Experimental Procedures

Animals and Materials.

Male Sprague-Dawley CDCrl:CD (SD)BR rats were purchased from Charles River (Cléon, France). Animals were fed with a normal standard diet (Autoclavé 113; Usine d’Alimentation Rationnelle, Villemoisson-sur-Orge, France) and used when they weighed 180 to 190 g.

Human recombinant IL-2 was purchased from Genzyme (Cambridge, MA).n-Butyric acid, DMSO, retinoic acid, genistein, and erythromycin were from Sigma Chemical Co. (St. Louis, MO). Collagenase type I and the transfection reagentN-[1-(2,3-dioleoyloxy)propyl)]-N,N,N-trimethylammonium methylsulfate (DOTAP) were obtained from Boehringer-Mannheim (Mannheim, Germany). Collagen-coated culture dishes were from Polylabo (Strasbourg, France). [³²P]Deoxycytidine triphosphate (3000 Ci/mmol) and enhanced chemiluminescence (ECL) kits were from Amersham (Les Ulis, France). Protein A-Sepharose was from Pharmacia (Uppsala, Sweden). Peroxidase-conjugated streptavidin was from DAKO (Glostrup, Denmark).

Anti-c-Myc Antibodies, c-myc DNA Probes, and c-myc P(S)ODNs.

A mouse monoclonal anti-c-Myc antibody and a rabbit polyclonal anti-c-Myc antibody were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). These two antibodies were used to immunoprecipitate and reveal c-Myc, respectively.

A cDNA probe covering the region from nucleotides 71 to 1053 of the human glyceraldehyde-3-phosphate dehydrogenase (G3PDH) gene was purchased from Clontech (Palo Alto, CA). TheClaI/EcoRI 1.4-kb probe for the c-mycthird exon was purchased from Appligene Oncor (Illkirch, France). Antisense (5′-CACGTTGAGGGGCAT-3′) and sense (5′-ATGCCCCTCAACGTG-3′) c-myc P(S)ODNs were prepared by Genset (Paris, France). These P(S)ODNs correspond to the sequence ranging from nucleotides 1995 to 2010 of rat c-myc (Hayashi et al., 1987). This is the initial translation initiation region for the second exon, which is the first to be translated because the first rat exon has no coding capacity (Hayashi et al., 1987). Equivalent 15-mer ODNs for human c-myc have been used in previous antisense studies (Holt et al., 1988; Shi et al., 1992). This region was selected because it is single-stranded in the human c-myc transcript (Holt et al., 1988), and the rat and human genes differ by only one base in this region.

Anti-CYP Antibodies and CYP Primers.

The anti-CYP antibodies used in this study and the primers used to assess the CYP2C11, CYP3A, and β-actin transcripts by reverse transcription (RT)-polymerase chain reaction (PCR) have been described (Thal et al., 1994; Tinel et al., 1995).

Cell Culture.

Hepatocytes were isolated by collagenase perfusion as previously described (Tinel et al., 1995). The viability of hepatocytes determined by trypan blue exclusion was greater than 85%. Isolated hepatocytes were cultured on collagen-coated culture dishes in William’s E medium supplemented with 0.1 mg/ml insulin (Sigma), 100 U/ml penicillin, 0.1 mg/ml streptomycin, 70 μM hydrocortisone, and 10% fetal calf serum (FCS). Hepatocytes were maintained at 37°C in a humidified incubator under 5% CO₂ in air. After the attachment period (3 h), the medium was removed and replaced by a fresh medium without FCS and with or without IL-2 (350 U/ml). In preliminary experiments, this dose of pure human recombinant IL-2 (Genzyme) was found to reproduce the CYP effects that we had observed with 10-fold higher doses of the human recombinant IL-2 that was mixed with a dextran vehicle (Roussel-Uclaf) (Tinel et al., 1995), suggesting that this vehicle may inhibit the binding of IL-2 with its receptor. In some culture dishes, genistein (20 μg/ml), retinoic acid (10 μM), n-butyric acid (3 mM), or DMSO (2%) also was added. Measurements were made after 24 h of treatment. In other experiments, treatments were applied from 24 to 48 h after cell attachment, and CYPs were studied 48 h after cell attachment.

For the cultures performed with the sense or antisense c-mycP(S)ODN, heat-inactivated (65°C for 30 min) FCS with minimal nuclease activity was used for cell attachment. The medium was then removed and replaced by serum-free medium containing the transfection reagent DOTAP (5 μg/ml) with or without IL-2 (350 U/ml) and the sense or antisense P(S)ODN (2 μM).

Northern Blot Analysis of c-myc mRNA.

Total RNAs were isolated from culture dishes as described previously (Tinel et al., 1995). Total RNA (20 μg) was subjected to 1.2% agarose gel electrophoresis in the presence of 6.6% formaldehyde, transferred by capillarity to Hybond N⁺ nylon membranes (Amersham), and baked for 2 h at 80°C. The c-myc and G3PDH probes were labeled with [³²P]dCTP by random priming. The nylon membranes were prehybridized for 4 h in 50% formamide, 5× Denhardt’s solution, 10% dextran sulfate, 6× standard sodium citrate, 1% SDS, and 20 μg/ml salmon sperm DNA and then hybridized for 24 h with the DNA probe (2 × 10⁶cpm/ml). After three washings, membranes were exposed to autoradiography film (5 days at −80°C), and RNA bands were quantified by laser scanning.

Immunoprecipitation and Immunoblots of c-Myc Protein.

Hepatocytes (10⁶ cells) were washed in PBS and lyzed with 0.1% Nonidet P-40 (BDH Laboratory Supplies, Poole, England), 250 mM NaCl, 5 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 0.5 mM sodium orthovanadate, 5 mM sodium fluoride, and 50 mM HEPES buffer, pH 7.8 (Robin et al., 1996). After protein measurement, the cell lysate was diluted to a protein concentration of 0.2 mg/ml, as described previously (Robin et al., 1996). Then, 1 ml was incubated with the anti-c-Myc mouse monoclonal antibody (2 μg) for 1 h at room temperature.

Immunocomplexes were precipitated with protein A-Sepharose, eluted, and subjected to SDS-9% polyacrylamide gel electrophoresis (PAGE) as described previously (Robin et al., 1996). After incubation with the blocking reagent (Amersham), nitrocellulose sheets were incubated for 1 h with the anti-c-Myc rabbit polyclonal antibody (diluted 1:3000 in Tween-20, Tris-buffered saline, pH 7.6). Sheets were then incubated for 1 h with a horseradish peroxidase-linked antibody raised in donkeys against rabbit immunoglobulins (Amersham) diluted 1:1500. Immunoreactive c-Myc was revealed by an ECL detection system (Amersham). Quantification of the protein bands was performed by laser densitometry.

Cellular CYP, Erythromycin Demethylation, and CYP Proteins.

Total cellular CYP was measured by the CO-difference spectrum of dithionite-reduced cell lysates as described previously (Tinel et al., 1995). Cell erythromycin N-demethylation, an activity supported by CYPs 3A, was measured with a whole-cell homogenate as described previously (Tinel et al., 1995).

To prepare microsomes, hepatocytes were washed with PBS, scraped, and centrifuged (Tinel et al., 1995). The cell pellet was resuspended in 0.1 M phosphate buffer, pH 7.4, and sonicated with a 80-W Vibracell sonicator (Bioblock, Illkirch, France). The microsomes were isolated by differential centrifugation at 10,000g and then 100,000g, and resuspended in 0.1 M phosphate buffer, pH 7.4, 1 mM EDTA, and 20% of glycerol. The microsomal suspension was kept at −80°C until use.

Unless otherwise indicated, 10 μg of microsomal proteins was fractionated by SDS-10% PAGE as described previously (Tinel et al., 1995). Electrophoretic transfer to nitrocellulose sheets was performed in a Bio-Rad semidry transfer cell (Bio-Rad, Ivry-sur-Seine, France). Unspecific binding sites were blocked with 1% Tween-20 and 1% polyvinylpyrrolidone in PBS, pH 7.2. Nitrocellulose sheets were first incubated with the anti-CYP sera diluted 1:1000 (anti-CYP1A and anti-CYP3A), 1:500 (anti-CYP2C11), or 1:100 (anti-CYP2B and anti-CYP2D1) in PBS containing 10% BSA and 10% neonatal calf serum and then incubated with the peroxidase-conjugated antibody against rabbit or mouse immunoglobulins (DAKO, Copenhagen, Denmark) and stained with diaminobenzidine and H₂O₂ in the presence of 0.015% CoCl₂ and 0.015% NiCl₂. In experiments performed with the sense or antisense P(S)ODN, immunoblots (1.5 μg of microsomal proteins) were revealed with an ECL system (Amersham). Protein bands were quantified by laser densitometry. The linearity of densitometric measurements was verified in preliminary experiments.

CYP2C11 and CYP3A mRNAs.

The transcripts of CYP2C11, CYP3A, and β-actin were assessed by RT-PCR as described previously (Tinel et al., 1995). Briefly, cDNAs were synthesized from 1 μg of total RNA using the Moloney murine leukemia virus RT system, followed by PCR amplification in the presence of the corresponding primers and [³²P]deoxycitidine triphosphate (1 μCi/tube) as described previously (Tinel et al., 1995). After 1.5% agarose gel electrophoresis, the amplified cDNA bands were excised under UV light and counted for ³²P radioactivity (Tinel et al., 1995).

Statistical Analysis.

ANOVA followed by a Dunnett’st test was used to assess the significance of differences between mean values.

Results

Prevention of IL-2-Mediated Increase in c-myc mRNA and Decrease in CYP by Retinoic Acid, n-Butyric Acid, or DMSO.

Because IL-2 has been shown to increase c-mycexpression in lymphocytes (Asao et al., 1994), we looked for similar effects in hepatocytes (Fig. 1). The c-myc mRNA was increased by 140% in hepatocytes cultured for 24 h with IL-2 (350 U/ml) (Fig. 1). Retinoic acid,n-butyric acid, and DMSO, three agents that have been shown to decrease transcription of the c-myc gene (Westin et al., 1982; Dony et al., 1985; Krumm et al., 1992; Strobl and Eick, 1992;Heruth et al., 1993), all prevented the IL-2-mediated increase in c-myc mRNA in IL-2-treated cells (Fig. 1). When added alone (without IL-2), these agents had limited effects on c-mycmRNA, which was not significantly decreased compared with control cells (Fig. 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Prevention of the IL-2-mediated increase in c-myc mRNA by retinoic acid, butyric acid, or DMSO. Rat hepatocytes were cultured for 24 h with or without IL-2 (350 U/ml), retinoic acid (10 μM), butyric acid (3 mM), or DMSO (2%) added after cell attachment. The cellular expression of c-myc mRNA and G3PDH mRNA was assessed by Northern blot analysis. Lane 1, control; lane 2, IL-2; lane 3, retinoic acid; lane 4, retinoic acid plus IL-2; lane 5, n-butyric acid; lane 6,n-butyric acid plus IL-2; lane 7, DMSO; and lane 8, DMSO plus IL-2. The c-myc mRNA band (mean ± S.E.M. for six cultures) was quantified by laser densitometry of autoradiograms and expressed as the percentage of control cells. c-mycmRNA was 100 ± 17% in control, 240 ± 43% with IL-2 (P < .05 versus control), 94 ± 15% with retinoic acid, 93 ± 22% with retinoic acid and IL-2, 68 ± 18% with n-butyric acid, 91 ± 5% withn-butyric acid and IL-2, 73 ± 4% with DMSO, and 84 ± 22% with DMSO and IL-2).

Total CYP was decreased in hepatocytes that were treated for 24 h with IL-2 (350 U/ml) alone (Fig. 2). This decrease was similar whether the IL-2 treatment was started either immediately or 24 h after cell attachment (Fig. 2). Retinoic acid,n-butyric acid, and DMSO all prevented the IL-2-mediated decrease in total CYP in IL-2-treated cells, although none of these agents had significant effects on CYP when added alone (Fig. 2).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Prevention of the IL-2-mediated decrease in total CYP by retinoic acid, n-butyric acid, or DMSO. Hepatocytes were treated for 24 h with IL-2 (350 U/ml), retinoic acid (RA) (10 μM), n-butyric acid (BUT) (3 mM), or DMSO (2%) added either after cell attachment or 24 h after cell attachment. Total cellular CYP (mean ± S.E.M. for five cultures) was measured from the absorbance of its CO complex (P-450). Significantly different from control, *P < .05, **P < .01.

IL-2 caused a nonsignificant decrease in CYP1A2 but significant decreases in CYP1A1, CYP2B1/2, CYP2C11, CYP2D1, and CYP3A or erythromycin demethylation (Table 1). Retinoic acid, n-butyric acid, and DMSO all prevented these IL-2-mediated decreases, although none of these agents had significant effects when added alone (Table 1).

Prevention by Genistein.

In lymphocytes or hepatocytes, the IL-2/IL-2R signal starts with the activation of tyrosine kinases, which are inhibited by genistein (Nishio et al., 1994; Tinel et al., 1995). Genistein (20 μg/ml) prevented the IL-2-mediated increase in c-myc mRNA and decrease in total CYP in IL-2-treated cells, although genistein alone had no significant effects on c-mycmRNA or CYP (Fig. 3).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Prevention of the IL-2-mediated increase in c-myc mRNA and decrease in total CYP by genistein. Hepatocytes were treated for 24 h with IL-2 (350 U/ml) and/or genistein (20 μg/ml) added after cell attachment. The expression of c-myc mRNA was assayed by Northern blot analysis. Total cellular CYP was measured from the absorption of its CO complex (P-450) in dithionite-reduced cell lysates. Results are mean ± S.E.M. for three cultures. *Significantly different from control,P < .05.

Prevention by an Antisense c-myc P(S)ODN.

A frequent mechanism of action of antisense ODNs is the rapid breakdown of the transcript due to the degradation of the mRNA/DNA duplex by cell RNase H (Scanlon et al., 1995). A c-myc antisense P(S)ODN markedly decreased c-myc mRNA in IL-2-treated cells (Fig.4 and Table2). Changes in c-Myc protein were about half the changes in c-myc mRNA (Fig. 4 and Table 2). IL-2 increased c-myc mRNA by 114%, although it increased c-Myc protein by 49%; the antisense P(S)ODN decreased c-myc mRNA by 71% in IL-2-treated cells, whereas this antisense caused a 42% decrease in c-Myc protein in these cells (Fig. 4 and Table 2). Thus, the antisense P(S)ODN just normalized the c-Myc protein in IL-2-treated cells, leaving residual levels that were not significantly lower than those in control cells (Fig. 4 and Table 2).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

Assessment of c-myc mRNA, c-Myc protein, and CYP3A protein in hepatocytes cultured with or without IL-2 and a sense or antisense c-myc P(S)ODN. After cell attachment, hepatocytes were cultured for 24 h with DOTAP and various agents. Lane 1, control; lane 2, IL-2 (350 U/ml); lane 3, IL-2 plus sense P(S)ODN (2 μM); and lane 4, IL-2 plus antisense P(S)ODN (2 μM). Densitometric quantifications for five successive cultures are reported in Table 2.

Concomitant with the prevention of the IL-2-mediated c-Myc overexpression, the anti-sense P(S)ODN mostly prevented the effects of IL-2 on CYPs (Fig. 4 and Table 2). Total CYP and CYP2C11 or CYP3A mRNAs and proteins were decreased in cells treated with IL-2 alone, but they were not significantly decreased in cells treated with both IL-2 and the antisense P(S)ODN (Table 2). In these cells, mean values for CYP2C11 mRNA, CYP3A mRNA, CYP2C11 protein, CYP3A protein, and total CYP were 61%, 80%, 74%, 98%, and 100%, respectively, of those in control cells (Table 2).

In contrast, the sense P(S)ODN had none of these preventive effects (Fig. 4 and Table 2). Instead, it even caused a further decrease in CYP2C11 mRNA (Table 2), possibly because the CYP2C11 mRNA contains a 5′-UUGAGG-3′ sequence (third exon) that might weakly hybridize with the 3′-AACTCC-5′ sequence of the sense c-myc P(S)ODN.

Discussion

It has been shown previously that the IL-2R is expressed by rat hepatocytes and that the IL-2/IL-2R interaction down-regulates the expression of CYPs in these cells (Tinel et al., 1995). However, the ultimate signal or signals involved in the down-regulation of CYPs have not been determined (Tinel et al., 1995).

In the present study, we show that the IL-2/IL-2R interaction causes the overexpression of c-myc mRNA and c-Myc protein in rat hepatocytes (Fig. 4 and Table 2). Three different pieces of evidence support a major role of c-myc overexpression in the IL-2-mediated CYP down-regulation. First, genistein, a tyrosine kinase inhibitor that blocks the initial IL-2R signal transduction (Nishio et al., 1994; Tinel et al., 1995), prevented the IL-2-mediated overexpression of c-myc mRNA and down-regulation of CYP (Fig. 3). Second, three agents (retinoic acid, n-butyric acid, and DMSO) that block c-myc transcription (Westin et al., 1982; Dony et al., 1985; Krumm et al., 1992; Strobl and Eick, 1992; Heruth et al., 1993) prevented the IL-2-mediated overexpression of c-myc mRNA and down-regulation of CYP (Figs. 1 and 2 and Table 1). Finally, an antisense c-myc P(S)ODN that may cause the rapid breakdown of the c-myc transcript due to the degradation of the mRNA/DNA duplex by cell RNase H (Scanlon et al., 1995) prevented the IL-2-mediated increase in c-myc mRNA and c-Myc protein and totally or mostly prevented the IL-2-mediated decreases in total CYP, CYP mRNAs, and CYP proteins (Fig. 4 and Table2). Thus, prevention of the IL-2-mediated c-myc overexpression by five different agents, acting at three different levels (signal transduction, c-myctranscription, c-myc mRNA stability), always prevented CYP down-regulation, suggesting a major role of c-mycoverexpression in IL-2-mediated CYP down-regulation.

In cells treated with IL-2 and either retinoic acid,n-butyric acid, DMSO or the c-myc antisense P(S)ODN, none of the CYP proteins were statistically different from values in untreated cells (Tables 1 and 2). However, although the mean values for CYP1A1, CYP1A2, CYP2B1/2, CYP2D1, or CYP3A ranged from 81% to 141% of control values in these protected cells, CYP2C11 values tended to remain slightly low, at 74%, 82%, 99%, and 74% of control values in cells protected by retinoic acid, n-butyric acid, DMSO, or the c-myc antisense P(S)ODN, respectively (Tables 1and 2). Thus, we cannot exclude that in addition to the major role of c-myc overexpression suggested by the present study, some other signal or signals may be also involved in the IL-2-mediated down-regulation of some particular CYPs, including CYP2C11.

In previous studies, c-Myc has been shown to repress the transcription of several genes, including those of major histocompatibility complex class I molecules, albumin, leukocyte function antigen-1α, neural cell adhesion molecule, the adenovirus-2 major late promoter, and the CCATT/enhancer binding protein α (C/EBPα) (Roy et al., 1993; Li et al., 1994; Antonson et al., 1995). Because several of these genes contain a TCA(+1) transcription initiator element (Inr), it has been initially suggested that Myc (alone or in combination with other factors) may interact with the Inr sequence to repress transcription (Li et al., 1994). However, it was shown later that mutation of the Inr site did not abolish these effects, suggesting that Myc may instead (or also) interact with other component or components of the basal transcription machinery (Antonson et al., 1995).

In addition to these direct effects, Myc may act indirectly, by inhibiting C/EBP-dependent transactivation (Mink et al., 1996). Myc has two effects on C/EBPs (Mink et al., 1996). First, it suppresses C/EBPα and C/EBPβ mRNA levels (Mink et al., 1996). Second, Myc suppresses C/EBP-mediated transactivation even when the amount of the C/EBP transcription factors remains constant (e.g., in cells transfected with a constitutively active C/EBPβ expression vector) (Mink et al., 1996). C/EBPs are important transcription factors in hepatocytes, and the 5′-flanking region of most CYPs, including those investigated in the present study, contains possible C/EBP-binding sequences. Indeed, C/EBPα-responsive elements have been recently implicated in the regulated expression of CYPs3A (Ourlin et al., 1996), CYP2B1 and CYP2B2 (Luc et al., 1996), or CYP2C11 and CYP2C12 (Buggs et al., 1998).

Interestingly, the direct or indirect role of c-mycoverexpression in CYP repression may also apply to several other conditions. First, ongoing studies in our laboratory show that the interaction of epidermal growth factor with its receptor also increases c-myc expression and decreases CYP in rat hepatocytes. These effects are prevented by genistein, retinoic acid, and DMSO. Second, several other cytokines, including IL-1, IL-6, tumor necrosis factor-α, and interferons, cause both c-myc overexpression in lymphocytes (Kovacs et al., 1986; Lomo et al., 1991; Escriou et al., 1992; Chaouchi et al., 1994; Georgakopoulos et al., 1996) and CYP down-regulation in hepatocytes (Abdel-Razzak et al., 1993; Chen et al., 1995). Finally, overexpression of c-myc, down-regulation of C/EBPα, and down-regulation of CYPs occur in neoplastic hepatocytes (Wiebel et al., 1984; Fujiwara et al., 1993; Habib et al., 1994, Jover et al., 1998) or the regenerating rat liver (Liddle et al., 1989;Mischoulon et al., 1992; Habib et al., 1994). Conceivably, the c-myc-induced down-regulation of CYPs in dividing hepatocytes may offer a survival advantage by decreasing the formation of DNA-damaging electrophilic metabolites during the critical period of DNA replication. Interestingly, forced reexpression of C/EBPα (through stable transfection with a C/EBPα vector containing a zinc-inducible metallothionein promoter) was shown to up-regulate several CYP mRNAs in HepG2 cells (Jover et al., 1998). This observation strengthens the view that c-myc overexpression may down-regulate CYP expression by first down-regulating C/EBPα expression.

IL-2 immunotherapy is used in patients with disseminated cancer (Lotze et al., 1985). However, high IL-2 doses are required, causing high serum IL-2 concentrations (4000 U/ml) (Lotze et al., 1985). In the present study, 350U/ml IL-2 caused significant CYP down-regulation in rat hepatocytes (Table 1). Thus, human IL-2 concentrations may be sufficiently high to cause CYP down-regulation. Indeed, IL-2 immunotherapy decreased hepatic CYP proteins and monooxygenase activities in patients with cancer (Elkahwaji et al., 1999), suggesting the possibility of drug interactions.

In conclusion, the IL-2/IL-2R interaction causes c-mycoverexpression and CYP down-regulation in cultured rat hepatocytes. CYP down-regulation is prevented by five different agents that prevent c-myc overexpression at three different levels (initial IL-2 signal transduction, c-myc transcription, c-mycmRNA stability), suggesting a major role of c-mycoverexpression in IL-2-mediated CYP down-regulation. It remains to be determined whether c-myc overexpression acts directly on CYP expression and/or first inhibits the expression and transactivating effects of C/EBPα.