear localization of SRC-1 in a variety of cell types (35, 36, 37, 40). Contradictory results in HeLa cells (36, 40) and COS-7 cells (37) strongly suggest that cell types and culture conditions may influence the physiology and subcellular localization of coactivators.
In both Schwann cells and astrocytes, the subcellular localization of SRC-2 was mainly nuclear. A very faint cytoplasmic staining of SRC-2 was observed in astrocytes, and this cytoplasmic fraction disappeared after 3 h of Dex treatment. An exclusive nuclear localization seems to be a general feature of SRC-2, as suggested by other studies (41, 42). However, a few exceptions have been described. Chen et al. (43) showed by confocal analysis a nucleocytoplasmic distribution of SRC-2 in proliferating myoblasts.
SRC-3 was mainly localized in the cytoplasm in the absence and 3 h after Dex treatment in Schwann cells. However, as late as 24 h after stimulation of these cells with Dex, a fraction of SRC-3 appeared nuclear. Thus, SRC-3 differs from SRC-1, which is rapidly translocated in the nucleus after Dex treatment. We hypothesize that SRC-1 may mediate rapid and transient effects of the GR, whereas SRC-3 would mediate prolonged GR transactivation: 1) SRC-1 exited the nucleus after prolonged time Dex treatment; 2) SRC-3 began to enter the nucleus when SRC-1 started to exit from it (
24 h of Dex treatment); 3) siRNA and overexpression studies have shown that both SRC-1 and SRC-3 intervene in GR signaling within MSC-80 cells. Thus, during Dex treatment, GR may first recruit SRC-1 and, if the induction lasts, SRC-3 instead of SRC-1.
In both control and Dex-treated astrocytes, SRC-3 unexpectedly localized to cytoplasmic layers on one side of the nucleus. The polarized localization of SRC-3 led us to hypothesize that the coactivator may be concentrated in the Golgi apparatus. This was confirmed by using an antibody against giantin, a vesicle-docking component of the Golgi membrane. A confocal microscopy study revealed a strict colocalization of giantin and SRC-3. Furthermore, cell fractioning of astrocytes confirmed its localization in the Golgi apparatus and raised the possibility of the presence of two forms of SRC-3 in these cells. Probably the presence of SRC-3 in the Golgi apparatus could account for a transformation and a maturation of this coactivator in this cellular compartment to be active in the nucleus. This hypothesis is currently under investigation. A confocal analysis showed that immunoreactive giantin surrounded SRC-3. From this observation, we concluded that SRC-3 may localize in the lumen of the Golgi apparatus, which may be a sequestrating organelle for the coactivator and regulate its nuclear availability during hormonal induction. Indeed, a weak nuclear staining of SRC-3 was observed after 3 h of Dex treatment, which was strongly enhanced after 72 h of treatment with Dex. The import of SRC-3 in the astrocyte nucleus correlates with its participation in GR signaling. As a matter of fact, the knockdown of SRC-3 by siRNA inhibited Dex stimulation at 72 h. Cytoplasmic or nuclear localization of SRC-3 has been shown in different cell lines (HeLa, MCF-7, mEF, or MDA-435) (44) and seems to be dependent on the proliferation status of the cells (i.e. nuclear in highly proliferative cells, cytoplasmic in differentiated cells) (44, 45). The striking differences in SRC-3 localization between astrocytes (Golgi apparatus) and MSC80 cells (cytoplasm) could account for its cell-specific actions. Such an unusual coactivator localization in the Golgi apparatus has already been reported for the putative coactivator TATA element modulatory factor/androgen receptor activator 160 in HeLa cells (46) and for the thyroid hormone receptor coactivator, thyroid hormone receptor interacting protein 230, in CV1 cells and human bladder carcinoma cells (47).
Another interesting observation we made was the absence of SRC-1a transcript in astrocytes. Furthermore, transfected SRC-1a failed to potentiate GR transactivation in these cells, whereas this isoform was active in the Schwann cell line. To understand this discrepancy, we have used deletion mutants of SRC-1a and found that the truncation of the NR2 box, which was described as a GR binding domain, allowed SRC-1a to potentiate GR signaling. Thus, in astrocytes, this NR2 box is not a GR-binding domain as previously described in two-hybrid systems (20, 48, 49), but seems to be a GR-excluding domain preventing SRC-1a to act as a GR coactivator. Moreover, in Schwann cells, the NR2 box was not the sole GR-binding domain, because its deletion did not alter GR signaling. In these cells, the two NR boxes present in SRC-1a (NR1 nd NR2) are exchangeable, as we have previously shown (35), whereas in astrocytes NR1 seems to be the exclusive GR-interacting domain. Such differences between SRC-1a and e transactivation potency were observed by Kalkhoven et al. (50): SRC-1e enhanced the ability of the ER to stimulate transcription to a greater extent than SRC-1a. Furthermore, Meijer et al. (51) have recently described that the cellular differences in the SRC-1a to SRC-1e ratio demonstrated in vivo might be involved in cell-specific responses to corticosteroids in a promoter- and ligand-dependent way. In conclusion, in glial cells, these two domains of SRC-1 harbor cell-specific GR-interacting functions, thus differing from previous data based on in vitro studies or the yeast two-hybrid system. The cell-specific actions of SRC-1a and SRC-1e could account for a specific regulation of gene expression in the different types of glial cells. The differential recruitment of the p160s by the GR in Schwann cells and astrocytes could have an important outcome on the patterns of gene regulation by glucocorticoids in glial cells.
Our observations of promoter-dependent (35) and cell-specific actions of the p160 coactivators in glial cells contribute to a better understanding of GR signaling within the nervous system. One study has described a receptor-dependent recruitment of p160s by the GR and the PR (52), whereas others have suggested that the presence and type of other response elements in the promoter may influence the interaction between the p160s and steroid receptors (50, 51). In conclusion, the interactions between a given p160 coactivator and a nuclear receptor are submitted to several constraints, dependent on multiple parameters such as the target tissue or cell type, the receptor (52), the promoter context (35), the nature of the steroid receptor-binding site (53), and the kinetics of the induction (54). To date, more than 150 coactivators and 30 corepressors have been cloned. The high number of possible combinations between those factors should be added to cellular and ligand parameters. This myriad of constraints is determinant for the specificity of gene regulation.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
The primary rat astrocytes culture was obtained using 2-d-old Sprague Dawley rats as described in Ref.55 and maintained in DMEM at 6 g/liter of glucose and supplemented with 10% fetal calf serum (Life Technologies, Inc., Gaithersburg, MD), 100 U/ml penicillin, 100 µl/ml streptomycin (Diament), and 1 µg/ml fungizone (Life Technologies). Cell fractioning of astrocytes was performed as described elsewhere (56, 57).
Plasmids
Expression vectors of wild-type SRC-1a, SRC-1e, and SRC-1a NR2 (amino acids 1–1208) have been described (37, 58, 59). They have been subcloned in the pSG5-HA expression vector (41). SRC-2/TIF-2 expression vector was a gift from H. Gronemeyer (Strasbourg, France), and SRC-3 was kindly provided by Ron Evans (San Diego, CA). Rip140 expression vector was a gift from V. Cavaillès (Montpellier, France) (60). The (GRE)2-TATA-CAT plasmid has been described previously (58). PGL2-SV40-luciferase vector was purchased from Promega Corp. (Madison, WI).
siRNA Preparation
siRNA expression vectors directed against SRC-1, SRC-2, and SRC-3 were prepared as described by Brummelkamp et al. (61) using pSuper vector. The oligonucleotides recognizing the p160s were designed as described previously (35).
RT-PCR Experiments
The total RNA from cultured MSC-80 cells or astrocytes was obtained using Rneasy mini-kit (QIAGEN, Chatsworth, CA) and reverse transcribed with random primers from Cell Signaling Technology (Beverly, MA) and reverse transcriptase M-MuLV-RT from Finnzymes (Espoo, Finland). PCR experiments were performed using Taq DNA polymerase purchased from Cell Signaling Technology, and primers specific to each SRCs from Proligo (Boulder, CO). Sequences are shown below.
SRC-1 forward: 5'-AGGAACAATGGGAAACAAC-3'
SRC-1 reverse: 5'-CCATCTGCGTCTGTTTG-3'
SRC-1e reverse: 5'-GTCACCACAGAGAAGAAC-3'
SRC-1a+e reverse: 5'-GCCTACCAGATTCACTGT-3'
SRC-2 forward: 5'-GACAGATCGTGCCAGTAACACAA-3'
SRC-2 reverse: 5'-TTCAGCTGTGAGTTGCATGAGG-3'
SRC-3 forward: 5'-GCAGATGAGTGGAGCTAGGTATG-3'
SRC-3 reverse: 5'-CACGATTACGAGGAGAAATCATG-3'
Antibodies
The antibody against either SRC-1 (mouse monoclonal) IgGk was purchased from Upstate Biotechnology, Inc. (Lake Placid, NY); SRC-2/TIF-2 (mouse monoclonal) and SRC-3/AIB-1 (mouse monoclonal) were obtained from BD Transduction Laboratories (Lexington, KY); and giantin antibody (rabbit polyclonal antibody) was purchased from Covance Research Products (Berkeley, CA). Fluorescent antibodies Alexa 488 (goat antimouse) and Cy-3 (goat antirabbit) were purchased from Molecular Probes (Eugene, OR) and Sigma Chemical Co. (St. Louis, MO), respectively. The Western blot procedure was described previously (35).
Transient Transfections
MSC80 cells and astrocytes were transiently transfected using the Polyethylenimine reagent (Sigma). MSC80 cells or astrocytes (4 x105 cells/6-cm dish) were seeded into 6 cm-dishes 1 d before the transfection and incubated in the DMEM culture medium containing 10% fetal calf serum. The (GRE)2-TATA-CAT plasmid (3 µg), the pGL2-SV40-luciferase expression vector (1 µg), and the coactivator expression vectors or mock vector at the concentrations indicated in the figure legends were mixed with a solution containing PEI (0.85 mg/ml) in serum-free DMEM. The mixture was then added on the cells for 5 h. One day after the transfection, the medium was replaced by DMEM containing 10% charcoal-treated fetal calf serum containing or not glucocorticoid agonist Dex (10–6 M).
Luciferase and CAT Assays
Luciferase activity was used to normalize the transfection efficiency in all culture dishes. The assay was performed as previously described (62). The CAT activity was determined using the two-phase assay (63).
Immunocytochemistry and Western Blot
MSC80 cells or astrocytes were seeded at the density of 2 x 105 cells in 4-cm2 glass Lab-Tek wells (Nalge Nunc, Naperville, IL). The cells were incubated with Dex (10–6 M) for 3 or 24 h (as indicated in the figure legends). The cells were then washed and fixed with Zamboni for 15 min at 4 C. After three washes, cells were permeabilized at –80 C for 30 min. The cells then were incubated with primary antibodies (dilution 1:250) overnight at 4 C and labeled with conjugated antibodies (dilution 1:400) for 45 min at room temperature. The slides were imaged using a confocal microscope LSM 510-Meta-Confocor2 (Carl Zeiss, Inc., Le Pecq, France) with a x40 (numeric aperture 1.2) lens and sequential excitation with laser lines 488 nm. Western blots were performed as described by Grenier et al. (35).
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
First Published Online September 22, 2005
Abbreviations: CAT, Chloramphenicol acetyl transferase; CBP, CREB-binding protein; CREB, cAMP response element-binding protein; Dex, dexamethasone; ER, estrogen receptor; GC, glucocorticoid; GR, glucocorticoid receptor; GRE, GC-response element; NR, nuclear receptor; PR, progesterone receptor; Rip140, receptor-interacting protein 140; siRNA, short interfering RNA; siSRC, short interfering SRC; SRC, steroid receptor coactivator; SV40, simian virus 40; TIF, transcriptional intermediary factor.
Received for publication January 25, 2005. Accepted for publication September 14, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
and steroid receptor coactivator-1. Mol Endocrinol 14:518–534 and BRG-1/hSNF2 ß and localizes in the Golgi apparatus. FEBS Lett 520:127–132[CrossRef][Medline] depends on the coactivator subtype, the type of estrogen response element, and the promoter context. Mol Endocrinol 16:2571–2581NURSA Molecule Pages Link:
This article has been cited by other articles:
 |
|
 |
|
  S. van der Laan, S. B. Lachize, E. Vreugdenhil, E. R. de Kloet, and O. C. Meijer Nuclear Receptor Coregulators Differentially Modulate Induction and Glucocorticoid Receptor-Mediated Repression of the Corticotropin-Releasing Hormone Gene Endocrinology, February 1, 2008; 149(2): 725 - 732. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Amazit, L. Pasini, A. T. Szafran, V. Berno, R.-C. Wu, M. Mielke, E. D. Jones, M. G. Mancini, C. A. Hinojos, B. W. O'Malley, et al. Regulation of SRC-3 Intercompartmental Dynamics by Estrogen Receptor and Phosphorylation Mol. Cell. Biol., October 1, 2007; 27(19): 6913 - 6932. [Abstract] [Full Text] [PDF]
|
|
|
