Glucocorticoid repression and basal regulation of the epoxide hydrolase promoter

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS

Vol. 279, No. 2, June, pp. 363-369,199O

Glucocorticoid Repression and Basal Regulation of the Epoxide Hydrolase Promoter’

Peter A. Bell, Charles N. Falany,2 Patricia McQuiddy, and Charles B. Kasper McArdle Laboratory for Cancer Research, 1400 University Avenue, Madison, Wisconsin 53706

Received November 6,1989; and in revised form February 23,199O

Through a series of promoter deletions and gene transfer experiments we have examined the basal regulation and glucocorticoid-mediated repression of the rat epoxide hydrolase gene. Three regions of the 5’ flanking sequence were found to influence the basal level of promoter function in H4IIE hepatoma cells. Re- gion A (-891 to -355 bp) contains an apparent repres- sor of epoxide hydrolase expression, while regions B (-271 to -171 bp) and C (-141 to -85) were found to contain important sequences required for optimal promoter activity. Previous work has demonstrated that dexamethasone represses epoxide hydrolase transcription by approximately 50% in isolated rat liver nuclei, and, in this study, we have demonstrated that the ability of the epoxide hydrolase promoter to drive CAT expression is similarly repressed in H4IIE cells treated with 1 pM dexamethasone. Furthermore, the level of endogenous epoxide hydrolase mRNA is decreased by 70- 88% in nontransfected H4IIE cells treated with dexamethasone. Interestingly, promoter activity was not decreased by dexamethasone in COS cells, which lack glucocorticoid receptors. The current data show that sequences from -42 to + 110 bp are sufficient to support the dexamethasone response, and, furthermore, they suggest that repression may not require direct interac- tion of the ligand-receptor complex with the promoter region. C 1990 Academic Press, Inc.

Microsomal epoxide hydrolase catalyzes the hydroly- sis of aliphatic and arene epoxides to trans-dihydrodiols. The enzyme is found in the endoplasmic reticulum (l- 3) and nuclear envelope (4-5) of mammalian cells, where it is associated with the cytochrome P450 mixed func-

’ This research was supported by Grants CA22484, CA09020, and CA09135 from the National Institutes of Health.

’ Current address: Department of Pharmacology, University of Rochester Medical Center, Rochester, NY 14642.

0003.9861/90 $3.00 Copyright 8 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.

tion oxidase system. Reaction of a number of xenobiotic compounds with cytochromes P450 can produce highly electrophilic epoxides which are capable of modifying cellular macromolecules. Epoxide hydrolase converts these intermediates into chemically less reactive diols. These diols can then serve as substrates in phase II reac- tions and ultimately be excreted in a variety of forms depending upon the conjugation reaction. It is ironic that in certain instances diols produced by epoxide hydrolase can be further metabolized by the cytochromes P450 to highly electrophilic diol epoxides that are in turn poor substrates for epoxide hydrolase (6) but form cova- lent derivatives with DNA as well as protein and RNA. Thus, microsomal epoxide hydrolase plays a central role in both the detoxication of xenobiotic compounds and in the metabolic activation of certain chemical carcino- gens.

A number of chemicals modulate epoxide hydrolase expression. For example, phenobarbital, 2-acetylaminofluorene, and trans-stilbene oxide are very effective inducers of epoxide hydrolase activity. Both phenobarbital (7) and 2-acetyl aminofluorene3 have been shown to act by augmenting transcription, while the mechanism of trans-stilbene oxide induction remains to be established. Similarly, intraperitoneal administration of lead acetate increases epoxide hydrolase transcription in isolated rat kidney nuclei.4 In contrast to transcriptional enhancement, synthetic glucocorticoids have been shown to repress epoxide hydrolase transcription in rat liver nuclei (8). In particular, dexamethasone inhibited epoxide hydrolase transcription by approximately 50%, and this decrease was reflected in lower levels of the enzyme in liver microsomes. Interestingly, administration of synthetic glucocorticoids to laboratory animals has been shown to protect against the effects of subsequent exposure to toxic chemicals (9), possibly via a decrease in the levels of microsomal epoxide hydrolase as well as

3 P. McQuiddy, unpublished results. 4 J. Sheehan and C. B. Kasper, manuscript in preparation.

363

364 BELL ET AL.

the induction of the steroid-inducible cytochromes P450. In general, modulation of the levels of drug metab- olizing enzymes would be expected to have a significant impact on the production of toxic and carcinogenic me- tabolites from xenobiotic compounds.

Of interest are the roles of specific DNA sequences on the basal and drug-influenced activity of the microsomal epoxide hydrolase promoter. Flanking sequences derived from the rat microsomal epoxide hydrolase gene (10) were used to drive expression of the CAT5 gene in transiently transfected rat, hepatoma H4IIE cells. We have examined the response of wild type and mutant promoter constructs to dexamethasone and have identified at least three &-acting elements important in regu- lating basal epoxide hydrolase expression.

MATERIALS AND METHODS

Steroids and DMSO were obtained from Sigma. DEAE-De&an was obtained from Pharmacia. Fetal calf serum and Dulbecco’s modified Eagle’s medium were purchased from GIBCO. [‘4C]Chloramphen- icol(60 mCi/mmol) was obtained from DuPont-NEN, and [32P]dCTP (800 Ci/mmol) was obtained from Amersham. Oligonucleotides were obtained from the University of Wisconsin Biotechnology Center. En- zymes and DNA linkers were obtained from reliable commercial sources.

Construction of plasmids. Plasmid pTZB18R was constructed by ligation of a BglII linker into the SmaI site of the phagemid pTZ18R (11). To construct pEHTZ18R, the PstI site at position +160 relative to the epoxide hydrolase transcription start site (10) was converted to Hind111 using the Klenow fragment of DNA polymerase I and Hind111 linkers. DNA was further digested with BamHI and HindHI, and the resultant 513-bp fragment was cloned into the BamHI and Hind111 sites of pTZ18R. Plasmid pSV&AT was constructed by excision of the AccI to Hind111 segment followed by conversion of the AccI site to Hind111 using the Klenow fragment of DNA polymerase I and Hind111 linkers. Plasmid pSVBglI1 was constructed by conversion of the NdeI site in pSV,CAT into a BglII site using T4 DNA polymerase and BglII linkers. pSVBamH1 was constructed by deletion of the BumHI site in pSV$AT and creation of a BamHI site by linker ligation at the mung bean nuclease treated NdeI site.

Epoxide hydrolase promoter mutants. Ba131 promoter deletion mutants were made as follows. pEHTZ18R was linearized with BamHI, and treated with B&31 nuclease to remove bases from the 5 end of the promoter region. DNA ends were blunted using T4 DNA polymerase, and B&II linkers were ligated to the repaired ends. Treated DNA was digested with BglII and Hind111 to produce deletion mutants of the promoter region. Fragments were gel isolated and re- cloned as BglII/HindIII inserts in pTZB18R. The ext.ent of Bat31 digestion was determined by DNA sequencing (12) in pTZB18R. Se- lected mutants were cloned into pSVBglI1 as BglII/HindIII fragments to produce plasmids p-355EH, p-271EH, p-171EH, p-141EH, p-85EH, and p-42EH. Plasmids p-891EH, p-398EH, p-144EH, and p-52EH were constructed using available restriction sites in the 5’ flanking sequence of the epoxide hydrolase gene. p-891EH was constructed by treatment of a PstI/PstI fragment (-891 bp to +160 bp) with the Klenow fragment of DNA polymerase I in the presence of nucleoside triphosphates. Hind111 linkers were ligated to the blunt ends, and, after digestion with HindIII, the fragment was cloned into the Hind111

’ Abbreviations used: CAT, chloramphenicol acetyl-transferase; bp, base pair; DMSO, dimethyl sulfoxide.

site of pSV&AT. BamHI/PstI, SacI/PstI, and PuuII/PstI digested epoxide hydrolase DNA was similarly used to construct p-398EH, p-144EH, and p-52EH, respectively.

To construct pIR-144EH, two oligonucleotides (5’ CACTGGACC- CTCCAGGTCACGAGCT 3’and 5’CGTGACCTGGAGGGTCCA. GTGAGCT 3’) which contain the inverted repeat between bases -140 and -121 relative to the epoxide hydrolase transcription start site were annealed and cloned directly into the Sac1 site of p-398EH. This resulted in a tandem duplication of the 20-bp repeat with 9 intervening base pairs.

To construct p-201/-85EH and p-144/-85EH, bases were removed from the 3’ end of the promoter region using ExoIII and Sac1 nucleases, respectively. DNA fragments were made blunt using mung bean nuclease, andBglI1 linkers were added. Following digestion with BglII, deleted fragments were ligated at the BglII site of p-85EH to produce internal deletion mutants p-201/-85EH and p-144/-85EH. Plasmid p+llOEH was prepared by linearizing p-398EH with Hind111 and removing bases from the 3’end of the promoter region with ExoIII. DNA ends were made blunt using Sl nuclease, and Hind111 linkers were added. DNA was digested with BamHI and HindIII, and the re- sulting fragment was cloned into pSV&AT to produce p+llOEH, which lacks the progesterone receptor consensus site at position +120. All plasmid constructs used for transfection of HIIIE cells were isolated with a CsCl density gradient according to standard procedures (13) and checked for integrity on agarose gels prior to use.

Treatment of cells. Glucocorticoid responsive H4IIE rat hepatoma cells (14) and COS monkey kidney cells were maintained in Dulbecco’s modified Eagle’s medium with high glucose supplemented with 10% heat-inactivated fetal calf serum. Cells were plated at a density of 2 X 10”per 35.mm dish 18 h before transfection. Transfection with plasmid DNA was accomplished by a modification (15) of the DEAE-Dex- tran method (16) followed by DMSO shock (17-19). Briefly, cells were washed two times with serum-free medium, incubated at 37°C under 5% COs in 50% serum-free medium-50% 50 mM TrissHCl, pH 7.3, containing 75 pg/ml DEAE-Dextran and 7 @g/ml plasmid DNA for 7.5 h. The medium was replaced with 10% DMSO shock solution for 2.5 min, and cells were washed once with serum-free medium and incubated for 3 days in medium containing 10% fetal calf serum and either 0.1% DMSO or drug in 0.1% DMSO. Medium was changed after the first 24 h.

CAl’assoy. Cells were collected from duplicate or triplicate dishes by trypsinization, washed with 150 mM NaCl, 40 mM Tris-HCl pH 7.6, 1 mM EDTA, and lysed in 100 ~1 of 250 mM Tris-HCl, pH 7.8, with three rounds of freezing and thawing. The supernatants were assayed for CAT activity (20) with modifications (19). Fifty microliters of cell supernatant, 27 ,ul of 250 mM TrisHCl, pH 7.8, and 3 ~1 (5 nmol) of [‘4C]chloramphenicol were mixed and prewarmed at 37°C. The reaction was started by the addition of 20 ~1 of 8 mM acetyl-CoA, incubated for 90 min at 37”C, and extracted twice with I ml ethyl acetate. Ex- tracts were dried, spotted on thin layer chromatography plates, and developed with 95:5 chloroform:methanol. After overnight exposure to film, spots corresponding to acetylated and unacetylated chloramphenicol were scraped into vials and counted. All counts obtained were corrected for total protein content in the cell lysates.

Isolation of RNA and Northern blotting. Total RNA was isolated from dexamethasone-treated and untreated H4IIE cells according to established methods (21). Poly(A)’ RNA was prepared via oligo(dT) cellulose chromatography as previously described (22). Poly(A)+ RNA (5 pg) from dexamethasone-treated and untreated cells was fraction- ated by agarose gel electrophoresis in formaldehyde (13). Gels were blotted onto nitrocellulose membranes and probed with a nick-trans- lated BamHI to Sal1 fragment derived from the rat epoxide hydrolase cDNA (23). Hybridization conditions were as previously described (24). Following hybridization, autoradiographs were prepared and sec- tions of the filter corresponding to epoxide hydrolase mRNA were excised and quantitated by liquid scintillation counting.

REGULATION OF THE EPOXIDE HYDROLASE PROMOTER 365

11 10 9 8 7 6 5 4

-log [Dexamethasonel

FIG. 1. Response of the epoxide hydrolase promoter to dexamethasone. Plasmid p - 398EH was used to transfect rat H4IIE hepatoma cells. Cells were incubated in medium containing various amounts of dexamethasone for ‘72 h as described under Materials and Methods. Each point represents the average of at least three trials, and CAT activities are expressed as percentage activity of transfected cells treated with vehicle alone.

RESULTS

Dose Response of the Epoxide Hydrolase Promoter to Dexamethasone

As a first step in studying the glucocorticoid-mediated

repression of the epoxide hydrolase promoter, the response of p-398EH to various concentrations of dexamethasone was examined (Fig. 1). H4IIE cells transfected with p-398EH were incubated with 10-l’ to 10e5 M dexamethasone for 72 h, and the degree of repression of transcription was measured by CAT assay as described under Materials and Methods. Transfected cells treated with 10-l’ and lo-’ M dexamethasone produced CAT activities which were essentially equal to control activities. At dexamethasone concentrations above lo-’ M, epoxide hydrolase promoter activity decreased approximately 20% with each lo-fold increase in the con- centration of steroid, until a maximum repression of about 50% was reached. This maximal repression of promoter function was obtained when H4IIE hepatoma cells were incubated for 72 h in the presence of 10m7 to lop6 M dexamethasone (Fig. 1).

In contrast to the repression noted in H4IIE cells, CAT expression was not repressed in p-891EH-transfected COS cells treated with 1 FM dexamethasone. Spe- cifically, COS cells treated with dexamethasone expressed CAT at 112 + 3 percent of control values. This observation is consistent with the reported absence of glucocorticoid receptors in this cell line (25) and suggests the participation of glucocorticoid receptor in epoxide hydrolase repression by dexamethasone. To determine if the observed dexamethasone repression of p- 398EH in H4IIE cells also held true for the endogenous epoxide hydrolase gene, a northern blot was done using poly(A)+ RNA from dexamethasone-treated and control

H4IIE cells (Fig. 2). Radioquantitation of the excised “‘P-labeled gel bands demonstrated that 1 PM dexamethasone repressed the level of cellular epoxide hydrolase mRNA by 70%, indicating that regulation of epoxide hydrolase expression in dexamethasone-treated H4IIE cells is similar to that observed in uiuo (8). In a separate experiment, 88% repression was noted.

Effect of Various Steroids on Epoxide Hydrolase Promoter Activity

In order to determine whether the observed repression of promoter function is due to the action of glucocorticoids or to a more general effect of steroid hormones, p- 891EH-transfected H4IIE cells were treated with 1 PM

concentrations of /3-estradiol, testosterone, and various glucocorticoids (Table I). All of the glucocorticoids tested repressed epoxide hydrolase promoter function from 55 to 66% relative to control cells while neither p- estradiol nor testosterone resulted in any change in promoter activity (Table I). These results are similar to those obtained for cytochrome P450pcN induction in pri- mary cultures of rat hepatocytes (26).

Activity of Epoxide Hydrolase Promoter Mutants

In order to determine the role of specific DNA sequences in both basal expression and dexamethasone repression, we constructed deletion mutants of the epoxide hydrolase promoter (Figs. 3 and 4). With the exception

A B C

2322 2027

FIG. 2. Dexamethasone-mediated repression of epoxide hydrolase mRNA levels in H4IIE hepatoma cells. Northern blot of poly(A)’ RNA from dexamethasone-treated and untreated H4IIE cells. The gel was blotted onto nitrocellulose and probed with radiolabeled epoxide hydrolase cDNA. (A) Poly(A)+ RNA (5 Kg) isolated from control cells. (B) poly(A)’ RNA (5 Kg) isolated from cells treated with 1 KM dexamethasone for 72 h. (C) 0.5 pg poly(A)+ RNA isolated from rat liver. Numbers at the left of the gel represent the location of DNA size mark- ers.

366 BELL ET AL.

TABLE I

Effect of Various Steroids on the Promoter Activity of Epoxide Hydrolase”

Steroid (1 pM) % Acetylated

chloramphenicol % Control

activity

Control Dexamethasone Betamethasone Triamcinolone Prednisolone Cortisol P-Estradiol Testosterone

4.7 k 0.6 100 1.6? 0.1 34 1.6 k 0.3 34 2.1 f 0.5 45 2.1 + 0.3 45 1.9 + 0.4 40 4.6 t- 0.4 98 4.6 i- 0.9 98

n H4IIE cells were transfected with p-891EH and treated with 1 pM

concentrations of the steroids listed. CAT activities were measured as described under Materials and Methods and are shown &SD. The results obtained for each steroid are also expressed as percentage of control activity.

of p+llOEH, all mutants include 161 bp of untranslated exon 1 and various amounts of 5’ flanking sequence linked to the bacterial CAT gene in pSV&AT. H4IIE cells were transfected with each construct and incubated in the presence or absence of 1 PM dexamethasone. In parallel experiments H4IIE cells were transfected with pSV,CAT and pSV&AT as controls. The ability of the various EH promoter mutants to drive transcription of the CAT gene was measured enzymatically as described under Materials and Methods. At least two dishes of p- 398EH-transfected cells were included in every experiment, and all CAT activities are expressed as a percentage of p-398EH activity measured in the absence of steroid.

Sequential deletion of nucleotides from the 5’ end of p-891EH resulted in a progressive increase in basal epoxide hydrolase promoter activity (Fig. 4). Removal of DNA sequences between -891 and -355 bp produced a twofold increase in CAT expression, while successive deletions between -271 and -85 nucleotides revealed at least two regions (B and C) important in maintaining high levels of epoxide hydrolase expression. Deletion of the region from -271 to -171 resulted in a 37% relative decrease in basal activity, and continued removal of nucleotides -141 to -85 further reduced the basal activity by 77%. Taken together, these observations suggest the presence of at least three regions important to the basal expression of epoxide hydrolase. The sequence between nucleotides -891 and -355 appears to contain an element(s) that represses EH promoter activity, while sequences between -271 and -171 and between -141 and -85 define two positive regulatory regions. There are no striking homologies to known c&-regulatory elements in the region from -271 to -171; however, the -141 to -85 region contains a conspicuous 20 nucleotide inverted repeat (5’~CACTGGACCCTCCAGGTCAC-3’) and a

possible enhancer sequence (5’-GTGGTTC-3’) (Figs. 3 and 4).

To confirm the apparent positive regulatory role of the region from -141 to -85, nucleotides -144 to -85 were deleted from p-398EH to produce p-144/-85EH. How- ever, instead of the expected decrease in activity, p-144/ -85EH and p-398EH displayed almost identical activities that were sevenfold greater than the promoter activity of p-85EH (Fig. 4). Comparison of constructs p- 398EH, p-85EH, and p-144/-85EH shows that the loss of activity observed when the 20-bp repeat and enhancer-like sequences are removed is compensated by restor-

-398 -355 r

GGATCCCAAGATC~CATAGTCTATCAGACTTAGCAGCAC CCTAGGGTTCTAGTTTGTATCAGATAGTCTGAATCGTCGTCGTTCACG~TGAGC~CTCG

CTTCTCAC~AGACTAATA~ACAGTTTTA~XAGAAT~ATTTCTAA~ATGTCTGTTTA GAAGAGTGGTCTGATTATGTGTCAAAATTTGGTTCTTAATAAAGATTTTTACXAUT

-27 1 I.

TACTTTCAGGCCACAGTC~TCCATGGGTiAACAAAACCkAATAAGGhTGGGCACTGG ATGAAAGTCCGGTGTCAGCAGGTACCCATTTGTTTTGGAGCC

-171 I. .

GGCAAGACCAAC4TC4GTCATCCTAATGACCTTTGCCTATATTCTGTG~~TTTTTTTT CCGTTCTGGTTGTACTCAGTAGGATTACTGGAAACGGATATAAGACACCCCW

-144 -141 I rr

CTTCCAATGGAGCTCCAGCACTGGACCCTCCAGGTCACTGTCT GAAGGTTACCTCGAGGTCGTGACCTGGGAGGTCCAGTGACAGTGACACCAAGTCAAACGA

-85 -52 -42 I . I I .

GTCTAAAGkCAGGGTGATAACTGCCCTGTGTGCACAA&ATCAGCTGAGTCATTGGGG C4GATTTCAGGTCCCACTATTGACGGGACACACGTGTTGTCTAGTC~CT~GT~CCCC

i;. . CTATGAAAiCATAAAAAGTCAGGTTTGG~AGGAATCAC ACTCAGCCACTCTACTTCGTG GATACTTTCGTATTTTTCAGTCCAAACCATCCTTAGTG TGAGTCGGTGAGATGAAGC

GGGATAATiCTTCCCAGAiAAGGMAAGbAGGCACTTCTGTTCCCAGGGAAAACAACAGG CCCTATTAAGAAGGGTCTTTTCCTTTTCGTCCGTGAAGACTCCCTTTTGTTGTCC

AGCACTTTGGACCTCCCTGCCAAGCTT TCGTGAAACCTGGAGGGACGGTTCGAA CAT I

FIG. 3. Sequence of the epoxide hydrolase promoter region and locations of deletion endpoints. Ba131 deletion endpoints are shown by dark vertical lines above the DNA sequence and are numbered relative to the major transcription start site. Locations of CCAAT (-155 to -151) and TATA (-29 to -22) consensus sequences are indicated. The 20-bp inverted repeat (-140 to -121), an enhancer-like element (-113 to -105), glucocorticoid receptor-like sequence (-68 to -56), progesterone receptor consensus sequence (+115 to +120), and a segment with homology to the 20-bp inverted repeat (-224 to -219) are shown underlined. BarnHI, SacI, and PuuII cleavage sites are located at -398, -144, and -52, respectively. The arrow indicates the previously determined transcription start site (10). The 3’ boundary of p+llOEH is positionedat +llO.

REGULATION OF THE EPOXIDE

lnvcrtcd

p-398EH

% p-398EH Control O/b Dcxamcthasonc 0 20 40 60 so 100 120 140 I-Cpl-eSSiOn

I II $1 1 IJ

I P 44 k 11

43* 11

p-355Ell

p-271EH

p-17lEll

p144EIl

p-141EH

47 + 7

35 + 6

49 t 5

45 * 8

39f 11

p-85EH F 38 + 7

p-52Ell 30 2 5

0 I I un Dun

III I n I II I 1

A BC -

FIG. 4. Activity and dexamethasone repression of epoxide

p-42EI I iii? 33 i 4

l-l p-20 l/-85ElH

46k 13

57 * 8

47 f 11

hydrolase promoter mutants. Schematic representation of portions of the EH promoter region are shown at left. Each construct includes various lengths of 5’ flanking sequence and 161 bp of untranslated exon 1 except for p+llOEH. All fragments were cloned 5’ to the CAT gene in pSV,CAT. The transcription start site is designated as tl, and the positions of CAAT, TATA, the enhancer element, the inverted repeat, and the progesterone receptor (PRE) sequences are indicated. At lower left are shown three functional regions labeled A, B, and C. The ability of the EH promoter mutants to drive CAT expression is given as a percentage of the activity obtained with p-398EH. At right, open bars represent control values, and stippled bars represent activity upon treatment with 10 6 M

dexamethasone. Each bar represents the average ? SD of at least nine dishes, except for dexamethasone-treated p-201/-85EH, which was assayed only twice (six dishes). At far right is shown the average extent of repression for each construct assayed +-SD.

p-89 1 EH

HYDROLASE PROMOTER 367

ing upstream elements. Furthermore, deletion of an additional 54 base pairs, including the CAAT sequence, from p-144/-85EH produces a construct (p-201/-85EH) with approximately one-third the promoter activity. These observations further support the positive role of the -271 to -171 region in maintaining transcription of the epoxide hydrolase gene. Duplication of the 20-bp inverted repeat at position -149 (pIR-144EH), 9 nucleotides 5’ from the native sequence, was ineffective in en- hancing promoter function.

Glucocorticoid-Mediated Repression

In parallel experiments, the mutant epoxide hydrolase promoter constructs were assayed for their ability to drive CAT expression in cells treated with dexametha-

sone. As shown in Fig. 4, dexamethasone treatment resulted in a 30 to 57% repression of CAT activity in all of the constructs tested. The differences in CAT activities between treated and untreated cells are statistically significant (P < 0.05) in every case. In contrast, dexamethasone treatment of H4IIE cells transfected with pSV&AT or pSV,CAT did not affect the level of CAT expression. Specifically, pSV&AT and pSV&AT were expressed at approximately 4 and 6%, respectively, of the p-398EH value in untreated cells. Interestingly, none of the sequences altered in the epoxide hydrolase promoter constructs can account for glucocorticoid repression of the epoxide hydrolase promoter.

DISCUSSION Epoxide hydrolase expression is induced by a number

of exogenous and endogenous compounds. This increase

368 BELL ET AL.

is the result of augmented transcription in animals treated with phenobarbital (7), and acetylaminofluorene or its metabolite N-OH-acetylaminofluorene.3 Sim- ilarly, treatment with lead acetate results in a fivefold increase in the rate of epoxide hydrolase transcription in rat kidney.4 In contrast, recent work has demonstrated a substantial decrease in transcription of the epoxide hydrolase gene in animals treated with glucocorticoids, specifically dexamethasone (8). Of particular interest is the mechanism by which these modulators of transcription act. The experiments presented here have examined the response of the epoxide hydrolase promoter region to various steroids and have attempted to define DNA sequence elements important to both the basal activity and the dexamethasone repression of the epoxide hydrolase promoter.

The dexamethasone response was assayed in rat hepatoma H4IIE cells (14) which were shown to be responsive to glucocorticoids in a manner that mimics re- sponses observed in uiuo. For example, transcription of the phosphoenol pyruvate carboxykinase gene is induced sevenfold by 5 X 10M7 M dexamethasone in H4IIE cells (27). Furthermore, we have shown that the endogenous level of epoxide hydrolase mRNA is markedly reduced following treatment of H4IIE cells with 1O-6 M

dexamethasone. Consequently, we conclude that these cells provide an appropriate system for studying modulation of the epoxide hydrolase promoter by glucocorticoids.

Treatment of p-398EH transfected cells with various concentrations of dexamethasone revealed a maximal repression in the range of 10d7 to 10e6 M dexamethasone (Fig. 1). This response to glucocorticoids is similar to that observed for induction of a number of transfected genes in cell culture (28,34) and is consistent with a glucocorticoid receptor-mediated process. Further, when COS cells, which lack glucocorticoid receptors (25), were transfected with p-891EH and treated with dexamethasone, no significant change in the level of CAT expression resulted. This observation lends support to the idea that repression of epoxide hydrolase transcription is mediated either directly or indirectly by the glucocorticoid receptor. Interestingly, expression of CAT in all of the constructs tested in HUIE cells was repressed by dexamethasone between 30 and 57%, similar to that noted for in vitro nuclear transcription experiments (8). Notably, glucocorticoid-mediated repression was not significantly lost in any of the constructs tested, and dexamethasone- mediated repression was not seen in cells transfected with pSV,CAT or pSV&AT. These observations suggest a number of mechanisms by which glucocorticoids may act to inhibit transcription from the epoxide hydrolase promoter. First, repression may occur as a second- ary effect of the hormone. Glucocorticoids may alter the synthesis or availability of factors critical to the transcription of the epoxide hydrolase gene in H4IIE cells.

Such an effect is consistent with the observation that deletion to within 14 bp of the TATA-like element still results in a 33% dexamethasone-mediated repression of epoxide hydrolase promoter driven CAT expression (Fig. 4, p-42EH). Second, ligand-induced binding of the glucocorticoid receptor to a site in the promoter could interfere with binding of other factors important to transcription. An analogous effect has recently been described for sterol-mediated repression of the low density lipoprotein receptor gene (29). It was suggested that ste- rols stimulate binding of a protein to promoter sequences and thereby inhibit binding of the constitutive activator SPl to an adjacent site. More recently, it has been shown that glucocorticoid receptor binding inter- feres with a CAMP responsive element in the human gly- coprotein hormone o-subunit gene (30). Third, glucocorticoid receptor binding to a site 3’ of the transcription start site could interfere either with passage of RNA polymerase or affect conformation-dependent binding of other factors further 5’.

Sequence analysis of the epoxide hydrolase promoter sequences 5’ to the transcription start site has revealed a segment with homology to a glucocorticoid receptor consensus site (31) from -69 to -57. This element is weakly functional, if at all, and does not account for the observed repression, since p-42EH, which lacks this sequence, is inhibited 33% by dexamethasone. Analysis of DNA sequences within exon 1 had identified a segment (f120 to f125) with absolute homology to a previously characterized progesterone receptor site (32) having the sequence TGTTCC. Glucocorticoid and progesterone receptor proteins are 91% homologous in their DNA binding domains (33) and display similar DNA binding properties (34). Removal of the progesterone receptor element, however, does not relieve dexamethasone repression (Fig. 4, p+llOEH). This result and the absence of other glucocorticoid receptor sites in the region from -42 to +lOO bp suggests that repression of epoxide hydrolase transcription may not be directly mediated by glucocorticoid receptor binding to the promoter region. Instead, glucocorticoid repression may be more closely associated with the reduction in liver of a factor(s) nec- essary for the initiation of transcription or with an increase in a negative regulatory factor.

Earlier analysis of the epoxide hydrolase promoter revealed sequences analogous to TATA, CAAT, and enhancer-like elements (10). In addition, a conspicuous 20- bp inverted repeat was noted. Deletion analysis of the promoter region suggests the presence of at least three DNA regions (A, B, and C) important to the basal expression of epoxide hydrolase (Fig. 4). DNA sequences between -891 and -355 bp (A) appear to define a repres- sor function, since a twofold increase in basal promoter activity is noted upon their deletion. Further deletion of the -141 to -85-bp region (C), which contains the 20- bp inverted repeat and enhancer-like elements resulted

REGULATION OF THE EPOXIDE HYDROLASE PROMOTER 369

in a 77% decrease relative to construct p-141EH. Inter- estingly, activity was restored to the level of p-398EH when sequences further 5’ were included (Fig. 4, p-1441 -85EH). Deletion of the region between -271 and -171 bp (B) resulted in a 37% decrease in basal promoter activity, and a significant portion of this reduction may be mediated by sequences between -201 and -171 (Fig. 4, compare activities of p-201/-85EH, p-144/-85EH and p- 171EH). Comparison of region B with a recent compila- tion of &-acting DNA sequence elements (35) has revealed no striking homologies. However, internal comparison of this sequence with other portions of the epoxide hydrolase promoter showed a 7 nucleotide sequence, GCACTGG (-225 to -219 bp), with overlap- ping homology to the ends of the 20-bp inverted repeat (-140 to -121 bp). It is possible that this element is important to the promoting activity of both the -271 to -171 bp and the -141 to -85 bp regions and will be the subject of further study. Additional mutations both 5’ and 3’ to the transcription start should further elucidate the basal and drug responsive DNA elements critical to epoxide hydrolase transcription.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23. ACKNOWLEDGMENTS

We thank Dr. Henry Pitot for providing H4IIE cells and Dr. Bill Sugden for kindly supplying pSV&AT DNA.

24.

25.

REFERENCES 26.

1. Oesch, F., Jerina, D. M., and Daly, J. (1971) Rio&m. Biophys. Acta 227,685-691. 27.

2. Lu, A. Y. H., Ryan, D., Jerina, D. M., Daly, J. W., and Levin, W. (1975) J. Biol. Chem. 250,8283-8288.

3. Oesch, F., Glatt, H., and Schmassman, H. (1977) Rio&em. Phar- mncol. 26, 603607.

28.

4. Fahl, W. E., Jefcoate, C. R., and Kasper, C. B. (1978) J. Biol. Chem. 253, 3106p3113.

5. Gonzalez, F. ,J., and Kasper, C. B. (1982) Mol. Pharmacol. 21, 51 l-516.

6. Yang, S. K., McCourt, D. W., Leutz, J. C., and Gelboin, H. V. (1977) Science 196,1199-1201.

7. Hardwick, J. P., Gonzalez, F. J., and Kasper, C. B. (1983) J. Biol. Chem. 258,8081-8085.

8. Simmons, D. L., McQuiddy, P., and Kasper, C. B. (1987) J. Biol. Chem. 262,326-332.

9. Selye, H. (1971) J. Pharm. Sci. 60, l-28.

29.

30.

31.

32.

33. 34.

35.

Falany, C. N., McQuiddy, P., and Kasper, C. B. (1987) J. Eliol. Chem. 262,5924-5930.

Mead, D. A., Szczesna-Skorupa, E., and Kemper, B. W. (1986) Protein Eng. 1,67-74. Sanger, F., Nicklen, S., and Coulson, A. R. (I 977) I’roc. N&l. Acad. Sci. USA 74,5463&5467.

Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory.

Pitot, H. (1964) N&l. Cancer Inst. Monogr. 13,229-242. Danna, K. J., and Sompayrac, L. M. (1982) J. Viral. Methods 5, 335-341.

McCutchan, J. H., and Pagano, d. (1968) J. N&l. Cancer Inst. 41, 351-357.

Lewis, W. H., Srinivansam, P. R., Stokoe, N., and Siminovitch, L. (1980) Somatic Cell Genet. 6,333-348. Sussman, D. J., and Milano, G. (1984) Mol. Cell. Riol. 4, 1641l 1643. Lopata, M. A., Cleveland, D. W., and Sollner-Webb, B. (1984) Nucleic Acids Res. 5,570775717. Gorman, C., Moffat, L., and Howard, B. (1982) Mol. Cell. Biol. 2, 1044~1051. Chirgwin, J. M., Przybyla, A. E., MacDonald, R. J., and Rutter, W. J. (1979) Biochemistry 18,5294-5299. Gonzalez, F. J., and Kasper, C. B. (1980) Biochemistry 19,1790- 1796. Porter, T. D., Beck, T. W., and Kasper, C. B. (1986) Arch. Bio- them. Biophys. 248,121-129.

Gonzalez, F. J., and Kasper, C. B. (1982) J. Biol. Chem. 257, 5962-5968. Giguere, V., Hollenberg, S. M., Rosenfeld, M. G., and Evans, R. M. (1986) Cell 46,645-652. Schuetz, E. G., Wrighton, S. A., Barwick, d. L., Guzelian, P. S. (1984) J. Biol. Chem. 259,1999-2006. Petersen, D. D., Magnuson, M. A., and Granner, D. K. (1988) Mol. Cell Biol. 8,96-104. Jantzen, H.-M., Stahle, U., Gloss, B., Stewart, F., Schmid, W., Boshart, M., Miksicek, R., Schutz, G. (1987) Cell 49,29-38. Dawson, P. A., Hofmann, S. L., van der Westhuyzen, D. R., Siid- hof, T. C., Brown, M. S., and Goldstein, J. S. (1988) J. Biol. Chrm. 263,3372-3379.

Akerblom, I. E., Slater, E. P., Beato, M., Baxter, J. D., and Mellon, P. L. (1988) Science 241,350-353. Scheidereit, C., Geisse, S., Westphal, H. M., and Beato, M. (1983) Nature (London) 304,749-752. Bailly, A., LePage, C., Rauch, M., and Milgrom, M. (1986) EMBO J. 5,3235-3241. Green, S., Chambon, P. (1986) Nature (London) 324,615-617. Strahle, U., Klock, G., Schutz, G. (1987) Proc. Natl. Acad. Sci. (ISA 84, 7871-7875.

Wingender, E. (1988) Nucleic Acids Res. 16.187991902.

Glucocorticoid repression and basal regulation of the epoxide hydrolase promoter

Documents

Transcript of Glucocorticoid repression and basal regulation of the epoxide hydrolase promoter