Dynamic Changes in Subcellular Localization of Mineralocorticoid Receptor in Living Cells: In Comparison with Glucocorticoid Receptor using Dual-Color Labeling with Green Fluorescent Protein Spectral Variants
Mayumi Nishi, Hiroshi Ogawa, Takao Ito, Ken-Ichi Matsuda, and Mitsuhiro Kawata Department of Anatomy and Neurobiology Kyoto Prefectural University of Medicine Kawaramachi Hirokoji, Kamigyo-ku Kyoto 602-8566, Japan
Mineralocorticoid receptor (MR) and glucocorticoid receptor (GR) are ligand-dependent transcription factors. Although it is generally accepted that GR is translocated into the nucleus from the cytoplasm only after ligand binding, the subcellular localization of MR is still quite controversial. We examined the intracellular trafficking of MR in living neurons and nonneural cells using a fusion protein of green fluorescent protein (GFP) and rat MR (GFP-MR). Corticosterone (CORT) induced a rapid nuclear accumulation of GFP-MR, whereas in the absence of ligand, GFP-MR was distributed in both cytoplasm and nucleus in the majority of transfected cells. Given the differential action of MR and GR in the central nervous system, it is important to elucidate how the trafficking of these receptors between cytoplasm and nucleus is regulated by ligand. To examine the simultaneous trafficking of MR and GR within single living cells, we use different spectral variants of GFP, yellow fluorescent protein (YFP) and cyan fluorescent protein (CFP), linked to MR and GR, respectively. In COS-1 cells, expressing no endogenous corticosteroid receptors, the YFP-MR chimera was accumulated in the nucleus faster than the CFP-GR chimera in the presence of 10ⴚ9 M CORT, while there was no significant difference in the nuclear accumulation rates in the presence of 10ⴚ6 M CORT. On the other hand, in primary cultured hippocampal neurons expressing endogenous receptors, the nuclear accumulation rates of the YFP-MR chimera and CFP-GR chimera were nearly the same in the presence of 0888-8809/01/$3.00/0 Molecular Endocrinology 15(7): 1077–1092 Copyright © 2001 by The Endocrine Society Printed in U.S.A.
both concentrations of CORT. These results suggest that CORT-induced nuclear translocation of MR and GR exhibits differential patterns depending on ligand concentrations or cell types. (Molecular Endocrinology 15: 1077–1092, 2001)
INTRODUCTION Adrenal corticosteroids readily enter the brain and affect the energy metabolism and stress response of the target neural cells. These effects are regulated via two receptor systems, the mineralocorticoid receptor (MR; type I corticosteroid receptor) and the glucocorticoid receptor (GR; type II corticosteroid receptor) (1, 2), both of which are ligand-dependent transcription factors (3, 4). Although it is mostly accepted that GR is translocated into the nucleus from the cytoplasm only after ligand binding (5–7), the subcellular localization of MR is still quite controversial (8–10). Recent studies have used a chimera of green fluorescent protein (GFP), a 27-kDa protein from the jellyfish Aequorea victoria, and GR (GFP-GR) showing dynamic cytoplasmic-to-nuclear translocation of GR upon ligand treatment (11, 12). Furthermore, Nishi et al. (13) demonstrated that there is no significant difference in the patterns of nuclear translocation of GFP-GR between neural cells and nonneural cells. Although fusion proteins of GFP and MR (GFP-MR) were also used to examine the dynamic changes in the subcellular localization in response to ligands using CV-1, CHO, and MDCK cells (14, 15), MR dynamics have never been studied in living neural cells. In the present study, we examined the intracellular trafficking of MR, focusing on 1) the trafficking time course, 2) 1077
MOL ENDO · 2001 1078
relationship with transcriptional sites in the nucleus, 3) ligand specificity, 4)association with the signal transduction system, 5) effects of microtubule disruption, and 6) comparison of these parameters between neurons and nonneural cells. Another interesting unsettled question is how MR interacts with GR in the brain. In the central nervous system, MRs are localized mainly in the hippocampus, while GRs are distributed throughout the brain (16–19). However, MR and GR are highly colocalized in the hippocampus, particularly in the pyramidal cells of the CA1 region (20–22). MR has a high affinity for corticosterone (CORT), a common endogenous ligand for MR and GR in rodents, and is extensively bound at low levels of circulating CORT, while GR has a lower affinity for CORT and is extensively bound at high CORT levels (1, 23–26). The differential actions of MR and GR in the hippocampal regions at different concentrations of CORT are intriguing. To elucidate possible differential responses of MR and GR to CORT, and also possible interactions between MR and GR, we investigated the subcellular distribution of MR and GR in response to CORT in single living cells using fusion proteins labeled with two different spectral variants of GFP, yellow fluorescent protein (YFP) and cyan fluorescent protein (CFP). We report here, for the first time, the results of the visualization of trafficking of MR and GR simultaneously in single living neurons in comparison with nonneural cells. Our findings clearly showed that the subcellular distribution of GFP-MR is dynamically changed in response to ligands, and that the trafficking patterns of GFP-MR in neurons and nonneural cells are nearly the same. We also demonstrated that CORT-induced nuclear accumulation of YFP-MR and CFP-GR exhibited differential time courses at different ligand concentrations and in different cell types. Furthermore, to scrutinize the different response of MR and GR to CORT, we examined the role of receptor association with heat shock proteins (hsp), particularly hsp90, in the nuclear translocation of MR and GR. Since hsp90 affects the hormone binding activity of MR (27) and GR (28), hsp90 association may provide an explanation for differential actions of CORT on MR and GR. We used geldanamycin that binds to hsp90 and inhibits hormone binding affinity to steroid hormone receptors (29). Different mechanisms that might be involved in the ligand-induced nuclear translocation of MR and GR regulated by the chaperoning effect of hsp90 in cultured hippocampal neurons and nonneuronal cells were discussed.
RESULTS Characterization of Fusion Proteins The GFP-MR, YFP-MR, or CFP-GR chimera construct was transiently transfected into COS-1 cells, and the expressed fusion proteins were analyzed by using im-
Vol. 15 No. 7
munocytochemistry, immunoblotting, and luciferase reporter assay. As a control, cells transfected with only GFP showed neither MR nor GR immunoreactivity, which demonstrates that COS-1 cells express no endogenous corticosteroid receptor proteins. GFP-MR- or YFP-MR-transfected cells showed MR immunoreactivity in both the cytoplasm and the nucleus. After treatment with CORT, MR immunoreactivity was detected within the nucleus. In the CFP-GR-transfected cells, GR immunoreactivity was found predominantly in the cytoplasm in the absence of ligand, whereas the treatment with CORT induced localization of GR immunoreactivity in the nucleus (data not shown). An immunoblot of GFP-MR- or YFP-MR-transfected COS-1cells showed a single band at 121 kDa labeled by anti-MR antibody (Fig. 1A, lane 2, and Fig. 1C, lane 1, respectively), which was at the expected size of GFP-MR and YFP-MR fusion proteins, while no specific band was observed in cells transfected with GFP alone (Fig. 1A, lane 1). The expected 108-kDa protein was detected in CFP-GR-transfected COS-1 cells using anti-GR antibody (Fig. 1C, lane 2). To elucidate functional properties of the fusion proteins, the construct was cotransfected with a reporter mouse mammary tumor virus (MMTV)-Luc (kindly provided by Dr. K. Umezono, Kyoto University) into COS-1 cells. As an internal standard, the -galactosidase gene was also cotransfected. When treated with 10⫺7 M CORT for 18 h, GFP-MR induced the activation of MMTV-Luc reporter DNA (Fig. 1B). YFP-MR and CFP-GR were also able to induce significant activation of MMTV-Luc reporter DNA (Fig. 1D). These results demonstrate that the GFP-MR, YFP-MR, and CFP-GR fusion proteins are effectively transfected into COS-1 cells and functionally active in CORT-mediated transcriptional activation of transiently transfected reporter plasmid DNA. Dynamics of GFP-MR in the Transfected Cells Time Course Study We then introduced GFP-MR fusion protein into primary cultured hippocampal neurons as well as COS-1 cells. In the absence of ligand, GFP-MR was distributed almost homogeneously in both cytoplasm and nucleus in the majority of transfected cells in both cell types (Fig. 2, A and B), although in about 10% of transfected cells, GFP-MR was observed in either the cytoplasmic or the nuclear region. There were no significant differences in the GFP-MR nuclear translocation patterns among the different cell types used in this study: COS-1 cells, which express no endogenous MR, and cultured hippocampal neurons, which express endogenous MR. Exposure to CORT caused a nuclear accumulation of GFP-MR in about 100% of the transfected cells. GFP-MR was mostly accumulated in the nuclear region within 60 min at 37 C in both COS-1 cells and hippocampal neurons (Fig. 2, A and B). In each cell type, the nuclear accumulation started at around 5 min
Real-Time Imaging of MR and GR
Fig. 1. Characterization of Fusion Proteins A, Immunoblotting of GFP-MR-transfected COS cells detected with anti-MR antibody. Lane 1, cells transfected with GFP alone showed no MR staining, whereas in lane 2, cells transfected with GFP-MR exhibited MR staining at the predicted molecular mass of 121 kDa. B, Transcriptional activation of GFP-MR in COS-1 cells. The glucocorticoid-inducible reporter MMTV-Luc was cotransfected with expression plasmids encoding GFP-MR. As an internal standard, pCH110, a mammalian positive control vector for the expression of -galactosidase, was also cotransfected in each case. The maximum induction obtained with 10⫺7 M CORT for 12 h was taken as 100 after normalization relative to -galactosidase, and the relative reporter luciferase activities were plotted. C, Immunoblotting of YFP-MR- (lane 1) and CFP-GR- (lane 2) transfected COS-1 cells detected with anti-MR and anti-GR antibody, respectively. Lane 1, YFP-MR-transfected cells showed MR staining at the predicted molecular mass of 121 kDa; while in lane 2, CFP-GR-transfected cells showed GR staining at the predicted molecular mass of 108 kDa. D, Transcriptional activities of YFP-MR- or CFP-GR-transfected cells with or without CORT induction were also examined in cells treated in the same way as described in panel B.
after CORT addition, and nearly half-maximal accumulation was observed at about 15 min. The images shown in Fig. 2, A and B, were treated with an image deconvolution procedure. The fluorescence appearance of GFP-MR in COS-1 cells became more heterogeneous in approximately 30–60 min after CORT treatment and finally showed heterogeneous dot-like
1079
distribution patterns in the nucleus, while there was no fluorescence distribution in the nucleoli. The appearances of the cells after recording images for 1 h were almost the same as those before recording images, indicating that the cells did not deteriorate during 1 h of time-lapse imaging. Relationship with Transcriptional Sites To examine the functional meaning of heterogeneous dot-like distributions of GFP-MR observed in the nuclear region after CORT treatment, we analyzed the colocalization of GFP-MR with hyperphosphorylated RNA polymerase II (pol II) by double-immunofluorescence staining using confocal laser scanning microscopy (CLSM). Nuclear distribution of pol II was detected with monoclonal antibody mara3, which specifically targets in the COOH-terminal domain of pol II (Ref. 30; kindly donated by Dr. Bart Sefton, Salk Institute, La Jolla, CA). There is a high degree of colocalization of hyperphosphorylated pol II with sites of nascent RNA, indicating that this form of pol II is in close vicinity to ongoing transcription (31). The staining of hyperphosphorylated pol II exhibited dot-like structures. The nuclear distribution of GFP-MR was detected with polyclonal anti-GFP antibody (CLONTECH Laboratories, Inc., Palo Alto, CA). The merged images showed that the dot-like staining patterns of GFP-MR partially overlapped with hyperphosphorylated pol II (Fig. 3, A and B). COS-1 cells showed less colocalization of GFP-MR with hyperphosphorylated pol II than that in cultured hippocampal neurons. Furthermore, the effect of transcriptional inhibition on the nuclear distribution pattern of GFP-MR was examined. COS-1 cells transfected with GFP-MR were pretreated with 5 g/ml of ␣-amanitin for 4 h before addition of CORT. A significant difference in the subnuclear localization patterns of GFP-MR was not observed in comparison with the case of CORT treatment alone (Fig. 3C, both panels). Ligand Specificity In addition to CORT, we employed treatment with aldosterone (ALD), dexamethasone (DEX), progesterone (PRG), or estradiol (E2) to examine the ligand specificity of GFP-MR. ALD is a specific ligand for MR with a very high affinity. DEX, a typical synthetic agonist for GR, has a chemical structure similar to that of ALD (32). Therefore, DEX shows binding affinity for MR, although the dissociation constant (Kd) value of DEX for MR is about 10-fold lower than that of ALD. It is also known that PRG, which also has a chemical structure similar to that of ALD, binds to MR with a Kd value mostly the same as that of ALD (33, 34). We observed that 1 ⫻ 10⫺ 7 M ALD, DEX, or PRG induced the nuclear translocation of GFP-MR in essentially the same manner as observed with CORT (Fig. 4, A, B, and C, respectively), and hippocampal neurons (data not shown). In contrast, E2 has a different structure at the ligand- binding site and cannot bind to MR. We found that GFP-MR was not translocated into the nuclear region in cells treated with 1 ⫻ 10⫺7 M E2 (Fig. 4D), or in hippocampal neurons (data not shown). Taken together, the present findings con-
MOL ENDO · 2001 1080
Vol. 15 No. 7
Fig. 2. Time-Lapse Imaging of GFP-MR-Transfected Cells The cells transfected with GFP-MR were cultured in the absence of serum and steroids for 24 h before observation, and then the culture medium was replaced with SFM buffered with HEPES. Representative fluorescence images of a GFP-MR-transfected COS-1 cell (A) and a hippocampal neuron (B) subjected to image deconvolution procedure were shown at 0 min, 5 min, 10 min, 15 min, 30 min, and 60 min after 10⫺9 M CORT treatment, indicating almost the same nuclear translocation manners. Nomarski differential interference contrast images of the cells (black arrow) and images of the nucleus visualized with Hoechst 33342 (HOE) (white arrow) are shown together. Note that heterogeneous dot-like distributions of GFP-MR (arrowheads) were observed in the nucleus at 30 min and 60 min after CORT treatment. GFP-MR was not distributed in the nucleolus (asterisks). Bar, 20 m.
firm that GFP-MR retains its native receptor structure and exhibits high specificity for ligands. Effects of Protein Kinase A (PKA) and Cytoskeletal Elements on Nuclear Accumulation of GFP-MR It has been shown that cAMP can promote the transcriptional activation of MR (35). This promotion
may be obtained by a direct effect on MR or through an indirect mechanism. To elucidate whether cAMP alone affects the nuclear accumulation of MR, cells were treated with forskolin (FK), an activator of cAMP-dependent protein kinase, and subjected to a time-lapse imaging study. We found that 10⫺6 M FK had no effect on the nuclear accumulation of GFP-MR in COS-1 cells (Fig. 4E) or in hippocampal neurons.
Real-Time Imaging of MR and GR
1081
Fig. 3. Relationship with Transcription Sites GFP-MR-transfected COS-1 cells and cultured hippocampal neurons were fixed and double labeled with anti-GFP antibody (green) and mara 3 against hyperphosphorylated pol II (red). Single optical sections of each fluorochrome were recorded using CLSM in independent scans. Rectangle-marked areas of 1 and 2 in COS-1 cell and cultured hippocampal neuron (A and B, respectively) are shown enlarged as green signal only (upper inset), red signal only (middle inset), or as the merged image (bottom inset). GFP-MR-transfected COS-1 cells were pretreated with ␣-amanitin (5 g/ml) to arrest transcription and then image was captured at 60 min after CORT addition at a deconvolution level (C, right). Control COS-1 cells transfected with GFP-MR were treated with CORT alone for 60 min (C, left). Bar: panel A, 20 m; panel B, 10 m; panel C, 20 m.
We next addressed the question of whether a certain pathway is involved in the receptor trafficking between the cytoplasm and the nucleus. We focused
on microtubules in the present study. We pretreated cells with 10 M colchicine or 10 g/ml nocodazole for 3 h before exposure to CORT. These doses of drugs
MOL ENDO · 2001 1082
Vol. 15 No. 7
Fig. 4. Representative Fluorescence Images of GFP-MR-Transfected COS-1 Cells Treated with 10⫺7 M ALD (A), DEX (B), PRG (C), E2 (D), or 10⫺6 M FK (E) In each treatment, cells were incubated for 0 min, 15 min, 30 min, and 60 min. Note that ALD, DEX, and PRG induced the nuclear accumulation of GFP-MR in essentially the same manner as that of CORT, whereas E2 and FK had no effect on the GFP-MR subcellular localization. Bar, 20 m.
were shown to induce an almost complete depolymerization of microtubules within 1 h (7, 36). We confirmed the microtubule disruption by immunocytochemistry using antibody against tyrosinated ␣-tubulin (37), as shown in a previous report (13). The pretreatment with colchicine or nocodazole did not cause significant changes in the manner of CORT-induced nuclear accumulation of GFP-MR in COS-1 cells and hippocampal neurons (Fig. 5, A and B, respectively). In both types of cells, GFP-MR was completely accumulated in the nuclear region within 60 min after CORT addition. Dynamics of YFP-MR and CFP-GR in Response to CORT in the Cotransfected Cells We investigated the trafficking patterns of YFP-MR and CFP-GR in response to the common ligand, CORT, in single living cells cotransfected with YFP-MR and CFP-GR. It is known that YFP makes an excellent double-label pair with CFP, which can be separately detected using conventional fluorescence microscopy (38). In the absence of CORT, CFP-GR was predominantly localized in the cytoplasmic regions, whereas YFP-MR was distributed in both the cytoplasm and nucleus. Exposure to CORT induced the nuclear localization of both CFP-GR and YFP-MR. The patterns of subcellular distribution observed in the cotransfected cells were the same as those observed in the singly transfected cells. Also, the dynamics of YFP-MR and CFP-GR trafficking were essentially the
same as those of GFP-MR and GFP-GR, respectively. Addition of an excess of CFP-GR had no effect on the localization of YFP-MR, and vice versa (data not shown). In COS-1 cells, the nuclear translocation of CFP-GR was started around 10 min after addition of 10⫺6 M CORT and was completed within 30 min (Fig. 6A). YFP-MR was also accumulated in the nuclear region within 30 min in the presence of 10⫺6 M CORT (Fig. 6B). The average nuclear/cytoplasmic ratios of YFP-MR and CFP-GR fluorescence intensities for each time point in the presence of 10⫺6 M CORT are shown in Fig. 6E (top). Each resulting curve reflected a relative increase in the nuclear intensity of YFP-MR or CFP-GR. These quantitative data and fluorescence images revealed that the rates of nuclear accumulation of YFP-MR and CFP-GR in the presence of 10⫺6 M CORT were nearly equal. In contrast, in the presence of 10⫺9 M CORT, CFP-GR started to translocate into the nuclear region after about 15 min, and its fluorescence remained in the cytoplasm even after 60 min (Fig. 6C), while YFP-MR began to accumulate in the nuclear region around 5 min after exposure to 10⫺9 M CORT, and was completely localized in the nucleus within 30 min (Fig. 6D). The results of quantitative analysis (Fig. 6E, bottom) and fluorescence images showed that the nuclear accumulation rate of YFP-MR was faster than that of CFP-GR in the presence of 10⫺9 M CORT.
Real-Time Imaging of MR and GR
1083
Fig. 5. Effects of Microtubule Disruption on the Subcellular Localization Pattern of GFP-MR GFP-MR-transfected COS-1 cells (A) and cultured hippocampal neurons (B) that were pretreated with 10 M colchicine for 3 h. Typical fluorescence images of the cells were captured at 0 min, 5 min, 10 min, 15 min, 30 min, and 60 min after exposure to 10⫺9 M CORT. GFP-MR was accumulated in the nuclear region in 60 min after CORT treatment both in COS-1 cells and cultured hippocampal neurons. To define the nuclear region of the hippocampal neuron in panel B, the image of nucleus visualized with HOE was shown. Bar, 20 m.
To examine whether YFP-MR and CFP-GR are colocalized in the nucleus or not in COS-1 cells, we analyzed the subnuclear localization of these two receptors at a deconvolution level (Fig. 6F). Both receptors showed heterogeneous dot-like distributions in the nucleus at 30 min and 60 min after addition of 10⫺6 M CORT. As shown in the merged image, the two receptors showed a partial colocalization. The colocalization rate after 60 min was higher than that observed after 30 min. In contrast, hippocampal neurons exhibited no significant differences between the rates of nuclear accumulation of CFP-GR and YFP-MR in the presence of 10⫺6 M or 10⫺9 M CORT (Fig. 7, A–D). Both CFP-GR and YFP-MR began to accumulate in the nuclear region about 10 min after the treatment and mostly localized in the nucleus within about 30 min at both low and high concentrations of CORT. These results were also confirmed by the quantitative analyses shown in Fig. 7E. The appearances of the cells after recording images for 60 min were almost the same as those before recording images, indicating that the cells were not deteriorated during 1 h of time-lapse imaging.
Effect of Geldanamycin Finally, we investigated the effects of hsp90 association on receptor trafficking using geldanamycin. COS-1 cells cotransfected with YFP-MR and CFP-GR were pretreated with 1 M geldanamycin for 3 h and then exposed to CORT. In the presence of both 10⫺6 ⫺9 M (Fig. 8, A and B) and 10 M CORT (data not shown), most of the fluorescent intensities of YFP-MR and CFP-GR remained in the cytoplasm even 60 min after CORT treatment. There were no significant differences in the inhibitory action of geldanamycin on the method of CORT-induced nuclear translocation between YFP-MR and CFP-GR. We also examined the effect of geldanamycin on cultured hippocampal neurons transfected with CFP-GR or YFP-MR in the same way as indicated for COS-1 cells. Unlike the case of COS-1 cells, both CFP-GR and YFP-MR were translocated into the nucleus within 60 min after addition of CORT, even though pretreated with 1 M geldanamycin (Fig. 8, C and D, respectively). To confirm that geldanamycin at the concentration used is actually accumulating within the primary cultured hippocampal neurons, we exam-
MOL ENDO · 2001 1084
Vol. 15 No. 7
Fig. 6. Dual-Color Imaging of GR and MR with GFP Spectral Variants in Single COS-1 Cells COS-1 cells cotransfected with CFP-GR and YFP-MR were cultured in the absence of serum and steroids for 24 h before observation, and then the culture medium was replaced with SFM buffered with HEPES for the time-lapse study. A and C, Representative fluorescence images of CFP-GR. B and D, Representative fluorescence images of YFP-MR. Cells shown in panels A and B were treated with 10⫺6 M CORT, while cells shown in panels C and D were exposed to 10⫺9 M CORT. Note that in the presence of 10⫺6 M CORT, CFP-GR and YFP-MR showed essentially the same nuclear accumulation rates, whereas YFP-MR was accumulated in the nuclear region faster than CFP-GR in the presence of 10⫺9 M CORT. E, Semiquantitative analysis of nuclear
Real-Time Imaging of MR and GR
1085
Fig. 7. Dual-Color Imaging of GR and MR with GFP Spectral Variants in Single Hippocampal Neurons Hippocampal neurons cotransfected with CFP-GR and YFP-MR were cultured in the absence of serum and steroids for 24 h before observation, and then the culture medium was replaced with SFM buffered with HEPES for the time-lapse study. A and C, Representative fluorescence images of CFP-GR. B and D, Representative fluorescence images of YFP-MR. Cells shown in panels A and B were treated with 10⫺6 M CORT, while cells in panels C and D were exposed to 10⫺9 M CORT. No significant differences in the nuclear accumulation manners of CFP-GR and YFP-MR were observed in the presence of either 10⫺6 M or 10⫺9 M CORT. E, Semiquantitative analysis of nuclear accumulation of CFP-GR and YFP-MR in hippocampal neurons. The average nuclear/cytoplasmic ratios were quantified for at least five transfected cells for each data point as described in Materials and Methods. Bar, 20 m.
ined whether geldanamycin causes induction of hsp70 by immunoblotting analysis with anti-hsp70 antibody. As shown in Fig. 8E, the treatment with 1 M geldanamycin for 3 h effectively induced hsp70 expression in both COS-1 cells (lane 2) and cultured hippocampal neurons (lane 4) as compared with control (lanes 1 and 3, respectively). The induction level of hsp70 in cultured hippocampal neurons was compared with that in COS-1 cells. Relative hsp70 levels determined from densitometric scanning of enhanced chemiluminescence-exposed film in COS-1 cells and cultured hippocampal neurons were nearly the same (i.e. 1.7-fold induction and 1.6-fold induction, respectively). As a control, we probed the samples with an antibody against glyceraldehyde-3-phosphate dehydrogenase to ensure equivalent protein loading (data not shown).
report, we first focused on the subcellular localization of GFP-MR fusion proteins in single living neurons in comparison with nonneural cells. We confirmed that our GFP fusion proteins were transcriptionally active using the MMTV-Luc reporter assay and had the expected molecular mass using immunoblotting. Although recent studies have used comparable GFPimaging systems for studying steroid receptor trafficking in living vertebrate cell lines (11, 12, 14, 15), no study has been done in living neurons. Furthermore, we first showed the dual-color labeling of MR and GR with YFP and CFP, spectral variants of GFP, respectively, in single living neurons and nonneural cells.
DISCUSSION
The subcellular distribution of MR in the absence of ligands remains quite controversial. Some studies have suggested mostly cytoplasmic (9, 10), whereas others have suggested mostly nuclear (39), localiza-
This work extends our previous study of real-time imaging of GR dynamics using GFP (13). In the present
Time-Lapse Study of GFP-MR Trafficking in Single Living Cells
accumulation of CFP-GR and YFP-MR in COS-1 cells. The average nuclear/cytoplasmic ratios were quantified for at least five transfected cells for each data point as described in Materials and Methods. F, Representative fluorescence images of COS-1 cell cotransfected with CFP-GR and YFP-MR subjected to image deconvolution procedure. Both images of CFP-GR and YFP-MR were presented in pseudocolor, green and red, respectively. Areas marked by a rectangle are shown enlarged as insets in the respective merged image. Bar, 20 m.
MOL ENDO · 2001 1086
Vol. 15 No. 7
tion. Our results observed in living hippocampal neurons and COS-1 cells showed that GFP-MR was distributed both in the cytoplasm and the nucleus in the absence of ligands in the majority of transfected cells. These results are mostly consistent with those of a previous study, which also employed GFP-MR fusion proteins in living CHO cells and CV-1 cells (15). They indicated that MR could belong to a third category within the steroid/thyroid receptor superfamily, in which the unliganded receptor is not predominantly in the cytoplasm like the GR (11–13) or in the nucleus as other members of this family (40). A shuttling of unliganded receptors between the cytoplasm and the nucleus has been reported for other steroid receptors (40–43), which is one of the possible mechanisms to explain the distribution pattern of GFP-MR in the absence of ligand observed in the present study. It is also possible that the subcellular localization of MR is altered depending on the cell cycle, although hippocampal neurons, which are postmitotic cells, are independent of cell cycles. Since COS-1 cells used in the present study were maintained in the absence of serum for 24 h before observation, most of the cells might be synchronized in the G1 phase. The cell cycle of COS-1 cells used in the present study was evaluated by using BrdU incorporation for S phase and lamin A/C or cyclin C1 antibody for G1 markers. We observed that approximately 70–80% of the cells were in G1 and about 20–30% cells were in S (data not shown). These results suggest that the heterogeneity of cell cycles observed in the present experiment causes the varieties of subcellular distribution of GFP-MR in COS-1 cells. Relationship with Transcriptional Sites
Fig. 8. Effects of Geldanamycin Treatment on CORTInduced Nuclear Translocation of YFP-MR and CFP-GR COS-1 cells cotransfected with CFP-GR and YFP-MR were cultured in the absence of serum and steroids for 24 h before observation and then pretreated with geldanamycin before 10⫺6 M CORT treatment. Panels A and B are representative fluorescence images of CFP-GR and YFP-MR, respectively, pretreated with 1 M geldanamycin for 3 h. Nuclear translocation of CFP-GR and YFP-MR were mostly inhibited. Almost all the fluorescent intensities remained in the cytoplasm even 60 min after treatment with CORT. Panels C and D are representative fluorescence images of cultured hippocampal neurons transfected with CFP-GR and YFPMR, respectively. They were pretreated with 1 M geldanamycin before 10⫺6 M CORT treatment. Note that both CFP-GR and YFP-MR were accumulated in the nucleus
After GFP-MR entered the nucleus, the fluorescence appeared to be accumulated in certain specific nuclear regions and was distributed in heterogeneous dot-like patterns in the nuclear region, like GFP-GR (12, 13). Fejes-Toth et al. (15) reported that agonistactivated MR accumulates in discrete clusters in the nucleus, and that this phenomenon occurs only with transcriptionally active MR. In contrast, van Steensel et al. (44, 45) demonstrated the spatial distribution of GR and MR in clusters in specific nuclear domains using an immunofluorescence technique with confocal microscopy. They indicated that there was no correlation between the GR clusters and the distribution of
within 60 min. Bar, 20 m. Panel E shows the effects of geldanamycin on hsp70 expression. COS-1 cells and hippocampal neurons were cultured in the presence of 0.05% DMSO (lanes 1 and 3, respectively) and 1 M geldanamycin for 3 h (lanes 2 and 4, respectively). In each lane, 30 g of proteins were loaded. Immunoblotting analysis was performed with hsp70-specific antibody. Geldanamycin increased the expression level of hsp70 in both COS-1 cells and cultured hippocampal neurons.
Real-Time Imaging of MR and GR
newly synthesized pre-mRNA, suggesting that the clusters are not directly involved in active transcription. We performed double immunofluorescence staining of GFP-MR and hyperphosphorylated pol II to examine whether the heterogeneous dot-like distribution of GFP-MR in the nucleus after CORT treatment is associated with transcription sites or not. Since hyperphosphorylated pol II is associated with sites of ongoing transcription (31), the colocalization studies could indicate the relationship between GFP-MR and active transcription sites in the nucleus. The heterogeneous dot-like nuclear distribution of GFP-MR partially coincides with that of hyperphosphorylated pol II both in COS-1 cells and cultured hippocampal neurons transfected with GFP-MR. Our results suggest that some of the receptors are associated with the transcription sites after hormone treatment, while others are not directly correlated with transcription sites. Furthermore, the effects of transcriptional inhibition on the nuclear distribution of GFP-MR were investigated by using ␣-amanitin. The nuclear distribution patterns were not significantly changed after treatment with transcription inhibitor. However, we cannot determine whether the dot-like distributions of GFP-MR induced after hormone treatment are involved in the active transcription sites or not from the transcription inhibition experiment. Because the interaction of transcription factor and active transcription sites could be very dynamic, it is difficult to evaluate whether heterogeneous dot-like patterns directly associate with active transcriptional sites or not. Recent studies showed that various nuclear proteins, such as transcription factors and splicing factors, continuously and rapidly associate and dissociate with nuclear compartments such as regulatory sites in living cells (46–49). These studies investigated the nuclear dynamics of GFPlabeled proteins in living cells using FRAP (fluorescence recovery after photobleaching) and FLIP (fluorescence loss in photobleaching) to determine how fast these nuclear proteins move within the nucleus. The techniques of FRAP and FLIP demonstrated the possibility that the nuclear proteins are freely diffusing or constrained by structure, perhaps actively recruited from one place to the next. Another possible reason for the discrepancies in the relationship between receptors and active transcription sites is that the observed focal distribution of receptors represents a primary step leading to transcriptional activation. The exact nature and the function of the nuclear distribution patterns of GFP-MR remain to be determined. Effect of PKA A recent study demonstrated that PKA, a major mediator of signal transduction pathways, modulates transcriptional activity of human MR (35). That study used transient transfection assays in HepG2 cells to show that PKA stimulates glucocorticoid response el-
1087
ement (GRE)-containing promoters in a ligand-independent manner. However, the present findings indicate that activation of MR with FK alone cannot induce nuclear accumulation of MR, the primary step in activation of specific transcription sites by MR. Massaad et al. (35) showed that cAMP alone induced a 3- to 4-fold activation of transcription, but this was about 35% of the activation induced by ALD alone. They also indicated that a synergistic activation was achieved when cells were treated with both ALD and cAMP. Taken together, these results suggest that cAMP alone cannot induce nuclear translocation of GFP-MR, but can activate small amount of MR that are distributed in the nucleus, even in the absence of ligand, and induce a lower transcriptional activation than that induced by ALD alone. When cells are treated with both ALD and cAMP, MR is fully translocated into the nucleus by binding with ALD, showing synergistic activation. Effects of Cytoskeletal Elements Extensive studies have recently focused on the nuclear-cytoplasmic transport of macromolecules including transcription factors (50, 51). These studies investigated mainly nuclear pore complexes and shuttling proteins, all of which are elements involved in transport in the region of the nuclear membrane or perinuclear sites. In contrast, very little is known about the pathway of transport of nuclear receptors from the cytoplasmic region to the perinuclear site. Some studies suggested that cytoskeletal elements are involved in nuclear translocation of receptors (52, 53), while others reported that microtubule disruption does not prevent nuclear translocation (54–56). The present study examined the effects of microtubule disruption on the nuclear translocation of GFP-MR in living cells. The GFP-MR was completely translocated into the nuclear region even after treatment with colchicine or nocodazole. These results suggest that the microtubules are not essentially involved in transporting GFP-MR from the cytoplasm to the nucleus. The same result was obtained in the case of GFP-GR (13). However, we cannot entirely rule out the possible interactions of microtubules and MR in the cells. Future studies will examine more precise relations between MR and specific organelles, including cytoskeletal elements, involved in the trafficking of nuclear receptors. Differential Responses of MR and GR to CORT We investigated the trafficking patterns by which MR and GR respond to a common natural ligand, CORT, in a single living cell using dual-color labeling with two different GFP spectral variants. The double labeling strategies with CFP and YFP have allowed simultaneous time-lapse imaging of two different receptors in single living cells. The dynamics of YFP-MR and CFP-GR were essentially the same as those of GFP-MR and GFP-GR, respectively. The trafficking
MOL ENDO · 2001 1088
characteristics of these fusion proteins in the cotransfected cells were basically the same as those in the singly transfected cells. In COS-1 cells, YFP-MR was accumulated in the nucleus faster than CFP-GR in the presence of 10⫺9 M CORT, a concentration between the Kd values of MR and GR. In contrast, no significant difference was observed in the accumulation rate in the presence of 10⫺6 M CORT, a concentration much higher than the Kd values of both receptors. Since COS-1 cells have no endogenous MR or GR, the difference in trafficking kinetics detected in the presence of 10⫺9 M CORT is considered to directly reflect the difference in affinities for CORT between MR and GR; more specifically, MR has about 10-fold higher affinity for CORT compared with GR. The findings suggest that both MR and GR are saturated in cells treated with 10⫺6 M CORT, causing the lack of difference in trafficking kinetics. These results are in agreement with the present finding that MR plays major roles at physiological concentrations of CORT, while GR is mainly effective at high concentrations of CORT (20, 23, 57, 58). Hippocampal neurons did not show any obvious difference in the nuclear accumulation rates of MR and GR in the presence of either 10⫺9 M or 10⫺6 M CORT. Since hippocampal neurons express endogenous MR and GR (20–22), these endogenous receptors may affect the trafficking of YFP-MR and CFP-GR. Another possible explanation for the present results is that hippocampal neurons, in contrast to COS-1cells, may have a unique nuclear transporting system for accumulating MR and GR in the nucleus together. A recent study showed that vesicles containing NMDA receptor 2B are transported along microtubules by KIF (kinesin superfamily)17, a neuron-specific molecular motor (59). Although our present and previous data indicated that microtubules are not essentially involved in the nuclear import of GFP-MR or GFP-GR, the results of Setou et al. (59) suggest that some receptors expressed in neuronal cells are transported by a neuronspecific molecular motor. These results led us to speculate that MR and GR could be translocated into the nucleus at mostly the same speed using specific motor molecules in cultured hippocampal neurons. To elucidate another possible mechanism of differential responses of MR and GR to CORT, we examined an effect of hsp90 association with the receptors. Since geldanamycin binds to hsp90 and inhibits hormone binding to GR and MR (29, 60), which inhibits ligand-dependent nuclear translocation of the receptors (54), it may be a good tool with which to analyze an effect of hsp90 on nuclear-cytoplasmic trafficking of corticosteroid receptors. The treatment with 1 M of geldanamycin mostly inhibited CORT-induced nuclear translocation of both YFP-MR and CFP-GR in COS-1 cells. We observed no significant differences in the inhibitory action of geldanamycin on the CORTinduced nuclear translocation manner between YFP-MR and CFP-GR. These results indicate that both receptors are associated with hsp90 in the cytoplasm,
Vol. 15 No. 7
and their ligand-binding affinities and ligand-induced nuclear translocations could be regulated by hsp90 in COS-1 cells in the same way. We also treated cultured hippocampal neurons with geldanamycin in the same way as described for COS-1 cells. To investigate whether the treatment with 1 M geldanamycin actually causes accumulation of geldanamycin within the primary cultured hippocampal neurons, we examined geldanamycin-induced induction of hsp70 (61, 62). We confirmed that geldanamycin at the concentration used highly induced hsp70 expression, showing mostly the same induction level as that in COS-1 cells. These results indicated that geldanamycin was effective in the cultured hippocampal neurons. Our data showed that geldanamycin had no effects on CORTinduced nuclear translocation of MR and GR in cultured hippocampal neurons, suggesting that the manner in which these receptors associate with chaperone protein for regulating ligand binding in hippocampal neurons differs from that of COS-1 cells. Taken together, these results suggest that neuronal cells have unique regulatory systems for nuclear translocation of GR and MR that differ from those of COS-1 cells for responding to diversified stress stimuli in the brain. Future studies are required to clarify whether hippocampal neurons utilize any specific nuclear import machinery as both receptors and the mechanism of hsp90 chaperoning in these receptors are different from nonneuronal cells. In conclusion, the present study using a GFP-MR chimera system revealed dynamic changes in the subcellular localization of GFP-MR in response to ligands. Furthermore, this is the first report to demonstrate the differential actions of two different corticosteroid receptors in a single living cell by employing dual-color imaging with YFP and CFP. This system should be quite useful for clarifying dynamic changes in the subcellular localization and interactions of functional molecules that cannot be detected in fixed cells.
MATERIALS AND METHODS Plasmid Construct The 6RMR vector (provided by Dr. P. D. Patel, Mental Health Research Institute, University of Michigan Medical School, Ann Arbor, MI) containing the rat hippocampus MR cDNA was digested with SalI. The obtained MR cDNA with a truncated 5⬘-coding region was ligated in frame into the XhoI site in the multiple cloning site of pEGFPC1 or pEYFPC1(CLONTECH Laboratories, Inc. Palo Alto, CA) using a ligation kit (Takara, Kusatsu, Japan). 6RGR vector (provided by Dr. K. R.Yamamoto, Department of Biochemistry and Biophysics, University of California, San Francisco, CA) containing the rat liver GR cDNA was digested at the BamHI sites and subcloned into pGEM-4 vector (Promega Corp., Madison, WI), resulting in pGEM-4-GR. The GR cDNA with a truncated 5⬘-coding region was isolated from pGEM-4-GR by EaeI-BamHI digestion and ligated in frame into the multiple cloning site (Bsp120I and BamHI) of pEGFPC1 or pECFPC1 (CLONTECH Laboratories, Inc.) using a ligation kit (Takara). In the resulting fusion proteins, the C terminus of pEGFPC1 or
Real-Time Imaging of MR and GR
pEYFPC1 was coupled to the 13th amino acid position of the N terminus of MR, and the C terminus of pEGFPC1 or pECFPC1 was coupled to the 20th amino acid position of the N terminus of GR. Cell Culture and Transfection Dissociated hippocampal primary neuronal cultures were prepared from 18-day-old Sprague Dawley rat fetuses according to the method of Azmitia and Hou (63). Briefly, the rat fetuses were removed from the placenta in a laminar flow hood and transferred to ice-cold dissecting solution (0.8% NaCl, 0.04% KCl, 0.006% Na2HPO4䡠12 H2O, 0.003% KH2PO4, 0.5% glucose, 0.00012% phenol red, 0.0125% penicillin G, and 0.02% streptomycin). The isolated hippocampus was mechanically dissociated by triturating through a fire-polished glass pipette. The dissociated cells were plated on 16-well glass slides precoated with 0.1 mg/ml polyethylenimine (Sigma, St. Louis, MO) at an initial plating density of 1 ⫻ 105 cells per well by adding 200 l of the cell suspension to each well (area of 0.28 cm2; Lab-Tek Chamber Slide, Nunc, Naperville, IL). The cultures were maintained in complete neuronal medium (CNM), consisting of 92.5% (vol/ vol) Eagle’s minimum essential medium (MEM, Sigma), 1% (wt/vol) nonessential amino acids (Life Technologies, Inc., Gaithersburg, MD), 0.16% (wt/vol) glucose, and 5% (vol/vol) FCS (Sigma) in a CO2 incubator at 37 C with 5% CO2/95% air. COS-1 cells were maintained in DMEM (Sigma), without phenol red, supplemented with 10% FCS overnight in a fourwell Multidish with 16-mm diameter (Nunc) at an initial plating density of 2 ⫻ 104 cells per well in 400 l of medium. Plasmid DNA was transiently transfected into cells by a liposome-mediated method using LipofectAMINE PLUS (Life Technologies, Inc.) according to the manufacturer’s instructions. Hippocampal cells were cultured in CNM for 48 h and then treated with 1 mM cytosine--D-arabinofuranoside for 24 h to suppress the proliferation of glial cells. They were cultured in serum-free medium (SFM) without steroids (MEM with 0.16% of glucose, 1% nonessential amino acids, 20 mM putrescine, 15 nM sodium selenite, 5 g/ml insulin, and 100 g/ml transferrin) for 48 h before transfection. For COS-1 cells, the medium was replaced with SFM 2 h before transfection. Cells were transfected with 200 l of OPTI MEM (Life Technologies, Inc.) containing 8 l of LipofectAMINE solution and 200 ng of plasmid DNA per well of 1 ⫻ 105 cells for 5 h at 37 C. For ligand stimulation, cells were washed in SFM and treated with 10⫺6 or 10⫺9 M CORT (Sigma), 10⫺7 M ALD (Sigma), DEX (Sigma), PRG (Sigma), or estradiol (E2) (Sigma) at 37 C. The effect of FK was examined using 10⫺6 M FK (Sigma). To analyze the role of microtubules in MR trafficking, cells were pretreated with 1 M or 10 M colchicine (Sigma) or 1 g/ml or 10 g/ml nocodazole (Sigma) for the indicated period of time and then treated with 10⫺ 9 M CORT at 37 C. To investigate the interactions between hsp90 and MR or GR, the cultured COS-1 cells were treated with 1 M or 10 nM geldanamycin (Sigma) for 3 h before CORT treatment. Immunocytochemistry and Immunoblotting The immunocytochemistry of the cultured cells was performed according to a previously described method (64). Briefly, the cells were fixed in 4% paraformaldehyde in PBS and then subjected to blocking with 2% BSA in PBS including 0.2% Triton X-100 for 1 h at room temperature. The fixed cells were incubated with rabbit polyclonal anti-GR antibody [1:2,000 dilution, (19)], rabbit polyclonal anti-MR antibody [1:500 dilution, (65)], rabbit polyclonal anti-GFP antibody (1: 1,000 dilution, CLONTECH Laboratories, Inc.), mouse monoclonal antihyperphosphorylated pol II antibody [1:10, kindly donated by Dr. Bart Sefton, Salk Institute, La Jolla, CA (30)] for 48 h at 4 C. After washing with PBS five times, cells were
1089
reacted with biotinylated goat antirabbit antibody (1:250 dilution; Roche Molecular Biochemicals, Mannheim, Germany) for 1 h at room temperature. The cultures were reacted with avidin-biotin-peroxidase complex (Vector Laboratories, Inc. Burlingame, CA) for 1 h at room temperature. The cells were visualized with 0.02% 3,3⬘-diaminobenzidine (Sigma) and 0.006% H2O2 in Tris-HCl buffered saline (pH 7.6). For dual immunofluorescence staining, primary antibodies from different sources (mouse or rabbit) were used simultaneously and detected with species-specific secondary antibodies linked to Alexa 488, Alexa 546, or Alexa 568 (Molecular Probes, Inc., Eugene, OR). To confirm disruption of microtubules, we used monoclonal antityrosinated ␣-antibody (Sigma) at a dilution of 1:800. Immunoblotting analysis was performed using COS-1cells transfected with GFP-MR, YFP-MR, or CFP-GR as described previously (13), or COS-1 cells and cultured hippocampal neurons treated with 1 M geldanamycin for hsp70 induction analysis. Proteins were separated by a 10% SDS-PAGE). Samples were electroblotted to polyvinylidene difluoride membranes (Immobilon-P, Millipore Corp., Bedford, MA) by using a semidry blotting apparatus (Transblot-SD, Bio-Rad Laboratories, Inc. Hercules, CA). The membranes were incubated with anti-GR (1:5,000 dilution), anti-MR (1:500 dilution), anti-GFP polyclonal antibody(1:1,000 dilution), or anti-hsp70 monoclonal antibody (1:1,000 dilution, StressGen Biotechnologies Corp., Victoria, British Columbia, Canada) overnight at 4 C. Secondary goat-antirabbit or mouse-horseradish peroxidase (HRP) (Bio-Rad Laboratories, Inc.) was added at 1:5,000 for 1 h at room temperature. Blots were visualized using enhanced chemiluminescence (ECL; Amersham Pharmacia Biotech, Buckinghamshire, UK). For a quantitative analysis of the effect of geldanamycin on hsp70 expression in COS-1 cells and cultured hippocampal neurons, cells were treated with 1 M geldanamycin for 3 h and then harvested on ice-cold lysis buffer (150 mM NaCl, 50 mM Tris-HCl, pH 7.6, 1% Triton X100) containing 1 M leupeptin, 1 M pepstatin A, and 1 mM phenylmethylsulfonyl fluoride PMSF at 4 C. The protein content of cell lysates was determined by using Bradford reagent (Bio-Rad Laboratories, Inc.) with BSA as standard. The following immunoblotting procedures were performed in the same way as described above except that 30 g of proteins were separated by SDS-PAGE. Relative hsp70 levels were determined from densitometric scanning of ECLexposed film using NIH image. To show that equivalent protein was loaded in each lane, the samples were probed with antiglyceraldehyde-3-phosphate dehydrogenase polyclonal antibody (1:3,000 dilution; Trevigen, Inc., Gaithersburg, MD). Examination of Transcriptional Activity COS-1 cells plated on 35-mm dishes were cotransfected with 1 g of MMTV-Luc reporter (11) and 1 g of GFP-MR, YFPMR, or CFP-GR by lipofection. One microgram of pCH110, a mammalian positive control vector for the expression of -galactosidase (66), was also cotransfected as an internal standard. Cells were transfected as described previously (13). Cell lysates were centrifuged at 12,000 rpm for 2 min at 4 C, and the luciferase activity of the resulting supernatants was assayed at 25 C using the luciferase assay system Pica Gene (Toyo Inki, Tokyo, Japan) according to the manufacturer’s protocol, and normalized to -galactosidase activity. The maximum induction obtained with 10⫺7 M CORT for 12 h was taken as 100 after normalization by the -galactosidase activity, and the relative reporter luciferase activities were plotted. Time-Lapse Image Acquisition and Analysis For the living cell imaging experiments, the culture medium was replaced with SFM buffered with 20 mM HEPES (Sigma), and the image acquisition was performed in a room temper-
MOL ENDO · 2001 1090
ature-controlled at 37 C. Images were acquired using a SenSys1400 high-resolution, cooled CCD camera (Photometrics, Tucson, AZ) attached to a microscope (IXL70, Olympus Corp., Tokyo, Japan) equipped with an epifluorescence attachment. For the observation of neurons, a 60⫻ objective lens was used, while COS-1 cells were observed with a 40⫻ objective lens. For the identification of the nuclear position, the chromatin was stained with 100 ng/ml Hoechst 33342 (Sigma). GFP fluorescence was observed using a filter set with 480-nm excitation and 515-nm emission, and a 505-nm dichroic mirror (Olympus Corp.); YFP fluorescence was viewed using a filter set with 500-nm excitation and 545-nm emission, and a 525-nm dichroic mirror (Omega Optical, Inc., Brattleboro, VT); and CFP fluorescence was observed using a filter set with 440-nm excitation and 480-nm emission, and a 455-nm dichroic mirror (Omega Optical Inc.). Data were evaluated with the image analysis software program, IPLab Spectrum (Signal Analytical Corp., Vienna, VA) or Meta Morph (Universal Imaging Corp., West Chester, PA). In the time-lapse analysis, images were captured using the timelapse program of IPLab Spectrum or Meta Morph. For highresolution analysis, an image deconvolution procedure (Meta Morph) was applied to series of images. A “nearest neighbor estimate” was calculated from the raw data. To measure nuclear/cytoplasmic ratios of the fluorescence intensity, data were collected and quantitated using a line intensity profile across the cell. For each set of conditions, the average intensities of the pixels were collected within the individual nuclei and cytoplasms of at least five cells from three independent experiments. Nuclear/cytoplasmic fluorescence ratios were calculated for each time point. The results were normalized relative to the value at 0 min taken as 1. Confocal scanning laser microscopic images were collected with a Radiance 2000 confocal microscope equipped with a 488 argon ion laser and 543 green helium-neon laser (Bio-Rad Laboratories, Inc.). An oil-immersion objective (⫻60; NA ⫽ 1.4; Nikon, Tokyo, Japan) was used.
Acknowledgments The authors thank Drs. P. D. Patel for the 6RMR plasmid, K. R. Yamamoto for the 6RGR plasmid, K. Umesono for the MMTV-Luc reporter, and B. Sefton for the hyperphosphorylated pol II antibody. The authors would also like to thank Dr. Yasunori Hayashi for critical comments on the manuscript and Masataka Sunaguchi for technical assistance.
Received June 26, 2000. Re-revision received February 1, 2001. Accepted March 16, 2001. Address requests for reprints to: Mayumi Nishi, Department of Anatomy and Neurobiology, Kyoto Prefectural University of Medicine, Kawaramachi Hirokoji, Kamigyo-ku, Kyoto 602-8566, Japan. E-mail:
[email protected]. This work was supported by Grants-in-Aid for Scientific Research from the Ministry of Education, Science, Sports and Culture, Japan (to M.N. and M.K.).
REFERENCES 1. Reul JMHM, deKloet ER 1985 Two receptor systems for corticosterone in rat brain: microdistribution and differential occupation. Endocrinology 117:2505–2511 2. Evans R, Arriza JL 1989 A molecular frame work for the action of glucocorticoid hormones in the nervous system. Neuron 2:1105–1112 3. Yamamoto KR 1985 Steroid receptor regulated transcription of genes and gene network. Annu Rev Genet 19:209–252
Vol. 15 No. 7
4. Tsai MJ, O’Malley BW 1994 Molecular mechanism of action of steroid/thyroid receptor superfamily members. Annu Rev Biochem 63:451–486 5. Gas, J-M, Baulieu EE 1987 In: Clark CR (ed) Steroid Hormone Receptors: Their Intracellular Localization. Ellis Horwood, Chichester, U.K., pp 233–250 6. Wikstrom A-C, Bakke O, Okret S, Bronnegard M, Gustafsson J-A 1987 Intracellular localization of the glucocorticoid receptor: evidence for cytoplasmic and nuclear localization. Endocrinology 120:1232–1242 7. Akner G, Wikstrom AC, Gustafsson J-A 1995 Subcellular distribution of the glucocorticoid receptor and evidence for its association with microtubules. J Steroid Biochem Mol Biol 52:1–16 8. Krozowski ZS, Rundle SE, Wallace C, Castell MJ, Shen J-H, Dowling J, Funder JW, Smith AI 1989 Immunolocalization of renal mineralocorticoid receptors with an antiserum against a peptide deduced from the complementary deoxyribonucleic acid sequence. Endocrinology 125:192–198 9. Robertson NM, Schulman G, Karnick S, Alnemri E, Litwack G 1993 Demonstration of nuclear translocation of the mineralocorticoid receptor (MR) using anti-MR antibody and confocal laser scanning microscopy. Mol Endocrinol 7:1226–1239 10. Lombes M, Binart N, Delahaye F, Baulieu EE, RafestinOblin M-E 1994 Differential intracellular localization of human mineralocorticoid receptor on binding of agonists and antagonists. Biochem J 302:191–197 11. Ogawa H, Inoue S, Tsuji F, Yasuda K, Umesono K 1995 Localization, trafficking, and temperature-dependence of the Aequorea green fluorescent protein in cultured vertebrate cells. Proc Natl Acad Sci USA 92:11899–11903 12. Htun H, Barsony J, Renyi I, Gould D, Hager GL 1996 Visualization of glucocorticoid receptor translocation and intranuclear organization in living cells with a green fluorescent protein chimera. Proc Natl Acad Sci USA 93: 4845–4850 13. Nishi M, Takenaka N, Morita N, Ito T, Ozawa H, Kawata M 1999 Real-time imaging of glucocorticoid receptor dynamics in living neurons and glial cells in comparison with non-neural cells. Eur J Neurosci 11:1927–1936 14. Chen S-Y, Wang J, Liu W, Pearce D 1998 Aldosterone responsiveness of A6 cells is restored by cloned rat mineralocorticoid receptor. Am J Physiol 274:C39–C46 15. Fejes-Toth G, Pearce D, Naray-Fejes-Toth A 1998 Subcellualr localization of mineralocorticoid receptors in living cells: effects of receptor agonists and antagonists. Proc Natl Acad Sci USA 95:2973–2978 16. de Kloet ER, Wallach G, McEwen BS 1975 Differences in corticosterone and dexamethasone binding to rat brain and pituitary. Endocrinology 96:598–609 17. McEwen BS, de Kloet ER, Rostene W 1986 Adrenal steroid receptors and actions in the nervous system. Physiol Rev 66:1121–1187 18. Funder JW, Sheppard K 1987 Adrenocorticosteroids and the brain. Annu Rev Physiol 49:397–411 19. Morimoto M, Morita N, Ozawa H, Yokoyama K, Kawata M 1996 Distribution of glucocorticoid receptor immunoreactivity and mRNA in the rat brain: an immunocytochemical and in situ hybridization study. Neurosci Res 26:235–269 20. Arriza JL, Simerly RB, Swanson LW, Evans RM 1988 The neuronal mineralocorticoid receptor as a mediator of glucocorticoid response. Neuron 1:887–900 21. van Eekelen JAM, Jiang W, de Kloet ER, Bohn MC 1988 Distribution of the mineralocorticoid and the glucocorticoid receptor mRNAs in the rat hippocampus. J Neurosci Res 21: 88–94 22. de Kloet ER 1991 Brain corticosteroid receptor balance and homeostatic control. Front Neuroendocrinol 12: 95–164
Real-Time Imaging of MR and GR
23. Krozowski ZS, Funder JW 1983 Renal mineralocorticoid receptors and hippocampal corticosterone binding species have identical intrinsic steroid specificity. Proc Natl Acad Sci USA 80:6056–6060 24. Beaumont K, Fanestil DD 1988 Characterization of rat brain aldosterone receptors reveals high affinity for corticosterone. Endocrinology 113:2043–2049 25. Rupprecht R, Reul JMHM, van Steensel B, Spengler D, Soder M, Berning B, Holsboer F, Damm K 1993 Pharmacological and functional characterization of human mineralocorticoid and glucocorticoid receptor ligands. Eur J Pharmacol 247:145–154 26. Kawata M 1995 Roles of steroid hormones and their receptors in structural organization in the nervous system. Neurosci Res 24:1–46 27. Couette B, Jalaguier S, Hellal-Levy C, Lupo B, Fagart J, Auzou G, Rafestin-Oblin ME 1998 Folding requirements of the ligand-binding domain of the human mineralocorticoid receptor. Mol Endocrinol 12:855–863 28. Yang J, DeFranco DB 1996 Assessment of glucocorticoid receptor-heat shock protein 90 interactions in vivo during nucleocytoplasmic trafficking. Mol Endocrinol 10: 3–13 29. Whitesell L, Cook P 1996 Stable and specific binding of heat shock protein 90 by geldanamycin disrupts glucocorticoid receptor function in intact cells. Mol Endocrinol 10:705–712 30. Patturajan M, Schulte RJ, Sefton BM, Berezney R, Vincent M, Bensaude O, Warren SL, Corden JL 1998 Growth-related changes in phosphorylation of yeast RNA polymerase II. J Biol Chem 273:4689–4694 31. Grande MA, van der Kraan I, de Jong J, van Driel R 1997 Nuclear distribution of transcription factors in relation to sites of transcription and RNA polymerase II. J Cell Sci 110:1781–1791 32. Joels M, de Kloet ER 1994 Mineralocorticoid and glucocorticoid receptors in the brain. Implication for ion permeability and transmitter systems. Prog Neurobiol 43:1–36 33. Funder JW 1998 Aldosterone: fact, failure and the future. Clin Exp Pharmacol Physiol Suppl 25:S47–S50 34. Turner BB 1997 Influence of gonadal steroids on brain corticosteroid receptors: a minireview. Neurochem Res 22:1375–1385 35. Massaad C, Houard N, Lombes M, Barouski R 1999 Modulation of human mineralocorticoid receptor function by protein kinase A. Mol Endocrinol 13:57–65 36. Sazapary D, Barber T, Dwyer NK, Blanchette-Mackie EJ, Simons Jr SS 1994 Microtubules are not required for glucocorticoid receptor mediated gene induction. J Steroid Biochem Mol Biol 51:143–148 37. Kreis TE 1987 Microtubules containing detyrosinated tubulin are less dynamic. EMBO J 6:2597–2606 38. Ellenberg J, Lippincott-Schwartz J, Presley JF 1999 Dual-colour imaging with GFP variants. Trends Cell Biol 9:52–56 39. Pearce PT, McNally M, Funder JW 1986 Nuclear localization of type I aldosterone binding sites in steroidunexposed GH3 cell. Clin Exp Pharmacol Physiol 13: 647–654 40. Georget V, Lobaccaro JM, Terouanne B, Mangeat P, Nicolas JC, Sultan C 1997 Trafficking of the androgen receptor in living cells with fused green fluorescent protein-androgen receptor. Mol Cell Endocrinol 129: 17–26 41. Tyagi RK, Lavrovsky Y, Ahn SC, Song CS, Chatterjee B, Roy AK 2000 Dynamics of intracellular movement and nucleocytoplasmic recycling of the ligand-activated androgen receptor in living cells. Mol Endocrinol 14: 1162–1174 42. Guiochon-Mantel A, Delabre K, Lescop P, Milgrom E 1996 Intracellular traffic of steroid hormone receptors. J Steroid Biochem Mol Biol 56:3–9
1091
43. Guiochon-Mantel A, Delabre K, Lescop P, Perrot-Applanat M, Milgrom E 1994 Cytoplasmic-nuclear trafficking of progesterone receptor. In vivo study of the mechanism of action of antiprogestins. Biochem Pharmacol 47:21–24 44. van Steensel B, Brink M, van der Meulen K, van Binnendijk EP, Wansink DG, de Jong L, de Kloet ER, van Driel R 1995 Localization of the glucocorticoid receptor in discrete clusters in the cell nucleus. J Cell Sci 108: 3003–3011 45. van Steensel B, van Binnendijk EP, Hornsby C, van der Voort TM, Krozowski ZS, de Kloet ER, van Driel R 1996 Partial colocalization of glucocorticoid and mineralocorticoid receptors in discrete compartments in nuclei of rat hippocampal neurons. J Cell Sci 109:787–792 46. Phair RD, Misteli T 2000 High mobility of profiles in the mammalian cell nucleus. Nature 404:604–609 47. McNally JG, Muller WG, Walker D, Wolford R, Hager GL 2000 The glucocorticoid receptor: rapid exchange with regulatory sites in the living cells. Science 287: 1262–1265 48. Kruhlak MJ, Lever MA, Fischle W, Verdin E, Bazett-Jones DP, Hendzel MJ 2000 Reduced mobility of the alternate splicing factor (ASF) through the nucleoplasm and steady state speckle compartments. J Cell Biol 150:41–51 49. Shopland LS, Lawrence JB 2000 Seeking common ground in nuclear complexity. J Cell Biol 150:F1–F4 50. Nakielny S, Dreyfuss G 1997 Import and export of the nuclear protein import receptor transportin by a mechanism independent of GTP hydrolysis. Curr Biol 8:89–95 51. Nigg EA 1997 Nucleocytoplasmic transport: signals, mechanisms and regulation. Nature 386:779–787 52. Akner G, Sundqvist KG, Denis M, Wikstrom AC, Gustafsson J-A 1992 Immunocytochemical localization of glucocorticoid receptor in human gingival fibroblasts and evidence for a colocalization of glucocorticoid receptor with cytoplasmic microtubules. Eur J Cell Biol 53: 390–401 53. Pratt WB 1990 Glucocorticoid receptor structure and the initial events in signal transduction. Prog Clin Biol Res 322:119–132 54. Perrot-Applanat M, Lescop P, Milgrom E 1992 The cytoskeleton and the cellular traffick of the progesterone receptor. J Cell Biol 119:337–348 55. Czar MJ, Lyons RH, Welsh MJ, Renoir JM, Pratt WB 1995 Evidence that the FK506-binding immunophilin heat shock protein 56 is required for trafficking of the glucocorticoid receptor from the cytoplasm to the nucleus. Mol Endocrinol 9:1549–1560 56. Galigniana MD, Scruggs JL, Herrington J, Welsh MJ, Carter-Su C, Housley PR, Pratt WB 1998 Heat shock protein 90-dependent (geldanamycin-inhibited) movement of the glucocorticoid receptor through the cytoplasm to the nucleus requires intact cytoskeleton. Mol Endocrinol 12:1903–1913 57. Cameron H, Woolley C, Gould E 1993 Adrenal steroid receptor immunoreactivity in cells born in the adult rat dentate gyrus. Brain Res 611:342–346 58. de Kloet ER, Rots NY, van Den Berg DTWM, Oitzl MS 1994 Brain mineralocorticoid receptor function. Ann NY Acad Sci 746:8–21 59. Setou M, Nakagawa T, Seog D-H, Hirokawa N 2000 Kinesin superfamily motor protein KIF 17 and mLin-10 in NMDA receptor-containing vesicle transport. Science 288:1796–1802 60. Bamberger CM, Wald M, Bamberger AM, Schulte HM 1997 Inhibition of mineralocorticoid and glucocorticoid receptor function by the heat shock protein 90-binding agent geldanamycin. Mol Cell Endocrinol 131:233–240 61. Xiao N, Callaway CW, Lipinski CA, Hicks SD, DeFranco DB 1999 Geldanamycin provides posttreatment protec-
MOL ENDO · 2001 1092
tion against glutamate-induced oxidative toxicity in a mouse hippocampal cell line. J Neurochem 72:95–101 62. Kim HR, Kang HS, Kim HD 1999 Geldanamycin induces heat shock protein expression through activation of HSF1 in K562 erythroleukemic cells. IUBMB Life 48: 429–433 63. Azmitia EC, Hou XP 1994 Steroid regulation of neurotrophic activity: primary microcultures of midbrain raphe and hippocampus. In: deKloet ER, Sutanto W (eds) Methods in Neuroscience: Neurobiology of Steroids. Academic Press, San Diego, CA, vol 22:359–371
Vol. 15 No. 7
64. Nishi M, Whitaker-Azmitia PM, Azmitia EC 1996 Enhanced synaptophysin immunoreactivity in rat hippocampal culture by 5-HT1A agonist, S100b, and corticosteroid receptor agonists. Synapse 23:1–9 65. Ito T, Morita N, Nishi M, Kawata M 2000 In vitro and in vivo immunohistochemistry for the distribution of mineralocorticoid receptor with specific antibody. Neurosci Res 37:173–182 66. Hall CV, Jacob PE, Ringold GM, Lee F 1983 Expression and regulation of Escherichia coli lacZ gene fusion in mammalian cells. J Mol Appl Genet 2:101–109
