ARCHIVES
OF BIOCHEMISTRY
AND
BIOPHYSICS
Vol. 298, No. 1, October, pp. 159-166, 1992
Constitutive Expression and Hormonal Regulation of Male Sexually Differentiated Cytochromes P450 in Primary Cultured Rat Hepatocytes’ Christopher Department
Liddle,2
Agneta
of Medical Nutrition,
Received January
Mode,3 Catherine Karolinska
Institute,
Legraverend, Huddinge
Press,
Gustafsson
Hospital, Nouum F60, Huddinge
S-141 86, Sweden
22, 1992, and in revised form May 27, 1992
Experiments, predominantly performed in uiuo, have shown that the pattern of growth hormone (GH) release from the pituitary gland is a major regulator of sexspecific cytochromes P450 (P450) in rats and other rodents. However, difficulty in constitutively expressing male-specific forms of P450 using in vitro models, such as primary cell culture, has impeded efforts to examine the direct actions of hormones on these enzyme forms. In the present study mRNA species for the malespecific P450 2Cll and 2C13, but not 3A2, were successfully expressed in primary hepatocytes cultured on a laminin-rich extracellular matrix (matrigel) in a serum-free, chemically defined medium containing insulin as the only hormone. When cells were exposed to GH (100 rig/ml), 2C 11 mRNA expression was virtually abolished and 2C13 expression decreased to approximately 50% of control values, demonstrating that the negative regulation of these P450 forms by GH is a direct action on hepatocytes. Dexamethasone (DEX, lO-‘M) increased the expression of 2Cll to 195% while decreasing 2C13 expression to 25% of control values. When GH and DEX were administered concurrently 2Cll was downregulated, indicating that GH is the dominant regulatory hormone for this form. L-Triiodothyronine (T3) (lo-’ M) suppressed 2Cll (46% of control) but had no effect on 2C13 mRNA expression. The positive regulatory effect of glucocorticoids on 2Cll was also found to occur in vivo and demonstrated to operate predominantly at the transcriptional level. This study demonstrates that primary cultures of hepatocytes are a suitable in vitro model for studies on regulation of some male-specific P450 forms and that GH, DEX, and T3 act directly on hepatocytes at a preo 1992 ACTtranslational level to regulate these forms. demie
University
and Jan-Ake
Inc.
0003-9S61/92 $5.00 Copyright 0 1992 by Academic Press, All rights of reproduction in any form
Hepatic cytochrome P450 (P450)4 constitutes a family of enzyme forms predominantly located in the smooth endoplasmic reticulum of hepatocytes. These enzymes are active in the metabolism of an extensive range of endogenous and xenobiotic compounds. A P450 can be classified according to if it is constitutive or inducible. Constitutive forms of P450 are found in normal untreated animals, whereas inducible P45Os increase in concentration, often from undetectable levels, following exposure to certain xenobiotic compounds. Constitutive P450 forms appear to be subject to complex hormonal regulation. In rats and some other rodents there is a pronounced sexual dimorphism of hepatic oxidative metabolism due to differential expression of P450 forms between the sexes. Previous studies in rats, predominantly performed in uiuo, have indicated that the pattern of growth hormone (GH) release from the pituitary gland is a major determinant of this sex difference (l-3). Other hormones, such as glucocorticoids (3) and iodothyronines (3-6), have also been implicated in the regulation of constitutive P45Os. However, a major lim-
i This research was supported by a Swedish Medical Research Council grant (03X-06807) and the Magn. Bergvall foundation. ’ Recipient of a J. J. Billings Fellowship, Royal Australasian College of Physicians; B. J. Amos Fellowship, Westmead Association, Westmead Hospital, Sydney, Australia; and a Visiting Scientist Scholarship, Karolinska Institute, Sweden. Present address: Department of Gastroenterology and Hepatology, Westmead Hospital, Westmead NSW 2145, Australia. a To whom correspondence should be addressed. 4 Abbreviations used: DEX, dexamethasone; GH, growth hormone; hGH, human growth hormone; P450, cytochrome P450; 2Cl1, P450 2Cll; 2C13, P450 2C13; 3A2, P450 3A2; [32P]UTP, uridine [a“Pltriphosphate; SDS, sodium dodecyl sulfate; Ta, L-triiodothyronine; TAT, tyrosine aminotransferase; tNA, total nucleic acids, [%]UTP, uridine 5’-[a-%]triphosphate; Hx, hypophysectomized, Ax, adrenalectomized.
159 Inc. reserved.
160
LIDDLE
itation of in uiuo models is the problem of determining whether the actions of hormones on individual hepatic P450 enzymes are direct or mediated via indirect, extrahepatic mechanisms. Difficulty in constitutively expressing mRNA or protein (7,8) for male-specific forms of P450 using in uitro models, such as primary cell culture, has impeded efforts to examine the direct actions of hormones on these enzyme forms. Previous investigators have succeeded in expressing mRNA for two closely related P450 cytochromes, forms 2Bl and 2B2, in cultured hepatocytes following phenobarbital treatment and demonstrated downregulation by GH (9,10). However, the relevance of GH action on P450 following xenobiotic induction for regulation of constitutively expressed P45Os by GH is open to question. Moreover, some male-specific P450 forms, such as 2Cll (11) and 2C13 (12), are not subject to induction by any known xenobiotic and cannot be studied using this approach. It has previously been demonstrated that when primary hepatocytes are cultured on a laminin-rich basement membrane matrix (matrigel), the cells maintain the ability to express specialized hepatic proteins, including albumin (13,14), and following the addition of GH to the culture medium, the constitutive female-specific P450 form 2C12 (7, 15). The aims of the present study were to determine if primary cultures of hepatocytes, such as the system previously used for the expression of P450 2C12, could be used for studies on some male-specific rat P450 enzyme forms (2Cll,2C13, and 3A2) as measured by steady-state mRNA concentrations and with reference to the direct regulatory actions of GH, glucocorticoids, and iodothyronines on these enzymes. Furthermore, the actions of glucocorticoids, which were found to regulate both 2Cll and 2C13 in vitro, were examined in intact, adrenalectomized (Ax) and hypophysectomized (Hx) rats. MATERIALS
AND
METHODS
Animals and muterzids. For preparationof primary hepatocytes,adult male Sprague-Dawleyrats, approximate weight 300 g, were obtained from Alab (Stockholm, Sweden). For in uiuo experiments, male SpragueDawley rats, Hx at 49 days of age, and male age-matched non-Hx rats were obtained from Mallegaards Avlslaboratorium (Skensved, Denmark). The animals were maintained under conditions of constant temperature and humidity and allowed chow and water ad libitum. Completeness of hypophysectomy was determined by monitoring weight gain for 2 weeks prior to the commencement of hormone treatment. Adrenalectomy was performed under ether anesthesia via a dorsal approach. Ax animals received drinking water containing 0.07 M NaCl to compensate for excessive urinary sodium loss. For the in uiuo study, animals, n = 4 per treatment, were divided into six groups. Animals not receiving hormone treatment were injected subcutaneously with an equal volume of 0.15 M NaCl to act as a methodological control. Group 1 (control), non-Hx rats were injected with 0.15 M NaCl daily (at approx 1800 h) for 4 days prior to sacrifice. Group 2, non-Hx rats were Ax, allowed to recover for 3 days, and then administered 0.15 M NaCl injections as above. Group 3, Hx rats received0.15 M NaCl injections as above. Group 4, Hx rats received 1 mg/kg body
ET AL. weight dexamethasone sodium phosphate (DEX) (Decadron, Merck, Sharp and Dohme) daily by subcutaneous injection. Group 5, Hx rats received recombinant human GH (hGH) by constant infusion for 7 days using a subcutaneously implanted osmotic minipump (Alzet Model 2001, Alza Corp., Palo Alto, CA). Group 6, Hx rats received a combination of DEX and hGH treatments as above. All animals were sacrificed by decapitation and the livers removed promptly for the preparation of cellular fractions. Liver samples for the preparation of total nucleic acids (200 mg) were homogenized in 4 ml of ice-cold buffer (see below) for IO-20 s using a Polytron tissue homogenizer (Kinematica GMBH, Luzern, Switzerland) and stored at -20°C. Collagenase (type XI), hormones, cell culture medium components, and other biochemicals were of cell culture grade and purchased from Sigma Chemical Company (St. Louis, MO) unless otherwise specified. L-Glutamine and minimum essential medium vitamins were obtained from Gibco BRL (Grand Island, NY). Recombinant bovine GH was a generous gift from American Cyanamid Corporation (Wayne, NJ) and recombinant hGH was a generous gift from Kabi AB (Stockholm, Sweden). Proteinase-K was obtained from Merck (Darmstadt, Germany) and RNase A and RNase Tl were from Boehringer-Mannheim (Mannheim, Germany). Restriction endonucleases, ligase, plasmid vectors, and reagents for in vitro transcription of cRNA probes were supplied by Promega Biotech (Madison, WI). [35S]UTP (>lOOO Ci/mmol) and [32P]UTP (400 Ci/mmol) were from Amersham International plc (Buckinghamshire, UK). Hepatocyte isolation and cell culture. Matrigel was prepared from Engelbreth-Helm-Swam sarcoma propagated in C57BL/6 female mice as previously described (16), and stored at -20°C. After thawing on ice, 400 pl of matrigel was evenly applied to 60-mm-diameter plastic culture dishes (A/S Nunc, Roskilde, Denmark) and allowed to gel at room temperature. Hepatocytes were isolated by nonrecirculating collagenase perfusion through the portal vein of ether-anesthetized rats according to the method of Bissell and Guzelian (16). Cells, 3.5 X lOa per plate, viability 8590% as determined by trypan blue exclusion, were applied in 3 ml of modified Waymouth medium (16) containing insulin (1 pg/ ml) as the only hormone. Cultures were maintained in an incubator at 37°C in an atmosphere containing 5% COx. Medium was replaced daily, commencing 24 h after the cells were plated. Hormones, alone or in combination, were added in medium after passage through a 0.2-pm filter, 2 h after plating. The final hormone concentrations were GH, 100 (T3), lo-’ M. Hormones were rig/ml; DEX, 10-e M; L-triiodothyronine renewed daily following change of medium. Cells were harvested at various times after plating. Medium was aspirated from culture dishes and replaced with 2 ml ice-cold phosphateacid (EDTA), pH 7.4. buffered saline, 5 mM ethylenediaminetetraacetic Cells and matrigel were scraped from the plates using a rubber spatula and transferred to 15-ml capped plastic tubes, then allowed to stand on ice for 45 min to dissolve the matrigel. Cells were then collected by centrifugation at 750g for 5 min, lysed in 4 ml of a buffer consisting of 1% (w/v) sodium dodecyl sulfate (SDS), 10 mM EDTA, and 20 mM Tris-HCl, pH 7.5, and stored at -20°C. Solution hybridization. Total nucleic acids (tNA) were prepared by digestion of the cell lysate with proteinase K followed by phenol-chloroform extraction as previously described (17). The concentration of tNA in the samples was determined spectrophotometrically and the DNA concentration was quantified using a specific fluorometric method (18). Total RNA concentration was determined by calculating the difference between the tNA and DNA concentrations. Abundance of the respective mRNA for P450 2Cl1, 2C12, and 3A2 was determined using [3SS]UTP-labeled cRNA probes transcribed in uitro from cDNA templates essentially according to the method of Melton et al. (19). All probes were directed against the 3’ noncoding region of the respective P450 mRNA as the sequence homology between related enzyme forms exhibits increased variability within this region. The cDNA constructs used as templates were as follows: P450 2Cl1, a 205bp BamHI-EcoRI fragment encompassing bases 1580-1884 of the
HORMONAL
REGULATION
full-length cDNA (20); P450 2C13, a 189-bp KpnI-EcoRI fragment encompassing bases 1537-1720 of the full-length cDNA (21); P450 3A2, oligonucleotides corresponding to bases 1689-1738 of the published sequence (22), synthesized with Hind111 and EcoRI ends using an Applied Biosystems 380B DNA synthesizer (Applied Biosystems, Foster City, CA) and annealed prior to ligation. Constructs for all three P45Os were ligated into corresponding restriction sites in the polylinker region of the pGEM-3Z plasmid vector. Optimal assay conditions with regard to temperature were determined for each probe. Hybridization of aliquots of tNA samples was performed in 40 ~10.6 M NaCl, 22 mM Tris-HCl (pH 7.5), 5 mM EDTA, 0.1% SDS 20% formamide with approximately 20,000 (w/v), 1 mM dithiothreitol, cpm of probe per incubation. After overnight incubation, at 75°C for 2Cll and 2C13, 65°C for 3A2, the samples were exposed to RNase A and RNase Tl. Hybrids were then precipitated by the addition of 100 ~16 M trichloroacetic acid and collected on glass-fiber filters (Whatman GF/C, Clifton, NJ), and the specific activity was determined by liquid scintillation spectrometry using a Wallac 1410 counter (LKB-Wallac, Turku, Finland) after the addition of 4 ml scintillant (Ready-Safe, Beckman Instruments Inc., Fullerton, CA). Standard curves were constructed for each assay using tNA extracted from normal male rat liver. All determinations of mRNA for experimental samples were within the linear region of the standard curve. An internal standard, consisting of diluted tNA obtained from a normal male rat, was used in each assay to allow comparisons of results between assays. Potential cross-reactivity of P450 2C12 with the probe for 2C13, due to the high sequence homology between these two forms (87% for the probe region), was excluded by performing a solution hybridization using the 2C13 probe and tNA samples from normal female rat liver, which is known to contain substantial amounts of mRNA for form 2C12. The 2Cll and 3A2 probes shared relatively low sequence homology with other closely related forms of P450 (approximately 50%). Cells from five culture plates were pooled for each experimental point so as to reduce intersample variability. All assays were performed in triplicate and the results expressed as means + SD. Cell experiments were performed at least twice, with cells obtained from different rats, to ensure reproducibility. Each figure represents the data from one of these experiments. Western blotting for P4.50 2Cll. Hepatocytes pooled from five culture plates were lysed by sonication for 5 s, lo-pm amplitude (MSE ultrasonic disintegrator, MSE Scientific Instruments, Crawley, U.K.), in 0.5 ml of a buffer consisting of 10 mM potassium phosphate, 1 mM EDTA, and 0.25 M sucrose, pH 7.4, and the volume was made up to 3 ml with the same buffer. Microsomes were prepared from the cell lysate by sequential centrifugation at 4°C using a Beckman TL-100 benchtop ultracentrifuge and a TLA-100.3 rotor (Beckman Instruments Inc.) as follows: 10,OOOg for 20 min followed by transfer of the supernatant to a new centrifuge tube and centrifugation at 105,OOOgfor 60 min. The microsomal pellet was resuspended in 0.2 ml of a buffer consisting of 50 mM potassium phosphate, 1 mM EDTA, and 20% glycerol using a l-ml motor-driven Potter-Elvehjem homogenizer and stored at -70°C. A small aliquot of the suspension was used for determination of protein concentration using the method of Bradford (23). Western blotting for 2Cll was carried out essentially as described by Morgan et al. (24). Microsomal protein samples, 20 /lg from both control and GH-treated cells, were reduced and then separated on a 7.5% SDSpolyacrylamide gel, electrophoretically transferred to a nitrocellulose filter, and probed with a specific monoclonal antibody for 2Cl1, designated M16. Blots were visualized by incubating with an anti-mouse second antibody coupled to alkaline phosphatase followed by exposure to a substrate (ProtoBlot AP, Promega Biotech). Nuclear run-on analysis. Preparation of nuclei from liver samples as well as run-on assay of in vitro transcription of 2Cll was performed as previously described (25). Nascent [32P]UTP-labeled RNA was hybridized to full-length cDNA probes immobilized on a nylon filter (Gene Screen Plus, New England Nuclear). Nonspecific binding of related P450
OF CYTOCHROME
161
P450
forms was removed by washing the filters under conditions of high stringency. A pGEM 32 plasmid containing no insert acted as a negative control and results were normalized against the transcription signal for P-actin. Tyrosine aminotransferase (TAT), a generous gift from Professor Gunther Schutz (Heidelberg, Germany), was included to act as a positive control for the action of DEX and P450 2C12 for the action of infused GH (25). Densitometric scans of the autoradiographs were performed to obtain numerical data. Statistical analysis. All results are expressed as means + SD. Comparisons between treatment groups for the in uiuo study were carried out using the Student-Newman-Keuls test following single-factor analysis of variance. RESULTS
Cell Morphology, Constitutive Expression of MaleSpecific P450, and Total RNA Content in Cultured Hepatocytes Throughout the period in primary culture, hepatocytes retained a differentiated oval-shaped appearance when viewed by phase-contrast light microscopy. Hepatocyte mRNA concentrations for all three male-specific P45Os decayed rapidly during the first 24 h in culture (Fig. 1). However, by Day 3 the constitutive expression of 2C13 mRNA had begun to increase, and by Day 7 the level was 70% of that observed in livers from normal male rats. Expression of 2Cll mRNA remained low until Day 3, but then increased, such that by Day 5 the level was 21% of normal. No further increase in 2Cll mRNA occurred after Day 5. The expression of P450 2Cll protein in microsomal fractions isolated from cells in culture for 6 days was confirmed by Western blotting (Fig. 2). In contrast, levels of
120
1
F 3 E
1 a
E
0
1
2
3
4
5
6
7
TIME IN CULTURE (days) FIG. 1. Expression of mRNA for P450 2Cll (0), 2C13 (Cl), and 3A2 (0) as determined by solution hybridization in primary cultures of male rat hepatocytes harvested over a 7-day period. Cells were maintained on a matrigel substratum in modified Waymouth medium. Each experimental point represents the mean of triplicate determinations of total nucleic acid samples from five pooled culture dishes. Values are expressed as a percentage of the respective normal mRNA level as determined in liver samples from three adult male rats.
162
LIDDLE
C
GH
DEX
FIG. 2. Western blot analysis of P450 2Cll protein in microsomes from hepatocytes maintained in the absence (C) or presence of growth hormone (GH) (100 rig/ml) or dexamethasone (DEX) (10-s M) for 6 days.
3A2 mRNA continued to decline over the experimental period such that mRNA was undetectable by Day 5. Thus, the culture system used in the present study proved suitable for the further study of hormonal regulation of P450 2Cll and 2C13 but not 3A2. Total cellular RNA concentrations, expressed as the RN&DNA ratio, increased over the first 5 days, being approximately twofold greater in cells harvested on Day 5 compared to those harvested on Day 1 (Fig. 3). Treatment with GH tended to suppress RNA levels when compared to untreated control cells, though this effect was relatively minor. However, because of the observed changes in the RNA:DNA ratio with both time and treatment, all results pertaining to the expression of specific mRNA species following hormone treatment were corrected for the total RNA content of the sample rather than for the DNA content.
ET AL.
duction in immunoquantifiable protein for 2Cll was also observed following GH treatment (Fig. 2). In contrast, GH only reduced 2C13 mRNA to approximately 50% of that found in untreated control cells. However, when cells were cultured in the presence of DEX, 2C13 mRNA levels were markedly reduced. This resulted in 2C13 mRNA being 33 and 25% of control on Days 5 and 7, respectively. Interestingly, the negative regulatory effect of DEX on 2C13 was more profound than that of GH. DEX treatment caused a moderate increase in 2Cll mRNA expression, being 137% of control on Day 5 and 195% on Day 7. This effect of DEX was also reflected at the protein level (Fig. 2). Treatment of cells with T3 depressed 2Cll expression to 46 and 66% of control on Days 5 and 7, but had little effect on the expression of mRNA for 2C13. When hepatocytes were treated with combinations of hormones the following patterns emerged. GH exerted the dominant regulatory effect on 2Cll expression, resulting in marked downregulation of mRNA for this form in the presence of both DEX and T3. In contrast, the combined action of GH and DEX on 2C13 expression appeared to be additive, resulting in marked suppression of this form. Again T3 was found to have no effect on 2C13. In addition to the continuous treatment regime with GH, various protocols for applying GH in pulses to the hepatocytes were tested to try to induce the expression of P450 2Cll. No treatment proved successful in this respect and 2Cll mRNA levels were invariably suppressed by such manipulations (data not shown). In Vivo Regulation of P450 2Cll and 2C13 by DEX Efficacy of hypophysectomy of rats was confirmed by the failure of Hx animals to gain weight compared to non-
91
0
13
Effect of Hormone Treatment on Male-Specific P450 mRNA Expression After 1 day in culture, hormone treatment had no clear effect on P450 mRNA expression in hepatocytes. However, by Day 3 an effect of treatments on both 2Cll and 2C13 could be discerned, patterns that were firmly established by Days 5 and 7 (Fig. 4). This indicated that the cells required an adaptation period following inoculation into the culture system, an observation made previously (14). The presence of GH in the culture medium almost completely suppressed the expression of 2Cll mRNA in hepatocytes when measured on Days 3, 5, and 7. A re-
//?
3J 0
t
l
2
3
4
5
6
7
TIME IN CULTURE (days) FIG. 3. Total cellular RNA concentrations over culture age, expressed as the RNA:DNA ratio, in hepatocytes maintained in the absence (0) or presence of growth hormone, 100 rig/ml (0).
HORMONAL
REGULATION
n CONTROL Cl GH 0 I
DEX
OF CYTOCHROME B
163
P450
300
n CONTROL 0 0
250 I
q T3 q GH+DEX
60
K
z E 8
GH DEX
60 40
5 TIME IN CULTURE (days)
7
TIME IN CULTURE (days)
FIG. 4. Expression of mRNA for P450 2Cll (A) and P450 2C13 (B) as determined by solution hybridization in hepatocytes treated with hormones (alone or in combination) and harvested at 5 and 7 days of culture age. Each experimental point represents the mean f SD of triplicate determinations of total nucleic acid samples from five pooled culture dishes. GH, growth hormone (100 rig/ml); DEX, dexamethasone (lo-* M); T3, L-triiodothyronine (low9 M).
Hx, control animals: mean weight gain for Hx rats was 1.9 f 1.3 g/week; for control rats, 35.9 f 2.2 g/week. Ax animals had diminished weight gain (12.6 +- 4.2 g/week) compared to control rats. As expected, hypophysectomy resulted in a marked fall in the steady-state expression of mRNA for 2Cl1, being 5% of that in livers from intact male rat controls (P < 0.001). Treatment of Hx rats with DEX resulted in a fourfold increase in mRNA concentrations for 2Cl1, such that expression was 21% of that in intact controls (P < 0.001). When endogenous glucocorticoid secretion was abolished by adrenalectomy in otherwise intact rats, 2Cll mRNA levels decreased to 80% of that in control animals (P < 0.001). When Hx animals treated with DEX also received a continuous hGH infusion via a subcutaneously implanted osmotic minipump, 2Cll expression was decreased so as to resemble that in Hx rats receiving a hGH infusion alone, demonstrating that GH downregulation takes precedence over positive regulation by glucocorticoids. These data are summarized in Fig. 5. Despite the marked downregulation of P450 2C13 mRNA by DEX in the primary cultured hepatocytes, no effect of adrenalectomy or DEX treatment on this P450 form was observed in viuo (data not shown). To determine the level at which glucocorticoids were regulating 2Cll expression, in vitro transcriptional activity of this gene was determined by run-on analysis in nuclei obtained from the livers of animals subjected to various treatments. Nuclei from two animals of each treatment were analyzed; Fig. 6 shows the autoradiography result from one rat analyzed from each group. Following normalization of the numerical data against Pactin we found that DEX treatment of Hx rats resulted
in a fourfold increase in 2Cll transcription, directly paralleling the increase observed in steady-state mRNA concentrations. In contrast, no major effect of adrenalectomy on 2Cll transcription could be discerned, activity being 95% of that found in control animals. hGH infusion totally abolished 2Cll transcription in keeping with our previous findings (25). TAT transcriptional activity, included as a positive control for glucocorticoid action, was reduced
DEX OH,. ...r
DEX WI,,,
FIG. 5. Expression of mRNA for P450 2Cl1, expressed per microgram of DNA, in livers of adult male rats subjected to various treatments. CONT, control; Ax, adrenalectomized; Hx, hypophysectomized; Hx DEX, hypophysectomized, treated for 4 days with dexamethasone 1 mg/kg/day; Hx GHmp, hypophysectomized, treated for 7 days with continuous GH infusion by minipump; Hx DEX GHmp, hypophysectomized, treated with a combination of dexamethasone and GH (as above). Results expressed as means + SD, n = 4 animals per group.
164
LIDDLE
pGEM
!
CONT
A,
,’
H,
8 -“..*“, i :;; _
H, H, DEX GH,,
FIG. 6. In vitro elongation of P450 2Cll (2Cll) as determined by runon analysis of bepatic nuclei from male rats subjected to various treatments. Additional probes were included to act as controls: pGEM, pGEM 32 plasmid vector containing no cDNA insert; TAT, tyrosine aminotransferase; 2C12, P450 2C12. CONT, control; Ax, adrenalectomized; Hx, hypophysectomized, Hx DEX, hypophysectomized, treated for 4 days with dexamethasone 1 mg/kg/day; Hx GHmp, hypophysectomized, treated for 7 days with continuous GH infusion by minipump.
following adrenalectomy and increased twofold following DEX treatment of Hx rats. DEX treatment of Hx rats brought TAT activity back to normal levels. P450 2C12 transcriptional activity, included as a positive control for GH action, increased markedly following GH infusion. Interestingly, adrenalectomy alone also increased 2C12 transcription to 40% of that observed in Hx rats receiving GH infusion, suggesting that glucocorticoids may be capable of suppressing this gene. This possibility is supported by our previous observation that corticosterone is capable of decreasing 2Cl2 mRNA expression in cultured rat hepatocytes (15). DISCUSSION The use of primary cultures of hepatocytes as a model to examine the effects of isolated stimuli on liver metabolism is attractive. However, finding appropriate culture conditions that allow expression of specialized hepatic genes has proven difficult. The two predominant variables in a culture system are the type of the attachment surface and the composition of the culture medium. Various groups of investigators have described conditions that allow expression of particular genes or groups of genes, including some forms of P450. Schuetz et al. (14) demonstrated that several forms of P450 could be expressed following xenobiotic induction in hepatocytes maintained on a substratum of matrigel in modified Waymouth medium, a system that was also used in the present study. However, other combinations have also proven to be successful for expression of certain P450 enzymes, including phenobarbital induction of forms 2Bl and 2B2 using type I collagen and either Williams’ E medium (26) or Chee’s medium (8) and 3-methylcholanthrene induction of forms
ET AL.
1Al and lA2 using type 1 collagen and supplemented RPM1 1640 medium (27). A limitation of the above studies is that they have been restricted to P450 enzyme forms inducible by xenobiotics. The only example to date of increased expression of a constitutive hepatic P450 in cell culture following exposure to an endogenous regulating substance is the constitutive female-specific form 2C12 on treatment of hepatocytes with GH (7,15). Some constitutive P450 forms, for example 2C13, have no known positive regulator (12, 28). In uiuo studies have indicated that this male-specific form is repressed by the continuous pattern of GH release from the pituitary in female rats (19, 28). A form that appears to be similarly regulated, 3A2 (29), is also male specific, though this P450 has been demonstrated to be subject to xenobiotic induction by phenobarbital in intact animals (22). The male-specific form, 2Cl1, appears to be under complex regulation by GH. In uiuo studies indicate that the male pulsatile pattern of GH release from the pituitary positively regulates this form, whereas the continuous female pattern of GH secretion results in downregulation (3). In the present study, a constitutive increase in the expression of mRNA was found after 3 days for P450 form 2C13 and after 5 days for 2Cll in hepatocytes cultured on matrigel in modified Waymouth medium. Although the level of expression was less than that in normal male rat liver, 70% for 2C13 and 21% for 2Cl1, this provided a suitable model to study direct hormonal effects on these forms. The reduced level of expression of 2Cll was not surprising given its aforementioned positive regulation by pulses of GH; however, application of GH pulses to the cultured hepatocytes did not increase 2Cll mRNA levels. This indicates either that it was not possible to completely eliminate GH from the culture system between pulses or that additional factors may be required for full expression of this form. In this respect the circulating growth hormone-binding protein found in many species, including the rat (30), may be of importance, though the overall function of this protein remains unknown. The mRNA concentration for P450 3A2 decayed rapidly with time in cultured hepatocytes and did not exhibit the constitutive increase of expression over time seen with the other forms studied. Again, some additional factors appear to be necessary for the expression of this form. Using hepatocytes cultured on matrigel in modified Waymouth medium, Guzelian et al. (7) demonstrated that the upregulation of the female-specific P450 2C12 by GH was due to a direct action of this hormone on hepatocytes. Similarly, we have shown that continuous treatment of hepatocytes with GH acts directly on the cells to demasculinize hepatic oxidative metabolism by downregulating at least two male-specific forms of P450 at a pretranslational level. In the case of P450 2Cll this downregulation was almost complete. Surprisingly, P450 2C13 expression
HORMONAL
REGULATION
was only partially abolished by continuous exposure to GH. Only in the presence of the synthetic glucocorticoid DEX did GH reduce expression of 2C13 mRNA to low levels. Indeed, DEX treatment alone was effective in substantially decreasing expression of 2C13. The significance of these findings is uncertain as earlier in vivo studies have shown that continuous secretion of GH from osmotic minipumps alone is sufficient to suppress this enzyme in Hx animals, where it is expressed at a high level (21). Indeed, DEX treatment of Hx rats had no apparent effect on 2C13, demonstrating at least some divergence between the culture model and the in vivo situation. In the present study, the synthetic glucocorticoid agonist DEX has been demonstrated to increase the expression of 2Cll both in uiuo and in vitro. Moreover, we have demonstrated that this effect occurs predominantly at the transcriptional level. This observation is of interest given that Morishima et al. (31) have described the presence of a glucocorticoid-responsive element in the 5’ flanking region of the gene for P450 2Cll. Evidence for a physiological role of glucocorticoids in 2Cll gene regulation is provided by the decrease in the expression of this gene in rats in which endogenous steroid secretion was removed by adrenalectomy. Interestingly, in absolute terms, the decrease in 2Cll expression following adrenalectomy was approximately the same as the increase in expression in Hx rats treated with DEX. Thus glucocorticoid regulation of the 2Cll gene accounts for 20% of the overall expression of this P450 form in the livers of intact male rats. While this may appear to be a relatively small effect, the role of 2Cll as the dominant P450 form in the male rat plus the involvement of this form in the metabolism of a range of endogenous steroid compounds (32-34) suggest that such a change may well be of importance. An additional finding of the present study is that GH downregulation of 2Cll takes precedence over glucocorticoid effects, in keeping with the view that GH secretion patterns are the predominant regulatory mechanism for 2Cll. The mechanism by which GH exerts regulatory effects on some P450 enzyme forms remains unknown though a recent study from our laboratory indicates that this occurs at the transcriptional level (25). Iodothyronines have previously been reported to contribute to the GH-dependent suppression of 3A2 in male rats (5) and the GH-dependent induction of 2C12 as occurs in female rats (3). In the present study, T3 had no potentiating effect on the ability of GH to downregulate either 2Cll or 2C13 mRNA in primary cultured hepatocytes, though T3 treatment alone appeared to repress 2Cll expression. This further emphasizes the differing hormonal interactions that occur to regulate individual P450 enzyme forms. In summary, the present study demonstrates that primary cultured hepatocytes are a suitable model to examine the regulation of at least some constitutive male-specific
OF CYTOCHROME
165
P450
rat P450 forms. GH has been shown to act directly on hepatocytes at a pretranslational level to downregulate the expression of these forms. Glucocorticoids and iodothyronines also exert direct pretranslational actions on these forms, an effect that was confirmed in uivo for 2Cll. The overall significance of the other hormonal actions with regard to in vivo regulation remains to be determined. ACKNOWLEDGMENTS The authors are indebted to Dr. Anders Strom and Dr. Peter G. Zaphiropoulos for assistance with the probes used in this study and to Eva Floby for her expert technical assistance.
REFERENCES 1. Mode, A., Gustafsson, J.-A., Jansson, J.-O., Eden, S., and Isaksson, 0. (1982) Endocrinology 111, 169221697.
2. Morgan, E. T., MacGeoch, C., and Gustafsson, J.-A. (1985) J. Biol. Chem. 260, 11895511898. E., Strom, A., Zaphiropoulos, P. G., 3. Mode, A., Wiersma-Larsson, and Gustafsson, J.-A. (1989) J. Endocrinol. 120, 311-317. 4. Yamazoe, Y., Murayama, N., Shimada, M., and Kato, R. (1989) Biochem. Biophys. Res. Commun. 160,609-614. 5. Waxman, D. J., Ram, P. A., Notani, G., LeBlanc, G. A., Alberta, J. A., Morrissey,
J. J., and Sundseth, S. S. (1990) Mol. Endocrinol.
4,447-454. 6. Ram, P. A., and Waxman, D. J. (1990) J. Biol. Chem. 265,1922319229.
I. Guzelian, P. S., Li, D., Schuetz, E. G., Thomas, P., Levin, W., Mode, A., and Gustafsson, 9783-9787.
J.-A. (1988) Proc. Natl. Acad. Sci. USA 85,
8. Waxman, D. J., Morrissey, J. J., Naik, S., and Jauregui, H. 0. (1990) Biochem. J. 271,113-119. 9. Schuetz, E. G., Schuetz, J. D., May, B., and Guzelian, P. S. (1990) J. Biol. Chem. 265, 1188-1192.
10. Murayama,
N., Shimada, M., Yamazoe, Mol. Pharmacol. 39, 811-817.
Y., and Kato, R. (1991)
11. Waxman, D. J., Dannan, G. A., and Guengerich, chemistry 24, 4409-4417.
12. Bandiera,
F. P. (1985) Bio-
S., Ryan, D. E., Levin, W., and Thomas, Arch. Biochem. Biophys. 248,658-676.
P. E. (1986)
13. Bissell, D. M., Arenson, D. M., Maher, J. J., and Roll, F. J. (1987) J. Clin. Inuest. 79, 801-812. 14. Schuetz, E. G., Li, D., Omiecinski, C. J., Muller-Eberhard, U., Kleinman, H. K., Elswick, Physiol. 134, 309-323.
B., and Guzelian,
15. Tollet,
P. S. (1988) J. Cell.
P., Enberg, B., and Mode, A. (1990) Mol. Endocrinol. 1934-1942.
4,
16. Bissell, D. M., and Guzelian, P. S. (1980) Ann. N. Y. Acad. Sci. 349, 85-98.
17. Durnham, D. M., and Palmiter, 383-393.
R. P. (1983) Anal. Biochem. 131,
18. Labarca, C., and Paigen, K. (1980) Anal. Biochem. 102, 344-352. 19. Melton, D. A., Krieg, P. A., Rehagliati, M. R., Maniatis, T., Zinn, K., and Green, M. P. (1984) Nucl. Acids Res. 12, 7035-7036. 20. Strom, A., Mode, A., Zaphiropoulos, P., Nilsson, A.-G., Morgan, E., and Gustafsson, J.-A. (1988) Acta Endocrinol. 118, 314-320. 21. Zaphiropoulos, P. G., Strom, A., Robertson, J. A., and Gustafsson, J.-A. (1990) Mol. Endocrinol. 4, 53-58.
166
LIDDLE
22. Gonzalez, F. J., Song, B.-J., and Hardwick, Biol. 6,2969-2976. 23. Bradford,
J. P. (1986) Mol. Cell.
M. M. (1976) Anal. Biochem. 72, 248-254.
24. Morgan, E. T., Ronnholm, chemistry 26,4193-4200.
R., and Gustafsson,
J.-A. (1987) Bio-
25. Legraverend, C., Mode, A., Westin, S., Strom, A., Eguchi, H., Zaphiropoulos, P. G., and Gustafsson, J.-A. (1992) Mol. Endocrinol. 6,259-256. 26. Sinclair, P. R., Bement, W. J., Haugen, S. A., Sinclair, Guzelian, P. S. (1990) Cancer Res. 60,5219-5224. 27. Silver, G., and Krauter, 11807.
J. F., and
K. S. (1988) J. Biol. Chem. 263,
11802-
ET AL. 28. McClellan-Green, P. D., Linko, P., Yeowell, H. N., and Goldstein, J. A. (1989) J. Biol. Chem. 264, 18960-18965. 29. Shimada, M., Nagata, K., Murayama, N., Yamazoe, Y., and Kato, R. (1989) J. Biochem. 106, 1030-1034. 30. Massa, G., Mulumba, N., Ketelslegers, J.-M., and Maes, M. (1990) Endocrinology 126,1976-1980. 31. Morishima, N., Yoshioka, H., Higashi, Y., Sogawa, K., and FujiiKuriyama, Y. (1987) Biochemistry 26,8279-8285. 32. Cheng, K.-C., Shenkman, J. B. (1984) Drug Metab. Dispos. 12,222234. 33. Dannan, G. A., Porubek, D. J., Nelson, S. D., Waxman, D. J., and Guengerich, F. P. (1986) Endocrinology 118,1952-1960. 34. Waxman, D. J. (1988) Biochem. Pharmacol. 37,71-84.
OF BIOCHEMISTRY
AND
BIOPHYSICS
Vol. 298, No. 1, October, pp. 159-166, 1992
Constitutive Expression and Hormonal Regulation of Male Sexually Differentiated Cytochromes P450 in Primary Cultured Rat Hepatocytes’ Christopher Department
Liddle,2
Agneta
of Medical Nutrition,
Received January
Mode,3 Catherine Karolinska
Institute,
Legraverend, Huddinge
Press,
Gustafsson
Hospital, Nouum F60, Huddinge
S-141 86, Sweden
22, 1992, and in revised form May 27, 1992
Experiments, predominantly performed in uiuo, have shown that the pattern of growth hormone (GH) release from the pituitary gland is a major regulator of sexspecific cytochromes P450 (P450) in rats and other rodents. However, difficulty in constitutively expressing male-specific forms of P450 using in vitro models, such as primary cell culture, has impeded efforts to examine the direct actions of hormones on these enzyme forms. In the present study mRNA species for the malespecific P450 2Cll and 2C13, but not 3A2, were successfully expressed in primary hepatocytes cultured on a laminin-rich extracellular matrix (matrigel) in a serum-free, chemically defined medium containing insulin as the only hormone. When cells were exposed to GH (100 rig/ml), 2C 11 mRNA expression was virtually abolished and 2C13 expression decreased to approximately 50% of control values, demonstrating that the negative regulation of these P450 forms by GH is a direct action on hepatocytes. Dexamethasone (DEX, lO-‘M) increased the expression of 2Cll to 195% while decreasing 2C13 expression to 25% of control values. When GH and DEX were administered concurrently 2Cll was downregulated, indicating that GH is the dominant regulatory hormone for this form. L-Triiodothyronine (T3) (lo-’ M) suppressed 2Cll (46% of control) but had no effect on 2C13 mRNA expression. The positive regulatory effect of glucocorticoids on 2Cll was also found to occur in vivo and demonstrated to operate predominantly at the transcriptional level. This study demonstrates that primary cultures of hepatocytes are a suitable in vitro model for studies on regulation of some male-specific P450 forms and that GH, DEX, and T3 act directly on hepatocytes at a preo 1992 ACTtranslational level to regulate these forms. demie
University
and Jan-Ake
Inc.
0003-9S61/92 $5.00 Copyright 0 1992 by Academic Press, All rights of reproduction in any form
Hepatic cytochrome P450 (P450)4 constitutes a family of enzyme forms predominantly located in the smooth endoplasmic reticulum of hepatocytes. These enzymes are active in the metabolism of an extensive range of endogenous and xenobiotic compounds. A P450 can be classified according to if it is constitutive or inducible. Constitutive forms of P450 are found in normal untreated animals, whereas inducible P45Os increase in concentration, often from undetectable levels, following exposure to certain xenobiotic compounds. Constitutive P450 forms appear to be subject to complex hormonal regulation. In rats and some other rodents there is a pronounced sexual dimorphism of hepatic oxidative metabolism due to differential expression of P450 forms between the sexes. Previous studies in rats, predominantly performed in uiuo, have indicated that the pattern of growth hormone (GH) release from the pituitary gland is a major determinant of this sex difference (l-3). Other hormones, such as glucocorticoids (3) and iodothyronines (3-6), have also been implicated in the regulation of constitutive P45Os. However, a major lim-
i This research was supported by a Swedish Medical Research Council grant (03X-06807) and the Magn. Bergvall foundation. ’ Recipient of a J. J. Billings Fellowship, Royal Australasian College of Physicians; B. J. Amos Fellowship, Westmead Association, Westmead Hospital, Sydney, Australia; and a Visiting Scientist Scholarship, Karolinska Institute, Sweden. Present address: Department of Gastroenterology and Hepatology, Westmead Hospital, Westmead NSW 2145, Australia. a To whom correspondence should be addressed. 4 Abbreviations used: DEX, dexamethasone; GH, growth hormone; hGH, human growth hormone; P450, cytochrome P450; 2Cl1, P450 2Cll; 2C13, P450 2C13; 3A2, P450 3A2; [32P]UTP, uridine [a“Pltriphosphate; SDS, sodium dodecyl sulfate; Ta, L-triiodothyronine; TAT, tyrosine aminotransferase; tNA, total nucleic acids, [%]UTP, uridine 5’-[a-%]triphosphate; Hx, hypophysectomized, Ax, adrenalectomized.
159 Inc. reserved.
160
LIDDLE
itation of in uiuo models is the problem of determining whether the actions of hormones on individual hepatic P450 enzymes are direct or mediated via indirect, extrahepatic mechanisms. Difficulty in constitutively expressing mRNA or protein (7,8) for male-specific forms of P450 using in uitro models, such as primary cell culture, has impeded efforts to examine the direct actions of hormones on these enzyme forms. Previous investigators have succeeded in expressing mRNA for two closely related P450 cytochromes, forms 2Bl and 2B2, in cultured hepatocytes following phenobarbital treatment and demonstrated downregulation by GH (9,10). However, the relevance of GH action on P450 following xenobiotic induction for regulation of constitutively expressed P45Os by GH is open to question. Moreover, some male-specific P450 forms, such as 2Cll (11) and 2C13 (12), are not subject to induction by any known xenobiotic and cannot be studied using this approach. It has previously been demonstrated that when primary hepatocytes are cultured on a laminin-rich basement membrane matrix (matrigel), the cells maintain the ability to express specialized hepatic proteins, including albumin (13,14), and following the addition of GH to the culture medium, the constitutive female-specific P450 form 2C12 (7, 15). The aims of the present study were to determine if primary cultures of hepatocytes, such as the system previously used for the expression of P450 2C12, could be used for studies on some male-specific rat P450 enzyme forms (2Cll,2C13, and 3A2) as measured by steady-state mRNA concentrations and with reference to the direct regulatory actions of GH, glucocorticoids, and iodothyronines on these enzymes. Furthermore, the actions of glucocorticoids, which were found to regulate both 2Cll and 2C13 in vitro, were examined in intact, adrenalectomized (Ax) and hypophysectomized (Hx) rats. MATERIALS
AND
METHODS
Animals and muterzids. For preparationof primary hepatocytes,adult male Sprague-Dawleyrats, approximate weight 300 g, were obtained from Alab (Stockholm, Sweden). For in uiuo experiments, male SpragueDawley rats, Hx at 49 days of age, and male age-matched non-Hx rats were obtained from Mallegaards Avlslaboratorium (Skensved, Denmark). The animals were maintained under conditions of constant temperature and humidity and allowed chow and water ad libitum. Completeness of hypophysectomy was determined by monitoring weight gain for 2 weeks prior to the commencement of hormone treatment. Adrenalectomy was performed under ether anesthesia via a dorsal approach. Ax animals received drinking water containing 0.07 M NaCl to compensate for excessive urinary sodium loss. For the in uiuo study, animals, n = 4 per treatment, were divided into six groups. Animals not receiving hormone treatment were injected subcutaneously with an equal volume of 0.15 M NaCl to act as a methodological control. Group 1 (control), non-Hx rats were injected with 0.15 M NaCl daily (at approx 1800 h) for 4 days prior to sacrifice. Group 2, non-Hx rats were Ax, allowed to recover for 3 days, and then administered 0.15 M NaCl injections as above. Group 3, Hx rats received0.15 M NaCl injections as above. Group 4, Hx rats received 1 mg/kg body
ET AL. weight dexamethasone sodium phosphate (DEX) (Decadron, Merck, Sharp and Dohme) daily by subcutaneous injection. Group 5, Hx rats received recombinant human GH (hGH) by constant infusion for 7 days using a subcutaneously implanted osmotic minipump (Alzet Model 2001, Alza Corp., Palo Alto, CA). Group 6, Hx rats received a combination of DEX and hGH treatments as above. All animals were sacrificed by decapitation and the livers removed promptly for the preparation of cellular fractions. Liver samples for the preparation of total nucleic acids (200 mg) were homogenized in 4 ml of ice-cold buffer (see below) for IO-20 s using a Polytron tissue homogenizer (Kinematica GMBH, Luzern, Switzerland) and stored at -20°C. Collagenase (type XI), hormones, cell culture medium components, and other biochemicals were of cell culture grade and purchased from Sigma Chemical Company (St. Louis, MO) unless otherwise specified. L-Glutamine and minimum essential medium vitamins were obtained from Gibco BRL (Grand Island, NY). Recombinant bovine GH was a generous gift from American Cyanamid Corporation (Wayne, NJ) and recombinant hGH was a generous gift from Kabi AB (Stockholm, Sweden). Proteinase-K was obtained from Merck (Darmstadt, Germany) and RNase A and RNase Tl were from Boehringer-Mannheim (Mannheim, Germany). Restriction endonucleases, ligase, plasmid vectors, and reagents for in vitro transcription of cRNA probes were supplied by Promega Biotech (Madison, WI). [35S]UTP (>lOOO Ci/mmol) and [32P]UTP (400 Ci/mmol) were from Amersham International plc (Buckinghamshire, UK). Hepatocyte isolation and cell culture. Matrigel was prepared from Engelbreth-Helm-Swam sarcoma propagated in C57BL/6 female mice as previously described (16), and stored at -20°C. After thawing on ice, 400 pl of matrigel was evenly applied to 60-mm-diameter plastic culture dishes (A/S Nunc, Roskilde, Denmark) and allowed to gel at room temperature. Hepatocytes were isolated by nonrecirculating collagenase perfusion through the portal vein of ether-anesthetized rats according to the method of Bissell and Guzelian (16). Cells, 3.5 X lOa per plate, viability 8590% as determined by trypan blue exclusion, were applied in 3 ml of modified Waymouth medium (16) containing insulin (1 pg/ ml) as the only hormone. Cultures were maintained in an incubator at 37°C in an atmosphere containing 5% COx. Medium was replaced daily, commencing 24 h after the cells were plated. Hormones, alone or in combination, were added in medium after passage through a 0.2-pm filter, 2 h after plating. The final hormone concentrations were GH, 100 (T3), lo-’ M. Hormones were rig/ml; DEX, 10-e M; L-triiodothyronine renewed daily following change of medium. Cells were harvested at various times after plating. Medium was aspirated from culture dishes and replaced with 2 ml ice-cold phosphateacid (EDTA), pH 7.4. buffered saline, 5 mM ethylenediaminetetraacetic Cells and matrigel were scraped from the plates using a rubber spatula and transferred to 15-ml capped plastic tubes, then allowed to stand on ice for 45 min to dissolve the matrigel. Cells were then collected by centrifugation at 750g for 5 min, lysed in 4 ml of a buffer consisting of 1% (w/v) sodium dodecyl sulfate (SDS), 10 mM EDTA, and 20 mM Tris-HCl, pH 7.5, and stored at -20°C. Solution hybridization. Total nucleic acids (tNA) were prepared by digestion of the cell lysate with proteinase K followed by phenol-chloroform extraction as previously described (17). The concentration of tNA in the samples was determined spectrophotometrically and the DNA concentration was quantified using a specific fluorometric method (18). Total RNA concentration was determined by calculating the difference between the tNA and DNA concentrations. Abundance of the respective mRNA for P450 2Cl1, 2C12, and 3A2 was determined using [3SS]UTP-labeled cRNA probes transcribed in uitro from cDNA templates essentially according to the method of Melton et al. (19). All probes were directed against the 3’ noncoding region of the respective P450 mRNA as the sequence homology between related enzyme forms exhibits increased variability within this region. The cDNA constructs used as templates were as follows: P450 2Cl1, a 205bp BamHI-EcoRI fragment encompassing bases 1580-1884 of the
HORMONAL
REGULATION
full-length cDNA (20); P450 2C13, a 189-bp KpnI-EcoRI fragment encompassing bases 1537-1720 of the full-length cDNA (21); P450 3A2, oligonucleotides corresponding to bases 1689-1738 of the published sequence (22), synthesized with Hind111 and EcoRI ends using an Applied Biosystems 380B DNA synthesizer (Applied Biosystems, Foster City, CA) and annealed prior to ligation. Constructs for all three P45Os were ligated into corresponding restriction sites in the polylinker region of the pGEM-3Z plasmid vector. Optimal assay conditions with regard to temperature were determined for each probe. Hybridization of aliquots of tNA samples was performed in 40 ~10.6 M NaCl, 22 mM Tris-HCl (pH 7.5), 5 mM EDTA, 0.1% SDS 20% formamide with approximately 20,000 (w/v), 1 mM dithiothreitol, cpm of probe per incubation. After overnight incubation, at 75°C for 2Cll and 2C13, 65°C for 3A2, the samples were exposed to RNase A and RNase Tl. Hybrids were then precipitated by the addition of 100 ~16 M trichloroacetic acid and collected on glass-fiber filters (Whatman GF/C, Clifton, NJ), and the specific activity was determined by liquid scintillation spectrometry using a Wallac 1410 counter (LKB-Wallac, Turku, Finland) after the addition of 4 ml scintillant (Ready-Safe, Beckman Instruments Inc., Fullerton, CA). Standard curves were constructed for each assay using tNA extracted from normal male rat liver. All determinations of mRNA for experimental samples were within the linear region of the standard curve. An internal standard, consisting of diluted tNA obtained from a normal male rat, was used in each assay to allow comparisons of results between assays. Potential cross-reactivity of P450 2C12 with the probe for 2C13, due to the high sequence homology between these two forms (87% for the probe region), was excluded by performing a solution hybridization using the 2C13 probe and tNA samples from normal female rat liver, which is known to contain substantial amounts of mRNA for form 2C12. The 2Cll and 3A2 probes shared relatively low sequence homology with other closely related forms of P450 (approximately 50%). Cells from five culture plates were pooled for each experimental point so as to reduce intersample variability. All assays were performed in triplicate and the results expressed as means + SD. Cell experiments were performed at least twice, with cells obtained from different rats, to ensure reproducibility. Each figure represents the data from one of these experiments. Western blotting for P4.50 2Cll. Hepatocytes pooled from five culture plates were lysed by sonication for 5 s, lo-pm amplitude (MSE ultrasonic disintegrator, MSE Scientific Instruments, Crawley, U.K.), in 0.5 ml of a buffer consisting of 10 mM potassium phosphate, 1 mM EDTA, and 0.25 M sucrose, pH 7.4, and the volume was made up to 3 ml with the same buffer. Microsomes were prepared from the cell lysate by sequential centrifugation at 4°C using a Beckman TL-100 benchtop ultracentrifuge and a TLA-100.3 rotor (Beckman Instruments Inc.) as follows: 10,OOOg for 20 min followed by transfer of the supernatant to a new centrifuge tube and centrifugation at 105,OOOgfor 60 min. The microsomal pellet was resuspended in 0.2 ml of a buffer consisting of 50 mM potassium phosphate, 1 mM EDTA, and 20% glycerol using a l-ml motor-driven Potter-Elvehjem homogenizer and stored at -70°C. A small aliquot of the suspension was used for determination of protein concentration using the method of Bradford (23). Western blotting for 2Cll was carried out essentially as described by Morgan et al. (24). Microsomal protein samples, 20 /lg from both control and GH-treated cells, were reduced and then separated on a 7.5% SDSpolyacrylamide gel, electrophoretically transferred to a nitrocellulose filter, and probed with a specific monoclonal antibody for 2Cl1, designated M16. Blots were visualized by incubating with an anti-mouse second antibody coupled to alkaline phosphatase followed by exposure to a substrate (ProtoBlot AP, Promega Biotech). Nuclear run-on analysis. Preparation of nuclei from liver samples as well as run-on assay of in vitro transcription of 2Cll was performed as previously described (25). Nascent [32P]UTP-labeled RNA was hybridized to full-length cDNA probes immobilized on a nylon filter (Gene Screen Plus, New England Nuclear). Nonspecific binding of related P450
OF CYTOCHROME
161
P450
forms was removed by washing the filters under conditions of high stringency. A pGEM 32 plasmid containing no insert acted as a negative control and results were normalized against the transcription signal for P-actin. Tyrosine aminotransferase (TAT), a generous gift from Professor Gunther Schutz (Heidelberg, Germany), was included to act as a positive control for the action of DEX and P450 2C12 for the action of infused GH (25). Densitometric scans of the autoradiographs were performed to obtain numerical data. Statistical analysis. All results are expressed as means + SD. Comparisons between treatment groups for the in uiuo study were carried out using the Student-Newman-Keuls test following single-factor analysis of variance. RESULTS
Cell Morphology, Constitutive Expression of MaleSpecific P450, and Total RNA Content in Cultured Hepatocytes Throughout the period in primary culture, hepatocytes retained a differentiated oval-shaped appearance when viewed by phase-contrast light microscopy. Hepatocyte mRNA concentrations for all three male-specific P45Os decayed rapidly during the first 24 h in culture (Fig. 1). However, by Day 3 the constitutive expression of 2C13 mRNA had begun to increase, and by Day 7 the level was 70% of that observed in livers from normal male rats. Expression of 2Cll mRNA remained low until Day 3, but then increased, such that by Day 5 the level was 21% of normal. No further increase in 2Cll mRNA occurred after Day 5. The expression of P450 2Cll protein in microsomal fractions isolated from cells in culture for 6 days was confirmed by Western blotting (Fig. 2). In contrast, levels of
120
1
F 3 E
1 a
E
0
1
2
3
4
5
6
7
TIME IN CULTURE (days) FIG. 1. Expression of mRNA for P450 2Cll (0), 2C13 (Cl), and 3A2 (0) as determined by solution hybridization in primary cultures of male rat hepatocytes harvested over a 7-day period. Cells were maintained on a matrigel substratum in modified Waymouth medium. Each experimental point represents the mean of triplicate determinations of total nucleic acid samples from five pooled culture dishes. Values are expressed as a percentage of the respective normal mRNA level as determined in liver samples from three adult male rats.
162
LIDDLE
C
GH
DEX
FIG. 2. Western blot analysis of P450 2Cll protein in microsomes from hepatocytes maintained in the absence (C) or presence of growth hormone (GH) (100 rig/ml) or dexamethasone (DEX) (10-s M) for 6 days.
3A2 mRNA continued to decline over the experimental period such that mRNA was undetectable by Day 5. Thus, the culture system used in the present study proved suitable for the further study of hormonal regulation of P450 2Cll and 2C13 but not 3A2. Total cellular RNA concentrations, expressed as the RN&DNA ratio, increased over the first 5 days, being approximately twofold greater in cells harvested on Day 5 compared to those harvested on Day 1 (Fig. 3). Treatment with GH tended to suppress RNA levels when compared to untreated control cells, though this effect was relatively minor. However, because of the observed changes in the RNA:DNA ratio with both time and treatment, all results pertaining to the expression of specific mRNA species following hormone treatment were corrected for the total RNA content of the sample rather than for the DNA content.
ET AL.
duction in immunoquantifiable protein for 2Cll was also observed following GH treatment (Fig. 2). In contrast, GH only reduced 2C13 mRNA to approximately 50% of that found in untreated control cells. However, when cells were cultured in the presence of DEX, 2C13 mRNA levels were markedly reduced. This resulted in 2C13 mRNA being 33 and 25% of control on Days 5 and 7, respectively. Interestingly, the negative regulatory effect of DEX on 2C13 was more profound than that of GH. DEX treatment caused a moderate increase in 2Cll mRNA expression, being 137% of control on Day 5 and 195% on Day 7. This effect of DEX was also reflected at the protein level (Fig. 2). Treatment of cells with T3 depressed 2Cll expression to 46 and 66% of control on Days 5 and 7, but had little effect on the expression of mRNA for 2C13. When hepatocytes were treated with combinations of hormones the following patterns emerged. GH exerted the dominant regulatory effect on 2Cll expression, resulting in marked downregulation of mRNA for this form in the presence of both DEX and T3. In contrast, the combined action of GH and DEX on 2C13 expression appeared to be additive, resulting in marked suppression of this form. Again T3 was found to have no effect on 2C13. In addition to the continuous treatment regime with GH, various protocols for applying GH in pulses to the hepatocytes were tested to try to induce the expression of P450 2Cll. No treatment proved successful in this respect and 2Cll mRNA levels were invariably suppressed by such manipulations (data not shown). In Vivo Regulation of P450 2Cll and 2C13 by DEX Efficacy of hypophysectomy of rats was confirmed by the failure of Hx animals to gain weight compared to non-
91
0
13
Effect of Hormone Treatment on Male-Specific P450 mRNA Expression After 1 day in culture, hormone treatment had no clear effect on P450 mRNA expression in hepatocytes. However, by Day 3 an effect of treatments on both 2Cll and 2C13 could be discerned, patterns that were firmly established by Days 5 and 7 (Fig. 4). This indicated that the cells required an adaptation period following inoculation into the culture system, an observation made previously (14). The presence of GH in the culture medium almost completely suppressed the expression of 2Cll mRNA in hepatocytes when measured on Days 3, 5, and 7. A re-
//?
3J 0
t
l
2
3
4
5
6
7
TIME IN CULTURE (days) FIG. 3. Total cellular RNA concentrations over culture age, expressed as the RNA:DNA ratio, in hepatocytes maintained in the absence (0) or presence of growth hormone, 100 rig/ml (0).
HORMONAL
REGULATION
n CONTROL Cl GH 0 I
DEX
OF CYTOCHROME B
163
P450
300
n CONTROL 0 0
250 I
q T3 q GH+DEX
60
K
z E 8
GH DEX
60 40
5 TIME IN CULTURE (days)
7
TIME IN CULTURE (days)
FIG. 4. Expression of mRNA for P450 2Cll (A) and P450 2C13 (B) as determined by solution hybridization in hepatocytes treated with hormones (alone or in combination) and harvested at 5 and 7 days of culture age. Each experimental point represents the mean f SD of triplicate determinations of total nucleic acid samples from five pooled culture dishes. GH, growth hormone (100 rig/ml); DEX, dexamethasone (lo-* M); T3, L-triiodothyronine (low9 M).
Hx, control animals: mean weight gain for Hx rats was 1.9 f 1.3 g/week; for control rats, 35.9 f 2.2 g/week. Ax animals had diminished weight gain (12.6 +- 4.2 g/week) compared to control rats. As expected, hypophysectomy resulted in a marked fall in the steady-state expression of mRNA for 2Cl1, being 5% of that in livers from intact male rat controls (P < 0.001). Treatment of Hx rats with DEX resulted in a fourfold increase in mRNA concentrations for 2Cl1, such that expression was 21% of that in intact controls (P < 0.001). When endogenous glucocorticoid secretion was abolished by adrenalectomy in otherwise intact rats, 2Cll mRNA levels decreased to 80% of that in control animals (P < 0.001). When Hx animals treated with DEX also received a continuous hGH infusion via a subcutaneously implanted osmotic minipump, 2Cll expression was decreased so as to resemble that in Hx rats receiving a hGH infusion alone, demonstrating that GH downregulation takes precedence over positive regulation by glucocorticoids. These data are summarized in Fig. 5. Despite the marked downregulation of P450 2C13 mRNA by DEX in the primary cultured hepatocytes, no effect of adrenalectomy or DEX treatment on this P450 form was observed in viuo (data not shown). To determine the level at which glucocorticoids were regulating 2Cll expression, in vitro transcriptional activity of this gene was determined by run-on analysis in nuclei obtained from the livers of animals subjected to various treatments. Nuclei from two animals of each treatment were analyzed; Fig. 6 shows the autoradiography result from one rat analyzed from each group. Following normalization of the numerical data against Pactin we found that DEX treatment of Hx rats resulted
in a fourfold increase in 2Cll transcription, directly paralleling the increase observed in steady-state mRNA concentrations. In contrast, no major effect of adrenalectomy on 2Cll transcription could be discerned, activity being 95% of that found in control animals. hGH infusion totally abolished 2Cll transcription in keeping with our previous findings (25). TAT transcriptional activity, included as a positive control for glucocorticoid action, was reduced
DEX OH,. ...r
DEX WI,,,
FIG. 5. Expression of mRNA for P450 2Cl1, expressed per microgram of DNA, in livers of adult male rats subjected to various treatments. CONT, control; Ax, adrenalectomized; Hx, hypophysectomized; Hx DEX, hypophysectomized, treated for 4 days with dexamethasone 1 mg/kg/day; Hx GHmp, hypophysectomized, treated for 7 days with continuous GH infusion by minipump; Hx DEX GHmp, hypophysectomized, treated with a combination of dexamethasone and GH (as above). Results expressed as means + SD, n = 4 animals per group.
164
LIDDLE
pGEM
!
CONT
A,
,’
H,
8 -“..*“, i :;; _
H, H, DEX GH,,
FIG. 6. In vitro elongation of P450 2Cll (2Cll) as determined by runon analysis of bepatic nuclei from male rats subjected to various treatments. Additional probes were included to act as controls: pGEM, pGEM 32 plasmid vector containing no cDNA insert; TAT, tyrosine aminotransferase; 2C12, P450 2C12. CONT, control; Ax, adrenalectomized; Hx, hypophysectomized, Hx DEX, hypophysectomized, treated for 4 days with dexamethasone 1 mg/kg/day; Hx GHmp, hypophysectomized, treated for 7 days with continuous GH infusion by minipump.
following adrenalectomy and increased twofold following DEX treatment of Hx rats. DEX treatment of Hx rats brought TAT activity back to normal levels. P450 2C12 transcriptional activity, included as a positive control for GH action, increased markedly following GH infusion. Interestingly, adrenalectomy alone also increased 2C12 transcription to 40% of that observed in Hx rats receiving GH infusion, suggesting that glucocorticoids may be capable of suppressing this gene. This possibility is supported by our previous observation that corticosterone is capable of decreasing 2Cl2 mRNA expression in cultured rat hepatocytes (15). DISCUSSION The use of primary cultures of hepatocytes as a model to examine the effects of isolated stimuli on liver metabolism is attractive. However, finding appropriate culture conditions that allow expression of specialized hepatic genes has proven difficult. The two predominant variables in a culture system are the type of the attachment surface and the composition of the culture medium. Various groups of investigators have described conditions that allow expression of particular genes or groups of genes, including some forms of P450. Schuetz et al. (14) demonstrated that several forms of P450 could be expressed following xenobiotic induction in hepatocytes maintained on a substratum of matrigel in modified Waymouth medium, a system that was also used in the present study. However, other combinations have also proven to be successful for expression of certain P450 enzymes, including phenobarbital induction of forms 2Bl and 2B2 using type I collagen and either Williams’ E medium (26) or Chee’s medium (8) and 3-methylcholanthrene induction of forms
ET AL.
1Al and lA2 using type 1 collagen and supplemented RPM1 1640 medium (27). A limitation of the above studies is that they have been restricted to P450 enzyme forms inducible by xenobiotics. The only example to date of increased expression of a constitutive hepatic P450 in cell culture following exposure to an endogenous regulating substance is the constitutive female-specific form 2C12 on treatment of hepatocytes with GH (7,15). Some constitutive P450 forms, for example 2C13, have no known positive regulator (12, 28). In uiuo studies have indicated that this male-specific form is repressed by the continuous pattern of GH release from the pituitary in female rats (19, 28). A form that appears to be similarly regulated, 3A2 (29), is also male specific, though this P450 has been demonstrated to be subject to xenobiotic induction by phenobarbital in intact animals (22). The male-specific form, 2Cl1, appears to be under complex regulation by GH. In uiuo studies indicate that the male pulsatile pattern of GH release from the pituitary positively regulates this form, whereas the continuous female pattern of GH secretion results in downregulation (3). In the present study, a constitutive increase in the expression of mRNA was found after 3 days for P450 form 2C13 and after 5 days for 2Cll in hepatocytes cultured on matrigel in modified Waymouth medium. Although the level of expression was less than that in normal male rat liver, 70% for 2C13 and 21% for 2Cl1, this provided a suitable model to study direct hormonal effects on these forms. The reduced level of expression of 2Cll was not surprising given its aforementioned positive regulation by pulses of GH; however, application of GH pulses to the cultured hepatocytes did not increase 2Cll mRNA levels. This indicates either that it was not possible to completely eliminate GH from the culture system between pulses or that additional factors may be required for full expression of this form. In this respect the circulating growth hormone-binding protein found in many species, including the rat (30), may be of importance, though the overall function of this protein remains unknown. The mRNA concentration for P450 3A2 decayed rapidly with time in cultured hepatocytes and did not exhibit the constitutive increase of expression over time seen with the other forms studied. Again, some additional factors appear to be necessary for the expression of this form. Using hepatocytes cultured on matrigel in modified Waymouth medium, Guzelian et al. (7) demonstrated that the upregulation of the female-specific P450 2C12 by GH was due to a direct action of this hormone on hepatocytes. Similarly, we have shown that continuous treatment of hepatocytes with GH acts directly on the cells to demasculinize hepatic oxidative metabolism by downregulating at least two male-specific forms of P450 at a pretranslational level. In the case of P450 2Cll this downregulation was almost complete. Surprisingly, P450 2C13 expression
HORMONAL
REGULATION
was only partially abolished by continuous exposure to GH. Only in the presence of the synthetic glucocorticoid DEX did GH reduce expression of 2C13 mRNA to low levels. Indeed, DEX treatment alone was effective in substantially decreasing expression of 2C13. The significance of these findings is uncertain as earlier in vivo studies have shown that continuous secretion of GH from osmotic minipumps alone is sufficient to suppress this enzyme in Hx animals, where it is expressed at a high level (21). Indeed, DEX treatment of Hx rats had no apparent effect on 2C13, demonstrating at least some divergence between the culture model and the in vivo situation. In the present study, the synthetic glucocorticoid agonist DEX has been demonstrated to increase the expression of 2Cll both in uiuo and in vitro. Moreover, we have demonstrated that this effect occurs predominantly at the transcriptional level. This observation is of interest given that Morishima et al. (31) have described the presence of a glucocorticoid-responsive element in the 5’ flanking region of the gene for P450 2Cll. Evidence for a physiological role of glucocorticoids in 2Cll gene regulation is provided by the decrease in the expression of this gene in rats in which endogenous steroid secretion was removed by adrenalectomy. Interestingly, in absolute terms, the decrease in 2Cll expression following adrenalectomy was approximately the same as the increase in expression in Hx rats treated with DEX. Thus glucocorticoid regulation of the 2Cll gene accounts for 20% of the overall expression of this P450 form in the livers of intact male rats. While this may appear to be a relatively small effect, the role of 2Cll as the dominant P450 form in the male rat plus the involvement of this form in the metabolism of a range of endogenous steroid compounds (32-34) suggest that such a change may well be of importance. An additional finding of the present study is that GH downregulation of 2Cll takes precedence over glucocorticoid effects, in keeping with the view that GH secretion patterns are the predominant regulatory mechanism for 2Cll. The mechanism by which GH exerts regulatory effects on some P450 enzyme forms remains unknown though a recent study from our laboratory indicates that this occurs at the transcriptional level (25). Iodothyronines have previously been reported to contribute to the GH-dependent suppression of 3A2 in male rats (5) and the GH-dependent induction of 2C12 as occurs in female rats (3). In the present study, T3 had no potentiating effect on the ability of GH to downregulate either 2Cll or 2C13 mRNA in primary cultured hepatocytes, though T3 treatment alone appeared to repress 2Cll expression. This further emphasizes the differing hormonal interactions that occur to regulate individual P450 enzyme forms. In summary, the present study demonstrates that primary cultured hepatocytes are a suitable model to examine the regulation of at least some constitutive male-specific
OF CYTOCHROME
165
P450
rat P450 forms. GH has been shown to act directly on hepatocytes at a pretranslational level to downregulate the expression of these forms. Glucocorticoids and iodothyronines also exert direct pretranslational actions on these forms, an effect that was confirmed in uivo for 2Cll. The overall significance of the other hormonal actions with regard to in vivo regulation remains to be determined. ACKNOWLEDGMENTS The authors are indebted to Dr. Anders Strom and Dr. Peter G. Zaphiropoulos for assistance with the probes used in this study and to Eva Floby for her expert technical assistance.
REFERENCES 1. Mode, A., Gustafsson, J.-A., Jansson, J.-O., Eden, S., and Isaksson, 0. (1982) Endocrinology 111, 169221697.
2. Morgan, E. T., MacGeoch, C., and Gustafsson, J.-A. (1985) J. Biol. Chem. 260, 11895511898. E., Strom, A., Zaphiropoulos, P. G., 3. Mode, A., Wiersma-Larsson, and Gustafsson, J.-A. (1989) J. Endocrinol. 120, 311-317. 4. Yamazoe, Y., Murayama, N., Shimada, M., and Kato, R. (1989) Biochem. Biophys. Res. Commun. 160,609-614. 5. Waxman, D. J., Ram, P. A., Notani, G., LeBlanc, G. A., Alberta, J. A., Morrissey,
J. J., and Sundseth, S. S. (1990) Mol. Endocrinol.
4,447-454. 6. Ram, P. A., and Waxman, D. J. (1990) J. Biol. Chem. 265,1922319229.
I. Guzelian, P. S., Li, D., Schuetz, E. G., Thomas, P., Levin, W., Mode, A., and Gustafsson, 9783-9787.
J.-A. (1988) Proc. Natl. Acad. Sci. USA 85,
8. Waxman, D. J., Morrissey, J. J., Naik, S., and Jauregui, H. 0. (1990) Biochem. J. 271,113-119. 9. Schuetz, E. G., Schuetz, J. D., May, B., and Guzelian, P. S. (1990) J. Biol. Chem. 265, 1188-1192.
10. Murayama,
N., Shimada, M., Yamazoe, Mol. Pharmacol. 39, 811-817.
Y., and Kato, R. (1991)
11. Waxman, D. J., Dannan, G. A., and Guengerich, chemistry 24, 4409-4417.
12. Bandiera,
F. P. (1985) Bio-
S., Ryan, D. E., Levin, W., and Thomas, Arch. Biochem. Biophys. 248,658-676.
P. E. (1986)
13. Bissell, D. M., Arenson, D. M., Maher, J. J., and Roll, F. J. (1987) J. Clin. Inuest. 79, 801-812. 14. Schuetz, E. G., Li, D., Omiecinski, C. J., Muller-Eberhard, U., Kleinman, H. K., Elswick, Physiol. 134, 309-323.
B., and Guzelian,
15. Tollet,
P. S. (1988) J. Cell.
P., Enberg, B., and Mode, A. (1990) Mol. Endocrinol. 1934-1942.
4,
16. Bissell, D. M., and Guzelian, P. S. (1980) Ann. N. Y. Acad. Sci. 349, 85-98.
17. Durnham, D. M., and Palmiter, 383-393.
R. P. (1983) Anal. Biochem. 131,
18. Labarca, C., and Paigen, K. (1980) Anal. Biochem. 102, 344-352. 19. Melton, D. A., Krieg, P. A., Rehagliati, M. R., Maniatis, T., Zinn, K., and Green, M. P. (1984) Nucl. Acids Res. 12, 7035-7036. 20. Strom, A., Mode, A., Zaphiropoulos, P., Nilsson, A.-G., Morgan, E., and Gustafsson, J.-A. (1988) Acta Endocrinol. 118, 314-320. 21. Zaphiropoulos, P. G., Strom, A., Robertson, J. A., and Gustafsson, J.-A. (1990) Mol. Endocrinol. 4, 53-58.
166
LIDDLE
22. Gonzalez, F. J., Song, B.-J., and Hardwick, Biol. 6,2969-2976. 23. Bradford,
J. P. (1986) Mol. Cell.
M. M. (1976) Anal. Biochem. 72, 248-254.
24. Morgan, E. T., Ronnholm, chemistry 26,4193-4200.
R., and Gustafsson,
J.-A. (1987) Bio-
25. Legraverend, C., Mode, A., Westin, S., Strom, A., Eguchi, H., Zaphiropoulos, P. G., and Gustafsson, J.-A. (1992) Mol. Endocrinol. 6,259-256. 26. Sinclair, P. R., Bement, W. J., Haugen, S. A., Sinclair, Guzelian, P. S. (1990) Cancer Res. 60,5219-5224. 27. Silver, G., and Krauter, 11807.
J. F., and
K. S. (1988) J. Biol. Chem. 263,
11802-
ET AL. 28. McClellan-Green, P. D., Linko, P., Yeowell, H. N., and Goldstein, J. A. (1989) J. Biol. Chem. 264, 18960-18965. 29. Shimada, M., Nagata, K., Murayama, N., Yamazoe, Y., and Kato, R. (1989) J. Biochem. 106, 1030-1034. 30. Massa, G., Mulumba, N., Ketelslegers, J.-M., and Maes, M. (1990) Endocrinology 126,1976-1980. 31. Morishima, N., Yoshioka, H., Higashi, Y., Sogawa, K., and FujiiKuriyama, Y. (1987) Biochemistry 26,8279-8285. 32. Cheng, K.-C., Shenkman, J. B. (1984) Drug Metab. Dispos. 12,222234. 33. Dannan, G. A., Porubek, D. J., Nelson, S. D., Waxman, D. J., and Guengerich, F. P. (1986) Endocrinology 118,1952-1960. 34. Waxman, D. J. (1988) Biochem. Pharmacol. 37,71-84.
