Selective suppression of cytochrome P-450 gene expression by interleukins 1 and 6 in rat liver

BB.

ELSEVIER

Biochimica et Biophysica Acta 1219 (1994) 475-483

Biochi~ic~a et BlophysicaA~ta

Selective suppression of cytochrome P-450 gene expression by interleukins 1 and 6 in rat liver Edward T. Morgan a,,, K. Bradford Thomas a, Rebecca Swanson a, Teresa Vales a, Jaiweon Hwang a, Kristina Wright b a Department of Pharmacology, Emory University Medical School, Atlanta, GA 30322, USA b Department of Biology, Georgia State University, Atlanta, GA, USA

Received 19 April 1994

Abstract

Inflammatory stimuli suppress constitutive hepatic expression of the CY P2 Cl l and CYP2C12 genes in male and female rat livers, respectively. We have shown previously that injection of interleukin-1 (ILl), but not interleukin-6 (IL6), to female rats also suppresses CYP2C12. In the present study, we examined the effects of these cytokines on CYP2C12 expression in rat hepatocyte cultures, and their in vivo effects on expression of multiple cytochrome P-450 (P450) gene products in male rat livers. ILl suppressed the expression of CYP2C12 mRNA and protein in hepatocytes cultured on Matrigel in the presence of growth hormone. No consistent effect of IL6 was observed. Maximal suppression of CYP2C12 m R N A after 24 h of ILl treatment reached 12 and 32% of control levels in two separate experiments. The approximate EDs0 for ILl was 5 ng/ml. CYP2C12 protein was suppressed to 28% of control levels as early as 12 h after ILl treatment. Injection of ILl, low doses of dexamethasone, or both, in male rats produced decreases in total P450, and in CYP3A2 and CYP2Cll mRNA and protein expression similar to effects previously seen for CYP2C12 expression in females. CYP2E1 mRNA and protein was significantly suppressed only by the combination of ILl and dexamethasone. IL6 treatment of male rats down-regulated the CYP2Cll and CYP2E1 mRNAs at a dose of 4.5/zg/kg, which was lower than that required to induce haptoglobin mRNA, a prototype acute phase gene product. CYP2Cll protein content of the microsomes was also decreased by IL6 treatment, with a slower time-course than for suppression of its mRNA. No significant effects of IL6 treatment were seen on CYP3A2 mRNA or CYP3A2/1 proteins. These results demonstrate that ILl and IL6 treatments in vivo differentially affect subsets of P450 gene products in rat liver. Keywords: Cytochrome P-450; Glucocorticoid; Interleukin-1; Interleukin-6; Inflammation; Endotoxin

I. Introduction

Clinical infections cause a reduction in total hepatic content of c y t o c h r o m e P-450 (P450) and m a n y of its attendant drug metabolizing activities [1]. In animals, infections [1-5] or agents that mimic an infected or inflammatory state [6] cause similar effects.

Abbreviations: ILl, interleukin-1; IL6, interleukin-6; LPS, bacterial lipopolysaccharide; Hx, hypophysectomized; DEX, dexamethasone; HpG, haptoglobin; GAP, glyceraldehyde-3-phosphate dehydrogenase; P450, cytochrome P-450. * Corresponding author. Fax:+ 1 (404) 7270365. E-mail: [email protected]. 0167-4781/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0167-4781(94)00127-0

W e have shown that administration of bacterial LPS to rats causes suppression of the constitutively expressed hepatic C Y P 2 C l l and C Y P 2 C 1 2 proteins in male and female rats, respectively [7]. For C Y P 2 C l l , this effect is achieved at the level of gene transcription [8]. Local inflammation caused by turpentine has comparable effects [7,8]. However, low doses of LPS potentiate the induction o f certain forms of inducible P450 in m o u s e liver [9]. Considerable effort has b e e n directed towards identifying the h u m o r a l mediators of the P450 suppression after host-defense activation. In female rats, I L l , but not IL6, causes a down-regulation of C Y P 2 C 1 2 m R N A and protein [10]. IL6 suppresses expression of p h e n o barbital-induced CYP2B1 in rat hepatocytes cultured

476

E. 7". Morgan et al. / Biochimica et Biophysica Acta 1219 (1994) 475-483

on Matrigel [11], and decreases levels of the mRNAs for CYP1A1, CYP1A2 and CYP3A2 in human hepatoma cell lines [12]. On the other hand, ILl, but not IL6, inhibits 2,3,7,8-tetrachlorodibenzo-p-dioxin-induced CYP1A1 expression in hepatocytes [13]. Interferon inducers or recombinant interferons also suppress specific P450 mRNAs in rat liver [14-17]. It is clear, then, that immune mediators are capable of down-regulating P450 gene expression in vivo and in vitro, and that some of these effects may be specific for certain P450 isozymes or groups of isozymes. In the present study, we tested whether ILl or IL6 could down-regulate the expression of CYP2C12 in hepatocyte cultures and obtained results consistent with our previous in vivo work. We also found that ILl, IL6 and dexamethasone each suppress multiple P450 gene products in male rat livers in vivo, and that the two cytokines selectively affected different subsets of P450 mRNAs.

every 24 h thereafter. At t = 48 h after plating, recombinant h G H (Genotropin, a gift from Kabi Peptide Hormones, Stockholm, Sweden) was added to the medium to a final concentration of 500 n g / m l , and all cultures contained G H thereafter. After a further 2472 h to allow establishment of CYP2C12 mRNA expression, cells were treated with ILl or IL6 as indicated. For harvesting, medium was aspirated and cells were lysed with cold guanidinium thiocyanate solution [19] applied directly to the tissue culture plate. Matrigel was purchased from Collaborative Research, Bedford, MA, or prepared from the murine Engelbreth-Holm-Swarm tumor as described [20]. Recombinant human ILl/3 preparations used in hepatocyte incubations were obtained from the National Cancer Institute Biological Response Modifiers Program or Life Technologies, Bethesda, MD (1.8 • 1(I4 U / / x g and 5 . 1 0 4 U//xg, respectively), and were diluted in phosphate-buffered saline containing 1% bovine serum albumin for addition to the cultures.

2. Materials and methods

2.3. Hepatic microsomes and RNA

2.1. Animals and treatments

Pyrophosphate-washed hepatic microsomes were prepared by differential centrifugation [21]. For microsomes from cultured hepatocytes, a Beckman Optima T L X tabletop ultracentrifuge was used. Total RNA was prepared from whole livers and cultured hepatocytes by the method of Chomczynski and Sacchi [19].

All procedures were approved by the Institutional Animal Care and Use Committee. Male Sprague-Dawley rats were obtained from Harlan Sprague-Dawley (Indianapolis, IN), and were 45-52-day-old at the beginning of the experiments. Recombinant human I L l a ( 2 . 1 0 4 units//xg) used in the in vivo studies was a gift from Dr. P.L. Simon of Smith, Kline and French Laboratories (Swedeland, PA), and was diluted in 10 mM sodium phosphate buffer (pH 7.4), 0.15 M NaC1, 0.1% bovine serum albumin. Recombinant human IL6 (4.103 units//xg) was donated by Genetics Institute (Cambridge MA), and was diluted in the same buffer. D E X (Sigma) was suspended in corn oil and injected subcutaneously. Control animals received the appropriate vehicle(s). Details of the injections for the individual experiments are given in the figure legends.

2.2. Hepatocyte cultures Primary hepatocytes were prepared by perfusion of male rat livers in situ with a calcium-free buffer containing E G T A followed by Waymouth's 752 medium (Gibco) containing collagenase (Boehringer Type B or Sigma type XII), essentially as described by Bissell and Guzelian [18]. Hepatocytes were plated at a density of 3 • 106 viable cells/plate on 60 mm dishes coated with 150 /zl of Matrigel which had been diluted to 6.7 m g / m l with Waymouth's medium. The basic culture medium was Waymouth's containing 0.15 /~M insulin as the only hormone, and 100 u n i t s / m l of penicillin G. Medium was changed at 4 h and 24 h after plating, and

2.4. Assays and statistical analyses Total microsomal protein was determined as described by Lowry et al. [22]. Total microsomal P450 content was determined from the CO difference spectrum of the reduced protein [23]. Relative levels of C Y P 2 C l l , CYP2C12, C Y P 3 A 1 / 2 and CYP2E1 apoproteins in microsomal preparations were measured by Western blot assay as described previously [7]. Antibodies to rabbit CYP2E1 were kindly donated by Dr. D.R. Koop of Oregon Health Sciences University or purchased from Oxygene (Dallas, TX). Antibodies to rat CYP3A2, which also recognize CYP3A1, were a generous gift of Dr. James Halpert of the University of Arizona. Antigen-antibody complexes were detected with horseradish peroxidase-conjugated second antibodies. All assays were done under conditions of proportionality between the amount of antigen on the filter and the band intensity. Band intensities, measured by densitometry, were expressed per mg microsomal protein, and the relative levels of P450 proteins in the individual samples were calculated per mg of microsomal protein, and expressed as a percentage of the mean for the control group. Northern blots were performed as described previously [24]. The relative levels of C Y P 2 C l l and

E.T. Morgan et aL / Biochimica et Biophysica Acta 1219 (1994) 475-483

CYP2C12 mRNAs in total hepatic RNA were quantified by slot blot assay using the cloned cDNAs 1.1 and C-6 as described [7]. For CYP2E1 slot blots, the fulllength cDNA [25], generously donated by Dr. F.J. Gonzalez of the National Institutes of Health, was used. GAP cDNA [26] was obtained from the American Type Culture Collection (Rockville, MD), and a 780 bp PstI/XbaI restriction fragment was used as probe. Relative levels of the mRNAs for CYP3A2 and

477

HpG were measured using 32p-5'-end labeled oligonucleotides complementary to nucleotides 1690-1729 of CYP3A2 [27] and nucleotides 210-239 of HpG [28]. The specificity of the CYP3A2 probe has been documented earlier [27]. The hybridization and high-stringency washing conditions for the oligonucleotide probes were calculated as described in [29]. The intensities of the autoradiographic bands were quantified by laser densitometry, and assays were performed under established conditions of linearity between the amount of the target mRNA on the filter and the densitometric response. Results were expressed as the ratio of a specific mRNA to total poly(A) + (measured using an oligodeoxythymidylate 30-mer probe [10,30]) in the samples. Data from Western blot assays and slot blot assays were expressed as the percentage of the mean of the control group in each experiment. One way analysis of variance and the Neumann-Keuls Test, or Dunnett's test where appropriate, were used to determine differences among treatment groups. Data are presented in the figures as the mean + S.E.

3. Results

3.1. Effects of ILl and IL6 on CYP2C12 expression in cultured hepatocytes

5.8 5.8 58 58

C

Since ILl, but not IL6, treatment of female rats was shown previously to down-regulate CYP2C12 expression [10], we tested the abilities of these agents to suppress CYP2C12 mRNA in cultured hepatocytes. Hepatocytes cultured on Matrigel retain the differentiated morphology of adult liver cells and express CYP2C12 mRNA and protein under the control of GH, its main physiological regulator, regardless of the

F

ILl, ng/ml _=

140

.....

I

.

.

.

.

.

'''I

'

'

'

' '''I

,

i

,,,,,I

> 120 100 qO "6

80

C;;

o ~) 60

g..

< z

40

E

20

ft.

~ 0 >. ¢J -20 O-

,

,,,,,I

,

0.1

I

i

iii,ii

,

1

[ILl], ng/ml

,

10

•

•

,,,

100

Fig. 1. Suppression of CYP2C12 m R N A by I L l in hepatocyte cultures. (Upper panel) Northern blot. At t = 0 (48 h after plating), G H was added to all cultures and was continuously present thereafter. 24 h after introduction of GH, recombinant h u m a n ILl/3 at the indicated concentrations, or vehicle (C, control cultures)) were added. Cells were harvested 24 h after the cytokine addition. Each lane represents 7.5/zg of pooled R N A from two or three culture plates. F, 7.5 /zg of female rat liver RNA. (Lower panel) Dose-response relationship for I L l - d e p e n d e n t suppression of CYP2CI2. Cultures were performed in the presence of different I L l concentrations as described above, except that I L l treatment was initiated 48 h after G H addition. Cells were harvested 24 h after addition of ILl. Each data point represents the m e a n + S.E. of at least three determinations, each from a different sample of R N A from two or three pooled plates. Results from three independent experiments are shown. Relative levels of CYP2C12 m R N A were determined by slot blotting, corrected for the a m o u n t of poly(A) + R N A applied to the filter, and expressed as a percentage of the m e a n of the control group ( G H treated, no ILl) for each experiment. * Significantly different from control group, P < 0.05.

478

E.T. Morgan et aL / Biochimica et Biophysica Acta 1219 (1994) 475-483

sex of the animal from which the cells are isolated [31]. In experiments not shown, we have determined that CYP2C12 m R N A expression is essentially maximal after 48 h of G H exposure. It can be seen from the Northern blot in Fig. 1 that the level of CYP2C12 m R N A at this time (lane C) is comparable to that in female rat liver (lane F). Fig. 1 shows that exposure of hepatocytes from male rats, cultured in the presence of GH, to ILl/3 for 24 h causes a dose-dependent reduction in CYP2C12 expression, with a maximum suppression reaching 12 and 32% of control levels at an ILl/3 concentration of 27 n g / m l in two separate experiments. CYP2C12 protein expression is also suppressed by I L l treatment of the hepatocytes (Fig. 2). The magnitude of this effect is comparable to that of the m R N A , and is achieved as early as 12 h after I L l exposure (Fig. 2) In one experiment, we detected a significant increase in CYP2C12 m R N A in cultures treated with 0.5-1.4 n g / m l IL1/3 (result not shown), followed by suppression at higher concentrations. We have not observed this inductive effect of ILl in subsequent experiments. In contrast to the consistent suppression by ILl, we have been unable to produce a significant effect of IL6 in this system, using doses as high as 1500 U / m l . In one experiment, a 50% suppression by IL6 was detected, but three subsequent experiments gave no effect (not shown). Hepatocytes cultured in the absence of G H had no detectable CYP2C12 m R N A (not shown).

3.2. Effect of ILl, dexamethasone, and their combination on expression of multiple P450 genes in male rat liuers In an earlier study, we showed that recombinant human I L l a suppressed CYP2C12 m R N A and protein in the livers of normal female rats [10]. CYP2C12 gene products were also down regulated by D E X injection, or by a combination of D E X and I L l . We now investigated whether the same agents could down-regulate the expression of C Y P 2 C l l and CYP 3A2, two predominant P450s in the livers of male rats, as well as CYP2E1, an alcohol- and diabetes-inducible form that is also expressed constitutively. T r e a t m e n t with DEX, I L l , or I L l + D E X reduced the total content of P450 in liver microsomes to 75, 75 and 63% of control levels respectively (Table 1). These decreases were accompanied by decreases in the levels of microsomal C Y P 2 C l l protein and C Y P 2 C l l and CYP 3A2 m R N A s (Table 1). D E X alone suppressed the C Y P 2 C l l gene products to less than 50% of control levels (Table 1). CYP3A2 m R N A and C Y P 3 A 1 / 2 proteins were more profoundly affected, reaching 27% and 37% of control values, repectively. I L l alone and I L l + D E X were as effective as D E X in suppressing

12 h • : Q ~ h ~M Control

ILl

175

t-0

150 125

°,o0, eJ 8 ~ 0 -~

"6

" 75 50

E 0

£

25

0

0 0

I

I

I

I

I

I

4

8

12

16

20

24

Time, hours Fig. 2. Suppression of microsomal CYP2C12 protein by ILl in hepatocyte cultures. ILl (10 n g / m l ) or vehicle was added to the medium 48 h after addition of growth hormone to the cultures, as described in Fig. 1. Cells were harvested at the indicated times after ILl addition, and microsomes were analyzed by Western blotting using an immunoabsorbed antiserum to CYP2C12. Each independent sample represents microsomes from two pooled plates. (Upper panel) Western blot. 20 /xg of hepatocyte microsomal protein was applied to each lane. Different amounts of microsomes from an untreated female rat liver were used for a standard curve. (Lower panel) Results from densitometric analysis of the Western blot. Data are expressed as the mean + S.E. from each group (n = 3), relative to the mean value of the control group (0 h). Levels of CYP2C12 protein in the microsomes of this group were calculated to be 16.5~ of those in the untreated female rat liver microsomes. The lack of visible error bars indicates that the S.E. is smaller than the symbol. * Significantly different from control group, P < 0.05.

the C Y P 2 C l l m R N A , but C Y P 2 C l l protein levels were less affected by these treatments (72% and 64% of control male levels, respectively). CYP2E1 m R N A and protein were not significantly affected by either D E X or I L l alone, but were reduced to 28% and 65% of control levels in rats treated with I L l + DEX. CYP3A2 m R N A was suppressed by D E X alone and I L l alone (27% and 56% of control, respectively), but the combination of ILl and D E X produced no further significant decrease. C Y P 3 A 1 / 2 protein levels closely paralleled the CYP3A2 m R N A expression in this experiment. In summary, the effects of these treatments upon C Y P 2 C l l and CYP3A2 expression in the male

479

E.T. Morgan et al. / Biochimica et Biophysica Acta 1219 (1994) 475-483

phase protein, HpG, was elevated by IL6 treatment (Fig. 4D). The m R N A s for C Y P 2 C l l and CYP2E1 were again suppressed, but exhibited different timecourses (Fig. 4A, C). CYP2E1 m R N A was maximally suppressed within 6 h of IL6 injection, and returned to control levels by 24 h, whereas C Y P 2 C l l was still suppressed 24 h after injection. C Y P 2 C l l protein levels were significantly suppressed by 24 h following IL6 injection, with the greatest effect seen after 48 h (56% of control levels, Fig. 4A). No significant effect of IL6 on CYP2E1 content of the microsomes was seen at any time point, although there was a tendency toward lower mean values at later times (Fig. 4C) CYP3A2 m R N A levels were not suppressed by IL6 treatment, and in fact the mean values tended to be increased (Fig. 4B). There was also no significant effect of IL6 treatment on CYP3A2 protein levels in the microsomes. In another experiment (not shown), we determined that the decrease in CYP2C11 protein caused by injection of IL6 was not affected by the concomitant administration of DEX. G A P m R N A is often used as a control gene in studies of gene regulation [33]. We found that levels of this m R N A were also significantly suppressed by the IL6 treatments, and in fact were still decreased 48 h after IL6 injection (Fig. 4D). Suppression of G A P m R N A after an inflammatory stimulus has been documented previously [34]. Since discovering that another 'control' gene product, actin m R N A , was significantly affected by some of our treatments in female rats [10,15], we have routinely normalized our results to the total poly(A) ÷ R N A content of the samples. We have rarely observed a significant difference in this parameter among any of our treatment groups. However, it should be noted that in the present experiment, the poly(A +) content of the total R N A from the t = 6 h group relative to its absorbance at 260 nm was increased compared to the control group (232 + 23 versus 100 + 30).

rat closely resemble the effects we already observed upon CYP2C12 expression in the female rat [10], whereas CYP2E1 showed a different pattern of effects. 3.3. Effects o f I L 6 on P450 gene expression in male rat livers

In contrast to I L l , we had seen no in vivo effect of IL6 alone upon CYP2C12 levels in female rat liver, measured 24 h after injection [10]. To examine the effects of IL6 upon P450 gene expression in male rat livers, we treated groups of male rats with various doses of IL6, and examined P450 gene expression 24 h later. Because of the short biological half-life of IL6 [32], rats were given three IL6 injections, spaced 2 h apart. The dose-dependencies of the responses of C Y P 2 C l l and CYP2E1 m R N A s to three injections of IL6 at t = 0, 2 and 4 h were determined 24 h after the first injection, and compared with the response of H p G m R N A (Fig. 3A). A dose-dependent increase in H p G m R N A was observed, becoming significant at 15 ~ g / k g IL6. In contrast, levels of C Y P 2 C l l and CYP2E1 mRNAs (49 and 50% of control, respectively) were significantly suppressed at the lowest dose of IL6 (4.5 ~ g / k g ) , and decreased only slightly further at the maximal IL6 dose (37% of control, both mRNAs). In this experiment, no significant effects of IL6 were seen on total microsomal P450, or on microsomal C Y P 2 C l l levels (Fig. 3B). CYP2E1 content of the microsomes was decreased to 80 and 76% of control levels at the 15 ~ g / k g and 150 ~ g / k g doses, respectively. No significant effects were seen on hepatic microsomal protein yield (not shown). We then used a dose of 15 /xg/kg IL6 to examine the time-course of the effects of IL6 on various hepatic P450 mRNAs. IL6 was injected in partial doses at t = 0, 2 and 4 h as before, and rats were killed for analysis of hepatic gene expression at the times indicated (Fig. 4). As expected, the m R N A for the acute Table 1 ILl and DEX injection suppress P450 gene expression in male rat liver Percent of control group mean Treatment control n:5 Total P450 (nmol/mg) CYP2CllmRNA protein CYP2E1 mRNA protein CYP3A2mRNA protein

100 5:4.1 1005:17.9 1005- 6 1005:21 100± 3 100±11 100± 3

DEX 4

ILl 5

74.5 5: 1.9 * 45 46 116 101 27 37

_+ 8" 5: 7" 5:15 + 8 5: 4" 5:10"

ILl + DEX 5

75.4 5:4.0 * 41 72 55 100 56 54

5: 6" 5: 7" 5- 10 5:10 5: 7" ±12"

63.4 5:5.7 * 31 64 28 65 14 26

5:3" 5:3" 5:6 * 5:7" +2" 5:9"

Rats were injected at t = 0, 2 and 4 h with either DEX (100/zg/kg, s.c.), ILh~ (1.5 p~g/kg, i.p), or both. Control animals received the vehicles only, and all animals were killed at 24 h. Total microsomalP450, microsomalcontent of CYP2Cll, CYP2E1 and CYP3A2proteins, and hepatic CYP2Cll, CYP2E1 and CYP3A2mRNA levels were measured as described in the text. The mean rnicrosomal P450 content of the control group was 0.96 nmol/mg, n = number of animals. * Significantly different from control group, P < 0.05.

E.T. Morgan et al. / Biochimica et Biophysica Acta 1219 (1994) 475-483

480

4. D i s c u s s i o n

The results of this study are consistent with the participation of circulating interleukins 1 and 6, as well as glucocorticoids, in the down-regulation of hepatic cytochrome P-450 gene expression following an inflammatory stimulus. Furthermore, it is shown that just as individual P450 mRNAs and proteins have different responses to different cytokines, individual cytokines can have P450 gene-specific effects. We have also demonstrated for the first time that ILl can suppress the expression of CYP2C12 m R N A and microsomal 250

/

I

A

I

o Haptoglobin

g_

/ /

, / /

J.

/

, 100:

s0

0

~

0.0

/

.__

-r

X.....--.-..._~~

/ i

v CYP2E1

< z n" E 150

"6

i

•

• c¥P2Cll

~> 200 ._J

p,

i

I

I

I

I

0.5

1.0

1.5

2.0

I

I

I

log (dose of IL6, ~tg/kg) ~

2.5

I

B

125

~o- 100"

>

75

0

C.~ 50

"6

o total P450 • CYP2C11 apopmtein

e-

~

25

0

v CYP2E1 apoprotein f/f

0.0

I

I

I

I

0.5

1.0

1.5

2.0

2.5

log (dose of IL6, lag/kg) Fig. 3. Dose dependency of ]L6 effects on hepatic P450 gent expression. IL6 was injected i.p. at t = 0, 2 and 4 h, and animals were killed at 24 h for analysis of their livers. Control animals received vehicle only. The doses shown reflect the totals of the three injections. (A) Hepatic contents of CYP2Cll, CYP2Cll and H p G mRNA; (B) Microsomal contents of total P450, CYP2CI1 and CYP2E1 proteins, were assayed as described in the text. The mean P450 content of the control microsomes was 0.88 n m o l / m g protein. Values are the mean_+ S.E. of data from five animals per experimental group (except for the lowest IL6 dose, which had only four animals). * Significantly different from control group, P < 0.05.

protein in hepatocyte cultures. Recently, Abdel-Razzak et al. [35] reported the suppression of several human P450s by various cytokines, including ILl, in cultured human hepatocytes. However, in the latter study the cells were exposed to the cytokines for 72 h before measurement of P450 expression. Furthermore, cytokine treatment was initiated in a period when the hepatocyte expression of the P450s was already temporarily suppressed. In contrast, the suppression of CYP2C12 expression by ILl in our system was relatively rapid, with a near-maximal decrease in the protein being observed within 12 h of initiation of ILl treatment. This time-course closely mimics that seen when rats are treated with bacterial endotoxin in vivo [7]. The ability of ILl, but not IL6, to suppress the expression of CYP2C12 in hepatocyte cultures is consistent with our in vivo findings in female rats [10]. Recently, Barker et al. showed that ILl, but not IL6 or TNF, reduces the rates of transcription of the CYPIAI and CYPIA2 genes in hepatocytes cultured in the presence of 2,3,7,8-tetrachlorodibenzo-p-dioxin [13]. In contrast, 1L6 attenuated phenobarbital induction of CYP2B1 and CYP2B2 in cultured hepatocytes [11]. The latter studies could not distinguish between direct effects of the cytokines on P450 gene expression, and effects mediated via the induction mechanisms of the chemical inducers. Similarly, although CYP2C12 is expressed in untreated female rats, it has G H as a physiological inducer, and in fact is only expressed in culture if G H is present. Thus, it remains to be seen whether ILl affects CYP2C12 expression through interfering with its induction by GH, or by another mechanism. From this work, an approximate EDs0 for 1L1/3 down-regulation of CYP2C12 expression can be estimated at about 5 n g / m l , corresponding to 0.3 riM. A K d of 4 nM was determined for ILl t~ binding to a primary hepatocyte preparation that responded to ILl with a suppression of total P450 [36]. In the latter study, the concentration dependence for down-regulation of total P450 was almost identical to what we report here. Similarly, Barker et al. [13] found halfmaximal suppression of CYP1A1 and 1A2 induction by ILl/3 (Boehringer) at about 0.5 to 2 n g / m l , respectively. The small differences between their results and ours may reflect a possible difference in the degree of ILl receptor occupancy needed to suppress Ah-receptor-mediated induction of CYP1A1 and 1A2 versus that needed to inhibit GH-dependent expression of CYP2C12. Another point that must be considered regarding studies on hepatocytes is that although the preparations are highly enriched in parenchymal cells, there is inevitably contamination with nonparenchymal cells, including Kupffer cells. Therefore, there remains the

481

E.T. Morgan et aL / Biochimica et Biophysica Acta 1219 (1994) 475-483

previously in females. It suppressed total P450 content and the expression of CYP2C11 and CYP3A2, both at the m R N A and protein level. DEX and ILl alone had no significant effects on CYP2E1 m R N A or protein, but a marked suppression was seen when these two agents were given in combination. These observations have significant implications for xenobiotic and steroid metabolism in the male rat, since CYP2C11 and CYP3A2 are two of the major forms expressed in that gender, and CYP2E1 is critical in the metabolism of many compounds of toxicological importance. Recently, others have reported that ILl suppresses P450 proteins of the CYPIA, CYP2B, CYP3A [37] and CYP2D [38] subfamilies in rats. In support of our

possibility that I L l and IL6 could be exerting their suppressive effects partially or wholly through stimulation of Kupffer cells to release other mediators. Supporting a direct effect on parenchymal cells, it has been shown that IL6 reduces the levels of steady-state mRNAs for CYPIA1, CYPIA2 and CYP3A3 in human hepatoma cells in the absence of drug inducers [12]. However, this was after relatively long-term IL6 treatment (96 h), and although hepatoma cell lines are homogeneous populations, even the most differentiated of them are very different in phenotype from hepatic parenchymal cells. In vivo effects of ILl on P450 expression in the male rat were in many respects similar to what we observed

140

,

-:1oo o~

i

,

~i--O~

=

*

i

' I

140

-I

1

~ ~

/

mRNA

t

,'o

20

I

'

'

l

,

I •

oor 4°I-

o;

0

,o

;

;

300/

Hours after IL6 injection 180 .

,

,

,6o /

,

,

-

20 o 0

HOO~ a f t ; IL " 6 i i l ~ i ~

. B. cvP3.2 I

,'GAP '|

IX

.o

~-

,

,

•

protein

so I-

0,

i,

-,e

O HpG

,

i

-!

,

,

,

,

,

ol

,

,

•

GAP

,I

10

20

30

40

50

0

10

20

30

40

50

Hours after IL6 injection

Hours after IL6 injection

Fig. 4. Time-course of effects of I1,6 injection upon hepatic mRNAs in male rats. Rats were injected at t = 0, 2 and 4 h with 5/~g/kg (15/zg/kg total) IL6, and killed at the indicated time after the first injection. Controls were injected with vehicle only, and killed at t = 24 h. Hepatic

mRNAs and microsomalP450 proteins were measured as described in the text. Values are the means _ S.E. of data from five animals(except for the 48 h time point, which had only four animals). * Significantlydifferent from control group, P < 0.05.

482

E.T. Morgan et al. / Biochimica et Biophysica .4cta 1219 (1994) 475-483

results, in a more chronic protocol, Ferrari et al. showed that injection of rats with 0.1-10 m g / k g ILl daily for 3 days suppressed testosterone 16a-hydroxylase activity, which is catalyzed by C Y P 2 C l l , and 6/3-hydroxylase activity (CYP3A2) in untreated male rats [39]. Conversely, the 6/3-hydroxylase activity was slightly increased in female rats. Although we had previously found no effect of IL6 treatment on total P450 and CYP2C12 gene products in females [10], IL6 treatment of male rats in the present study produced clear suppression of CYP2C11 and CYP2E1 mRNAs. IL6 effects on CYP2E1 and C Y P 2 C l l proteins were smaller and less consistent. This may be due to the fact that the time-courses of the effects on P450 mRNAs and proteins are different (Fig. 4). However, a similar discrepancy between effects of endotoxin on C Y P 2 C l l m R N A and protein was observed previously in this laboratory, and we have suggested that the liver may have a mechanism to maintain expression of P450 proteins despite decreased transcription of the genes [7]. If the differences in our findings between the two sexes is due to a sex difference in the dose dependency of the effect of IL6, then the sex difference must be quite large, since the dose used in females ( 3 3 / z g / k g , twice [10]) was greater than that used in Fig. 4 and about an order of magnitude higher than that which gave a maximal suppression of C Y P 2 C l l and CYP2E1 mRNAs in males (Fig. 3). The P450 isozyme selectivity of the in vivo effects of IL6 is different from that of ILl. Thus, C Y P 2 C l l and CYP2E1 were suppressed by IL6, while CYP3A2 (male) and CYP2C12 (female [10]) were not. In comparison, C Y P 2 C l l , CYP3A2 and CYP2C12 (female [10])were suppressed by ILl, whereas CYP2E1 was unaffected. The isozyme selectivity, or lack thereof, of inflammatory suppression of P450s is an important issue with regard to our ability to predict effects of various agents or inflammatory reactions in humans, since humans have a quite different profile of P450 genes than do rodents [40]. There is other evidence for isozymespecific effects of immune mediators on P450 gene expression. For example, in direct contrast to the present study, IFN7 administration to male rats suppressed CYP3A2 but not C Y P 2 C l l expression [17]. There is an emerging picture that different cytokines can affect different subsets of the P450 superfamily. It is not surprising, however, that an inflammatory stimulus like endotoxin would produce a more general suppression of P450 gene products, since this evokes a response involving many different cytokines and interferons. So far, endotoxin administration has suppressed the expression of every P450 we have studied [7,41]. Recently, Chen et al. [37] described the reduction of hepatic cytochrome P-450 content in male rats by doses of IL6 (100 /xg/kg) that did not affect levels of

CYPIA1, C Y P 2 B 1 / 2 or C Y P 3 A 1 / 2 apoproteins. A dose of 2 0 0 / x g / k g was claimed to reduce the expression of CYP3A proteins, but this was not substantiated by quantification or statistical analysis [37]. The source and specific activity of the 1L6 was identical to that used here. In the present study we found significant, near-maximal effects of IL6 on CYP2CI1 and CYP2E1 mRNAs at 4.5 mg/kg, a dose lower than that needed to maximally induce the m R N A of the acute phase protein HpG. In our earlier work in female rats [10], tow doses of DEX caused decreases in total P450 content of liver microsomes and in CYP2C12 gene products. These findings are extended in the present study to show that low doses of the glucocorticoid also down-regulate expression of the male-specific P450s C Y P 2 C l l and CYP3A2. In contrast, CYP2E1 was clearly unaffected by the same treatment. High doses of DEX and other glucocorticoids in rats are well known to induce certain P450 enzymes [42,43], and to suppress others including C Y P 2 C l l [44]; these previous studies have typically used doses of D E X ranging from 10-200 m g / k g administered daily for several days. Here, we report suppression by doses of DEX 30-100-fold lower than those required to induce CYP3A enzymes [42]. This suggests that glucocorticoids may have a physiological role in down regulation of specific P450 gene products, at least during episodes of increased secretion such as occurs during an inflammatory response. Blood levels of corticosterone in rats reach about 1.3 /xM after endotoxin treatment [45]. Assuming that the volume of distribution of D E X is the same in rats as in humans (0.82 I/kg, [46]), we can calculate that a dose of about 60 /xg/kg D E X is needed to achieve a peak plasma level of 18.6 nM. Since DEX is about 70-fold more potent than corticosterone [47], this is roughly equivalent to 1.3/xM of the endogenous glucocorticoid. Thus, the doses of D E X we have used are highly appropriate for the inflammatory model. Although the generally good agreement between the effects of cytokines on P450 mRNAs and proteins suggests that the down-regulation is effected primarily at a pretranslational level, the mechanisms by which this effect is achieved are yet to be elucidated. On one hand, cytokine-regulated transcription factors could be involved in the direct down-regulation of the P450 genes. Another possibility is that cytokine induction of heme oxygenase [48] is involved. By depleting cells of their heme, induction of this enzyme could impair P450 synthesis, contributing directly to the decrease in microsomal protein. It could also affect transcription of individual P450 genes [49], although the regulation of P450 gene transcription by heme is controversial [50]. Neither ILl nor IL6 alone, nor either cytokine in combination with DEX, has in our hands produced as great a suppression of CYP2CI 1 or CYP2C12 mRNAs

E.T. Morgan et al. /Biochimica et Biophysica Acta 1219 (1994) 475-483

in vivo as we can obtain with LPS injection [7]. Since ILl and IL6 act synergistically to stimulate transcription of various acute phase genes [51], it is interesting to speculate that both of these agents, and possibly other cytokines and glucocorticoids, act upon the liver in concert to produce a maximal suppression of P450 gene expression in inflammation. Hepatocyte culture systems should be useful tools to approach this question.

Acknowledgments This project was supported by grants DK39968 and GM46897 from the National Institutes of Health (E.T.M.). We are grateful to Hel~ne Gravel, Judy Wilson and Kirsten Tenney for excellent technical assistance.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17]

[18] [19] [20]

[21] [22]

Renton, K.W. (1986) Clin. Biochem. 19, 72-75. Renton, K.W. (1981) Biochem. Pharmacol. 30, 2333-2336. Brattsten, L.B. (1987) Experientia 43, 451-454. Batra, J.K., Venkitasubramanian, T.A. and Raj, H.G. (1987) Eur. J. Drug Metab. Pharmacokinetics 12, 109-114. Azri, S. and Renton, K.W. (1991) Int. J. Immunopharmacol. 13, 197-204. Mahu, J.-L. and Feldman, G. (1984) Enzyme 31, 234-240. Morgan, E.T. (1989) Mol. Pharmacol. 36, 699-707. Wright, K. and Morgan, E.T. (1990) FEBS Lett. 271, 59-61. Stanley, L.A., Adams, D.J., Lindsay, R., Meehan, R.R., Liao, W. and Wolf, C.R. (1988) Eur. J. Biochem. 174, 31-36. Wright, K. and Morgan, E.T. (1991) Mol. Pharmacol. 39, 468474. Williams, J.F., Bement, W.J., Sinclair, J.F. and Sinclair, P.R. (1991) Biochem. Biophys. Res. Commun. 178, 1049-1055. Fukuda, Y., Ishida, N., Noguchi, T., Kappas, A. and Sassa, S. (1992) Biochem. Biophys. Res. Commun. 184, 960-965. Barker, C.W., Fagan, J.B. and Pasco, D.S. (1992) J. Biol. Chem. 267, 8050-8055. Morgan, E.T. and Norman, C.A. (1990) Drug Metab. Dispos. 18, 649-653. Morgan, E.T. (1991) Biochem. Pharmacol. 42, 51-57. Knickle, L.C., Spencer, D.F. and Renton, K.W. (1992) Biochem. Pharmacol. 44, 604-608. Craig, P.I., Mehta, I., Murray, M., McDonald, D., ,~kstr6m, A., Van der Meide, P.H. and Farrell, G.C. (1990) Mol. Pharmacol. 38, 313-318. Bissell, D.M. and Guzelian, P.S. (1980) Ann. N.Y. Acad. Sci. 349, 85-98. Chomczynski, P. and Sacchi, N. (1987) Anal. Biochem. 162, 156-159. Kleinman, H.K., McGarvey, M.L., Hassell, J.R., Star, V.L., Cannon, F.B., Laurie, G.W. and Martin, G.R. (1986) Biochemistry 25, 312-318. Haugen, D.A. and Coon, M.J. (1976) J. Biol. Chem. 251, 79297939. Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275.

483

[23] Omura, T. and Sato, R. (1964) J. Biol. Chem. 239, 2370-2378. [24] Donahue, B.S., Skottner-Lundin, A. and Morgan, E.T. (1991) Endocrinology 128, 2065-2076. [25] Song, B.J., Gelboin, H.V., Park, S.S., Yang, C.S. and Gonzalez, F.J. (1986) J. Biol. Chem. 261, 16689-16697. [26] Tso, J.Y., Sun, X.H., Kao, T.H., Reece, K.S. and Wu, R. (1985) Nucleic Acids Res. 13, 2485-2502. [27] Gonzalez, F.J., Song, B.J. and Hardwick, J.P. (1986) Mol. Cell. Biol. 6, 2969-2976. [28] Goldstein, L.A. and Heath, E.C. (1984) J. Biol. Chem. 259, 9212-9217. [29] Henderson, G.S., Conary, J.T., Davidson, J.M., Stewart, S.J., House, F.S. and McCurley, T.L. (1991) BioTechniclues 10, 190197. [30] Hollander, M.C. and Fornace, A.J.Jr. (1990) BioTechniques 9, 174-179. [31] Guzelian, P.S., Li, D., Schuetz, E.G., Thomas, P., Levin, W., Mode, A. and Gustafsson, J.-A. (1988) Proc. Natl. Acad. Sci. USA 85, 9783-9787. [32] Castell, J.V., Klapproth, J., Gross, V., Walter, E., Andus, T., Snyers, L., Content, J. and Heinrich, P.C. (1990) Eur. J. Biochem. 189, 113-118. [33] Schulz, W.A., Eickelman, P., Hallbrucker, C., Sies, H. and H~iussinger, D. (1991) FEBS Lett. 292, 264-266. [34] Milland, J., Tsykin, A., Thomas, T., Aldred, A.R., Cole, T. and Schreiber, G. (1990) Am. J. Physiol. 259, G340-G347. [35] Abdel-Razzak, Z., Loyer, P., Fautrel, A., Gautier, J-C., Corcos, L., Turlin, B., Beaune, P. and GuiUouzo, A. (1993) Mol. Pharmacol. 44, 707-715. [36] Sujita, K., Okuno, F., Tanaka, Y., Hirano, Y., Inamoto, Y., Eto, S. and Arai, M. (1990) Biochem. Biophys. Res. Commun. 168, 1217-1222. [37] Chen, Y.L., Florentin, I., Batt, A.M., Ferrari, L., Giroud, J.P. and Chauvelot-Moachon, L. (1992) Biochem. Pharmacol. 44, 137-148. [38] Kurokohchi, K., Yoneyama, H., Matsuo, Y., Nishioka, M. and Ichikawa, Y. (1992) Biochem. Pharmacol. 44, 1669-1674. [39] Ferrari. L., Herber, R., Batt, A.-M. and Siest, G. (1993) Biochem. Pharmacol. 45, 2269-2277. [40] Guengerich, F.P. (1992) Annu. Rev. Pharmacol. Toxicol. 29, 241-64. [41] Morgan, E.T. (1993) Biochem. Pharmacol. 45, 415-419. [42] Heuman, D.M., Gallagher, E.J., Barwick, J.L., Elshourbagy, N.A. and Guzelian, P.S. (1982) Mol. Pharmacol. 21, 753-760. [43] Meehan, R.R., Forrester, L.M., Stevenson, K., Hastie, N.D., Buchmann, A., Kunz, H.W. and Wolf, C.R. (1988) Biochem. J. 254, 789-797. [44] Shimada, M., Murayama, N., Yamauchi, K., Yamazoe, Y. and Kato, R. (1989) Arch. Biochem. Biophys. 270, 578-587. [45] Rivier, C. (1990) in Circulating Regulatory Factors and Neuroendocrine Function (Potter, J.C. and Jezova, D., eds.), pp. 295-301, Plenum Press, New York. [46] Benet, L.Z. and Williams, R.L. (1990) in The Pharmacological Basis of Therapeutics (Gilman, A.G., RaU, T.W., Nies, A.S. and Taylor, P., eds.), 8th Edn., pp. 1650-1735, Pergamon Press, New York. [47] Haynes, R.C.Jr. (1990) in The Pharmacological Basis of Therapeutics (Gilman, A.G., Rail, T.W., Nies, A.S. and Taylor, P., eds.), 8th Edn., pp. 1431-1462, Pergamon Press, New York. [48] Rizzardini, M., Terao, M., Falciani, F., and Cantoni, L. (1993) Biochem. J. 290, 343-347. [49] Venkateswar, V. and Padmanaban, G. (1991) Arch. Biochem. Biophys. 290, 167-172 [50] Srivastava, G., Bawden, M.J., Hansen, A.J. and May, B.K. (1989) Eur. J. Biochem. 178, 689-692. [51] Baumann, H. and Gauldie, J. (1990) Mol. Biol. Med. 7, 147-159.