Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The cytochrome P450 superfamily of enzymes (P450) catalyze monooxygenation of both xenobiotic and endobiotic, lipophilic compounds. They contribute both to the detoxification of a variety of xenobiotic compounds and to the activation of many compounds to toxic, mutagenic, or carcinogenic derivatives. Several P450s are inducible and a rather complex regulation of P450s can take place at the transcriptional, post-transcriptional, and post-translational level. The constitutive expression of cytochrome P450 1A1 (CYP1A1) in human and rat liver is hardly detectable, but CYP1A1 is highly inducible in liver and other tissues by a variety of compounds, including a class of ubiquitous environmental chemicals, the polycyclic aromatic hydrocarbons (PAHs) and halogenated aromatic hydrocarbons (for review, see Gonzalez, 1988). Regulation of the CYP1A1 gene in response to PAHs and halogenated aromatic hydrocarbons occurs at the transcriptional level and is mediated via the ligand-dependent activation of the aryl hydrocarbon receptor (AhR). In the inactive form, AhR has been shown to be part of a cytosolic complex consisting of hsp90, hsp70, the immunophilin ARA9/AIP and other proteins (Perdew and Whitelaw, 1991; Carver and Bradfield, 1997; Ma and Whitlock, 1997). Upon binding of a ligand to the AhR, the receptor is translocated to the nucleus where it dimerizes to the bHLH-PAS protein arnt and binds to the xenobiotic response element (XRE) in the regulatory region of the CYP1A1 gene (for review, see Hankinson, 1995). Recently, several nonplanar compounds with little aromaticity, such as omeprazole and other benzimidazole compounds (Backlund et al., 1999 and references therein), the thiazolium compound YH439 (Lee et al., 1996), carbaryl (Ledirac et al., 1997), and bilirubin (Sinal and Bend, 1997) have been reported to induce CYP1A1 mRNA expression. The mechanism of induction of the CYP1A1 gene by these compounds is not clearly understood, since they are structurally different from classical AhR ligands.

Whereas regulation of CYP1A1 has been studied in great detail at the transcriptional level, little is known about the possible post-translational regulation of this protein. Post-translational regulation of P450 enzymes has been described for CYP2E1 and CYP3A, which have been shown to be protected from degradation by their substrates (Watkins et al., 1986; Eliasson et al., 1988) and ligands (Zhukov and Ingelman-Sundberg, 1997). For CYP2E1, this is the major mechanism for substrate-dependent elevation of the enzyme. Some reports also indicate that CYP1A2 can be protected from degradation by its substrate isosafrole (Steward et al., 1985) and methylcholanthrene (Silver and Krauter, 1988).

Primaquine (Fig. 1) interferes with the energy and reproductive abilities of malaria parasites and neutralizes acidic intracellular compartments, thus inhibiting lysosomal proteases and the budding of transport vesicles in the trans Golgi. It can also intercalate into lipid bilayers causing membrane deformation (Hiebsch et al., 1991; Reif et al., 1991). Primaquine is the most effective drug used against malaria relapses as the population rapidly develops an increasing chloroquine resistance. However, primaquine has a restricted use because of its toxicity in certain individuals. Here we report that primaquine, in the rat hepatoma H4IIE cells, is a potent inducer of CYP1A1. Our data indicate that primaquine uses the AhR for the transcriptional activation of the CYP1A1 gene and causes an inhibition of CYP1A1 enzyme degradation. This is to our knowledge the first report of a post-translational mechanism for CYP1A1 regulation.

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Structures of primaquine and TCDD.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Culture Conditions. The rat hepatoma H4IIE cells were grown as previously described (Backlund et al., 1997), and treatment was started when the cells were 80 to 90% confluent. Primaquine (Sigma, St. Louis, MO) and TCDD (Larodan Fine Chemicals, Malmö, Sweden) were dissolved in dimethyl sulfoxide (DMSO), and this solvent was added to control cells at 0.1%. V79 hamster fibroblast cells stably transfected with human CYP1A1 cDNA (Schmalix et al., 1993) were grown in Dulbecco's modified Eagle's medium containing 10% (v/v) fetal bovine serum, penicillin (100 IU/ml), streptomycin (100 µg/ml), and G418 (200 µg/ml). The cells were grown on 35-mm Petri dishes in a 5% CO₂ humidified incubator at 37°C. The medium was changed every 2 to 3 days and specifically the day before an experiment. Treatment was performed at 90 to 100% confluency. All cell culture media were obtained from Life Technologies (Rockville, MD).

Western Blot Analysis. After treatment of H4IIE cells for 24 h, the cells were harvested by scraping into phosphate-buffered saline (137 mM NaCl, 2.7 mM KCl, 4.3 mM Na₂HPO₄, 1.4 mM KH₂PO₄) and the 10,000g supernatant was prepared as previously described (Backlund et al., 1997). The V79 cells were treated for 8 h, rinsed, and then harvested by scraping into 800 µl of 10 mM Hepes (pH 7.5) containing 0.25 M sucrose. Microsomes were then prepared as described previously (Zhukov and Ingelman-Sundberg, 1999) in 10 mM Tris-HCl (pH 7.5). Microsomal proteins were separated on an 8 or 10% (w/v) bis/acrylamide SDS-polyacrylamide gel electrophoresis gel in a Bio-Rad Mini-Protean II cell. The proteins were transferred from the gel to a Hybond-C extra filter (Amersham International, Bucks, UK), which was incubated with an anti-rat CYP1A1 antibody (Gentest Corporation, Woburn, MA) and, in CYP1A1 degradation studies, additionally with antisera against rat NADPH-cytochrome P450 reductase. The degradation rate of the reductase is close to that of total microsomal protein, giving an internal control in the CYP1A1 degradation experiments. Possible sampling errors and contamination with nonmicrosomal protein are compensated for in this manner (Zhukov and Ingelman-Sundberg, 1997). The filters were then incubated with horseradish peroxidase-conjugated anti-goat immunoglobulins (Dako AS, Glostrup, Denmark) as the secondary antibody and with enhanced chemiluminescence reagent (Amersham) or SuperSignal reagent (Pierce, Rockford, IL) for detection. Band densities were measured on a Personal Densitometer (Molecular Dynamics, Sunnyvale, CA) with ImageQuant version 3.2 software.

CYP1A1 Degradation Assay. CYP1A1 degradation was studied using cycloheximide chase essentially as described earlier (Zhukov and Ingelman-Sundberg, 1997). On the day of the experiment, the cells were supplied with fresh medium containing 10 µg/ml cycloheximide and test substances as indicated. All substances were added as aqueous solutions. Each time and concentration point was run in triplicate. The duration of the incubations was 8 h unless otherwise stated. After incubation, the cells were harvested, microsomes were isolated, and Western blot analysis was performed as described above. IC₅₀ was estimated on the assumption that inhibition results from the equilibrium binding of PQ to its target. In this case IC₅₀ is given by the following equation: IC₅₀ = PQ (100  I)/I, where PQ and I are the primaquine concentration and percentage of inhibition, respectively. The data were fitted to a straight line on the PQ/I versus PQ plot and IC₅₀ was determined as the negative of the x-intercept.

Total Protein Degradation Assay. Degradation of total long-lived cell protein was followed using L-[U-¹⁴C]valine (specific activity 250 µC_i/mol; Amersham) in a pulse-chase experiment. When the cells were 80 to 85% confluent, the cell protein was labeled for 40 to 55 h in medium supplemented with 0.2 µCi/ml [¹⁴C]valine. The cells were then incubated for 9 to 10 h in fresh medium supplemented with 2 mM unlabeled valine to allow short-lived protein to degrade, after which the different effectors were added and incubation was continued for another 8 h. Aliquots of the media (0.2 ml in duplicate) were taken after 0 and 8 h and mixed with 0.2 ml of 1% bovine serum albumin and 0.4 ml of 40% trichloroacetic acid. After 15 min on ice, the mixtures were centrifuged at 17,000g for 10 min and the supernatant was mixed with 4 ml of scintillation liquid. Radioactivity in the aliquots was measured on a Beckman liquid scintillation system. After 8 h, the cells were harvested in the remaining medium, scintillation liquid was added, and the radioactivity was measured. The results are expressed as percentage of inhibition of degradation where degradation is the mean radioactive release at 8 h divided by the mean total incorporated radioactivity.

Northern Blot. Total RNA was isolated from H4IIE cells according to the method of Chomczynski and Sacchi (1987) and 15 µg of RNA was subjected to Northern blot analysis using standard procedures. Probes for CYP1A1 and -actin were obtained by reverse transcription-PCR amplification (Backlund et al., 1997), and labeled with [-³²P]dCTP using the RadPrime DNA labeling system (Life Technologies).

Plasmid Construction. The pGL3-XRE reporter plasmid has previously been described (Sogawa et al., 1986; Backlund et al., 1997). The 1195pGL3 and 917pGL3 reporter constructs (extending from nucleotides 1195 to +61 and 917 to +61, respectively, relative to the transcription start site) were prepared by PCR amplification of rat genomic DNA with specific primers and Pfu DNA polymerase (Stratagene, La Jolla, CA). The 5' primers contained, reading 5' to 3', a 3-nt random sequence, a MluI restriction site, and 19 to 20 nt corresponding to the appropriate target sequence in the rat CYP1A1 5'-flanking region. The 3' primers contained a 3-nt random sequence, an XhoI restriction site, and 19 nt of the appropriate CYP1A1 sequence (primer for 1195pGL3: 5'-ACT ACG CGT GCC CTT GCA AAG CTT AAG AC-3'; primer for 917pGL3: 5'-ACT ACG CGT CTC CAG GAA CCT GTG TGC A-3'; reverse primer: 5'-CTA CTC GAG GAA GAG TGT TCT CTA GGA C-3'). PCR products were digested with MluI and XhoI and ligated into the fitting cloning site of the pGL3-promoter vector (Promega, Madison, WI). The Gal4-AhR hybrid protein was created by cloning of rat AhR into the mammalian expression vector pFA-CMV (Stratagene), containing the cDNA for the yeast Gal4 DNA-binding domain (amino acids 1-147). The AhR construct pFA-AhR37, starting from amino acid 37 of AhR, was prepared by PCR amplification of a pCMV4-AhR plasmid using Pfu DNA polymerase. The forward primer containing a XbaI site was 5'-GAATTCTAGAACGACACAGAGACCGGCTG-3' and the reverse primer containing a KpnI site was 5'-TAATGGTACCTACAGGAATCCGCTGGGTG-3'. The PCR products were digested with XbaI and KpnI and ligated into the fitting clone site of the pFA-CMV vector. The sequence of the plasmid constructs was confirmed by sequencing.

Transfection. Transient transfection of H4IIE cells (with pGL3 vectors) was carried out in 12-well plates, using 1.3 µg of pGL3 reporter plasmid, 0.2 µg of pRL-TK control plasmid (Promega). For pFA-plasmids, cells were transfected in 24-well plates using 240 ng of pFA-AhR37, 250 ng of pFR-Luc reporter plasmid (Stratagene), and 5 pg of pRL-CMV control plasmid (Promega). The Tfx-20 transfection reagent was used according to the manufacturer's protocol (Promega). The day after transfection, the culture medium was exchanged with fresh media containing the inducers. After 24 h the cells were harvested and luciferase activity was analyzed according to the protocol for the Dual Luciferase Assay system (Promega) on a Turner Designs TD-20/20 luminometer. The results are expressed as the ratio between the firefly luciferase activity of the reporter plasmid to the Renilla luciferase activity of the control plasmid constituting the control for the transfection efficiency.

Cytosol Preparation and Electrophoretic Mobility Shift Assay. Confluent H4IIE cells were harvested by scraping into phosphate-buffered saline and cytosol was prepared as previously described (Backlund et al., 1997). For activation of cytosolic AhR to a DNA binding form, 32.5 µg of cytosolic protein was incubated in the presence of different concentrations of the ligand at 28°C for 3 to 4 h followed by electrophoretic mobility shift assay (EMSA) using a ³²P-labeled double-stranded XRE oligonucleotide, carried out as described in Backlund et al. (1997). Competition experiments were performed using 40-fold molar excess of unlabeled double-stranded XRE oligonucleotide. The retarded [³²P]XRE AhR-arnt complexes were identified by using specific antiserum against AhR (Affinity Bioreagents Inc., Golden, CO) and arnt proteins (Backlund et al., 1997) in the incubation reaction.

Kinetics of CYP1A1 Inhibition by Primaquine. Inhibition of CYP1A1 by primaquine was studied using microsomes from yeast expressing human CYP1A1 (Persson et al., 1997) with 7-ethoxyresorufin as a substrate. The 0.6-ml incubation mixture contained 0.1 M potassium phosphate buffer, pH 7.4, 0.1 mg/ml microsomal protein, 0.02% DMSO, and varying concentrations of ethoxyresorufin and primaquine. The reaction was started by the addition of NADPH (Sigma) to a final concentration of 1 mM. The mixture was incubated at 37°C for 10 min, and the reaction was stopped with 0.2 ml of 15% ZnSO₄ and 0.025 ml of 40% trichloroacetic acid. The mixture was extracted with 1 ml of diethyl ether, and the amount of resorufin formed was measured by high pressure liquid chromatography as described earlier (Leclercq et al., 1996). To determine the type of inhibition and estimate the inhibition constant the results were plotted in 1/v versus PQ and ER/v versus PQ coordinates where v is the reaction rate and PQ and ER are the concentrations of primaquine and ethoxyresorufin, respectively. The abscissa of the common intersection point of 1/v versus PQ graphs obtained at three different concentrations of ethoxyresorufin was taken as the estimate of the inhibition constant.

Primaquine Binding Spectrum. Yeast microsomes expressing CYP1A1 (Persson et al., 1997) were suspended in 0.1 M potassium phosphate buffer, pH 7.4, at the protein concentration of 0.5 mg/ml and divided between the sample and reference cells of a Cary 400 double beam spectrophotometer. The baseline was recorded, 10 µM primaquine added to the sample cell, and the difference spectrum recorded. To compensate for the absorption of primaquine below 400 nm, the spectrum of 10 µM primaquine was subtracted from the difference spectrum obtained with the microsomes.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Induction of CYP1A1 Protein and mRNA Levels. Immunoblot analysis was performed to examine the effect of primaquine on CYP1A1 protein expression in the H4IIE rat hepatoma cell line. As shown in Fig. 2A, no immunochemically detectable CYP1A1 was found in untreated cells. However, primaquine caused a dose-dependent increase of CYP1A1 protein already apparent at a concentration of 1 µM, whereas maximum induction was obtained at about 45 µM. The extent of CYP1A1 protein induction by 45 µM primaquine was approximately 70% of the maximum induction caused by TCDD. At a concentration of 30 µM primaquine caused a time-dependent increase in CYP1A1 protein levels, which remained elevated for at least 24 h (Fig. 2B). To determine whether the increase in CYP1A1 protein was accompanied by an increase in CYP1A1 mRNA levels, Northern blot analysis was carried out. As shown in Fig. 2, C and D, treatment of H4IIE cells with 30 µM primaquine as well as 10 nM TCDD caused a time-dependent increase in CYP1A1 mRNA, reaching maximum levels at 6 h for primaquine. Whereas the maximum level of CYP1A1 mRNA achieved by TCDD treatment showed further increase after 12 h, the CYP1A1 mRNA levels in primaquine-treated cells were clearly diminished at 9 h and further at 12 h.

View larger version (31K): [in this window] [in a new window]   Fig. 2.   Dose- and time-dependent induction of CYP1A1. A, H4IIE cells were treated with 0 to 60 µM primaquine (PQ) and 0.1 and 10 nM TCDD for 24 h and analyzed by Western blot. B, H4IIE cells were treated with 30 µM primaquine for the indicated time points. For analysis of CYP1A1 protein expression by Western blot, 30 µg of cellular protein was used. C, H4IIE cells were treated during the indicated time with DMSO (CTRL), 30 µM primaquine, 10 nM TCDD, and total RNA was prepared. Total RNA (15 µg) was subjected to Northern blot analysis and blots were hybridized with 32P-labeled CYP1A1 and -actin probes. One representative blot of CYP1A1 and -actin expression is shown. D, densitometric scanning of the autoradiograms was done to estimate the mRNA levels of CYP1A1. The diagram represents the relative amount of CYP1A1 after correction for -actin levels. The values are mean ± S.E. of four independent experiments.

Transcriptional Activation of the CYP1A1 Gene. To investigate the effect of primaquine on the transcriptional activation of the CYP1A1 gene, transient transfection assays were carried out in H4IIE cells using three different reporter constructs derived from the rat CYP1A1 5'-flanking region. The 1195pGL3 plasmid contains most of the 5'-flanking sequence of the CYP1A1 gene shown to be important for high inducibility of the rat CYP1A1 gene (Sogawa et al., 1986; Fujisawa-Sehara et al., 1987). As shown in Fig. 3, H4IIE cells transiently transfected with the 1195pGL3 reporter plasmid and treated with primaquine and TCDD for 24 h, exhibited an increase in luciferase activity 3.5- and 6.2-fold, respectively, compared with control (DMSO-treated) cells. The pGL3-XRE plasmid (extending from nucleotides 1140 to 848 relative to the transcription start site) contains several copies of the XRE enhancer element that can drive inducibility by TCDD and other AhR ligands in an AhR-dependent manner (Fujisawa-Sehara et al., 1987). H4IIE cells transfected with pGL3-XRE clearly showed a 3.7- and 7.2-fold induction of luciferase activity after exposure to primaquine and TCDD, respectively, suggesting an involvement of the AhR in primaquine-mediated CYP1A1 induction. In contrast, transfection of H4IIE cells with the 917pGL3 plasmid, which contains elements implicated in negative regulation of CYP1A1 gene as well as a PAH (4S)-protein binding region (Hines et al., 1988; Sterling et al., 1993), did not display increased luciferase activity by primaquine and TCDD compared with control cells.

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Transient transfection with pGL3 luciferase reporter constructs in H4IIE cells. H4IIE cells were transiently transfected with the CYP1A1 5'-flanking reporter plasmids 1195pGL3, pGL3-XRE, or 917pGL3, respectively, together with the pRL-TK control plasmid. The next day, the medium was changed and the cells were treated for 24 h with DMSO, 30 µM primaquine, or 10 nM TCDD. The cells were harvested and firefly and Renilla luciferase activity was analyzed as described under Materials and Methods. The values are mean ± S.E. of four independent experiments.

Activation of Gal4-AhR Hybrid Protein. The involvement of the AhR in primaquine-dependent transcriptional activation of the CYP1A1 gene was further assessed by investigating activation of a Gal4-AhR hybrid protein by primaquine. In the pFA-AhR37 construct, the first 36 amino acids of the rat AhR, e.g., the DNA binding domain, are replaced by the yeast Gal4 DNA binding domain. Transfection of H4IIE cells with the pFA-AhR37 plasmid and the appropriate pFR-Luc reporter gene, generated a transient activation of the Gal4-AhR hybrid protein, as displayed by an temporary increase in luciferase activity, in response to both primaquine and TCDD (Fig. 4). The basal luciferase activity of cells transfected with the pFA-AhR37 construct was approximately 10-fold higher than the activity of the empty pFA-CMV plasmid and primaquine did not have an effect on the vector alone (data not shown).

View larger version (30K): [in this window] [in a new window]   Fig. 4.   Activation of Gal4-AhR hybrid protein. H4IIE cells were cotransfected with the pFA-AhR37 plasmid, the pFR-Luc reporter plasmid and the pRL-CMV control plasmid. The next day, the medium was exchanged and the cells were treated with DMSO, 30 µM primaquine, or 10 nM TCDD for the time indicated. The cells were harvested and firefly and Renilla luciferase activity was analyzed. The diagram represents mean ± S.D. values of one representative experiment of four independent experiments.

Cytosolic EMSA. Ligand-dependent activation of the cytosolic AhR to a DNA-binding form, so called transformation, can be monitored by incubation of cytosol with the ligand, in vitro followed by EMSA. H4IIE cytosol was incubated with different concentrations of primaquine and TCDD as indicated in Fig. 5 and analyzed by gel shift assay. In control cytosol (exposed to DMSO) a retarded ³²P-labeled band was apparent (lane 1), which increased in intensity with increasing concentrations of primaquine (lanes 2-4). The [³²P]XRE-protein complex comigrated with the complex formed by cytosol exposed to TCDD (lanes 6-8). Furthermore, competition with 40-fold molar excess of cold XRE oligonucleotide abolished the formation of the [³²P]XRE-protein complex of primaquine-treated cytosol (lane 10). The identity of proteins in the [³²P]XRE complex could be verified as AhR and arnt, since addition of an antibody against AhR to the EMSA reaction abolished formation of the retarded radiolabeled band (lane 11) and addition of antisera against arnt produced a supershift of the ³²P-labeled band (lane 12).

View larger version (58K): [in this window] [in a new window]   Fig. 5.   Cytosolic EMSA. Cytosolic protein was incubated with DMSO (lane 1), primaquine (lane 2, 0.3 µM; lane 3, 3 µM; lane 4, 30 µM) or TCDD (lane 6, 0.01 nM; lane 7, 0.1 nM; lane 8, 1 nM) at 28°C for 3 h followed by incubation with a 32P-labeled XRE probe for 20 min before subjecting part of the incubation reaction to EMSA. In lane 10 to 12, the effect of a 40-fold excess of unlabeled XRE (lane 10) or the addition of 3 µl of AhR antibody (lane 11) or 1.5 µl of arnt antiserum (lane 12) on the gel shift caused by 30 µM primaquine (lane 9) is shown. Lane 5, probe alone; R, AhR-arnt complex; S, supershifted receptor complex.

Inhibition of CYP1A1 Protein Degradation. The transient increase of the CYP1A1 mRNA levels, with a peak at 6 h (Fig. 2C), but a sustained elevation of the CYP1A1 protein for at least 24 h (Fig. 2B), prompted a further investigation of whether primaquine could influence the stability of the CYP1A1 protein. Hamster fibroblast V79 cells, stably transfected with the human CYP1A1 cDNA (Schmalix et al., 1993), were used as a model system. In these transfected cells, in contrast to hepatocytes, uninduced CYP1A1 is expressed at levels allowing detection by immunoblot. In cycloheximide chase experiments, CYP1A1 was rapidly degraded in V79 cells (Fig. 6) with a half-life of 2.8 h (data not shown), whereas reductase remained stable (Fig. 7). Primaquine caused a dose-dependent inhibition of CYP1A1 degradation with an IC₅₀ of 3.3 µM (Fig. 8).

View larger version (12K): [in this window] [in a new window]   Fig. 6.   Dose-response relationship for inhibition of 1A1 degradation by primaquine. Semilogarithmic plot of CYP1A1 fraction remaining after an 8-h cycloheximide chase as a function of primaquine concentration. See Materials and Methods for experimental details. The data represent the mean ± S.E. of three or four experiments. The inset shows the immunoblot of one of the experiments. C, zero time control.

View larger version (18K): [in this window] [in a new window]   Fig. 7.   Western blot of CYP1A1 and reductase. After the indicated incubation times in the presence of cycloheximide, microsomes were isolated and the proteins were assayed using a mixture of anti-CYP1A1 and anti-reductase sera.

View larger version (10K): [in this window] [in a new window]   Fig. 8.   Hanes plot for inhibition of CYP1A1 degradation by primaquine. The mean values from Fig. 6 were used when percentage of inhibition (I) was calculated relative to the fraction degraded in the absence of primaquine (91.5%). The data was fitted to a straight line on the PQ/I versus PQ plot and IC50 was determined as the negative of the x-intercept (3.3 µM).

Effect of Other Lysosomotropic Agents on CYP1A1 Degradation. Experiments were designed to evaluate whether the inhibition of CYP1A1 degradation by primaquine was due to the lysosomotropic activity of the drug. Primaquine as well as two other lysosomotropic agents, chloroquine and ammonium chloride, were compared for their inhibitory effect on the degradation of total long-lived protein and CYP1A1 in V79 cells. Like primaquine, both chloroquine and ammonium chloride are membrane-permeable weak bases that neutralize pH in lysosomes, thus inhibiting lysosomal proteases (Kalina and Socher, 1991). The concentrations of chloroquine and ammonium chloride used in this experiment were similar to those shown to inhibit lysosomal proteases (Fritsch et al., 1988; Eleftheriades et al., 1995). Correlation between the effects of these three substances on the total protein and CYP1A1 degradation would be consistent with the involvement of lysosomes in CYP1A1 degradation. At the concentrations used, chloroquine showed the same, and ammonium chloride twice as strong inhibition of total protein degradation as 60 µM primaquine (Table 1). Yet, CYP1A1 degradation was affected by chloroquine only slightly and was not affected at all by ammonium chloride. These results are inconsistent with lysosome-mediated degradation of CYP1A1 and indicate that primaquine protects the enzyme from degradation by a different mechanism.

Primaquine Binding to CYP1A1. To find out whether primaquine binds to CYP1A1, its effect on 7-ethoxyresorufin-O-deethylase activity, a specific probe for CYP1A1 and CYP1A2 (Pastrakuljic et al., 1997), was investigated. As shown in Fig. 9, primaquine acts as a competitive inhibitor with a K_i of 1.3 µM, thus demonstrating that the drug is a high-affinity ligand of CYP1A1. This was further confirmed by the finding that primaquine elicited a type II binding spectrum with the yeast microsomes expressing CYP1A1. Thus, a maximum absorbance at 429 nm and a minimum at about 400 nm were found (Fig. 10).

View larger version (14K): [in this window] [in a new window]   Fig. 9.   Kinetics of CYP1A1 inhibition by primaquine. Dixon (A) and Cornish-Bowden (B) plots for inhibition of ethoxyresorufin O-deethylase activity by primaquine. Ethoxyresorufin concentrations were 0.02 µM (), 0.05 µM (), and 0.1 µM (). The results are calculated from the mean of two or three experiments. See Materials and Methods for experimental details.

View larger version (10K): [in this window] [in a new window]   Fig. 10.   Primaquine binding spectrum. The binding spectrum elicited by 10 µM primaquine with yeast microsomes expressing CYP1A1 (0.5 mg/ml protein). See Materials and Methods for details.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The data of the present study show that primaquine causes a time- and dose-dependent increase in the amount of CYP1A1 mRNA and protein in the rat hepatoma H4IIE cells. The increase in CYP1A1 mRNA was transient, reaching maximal levels (90% of the TCDD response) 6 h after primaquine treatment. Although the CYP1A1 mRNA levels declined after 6 h, the protein levels remained elevated for at least 24 h. The transient induction on the mRNA level, and more persistent action on the protein level in H4IIE cells, has previously been seen with benzo[a]pyrene as the inducer. The CYP1A1 mRNA levels peaked at 4 h after treatment and then declined, whereas the enzymatic activity was maximal at 12 h and still remained elevated 72 h after treatment (Xu et al., 1993). Similar transient induction of CYP1A1 has been obtained with other compounds that are metabolized or inactivated by the cells. On the other hand, dioxin congeners produce a much more sustained effect on CYP1A1 induction probably due to the stability of these compounds in biological systems (Riddick et al., 1994).

Interestingly, primaquine caused a rapid CYP1A1 mRNA induction response already evident after 3 h of treatment, similar to the TCDD response. In contrast, we have previously shown that omeprazole causes a much slower response, indicating a different mechanism of induction exerted by this drug (Backlund et al., 1997).

To address the question of transcriptional induction of the CYP1A1 gene by primaquine, transient transfections using different luciferase reporter constructs derived from the rat CYP1A1 5'-regulatory region were carried out. Both the 1195pGL3 and pGL3-XRE constructs showed a substantial increase in luciferase activity in cells treated with primaquine. These constructs contain several copies of the cis-regulatory element XRE (nt 1140 to 848 of the rat CYP1A1 gene), known to be necessary for transcriptional activation of the CYP1A1 gene (Sogawa et al., 1986; Fujisawa-Sehara et al., 1987). In contrast, the 917pGL3 plasmid did not display increased luciferase activity by primaquine or TCDD, explained by the lack of XREs in this construct. Upon stimulation, the transcription factor AhR and its dimerization partner arnt have been shown to bind to the XRE elements and transactivate the CYP1A1 gene. Therefore, an increase in pGL3-XRE luciferase activity as an effect of primaquine treatment reflects recruitment of the AhR-arnt transcription factor.

The involvement of the AhR was further examined by using a Gal4-AhR fusion protein, which upon activation causes an increase in luciferase reporter activity in transfected cells. The results of the present study clearly demonstrate an activation of the AhR fusion protein by primaquine that is consistent with a utilization of the AhR. However, activation of the Gal4-AhR hybrid protein is not direct evidence for binding of primaquine to the AhR. To address this question we examined transformation of cytosolic AhR to a DNA binding form in vitro, a sensitive assay, which has been extensively used to assess binding and affinity of ligands to the Ah receptor. Primaquine was capable of transforming cytosolic AhR, indicating that primaquine induced dissociation of AhR from the hsp90 complex and binding to XRE. Fontaine et al. (1999) recently showed that primaquine significantly induced CYP1A1 gene expression in hepatocytes and HepG2 cells. Although an AhR antagonist, -naphthoflavone, did suppress CYP1A1 induction, primaquine was not able to displace [³H]TCDD in competitive binding studies using 9S-enriched fractions of human cytosol. Thus, they suggest that CYP1A1 induction involves participation of the AhR, but no direct primaquine-receptor interaction. In contrast, our results showed that primaquine was capable of transforming cytosolic AhR to a DNA binding form, in vitro, indicating an interaction of primaquine with the AhR-arnt complex. The reason for the discrepancy between our results and those of Fontaine et al. (1999) may be due to the different cell systems used, rat versus human, and/or the different assay methods used. Displacement of [³H]TCDD by a weak ligand, such as primaquine, would require that the unbound concentration of [³H]TCDD in the assay should not be greater than needed for saturation of the AhR. The equilibrium constant (K_d) for [³H]TCDD to HepG2 and human AhR has been reported to 9 and 18 nM, respectively (Roberts et al., 1990). The concentration of [³H]TCDD used by Fontaine and coworkers was 50 nM, which is likely to be greater than that necessary to saturate human AhR. Although we have not directly compared the sensitivity of the two assays in this study, our results clearly indicate that primaquine directly activates the receptor complex in vitro. The mechanism of activation of the AhR by primaquine is in sharp contrast to what we previously found for omeprazole. This benzimidazole, apparently not being able to displace TCDD from the AhR, caused a nuclear translocation of the AhR but no cytosolic transformation, in vitro (Daujat et al., 1992; Backlund et al., 1997). One can therefore distinguish three types of compounds that can activate the AhR: 1) agents such as TCDD that bind with high affinity and can transform the receptor in vitro; 2) agents such as primaquine that probably are weak ligands and can transform the receptor in vitro; and 3) agents such as omeprazole that apparently cannot displace TCDD from the receptor and not transform the receptor in vitro. The mechanism of AhR activation by these different types of compounds could include direct binding to the AhR, either to the TCDD binding site or a second binding site, indirect binding followed by activation of the receptor, or signal transduction-mediated activation of the receptor, causing a similar conformational change as that of ligand binding. Further research is needed to elucidate the molecular basis for these different mechanisms of AhR activation.

It is well known that CYP1A1 expression is regulated at the transcriptional level and regulation of CYP1A1 at the post-transcriptional level has also been reported (Silver and Krauter, 1988; Xu et al., 1993). Yet, no direct evidence on the possibility of post-translational regulation of CYP1A1 is available. Regarding the P450 enzymes CYP2E1 and CYP3A, post-translational regulation is an important mechanism for enzyme induction and degradation is inhibited by their substrates and ligands (Watkins et al., 1986; Eliasson et al., 1988; Zhukov and Ingelman-Sundberg, 1997). Since primaquine caused a transient increase in CYP1A1 mRNA, yet gave a sustained increase of CYP1A1 protein level, the question of whether primaquine also can stabilize CYP1A1 post-translationally was investigated. Hamster fibroblast V79 cells expressing no P450s of their own, which have been stably transfected with human CYP1A1 cDNA (Schmalix et al., 1993), were used. In this system, CYP1A1 degradation was studied using a cycloheximide chase, and primaquine was shown to inhibit the degradation in a dose-dependent manner.

Primaquine is known to neutralize pH in lysosomes, thus inhibiting lysosomal proteolysis (Reif et al., 1991). To rule out the possibility of degradation inhibition being due to lysosome dysfunction, two other lysosomotropic agents were tested. Only primaquine, but neither chloroquine nor ammonium chloride caused an appreciable inhibition of CYP1A1 degradation in V79 cells. In contrast, both chloroquine and ammonium chloride were efficient in inhibiting degradation of long-lived proteins in the cells. This shows that the inhibition of CYP1A1 degradation by primaquine is not due to the inhibition of lysosomal proteases.

As mentioned earlier, CYP2E1 and CYP3A4 are protected from degradation by their substrates and ligands (Watkins et al., 1986; Eliasson et al., 1988; Zhukov and Ingelman-Sundberg, 1997). A similar mechanism may be relevant for the ability of primaquine to inhibit CYP1A1 degradation. Murray (1984) showed that primaquine inhibits aminopyrine N-demethylase activity in rat liver microsomes. Also, primaquine but not chloroquine inhibited CYP1A2-catalyzed antipyrine metabolism in vivo (Back et al., 1983). In the present study, primaquine was found to act as a competitive inhibitor of CYP1A1-dependent ethoxyresorufin O-dealkylase activity with a K_i of 1.3 µM, showing it to be a high-affinity ligand of CYP1A1. A similar IC₅₀ value of 3.3 µM obtained in the degradation inhibition experiments with V79 cells supports the view that primaquine binding in the CYP1A1 active site can protect the enzyme from degradation.

All substances reported to protect CYP2E1 from degradation are type II or reverse type II substrates. Type II ligands bind to the heme iron and compete with oxygen, thus preventing oxygen activation by P450. Reactive oxygen species cause oxidative modifications of proteins that can increase susceptibility to proteolytic attack (for review, see Karuzina and Archakov, 1994). Blocking the supply of reducing equivalents to CYP2E1 protects the enzyme from degradation (Zhukov and Ingelman-Sundberg, 1999). We found that primaquine elicits a type II difference spectrum with yeast CYP1A1 microsomes and was a high-affinity ligand of CYP1A1. It may therefore protect CYP1A1 by a mechanism similar to that demonstrated for CYP2E1, indicating that protection from degradation by type II substrates could be a general phenomenon for a number of cytochrome P450 forms.

In conclusion, the present study shows that primaquine causes activation of the AhR complex and transcriptional induction of the CYP1A1 gene, but also prevents degradation of the CYP1A1 enzyme by binding to its active site. This indicates that this enzyme of great importance for bioactivation of precarcinogens, is subject to multifactorial regulation on several different cellular levels.