Biochemical Pharmacology 86 (2013) 231–241

Pregnane X receptor dependent up-regulation of CYP2C9 and CYP3A4in tumor cells by antitumor acridine agents, C-1748 and C-1305,selectively diminished under hypoxia§

Magdalena Niemira a, Jarosław Dastych b, Zofia Mazerska a,*a Department of Pharmaceutical Technology and Biochemistry, Chemical Faculty, Gdansk University of Technology, Narutowicza 11/12, 80-233, Gdansk,

Polandb Laboratory of Cellular Immunology, Institute of Medical Biology, Polish Academy of Sciences, Lodowa 106, 93-232, Lodz, Poland

A R T I C L E I N F O

Article history:

Received 16 March 2013

Accepted 8 May 2013

Available online 18 May 2013

Keywords:

Drug–drug interactions

Antitumor acridines

PXR receptor induction

Enzyme induction

CYP3A4

CYP2C9

UGT

P-gp induction

Cell hypoxia

A B S T R A C T

Induction of proteins involved in drug metabolism and in drug delivery has a significant impact on drug–

drug interactions and on the final therapeutic effects. Two antitumor acridine derivatives selected for

present studies, C-1748 (9-(20-hydroxyethylamino)-4-methyl-1-nitroacridine) and C-1305 (5-dimethy-

laminopropylamino-8-hydroxy-triazoloacridinone), expressed high and low susceptibility to metabolic

transformations with liver microsomes, respectively. In the current study, we examined the influence of

these compounds on cytochrome P450 3A4 (CYP3A4) and 2C9 (CYP2C9) enzymatic activity and gene

expression in HepG2 tumor cells. Luminescence and HPLC examination, real-time RT-PCR and western

blot analyses along with transfection of pregnane X receptor (PXR) siRNA and CYP3A4 reporter gene

assays were applied. We found that both compounds strongly induced CYP3A4 and CYP2C9 activity and

expression as well as expression of UGT1A1 and MDR1 in a concentration- and time-dependent manner.

C-1748-mediated CYP3A4 and CYP2C9 mRNA induction equal to rifampicin occurred at extremely low

concentrations (0.001 and 0.01 mM), whereas 10 mM C-1305 induced three-times higher CYP3A4 and

CYP2C9 mRNA levels than rifampicin did. CYP3A4 and CYP2C9 expressions were shown to be PXR-

dependent; however, neither compound influenced PXR expression. Thus, the observed drug-mediated

induction of isoenzymes occurs on a PXR-mediated regulatory level. Furthermore, C-1748 and C-1305

were demonstrated to be selective PXR agonists. These effects are hypoxia-inhibited only in the case of C-

1748, which is sensitive to P450 metabolism. In summary, PXR was found to be a new target of the

studied compounds. Thus, possible combinations of these compounds with other therapeutics might

lead to the PXR-dependent enzyme-mediated drug–drug interactions.

� 2013 Elsevier Inc. All rights reserved.

Contents lists available at SciVerse ScienceDirect

Biochemical Pharmacology

jo u rn al h om epag e: ww w.els evier .c o m/lo cat e/b io c hem p har m

1. Introduction

Drug–drug interactions have a major impact on treatmentoutcomes in cancer patients. These patients are usually treatedwith multiple cytotoxic anticancer drugs, hormonal agents, andpain relievers as well as prophylactic anti-emetics and analgesics.Therefore, they are at high risk for adverse interactions betweenthese drugs [1]. One of the cellular mechanisms underlying drug–drug interactions is the induction of enzymes or efflux transporters

§ Part of this work was previously presented at the conference: Niemira M,

Brillowska-Dabrowska A, Konopa J, Mazerska Z. The antitumor derivatives C-1305

and C-1748, were different in effects on CYP3A4 expression and catalytic activity in

HepG2 cells. Abstarcts 37th FEBS Congress, Gothenburg, 23–26 June, FEBS J, 278,

Suppl. 1, p. 26, 2011.

* Corresponding author. Tel.: +48 58 347 2407; fax: +48 58 347 1516.

E-mail addresses: m.niemira@gmail.com (M. Niemira), jdastych@cbm.pan.pl

(J. Dastych), zofia.mazerska@pg.gda.pl (Z. Mazerska).

0006-2952/$ – see front matter � 2013 Elsevier Inc. All rights reserved.

http://dx.doi.org/10.1016/j.bcp.2013.05.008

involved in the biotransformation and clearance of drugs. Thepregnane X receptor (PXR), which belongs to the nuclear hormonereceptor family (NR), is involved in the induction of cytochromesP450 (P450s), particularly the CYP3A family, phase II enzymes asUDP-glucuronyltransferases (UGTs) and sulfotransferases (SULTs),drug efflux transporters including multidrug resistance proteins(MRPs), breast cancer resistance protein (BCRP) and P-glycoprotein(P-gp, encoded by MDR1) [2]. Several studies have stronglysuggested that PXR also plays an important role in endobioticmetabolism in humans, mice and rats [3,4]. Although PXR is mainlyexpressed in liver and intestinal tissues, its expression has alsobeen detected in breast, prostate and gastrointestinal cancers [5].

The influence of PXR induction on tumor cell propagation isnot unequivocal [6]. PXR activation has been found to inhibitbreast cancer cell proliferation by inducing apoptosis [7].However, other reports revealed that high levels of PXR in breastand prostate cancer stimulated the expression of MDR1 andincreased cancer cell resistance to chemotherapeutics [8],

M. Niemira et al. / Biochemical Pharmacology 86 (2013) 231–241232

Furthermore, chemotherapeutic agents can activate human PXR,emphasizing the importance of PXR in the tumor response tochemotherapy. Both, the native and metabolized form of theantitumor drug tamoxifen, have been shown to activate PXR[9,10]. A similar activation was demonstrated with the anti-mitotic agent paclitaxel, whereas its semi-synthetic analogdocetaxel was not a PXR activator [11,12]. An example of aPXR mediated drug–drug interaction increasing compoundcytotoxicity is the interaction of cyclophosphamide with PXRagonists phenobarbital and rifampicin. These compounds in-creased the rate of cyclophosphamide transformation by CYP3A4and CYP2B6 into highly cytotoxic 4-hydroxymetabolite [13].Furthermore, it has been shown that either, direct interaction oferlotinib or flutamide anticancer drugs with PXR, or theinteraction of anticancer compounds with PXR agonists is ableto up-regulate PXR-mediated CYP3A4 and MDR1 expression. As aresult the activity of anticancer drug is diminished [14,15]. Inlight of such PXR-mediated drug resistance, the possibleadvantages of down regulation of PXR for cancer therapy havebeen considered [16]. However, PXR activation might also inducenecessary metabolic activation, when a prodrug is delivered intothe organism.

In this study, we investigated two acridine antitumor agents, C-1748 and C-1305 (Fig. 1), which are characterized by multidirec-tional mechanisms of antitumor activity. These compounds aredissimilar in structure, P450-mediated metabolism and suscepti-bility to UGT-glucuronidation. The first compound, 9-(20-hydro-xyethylamino)-4-methyl-1-nitroacridine (C-1748), exhibits strongcytotoxic activity against colon cancer cell lines [17], highantitumor activity against several prostate and colon carcinomaxenografts in nude mice [18], and slight myelosuppressiveproperties [19]. Based on this information, C-1748 has beenselected for preclinical studies. We showed earlier that C-1748 wasmetabolized with rat liver microsomes and with oxidoreductaseenzymes DT-diaphorase and xanthine oxidase. A crucial role ofhypoxic conditions in the metabolism of this agent has also beenobserved [20]. The second compound, 5-dimethylaminopropyla-mino-8-hydroxytriazoloacridinone [21] (C-1305), expressed highcytotoxic activity toward tumor cell lines [22] lines and activityagainst several experimental tumors and has also been selected forpreclinical trials. We have shown that the involvement of P450s inthe microsomal metabolism of C-1305 is negligible. However, C-1305 was found to be an effective inhibitor of CYP1A2, slight one ofCYP3A4 but was not effective toward CYP2C9 activity innoncellular systems [23].

In the present work the effects of these compounds on CYP3A4and CYP2C9 enzymatic activity and on CYP3A4, CYP2C9, UGT1A1and MDR1 gene expression in tumor cells were studied andcompared with standard inducer rifampicin. Hepatic tumor cells,HepG2, were selected, because, they do not have the drawback ofhepatocytes, as high interindividual donor variation and rapiddecline of metabolic enzyme expression [24,25]. Additionally,tumor cells are appropriate for studies on the effects of antitumoragents. The role of PXR in gene expression of listed above proteinswas elucidated applying reporter gene assay with nhrtox-hepg2cell line. We also present here the data on the influence of hypoxiaon the observed drug-mediated modulation of P450, UGT, MDR1and PXR gene expression.

2. Materials and methods

2.1. Chemicals and antibodies

5-dimethylaminopropylamino-8-hydroxytriazoloacridinone,C-1305 [21], was synthesized as dihydrochloride and 9-(20-hydroxyethylamino)-4-methylamino)-4-methyl-1-nitroacridine,

C-1748 [26], was synthesized as a monochloride in the Departmentof Pharmaceutical Technology and Biochemistry at the TechnicalUniversity of Gdansk. Both compounds were of more than 98% pureas determined by HPLC and NMR. Testosterone, rifampicin,ketoconazole, aflatoxin, MTT and antibodies were purchased fromSigma–Aldrich (St. Louis, MO, USA). Antibodies: (1) mousepolyclonal anti-CYP3A4 antibody (cat. no. SAB1400065, dilution1:500), (2) rabbit polyclonal anti-CYP2C9 antibody (cat. no.AV41809, 1:4000), (3) rabbit polyclonal anti-UGT1A1 antibody(cat. no. SAB2102640, dilution 1:1000), (4) mouse monoclonalanti-P-gp (clone 1F11, cat. no. WH0005243M1, dilution 1:1000),(5) goat polyclonal anti-PXR/NR1I2 (cat. no. SAB2500844, dilution1;500) (6) mouse monoclonal anti-b-actin antibody (cat. no.A5441, clone AC-15, dilution 1:6000) and (7) anti-mouse IgG(whole molecule) – peroxidase antibody produced in goat (cat. no.A4416, dilution 1:10,000), (8) anti-rabbit IgG (whole molecule) –peroxidase antibody produced in goat (cat. no. A0545, dilution –1:10,000), (9) anti-goat IgG (whole molecule) – peroxidaseantibody produced in rabbit (cat. no. A5420, dilution –1:10,000). Luciferin was purchased from BD Biosciences (San Jose,USA), and LipofectamineTMRNAiMAX was obtained from Invitro-gen (Carlsbad, CA, USA). The following chemicals were obtainedfrom Roche Diagnostics (Basel, Switzerland): the High FidelitycDNA Synthesis Kit, the Transcription High Fidelity cDNA SynthesisKit and FastStart DNA Master SYBR Green I.

2.2. Cell culture and growth inhibition assay

The human hepatocarcinoma HepG2 cell line was kindlyprovided by Adam Holownia from the Department of ClinicalPharmacology at the Medical University of Bialystok (Poland). Thenhrtox-hepg2 reporter cell line contained the luciferase ORF underthe control of an artificial promoter with 9 repeats of the PXRbinding motif (CYP3A4-ER6) was kindly provided by MarcinRatajewski from the Laboratory of Transcriptional Regulation,Institute of Medicinal Biology of Polish Academy of Sciences (Lodz,Poland). The HepG2 cell line was maintained under standardconditions in MEM (Sigma–Aldrich, St. Louis, MO, USA) containing10% fetal bovine serum and antibiotics (100 mg/ml streptomycin,100 U/ml penicillin). The nhrtox-hepg2 cell line was maintainedunder standard conditions in high-glucose DMEM medium(Sigma–Aldrich, St. Louis, MO, USA) containing 10% fetal bovineserum and hygromycin-B at a concentration of 100 mg/ml. All ofthe experiments were performed with cells in the exponentialphase growth.

To estimate cell viability, the MTT assay was conducted. Cellswere seeded into 24-well plates and after 24 h were treated withincreasing concentration of the investigated compounds. Afterthe treatment, MTT solution was added and the plates wereincubated for 4 h at 37 8C. The purple formazan crystals weredissolved in DMSO per well. After 30 min, the plates were read onmicroplate reader (Victor 3V, Perkin-Elmer, Walthman, MA, USA)at 540 nm. The concentration of the drug required to inhibit cellsgrowth by 50% compared with untreated control cells wasdetermined from the curves plotting survival as a function ofdose.

The cells were maintained at 37 8C under a 5% CO2 atmosphere.For hypoxia exposure, the cells were placed into a hypoxicchamber that was thoroughly flushed for 20 min with a mixture of1% O2, 5% CO2 and 94% N2. After flushing, the cells were tightlysealed and placed at 37 8C.

2.3. Cell treatment

Both cell lines were treated with pre-defined concentrationsof compounds as follows: C-1748 (0.0001, 0.001, 0.01 and 0.1 mM),

M. Niemira et al. / Biochemical Pharmacology 86 (2013) 231–241 233

C-1305 (0.1, 1, 5 and 10 mM), rifampicin (10 mM) and aflatoxin B1(10 mM), for 3- to 48-h. All of these compounds were prepared as adimethylsulfoxide (DMSO) stock solutions and were directly addedto the cultures. The final concentration of DMSO was 0.1%.

2.4. Influence of C-1748 and C-1305 on CYP3A4 and CYP2C9

enzymatic activity in HepG2 cells

2.4.1. Luminescence analysis

The activities of the CYP2C9 and CYP3A4 enzymes in HepG2cells were measured using the P450-GloTM CYP3A4 and CYP2C9assay kits (Promega GMBH, Fitchburg, WI, USA). This assayemploys luciferin-IPA and luciferin-H, which in turn reacts withluciferase which are specifically converted by P450 CYP3A4 andCYP2C9, respectively. The quantity of light produced is directlyproportional to P450 CYP3A4 and CYP2C9 activities. Luciferin IPAmetabolism was assessed by measuring the luciferase activitydirectly in cell culture, according to the manufacturer’s instruc-tions. HepG2 cells (8 � 103) were seeded in 96-well microcultureplates and treated with various C-1748 and C-1305 concentra-tions for 24 h. Next, 3 micromoles of luciferin IPA was added to thefresh culture medium, and the cells were incubated at 37 8C for60 min. Then, 50 ml of the luciferin detection reagent was added to50 ml of cell culture medium containing luciferin IPA. Theluminescence was recorded 30 min later, after cell lysis, using aWallac 1450 MicroBeta luminometer (Perkin-Elmer, Walthman,MA, USA). All of the experiments were repeated at least 3 times.The results were expressed as the relative activities of CYP3A4 andCYP2C9.

2.4.2. Chromatographic analysis

The activity of CYP3A4 was measured by the formation of 6b-hydroxytestosterone in the HepG2 cells. After 24 h of exposure tochemicals, the culture medium was replaced with fresh MEMmedium containing 50 mM testosterone. Aliquots of the mediawere removed after 24 h of incubation at 37 8C. Testosterone andthe formed metabolites were extracted with acetonitrile (1:1, v/v)and were analyzed by high-performance liquid chromatographywith a reversed-phase 5 mm Suplex pKb-100 analytical column(0.46 cm � 25 cm, C18, Supelco, Bellefonte, PA, USA) with WatersBreeze System (Milford, MA, USA). The HPLC analyses werecarried out at a flow rate of 1 mL � min�1 with the followingeluent system: a linear gradient from 15 to 80% methanol inammonium formate (0.05 M, pH 3.4) for 25 min, followed by alinear gradient from 80 to 100% methanol in ammonium formatefor 3 min.

The eluent was detected by its absorbance at 245 nm.All of the experiments were performed in triplicate, and the

A.

B.

N

CH3

NH NO2

OH

N

O

OH

N N

NH N

Fig. 1. Chemical structures of (A) C-1748 and (B) C-1305.

relative amounts of CYP3A4 activity are presented in Fig. 2Cand D.

2.5. Western blot analysis

Microsomal fractions were isolated by sonication of the HepG2cells. The homogenates were centrifuged at 12,000 � g. Theresulting supernatants were then centrifuged at 100,000 � g toobtain the microsomal pellet. The contents of the microsomalfractions were measured by the Bradford protein assay. Theproteins were separated by SDS-polyacrylamide gel electrophore-sis and then transferred onto a PVDF membrane. The membranewas probed with the corresponding, described above primary anti-CYP3A4, anti-CYP2C9, anti-UGT1A1, anti-P-gp, anti-PXR/NR/1/2 oranti-actin antibodies. The membranes were then incubated withthe appropriate secondary antibodies and visualized with theenhanced chemiluminescence kit, SuperSignal West PICO Chemi-luminescent Substrate (Fisher Scientific, Pitsburgh, PA, USA). Toquantify the bands obtained via Western blot analysis, we appliedImageJ software. The area under curve (AUC) of the specific signalwas corrected for the AUC of the loading control (beta-actin). Thevalue for the ‘control’ condition was set as 1 and other conditionswere recalculated correspondingly to allow ratio comparisons.

2.6. Real-time RT-PCR

Total RNA was isolated from the cells using a High Pure IsolationKit (Roche, Basel, Switzerland) and reverse-transcribed with aTranscription High Fidelity cDNA Synthesis Kit (Roche, Basel,Switzerland) using an anchored oligo-dT18 primer. The level ofcDNA was measured by real-time RT-PCR amplification performedon a LightCycler 1.5 from Roche, using FastStart DNA Master SYBRGreen I (Roche) to detect the PCR products. The cycling conditionswere as follows: 95 8C for 10 min followed by 45 cycles at 95 8C for10 s, 59 8C for 15 s and 72 8C for 20 s. All of the primers wereentered into the NCBI Blast program to ensure specificity. Thesequence of these primers were as follows: CYP3A4 forward:50AAGAAACTGAGTCCCACAAAGC30, CYP3A4 reverse: 50ACACTGCT-ACACTGCTCGTGGTTTCACA30; CYP2C9 forward: 50GGACAGAGA-CGACAAGCACA30, CYP2C9 reverse: 50AATGGACATGAACAACCCTCA30;UGT1A1 forward: 50TGACGCCTCGTTGTACATCAG30, UGT1A1 reverse:50CCTCCCTTTGGAATGGCAC30; MDR1 forward: 50AGGAAGACATGA-CCAGGTATG30, MDR1 reverse: 50CCAACATCGTGCACATCAAAC30; PXRforward: 50CCAGGACATACACCCCTTTG30, PXR reverse: 50CTACC-TGTGATGCCGAACAA30; GAPDH forward: 50TGCACCACCAACTGC-TTAGC30, GAPDH reverse: 50GGCATGGACTGTGGTCATGAG30. Thelevels of CYP3A4, CYP2C9, UGT1A1, MDR1 and PXR mRNAs werenormalized to the GAPDH values to control for the quality of mRNA.The quantitative PCR values were determined for each of the mRNAslevels using the comparative CT (DDCT) method.

2.7. Transfection of siRNA in the HepG2 cell line

HepG2 cells were plated into 24-well plates in MEM at adensity of 2.5 � 104 cells per well. LipofectamineTMRNAiMAX(Invitrogen, Carlsbad, CA, USA) and Opti-MEM1 (GIBCO LifeTechnology, Paisley, Scotland) were used for transfecting 10 nMquantities of PXR siRNA (Applied Biosystems, Carlsbad, CA, USA)according to the manufacturer’s protocol for LipofectamineTMR-NAiMAX. The transfections in HepG2 cells were allowed toproceed for 24-h, after which fresh medium (containing C-1748and C-1305 derivatives) was added to the transfected cells asdescribed above. At the end of the drug exposure, total RNA wasisolated from the cells. CYP3A4, CYP2C9 and PXR-specific mRNAlevels were then determined using the real-time PCR method asdescribed above.

0

1

2

3

4

5

6

7

8

9

10

Rif 0,000 1 0,001 0,01 0,1

Rela

tive

acti

vit

y o

f C

YP

3A

4 a

nd

CY

P2C

9

Concen tratio n of C-1748 [μM]

CYP3 A4

CYP2 C9

A.

C.

0

100

200

300

400

500

600

700

800

900

1000

0,000 1 0,001 0,01 0,1

Ac

tivit

y o

fC

YP

3A

4[%

]

Concen tratio n of C-1748 [μM]

B.

0

1

2

3

4

5

6

7

8

9

10

Rif 0,1 1 5 10

Re

lati

ve

ac

tivit

y o

f C

YP

3A

4 a

nd

CY

P2C

9

Concen tration of C-1305 [μM]

CYP3 A4

CYP2 C9

D.

0

100

200

300

400

500

600

700

800

900

1000

0,1 1 5 10

Ac

tivit

y o

f C

YP

3A

4 [

%]

Concen tratio n of C-1305 [μM]

******

**

**** **

**

** **

******

****

**

***

C C

C C

Fig. 2. Effect of C-1748 and C-1305 on P450 3A4 and P450 2C9 activities. (A and B) P450 3A4 and P450 2C9 activities were measured using P450-GloTM 3A4 and P450-GloTM

2C9 (Promega), respectively. These assays employ luciferin IPA and luciferin H, which are specifically converted by P450 3A4 and P450 2C9 to luciferin, which, in turn, reacts

with luciferase; the quantity of light produced is directly proportional to P450 3A4 and P450 2C9 activities. Rifampicin (10 mM) was used as a positive control. (C and D) The

rate of 6b-hydroxytestosterone formation by untreated and drug-treated cells was used as a marker for P450 3A4 activity. Following the drug treatment, HepG2 cells were

incubated with medium containing testosterone (50 mM) for 24-h. Testosterone and the formed metabolites were extracted with acetonitrile (1:1, v/v) and were measured by

high-performance liquid chromatography. The results are expressed as the mean � SEM (n = 3). Significantly different from the control at: **p < 0.01; ***p < 0.001.

M. Niemira et al. / Biochemical Pharmacology 86 (2013) 231–241234

2.8. CYP3A4 reporter gene assay

The nhrtox-hepg2 cells were seeded onto 96-well plates at adensity of 8 � 103 cells per well. The next day, the cells wereincubated with C-1748, C-1305, rifampicin and aflatoxin B1 asdescribed above. Following incubation, the cells were harvestedand lysed. The luciferase activity in the cell lysates was determinedusing an EnVision luminometer plate reader (Perkin-Elmer,Walthman, MA, USA) with commercial luciferase substrate (BDBiosciences, San Jose, CA, USA).

2.9. Data analysis

All of the data reported represent the mean � the standard errormean. The statistical significance of the differences between thevariables was determined by Student’s unpaired t test. Analysis ofvariance (one-way ANOVA) was carried out using GraphPad Software,Inc. (San Diego, CA) p < 0.05 was used as the cutoff for statisticalsignificance.

3. Results

3.1. Cytotoxic effects of C-1748 and C-1305 toward HepG2 cells and

nhrtox-hepg2 cell lines

The studies presented here were preceded by tests of cytotoxiceffects induced by both compounds in HepG2 cell line and in thevalidated nhrtox-hepg2 reporter cell line used to test PXR

dependency of CYP3A4 expression. EC50 values determined forC-1748 were equal to 0.54 � 0.08 mM and 0.47 � 0.07 mM forHepG2 and nhrtox-hepg2 cell line, respectively. Therefore, they werenot different significantly. Equitoxic concentrations for C-1305 werealso similar for both cell lines, 18.9 � 0.8 mM for HepG2 and17.3 � 0.9 mM for nhrtox-hepg2. The above results demonstratedthat C-1748 expressed 40 times higher cytotoxicity than C-1305 inboth cell lines.

The present study intended to reflect the functional but notcytotoxic effects of the antitumor agents. Therefore, we selectedconcentrations lower than IC50 for both compounds. The concen-tration ranges 0.0001–0.1 mM for C-1748 and 0.1–10 mM for C-1305 were shown to be markedly lower than the EC30 effectsdetermined by the MTT assay for both compounds. Therefore,these concentrations were applied in the studies presented below.

3.2. Modulation of cytochrome P450 3A4 and 2C9 activities by C-1748

and C-1305 in HepG2 tumor cells

The influence of the studied compounds on the enzymaticactivities of the CYP2 and CYP3 families was investigated bydetermining the relative CYP2C9 and CYP3A4 activities in humanHepG2 cells using two methods. In the first method, a specificsubstrates of CYP3A4 and CYP2C9 were converted to luciferine, andthe level of luminescence was a measure of enzymatic activity inFig. 2A and B. The metabolic rate of testosterone, a standardsubstrate of CYP3A4, was analyzed in the second method.Testosterone was incubated with HepG2 cells in the presence or

M. Niemira et al. / Biochemical Pharmacology 86 (2013) 231–241 235

absence (control incubation) of the drugs. The formation of areaction product (6-b-hydroxytestosterone) was followed byHPLC, and the relative CYP3A4 activity are presented in Fig. 2Cand D for C-1748 and C-1305, respectively.

In the case of C-1748, the relative activities of CYP2C9 andCYP3A4 determined by luminescence method (Fig. 2A) increasedat very low drug concentrations 0.001 and 0.01 mM and thendecreased at a concentration of 0.1 mM. The increases in theenzymatic activities were statistically significant compared to thecontrol incubation and were comparable to that demonstrated fora standard inductor of CYP3A4 activity, rifampicin. Chro-matographic analysis confirmed that C-1748 induced enzymeactivities at 0.001 and 0.01 mM up to 400% of the control value(Fig. 2C). At a concentration of 0.1 mM, enzyme activity decreasedas it was shown by the luminescence method. These resultsdemonstrated that C-1748 was a strong inducer of CYP3A4 andCYP2C9 enzymatic activity to the level comparable to thatobserved with the prototypical inducer, rifampicin at 10 mM.What is more, this induction was observed at very lowconcentrations, in comparison to rifampicin.

The results of an identical procedure performed for C-1305 areshown in Fig. 2B and D. The luminescence method, as well as theHPLC analysis of the enzyme activities, revealed that thiscompound induced CYP3A4 and CYP2C9 activities 10 times higherthan the control value. In other words, these activities were 3 timeshigher than the level induced by rifampicin under identicalconcentration (10 mM).

0

1

2

3

4

5

6

7

8

9

10

K Rif 10 µM 0,000 1 0,001 0,01 0,1

CY

P m

RN

A /

GA

PD

H m

RN

A

Concen tratio n of C-1748 [μM]

CYP3 A4

CYP2 C9

C.

A.

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

0 10 20 30 40 50

CY

P m

RN

A /

GA

PD

H m

RN

A

Incuba tio n ti me [h]

CYP3 A4

CYP2C9

Rif / CYP3 A4

****

**** ****

*

**

****

*

*

*

C

Fig. 3. Effect of C-1748 and C-1305 on P450 3A4 and P450 2C9 mRNA in HepG2 cells. (A a

compounds for 24-h. (C and D) HepG2 cells were incubated with C-1748 (0.001 mM) an

positive control. The relative amounts of P450 3A4 and P450 2C9 mRNAs were measured b

mRNA. The quantitative PCR values were determined for each mRNA level using the co

Significantly different from the control at: **p < 0.01; ***p < 0.001.

3.3. Influence of the antitumor compounds C-1748 and C-1305 on

P450 3A4 and 2C9 gene expression and protein levels

Because we demonstrated above that C-1748 and C-1305specifically induced CYP3A4 and CYP2C9 enzymatic activities intumor cells, the question whether this higher enzymatic activitymight result from the compound-mediated induction of CYP3A4and CYP2C9 gene expression was stated here. The results of thereal-time RT-PCR presented in Fig. 3A shows that incubation ofHepG2 cells with 0.001 and 0.01 mM C-1748 resulted in asignificant increase (3.0, and 3.8-fold, respectively) in the levelsof CYP3A4 and CYP2C9 mRNA, while the CYP3A4 inducerrifampicin (Rif), showed a 3.5-fold increase at 10 mM. At 0.1 mMC-1748, the CYP mRNA levels were reduced to that of the control.Cell incubation with 5 and 10 mM C-1305 (Fig. 3B) induced aprominent increase in mRNA levels, up to 9-fold in relation tountreated cells and 3-fold in relation to rifampicin treated cells.

The profiles of time-dependent modulation of CYP3A4 andCYP2C9 mRNA were different for each compound. Fig. 3C demon-strates that HepG2 incubation with C-1748 induced 3.5-fold CYP3A4and CYP2C9 mRNA levels up to 24 h, and then the levels of bothdecreased. In contrast, Fig. 3D presents that the relative levels ofmRNA increased up to 24 h of drug treatment and were stable at thelevels of 9-fold for C-1305 and 3-fold for rifampicin. Therefore, C-1748 induced the time-profile different from that of rifampicin,whereas time-profile of C-1305 were close to that of rifampicin butreach much higher level of the transcription.

0

1

2

3

4

5

6

7

8

9

10

K Rif 10 µM 0,1 1 5 10

CY

P m

RN

A /

GA

PD

H m

RN

A

Concen tratio n of C-1305 [μM]

CYP3 A4

CYP2 C9

0

1

2

3

4

5

6

7

8

9

10

0 10 20 30 40 50

CY

P m

RN

A /

GA

PD

H m

RN

A

Incuba tio n ti me [h]

CYP3 A4

CYP2C9

RIF / CYP3 A4

****

****

******

**

*** ***

**

****

B.

D.

C

nd B) HepG2 cells were treated with increasing concentrations of the experimental

d C-1305 (10 mM) for 3-, 6-, 12-, 24-, and 48-h. Rifampicin (10 mM) was used as a

y real-time RT-PCR and normalized to the GAPDH values to control for the quality of

mparative CT (DDCT) method. The results are expressed as the mean � SEM (n = 2).

Fig. 4. Effect of C-1748 (A) and C-1305 (B) on P450 3A4 and P450 2C9 protein levels in HepG2 cells. The cells were treated with increasing concentrations of the experimental

compounds. After 24-h, the microsomal fraction was separated by SDS-PAGE, electrotransferred to nitrocellulose membrane, and subjected to immunoblotting with anti-

CYP3A4 and anti-CYP2C9 antibodies. b-actin was used as an internal control for protein loading and was detected with an antibody to mouse actin. The WB images are

representative of three separate blots. The results are expressed as the mean � SEM (n = 3). Significantly different from the control at: *p < 0.05; **p < 0.01; ***p < 0.001.

M. Niemira et al. / Biochemical Pharmacology 86 (2013) 231–241236

The concentration-dependent influence of the compoundson the CYP3A4 and CYP2C9 protein levels are presented in Fig. 4,as a result of CYP immunostaining assay. The levels ofthese proteins in Hep G2 cells treated for 24-h with 0.0001,0.001 and 0.01 mM C-1748 increased up to 2-times highercompared to the control and decreased to the same levels asthe control reaction at a concentration of 0.1 mM. C-1305induced a regular increase in the protein levels from 0.1 to10 mM and achieved 6-fold higher protein levels than thecontrol probe.

In conclusion, C-1748 and C-1305 strongly induced CYP3A4 andCYP2C9 gene expression and protein levels in HepG2 cells, but in avariable manner.

3.4. Impact of hypoxia on induced by C-1748 and C-1305 P450 3A4

and 2C9 gene expression and protein levels

Compared with metabolism under normoxia, the cellularmetabolism of C-1748 under hypoxia led to higher levels ofselected metabolites. The direct involvement of cytochrome

A.

0

1

2

3

4

5

6

7

8

9

10

K 0,000 1 0,001 0,01 0,1 K 0,000 1 0,001 0,01 0,1

CY

P m

RN

A /

GA

PD

H m

RN

A

Concen tratio n of C-1748 [μM]

CYP3 A4

CYP2 C9

NORMOXIA HYPOXIA

**** ** **

C C

Fig. 5. P450 3A4 and P450 2C9 gene expression in HepG2 cells treated with increasing co

The cells were incubated with these two compounds for 24-h at 37 8C with a 5% CO2 atmo

5% CO2 and 94% N2. The effects of the compounds on the mRNA levels of the tested genes

compared to the control vehicle (DMSO)-treated cells. The results are expressed as the m

P450-reductase (POR) in the metabolism of C-1748 was alsodemonstrated [20]. The contribution of POR in C-1748 metabo-lism was also shown by experiments where POR was over-expressed in the HepG2 cell line. The metabolism of C-1305 inHepG2 cells was not sensitive to hypoxia nor to POR over-expression (data not shown).

In light of the above results, the expression of CYP3A4 andCYP2C9 mRNAs was determined by a real-time RT-PCR assay afterincubation with both compounds under hypoxia. The results underhypoxia were compared to the results under normoxia. The resultspresented in Fig. 5A clearly demonstrate that hypoxia significantlyreduced the induction effects observed for C-1748 compared tonormoxia. However, C-1305-mediated induction under hypoxiawas preserved on a comparable level to that under normoxia, whatis demonstrated in Fig. 5B. In contrast to normoxia, hypoxia-dependent effects of C-1748 were weakly dependent on time andconcentration (Fig. 6A). In addition, under hypoxia, the inductioneffects of C-1748 on CYP3A4 and CYP2C9 protein levels werediminished significantly in comparison to normoxia, what isshown in Fig. 6B.

B.

0

1

2

3

4

5

6

7

8

9

10

K 0,1 1 5 10 K 0,1 1 5 10

CY

P m

RN

A /

GA

PD

H m

RN

A

Concen tratio n of C-1305 [μM]

CYP3 A4

CYP2 C9

NORMOXIA HYP OXIA

****

******

****

******

C C

ncentrations of (A) C-1748 and (B) C-1305 under normoxic and hypoxic conditions.

sphere for normoxia. For hypoxia, the cells were incubated with a mixture of 1% O2,

were normalized to the GAPDH mRNA and are presented as the relative mRNA levels

ean � SEM (n = 3). Significantly different from the control at: **p < 0.01; ***p < 0.001.

Fig. 6. Effects of C-1748 on P450 3A4 and P450 2C9 gene expression and protein levels under normoxic and hypoxic conditions (A). Real-time RT-PCR analysis of the total RNA

isolated from HepG2 cells. The HepG2 cells were treated with C-1748 (0.001 mM) for 3-, 6-, 12-, 24-, or 48-h. The effects of the compounds on the mRNA levels of the tested

genes were normalized to the GAPDH mRNA and are presented as the relative mRNA expression compared to control vehicle (DMSO)-treated cells. (B) P450 3A4 and P450 2C9

proteins were detected by western blotting in HepG2 microsomal fractions after 24-h of treatment with increasing concentrations of C-1748. Equal amounts of the

microsomal fraction (30 mg) were loaded in the gel. b-actin was used as a loading control. The histogram reflects the quantitative densitometric analysis of the average fold

change in P450 3A4 and P450 2C9 bands normalized to the control (DMSO) from three separate analyses, respectively. The results are expressed as the mean � SEM (n = 3).

Significantly different from the control at *p < 0.05; **p < 0.01.

M. Niemira et al. / Biochemical Pharmacology 86 (2013) 231–241 237

3.5. PXR-mediated P450 3A4 induction of transcription in nhrtox-

hepg2 cells

We examined whether the increasing CYP3A4 gene expressioninduced in HepG2 cells by the studied compounds is PXR-dependent. The validated nhrtox-hepg2 reporter cell line, contain-ing the luciferase ORF under the control of an artificial promoterwith 9 repeats of the PXR binding motif (CYP3A4-ER6), [27] wasused to test PXR-dependency. Fig. 7A shows the relative luciferaseactivity, which is a measure of the PXR – mediated induction ofreporter gene transcription. Incubation with 0.001 and 0.01 mM C-1748 resulted in significant increase (2.0 and 2.3-fold, respective-ly) in luciferase expression in the nhrtox-hepg2 cells. At 0.0001 and0.1 mM C-1748, the luciferase activity was comparable to that ofthe untreated cells. In contrast, C-1305 induced a systematicincrease in the relative luciferase activity (from 0.01 to 10 mM

A.

0

1

2

3

4

5

6

7

8

9

10

Re

lati

ve

lu

cif

era

se

ac

tivit

y

C-1748 [μM] C-1305 [μM]

*

***

***

***

C

Fig. 7. Identification of C-1748 and C-1305 as potent activators of PXR in nhrtox-hepg2 c

reporter gene expression in the nhrtox-hepg2 cells for 24-h incubation. (B) Time course

expression in the nhrtox-hepg2 cells. Rifampicin (10 mM) was used as a positive control

using a luminometer with luciferase substrate. The results are expressed as the me

***p < 0.001.

concentrations) leading to 9-fold increase in relation to untreatedcells and nearly 4 and 2 times higher than those for rifampicin andaflatoxin, respectively. Rifampicin and aflatoxin were applied asthe reference strong inducers of PXR-dependent CYP3A4 expres-sion being the representatives of the drugs and environmentalpollutants, respectively. The graph of time dependence (Fig. 7B)showed that the profile of C-1748 was close to that of rifampicin upto 24 h, and a prolonged incubation time reduced PXR activation tothe untreated level. The C-1305-mediated luciferase activityreached the maximum value after 24 h (a 9-fold increase) andthen plateaued up to 48 h. Therefore, this profile strongly differsfrom the profile of C-1748 and rifampicin.

To confirm the role of PXR in the induction of CYP3A4 geneexpression, we silenced the PXR gene. Knock-down of PXR mRNA(80%) in HepG2 cells transfected with siRNA directed againsthuman PXR (insert of Fig. 8) resulted in a substantial decrease of

B.

0

1

2

3

4

5

6

7

8

9

10

0 10 20 30 40 50

Re

lati

ve

lu

cif

era

se

ac

tivit

y

Incuba tio n ti me [h ]

C-1748

C-1305

Rif

**

*** ***

* *

ells. (A) Effect of increasing concentrations of C-1748 and C-1305 on PXR-mediated

of C-1748 (0.001 mM) and C-1305 (10 mM) effects on PXR-mediated reporter gene

. After the treatment, the cells were lysed and the luciferase activity was measured

an � SEM (n = 3). *Significantly different from the control at: *p < 0.05; **p < 0.01;

0

1

2

3

4

5

6

7

8

9

10

K Rif C-174 8 C-130 5 K Rif C-174 8 C-130 5

CY

P m

RN

A /

GA

PD

H m

RN

A

CYP3 A4

CYP2C9

0

0.2

0.4

0.6

0.8

1

1.2

K siRNA PXR

PX

R m

RN

A/G

AP

DH

m

RN

A

** **** **

******

C C

control siRN A PXR siRN A

C

Fig. 8. Effect of PXR knock-down on C-1748-mediated and C-1305-mediated P450

3A4 and P450 2C9 induction. PXR mRNA expression in HepG2 cells was detected by

real-time RT-PCR using the total RNA from HepG2 cells transfected either with

10 nM control siRNA or 10 nM PXR siRNA (insert). The extent of P450 3A4/2C9

mRNA induction after the HepG2 cells not transfected (control siRNA) or

transfected (PXR siRNA) were treated for 24-h with C-1748 (0.001 mM) and C-

1305 (10 mM). Rifampicin (10 mM) was used as a positive control. The relative

amounts of P450 3A4, and P450 2C9 mRNAs were measured by real-time RT-PCR

and normalized to the GAPDH values to control for mRNA quality. The quantitative

PCR values were determined for each mRNA level using the comparative CT (DDCT)

method. The results are expressed as the mean � SEM (n = 2). Significantly different

from the control at **p < 0.01; ***p < 0.001.

M. Niemira et al. / Biochemical Pharmacology 86 (2013) 231–241238

CYP3A4 and CYP2C9 gene expression after drug treatment, which isshown in Fig. 8. This decrease was independent of the differencesbetween the C-1748 and C-1305 induction profiles, describedabove (Figs. 3 and 4). Therefore, the absence of PXR geneexpression abolishes any effects of the studied compounds onCYP3A4 expression.

3.6. Effects of C-1748 and C-1305 on UGT1, MDR1 and PXR gene

expression

Considering that PXR activation leads to up-regulation of notonly cytochrome P450 but also of UGT [28] and other phase I and

Fig. 9. Expression of UGT1A1, MDR1 (P-gp) and PXR genes and protein levels in HepG2 cell

The relative amounts of UGT1A1, MDR1 and PXR mRNAs were measured by real-time

quantitative PCR values were determined for each mRNA level using the comparative CT

and PXR were performed using the total proteins and antibodies as described in the mate

The results are expressed as the mean � SEM (n = 3). Significantly different from the con

phase II enzymes [29] as well as drug transporters [30], weevaluated the impact of C-1748 and C-1305 on the gene expressionof these proteins. We also tested whether the studied compoundsinduce PXR gene expression. The latter experiment is intended toexplain whether the impact of the studied compounds on PXRaction depends only on PXR function mediated by an agonist orgene expression up-regulation of PXR should be also considered inthis respect. The results of the relative real-time RT-PCR analysis(Fig. 9A) demonstrated that incubation of both compounds atconcentrations identical to those applied above (in the range of0.0001–0.1 mM for C-1748 and 0.1–10 mM for C-1305) with HepG2cells resulted in profile of UGT1A1 and MDR1 mRNA levels similarto that of CYP3A4 and CYP2C9 presented in Fig. 3. Therefore, bothcompounds were shown to induce significantly UGT1A1 and MDR1gene expression. C-1748 and C-1305 also affected UGT1A1 proteinlevels (Fig. 9B) and the effect of this induction was 2-fold and 5-foldhigher in relation to control in the case of C-1748 and C-1305,respectively. However, the increase of P-gp protein level wassignificant only for C-1305 not for C-1748. What is crucial, Fig. 9Aand B also demonstrated that neither PXR gene expression nor PXRprotein levels, respectively, in the HepG2 cells were influenced byboth compounds.

3.7. Modulation of acridine-mediated PXR-dependent induction of

CYP3A4 gene expression by rifampicin

To characterize the type of interactions between drugs and PXR,studied here, we elucidate the impact of a well-known PXRinductor, rifampicine, on PXR-dependent CYP3A4 expressioninduced by acridine derivatives. Fig. 10 presents the relativelevels of luciferase activity found in nhrtox-hepg2 cells treatedwith combinations of 0.001 mM C-1748 and 10 mM C-1305 with10 mM rifampicin (Rif). There were demonstrated that the increaseof luciferase activity induced individually by rifampicin and C-1748 were comparable, (2-fold), whereas the common effect ofthese compounds was nearly 2 times higher than individual.Therefore, the effect of C-1748 and rifampicin on PXR-dependentCYP3A4 expression seems to be additive. In the case of C-1305,Fig. 10 demonstrated that the individual influence of this drug onluciferase activity was 4 times higher than that of rifampicin. Incontrast to C-1748, the addition of rifampicin slightly enhanced the

s following exposure to C-1748 and C-1305 at increasing concentrations for 24-h. (A)

RT-PCR and normalized to the GAPDH values to control for the mRNA quality. The

(DDCT) method. (B) Western blotting and densitometric analyses of UGT1A1, P-gp

rials and methods section. The WB images are representative of three separate blots.

trol at: *p < 0.05; **p < 0.01; ***p < 0.001.

0

2

4

6

8

10

12R

ela

tive

lu

cif

era

se

ac

tivit

y

Rif - + - - + + C-1748 - - + - + -C-1305 - - - + - +

** **

***

***

***

ns

**

Fig. 10. Effect of rifampicin on PXR-mediated reporter gene expression induced by

C-1748 and C-1305. Nhrtox-hepg2 cells were treated for 24 h with rifampicin

(10 mM) as standard inductor of PXR with or without C-1748 (0.001 mM) and C-

1305 (10 mM) expression. After the treatment, cells were lysed and luciferase

activity was measured using luminometer with luciferase substrate. Mean � SEM,

n = 3, Significantly different from control or each from other at: **p < 0.01;

***p < 0.001; ns, nonsignificantly different.

M. Niemira et al. / Biochemical Pharmacology 86 (2013) 231–241 239

level of individual induction of PXR-dependent CYP3A4 expressionby C-1305.

4. Discussion

Interactions of antitumor agents with metabolic enzymespresent the potential for drug–drug interactions. Here, we testedthis potency of two acridine derivatives, which were selected topreclinical studies and were characterized by multidirectionalmechanisms of antitumor activity [17,22]. They are also dissimilarin their susceptibility to P450-mediated metabolism [20,23].Because of the above similarities and differences betweencompounds, studies on them gave good models to examine therole of PXR in mechanisms of drug–enzyme and drug–druginteractions.

We showed in this work that C-1748 and C-1305 increasedCYP3A4 and CYP2C9 enzymatic activities in HepG2 cells. Thisresult is in contrast with that obtained for C-1305 with humanrecombinant isoenzyme, where this compound was the slightinhibitor of CYP3A4 [23]. The postulated reason of suchdiscrepancy is the different conditions of enzymatic catalysis incell culture and in model enzymatic systems applied previously.The mechanism of C-1305-mediated inhibition of CYP3A4 wasshown there to be irreversible, probably mechanism-basedand was observed under concentration close to cytotoxic value

Fig. 11. Scheme of the effects induced by the antitumor agents C-1748 and C-1305 propos

CYP3A4, CYP2C9 UGT1A1 and MDR1 gene expression. Hypoxia diminished the effect of C

compounds.

of IC50. In contrast, the observed here increase of P450 enzymaticactivity occurred at very low concentrations and resulted, as wepostulate, from the induction of gene expression of enzymaticproteins. This conclusion was supported by the fact that theprofiles of concentration dependent CYP3A4 and CYP2C9 enzy-matic activity and relative mRNA levels were very similar.

Our studies, results of which are summarized in Fig. 11demonstrated that both compounds, induced not only CYP2C9 andCYP3A4, but also UGT1A1 and MDR1 gene expression. Thisinduction was shown to be mediated by PXR receptor. However,the level of PXR gene expression was not changed after incubationof HepG2 cells with C-1748 and C-1305. Therefore, thesecompounds modulate PXR-dependent gene expression of theCYP and UGT proteins rather by regulatory mechanisms thatdepend on up-regulation of PXR transactivating activity than bymodulation of PXR gene expression. In conclusion, we found that C-1748 and C-1305 are the ligands for PXR and up-regulators ofcytochrome P450 and UGT isoenzymes. In relation to rifampicinthe effect of C-1748 on PXR activity was comparable, whereas C-1305 effect was 4 times higher (Fig. 10). The composed result of C-1748 and rifampicin inductions of PXR activity was shown to beadditive. In contrast, the level of induction by C-1305 was such ahigh that the addition of rifampicin did not changed it significantly.One can suggest that C-1305 nearly saturate the place of ligandbinding and then rifampicin is not able to fortify this effect.

Several studies have reported the results of PXR-mediatedinduction by toxic agents, such as aflatoxins [27], pesticides andinsecticides [31] as well as therapeutic agents [32,33]. However,having the results of high induction of PXR-mediated CYP2C9CYP3A4, UGT1A1 and MDR1 gene expression by antitumor agents,we have to consider some additional specific aspects indicatingthat: (i) the induction of drug metabolizing enzymes can result indrug–drug interactions, therefore can affect the pharmacokineticsof other agents applied in multidrug antitumor therapy; (ii) cancerpatients are of high risk of PXR-mediated modulation, because ofvery low ‘‘therapeutic window’’ and multidrug therapy treatment;(iii) other consequences of PXR induction can be observed in healthand in tumor tissue, and (iv) PXR is able to modulate not only drugmetabolizing enzymes and drug-transporters, but also endoge-nous lipids and bile acids.

In addition, PXR-mediated up-regulation of metabolic enzymesmay play a specific role in antitumor therapy, because PXRexpression is detected not only in the liver and intestinal tissue butalso in breast, prostate and gastrointestinal cancer. PXR agonists issuggested to act on at least three ways. First of them, it is theactivation of prodrugs to their active forms, as it was observed inthe case of cytochrome P450 drug activation. These includealkylating agents ifosfamide and cyclophosphamide (2B6, 3A4),procarbazide (1A1), the fluoropirimidine prodrug, tegafur (2A6) orthe endocrine agents, flutamide (1A2, 3A4) and tamoxifen (3A4,

ed in this study. Drug (D)-mediated PXR induction, not PXR expression up-regulates

-1748 but not of C-1305. PXR-knock-out cells silenced induction properties of both

M. Niemira et al. / Biochemical Pharmacology 86 (2013) 231–241240

2D6) [24,34]. Therefore, the studied C-1748 and C-1305 canstimulate the prodrug transformation to its active form inmultidrug therapy.

The second function of PXR agonists is the activation of phase IImetabolic enzymes, including UGT isoforms, and the third, theinduction of phase III transporters, MDR proteins. The addition ofpolar glucuronid acid moiety generally decreases the biologicalactivity of the acceptor substrate, increases its water solubility, andyields metabolites that are readily recognized by varioustransporters, thereby facilitating its excretion. The PXR-mediatedup-regulation of both UGT and MDR fortifies these effects, whatlead to drug resistance. It was reported that up-regulation of UGTin human lung and colon cancer cells resulted in irinotecan (CPT-11)-resistant cell lines [35]. Similarly, overexpression of UGT1A6was shown to confer resistance to methotrexate in breast cancercell cultures [36]. These results strongly indicate that studiedcompounds, C-1748, and particularly C-1305 could diminish theanticancer effects of other drugs. However, this conclusion is notunequivocal. Firstly, UGT overexpression can decrease the level oftoxic metabolites of coadministered drugs, secondly, it is ofparticular interest, that PXR-modulated UGT-overexpression isalso postulated to have protective effect in hormone dependenttumor proliferation. Recent research has indicated that stableexpression of UGT2B7 in ovarian cells decreases cell proliferation(Radominska-Pandya A, unpublished data). There is also the simpleadvantage of UGT and MDR gene overexpression. This can reducethe drug toxicity induced in health tissue, particularly in the liver.

Other aspects should be care about the anticancer therapy withdrugs, which have been shown to induce PXR activity [37]. Thedifferential expression of PXR is observed in normal and canceroustissue. Some studies reported exclusively PXR expression incancerous tissue, other reports also provided evidence for higherPXR expression in cancerous tissue than in adjacent normal tissue.Markedly, PXR expression was found to be higher in the invasivestage than in early phase of breast cancer patients [38]. In addition,it is interesting to note the antiapoptotic role of PXR in tumortissue. For instance, the activation of PXR by rifampicin inhibitedthe proliferation of MCF-7 human breast cancer cells inducing cellcycle arrest at G1/S phase followed by apoptosis. The knockdownof PXR (using the targeted siRNA) blocked PXR activation inducedapoptosis of MCF-7 cells, suggesting that PXR activation cancontrol growth and can induce apoptosis [7]. Conversely, thetreatment of primary cultured hepatocytes with PXR agonistsincreased hepatocyte viability and protects them from druginduced apoptosis and this effect was correlated with theinduction of antiapoptotic protein genes Bcl-2 and Bcl-xL [39].

The above results indicate that studies on the role of PXRinduction on drug–drug interaction need strong improvement inmany aspects of PXR drug induction related to activation ofprodrugs, induction of phase II enzymes and phase III proteins aswell as the potential differences between normal and tumor cells.There were concluded that PXR antagonists are urgently needed[9], as they can potentially prevent PXR induced drug–druginteractions. However, it should be considered that PXR playsconserved roles in metabolism of endobiotics, including bile acidsand steroid hormones. Furthermore, the intensive inhibition ofPXR may cause hypersensitivity to xenobiotics and other toxins.

To support the option, which intends to reduce drug inducedPXR activity, we examined the impact of hypoxic conditions on thecompound-mediated induction of CYP3A4 and CYP2C9 geneexpression. This induction was minimized under hypoxia aftertreatment with C-1748. It is very profitable effect and found thefirst time for anticancer agent in respect to PXR induced drug–druginteractions. It is crucial result because hypoxia conditions arespecific for no susceptible to therapy solid tumors. However, C-1305 induction was not influenced by hypoxia. To explain this

phenomenon, one should consider that C-1305 expressed slightsusceptibility to metabolism in HepG2 cells [23], whereas C-1748was sensitive to metabolic transformations in vitro as well as inhepatoma cells. In addition, the rate of C-1748 metabolism wassignificantly higher under hypoxic conditions than under nor-moxia [20]. In light of above results, we suggest that metabolites ofC-1748 are not able to function as a PXR agonist. Only nativecompound is able to interact with PXR. Therefore, C-1748 seems tobe highly specific ligand of PXR, which is active at extremely lowconcentrations before its metabolic transformations. In accor-dance, the unexpected drop in C-1748-mediated modulation ofCYP2C9, CYP3A4, UGT1A1 and MDR1 gene expression undernormoxia at higher C-1748 concentration (Figs. 3, 4 and 9) mightresult from a higher level of metabolites inducing cytotoxic effects.

In conclusion, all of the above results indicate that bothantitumor agents C-1748 and C-1305 demonstrate significant up-regulation of CYP2C9, CYP3A4, UGT1A1 and MDR1 expression, in theprocess that involves the PXR transcription factor (Fig. 11).Therefore, PXR was shown to be a new target for acridineantitumor agents. The potential capabilities of these compounds toinduce PXR-mediated drug resistance is likely although suchresistance could be dimished in the case of C-1748 under hypoxicconditions observed in solid tumors.

Acknowledgement

This work was supported by the Ministry of Science and Highereducation (Poland) [Grant no. N405 355039]

References

[1] Harmsen S, Meijerman I, Beijnen JH, Schellens JHM. The role of nuclearreceptors in pharmacokinetic drug–drug interactions in oncology. CancerTreat Rev 2009;30:297–343.

[2] Choi HK, Yang JW, Roh SH, Han CY. Induction of multidrug resistance associ-ated protein 2 in tamoxifen-resistant breast cancer cells. Endocr Relat Cancer2007;14:293–303.

[3] Cho JY, Kang DW, Ma X, Ahn SH, Krausz KW, Luecke H, et al. Metabolomicsreveals a novel vitamin E metabolite and attenuated vitamin E metabolizmupon PXR activations. J Lipid Res 2009;50:924–57.

[4] Xie W, Radominska-Pandya A, Shi Y, Simon CM, Nelson MC, Ong ES, et al. Anessential role for nuclear receptors SXR/PXR in detoxification of cholestaticbile acid. Proc Natl Acad Sci USA 2001;98:3375–80.

[5] Albermann N, Schmitz-Winnenthal FH, Z’Graggen K, Volk C, Hoffman MM,Haefeli WE, et al. Expression of the drug transporters MDR1/ABCB1, MRP1/ABCC1, MRP2/ABCC2, BCRP/ABCG2 and PXR in peripheral blood mononuclearcells and their relationship with the expression in intestine and liver. BiochemPharmacol 2005;70:949–58.

[6] Pondugula SR, Mani S. Pregnane xenobioticreceptor in cancer pathogenesisand therapeutic response. Cancer Lett 2013;328:1–9.

[7] Verma S, Tabb MM, Blumberg B. Activation of the steroid and xenobioticreceptor, SXR, induces apoptosis in breast cancer cells. BMC Cancer 2009;9:3.

[8] Chen T. Overcaming drug resistance by regulating nuclear receptors. Adv DrugDeliv Rev 2004;62:1257–64.

[9] Chen Y, Tang Y, Guo C, Wang J, Boral D, Nie D. Nucler receptors in the multidrugresistance through the regulation of drug-metabolizing enzymes and drugtransporters. Biochem Pharmacol 2012;83:1112–26.

[10] Sane RS, Buckley DJ, Buckley AR, Nallani SC, Desai PB. Role of human pregnaneX receptor in tamoxifen and 4-hydroxytamoxifen–mediated CYP3A4 induc-tion in primary human hepatocytes and LS174 T. Drug Metab Dispos2008;36:946–54.

[11] Nallani SC, Goodwin B, Maglich JM, Buckley AR, Buckley DJ, Desai PB. Inductionof cytochrome P450 3A by paclitaxel in mice: pivotal role of the nuclearxenobiotic receptor, pregnane X receptor. Drug Metab Dispos 2003;31:681–4.

[12] Nallani SC, Goodwin B, Buckley AR, Buckley DJ, Desai PB. Differences in theinduction of cytochrome P450 3A by taxane anticancer drugs, docetaxel andpaclitaxel, assessed employing primary human hepatocytes. Cancer Che-mother Pharmacol 2004;54:219–29.

[13] Lindley C, Hamilton G, McCune JS, Faucette S, Shord SS, Hawke RL. The effect ofcyclophosphamide with and without dexamethasone on cytochrome P4503A4 and 2B6 in human hepatocytes. Drug Metab Dispos 2002;30:814–22.

[14] Harmsen S, Meijerman I, Beijnen JH, Schellens JHM. Nuclear receptor mediatedinduction of cytochrome P450 3A4 by anticancer drugs: a key role for thepregnane X receptor. Cancer Chemother Pharmacol 2009;64:35–43.

[15] Harmsen S, Meijerman I, Febus CL, Mass-Bakker RF, Beijnen JH, Schellens JHM.PXR-mediated induction of P-glycoprotein by anticancer drugs in a human

M. Niemira et al. / Biochemical Pharmacology 86 (2013) 231–241 241

colon adenoracinoma-derived cell line. Cancer Chemother Pharmacol 2010;66:765–71.

[16] Masuyama H, Nakatsukasa H, Takamoto N, Hiramatsu Y. Down regulationof pregnane X receptor contributes to cell growth inhibition and apoptosisby anticancer agents in endometrial cancer cells. Mol Pharmacol 2007;72:1045–53.

[17] Augustin E, Mos-Rompa A, Nowak-Ziatyk D, Konopa J. Antitumor 1-nitroa-cridine derivative, C-1748, induces apoptosis, necrosis or senescence in hu-man colon carcinoma HCT8 and HT29 cells. Biochem Pharmacol 2010;79:1231–41.

[18] Tadi K, Ashok BT, Chen YT, Banerjee D, Wysocka-Skrzela B, Konopa J, et al. Pre-clinical evaluation of 1-nitroacridine derived chemotherapeutic agent that haspreferential cytotoxic activity towards prostate cancer. Cancer Biol 2007;6:1632–7.

[19] Narayanan R, Tiwari P, Inoa D, Ashok BT. Comparative analysis of mutagenicpotency of 1-nitro-acridine derivatives. Life Sci 2005;77:2312–23.

[20] Wisniewska A, Niemira M, Jagiello K, Potega A, Swist M, Henderson C, et al.Diminished toxicity of C-1748, 4-methyl-9-hydroxyethyloamino-1-nitroacri-dine, compared with its demethyl analog, C-857, correspond to its resistanceto metabolism in HepG2. Biochem Pharmacol 2012;84:30–42.

[21] Cholody WM, Martelli S, Konopa J. 8-Substituted 5-[(aminoalkyl)amino]-6H-v-triazolo[4,5,1-de]acridin-6-ones as potential antineoplastic agents. Synthe-sis and biological activity. J Med Chem 1990;33:2852–6.

[22] Augustin E, Mos-Rompa A, Skwarska A, Witkowski JM, Konopa J. Induction ofG2/M phase arrest and apoptosis of human leukemia cells by potent antitumortriazoloacridinone C-1305. Biochem Pharmacol 2006;72:1668–79.

[23] Fedejko-Kap B, Niemira M, Radominska-Pandya A, Mazerska Z. Flavin-mono-oxygenases FMO1 and FMO3, not cytochrome P450 isoenzymes, contribute tometabolism of antitumor triazoloacridinone, C-1305, in liver microsomes andHepG2 cells. Xenobiotica 2011;41:1044–55.

[24] Harmsen S, Meierman I, Beijnen JH, Schellens JHM. The role of nuclearreceptors in pharmacokinetic drug–drug interactions in oncology. CancerTreat Rev 2007;33:369–80.

[25] Raucy JL, Lasker JM. Cell-based systems to assess nuclear receptor activationand their use in drug development. Drug Metab Rev 2013;45:101–9.

[26] Konopa J, Wysocka-Skrzela B, Tiwari RK. 9-Alkylamino-1-nitroacridine deri-vatives, European patent, 01910914.9-2101-US0105199; 1981, Americanpatent, 6.589.961 B2.

[27] Ratajewski M, Walczak-Drzewiecka A, Sokolowska A, Dastych J. Aflatoxinsupregulate CYP3A4 mRNA expression in a process that involves the PXRtranscription factor. Toxicol Lett 2011;205:146–53.

[28] Ma Q. Xenobiotic-activated receptors: from transcription to drug metabolismto disease. Chem Res Toxicol 2008;21:1651–71.

[29] Xu Ch, Yong-Tao Li Ch, Tony Kong AN. Induction of phase I, II and III drugmetabolism/transport by xenobiotics. Arch Pharm Res 2005;28:249–68.

[30] Mackenzie PI, Hu DG, Garden-Stephen DA. The regulation of UDP-glucurono-syltransferase genes by tissue-specific and ligand-activated transcriptionfactors. Drug Metab Rev 2010;42:95–105.

[31] Casabar RTC, Das PC, DeKrey GK, Gardiner CS, Cao Y, Rose RL, Wallace AD.Endosulfan induces CYP2B6 and CYP3A4 by activating the pregane X receptor.Toxicol Appl Pharmacol 2010;245:335–43.

[32] Vrzal R, Kubesowa K, Pavek P, Dvorak Z. Benzodiazepine medazepam andmidazolam are activators of pregnane X receptor and weak inducers ofCYP3A4: investigation in primary cultures of human hepatocytes and hepa-tocarcinoma cell lines. Toxicol Lett 2010;193:183–8.

[33] Rigalli JP, Ruiz ML, Perdomo VG, Villanueva SSM, Mottino AD, Catania VA.Pregnane X receptor mediates the induction of P-glycoprotein by spironolac-tone in HepG2 cells. Toxicology 2011;285:18–24.

[34] Michael M, Doherty MM. Tumoral drug metabolism: overview and its impli-cations for cancer therapy. J Clin Oncol 2005;23:205–29.

[35] Cummings J,Ethell BT, JardineL,Boyd G,Macpherson JS,Burchell B,Smyth JF, JodrellDI. Glucuronidation as a mechanism of intrinsic drug resistance in human coloncancer: reversal of resistance by food additives. Cancer Res 2003;63:8443–50.

[36] de Almagro MC, Selga E, Thibaut R, Porte C, Noe V, Ciudad CJ. UDP glucur-onosyltransferase1A6 overexpression in breast cancer cells resistant to meth-otrexate. Biochem Pharmacol 2011;81:60–70.

[37] Pondugula SR, Mani S. Pregnane xenobiotic receptor in cancer pathogenesisand therapeutic response. Cancer Lett 2013;328:1–9.

[38] Conde I, Lobo MV, Zamora J, Perez J, Gonzales FJ, Alba E, Fraile B, Paniagua R,Arenas MI. Human pregnane X receptor is expressed in breast carcinomas,potential heterodimers formatiom between hPXR and RXR-alpha. BMC Cancer2008;8:174.

[39] Zucchini N, de Sousa G, Bailly-Maitre B, Gugenheim J, Bars R, Lemaire G,Rahmani R. Regulation of Bcl-2 and Bcl-xL antiapoptotic protein expression bynuclear receptor PXR in primary cultures of human and rat hepatocytes. BBA2005;1745:48–58.