Differential effects of selected natural compounds with anti-inflammatory activity on the glucocorticoid receptor and NF-κB in HeLa cells

Chemico-Biological Interactions 159 (2006) 117–128

Differential effects of selected natural compounds withanti-inflammatory activity on the glucocorticoid

receptor and NF-�B in HeLa cells

Zdenek Dvoraka,∗, Radim Vrzala,b, Patrick Maurelb, Jitka Ulrichovaa

a Institute of Medical Chemistry and Biochemistry, Medical Faculty, Palacky University Olomouc, Hnevotınska 3,77515 Olomouc, Czech Republic

b INSERM – UMR632, 1919 Route de Mende, 34293 Montpellier, France

Received 20 July 2005; received in revised form 27 September 2005; accepted 5 October 2005Available online 10 November 2005

Abstract

Natural compounds have been used in the treatment of various diseases for centuries. Herein, we investigated the effects of struc-turally diverse alkaloids with anti-inflammatory activity (berberine, sanguinarine, chelerythrine, and colchicine) on two importantanti-inflammatory and pro-inflammatory players, i.e. glucocorticoid receptor (GR) and nuclear factor kappa B (NF-�B), respec-tively. Sanguinarine and chelerythrine elicited nuclear translocation of the p65 subunit of NF-�B. The nuclear import of p65 wasstrongly augmented by these akaloids in non-stimulated cells as well as in cells challenged with tumor necrosis factor alpha (TNF�).Colchicine and berberine had no effect on p65 nuclear translocation regardless of the presence or absence of TNF�. Colchicine

rine andscrip-onsedand

dst.

ralonlarerti-

caused rapid degradation of the GR protein, whereas berberine had no effect on GR content or cellular localization. Sanguinachelerythrine induced accumulation of GR in the nucleus with concomitant diminution of cytosolic GR. Analyses on the trantional activity of GR and NF-�B monitored by reporter assays using HeLa cells transiently transfected with glucocorticoid respelement (pGRE-LUC) and/or NF-�B elements fused to luciferase gene (pNF-�B-luc) showed that none of the compounds tested hathe capability to trigger GR and/or NF-�B transcriptional activities, respectively. The ligand binding assay showed that colchicineberberine are not GR ligands whereas sanguinarine and chelerythrine significantly decreased binding of3H-labelled dexamethasoneto GR. In conclusion, structurally diverse natural antiflogistics displayed differential effects on GR and NF-�B in HeLa cells.© 2005 Elsevier Ireland Ltd. All rights reserved.

Keywords: Berberine; Colchicine; Sanguinarine; Chelerythrine; Glucocorticoid receptor; Nuclear factor kappa B; HeLa cells

Abbreviations: BER, berberine; CHE, chelerythrine; COL,colchicine; COS-1, Simian kidney cells; DEX, dexamathasone; GR,glucocorticoid receptor; GRE-LUC, glucocorticoid response elementfused to luciferase; HeLa, Human Negroid cervix epitheloid carcinomacells; LDH, lactate dehydrogenase; NF-�B, nuclear factor kappa B;QBAs, quaternary benzo[c]phenanthridine alkaloids; SA, sanguinar-ine; TNF�, tumor necrosis factor alpha

∗ Corresponding author. Tel.: +420 58 5632324;fax: 420 58 5632302.

E-mail address: [email protected] (Z. Dvorak).

1. Introduction

The use of natural compounds in the treatment anprevention of various disorders has advanced in the latwo decades, following the boom of synthetic drugsDietary supplements and remedies derived from natusubstances are of growing importance in the preventiand management of such conditions as cardiovascudiseases, immune system disorders and Alzheimdisease among others. Natural compounds with an

0009-2797/$ – see front matter © 2005 Elsevier Ireland Ltd. All rights reserved.doi:10.1016/j.cbi.2005.10.105

118 Z. Dvorak et al. / Chemico-Biological Interactions 159 (2006) 117–128

Fig. 1. Structure of colchicine, berberine, sanguinarine, and chelerythrine.

inflammatory activities are an important group oftherapeutics, as inflammation is common denominatorin a variety of diseases. In the present paper, we focuson four structurally diverse, commonly used naturalcompounds with anti-inflammatory activity, namelycolchicine, berberine, sanguinarine, and chelerythrine(for structures seeFig. 1). The tropolone alkaloidcolchicine was used in the treatment of gout as earlyas from the 6th century A.D. It possesses anti-mitotic,anti-inflammatory and anti-fibrotic properties[1].Currently, it is used as a drug of choice for acute goutattacks and to treat diseases of autoimmune origin suchas Familial Mediterranean Fever, Behcet’s disease,and scleroderma. Colchicine inhibits the chemotaxisof leucocytes and prevents degranulation of polymor-phonuclear neutrophiles[1]. However, the mechanismof colchicine anti-inflammatory activities is not fullyunderstood yet[2]. The quaternary isoquinoline alkaloidberberine has been extensively studied as a promisingantimalarial drug[3,4]. Owing to its anti-inflammatory,anti-microbial, anti-motility, and anti-secretory activi-ties, extracts containing berberine were used as a remedyfor the treatment of diarrhea and gastro-enteritis for cen-turies[5]. However, there is dearth of information on themechanisms involved in its anti-inflammatory activity.It is known to inhibit gene expression of prostaglandinH synthase (COX-2), an important enzyme in the

inflammatory response[6]. COX-2 down-regulation ismediated probably via suppression of activator protein1 [7]. The quaternary benzo[c]phenanthridine alkaloids(QBAs) sanguinarine and chelerythrine exert a widespectrum of biological activities, e.g. anti-microbial,anti-fungal, anti-inflammatory, adrenolytic, sympa-tholytic and local anesthetic. Due to their biologicalactivities, QBAs are used in dental care products[8].The capability of sanguinarine but not chelerythrine toinhibit activation of NF-�B has been reported[9]. Asin the case of colchicine and berberine, however, thedetails of QBAs anti-inflammatory action is not exactlyknown, as QBAs have multiple cellular targets[8]. Thecompounds described above share the following: (i)They have been traditionally used as phytotherapeuticsfor their anti-inflammatory activities. (ii) The molecularmechanisms involved are not fully understood yet.

The aim of this work was to investigate the putativeeffects of four structurally diverse natural therapeuticswith anti-inflammatory activities (colchicine, berberine,sanguinarine and chelerythrine) on two key factors in-volved in the inflammatory response; the glucocorticoidreceptor (GR) and the nuclear factor kappa B (NF-�B).GR plays an important role in a variety of cellular pro-cesses such as development, differentiation and home-ostasis[10]. One of the most prominent functions of GRis its unique entanglement in inflammation control[11].

Z. Dvorak et al. / Chemico-Biological Interactions 159 (2006) 117–128 119

During inflammation several transcriptional factors, NF-�B being an example, are activated in response to therelease of cytokines by macrophages. These serve as atoggle switch for a whole battery of genes while GR hasthe opposite effect[12]. Several mechanisms of inter-action between GR and pro-inflammatory factors havebeen proposed or shown[11,13,14].

In this work, we have examined the effects ofcolchicine, berberine, sanguinarine, and chelerythrineon: (1) The activation of NF-�B, assessed as (a) nucleartranslocation of the p65 subunit; (b) luciferase activity inthe cells transiently transfected with the NF-�B respon-sive element fused to the luciferase gene. (2) Activationof the glucocorticoid receptor, assessed as (a) nucleartranslocation of the GR protein; (b) luciferase activityin cells transiently transfected with the glucocorticoidresponsive element fused to the luciferase gene. HeLacells which possess endogenous and functional GR andNF-�B were the model of choice.

2. Materials and methods

Cell culture media and supplements, dexametha-sone (DEX), colchicine (COL), berberine hemisulphate(BER, purity 95%), tumor necrosis factor alpha (TNF�),dithiothreitol, phenylmethanesulfonyl fluoride (PMSF),horseradish peroxidase conjugated secondary antibod-ies, and foetal calf serum were purchased from Sigma(St. Louis, MO, USA). An ECL kit including hyperfilmphotographic paper and [1,2,4-3H]-dexamethasone werep d).S iso-l c.,B ona .p.2 ty,mp ularB mi-c mer-c

2

ellsH inD p-p cin,1 -e 7i et mm

i.d.) in a density of 2× 106 cells/dish. Cells were al-lowed to grow up to 70–80% of confluence. Followingexchange for a serum-free medium, the culture was sta-bilized for an additional 4–6 h prior to the treatments.

2.2. Cytotoxicity assays

HeLa cells were seeded on 24-well dishes in a den-sity of 1× 105 cells/well using culture media enrichedwith foetal calf serum (10% v/v). Following 16 h of sta-bilization, the medium was exchanged for a serum-freeone and the cells were treated 30 min, 12 h and 24 h withcolchicine, berberine, sanguinarine and chelerythrine inconcentrations of up to 10�M. In parallel, cultures weretreated with DMSO (vehicle) and 1% v/v Triton X-100to assess the minimal and maximal cell damage, respec-tively. Lactate dehydrogenase (LDH) leakage into themedium (kit Promega) and MTT assay were measuredas the indicators of membrane damage and viability,respectively.

2.3. Preparation of nuclear and cytosolic extracts

Nuclear and cytosolic extracts were prepared as de-scribed elsewhere[17] with minor modifications. Briefly,following the treatments HeLa cells were washed twicewith 2 mL of ice-cold PBS and scraped into 1 mL of PBS.The suspension was centrifuged (1500× g/5 min/4◦C)and the pellet was resuspended by gentle pipetting in300�L of ice-cold buffer A (10 mM Hepes pH 7.9;

ice

cy-dedr BA;ven-u-ed atined

zedn adureaneel of

urchased from Amersham (Little Chalfont, Englananguinarine (SA) and chelerythrine (CHE) were

ated from sanguiritrine (CAMAS Technologies, Inroomfield, USA) using column chromatographylumina[15]. Sanguinarine (SA) in 98.1% purity, m79–282◦C[16] and chelerythrine (CHE) in 95% puri.p. 200–204◦C [16] were obtained. CompleteTM Minirotease inhibitor was purchased from Roche Moleciochemicals (Mannheim, Germany). All other cheals and reagents were of the highest quality comially available.

.1. Cell cultures

Human Negroid cervix epitheloid carcinoma ceLa (ECACC No. 93021013) were culturedulbecco’s modified Eagle’s medium (DMEM) sulemented with 10% FCS, 100 U/mL streptomy00�g/mL penicillin, 4 mM l-glutamine, and nonssential amino acids. Cultures were maintained at 3◦C

n 5% CO2 (air:CO2, 95:5) humidified incubator. For threatments, the cells were plated on Petri dishes (100

10 mM KCl; 1.5 mM MgCl2; 0.5 mM DTT; 0.1% v/vNP-40). The mixture was incubated for 10 min onand then centrifuged (12000× g/10 min/4◦C). Follow-ing collection of the supernatant which containedtosolic fraction, the pellet was vigorously resuspenby syringe/needle in three volumes of ice-cold buffe(20 mM Hepes pH 7.9; 420 mM NaCl; 0.2 mM EDT1.5 mM MgCl2; 0.5 mM DTT; 0.5 mM PMSF; 25% v/glycerol) and incubated for 30 min on ice. Following ctrifugation (12,000 g/20 min/4◦C), the supernatant (nclear extract) was collected. Both extracts were stor−80◦C. The protein content in extracts was determby the biscinchoninic acid method[18].

2.4. Western blots

The cytosolic and nuclear extracts were analyas follows: SDS-PAGE gels (7.5%) were run oHoefer apparatus according to the general proce[19]. Protein transfer onto nitrocellulose membrwas carried out as described[20]. The membranwas stained with Ponceau S red dye for contro


transfer and then saturated with 8% non-fat dried milkovernight. Blots were probed with primary antibodiesagainst: human glucocorticoid receptor (GR(E-20)Xrabbit polyclonal; dilution 1/1000); p65 (NF-�B p65(A)rabbit polyclonal; dilution 1/500), and�-actin (Actin(I-19) goat polyclonal; dilution 1/1000); all purchasedfrom Santa Cruz Biotechnology, Inc. (Santa Cruz,USA). Chemiluminescence detection using horseradishperoxidase conjugated secondary antibodies and anAmersham ECL kit was performed.

2.5. Reporter assays

The pTAT-(GRE)2-TK-luc plasmid containing twocopies of the consensus GRE upstream of a minimalherpes simplex virus thymidine kinase promoter and aluciferase reporter gene was kindly provided by Dr. L.Poellinger (Karolinska Institute, Stockholm, Sweden).The pNF-�B-luc reporter plasmid containing direct re-peats of the transcription recognition sequences for NF-�B (PathDetect® NF-�B Cis-Reporting System) waspurchased from Stratagene (La Jolla, CA, USA). Forreporter assays, HeLa cells were transiently transfectedby lipofection (FuGENE 6) with 300 ng of luciferase re-porter construct pTAT-(GRE)2-tkLUC or pNF-�B-luc.Following a 16 h incubation period, cells were treated1, 3, 6, 12, and 24 h with tested compounds. In parallel,dexamethasone and TNF� were applied to the cells aspositive controls for the assessment of GR and NF-�Bactivation, respectively. After the treatments, cells were

dard-

ro-Hs aex-

hedthe

singlas,reingd in

%

cen-u-5

DEX alone or in combination with DMSO, 1�M un-labelled DEX, 10�M colchicine, 10�M sanguinarine,10�M chelerythrine or 10�M berberine, and incubatedovernight at 4◦C. With 100�L of the incubation reac-tion, bound and free steroids were separated by gel fil-tration using Sephadex LH-20 (Pharmacia, Guyancourt,France) and radioactivity was measured with 600�Lof the total eluate of 1000�L in a liquid scintillationcounter Packard Tri-Carb 2100TR (Canberra Packard,Savigny-le-Temple, France). The presence of GR in thecytosolic fraction was verified using Western blot.

3. Results

Since the alkaloids tested were without doubt toxiccompounds, we determined their cytotoxic potentialagainst human cervix carcinoma cells (HeLa) prior tothe treatments. Lactate dehydrogenase (LDH) activitymeasurements in the medium and MTT assay wereperformed to assess cell membrane damage and themetabolic status of cells, respectively. The concentra-tions under investigation ranged from 0.1 to 10�M; thetime periods selected for testing were 30 min, 12 h, and24 h (Table 1). The tested compounds were not cyto-toxic after 30 min of incubation in any concentration.In contrast, the viability of the cells challenged for 12and 24 h with sanguinarine, chelerythrine and colchicinein 10�M concentrations was significantly decreased(Table 1). The alkaloid doses that caused no more than a10% decrease in cell viability were considered non toxic

in-

tororTheatures ofminion.

tion

p65a

lysed and luciferase activity was measured and stanized per milligram of protein.

2.6. Ligand binding assay

Ligand binding was performed according to the pcedure of Herr[21]. Simian kidney cells (COS-1; NIATCC, Bethesda, MD) were the model of choice aconvenient system for preparative over-expressingperiments. Cells were cultured in a medium enricwith foetal calf serum (10% v/v) as recommended bysupplier and transfected with wild type hGR-expresplasmid pSG5-hGR (kindly provided by Dr. J.C. NicoINSERM, Montpellier, France). After 16 h, cells wewashed twice with PBS, scraped in 1 mL of bindbuffer/10 cm petri dish, centrifuged and resuspendetotal 1 mL of binding buffer (5 mM Tris–HCl, pH 7.4; 5glycerol; 1 mM EDTA; 10 mM Na2MoO6(2H2O)). Cellswere incubated 10 min on ice, well homogenized,trifuged (35,000 rpm/4◦C/ 1 h; Ti 70.1 rotor), and the spernatant, the cytosolic fraction, was collected. 143.�Lof the cytosol was then mixed with 5�L of 300 nM [3H]-

for further experiments.

3.1. Effects of studied alkaloids on NF-κB nucleartranslocation in HeLa cells

HeLa cells were pre-incubated for 10 min withcreasing doses of tested compounds (0.01–10�M) andthen challenged for 30 min with tumor necrosis facalpha (TNF�, 20 ng/mL) and/or DMSO as a vehicle fcontrol in the presence of the tested compounds.exposure times were chosen based on reported literand on our own observations. The nuclear levelp65 and/or GR proteins reach a maximum at 15–40and this is followed by nuclear export and degradatColchicine and berberine had no effects on NF-�B (p65)nuclear content in either unstimulated or TNF� treatedHeLa cells, while TNF�, a typical NF-�B activator,clearly induced p65 nuclear translocation (Fig. 2Aand B). Despite reports suggesting that disrupof microtubules leads to NF-�B activation [22], wefailed to observe such an effect of colchicine onnuclear levels in our experiments[23,24]. These dat


Table 1Cytotoxicity of tested compounds in HeLa cells

Alkaloid c (�M) LDH (% of damage) MTT (% of contol)

30 min 12 h 24 h 30 min 12 h 24 h

Sanguinarine 0.1 0.9± 0.1 1.1± 0.2 1.5± 0.5 100.3± 5.9 99.8± 7.7 101.7± 8.41 1.2± 0.2 2.1± 0.4 4.1± 0.7 98.9± 8.3 95.1± 6.9 92.2± 7.3

10 2.1± 0.1 28.3± 3.1 47.2± 4.7 97.1± 6.7 55.1± 6.7 35.1± 3.7

Chelerythrine 0.1 1.3± 0.2 1.1± 0.3 1.9± 0.4 98.7± 4.8 101.4± 6.3 99.9± 8.61 1.2± 0.3 1.9± 0.3 3.2± 0.5 102.5± 7.1 101.8± 7.9 95.1± 6.4

10 1.8± 0.1 22.4± 4.0 36.E± 5.1 98.2± 5.8 72.3± 7.9 39.2± 4.3

Berberine 0.1 1.9± 0.2 2.4± 0.4 1.3± 0.2 100.1± 4.9 100.8± 6.7 99.8± 6.21 1.3± 0.3 1.1± 0.3 2.3± 0.3 99.6± 7.6 100.2± 9.3 100.2± 7.3

10 1.2± 0.1 1.2± 0.3 1.1± 0.2 99.1± 6.8 99.8± 5.9 100.1± 8.4

Colchicine 0.1 1.5± 0.1 1.4± 0.1 1.7± 0.3 101.1± 5.6 98.8± 8.7 100.1± 7.41 1.3± 0.2 2.1± 0.3 2.3± 0.5 99.8± 6.9 91.1± 8.8 71.2± 6.7

10 1.4± 0.2 1.6± 0.1 1.1± 0.2 98.1± 5.8 38.3± 4.9 27.1± 4.2

Control 0 0 0 100 100 100Triton X-100 100 100 100 0 0 0

The activity of lactate dehydrogenase was measured in the media. The cytotoxicity was calculated as follows: LDH (% of dam-age) = 100× (LDHsample− LDHcontrol)/(LDHTriton− LDHcontrol) MTT (% of control) = 100× MTTsample/MTTcontrol.

imply that the inhibition of cytokines elicited NF-�Bnuclear translocation is not the mechanism of the anti-inflammatory activity of colchicine and berberine. Incontrast, sanguinarine induced dose-dependent nucleartranslocation of p65 that was remarkable at 10�Mconcentrations of sanguinarine (Fig. 2C). Chelerythrine

slightly augmented the level of p65 protein in the nu-cleus at a concentration of 1�M, however, its effect wasmuch weaker than that of sanguinarine and the processwas not concentration dependent (Fig. 2C). NF-�B(p65) nuclear translocation by TNF� was potentiatedby both alkaloids tested; the effect of each displayed

Fig. 2. Effect of studied alkaloids on the nuclear translocation of NF-�B in HeLa cells. Shown are representative Western blots of p65 proteindetected in nuclear extracts from HeLa cells. Cultures were pre-treated for 10 min with colchicine (COL; 1 and 10�M), berberine (BER; 1 and10�M), sanguinarine (SA; 1 and 10�M), chelerythrine (CHE; 1 and 10�M) and/or DMSO as vehicle for control. Thereafter, HeLa cells werechallenged for 30 min with tumor necrosis factor alpha (TNF�, 20 ng/mL) or DMSO in the presence of the alkaloids. Nuclear extracts from the cellswere isolated, and following Western blot analysis the membranes were probed with antibody against p65. Similar behavior was observed in fivei
ndependent experiments.


Fig. 3. Effect of studied alkaloids on NF-�B activation in HeLa cells. Bar graphs summarizing the transcriptional activity of NF-�B in transientlytransfected HeLa cells. Cells transfected with reporter plasmid containing direct repeats of the transcription recognition sequences for NF-�B(pNF-�B-luc) were incubated for 1 h (Panel A), 3 h (Panel B), 6 h (Panel C), 12 h (Panel D) and 24 h (Panel E) with colchicine (COL; 1 and 10�M),berberine (BER; 1 and 10�M), sanguinarine (SA; 1 and 10�M), chelerythrine (CHE; 1 and 10�M), tumor necrosis factor alpha (TNF�; 20 ng/mL)and/or vehicle (DMSO). Chemiluminescent activity of luciferase reporter gene in cell lysates was measured and normalized to protein concentration.The data are expressed as the mean± S.D. from three independent experiments. (*) Value is significantly different from the activity of DMSO-treatedcells (p < 0.05). Due to cytotoxicity, COL, SA, and CHE were applied to the cells in maximal concentrations of 1�M in 12 and 24 h treatments.n.d. = not determined.


a pattern similar to non-stimulated cells (Fig. 2C).The elevation of p65 nuclear levels by sanguinarineand chelerythrine is more a sign of pro-inflammatorybehavior at a molecular level than the opposite, whichis paradoxical regarding the anti-inflammatory activityof both agents. In addition, the difference betweenthe potency and the dose-dependence of sanguinarineand chelerythrine to induce nuclear translocation ofNF-�B (p65) is interesting from a mechanistic point ofview.

3.2. Effects of studied alkaloids on thetranscriptional activity of NF-κB in HeLa cellstransiently transfected with pNF-κB-luc

In order to determine whether sanguinarine and chel-erythrine induce nuclear translocation of p65 resultingin NF-�B transcriptional activity we analyzed the effectsof the tested compounds in a reporter assay. HeLa cellswere transiently transfected with pNF-�B-luc plasmidcontaining direct repeats of the transcription recognitionsequences for NF-�B fused to the luciferase gene. Fol-lowing 16 h of incubation, cells were challenged withtested compounds (1 and 10�M), tumor necrosis fac-tor alpha (TNF�, 20 ng/mL) and/or vehicle (DMSO) for1, 3, 6, 12, and 24 h. Treatment with TNF� resultedin strong time-dependent elevation of luciferase activ-ity that reached a plateau between 6 and 12 h (Fig. 3).Studied compounds did not trigger transcriptional activ-ity of NF-�B in any time periode tested, as luciferasea dieda sea r-i f p65i lyi

3t

ion,t thec oteinc re int NAs thet GR.H seso -t sv icitn nu-

tion of cytosolic GR (Fig. 4C). The tropolone alkaloidcolchicine caused concentration dependent degradationof the GR protein that was observed in both cytosolicand nuclear compartments (Fig. 4A). As we have de-scribed elsewhere, this rapid degradation was also timedependent and mediated via proteasome-ubiquitine sys-tem [24]. This “anti-glucocorticoid” behaviour of theanti-inflammatory drug colchicine is paradoxical, takinginto account the role of GR in combating inflammation.The isoquinoline alkaloid berberine had no effect on theglucocorticoid receptor subcellular localization and pro-tein stability in HeLa cells (Fig. 4B). Hence, the mecha-nism of berberine anti-inflammatory activity is not likelyto be linked to the activation of GR. The treatment ofHeLa cells with the quaternary QBAs sanguinarine andchelerythrine resulted in massive accumulation of GR inthe nucleus with concomitant disappearance of GR fromthe cytosol (Fig. 4C). This effect was concentration de-

glu-tativetractsOL;d

10�M), chelerythrine (CHE; 1 and 10�M), dexamethasone (DEX;0.1�M) and/or DMSO as vehicle for control. Cytosolic and nuclearextracts from the cells were isolated, and after Western blot analysis themembranes were probed with antibody against GR. Similar behaviorwas observed in five independent experiments.

ctivities measured in the cells challenged with stulkaloids were not significantly different from luciferactivity in vehicle-treated cells (Fig. 3). Thus, sanguina

ne and chelerythrine cause nuclear accumulation on HeLa cells, whereas NF-�B remains transcriptionalnactive.

.3. Effects of the studied alkaloids on GR nuclearranslocation in HeLa cells

In the process of glucocorticoid receptor activathe ligand (glucocorticoid) binds to the receptor,haperone proteins are then dissociated from the promplex and GR translocates to the nucleus whehe homodimeric form GR/GR binds its cognate Dequences[25]. Thus, we examined the effects ofested compounds on subcellular localization ofeLa cells were treated for 30 min with increasing dof tested compounds (0.01–10�M), synthetic glucocor

icoid dexamethasone (DEX, 0.1�M) and/or DMSO aehicle for control. The addition of DEX induced expluclear translocation of GR with concomitant dimi

Fig. 4. Effect of studied alkaloids on nuclear translocation ofcocorticoid receptor (GR) in HeLa cells. Shown are represenWestern blots of GR protein detected in cytosolic and nuclear exfrom HeLa cells. Cultures were treated 30 min with colchicine (C0.01–10�M), berberine (BER; 1 and 10�M), sanguinarine (SA; 1 an


pendent and was remarkable in concentrations as high as10�M. The activation of GR by QBAs may explain, atleast partly, the mechanism of QBAs’ anti-inflammatoryactivity.

3.4. Effects of the studied alkaloids on thetranscriptional activity of GR in HeLa cellstransiently transfected with GRE-LUC

The presence of GR protein in the nucleus does notnecessarily mean that GR is transcriptionally active.Different GR ligands have distinct effects on GR interms of transcriptional activity and their capabilityto combat with NF-�B [26]. For instance, the anti-glucocorticoid mifepristone (RU486) binds GR, inducesGR nuclear translocation, but in this case GR fails totrigger transcription of target genes and inhibit NF-�B[26]. We tested the effects of the studied alkaloids on thetranscriptional activity of GR. HeLa cells that possessendogenous GR were transiently transfected with GRE-LUC reporter, and the transcriptional activity of GR wasmonitored as luciferase activity. Cells were treated 6, 12,and 24 h with the tested compounds, DEX and/or DMSOas vehicle for control. Glucocorticoid DEX caused 6–10,20–30 and 50–60 fold increase in luciferase activityafter 6, 12, and 24 h, respectively (Fig. 5). No significantincrease in luciferase activity was observed for anyof the alkaloids tested (Fig. 5). Thence, sanguinarineand chelerythrine elicit nuclear translocation of GR butlack the ability to turn-on the transcriptional activity of

thet GRts ofe on-

ob-r-ing.t the

fecton-ind-of

re GR

4. Discussion

In the present work we have examined how selectnatural compounds with anti-inflammatory activity canmodulate cellular pathways involved in the inflamma-tory process. We investigated the effects of colchicine,berberine, sanguinarine and chelerythrine on the gluco-corticoid receptor and the nuclear factor kappa B in HeLacells. The tested alkaloids displayed differential effectson GR and NF-�B activation: (i) Berberine had no ef-fect on GR and NF-�B in terms of nuclear translocationand transcriptional activity of these factors. Apart fromthe capability of berberine to inhibit COX-2 expression[6], the putative molecular mechanism of berberine anti-inflammatory activity remains to be uncovered. Whileberberine down-regulates COX-2 via inhibition of AP-1protein binding to DNA[7], it does not affect the NF-�Bpathway. (ii) Colchicine caused rapid, dose- and time-dependent degradation of the GR protein, as we have re-ported previously[24]. In addition, inhibition of the GRtranscriptional activity in primary cultures of human hep-atocytes was observed[23,27]. The anti-glucocorticoidbehavior of colchicine at a molecular level is paradoxi-cal and does not explain the anti-inflammatory effects ofcolchicine in vivo[28,29]. Furthermore, the mechanismof colchicine anti-inflammatory activity does not involveinterference with the NF-�B pathway. (iii) Sanguinarineand chelerythrine caused nuclear translocation of the p65subunit of NF-�B. However, these compounds failed toinduce transcriptional activity of NF-�B as revealed by

thatation,-

wnucey tose

andBAsF-

GRet-eus.NF-

ionnd

then-

GR.

3.5. Effects of the studied alkaloids onglucocorticoid binding to GR

In the last series of experiments we investigatedeffects of the tested alkaloids on ligand-dependenactivation. For this purpose, we analyzed the effecsanguinarine, chelerythrine, berberine and colchicinthe binding of [3H]-DEX to in vitro translated GR. Nonlabelled DEX decreased the binding of [3H]-labelledDEX to GR approximately down to 7% of the valuetained for vehicle alone (Fig. 6). Colchicine and berbeine did not have any significant effect on this bindThese data imply that colchicine and berberine are noligands for the glucocorticoid receptor and do not afthe binding of glucocorticoids to the receptor. In ctrast, sanguinarine and chelerythrine inhibited the bing of [3H]-DEX to GR approximately down to 45%the value obtained for the vehicle alone (Fig. 6). Thesedata suggest that sanguinarine and chelerythrine aligands.

reporter assay. In addition, it was reported elsewheresanguinarine blocked phosphorylation and degradof I�B� in the process of NF-�B activation by cytokinesphorbol esters and lipopolysaccharide[9], and that sanguinarine restricts NF-�B binding to its cognate DNAsequence[30]. In our experiments we have also shothat QBAs have the capability to bind to GR and indits nuclear translocation, but they lack the potencturn-on the transcriptional activity of GR. Taking thedata together, led us to the following hypothesis (Fig. 7):QBAs activate GR by a ligand dependent mechanisminduce GR nuclear translocation. At the same time Qelicit nuclear translocation of the p65 subunit of N�B by an unknown mechanism. The free forms ofand NF-�B (i.e. GR/GR homodimer and p50/p65 herodimer, respectively) then may interact in the nuclSeveral mechanisms of interaction between GR and�B exist, e.g. protein–protein interaction, competitfor DNA binding sites, physical interaction in DNA, afunctional transcriptional combat[12–14,31]. We ob-served an elevation in p65 and GR protein levels innucleus (Figs. 2 and 4); however, both factors were tra


Fig. 5. Effect of studied alkaloids on glucocorticoid receptor activation in HeLa cells. Bar graphs summarizing the GR transcriptional activity intransiently transfected HeLa cells. Cells transfected with glucocorticoid responsive element fused to the luciferase gene (GRE-LUC) were incubatedfor 6 h (Panel A), 12 h (Panel B) and 24 h (Panel C) with dexamethasone (DEX; 0.1�M), colchicine (COL; 1 and 10�M), berberine (BER; 1 and10�M), sanguinarine (SA; 1 and 10�M), chelerythrine (CHE; 1 and 10�M) and/or DMSO as vehicle for control. The chemiluminescent activityof luciferase reporter gene in cell lysates was measured and normalized to protein concentration. (*) Value is significantly different from the activityof DMSO-treated cells (p < 0.05). Owing to cytotoxicity, COL, SA and CHE were applied to the cells in maximal concentrations of 1�M in 24 htreatments. n.d. = not determined.


Fig. 6. Effect of studied alkaloids on dexamethasone binding to glucocorticoid receptor. Cytosolic fraction from COS-1 cells expressing gluco-corticoid receptor was incubated overnight with 10 nM [3H]-DEX in the presence of DMSO, unlabelled dexamethasone (DEX; 1�M), colchicine(COL; 10�M), berberine (BER; 10�M), sanguinarine (SA; 10�M) and chelerythrine (CHE; 10�M) as described in the Materials and Methodssection. The data are expressed as percent of relative [3H]-DEX binding normalized per total protein in sample. Bars are mean± S.D. from duplicatemeasurements for three independent cytosolic fraction preparations. (*) Value is significantly different from control (p < 0.01).

scriptionally inactive (Figs. 3 and 5) [9,30]. In addition,we observed that the tested QBAs inhibited binding ofradio-labeled dexamethasone to GR. Thus, they are GRligands. However, different compounds that bind to GR

and induce its nuclear translocation (i.e. RU486) havedistinct effects on GR transcriptional activity and the ca-pability of GR to combat with NF-�B [26]. This maybe a cause of the absence of GR transcriptional activ-

Fig. 7. Effect of sanguinarine and chelerythrine on GR and NF-�B signaling in HeLa cells. Sanguinarine and chelerythrine cause an increase innuclear levels of NF-�B (p65) and GR. Pro-inflammatory NF-�B and anti-inflammatory GR can interact by several mechanisms: (i) protein–proteininteraction of unbound forms; (ii) physical interaction in DNA – disabling transcriptional activation of controlled genes; (iii) functional transcriptional

ed, the shock
combat. Albeit the nuclear levels of both, GR and NF-�B are elevatprotein, MW 70/90; I�B�, inhibitor kappa B alpha; Uq, ubiquitin.
y are not transcriptionally active. Abbreviations: Hsp70/90, heat


ity in the cells treated with QBAs. The activation of GRand consequent interaction with NF-�B does not entirelyexplain the anti-inflammatory activity of QBAs, sincethese effects were attained in concentrations of QBAsthat do not occur in vivo. On the other hand, the data ob-tained are mechanistically consistent with those reportedby other authors[30]. The effects of QBAs are ratherpleiothropic, anti-inflammatory activity of sanguinarineand chelerythrine being an example from a spectrum ofeffects exerted by these alkaloids. Numerous studies onthe biological activities of QBAs have been performed.However, the true cellular target for these alkaloids hasnot been found yet. The data presented here support theidea that QBAs have a number of targets within the cell.The overall effect of the alkaloids is the result of theQBAs interactions with individual targets.

In conclusion, structurally diverse natural compoundswith anti-inflammatory activity displayed differential ef-fects on GR and NF-�B activities in HeLa cells: The bi-ological activity of isoquinoline alkaloid berberine doesnot comprise GR and/or NF-�B signaling pathways interms of nuclear translocation and transcriptional activ-ity. Tropolone alkaloid colchicine causes rapid degrada-tion of GR protein in both cytosol and nucleus, havingno effect on NF-�B. Benzophenanthridine alkaloids san-guinarine and chelerythrine act at least partly via GR andNF-�B pathways. Both alkaloids induce nuclear translo-cation of these two factors, however, they remain tran-scriptionally inactive.

A

SM6 ndS 03/0 ub-l e,P an-g

R

hri-

ma-

entmbi-Afr.

it-

002)

[5] A.W. Baird, C.T. Taylor, D.J. Brayden, Non-antibiotic anti-diarrhoeal drugs: factors affecting oral bioavailability of berber-ine and loperamide in intestinal tissue, Adv. Drug Deliv. Rev. 23(1997) 111–120.

[6] K. Fukuda, Y. Hibiya, M. Mutoh, M. Koshiji, S. Akao, H. Fu-jiwara, Inhibition by berberine of cyclooxygenase-2 transcrip-tional activity in human colon cancer cells, J. Ethnopharmacol.66 (1999) 227–233.

[7] C.L. Kuo, C.W. Chi, T.Y. Liu, The anti-inflammatory potential ofberberine in vitro and in vivo, Cancer Lett. 203 (2004) 127–137.

[8] D. Walterova, J. Ulrichova, I. Valka, J. Vicar, C. Vavreck-ova, E. Taborska, R.J. Harjrader, D.L. Meyer, H. Cerna, V.Simanek, Benzo[c]phenanthridine alkaloids sanguinarine andchelerythrine: biological activities and dental care applications,Acta Univ. Palacki. Olomuc. Fac. Med. 139 (1995) 7–16.

[9] M.M. Chaturvedi, A. Kumar, B.G. Darnay, G.B. Chainy, S. Agar-wal, B.B. Aggarwal, Sanguinarine (pseudochelerythrine) is a po-tent inhibitor of NF-kappaB activation, IkappaBalpha phosphory-lation, and degradation, J. Biol. Chem. 272 (1997) 30129–30134.

[10] L.I. McKay, J.A. Cidlowski, Molecular control of im-mune/inflammatory responses: interactions between nuclearfactor-kappa B and steroid receptor-signaling pathways, Endocr.Rev. 20 (1999) 435–459.

[11] K.A. Smoak, J.A. Cidlowski, Mechanisms of glucocorticoid re-ceptor signaling during inflammation, Mech. Ageing Dev. 125(2004) 697–706.

[12] I.M. Adcock, G. Caramori, Cross-talk between pro-inflammatorytranscription factors and glucocorticoids, Immunol. Cell Biol. 79(2001) 376–384.

[13] M. Gottlicher, S. Heck, P. Herrlich, Transcriptional cross-talk, thesecond mode of steroid hormone receptor action, J. Mol. Med. 76(1998) 480–489.

[14] K. De Bosscher, W. Vanden Berghe, G. Haegeman, The interplaybetween the glucocorticoid receptor and nuclear factor-kappaB oractivator protein-1: molecular mechanisms for gene repression,

[ atog-nos-

[ an

[ , Vi-ation41

[ 182

[ sem-680–

[ r ofpro-A 76

[ r, T.tor,

[ sion:Cell

[ uletran-

cknowledgements

This research was supported by grant M198959216 from the Ministry of Education, Youth aports of the Czech Republic, and by grant GACR 34/P074 from the Grant Agency of the Czech Rep

ic. We thank Dr. Jaroslav Vicar (Faculty of Medicinalacky University Olomouc) for preparation of suinarine and chelerythrine.

eferences

[1] E. Ben-Chetrit, M. Levy, Colchicine: 1998 update, Semin. Arttis Rheum. 28 (1998) 48–59.

[2] R. Bibas, N.K. Gaspar, M. Ramos-e-Silva, Colchicine for dertologic diseases, J. Drugs Dermatol. 4 (2005) 196–204.

[3] W.D. Sheng, M.S. Jiddawi, X.Q. Hong, S.M. Abdulla, Treatmof chloroquine-resistant malaria using pyrimethamine in conation with berberine, tetracycline or cotrimoxazole, EastMed. J. 74 (1997) 283–284.

[4] N. Sriwilaijareon, S. Petmitr, A. Mutirangura, M. Ponglikmongkol, P. Wilairat, Stage specificity ofPlasmodium falciparumtelomerase and its inhibition by berberine, Parasitol. Int. 51 (299–103.

Endocr. Rev. 24 (2003) 488–522.15] J. Dostal, E. Taborska, J. Slavik, Preparative column chrom

raphy of quaternary benzophenanthridine alkaloids of Dicratigma lactucoides, Fitoterapia 63 (1992) 67–70.

16] I.W. Southon, J. Buckingham, Dictionary of Alkaloids, Chapmand Hall, London, 1989.

17] J.M. Carcamo, A. Pedraza, O. Borquez-Ojeda, D.W. Goldetamin C suppresses TNF alpha-induced NF kappa B activby inhibiting I kappa B alpha phosphorylation, Biochemistry(2002) 12995–13002.

18] C.M. Stoscheck, Quantitation of protein, Methods Enzymol.(1990) 50–68.

19] U.K. Laemmli, Cleavage of structural proteins during the asbly of the head of bacteriophage T4, Nature 227 (1970)685.

20] H. Towbin, T. Staehelin, J. Gordon, Electrophoretic transfeproteins from polyacrylamide gels to nitrocellulose sheets:cedure and some applications, Proc. Natl. Acad. Sci. US(1979) 4350–4354.

21] A.S. Herr, G.M. Wochnik, M.C. Rosenhagen, F. HolsboeRein, Rifampicin is not an activator of glucocorticoid recepMol. Pharmacol. 57 (2000) 732–737.

22] C. Rosette, M. Karin, Cytoskeletal control of gene expresdepolymerization of microtubules activates NF-kappa B, J.Biol. 128 (1995) 1111–1119.

23] Z. Dvorak, P. Maurel, J. Ulrichova, M. Modriansky, Microtubdisarray in primary cultures of human hepatocytes inhibits


scriptional activity of the glucocorticoid receptor via activation ofc-jun N-terminal kinase, Biomed. Pap. Med. Fac. Univ. Palacki.Olomouc. Czech Republ. 148 (2004) 135–139.

[24] Z. Dvorak, M. Modriansky, J. Ulrichova, P. Maurel, M.J. Vilarem,J.M. Pascussi, Disruption of microtubules leads to glucocorticoidreceptor degradation in HeLa cell line, Cell. Signal. 17 (2005)187–196.

[25] W.B. Pratt, M.D. Galigniana, J.M. Harrell, D.B. DeFranco, Roleof hsp90 and the hsp90-binding immunophilins in signalling pro-tein movement, Cell. Signal. 16 (2004) 857–872.

[26] H. Garside, A. Stevens, S. Farrow, C. Normand, B. Houle, A.Berry, B. Maschera, D. Ray, Glucocorticoid ligands specify dif-ferent interactions with NF-kappaB by allosteric effects on theglucocorticoid receptor DNA binding domain, J. Biol. Chem. 279(2004) 50050–50059.

[27] Z. Dvorak, M. Modriansky, L. Pichard-Garcia, P. Balaguer, M.J.Vilarem, J. Ulrichova, P. Maurel, J.M. Pascussi, Colchicine down-regulates cytochrome P450 2B6, 2C8, 2C9, and 3A4 in human

hepatocytes by affecting their glucocorticoid receptor-mediatedregulation, Mol. Pharmacol. 64 (2003) 160–169.

[28] Z. Dvorak, M. Modriansky, J. Ulrichova, P. Maurel, Speculationson the role of the microtubule network in glucocorticoid receptorsignaling, Cell Biol. Toxicol. 20 (2004) 333–343.

[29] Z. Dvorak, J. Ulrichova, M. Modriansky, Role of microtubulesnetwork in CYP genes expression, Curr. Drug Metab. (2005).

[30] A. Duvoix, S. Delhalle, R. Blasius, M. Schnekenburger, F.Morceau, M. Fougere, E. Henry, M.M. Galteau, M. Dicato, M.Diederich, Effect of chemopreventive agents on glutathione S-transferase P1-1 gene expression mechanisms via activating pro-tein 1 and nuclear factor kappaB inhibition, Biochem. Pharmacol.68 (2004) 1101–1111.

[31] K. De Bosscher, W. Vanden Berghe, L. Vermeulen, S. Plaisance,E. Boone, G. Haegeman, Glucocorticoids repress NF-kappaB-driven genes by disturbing the interaction of p65 with the basaltranscription machinery, irrespective of coactivator levels in thecell, Proc. Natl. Acad. Sci. USA 97 (2000) 3919–3924.

Documents

Differential effects of selected natural compounds with anti-inflammatory activity on the glucocorticoid receptor and NF-κB in HeLa cells