L ' L : " .~, ~, ~Jl

E L S E V I E R Biochimica et Biophysica Acta 1260 (1995) 308-314

Biochi~ic~a et Biophysica A~ta

DNA binding activity of the glucocorticoid receptor is sensitive to redox changes in intact cells

Franca Esposito, Franca Cuccovillo, Fernando Morra, Tommaso Russo, Filiberto Cimino *

Dipartimento di Biochimica e Biotecnologie Mediche, Universith degli Studi di Napoli 'Federico H', via S. Pansini 5, 80131 Napoli, Italy

Received 23 March 1994; revised 26 July 1994; accepted 15 September 1994

Abstract

The effect of changes of redox conditions on glucocorticoid receptor (GR) activity in intact cells has been studied using two approaches. One was to evaluate the GR-DNA binding in extracts of COS2 cells transiently overexpressing GR and in which reactive oxygen intermediates (ROD accumulate as a consequence of glutathione (GSH) depletion. GR-DNA binding was significantly decreased in COS2 cells treated with diethylmaleate (DEM), which causes GSH depletion by forming GSH-DEM complexes. A similar effect was observed for Spl, another Zn-finger transcription factor, whereas no difference was observed for the C/EBP transcription factor, which is known to be unaffected by redox changes in vitro. N-Acetylcysteine (NAC), which counteracts the effects of DEM by increasing GSH biosynthesis, prevents the decrease of GR-DNA binding in cells treated with DEM. The GR-DNA binding efficiency was similarly decreased using extracts from H 202-treated COS2 cells and from COS2 cells treated with buthionine sulphoximine, which causes GSH depletion via a mechanism different from that of DEM. The other approach was to evaluate the efficiency of a GR-regulated promoter under different redox conditions. In HeLa cells, transfected with a plasmid containing the CAT gene under the control of the glucocorticoid responsive element (GRE) within the mouse mammary tumor virus promoter, and treated with dexamethasone to activate GR, exposure to DEM significantly impaired the activation of CAT gene expression induced by dexamethasone. Also in this case NAC treatment inhibited the effects of DEM.

Keywords: Glucocorticoid receptor; Transcriptional factor; Oxidant; Anti-oxidant; DNA-protein interaction

1. Introduction

Recent evidence suggests that the activity of many transcriptional factors is under redox control. In vitro experiments have demonstrated that the transcriptional fac- tor USF efficiently interacts with DNA only if two cys- teines, present in the helix-loop-helix domain, are reduced [1], and that p53 binds to DNA only in the presence of reducing agents [2,3]. Fos-Jun heterodimers bind the AP-1 DNA consensus element only if two conserved cysteines (Cys-154 in Fos and Cys-272 in Jun), flanked by basic amino acids, are in a reduced state [4]. Furthermore,

Abbreviations: CAT, chloramphenicol acetyltransferase; DEM, dieth- ylmaleate; GR, glucocorticoid receptor; GRE, glucocorticoid responsive element; GSH, glutathione; NAC, N-acetylcysteine; ROI, reactive oxygen intermediates.

* Corresponding author. Fax: + 39 81 7463650.

0167-4781/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSDI 0167-4781(94)00209-6

transactivators encoded by viral genomes are also sensitive to redox changes; this is the case for BZLF1 [5], a leucine zipper factor encoded by the Epstein Barr viral genome, and for E2 protein, a transacting factor of type 1 bovine papilloma virus [6]. Not all the transcriptional factors examined show the same behaviour as those mentioned above. In fact, C / E B P is not affected by redox changes [5] and HoxB5, a human homeodomain gene, is activated by oxidation [7].

Zn-finger-containing transcriptional factors seem to be ideal candidates for regulation by redox conditions. The DNA binding domain of GR contains two Zn-finger struc- tures in which two groups of four cysteines are tetrahedri- cally coordinated with two Zn(II) ions [8,9]. Chemical modification of the sulfhydryl moieties prevents the bind- ing of the protein to DNA [10], thus suggesting an interfer- ence with the formation of Zn-finger structures. Further- more, the formation of intramolecular disulfide bonds be- tween the cysteines, obtained by treating the protein with

F. Esposito et al. / Biochimica et Biophysica Acta 1260 (1995) 308-314 309

oxidizing agents, results in a complete loss of GR-DNA binding activity [11]. Here we describe the activity of GR and of Spl, another Zn-finger transcription factor [12], and observe the effect of manipulating the redox state of the intracellular environment. The DNA binding efficiency of both factors decreased as a consequence of increased levels of ROI induced by glutathione depletion. Under these conditions the transcription driven by the promoter present in the mouse mammary tumor virus-long terminal repeat, containing a glucocorticoid responsive element, is significantly affected.

2. Materials and methods

2.1. Cell cultures

cellular extracts as well as Cat assays were performed as described in [13] using 10 /zg of protein per assay, [14C]chloramphenicol (Amersham, specific activity 54 mCi/mmol) and acetyl-CoA (Sigma, 0.5 mM final con- centration). Protein concentrations were determined ac- cording to Bradford [16]. The amount of acetylated chlor- amphenicol was assayed by scraping the radioactive spots from the thin layer chromatographic sheets and counting the 14C-labeled compounds in a liquid scintillation counter (Beckman). COS2 cells (4 .106 cells/100 mm plastic dish) were transfected with 20 /zg of the plasmid pSTC407-556 [17] by electroporation (Gene Pulser, Bio- Rad; 220 V, 250 /zFD). Where indicated, 36 h after transfection, the oxidizing and/or antioxidant agents were added to the culture medium for 6 h.

HeLa cells were cultured at 37°C, in a 5% CO 2 atmo- sphere in minimal essential medium (MEM, Gibco-BRL) supplemented with 8% calf serum (Gibco-BRL), penicillin (50 IU/ml) and streptomycin (50 /zg/ml). COS2 cells were cultured in Dulbecco's modified Eagle's medium (Gibco-BRL) supplemented with 10% fetal calf serum (Hiclone), plus antibiotics.

The amount necessary to reach the desired concentra- tion of diethylmaleate (DEM, Sigma) was added directly to the medium in the plate, using a stock solution of 6 M DEM in dimethylsulfoxide, and the plate was gently ro- tated to obtain uniform distribution of the compound. Similarly, the required amount of a freshly prepared solu- tion of H202 (50 mM, Aldrich) was added directely to the cells (200 /~M final concentration). Buthionine sulfox- imine (0.5 M, Sigma), in aqueous solution and NAC (0.5 M, Sigma) in PBS were added to the culture medium to obtain final concentrations of 5 mM and 30 mM, respec- tively. Cell viability, measured at different times and con- centrations of DEM, was determined microscopically by counting cells that exclude the trypan blue.

2.2. Transfections

HeLa cells were plated at a density of 2 • 105 cells/60 mm plastic dish. The following day, the medium was changed at least 4 h before transfection, and 10 /zg of plasmid pSV2-CAT [13] or MMTV-CAT [14] were trans- fected using the calcium phosphate method [15]. Where indicated, 36 h after transfections either DEM or H202 were added to the culture medium for an additional 6 h. Where indicated, cells were exposed to NAC 2 h before the addition of DEM to the culture medium. In the experi- ments where the MMTV-CAT plasmid was transfected, 36 h after transfection dexamethasone (Sigma) was also added to the cells, in 0.5% calf serum instead of 8% calf serum, to obtain a final concentration of 1 /zM. Cells were harvested 42 h after transfection in TEN buffer (40 mM Tris-HC1, pH 7.5, 1 mM EDTA, 150 mM NaC1) and

2.3. Gel retardation assays

Whole cellular extracts used for mobility shift assays were prepared according to Ref. [18]. Briefly, HeLa cell plates were washed twice with phosphate buffered solu- tion, the cells were scraped from the plates and lysed by the addition of 100 p,l/100 mm plate of a buffer contain- ing 20 mM Tris-HC1, pH 7.5, 20% glycerol, 2 mM DTT and 400 mM KCI. Two freezing and thawing cycles were performed and, after centrifugation for 10 min at 4°C, the supernatants were stored in aliquots at -80°C.

Whole cellular extracts were similarly prepared from COS2 cells 42 h after transfection with the pSTC 407-556 plasmid. The different oligonucleotides used as probes in the gel shift experiments were prepared by labeling dou- ble-stranded oligonucleotides (1 pmol) with 16 U of T4 polynucleotide kinase (Promega) and 40 /zCi of [3- 32p]ATP (Amersham, specific activity 3000 Ci/mmol) at 37°C for 30 min. Labeled oligonucleotides were purified on a 12% non-denaturing polyacrylamide gel, and 20 000- 40 000 cpm corresponding to 5 fmol) of the eluted radioac- tivity were used for each reaction. Binding reactions were obtained by incubating on ice 5 /zg of protein from cellular extracts for 15 rain in 18 /xl of a buffer containing 10 mM Tris-HC1 pH 7.5, 100 mM KC1, 2% Ficoll, 6% glycerol and 1 mM DT-I', 2 /xg poly(dI/dC) (Boehringer) and then for an additional 15 min at room temperature after the addition of 2 /zl labeled oligonucleotide. Where indicated, competitor oligonucleotides were added to the mixture during the first 15 min of incubation. The DNA-protein complexes were separated from the unbound DNA probe on a 4% non-denaturing polyacrylamide gel in 0.5 X TBE (44 mM Tris-HC1, 44 mM boric acid, 12.5 mM EDTA). Electrophoresis (20 mA/150 V) was carried out at room temperature. The sequences of double stranded oligo- nucleotides were: 5 ' -TCGACTGTACAGGATGTTC- TAGCTACT-3' for GRE; 5'-ATCGGGGCGGGGCGGG- GCGGGGCGGGGC-3' for Spl and 5'-AATTCAATTGG- GGCAATCAGG-3' for C /EBP (all supplied by GENSET).

310 F. Esposito et al. / Biochimica et Biophysica Acta 1260 (1995) 308-314

3. Results

3.1. DNA binding activity of GR and Spl in protein extracts of cells exposed to different oxidizing agents

To study the effect of the intracellular redox environ- ment on transcriptional factor efficiency we used DEM which acts on GSH levels by forming a GSH-DEM conju- gate in a reaction catalyzed by glutathione-S-transferase [19]. The decrease of intracellular GSH results in severe impairment of the molecular mechanisms leading to the inactivation of ROI, because some enzymes catalyzing this inactivation, mainly glutathione-peroxidase and glu- tathione-S-transferase, require GSH as a cofactor [20]. Because ROI, including free radicals, peroxides and elec- trophiles, are continuously produced in the cell, the addi- tion of DEM changes the intracellular redox condition into an oxidizing environment. DEM was selected as GSH-de- pleting agent because it is a weak electrophile and its direct oxidizing effects on other cellular substrates (mem- brane lipids) are negligible [21].

a b c d e f g h

g

To investigate how G R - D N A binding activity in vivo is affected by changing the redox conditions of the cells, we used the COS2 cell line in which the overexpression of GR was obtained by transfecting these cells with the pSTC 407-556 plasmid that carries the cDNA encoding the rat GR under the control of the cytomegalovirus promoter [17]. 42 h after the transfection of the pSTC plasmid, cell extracts were prepared and tested in gel shift experiments using as a probe the oligonucleotide GRE (see Materials and methods) containing the D N A consensus sequence for GR [22]. The oligonucleotide was shifted as a consequence of the specific binding of GR present in cellular extracts (Fig. 1; lanes b, c and d); however the treatment of the cells for 6 h with 1 mM DEM, before cell harvesting, almost completely abolished the shift (Fig. 1, lane e). This effect was already detectable at either 1 or 3 h after the exposure to DEM (Fig. 1, lanes j and k, respectively). Lower DEM concentrations (0.1 mM) had only a weak effect on D N A binding efficiency (data not shown). The cytotoxic effect of various concentrations of DEM was monitored by measuring cell viabil i ty at different times of treatment. 6 h after the exposure to 1 mM DEM (when the GR DNA binding efficiency is almost completely abol- ished) cell viabili ty was approx. 80% compared to un- treated cells (Fig. 2).

The decreased GR-DNA binding is not a consequence of a direct oxidizing effect of DEM on GR protein. In fact, the same decrease was observed when protein extracts from untreated COS2 cells were incubated for 1 h with 20

i i k I m n o p q

i / i ̧

Fig. 1, Effects of DEM treatment on DNA binding of GR and Spl. Lanes a and f: labeled GRE oligonucleotide without the addition of cellular extracts; lane m, labeled Spl oligonucleotide without the addition of cellular extracts; lanes b to l: gel retardation assays performed with protein extracts from COS2 cells harvested 42 h after transfection with the pSTC plasmid carrying the rat cDNA for GR (see Materials and methods); lane b: labeled GRE oligonucleotide incubated with cellular extracts from transfected COS2 cells; lanes c and d: as in lane b, the cellular extract was preincubated with a 100-fold molar excess of unla- beled GRE oligonucleotide (specific competitor) or of unlabelled Spl oligonucleotide (not specific competitor), respectively; lane e: labeled GRE oligonucleotide incubated with cellular extracts prepared from DEM-treated cells (36 h after transfection cell cultures were exposed to 1 mM DEM for 6 h and then harvested); lane g: labeled GRE oligonucleo- tide incubated with cellular extract from untreated cells, the extracts were incubated for 60 rain on ice with 1 mM DEM; lane h: as lane g, but the extracts were incubated with 20 mM H202 for 60 min on ice; lane i: labeled GRE oligonucleotide incubated with cellular extracts from COS2 cells transfected with the pSTC plasmid; lanes j, k and l, labeled GRE with cellular extracts prepared from COS2 cells treated with 1 mM DEM for 1, 3 and 6 h before harvesting, respectively; lanes n to q: protein extracts from HeLa cells; lane n, labeled Spl oligonucleotide incubated with HeLa cell extracts; lane q: labeled Spl oligonucleotide incubated with protein extracts from DEM-treated HeLa cells (6 h with 1 mM DEM); lanes o and p: as in lane n, the cellular extracts were preincubated with a 100-fold molar excess of unlabeled Spl oligonucleotide (specific competitor) or of unlabeled GRE oligonucleotide (non specific competi- tor) respectively. The arrow indicates the position of the specific DNA- protein complexes.

F. Esposito et al. / Biochimica et Biophysica Acta 1260 (1995) 308-314 311

1001 75

=

,~ so

-t M 25

3 6 9 2 4 no DEM 0.5 mr1 DEM Hours ! mM DE~I 1.5 mrl DEM

Fig. 2. Cell viability during DEM treatment of HeLa cells. HeLa cell cultures were treated with different concentrations of DEM. At the times indicated, total cell numbers were microscopically counted after a trypan blue treatment and the number of living cells (excluding those stained with the dye) was plotted as percentage. Each value is the average of the cell number contained in four plates for each DEM concentration.

mM (final concentration) H202 (Fig. 1; lane h), whereas no effect was observed when these extracts were treated for 1 h with 1 mM (final concentration) DEM (Fig. 1, lane g). Treatment with NAC is known to protect the cells from oxidizing radicals by increasing intracellular levels of GSH [23]. Incubation of COS2 cells with 30 mM NAC for 2 h before exposure to DEM almost completely prevented the decrease of GR DNA binding (Fig. 3). These results further reinforce the hypothesis that the decrease of GR DNA binding observed upon treatment of cells with DEM (see Fig. 1) is a consequence of GSH depletion, which in turn results in an increase in the level of ROI, rather than that of a direct effect of DEM.

Similarly to GR, Spl present in extracts from HeLa cells treated with 1 mM DEM for 6 h binds to its cognate cis-element with a significantly decreased efficiency (Fig. 1, lane q), although the phenomenon was less evident than with the GR binding in COS2 cells. To evaluate the specificity of the GR- and Spl-DNA binding decrease observed upon intracellular redox manipulations, the pro- tein extracts from HeLa and COS2 cells that were used in the experiments of Fig. 1 were tested using as a probe an

a b c d a b c d e

Fig. 3. Effects of NAC on DNA binding activity of GR during DEM treatment in COS2 cells. Lane a: labeled GRE oligonucleotide without the addition of cellular extracts; lanes b to d: gel retardation assays of GRE oligonucleotide challenged with protein extracts from COS2 cells har- vested 42 h after transfection with the pSTC plasmid (see Materials and methods); lane b: labeled GRE oligonucleotide incubated with cellular extracts from untreated COS2 cells; lane c: as in lane b but cellular extracts were prepared from DEM-treated cells (1 mM for 6 h before harvesting); lane d: as lane c, but cellular extracts were prepared from cells preincubated with 30 mM NAC, 2 h before DEM treatment.

Fig. 4. Effects of DEM treatment on DNA binding of C/EBP. Lane a: labeled C / E B P oligonucleotide without the addition of cellular extracts; lane b: labeled C / E B P oligonucleotide incubated with protein extracts from untreated COS2 cells (same extracts as described in Fig. 1.); lanes c and d: as lane b, but the cellular extracts were incubated with 100-fold molar excess of unlabeled C / E B P oligonucleotide (specific competitor) or unlabeled GRE oligonucleotide (not specific competitor), respectively; lane e: labeled C / E B P incubated with protein extracts from cells treated with 1 mM DEM, 6 h before harvesting the cells.

312 F. Esposito et al. / Biochimica et Biophysica Acta 1260 (1995) 308-314

oligonucleotide containing the consensus sequence of C / E B P which is a transcriptional factor not affected in vitro by changes in redox conditions [5]. Gel-shift experi- ments demonstrated that C /EBP-DNA binding is un- changed in the extracts from cells treated with DEM, compared to those from untreated cells (Fig. 4).

The DNA binding efficiency of GR overexpressed in COS2 cells was also evaluated using two other ways to change intracellular redox conditions: cell exposure to H20 2, which acts directly as an oxidizing agent, or to buthionine sulfoximine, a compound that induces GSH depletion by a mechanism different from that of DEM, i.e., by inhibiting 7-glutamylcysteine synthetase that catalyzes the biosynthesis of a GSH precursor [24]. Fig. 5 shows the gel retardation assays of GRE oligonucleotide challenged with extracts from COS2 cells exposed to 200 p~M H20 2 for 6 h or to 5 mM buthionine sulfoximine for 24 h: in both cases a significant decrease of GR-DNA binding was observed. Using the same extracts from cells treated with H20 2 or with buthionine sulfoximine, C /EBP-DNA bind- ing was identical to that observed in untreated cells (data not shown).

a b c d e f

Fig. 5. Effects of H202 and buthionine sulphoximine on DNA binding of GR. Lane a: labeled GRE oligonucleotide without the addition of cellular extracts; lane b: labeled GRE oligonucleotide incubated with protein extracts from untreated COS2 cells; lane c: labeled GRE oligonucleotide incubated with protein extracts from COS2 cells exposed to 200 /xM H202 for 6 h; lane d: labeled GRE oligonucleotide incubated with protein extracts from COS2 cells exposed to 5 mM buthionine sulphox- imine for 24 h; lanes e and f: as in lane b, but the cellular extracts were incubated with a 100-fold molar excess of unlabeled GRE oligonucleotide (specific competitor) or unlabeled C/EBP oligonucleotide (not specific competitor), respectively.

3.2. The intracellular redox conditions influence the trans-

activating efficiency o f GR

The in vivo changes of the redox conditions should also affect the transcription of the genes controlled by the GR naturally occurring in the cells, in analogy with the de- creased DNA binding efficiency observed for GR overex- pressed in COS2 cells. To verify this, HeLa cells were transfected with a plasmid carrying the CAT reporter gene under the control of the promoter present in the LTR of the MMTV, which contains a GRE. Upon exposure to dexa- methasone, the endogenous GR present in HeLa cells is activated and interacts with GRE, inducing the expression of the CAT gene. As shown in Fig. 6A, MMTV promoter induction, observed after the dexamethasone treatment, is reduced by treating the cells 6 h before harvesting with 200 /xM H20 2 or with 1 mM DEM. As in the case of the DNA binding experiments, the treatment of HeLa cells with 30 mM NAC, 2 h before exposure to DEM, prevents the decrease of CAT gene transcription observed with DEM alone (see Fig. 6B). In the absence of dexametha- sone induction, no differences in the transcriptional effi- ciency of the CAT gene were observed after the oxidizing and the reducing treatments (see Fig. 6C). Finally, to exclude the possibility that the molecules used (DEM, NAC and H20 2) affected the Cat assay per se, we mea- sured Cat activity in the presence of 200/xM H202, 1 mM DEM or 30 mM NAC, and obtained the same activity as the control (data not shown).

4. D i s c u s s i o n

This study shows that the exposure of intact cells to oxidizing treatments induces impairment of GR-DNA binding efficiency and in turn of GR-regulated transcrip- tion. This confirms the hypothesis, until now based only on in vitro evidence (see Introduction), that GR activity is sensitive to intracellular redox conditions. Very little is known about the occurrence and relevance of the redox regulation of the transcriptional factors in vivo. Indirect evidence suggests that the in vitro sensitivity of at least some transcriptional factors to redox changes are related to in vivo phenomena. In fact, mutation of both Cys-154 in c-fos and Cys-272 in c-jun, which are the residues in- volved in the in vitro redox sensitivity, increases the DNA binding efficiency of AP-1 and enhances the transforming activity of these protooncogenes [4]. Furthermore, the Cys

Ser mutation naturally occurring in v-jun is probably involved in the transforming activity of the oncogene [25]. These data are in agreement with the direct observation that in vivo antioxidant treatments increase the DNA-bind- ing and the transacting efficiency of AP-1 [26].

Whereas the in vitro and in vivo results obtained with AP-1 are in good agreement, those obtained with NFkB are in a striking contrast. In fact, while NFkB-DNA bind-

F. Esposito et al. /Biochimica et Biophysica Acta 1260 (1995) 308-314 313

A 120 B 120 C 60

:.; 90 ";

60 ~ 30 6o

'~ 30 ~ 30 o

DEX ~ + .~ DEX ~ + ~ ¢. H20 2 H202 ~ NAC ~ ~ NAC ~ DEM + DEM # ~ DEM + +

a b c d a b c d e a b c d e

i e

• • • • •

Fig. 6. Effect of H202, DEM and NAC on GR transactivation of the MMTV promoter. HeLa cells were transfected with the MMTV-CAT plasmid and were harvested 42 h after transfection. In the upper part of each panel the mean value of the different CAT experiments is reported, and in the lower part is a representative example of CAT assay. (A) Lane a: untreated cells; lanes b to d: cells treated with 1 p.M dexamethasone; lane c: cells treated with 200 /zM H202; lane d: cells treated with 1 mM DEM. (B) Lane a: untreated cells; lanes b to e: cells treated with 1 /zM dexamethasone; lanes c and d: cells treated with 30 mM NAC; lanes d and e: cells treated with 1 mM DEM. (C) Lane a: untreated cells; lane b: cells treated with 200/zM H202 for 6 h before harvesting the cells; lane c and e: cells treated with l mM DEM; lane d and e: cells treated with 30 mM NAC.

ing is almost completely abolished in vitro by oxidant treatments and a complete recovery of this binding is obtained by reducing agents [27], the exposure of intact cells to oxidant or antioxidant treatments results in a completely different effect. Indeed, in vivo, DNA binding and transactivation by NFkB are enhanced by H202 whereas antioxidants have no effect [26]. The demonstra- tion that NFkB nuclear translocation is activated by oxi- dant treatments and inhibited by antioxidant agents [28] suggests a possible explanation for in vitro and in vivo conflicting results on NFkB redox regulation. Also in the case of GR the regulatory machinery is essentially based on the nuclear translocation of the activated receptor [29]. However, our experimental approach does not allow us to ascertain whether the DNA binding alone or whether also the nuclear translocation mechanism of GR is affected by redox changes.

Further support for a redox regulation of transcriptional factors comes from the complex machinery, based on the Ref-1 protein, that controls the redox state of AP-1 and of other transcription factors [25]. This protein is itself under redox control and its reducing potential is regenerated by the thioredoxin system [25]. GR is not sensitive to Ref-1 [25], hence it is feasible that another mechanism controls the GR redox state. The inactive GR present in the cyto- plasm is a 300 kDa complex in which GR is associated with two molecules of hsp90 [30], a non steroid-binding protein of 59 kDa [30], a phospholipid [31], and in many

cases with another heat shock protein of 70 kDa [32]. The binding of the hormone results in the disassembling of this multimeric structure, so that GR, in its activated (trans- formed) form, can interact with DNA. A possible relation- ship between this multimeric complex and the redox regu- lation of GR is suggested by the finding that thioredoxin is often associated with the GR complex [33].

Because there are many Zn-binding proteins requiring Cys residues in a reduced form to allow the formation of metal-protein complexes, it will be of interest to evaluate whether the redox-dependent phenomena described in this article affect all the Zn-finger transcriptional factors and if it might involve also other Zn-binding proteins. In this context, we studied another Zn-finger transcription factor, Egr-1, and demonstrated that it is sensitive to redox changes in vivo [34]. In fact, the ROI accumulation induced by DEM treatment in HL-60 cells decreases the Egr-I-DNA binding and prevents the appearance of the differentiated phenotype induced by TPA in this and in another human myeloid cell line [34].

It is well demonstrated that ROI accumulation is associ- ated with aging [35], and several phenomena observed in senescent animals could be explained by changes in pro- tein structure caused by the oxidizing environment. Tran- scription factor-DNA binding is indeed affected in aged tissues, as in the case of Spl [36], which suggests that the alteration of gene expression occurring during aging could be also due to an alteration of the transcriptional machin-

314 F. Esposito et al. / Biochimica et Biophysica Acta 1260 (1995) 308-314

ery. GR could be involved in this age-related alteration; in fact, impaired glucocorticoid responsiveness during aging [37] is not always paralleled by a reduction in receptor number. In the light of our results this decreased respon- siveness to glucocorticoid could be explained as a conse- quence of the sensitivity of GR to intracellular redox conditions.

Acknowledgements

The authors thank Dr S. Rusconi for providing the plasmid pSTC407-506 and Dr. H.M. Bond for critical reviewing of the manuscript. This work has been supported by grants from CNR Special Projects 'Invecchiamento' No. 941490 and 'Ingegneria genetica', from AIRC and from MURST 40% and 60%.

References

[1] Pognonec, P., Kato, H. and Roeder, R.G. (1992) J. Biol. Chem. 267, 24563-24567.

[2] Hainaut, P. and Milner, J. (1993) Cancer Res. 53, 4469-4473. [3] Hupp, T.R., Meek, D.W., Midgley, C.A. and Lane, D.P. (1993)

Nucleic Acids Res. 21, 3167-3174. [4] Abate, C., Patel, L., Rauscher III, F.J. and Curran, T. (1990) Science

249, 1157-1161. [5] Bannister, A.J., Cook, A. and Kouzarides, T. (1991) Oncogene 6,

1243-1250. [6] McBride, A.A., Klausner, R.D. and Howley, P.M. (1992) Proc. Natl.

Acad. Sci. USA 89, 7531-7535. [7] Galang, C.K. and Hauser, C.A. (1993) Mol. Cell. Biol. 13, 4609-

4717. [8] Freedman, L.P., Luisi, B.F., Korszun, Z.R., Basavappa, R., Sigler,

P.B. and Yamamoto, K.R. (1988) Nature 334, 543-546. [9] Hard, T., Kellenbach, E., Boelens, R., Maler, B.A., Dahlman, K.,

Freedman, L.P., Carlstedt-Duke, J., Getafsson, J.A. and Kaptein, R. (1990) Science 249, 157-160.

[10] Bodwell, J.E., Holbrook, N.J. and Munch, A. (1984) Biochemistry 23, 4237-4242.

[11] Tienrungroj, W., Meshinchi, S., Sanchez, E.R., Pratt, S.E., Grippo, J.F., Holmgren, A. and Pratt, W.B. (1987) J. Biol. Chem. 262, 6992-7000.

[12] Kadonaga, J.T., Carner, K.R., Masiarz, F.R. and Tjian, R. (1987) Cell 51, 1079-1090.

[13] Gorman, C.M., Moffat, L.F. and Howard, B.H. (1982) Mol. Cell. Biol. 2, 1044-1051.

[14] Cato, A.C.B., Miksicek, R., Schutz, G., Arnermann, J. and Beato, M. (1986) EMBO J. 5, 2237-2240.

[15] Graham, F.L. and Van der Eb, A.J. (1973) Virology 52, 456-462. [16] Bradford, M. (1976) Anal. Biochem. 72, 248-254. [17] Severne, Y., Wieland, S., Schaffner, W. and Rusconi, S. (1988)

EMBO J. 7, 2503-2508. [18] Kumar, V., and Chambon, P. (1988) Cell 55, 145-156. [19] Boyland, E. and Chasseaud, L.F. (1970) Biochem. Pharmacol. 19,

1526-1528. [20] Meister, A. and Anderson, M.E. (1983) Annu. Rev. Biochem. 52,

711-760. [21] Wendel, A., Feuerstein, S. and Konz, K.H. (1978) in Functions of

Glutathione in Liver and Kidney (Sies, H. and Wendel, A., eds.), pp. 183, Springer-Verlag, Berlin.

[22] Yamamoto, K.R. (1985) Annu. Rev. Genet. 19, 209-252. [23] Staal, F.J.T., Roederer, M. and Herzenberg, L.A. (1990) Proc. Natl.

Acad. Sci. USA 87, 9943-9947. [24] Meister, A. (1991) Pharmac. Ther. 51, 155-194. [25] Xanthoudakis, S., Miao, G., Wang, F., Pan, Y.E. and Curran, T.

(1992) EMBO J. 11, 3323-3335. [26] Meyer, M., Schreck, R. and Bauerle, P.A. (1993) EMBO J. 12,

2005-2015. [27] Toledano, M.B. and Leonard, W.J. (1992) Proc. Natl. Acad. Sci.

USA 88, 4328-4332. [28] Menon, S.D., Qin, S., Guy, G.R. and Tan, Y.H. (1993) J. Biol.

Chem. 268, 26805-26812. [29] Vedeckis, W.V., (1983) Biochemistry 22, 1983-1989. [30] Rexin, M., Bush, W. and Gehring, U. (1991) J. Biol. Chem. 266,

24601-24605. [31] Bodine, P.V. and Litwack, G. (1990) J. Biol. Chem. 265, 9544-9554. [32] Diehl, E.E. and Schmidt, T.J. (1993) Biochemistry 32, 13510-13515. [33] Grippo, J.F., Tienrungroj, W., Dahmer, M.K. et al. (1983) J. Biol.

Chem. 258, 13658-13664. [34] Esposito, F., Agosti, V., Morrone, G., Morra, F., Cuomo, C., Russo,

T., Venuta, S. and Cimino, F. (1994) Biochem. J. 301, 649-653. [35] Stadtman, E.R. (1992) Science 257, 1220-1224. [36] Ammendola, R., Mesuraca, M., Russo, T. and Cimino, F. (1992) J.

Biol. Chem. 267, 17944-17948. [37] Roth, G.S. and Livingstone, J.N. (1979) Endocrinology 104, 423-

428.