Disruption of microtubules leads to glucocorticoid receptor degradation in HeLa cell line

Cellular Signalling 17 (2005) 187 – 196 www.elsevier.com/locate/cellsig

Disruption of microtubules leads to glucocorticoid receptor degradation in HeLa cell line Zdeneˇk Dvorˇa´ka,b,*, Martin Modriansky´a, Jitka Ulrichova´a, Patrick Maurelb, Marie-Jose Vilaremb, Jean-Marc Pascussib a

Institute of Medical Chemistry and Biochemistry, Medical Faculty, Palacky´ University Olomouc, Hneˇvotı´nska´ 3, 77515 OLOMOUC, Czech Republic b CNRS-INSERM-U128, 1919 Route de Mende, 34293 MONTPELLIER, France Received 7 May 2004; accepted 30 June 2004 Available online 26 August 2004

Abstract The role of microtubules (MTs) in steroid hormone-dependent human glucocorticoid receptor (hGR) activation/translocation is controversial. It was demonstrated recently that colchicine (COL) down-regulates hGR-driven genes in primary human hepatocytes by a mechanism involving inhibition of hGR translocation to the nucleus. To investigate whether inhibition of hGR translocation is the sole reason for its inactivation, we used human cervical carcinoma cells (HeLa) as a model. Herein we present evidence that perturbation of microtubules in HeLa cells leads to rapid time- and dose-dependent degradation of hGR protein. Degradation is proteasome mediated as revealed by its reversibility by proteasome inhibitor MG132. Moreover, degradation was observed for neither wt-hGR nor hGR mutants S226A and K419A in transiently transfected COS-1 cells. On the other hand, c-jun N-terminal kinase (JNK) seems not to be involved in the process because JNK inhibitor 1,9-Pyrazoloanthrone (SP600125) does not reverse hGR degradation. Similarly, another hGR functional antagonist, nuclear factor kappa beta (NFnB), did not play any role in the degradation process. D 2004 Elsevier Inc. All rights reserved. Keywords: Microtubules; Glucocorticoid receptor; Proteasome; Degradation; Translocation; c-jun N-terminal kinase; HeLa cells

1. Introduction

Abbreviations: AP-1, activator protein 1; BSA, bovine serum albumin; CAR, constitutive androstane receptor; COL, colchicine; CYPs, cytochromes P450; DEX, dexamethasone; DTT, dithiothreitol; GCs, glucocorticoids; GFP-GR, glucocorticoid receptor–green fluorescent protein chimera; GRE, glucocorticoid responsive element; hsp70 and hsp90, heat shock proteins 70 and 90 kD; h(m,r)GR, human (mouse, rat) glucocorticoid receptor; IL-2 and IL-6, interleukin 2 and 6; JNK, c-jun N-terminal kinase; MDCs, microtubules-disrupting compounds; MT, microtubules; NFnB, nuclear factor kappa beta; NOC, nocodazole; PMSF, phenylmethanesulfonyl fluoride; PXR, pregnane X receptor; SP600125, 1,9-Pyrazoloanthrone; TNFa, tumor necrosis factor alpha; VIN, vincristine. * Corresponding author. Institute of Medical Chemistry and Biochemistry, Medical Faculty, Palacky´ University Olomouc, Hneˇvotı´nska´ 3, 77515 OLOMOUC, Czech Republic. Tel.: +420 58 5632324; fax: +420 58 5632302. E-mail address: [email protected] (Z. Dvorˇa´k). 0898-6568/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.cellsig.2004.06.010

Human glucocorticoid receptor (hGR) plays an important role in variety of cellular processes such as development, differentiation and homeostasis. It exerts diverse effects in transcriptional control of wide spectrum of genes taking part in their up- and down-regulation [1]. One of the most prominent hGR functions is its unique entanglement in the inflammation process control. During inflammation, several transcriptional factors, such as nuclear factor kappa beta (NFnB) or activator protein 1 (AP-1), are activated in response to the release of cytokines by macrophages, e.g., interleukin 6 (IL-6) and tumor necrosis factor alpha (TNFa). These factors serve as a toggle switch for whole battery of genes with hGR having the opposite effect [2]. Several other mechanisms of contest between hGR and proinflammatory factors were proposed or shown [3]. hGR is an indispensable and crucial factor in hepatic drug metabolism: it controls

188

Z. Dvorˇa´k et al. / Cellular Signalling 17 (2005) 187–196

expression of orphan receptors, pregnane X receptor (PXR) and constitutive androstane receptor (CAR), and consequently, the expression of phase I [4], phase II [5], and phase III enzymes [6]. Any alteration in hGR activity/expression may have potentially deleterious impact on the detoxification capacity of the organism. Indeed, it was demonstrated that the expression of drug-metabolizing enzymes is reduced due to glucocorticoid receptor (GR) activity suppression by cytokines [7]. Thorough investigation of interplay between hGR and various pathophysiological stimuli is therefore beneficial. In the absence of a ligand, hGR is located predominantly in cytosol in a complex with chaperone proteins heat shock proteins 90 and 70 kD (hsp90, hsp70) and p60 [8]. Upon ligand binding, hGR dissociates from the complex, translocates into the nucleus, forms a homodimer, binds its DNA consensus sequence, and triggers transcription. Regardless of glucocorticoid (GC) presence, there exists a dynamic equilibrium between the cytosolic and nuclear hGR. However, the process is not fully understood yet. Although considered to be a constitutively expressed gene, there are reports on the autoregulation of hGR mRNA expression by glucocorticoids [9]. One mode of regulation of hGR activity is provided by glucocorticoids and proinflammatory elements, the other is the reduction of hGR protein half-life by glucocorticoids. In the presence of glucocorticoids, it drops down to 30–50% of its initial value [10]. It was reported recently that synthetic glucocorticoid dexamethasone (DEX) causes mouse GR protein degradation mediated by proteasome–ubiquitin pathway [11]. Hormones, drugs, and cytokines aside, there is little evidence on the role of cytoskeleton and/or microtubular network in hGR function and regulation. Data published on the importance of microtubules (MTs) in GR functionality are a bit controversial; some authors maintain the necessity of intact microtubules for proper GR function in connection with nuclear translocation [12,13]; others claim the opposite [14]. It is noteworthy that in majority of these studies, over-expression systems were employed which do not exactly correspond to the physiological state. Adding to the dilemma, we have recently demonstrated that microtubules disruption in primary human hepatocytes by colchicine (COL) perturbs hGR dependent PXR/CAR-cytochromes P450 (CYPs) cascade by a mechanism involving inhibition of hGR translocation to the nucleus [15]. We have also observed inhibition of hGR-GFP chimera translocation by colchicine in transiently transfected HEK293 cells. On the other hand, there was no effect in transiently transfected [hGR/glucocorticoid responsive element (GRE)–LUC] human hepatoma cell line HepG2, while in human cervical carcinoma cell line, human cervical carcinoma cells (HeLa) stably transfected with GRE–LUC construction, we observed inhibition of hGR activity by colchicine and nocodazole (NOC). In parallel, we have shown that colchicine has no effect on hGR mRNA expression and in 3H-dexamethasone ligand binding assay. Because activation of c-jun N-terminal kinases (JNK) by agents disrupting microtubules was described [16], we

anticipated that JNK may participate in hGR activity regulation by microtubules-disrupting compounds (MDCs). Moreover, it was demonstrated that activation of JNK leads to the inactivation of rat GR [17]. HeLa cells, which possess intrinsic hGR, were the model of choice in this work primarily to have hGR levels close to physiological values. The aim of our work was to study the effect of three structurally diverse microtubules disrupting compounds, i.e., colchicine (COL), nocodazole (NOC) and vincristine (VIN), on the hGR activity in terms of: (i) nuclear translocation of hGR; (ii) involvement of the proinflammatory nuclear factor kappa-beta (NFkB); and (iii) role of JNK activation. Surprisingly, in response to microtubules disrupting compounds treatment, we have observed rapid hGR degradation, which is linked to none of the above mechanisms but is proteasome–ubiquitin dependent.

2. Materials and methods Cell culture media and supplements, dexamethasone (DEX), colchicine (COL), nocodazole (NOC), vincristine (VIN), dithiothreitol (DTT), MG132 (Z–Leu–Leu–Leu–al), SP600125 (1,9-Pyrazoloanthrone), N-ethylmaleinimide, Hoechst 33258, horseradish peroxidase-conjugated secondary antibodies, and foetal calf serum were purchased from Sigma (St. Louis, MO, USA). An ECL kit including hyperfilm photographic paper were purchased from Amersham (Little Chalfont, England). Complete Mini protease inhibitor and FuGENE 6 transfection reagent were purchased from Roche Molecular Biochemicals (Mannheim, Germany). Protein A Sepharose 4 Fast Flow was from Amersham Pharmacia (Uppsala, Sweden). All other chemicals and reagents were of the highest quality commercially available. 2.1. Cell lines and transfections 2.1.1. Plasmids The human glucocorticoid receptor expression vector (pSG5–hGR) was kindly provided by Dr J.C. Nicolas (INSERM U439, Montpellier, France). The pSG5 vector was from Stratagene (Amsterdam, The Netherlands). Mutation of human glucocorticoid receptor cDNA from Lys419 to K419A or Ser226 to S226A were done by polymerase chain reaction site-directed mutagenesis using the QuickChange Site-Directed mutagenesis kit from Stratagene. Primers used were : hGRL419A: 5V–ccacaggaccacctcccacactctgcctggtgtgc and 5V–gcacaccaggcagagtgcgggaggtggtcctgttg; hGRS226A: 5V–ctgtttgcttgctcctctgg and 5V–ccagaggagcaagcaaacag. 2.1.2. HeLa Human cervical carcinoma cells (HeLa) were seeded on Petri dishes (100 mm I.D.) in density 2106 cells/dish using

Z. Dvorˇa´k et al. / Cellular Signalling 17 (2005) 187–196

Dulbecco’s modified eagle medium (DMEM) culture media supplemented with glucose (4.5 g/l), pyruvate (1 mM final), nonessential amino acids (100 Gibco), penicillin (100 U/ ml final), streptomycin (100 Ag/ml final), glutamine (2 mM final), and foetal calf serum (10 vol.%/vol.%). Cells were allowed to grow up to 70–80% of confluence. Following exchange for a serum-free medium, the culture was stabilized for additional 6–8 h prior to the treatments. 2.1.3. COS-1 COS-1 cells were transiently transfected by lipofection (FuGENE 6) with wild type hGR (wt) or hGR mutants mhGR(S226A) and m-hGR(K419A). Then, cells were seeded in the same fashion as for HeLa cells and allowed to stabilize for 24 h. Medium was then exchanged for a serumfree one 4–6 h prior to treatment. Cells were treated with colchicine (1 AM) or DMSO as vehicle for 45 min. 2.2. Microtubules disruption assessment HeLa cells were plated on 12-well culture dishes with immersed glass cover-slips in a density 5105 cells/well. Upon reaching 70–80% of confluence, the culture medium was exchanged for a serum-free one and the culture was allowed to further stabilize for 6–8 h. Cells were treated with colchicine (1 AM), nocodazole (1 AM), vincristine (1 AM) or DMSO as vehicle for 30 min, fixed (60 mM PIPES, 25 mM HEPES, 10 mM EDTA, 10 mM MgCl2, 3 vol.%/ vol.% formaldehyde, 0.05 vol.%/vol.% glutaraldehyde, pH 6.9), permeabilized (150 mM NaCl, 50 mM Tris–HCl, 0.2 vol.%/vol.% Triton-X100, pH 7.5), and their nonspecific sites were saturated with bovine serum albumin (BSA; 3 vol.%/vol.% in PBS). As the first antibody, mouse monoclonal anti-a tubulin (CLON DM1A; T9026; Sigma) was used in a dilution of 1:1000 supplemented with 1% BSA for 1 h; as the second antibody served Alexa FluorR488 conjugated goat anti-mouse IgG (Molecular Probes, Eugene, Oregon, USA), in a dilution of 1:1000 again incubated for 1 h in the presence of 1% BSA. Staining of nuclei with Hoechst 33258 was performed in parallel (1 Ag/ ml; 5 min). Microtubules status was then evaluated using fluorescent microscopy with a Leica DMRA microscope equipped with a 40 or 100 objective at the dIntegrated Image FacilityT (CRBM) directed by Dr P. Travo. 2.3. Preparation of nuclear and cytosolic extracts Nuclear and cytosolic extracts were prepared as described elsewhere [18], with minor modifications. Briefly, following the treatments HeLa cells were washed twice with 2 ml of icecold PBS and scraped into 1 ml of PBS. Suspension was centrifuged (1500g/5 min/4 8C) and the pellet was resuspended by gentle pipetting in 300 Al of ice-cold buffer A (10 mM HEPES pH 7.9; 10 mM KCl; 1.5 mM MgCl2; 0.5 mM DTT; 0.1 vol.%/vol.% NP-40). Mixture was incubated for 10 min on ice and then centrifuged (12 000g/10 min/4

189

8C). Following collection of supernatant, which contains cytosolic fraction, the pellet was vigorously resuspended by syringe/needle in 3 vol. of ice-cold buffer B (20 mM HEPES, pH 7.9; 420 mM NaCl; 0.2 mM EDTA; 1.5 mM MgCl2; 0.5 mM DTT; 0.5 mM phenylmethanesulfonyl fluoride (PMSF); 25 vol.%/vol.% glycerol) and incubated for 30 min on ice. Following centrifugation (12 000g/20 min/4 8C), the supernatant (nuclear extract) was collected. Both extracts were stored in 80 8C. Protein content in extracts was determined by biscinchoninic acid method [19]. 2.4. hGR immune-precipitation Cytosolic extracts from HeLa cells obtained as described above were prepared in I.P. buffer (10 mM Tris pH 7.4, 150 mM NaCl, 2 mM EDTA, 0.5% NP-40, 0.5% sodium deoxycholate, 5 mM N-ethylmaleinimide) containing Completek Mini protease inhibitor (1 tablet/10 ml). GR(E-20)X rabbit polyclonal antibody (Santa Cruz Biotechnology, USA) was added to 200 Ag of cytosolic extracts and rotated overnight at 4 8C to immune-precipitate hGR. Protein A Sepharose 4 Fast Flow was added and mixture rotated for additional 2 h. Resulting suspension was centrifuged and the pellet washed five times with 0.5 ml of i.p. buffer. Final pellets were dissolved in 50 Al of Laemmli loading buffer, boiled for 5 min, and analyzed by Western blot. 2.5. Western blots Cytosolic and nuclear extracts, and immunoprecipitated samples were all analyzed using the following conditions: SDS-PAGE gels (7.5%) were run on a Hoefer apparatus according to the general procedure [20]. Protein transfer onto nitrocellulose membrane was carried out as described [21]. The membrane was stained with amidoblack dye for control of transfer and then saturated with 8% nonfat dried milk overnight. Blots were probed with primary antibodies against: human glucocorticoid receptor (GR(E-20)X rabbit polyclonal; dilution 1/500); p50 (NFnB p50(H-119)X rabbit polyclonal; dilution 1/300); p65 (NFnB p65(A) rabbit polyclonal; dilution 1/300); ubiquitin (Ub(P4D1)AC mouse monoclonal; dilution 1/1500), and h-actin (Actin (I19) goat polyclonal; dilution 1/500); all purchased from Santa Cruz Biotechnology (Santa Cruz, USA). Chemiluminescence detection using horseradish peroxidase-conjugated secondary antibodies and an Amersham ECL kit was performed.

3. Results 3.1. Disruption of microtubules network in HeLa cells by microtubules-disrupting compounds Colchicine (COL), nocodazole (NOC), and vincristine (VIN) are well known to depolymerize microtubules in all

190

Z. Dvorˇa´k et al. / Cellular Signalling 17 (2005) 187–196

types of cells. To verify microtubules disruption in HeLa cells, we performed immunofluorescence evaluation of microtubules network status following treatment of the cells by the tested compounds. This demonstration is crucial because of relatively short time course of events under investigation. Fluorescent micrographs in Fig. 1 show that microtubules network of HeLa cells treated with three structurally diverse microtubules disrupting compounds, i.e., tropolone alkaloid colchicine, synthetic substance nocodazole, and vinca alkaloid vincristine (all 1 AM final), is clearly altered after 30 min of treatment. 3.2. Dose- and time-dependent effect of microtubules disrupting compounds on hGR degradation in HeLa cells In the initial series of experiments, we evaluated the effect of increasing dose of microtubules disrupting compounds on hGR protein levels in cultured HeLa cells. Colchicine, vincristine, and nocodazole all in concentrations 0, 0.01, 0.1, and 1 AM were administered to the cells for 25 min. All three compounds caused concentration-dependent hGR protein degradation in both cytosolic and nuclear fractions (Fig. 2). The potency of microtubules-disrupting

compounds used was comparable; hGR disappearance is obvious at 1 AM concentrations. Note how rapidly degradation occurs with about 50% of the protein degraded during the incubation period. Thus, for further experiments, the concentration of 1 AM for microtubules-disrupting compounds tested was used. Because microtubules disrupting compounds display dose-dependent effect on hGR protein content in HeLa cells, we next investigated the kinetic profile of this event within the time range of 0–60 min. Degradation of hGR by the tested compounds was time-dependent (Fig. 3), resembling first-order kinetics. It is faster in the cytosol than in the nucleus, a fact attributable to the predominant cytosolic localization of hGR in the absence of a ligand. hGR content apparently high in the nucleus in the absence of dexamethasone is linked to the different GR/total protein ratio in the nuclear versus cytosolic extracts. Residual pellets remaining after cytosolic and nuclear extract preparation were probed for GR as well and did not display any GR present (data not shown). Again, degradation potency of microtubules-disrupting compounds used was comparable, exerting half-time of degradation around 20 and 25 min for cytosolic and nuclear fraction of GR, respectively. Finally, a significant

Fig. 1. Disruption of microtubules by MDCs in HeLa cells. HeLa cells were treated with colchicine (COL; 1 AM), nocodazole (NOC; 1 AM), vincristine (VIN; 1 AM) or DMSO as vehicle for 30 min. Microtubular network status was then evaluated using immunohistochemical detection and fluorescent microscopy imaging.

Z. Dvorˇa´k et al. / Cellular Signalling 17 (2005) 187–196

191

Fig. 2. Concentration effect of microtubules disrupting compounds on hGR degradation in HeLa cells. HeLa cells were treated 25 min with colchicine (COL), nocodazole (NOC) and/or vincristine (VIN) in concentrations 0.01, 0.1 and 1 AM, and/or with vehicle (DMSO) for control (concentration 0 AM). Cytosolic and nuclear extracts were isolated, and after western blot analysis, the membrane was probed with anti-hGR and anti-actin antibodies. Similar profile was obtained in two independent experiments.

decrease in hGR protein content was observed even after 5 min of treatment. 3.3. Hormone-dependent hGR translocation in HeLa cells co-treated with microtubules-disrupting compounds Located predominantly in the cytosol in the absence of a ligand, hGR is activated and is translocated into the nucleus upon dexamethasone addition to the cell culture. Following dexamethasone treatment, the maximum amount of GR present in the nucleus, accompanied by a drop in cytosolic GR, is reached at about 16 min, followed by a decrease due

to subsequent nuclear regulatory processes (Fig. 4A, B). In HeLa cells co-treated with dexamethasone and microtubules-disrupting compounds, effects on both translocation and degradation processes are combined resulting in: (i) accelerated cytosolic GR depletion; and (ii) an apparent shift of hGR peak in the nucleus to shorter times (7 min vs. 16 min) (Fig. 4A, B). Similar behavior was observed for all tested compounds; the apparent time shift of GR peak time being always significant (Fig. 4C). It must be stressed at this point that the degradation process proceeds faster in the cytosol than in the nucleus (Fig. 4B). 3.4. Effect of proteasome inhibitor MG132 on microtubulesdisrupting compounds-mediated hGR degradation

Fig. 3. Time course of hGR degradation by microtubules disrupting compounds in HeLa cells. HeLa cells were treated with 1 AM colchicine (COL), nocodazole (NOC) and/or vincristine (VIN) for 5; 15; 25; 40 and 60 min, and with DMSO (60 min) as vehicle for control (time, 0 min in blot). Cytosolic and nuclear extracts were isolated, and after western blot analysis, the membrane was probed with anti-hGR and anti-actin antibodies. The bar plot compares the times needed for 50% of hGR protein degradation by MDCs in 1 AM concentration. The data are expressed as meanFS.D. from two independent experiments.

Proteasome–ubiquitin enzyme complex participates in the homeostasis maintenance of many intracellular proteins holding office of an important system involved in the degradation of these proteins. In response to various stimuli, e.g., aging, oxidative damage, misfolding, etc., the approximately 7-kD protein ubiquitin is bound to lysine residues within the target protein in a reaction catalyzed by proteasome constituent E2 and E3 ligases. Thus modified, i.e., bubiquitinatedQ, proteins are recognized by the proteasome complex and consequently degraded by proteolysis. As was recently demonstrated, glucocorticoid-activated mouse GR is a subject of proteasome–ubiquitin degradation pathway [11]. Extrapolating the role of proteasome in human GR regulation as well because of identical sequence in the PEST regulatory region, PEST standing for amino acid sequence of Pro (P), Glu (E), Ser (S), and Thr (T), we tested whether a known proteasome inhibitor MG132 can affect hGR degradation due to microtubules network collapse. Indeed, in HeLa cells pretreated with 1 AM MG132 for 1 h, hGR degradation provoked by all tested microtubules-disrupting compounds was entirely reverted in both the cytosol and the nucleus (Fig. 5, data shown for colchicine only). MG132 pretreatment had no effect on hGR function in terms of hormone-dependent nuclear trans-

192

Z. Dvorˇa´k et al. / Cellular Signalling 17 (2005) 187–196

Fig. 4. Translocation of hGR by dexamethasone in HeLa cells co-treated with microtubules disrupting compounds. (A) HeLa cells were treated with dexamethasone (DEX; 100 nM) itself and/or co-treated with DEX (100 nM) and colchicine (COL; 1 AM), nocodazole (NOC; 1 AM), and/or vincristine (VIN; 1 AM) for the period of time 5, 15, 25, 40 and 60 min. DMSO was used as vehicle for control (time, 0 min). Cytosolic and nuclear extracts were isolated, and after WB analysis, the membrane was probed with anti-hGR and anti-actin antibodies. (B) The plot compares time course of hGR in cytosol and nucleus in presence of 100 nM DEX (D; E), 1 AM COL (5; n) and/or COL+DEX (o; .). Data are means from two independent experiments. S.D. values are not shown for better orientation. Hollow symbols stand for cytosol and full ones for nuclear extract. (C) The bar plot shows the times when hGR reaches maximum in nucleus. The data are expressed as meanFS.D. from two independent experiments. *Value is significantly different from DEX treatment at pb0.05.

location because the disappearance of the receptor from the cytosol and concomitant massive increase in the nucleus (Fig. 5). In some experiments, we observed higher hGR content in control cells treated with MG132 than in untreated cells, hinting at a very dynamic hGR processing. To confirm the hypothesis that hGR is ubiquitinated after microtubules perturbation and consequently degraded by proteasome, we immune-precipitated (antibody against hGR) cytosolic extracts from control cells and cells treated with colchicine in the presence and

absence of MG132. In both cases, the original cytosolic extracts (Fig. 5) and the immune-precipitates from cytosols (Fig. 6) gave identical profile when probing the membrane with anti-hGR antibody. If using anti-ubiquitin antibody in whole cytosols, the presence of MG132 causes an increased level of all ubiquitinated proteins, while colchicine has no effect. The crucial observation is the profile obtained after probing immune-precipitates from cytosols with anti-ubiquitin antibody: in MG132 pretreated cells, the level of Ubi–hGR form is elevated in

Z. Dvorˇa´k et al. / Cellular Signalling 17 (2005) 187–196

Fig. 5. Effect of proteasome inhibitor MG132 on hGR degradation by microtubules disrupting compounds in HeLa cells. HeLa cells were preincubated with proteasome inhibitor MG132 (1 AM) and/or DMSO as vehicle for 1 h prior to the treatment with dexamethasone (DEX; 100 nM) and/or colchicine (COL; 1 AM). Treatment duration was 25 min and MG132 was presented in this period as well. Cytosolic and nuclear extracts were isolated, and after WB analysis, the membrane was probed with antihGR and anti-actin antibodies. Similar behavior was observed in five independent experiments.

the colchicine treated cells as compared to the control ones (Fig. 6). Taken together, microtubules disruption in HeLa cells leads to a specific enhancement of hGR ubiquitination and proteasome-dependent degradation which is reverted by MG132. 3.5. Effect of microtubules-disrupting compounds on hGR degradation in transiently transfected COS-1 cells Synthetic glucocorticoid dexamethasone (DEX), as was described recently, induces proteasome–ubiquitin-dependent mGR degradation in transiently transfected COS-1 cells [11]. The key residue involved in the degradation was identified close to a conserved PEST motif as Lys426,

193

Fig. 7. Effect of microtubules disrupting compounds on hGR degradation in transiently transfected COS-1 cells. COS-1 cells were transiently transfected with empty pSG5 plasmid, wt-hGR, m-hGR(K419A) and/or mhGR(S226A) by lipofection (FuGENE 6), plated on Petri dishes (100 mm I.D.) in density 2106 cells/dish and stabilized for 24 h. Following serum withdrawal from the medium (4–6 h prior to the treatments), cells were treated with colchicine (COL; 1 AM) and/or DMSO as vehicle for 45 min. Cytosolic extract was isolated, and after WB analysis, the membrane was probed with anti-hGR and anti-actin antibodies. Similar behavior was observed in three independent experiments.

which is ubiquitinated thus triggering hGR degradation by proteasome. Indeed, mutation of this residue resulted in the abrogation of dexamethasone-induced mGR degradation [11]. Other authors deal with the role of Ser246 in rGR transcriptional activity, demonstrating rGR inhibition when phosphorylated by c-jun N-terminal kinase on this residue [17]. Thus, we designed an experiment using COS-1 cells transiently transfected with wild-type hGR (wt-hGR), or mutants on Lys419, a residue corresponding to mouse Lys426 (m-hGR(K419A)), and on Ser226, a residue corresponding to rat Ser246 (m-hGR(S226A)). Surprisingly, degradation of hGR did not occur after colchicine treatment of cells transfected with any plasmid (Fig. 7). However, in our recent work, we observed the same phenomenon, i.e., strong microtubules-disrupting compounds effect in bnormalQ and no effect in over-expression systems [15]. Note that there is no hGR in nontransfected cells and that the level of hGR is dramatically increased in the case of

Fig. 6. Effect of microtubules disrupting compounds on hGR ubiquitination in HeLa cells. HeLa cells were pretreated with MG132 (1 AM; 1 h) and/or DMSO prior to the treatment with dexamethasone (DEX; 100 nM; 25 min) and/or colchicine (COL; 1 AM; 25 min). Cytosolic extract from the cells was subjected to immune-precipitation using anti-hGR antibody, as described in experimental part. After WB analysis, the membrane was probed with antibodies against ubiquitin, hGR and actin. Similar results were obtained from three independent experiments.

194

Z. Dvorˇa´k et al. / Cellular Signalling 17 (2005) 187–196

tubules-disrupting compounds on the activation of NFnB in HeLa cells. NFnB activation was estimated as nuclear translocation of p50 and p65 factors, the principal functional constituents of NFnB. None of the compounds tested had any effect on NFnB status in HeLa cells, while tumor necrosis factor alpha (TNFa), a typical NFnB activator, clearly induced p50/p65 nuclear translocation (Fig. 9). Thus, the putative NFnB activation after microtubules disruption is not the case in HeLa cells. 3.8. Effect of c-jun N-terminal kinase inhibitor SP600125 on hGR degradation in HeLa cells Fig. 8. Effect of microtubules disrupting compounds on hGR translocation in HeLa cells preincubated with MG132. HeLa cells were preincubated with proteasome inhibitor MG132 (1 AM) for 1 h prior to the treatment with colchicine (COL; 1 AM), nocodazole (NOC; 1 AM), vincristine (VIN; 1 AM) and/or DMSO as vehicle. Treatment duration was 25 min and MG132 was presented in this period. Thereafter, dexamethasone (DEX; 100 nM) and/or DMSO as vehicle were added for further 25 min. Cytosolic and nuclear extracts were isolated, and after WB analysis, the membrane was probed with anti-hGR and anti-actin antibodies. Similar results were obtained from two independent experiments.

bnondegradableQ K419A mutant, suggesting high turnover of hGR protein in the cell. 3.6. Effect of microtubules-disrupting compounds on dexamethasone-mediated hGR nuclear translocation in HeLa cells protected by MG132 As microtubules-disrupting compounds caused rapid hGR degradation in HeLa cells, evaluation of their effect on hGR hormone-dependent nuclear translocation was rather argumentative due to two simultaneously occurring events, i.e., degradation and translocation (Fig. 4). We took an advantage from the MG132-mediated reversibility of hGR degradation by microtubules-disrupting compounds and keeping in mind that dexamethasone-induced GR translocation is unaffected (Fig. 5). MG132 protected HeLa cells (1 AM, 1 h preincubation) were pretreated with colchicine, nocodazole, or vincristine (1 AM; 25 min) prior to dexamethasone (100 nM; 25 min) or vehicle (DMSO) addition. Dexamethasone-dependent hGR nuclear translocation was not affected by any of the microtubulesdisrupting compounds used (Fig. 8).

Because microtubules-disrupting compounds were described to activate JNK [16] and JNK involvement in rGR inhibition was demonstrated [17], we examined the effect of JNK inhibitor SP600125 on microtubules-disrupting compounds-mediated hGR degradation in HeLa cells. Because the IC50s for JNK inhibition by SP600125 are an order of magnitude less than 10 nM [22], we pretreated the cells were with 0, 1, and 10 AM for 1 h. Thereafter, cells were treated for 25 min with colchicine (1 AM) or DMSO as vehicle. Pretreatment with SP600125 had no effect on hGR degradation induced by colchicine (Fig. 10); therefore, JNK is not likely connected to this process. Furthermore, no effect of JNK activator TNFa on hGR protein degradation favors this hypothesis.

4. Discussion Role of microtubules as a cellular instrument supporting the cell’s proper function is often regarded as necessary. The presented work aims at strengthening the notion. We demonstrate that disruption of microtubules by three structurally diverse compounds, i.e., colchicine, vincristine, and nocodazole, leads to a swift, dose- and time-dependent degradation of human glucocorticoid receptor (hGR) in human cervical carcinoma cell line (HeLa). This process is proteasome–ubiquitin-dependent because: (a)

hGR degradation was reversed by proteasome inhibitor MG132; and

3.7. NFB activation by microtubules-disrupting compounds in HeLa cells One of the crucial factors determining hGR transcriptional activity is a well-known nuclear factor kappa B (NFnB). Indeed, it was shown that disruption of microtubules leads to NFnB activation [25] and several mechanism of competition between hGR and NFnB were proposed [3]. Therefore, we evaluated the effect of micro-

Fig. 9. Activation of NFnB in HeLa cells by microtubules disrupting compounds. HeLa cells were treated with TNFa (20 ng/ml), colchicine (COL; 1 AM), nocodazole (NOC; 1 AM), vincristine (VIN; 1 AM) and/or DMSO as vehicle for 25 min. Nuclear extracts from HeLa cells were isolated, and after WB analysis the membranes were probed with antibodies against p65 and actin. Similar behavior was observed in three independent experiments.

Z. Dvorˇa´k et al. / Cellular Signalling 17 (2005) 187–196

Fig. 10. Effect of c-jun N-terminal kinase inhibitor SP600125 on hGR degradation in HeLa cells. HeLa cells were pre-incubated with JNK inhibitor SP600125 (1 and 10 AM) and/or DMSO as vehicle for 1 h prior to the treatment with colchicine (COL; 1 AM) and/or vehicle. Treatment duration was 25 min and SP600125 was presented in this period as well. In parallel, cells were treated with JNK activator TNFa (20 ng/ml) for 25 min. Cytosolic extract was isolated, and after WB analysis, the membrane was probed with anti-hGR and anti-actin antibodies. Similar behavior was observed in two independent experiments.

(b)

an augmented hGR ubiquitination in response to colchicine treatment was observed in immune-precipitation assays.

The data are in contrast with the observation made in primary human hepatocytes where hGR nucleo-cytoplasmic shuttling was inhibited by colchicine [15], suggesting the coexistence of two different mechanisms of microtubulesdisrupting compounds anti-glucocorticoid activity in HeLa cells and primary human hepatocytes, i.e., protein degradation and inhibition of translocation. While in HeLa cells protected by MG132 pretreatment, hGR shuttling in response to dexamethasone was unaffected by microtubules-disrupting compounds (Fig. 7); the rapid hGR protein degradation, apparent already at 5 min, was observed in microtubules-disrupting compounds treated HeLa cells even after prolonged treatment periods and increased compounds doses (Figs. 2, 3). Thus, there is a clear line of evidence that disruption of microtubules results in the inactivation of hGR by at least two different mechanisms which are intertwined and may proceed at a variable pace depending on a cell type. Because dose- and time-dependent hGR protein degradation was observed in HeLa cells treated by three structurally different compounds which cause unequivocal microtubules depolymerization (Fig. 1), the disruption of microtubules constitutes, in our view, a primary event followed by subsequent hGR degradation. On the other hand, the drop in hGR content during kinetic experiments was remarkable as early as after 5 min of treatment, suggesting that other common cellular target might be affected by microtubulesdisrupting compounds tested (e.g., hsp90, a kinase/phosphatase, etc.). The data presented in this work show that hGR is degraded by proteasome–ubiquitin pathway in response to microtubules disruption. Degradation reversal by proteasome inhibitor MG132 and increased level of ubiquitinated forms of hGR testify strongly for this fact. Indeed, proteasome–ubiquitin complex was reported to be a regulatory tool in mGR homeostasis, serving as a negative feedback in the presence of a ligand [11]. So-called PEST degradation motifs involved in proteasome-controlled GR turnover were found in human, rat, and mouse GR. In case

195

of mGR, Lys426 was identified as a pivotal for ubiquitination by E2/E3 ligases. The ubiquitin-labeled mGR is recognized by proteasome complex and degraded [11]. Similarly, Lys419 in hGR, a residue homologous to mouse Lys426, likely plays the same role in hGR degradation. However, the processes, which precede the ubiquitin ligation step and are associated with GR regulation, activation, and degradation, are not completely understood yet. We anticipate that microtubules disruption triggers an unidentified process leading to hGR ubiquitination and consequent degradation by proteasome. hGR contains multiple phosphorylation sites in its Nterminal regulatory region, phosphorylation of which plays both positive and negative role in the receptor activation [23]. There are four main phosphorylation sites within rGR identified as Thr171, Ser224, Ser232, and Ser246 [24]. Furthermore, antagonism of rGR transcriptional activation by JNK was observed, with the phosphorylation of rGR Ser246 acting the critical part [17]. We can speculate that phosphorylation of Ser226 in hGR, a homologue of rGR Ser246, by activated JNK could explain hGR transcriptional inhibition caused by microtubules disruption because activation of JNK by microtubules interfering agents was published [16]. Moreover, this hypothesis is supported by observed time-lag (6–12 h) in hGR inhibition by COL in human hepatocytes [15] corresponding with peak-times of JNK activation (2–8 h) [16]. However, hGR degradation by microtubules-disrupting compounds in HeLa cells does not require active JNK hinting at the existence of a distinct mechanism leading to hGR degradation. Indeed, we did not observe any effect of typical JNK activator TNFa on hGR content in HeLa cells (Fig. 10). We also eliminated the possibility of NFnB pathway involvement in the interaction between microtubules network and hGR in HeLa cells (Fig. 8). Although negative, this finding is important as NFnB is a prominent hGR rival [3], and NFnB activation by microtubules disruption was published [25]. Taken together, the data obtained signalise that intact microtubules network is essential for proper hGR function; however, the mechanism of microtubules disrupting compounds action on hGR is distinct in various cells, emphasising again inhibition of nucleo-cytoplasmic shuttling in human hepatocytes vs. protein degradation in HeLa cells. In conclusion, the results presented here bring novel insight at hGR regulation and homeostasis. Furthermore, this knowledge has important implication and impact in terms of drug metabolism, inflammatory affairs and understanding of molecular mechanism of steroid receptors regulation.

Acknowledgements We thank Dr. Andrew Wallace for valuable hints and helpful discussion. This research was supported by grant MSM 151100003 from the Ministry of Education, Youth

196

Z. Dvorˇa´k et al. / Cellular Signalling 17 (2005) 187–196

and Sports of the Czech Republic, and by grant GACR 303/ 04/P074 from the Grant Agency of the Czech Republic.

References [1] [2] [3] [4] [5] [6] [7]

[8] [9] [10] [11] [12]

L.I. McKay, J.A. Cidlowski, Endocr. Rev. 20 (1999) 435 – 459. I.M. Adcock, G. Caramori, Immunol. Cell Biol. 79 (2001) 376 – 384. M. Gottlicher, S. Heck, P. Herrlich, J. Mol. Med. 76 (1998) 480 – 489. J.M. Pascussi, S. Gerbal-Chaloin, L. Drocourt, P. Maurel, M.J. Vilarem, Biochim. Biophys. Acta 1619 (2003) 243 – 253. T.H. Rushmore, A.N. Kong, Curr. Drug Metab. 3 (2002) 481 – 490. R.G. Tirona, B.F. Leake, A.W. Wolkoff, R.B. Kim, J. Pharmacol. Exp. Ther. 304 (2003) 223 – 228. J.M. Pascussi, S. Gerbal-Chaloin, L. Pichard-Garcia, M. Daujat, J.M. Fabre, P. Maurel, M.J. Vilarem, Biochem. Biophys. Res. Commun. 274 (2000) 707 – 713. K.D. Dittmar, D.R. Demady, L.F. Stancato, P. Krishna, W.B. Pratt, J. Biol. Chem. 272 (1997) 21213 – 21220. M. Shimojo, N. Hiroi, F. Yakushiji, H. Ueshiba, N. Yamaguchi, Y. Miyachi, Endocr. J. 42 (1995) 629 – 636. D.L. Bellingham, M. Sar, J.A. Cidlowski, Mol. Endocrinol. 6 (1992) 2090 – 2102. A.D. Wallace, J.A. Cidlowski, J. Biol. Chem. 276 (2001) 42714 – 42721. M.D. Galigniana, J.L. Scruggs, J. Herrington, M.J. Welsh, C. CarterSu, P.R. Housley, W.B. Pratt, Mol. Endocrinol. 12 (1998) 1903 – 1913.

[13] M.D. Galigniana, P.R. Housley, D.B. DeFranco, W.B. Pratt, J. Biol. Chem. 274 (1999) 16222 – 16227. [14] M. Nishi, N. Takenaka, N. Morita, T. Ito, H. Ozawa, M. Kawata, Eur. J. Neurosci. 11 (1999) 1927 – 1936. [15] Z. Dvoba´k, M. Modriansky´, L. Pichard-Garcia, P. Balaguer, M.J. Vilarem, J. Ulrichova´, P. Maurel, J.M. Pascussi, Mol. Pharmacol. 64 (2003) 160 – 169. [16] T.H. Wang, H.S. Wang, H. Ichijo, P. Giannakakou, J.S. Foster, T. Fojo, J. Wimalasena, J. Biol. Chem. 273 (1998) 4928 – 4936. [17] I. Rogatsky, S.K. Logan, M.J. Garabedian, Proc. Natl. Acad. Sci. U. S. A. 95 (1998) 2050 – 2055. [18] J.M. Carcamo, A. Pedraza, O. Borquez-Ojeda, D.W. Golde, Biochemistry 41 (2002) 12995 – 13002. [19] C.M. Stoscheck, Quantitation of protein, Methods Enzymol. 182 (1990) 50 – 68. [20] U.K. Laemmli, Nature 227 (1970) 680 – 685. [21] H. Towbin, T. Staehelin, J. Gordon, Proc. Natl. Acad. Sci. U. S. A. 76 (1979) 4350 – 4354. [22] B.L. Bennett, D.T. Sasaki, B.W. Murray, E.C. O’Leary, S.T. Sakata, W. Xu, J.C. Leisten, A. Motiwala, S. Pierce, Y. Satoh, S.S. Bhaqwat, A.M. Manning, D.W. Anderson, Proc. Natl. Acad. Sci. U. S. A. 98 (2001) 13681 – 13686. [23] J.C. Webster, C.M. Jewell, J.E. Bodwell, A. Munck, M. Sar, J.A. Cidlowski, J. Biol. Chem. 272 (1997) 9287 – 9293. [24] M.D. Krstic, I. Rogatsky, K.R. Yamamoto, M.J. Garabedian, Mol. Cell. Biol. 17 (1997) 3947 – 3954. [25] C. Rosette, M. Karin, J. Cell Biol. 128 (1995) 1111 – 1119.