| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Max Planck Institute of Psychiatry, 80804 Munich, Germany; and
* Faculty of Medicine, Louis Pasteur University, 67000 Strasbourg, France
1Correspondence: Max Planck Institute of Psychiatry, Kraepelinstrasse 2–10, D-80804 Munich, Germany. E-mail: osa{at}mpipsykl.mpg.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: corticosteroid • cell cycle arrest • apoptosis • HT-22 neural cell line
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 2)
. GC are increasingly being recognized as key players in the development and maintenance of brain structures, particularly the hippocampus. Thus, in vivo activation of glucocorticoid receptors (GR) attenuates neurogenesis (3)
, possibly by exerting antiproliferative effects on neuronal progenitor cells. At the same time, GR activation results in significant hippocampal cell loss (4
5
6
7)
at least partly through apoptotic mechanisms (8
, 9)
. Since a sizable number of hippocampal cells succumbing to GR-mediated apoptosis are located in the germinative or subgranular layer of the dentate gyrus, we had suggested that dividing or newly born granule cells might be glucocorticoid targets (8)
. Besides their actions on neural cell proliferation and death, glucocorticoids are involved in the differentiation of neural cells (10
, 11)
.
The molecular mechanisms underlying the effects of GC on neuronogenesis and neuronal survival are unknown. Since GR are ligand-dependent transcription factors (12)
, elucidation of GR-related transduction pathways is a challenging task. Our group has shown that administration of the highly specific GR agonist dexamethasone (DEX) to rats results in apoptosis in the hippocampus; the apoptosis was associated with alterations in the ratio of pro- (Bax) to anti-apoptotic (Bcl-2, Bcl-XL) molecules (9)
. We also demonstrated that Bax is essential for GC-induced apoptosis and that DEX treatment results in increased expression of the tumor suppressor protein p53. Other studies have implicated p53 in regulating neuronal cell numbers after manipulation of the glucocorticoid milieu. Adrenalectomy, which induces both cell death (13
, 14)
and neurogenesis (3)
in the hippocampus, is accompanied by increased p53 gene expression (15)
. Against this background and because p53 is a key player in the regulation of the proliferation and death of various neural and non-neural cells (16
17
18)
, we hypothesized that p53 may serve as the key link in GC-associated cell death, ultimately effected by its downstream target Bax in the hippocampus. In proliferating cells (including neurons), however, another p53-responsive gene p21WAF1/CIP1 is a critical player in establishment of the postmitotic/differentiated state (19
20
21
22)
.
The present studies were carried out on a cycling mouse neural cell line (HT22). We observed that rather than leading to cell death, DEX treatment results in an arrest of the cell cycle. The cells were arrested in G1, which, at least in non-neural cells, represents a critical phase when the cell must choose between cell cycle progression and quiescence (16
17
18
, 23)
. Since p53 plays a major role in this decision process, we examined the potential of GR to influence the transcriptional activity of p53 on cell cycle- or apoptosis-related genes. Indeed, GR activation by DEX was found to induce the translocation of p53 to the nucleus, enhance transcription of p53-responsive genes p21, GADD45, and bax (24
, 25)
, and increase the trans-activation potential of exogenous p53. Understanding how hormonal signals are transduced to the apoptotic or cell cycle regulatory machinery is important for elucidation of the mechanisms underlying neurogenesis/differentiation and neurodegeneration (e.g., priming cells for death induced by other signals).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 27)
and a human osteosarcoma cell line (Saos-2) were used. Cells were grown on either plastic or glass (Cell Locate®, Eppendorf, Hamburg, Germany) surfaces. All cell culture reagents were purchased from Life Technologies (Eggenstein, Germany) unless specifically mentioned otherwise. Cells were initially grown at 37°C (5% CO2) in Dulbecco’s modified medium (DMEM) supplemented with 4.5 g/L D-glucose, 1% kanamycin, 110 mg/L sodium pyruvate, Glutamax I, and 10% fetal calf serum (FCS). After 24 h, cells were transferred to the above medium supplemented with 10% dextran/charcoal-stripped bovine calf serum (BCS; Sigma Chemical Co., Deisenhofen, Germany; cortisol and corticosterone concentrations<1 nmol).
Treatments
Cells were exposed to the GR agonist DEX for 1–3 days. DEX (Fortecortin®, Merck, Darmstadt, Germany) was used at a final concentration of 10-6 M; this dose was found to produce consistent results in preliminary dose-response studies. GR activation was antagonized by pretreating cells for 2 h with RU38,486 (10-6 M in ethanol at a final concentration of 0.001%; Roussel-Uclaf, Romainville, France).
Cell counts and cell cycle analysis
At various intervals, HT22 cells were trypsinized (0.25% trypsin in Ca2+/Mg2+-free phosphate-buffered saline [PBS]) and resuspended in 1 mL of DMEM containing 0.04% trypan blue. Viable cells were counted using a hemocytometer.
For FACS analysis, cells were washed once with PBS, trypsinized as above, harvested in DMEM/10% FCS, and centrifuged (150 g, 3 min, RT). The resulting pellet was resuspended in 100 µL 0.1M PBS and added to 900 µL of cold lysis buffer (0.1M PBS containing 0.1% sodium citrate and 0.1% Triton X-100). After addition of propidium iodide (10 µg/mL; Sigma), the cell suspension was vortexed and incubated in the dark for at least 1 h (4°C). Cell cycle analysis was performed using a Coulter EPICS XL flow cytometer (flow rate of 250 cells/s; 15,000 events recorded, with single events representing 90–98% of the total populations analyzed). All parameters were measured on a linear scale and data were analyzed using Multicycle Software® (Phoenix Flow Systems, San Diego, CA).
Plasmids
The following plasmids were used: pBax-CAT (25)
, p53wt (expressing wild-type p53), p53
(expressing a point mutant of p53 that does not bind DNA) (28)
, p53-RE-CAT (PG13-CAT, a typical p53 reporter construct which contains multiple p53 binding sites linked to a basal promoter) and its negative control p53-RE-CAT (MG15-CAT) (29)
, pCMV-hGR (30)
, and pCMV-EGFP (Clontech, Heidelberg, Germany).
Transfection and reporter assays
One day before transfection with polyethylenimine (PEI; Eurogentec, Illkirch, France), HT-22 cells were seeded in DMEM-10% FCS at 70% confluence (150,000 cells/mL) in 6-well plates. The transfection procedure was performed as described (31)
. Transfection mixture containing 0.2–1.5 µg (see figure legends) was added to each culture well. After centrifugation (5 min, 250 g, RT), cells were returned to the incubator (37°C, 5% CO2) for 70 min, washed with DMEM, and incubated with stripped BCS-DMEM. DEX (10-6 M) was added to the cells 30 h before assessment of chloramphenicol acetyltransferase (CAT) activity (32)
. Cells were harvested with a rubber policeman and resuspended in 110 µL Tris HCl (250 mM, pH 7.4) before being subjected to 5 freeze-thaw cycles (liquid N2/37°C). The resulting extract was heated at 65°C for 5 min and centrifuged (20,000 g, 8 min, 4°C). The supernatant was collected, one aliquot of which was used to determine protein content (33)
. To another aliquot (90 µL), 10 µL Tris-HCl (250 mM pH 7.4) containing 0.1 mCi [14C]chloramphenicol (47 mCi/mmol; Amersham, Braunschweig, Germany) and 4 mM acetyl-coenzyme A (Sigma) were added. This latter mixture was incubated for 1 h (37°C) before adding 200 µL TMPD/xylene (1 part 2,6,10,14-tetramethyl-pentadecane: 2 parts xylene; Sigma). After thorough mixing (2x15 s) and centrifugation (10000 g, 3 min, RT), 150 µL of the aqueous phase was removed and prepared for liquid scintillation counting.
RT-PCR
Relative levels of bax, p53, p73, p21, and GADD45 mRNA were measured using semiquantitative reverse transcriptase polymerase chain reaction (RT-PCR), where GAPDH mRNA levels served as the internal standard. Total RNA, free from chromosomal DNA contamination, was isolated and reverse transcribed with SUPERSCRIPTTM II RNA H-reverse transcriptase (Life Technologies) using oligo-dT12–18 primers (MWG Biotech, Ebersberg, Germany). The gene-specific primers used were:
Murine bax (315 bp) sense 5' GGATGCGTCCACCAAGAAGC 3' antisense 5' CACCCTGGTCTTGGATCCAG 3';
Murine p21(313 bp) sense 5' AGTATGCCGTCGTCTGTTCGGTC 3' antisense 5' CCAGAGTGCAAGACAGCGACAAG 3';
Murine p73 (380 bp) sense 5' AGTTCAATTTGCTCAGCAGTGC 3' antisense 5' TGTCGGTCACATGCTCTGCCTT 3';
Murine p53 (380 bp) sense 5'CCGAGGCCGGCTCTGAGTATACCACCATCC 3' antisense 5' CTCATTCAGCTCCCGGAACATCTCGAAGCG 3';
Murine GADD45 sense 5' CTGGAGGAAGTGCTCAGCAAGG 3' antisense 5' CTGATCCATGTAGCGACTTTCC 3'.
RT-PCR was performed as described (9)
. Optimal amplifications of bax, p53, p73, p21, and GADD45 were achieved after 24, 28, 30, 28, and 26 PCR cycles, respectively, using the following conditions: denaturation for 40 s at 94°C; annealing for 40 s at 56°C (bax), 62°C (p21WAF1/CIP1), 65°C (p53), 58°C (p73), or 59°C (GADD45); primer extension for 2 min at 72°C. PCR products were electrophoresed on agarose gels, stained with ethidium bromide, and visualized under UV light. Intensities of amplified bax, p21, p53, p73, and GADD45 were normalized against those obtained for GAPDH in the same sample. Linearity of PCR amplifications was verified by amplifying serial dilutions of reverse transcribed cDNA.
Nuclear and cytoplasmic extracts were made as described previously (34)
. For total protein extracts, cells were lysed (0.1M Tris HCl, 250 mM NaCl, 1% Nonidet P-40, 1 mM EDTA, and protease inhibitor mixture; Perbio, Bonn, Germany), centrifuged (8 min, 20,000 g, 4°C), and supernatants containing equal amounts of protein were subjected to gel electrophoresis and transferred onto PVDF membranes (Bio-Rad, Munich, Germany). Blots were incubated in blocking buffer (5% dried nonfat milk in 1x PBS/0.2% Tween 20) and incubated with one of the following antisera: polyclonal anti-GR (1:2000; Santa Cruz Biotechnologies, Santa Cruz, CA), monoclonal anti-p53 (1:2000; Santa Cruz Biotechnology), monoclonal anti-p21 (1:500; PharMingen, Heidelberg, Germany), polyclonal anti-Bax (1:1000; PharMingen) and monoclonal anti-actin (0.2 µg/mL; Roche, Mannheim, Germany). Blots were washed before incubation with horseradish peroxidase-labeled secondary antibodies (Amersham; 1:2500 for 1 h at RT) and specific protein bands were detected using chemoluminescence (ECL-Plus, Amersham). Densitometry was used to obtain semiquantitative levels of the various proteins after correction for differences in actin levels in individual samples.
TUNEL reaction
Apoptosis was detected using TUNEL histochemistry (35)
on cells grown on gelatin/poly-D-lysine-coated glass coverslips and fixed in 0.1 M PBS containing 4% paraformaldehyde (10 min at RT). After rinsing in PBS and permeabilization (2 min in ice-cold PBS+0.1% sodium citrate+0.1% Triton X-100), coverslips were incubated in PBS containing 1% H2O2 (3 min) and rinsed twice with PBS. TUNEL reaction was performed for 1 h at 37°C in a solution containing terminal deoxynucleotidyl transferase (TdT; 0.15 U/µL; MBI Fermentas, Heidelberg, Germany), TdT buffer, and biotin-16-dUTP (0.01 pmol/µL; Roche). After thorough washing, coverslips were washed with PBS and incubated (1 h; RT) in streptavidin-coupled peroxidase (1:200; Sigma), 0.1% Triton X-100 in 4x SSC. After three washes, peroxidase activity was revealed with Tris-buffered 3,3'-diaminobenzidine tetrahydrochloride (0.025%; Sigma). Cell preparations were then dehydrated, cleared, mounted, and examined microscopically.
Detection of nuclear p53 by immunofluorescence
Control and DEX-treated HT-22 cells were fixed briefly in methanol/acetone (RT), washed, permeabilized with PBS+1% horse serum+0.5% BSA+0.1% Triton X-100, washed, and incubated with anti-p53 (Ab1, 2 µg/mL, 2 h, RT; Oncogene, Schwalbach, Germany). After incubation with avidin-conjugated anti-mouse IgG (Sigma), specimens were incubated with fluorescein-conjugated biotin (Vector, Burlingame, CA) and mounted in anti-fading medium before microscopic examination. HT-22 cells treated with doxorubicin (0.2 µg/mL; Adrimedac® Medac, Hamburg, Germany) served as positive controls.
Data analysis
All experiments were repeated (3–6 replicates per run) on at least three separate occasions. Quantitative and semiquantitative values were subjected to ANOVA, followed by appropriate post hoc testing (P<0.05).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Using a mouse neural cell line (HT-22) demonstrated to express functional GR (36)
, we observed that treatment with the GR agonist DEX (1 µM in charcoal-stripped, steroid-depleted medium) results in pronounced inhibition of cell proliferation. Phase contrast microscopy revealed that in the presence of DEX, HT-22 cells failed to reach confluency even after an extended period in culture (Fig. 1
A). As measured by the trypan blue exclusion test, the number of cells in DEX-treated cultures was reduced by 30% (48 h) and 42% (72 h) (Fig. 1B
; lower panel). Inhibition of cell proliferation was not observed when DEX was added to HT-22 cells growing in unstripped serum-supplemented medium (Fig. 1B
, upper panel), suggesting that steroids present in normal medium may counteract actions of the GR agonist. The DEX-induced reduction in total cell number was not due to a demise of cells through apoptosis (Fig. 1C
).
|
GR activation leads to reversible arrest in the G1 phase of the cell cycle
Flow cytometric analysis revealed that DEX-treated HT-22 cells arrest in G1 phase of the cell cycle with a concomitant reduction in the fraction of cells in the S phase (Fig. 2
A). As shown in Fig. 2B
, the percentage of cells in G1 steadily increased over time, whereas that of cells in S phase declined; the earliest inhibition of cell proliferation was observed at 12 h. Next, we examined whether the DEX-induced cell cycle arrest in HT-22 cells was due to the trans-activation properties of GR by comparing the effects of DEX with those of the GR antagonist RU38,486 (37)
. As shown in Fig. 2C
, the ability of DEX to inhibit cell proliferation was significantly attenuated (P
0.05) in the presence of RU38,486, indicating that the DEX effects observed result from activation of the glucocorticoid receptor. The slight increase of cells arrested in G1 phase observed after treatment with RU38,486 alone is consistent with the antagonist’s partial agonistic properties (38)
.
|
GR activation has been shown to induce reversible arrest of the cell cycle in G1 in many cell types of non-neural origin (39
40
41)
. Glucocorticoids have also been shown to cause irreversible apoptotic cell death (42
43
44
45)
. Our failure to observe apoptosis in HT-22 cells after DEX treatment (Fig. 1C
) suggested that the cell cycle arrest observed might be reversible after glucocorticoid withdrawal. To test this, we examined whether different durations of exposure to and withdrawal from DEX influenced the proportion HT-22 cells in the G1 phase. As depicted in Fig. 3
A, even after 48 h of continuous exposure to DEX, HT-22 cells were able to progress through G1 after removal of the glucocorticoid, restarting mitosis within 24 h of DEX withdrawal (columns A and D in Fig. 3A
).
|
In line with earlier observations (46)
immunoblot analysis revealed significant DEX-induced down-regulation of GR (Fig. 3B
). This down-regulation did not, however, interfere with the antiproliferative actions of DEX; as shown in Fig. 3A
, DEX must be continuously present to keep cells arrested at G1.
GR activation is associated with nuclear translocation and activation of p53
GR-induced growth arrest suggested that essential cell cycle regulatory components might be influenced. This prompted us to test whether GR activation might influence the activity of p53 (see ref 9
). RT-PCR analysis revealed that DEX treatment resulted in small but significant increases in the endogenous mRNA levels of three key downstream transcriptional targets of p53: GADD45, p21, and bax (Fig. 4
A).
|
We next examined whether the increased response to p53 resulted from its increased expression and/or translocation into the nucleus. No significant changes in p53 mRNA (data not shown) or protein levels (Fig. 4B
) were observed in DEX-treated HT-22 cells at the times investigated. However, immunofluorescence microscopy revealed translocation of p53 to the nucleus in 80–90% of cells after a 24 h exposure to DEX (Fig. 4C
). As a positive control in these studies, HT-22 cells were treated with doxorubicin (0.2 µg/mL), an established inducer of p53 activity (Fig. 4C
). These morphological observations were further supported by immunoblot analysis of cytoplasmic and nuclear extracts shown in Fig. 4D, E
. Mobilization of p53 was temporally associated with translocation of GR to the nucleus. Increased nuclear localization of GR and p53 first occurred within 30 min of DEX application; maximum transport of GR and p53 to the nuclear compartment was observed by 6 h after exposure to DEX, i.e., preceding the first indications of cell growth arrest by 
2 h (Fig. 2B
).
Activation of GR enhances the transcriptional activity of p53 in HT-22 cells
To assess whether activation of GR directly influences the trans-activation potential of p53, we performed transfection experiments with vectors expressing wild-type (p53wt) or a mutated, inactive form of p53 (p53; ref 28
) and a plasmid carrying the p53-responsive promoter of the human bax gene (25)
upstream of the bacterial chloramphenicol CAT gene as the reporter (pBax-CAT). As expected, expression of wild-type, but not mutant, p53 resulted in > 10-fold induction of the pBax-CAT reporter (Fig. 5
A). The presence of DEX substantially increased p53 activity ( 4.5-fold); this effect was efficiently blocked by the GR antagonist RU38,486 (Fig. 5A
). The same analysis was performed using an artificial p53-responsive promoter (29)
containing multiple p53 binding sites different from those of bax but similar to the consensus sequence (p53-RE-CAT). Again, DEX treatment resulted in a significant (threefold) stimulation of the p53-mediated induction of the p53-RE-CAT reporter (Fig. 5B
). These observations indicate that activated GR enhances the trans-activation properties of p53 independent of the promoter context and is not limited to a single target sequence (see ref 47
). In contrast to these observations with HT-22 cells, when p53 was overexpressed in p53-deficient Saos-2 cells (48)
, a decrease rather than an increase in the trans-activation potential of p53 was observed (Fig. 5C
), reminiscent of earlier observations with fibroblasts (49)
. The contrasting findings in Saos-2 and HT-22 cells underscore the cell type specificity of p53–GR interactions.
|
Mechanisms operating downstream of p53 provide an additional point of control on cell fate
Activation of p53 can lead to cell cycle arrest or apoptosis. These two phenomena seem to be mediated by different pathways (50)
. Induction of Bax by p53 is considered a crucial step in the promotion of apoptosis, whereas cell cycle arrest is frequently associated with induction of the cyclin-dependent kinase inhibitor p21 (51)
. Although p21 and bax mRNA levels were both elevated in HT-22 cells after stimulation of p53 by DEX (see Fig. 4A
), our results show that the pathway(s) leading to cell cycle arrest prevail over those leading to apoptosis. However, exposure of HT-22 cells to DEX for 24 and 48 h resulted in a significant up-regulation of p21 protein but not of Bax (Fig. 6
A). Cotransfection studies using a Bax expression vector and a fluorescent marker (EGFP) demonstrated that Bax-dependent apoptotic pathway(s) are intact in HT-22 cells. As shown in Fig. 6B
, the number of surviving HT-22 cells was markedly reduced after overexpression of Bax, implying that Bax levels might be critical in determining cell survival. These observations may explain the inability of DEX to promote cell death in HT-22 cells, suggesting an additional level for controlling Bax expression, downstream of its transcriptional induction by p53 (see ref 52
).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 9)
, albeit through as yet undefined mechanisms. Here we report that contrary to our expectation, DEX fails to induce apoptosis in the mouse neural cell line HT-22, but instead leads to a marked reduction in cell proliferation, with cells being arrested in the G1 phase of the cell cycle. The cell cycle arrest is a GR-mediated event insofar as it can be attenuated by the GR antagonist RU38,486. To our knowledge, this is the first report demonstrating GR-mediated cell cycle arrest in a neural cell.
Our study also shows that maintenance of cells in the arrested state depends on the continuous presence of DEX in the culture medium. This requirement was sustained even after GR had been substantially down-regulated. This observation suggests that a relatively small number of receptors is sufficient for the antiproliferative action of DEX to be manifest. The issue of GR concentrations is important because the levels of glucocorticoid receptors fluctuate during development (53)
, aging (54)
, and the cell cycle (55
, 56)
. Other authors working with lymphoid cells have suggested that GR down-regulation may serve to protect cells from glucocorticoid-induced cell death (57)
. Since cell cycle arrest might be a way to circumvent cell death, the present results imply that similar, but not identical (e.g., the need for continuous ligand availability), mechanisms might operate in neural cells.
On the basis of the observations that GR activation leads to either cell cycle arrest or apoptosis, it was deemed useful to analyze which regulatory components of these two processes might be involved. A key molecule that has the potential to regulate both of the above processes is the tumor suppressor p53, a transcription factor that binds to DNA in a sequence-specific manner, inducing or repressing the expression of target genes (16
, 58
, 59)
. Previous in vivo findings that DEX treatment results in elevated levels of p53 in the rat hippocampus (9)
also suggest p53 to be a likely player in GR-mediated growth arrest. Among the established p53-responsive genes that mediate the effects of p53 are p21, bax, and GADD45. In agreement with our hypothesis, significant increases in the mRNA levels of all these genes were observed. The p53 homologue p73, which trans-activates p53 target genes such as p21 and bax and induces apoptosis and growth suppression (60
, 61)
, was not detected in HT-22 cells under any culture conditions (T. M. Michaelidis, C. Crochemore, and D. Fischer, unpublished observations). Thus, the increased expression of p21, bax, and GADD45 most likely results from stimulation of p53 transcriptional activity by DEX. Subsequent transfection experiments revealed that GR activation indeed causes a substantial increase in the activity of exogenously introduced p53 on two different promoters: the bax promoter (25)
and a synthetic promoter (29)
. In the first case, transcriptional activation is mediated by a cooperative induction of three adjacent p53 half-sites that are not very similar to the consensus (47)
, whereas the synthetic promoter contains multiple p53 binding sites similar to the consensus sequence (29)
. The observation that DEX enhanced the transcriptional activity of p53 on both promoters ruled out the possibility that potentially cryptic cis-acting elements might be responsible for the stimulation of the p53-mediated transcription, supporting our conjecture that activation of GR directly influences the trans-activation potential of p53 protein.
How can activated GR potentiate the activity of p53? It is generally accepted that normally inactive p53 can be rapidly activated by a variety of stimuli (16
17
18
, 62)
. Activation of p53 usually involves alterations in the stability of the protein and its mobilization to the nucleus. Here we show that activation of GR in HT-22 cells did not increase mRNA or protein levels of p53. However, the treatment induced rapid translocation of p53 into the nucleus, as revealed by immunofluorescence and immunoblot analysis. Nuclear translocation of GR and p53 occurred contemporaneously, suggesting that physical interactions between these two transcriptional factors can occur (see ref 63
). Post-translational modifications are necessary for converting p53 into a transcriptionally competent form (64
, 65)
. Our data cannot rule out the possibility that GR activation may induce a qualitative conversion of p53 from a latent to an active form by influencing its phosphorylation, acetylation, or glycosylation status.
In contrast to what we observed with mouse-derived HT-22 cells, activation of GR in a non-neural human cell line (Saos-2) revealed an inhibitory effect on p53 activity. Together with previous reports showing an inhibitory effect of GR on p53 activity in fibroblasts (49)
and human neuroblastoma cell lines (66)
, these observations suggest a cell type- and/or species-specific mode of interaction between GR and p53. In addition to their DNA-recognition sequences, both transcription factors require protein–protein interaction with basal transcription factors and numerous accessory proteins (67)
. The existence of tissue-specific sets of transcriptional coregulators shared by different families of transcription factors (68
, 69)
may explain the distinct responses to GR activation by potentiation/attenuation of p53 in different experimental models. Exploitation of the differing results from various studies should provide a better appreciation of the cell type-specific actions of glucocorticoids and p53.
Arrest of cells in G1 accounts for most of the antiproliferative effects of glucocorticoids (48
, 70
71
72)
. Recent studies in fibroblasts (73)
, osteosarcoma (48)
, and hepatoma (74)
cells pointed to p21 as a critical component in GR-induced growth arrest. Our findings extend these data by showing that the same might be true in neural cells and that these GR-induced effects may be caused by a potentiation of the transcriptional activity of p53. Direct induction of p21 expression by GR appears to be relatively complex: its transcriptional activation has been proposed to occur via direct protein–protein interactions of the activated GR with preexisting p21 promoter-bound transcription factors (74)
and/or by the induction of C/EBP
, which binds directly to cognate DNA binding sites on the p21 promoter (75)
. On the other hand, p53 is an established regulator of p21 expression in response to several stimuli through its direct interaction with the cognate binding site on the p21 promoter (76)
. Moreover, DEX-mediated induction of p21 was recently shown to be associated with increased p53 phosphorylation, suggesting that p53 may be a critical mediator in GR-induced signaling networks that control p21 expression (77)
. Thus, although our experiments do not rule out a direct contribution of GR to the induction of p21 (41
, 74)
, the above-mentioned observations (76
, 77)
, together with the demonstration reported here of concurrent activation of two other p53-responsive genes—bax and GADD45—favor a p53-mediated effect in HT-22 cells.
Introduction of p53 in various tumor cell lines results in apoptosis whereas in normal cells it induces a reversible arrest of the cell cycle (78
79
80)
. The differential effects of p53 on normal vs. tumor cells raise the interesting question of what determines which outcome will prevail. The recent identification and analysis of mutants dissociating the apoptotic and cell cycle-arresting properties of p53 (50
, 81)
revealed that the response to p53 depends on cell type, cell environment, and genetic composition/integrity (82)
. Consistent with the notion that low levels of p53 favor activation of cell cycle arrest whereas higher levels of p53 expression are necessary for the induction of apoptosis (83)
, the moderate potentiation of p53 by GR observed in the present study might be of considerable biological significance. This study demonstrates that functional cross-talk between GR and p53, which, at least in HT-22 cells, potentiates the growth arrest properties of p53. The present work was initiated in an attempt to define the mechanisms by which glucocorticoids induce apoptosis in the hippocampus in vivo, but the model we used (HT-22 cells) led us to the discovery that at least in some neural cells, GR activation can result in cell cycle arrest. Resolving the different players and pathways leading to GR-mediated cell cycle arrest or apoptosis remains a challenging task for the future, especially in the context of neurogenesis (3)
and neural repair vs. neurodegeneration (36
; see ref 84
). Last, it merits mention that glucocorticoids are used to reduce inflammation and edema in neuro-oncology (11
, 85)
; a hitherto unrecognized benefit of such treatment may be proliferation arrest.
|
ACKNOWLEDGMENTS |
|---|
Received for publication July 10, 2001.
Revision received January 2, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
