Transcription activation of the tyrosine aminotransferase gene by glucocorticoids and cAMP in...

Eur. J. Biochem. 165,499- 506 (1987) 0 FEBS 1987

Transcription activation of the tyrosine aminotransferase gene by glucocorticoids and cAMP in primary hepatocytes Erika SCHMID', Wolfgang SCHMID', Michael JANTZEN', Doris MAYER', Bernd JASTORFF3 and Gunther SCHUTZ'

Institut fur Zell- und Tumorbiologie Institut fur Experimentelle Pathologie am Deutschen Krebsforschungszentrm, Heidelberg Fachbereich Biologie/Chemie, Universitit Bremen

(Received October l6,1986/December 31,1986) - EJB 86 1103

The expression of the tyrosine aminotransferase (TAT) gene of the rat was analyzed in primary hepatocytes. The TAT gene remains active in primary cultured cells at a level similar to that in liver cells. Expression can be induced by glucocorticoids and CAMP, glucocorticoids lead to a 8 - 10-fold increase in TAT mRNA level, CAMP to a 20 - 30-fold increase. The elevation of the TAT mRNA is preceeded by a rise in the relative rate of transcription of the gene. Surprisingly transcription of the albumin gene, which steadily declines with the age of the culture, can also strongly be stimulated by glucocorticoids in primary hepatocytes. CAMP antagonists, which act as competitive inhibitors of the CAMP-dependent protein kinase, prevent induction of transcription of the tyrosine aminotransferase gene by CAMP suggesting that the effect of CAMP on expression of the tyrosine aminotransferase gene is mediated by a CAMP-dependent protein kinase. The cAMP antagonist does not interfere with induction by glucocorticoids which suggests that phosphorylation of the glucocorticoid receptor by the CAMP-dependent protein kinase is not required for its function. We thus conclude that the two inducers affect transcription by independent mechanisms.

We are interested in the analysis of expression of the tyrosine aminotransferase (TAT) gene since this system illustrates well the complexity of gene control mechanisms occurring in higher eucaryotes [I]. The enzyme is synthesized exclusively in parenchymal cells of the liver [2]. Enzyme synthesis is under hormonal control: glucocorticoids and glucagon or its intracellular mediator cAMP lead to a rapid increase in enzyme synthesis which results from a concomitant increase in the TAT mRNA level [I, 3, 41. The enzyme is not synthesized in appreciable quantities before birth. The expression of the TAT gene is rapidly activated within a few hours following birth and is thus a typical representative of the gene cluster activated in the neonatal period [5]. Further- more, the activity of the gene is affected by two distinct genetic loci whch appear to act in trans since their chromosomal localization is different from that of the TAT structural gene [6]. One locus is represented by the albino lethal mutations: deletions in the vicinity of the c locus on chromosome 7 lead to deficient synthesis of several liver enzymes including TAT [7]. The deficiency of the enzyme is due to markedly reduced levels of TAT mRNA [8]. The second locus affecting TAT expression is illustrated by the phenomenon of extinction. If chromosome 11 from fibroblasts is introduced into hepatoma cells, TAT enzyme activity and TAT mRNA disappear. The regulatory locus involved has been called a tissue-specific extinguisher, since it appears to interfere with the expression of such liver-specific genes as TAT [9].

This manuscript is dedicated to Professor Peter Carlson on the occasion of his 70th anniversary.

Correspondence to G. Schutz, Institut fur Zell- und Tumorbiologie am Deutschen Krebsforschungszentrum, Im Neuenheimer Feld 280, D-6900 Heidelberg, Federal Republic of Germany

Abbreviation. CPT-CAMP, 8-(4-~hlorophenylthio)-cAMP.

To elucidate the regulatory mechanism of these complex control events it is imperative to determine at which level hormonal effectors or developmentally important regulators act. In previous experiments using intact animals we could demonstrate that glucocorticoids and CAMP lead to an increased level of TAT mRNA [3]. An increased rate of transcription of the TAT gene, is at least in part, responsible for CAMP-dependent induction of TAT mRNA [4]. The observed increase in transcriptional rate following cAMP treatment was smaller than the rise of TAT mRNA level, possibly because of an increased basal level induced by experimental handling. To avoid this problem we used primary cultures of hepatocytes to study in detail the effect of these two inducers, dexamethasone and CAMP. The collagenase perfusion technique for dissociation of hepatocytes has allowed a highly reproducible way of establishing primary hepatocyte cultures with high viability and maintenance of many normal liver functions [lo]. Using cultured hepatocytes we were able to investigate the influence of the synthetic glucocorticoid dexamethasone and the CAMP derivative 8-(4-chlorophenylthio)-cAMP (CPT- CAMP) on TAT gene expression under the more defined conditions of the in vitro system.

Whereas glucocorticoid induction of gene expression is relatively well understood, the way cAMP induces specific gene transcription is obscure. Most if not all effects of CAMP in mammalian cells are mediated by CAMP-dependent protein kinases. It is, however, unknown whether the kinases are involved in the control of gene transcription. The availability of cAMP antagonists allows a more direct analysis of the role of the CAMP-dependent protein kinases in the induction of TAT gene expression [I 1 - 151. Studies on the effect of the diastereomeric forms of adenosine cyclic 3',5'-phosphorothioates, (Sp)-cAMP[S] and (Rp)-cAMP[S], have demonstrat-

500

ed that (Sp)-cAMP[S] and (Rp)-cAMP[S] are agonists and antagonists, respectively, of the mammalian CAMP-dependent protein kinase [I3 - 151. The activity of mixtures of different concentrations of the antagonist (Rp)-cAMP[S] with cAMP or the agonist (Sp)-cAMP[S] reveals that (Rp)- cAMP[S] is a competitive antagonist of cAMP [15]. These cyclic nucleotide derivatives are thus suitable as tools for the elucidation of the molecular interaction between cAMP and its receptor protein(s) which is required for transcriptional activation of the TAT gene. Here we show that dexamethasone and cAMP affect TAT gene expression by control of the rate of transcription of the gene, that these inducers exert their effects by independent mechanisms and that cAMP effects on TAT gene expression are mediated by the protein kinase.

EXPERIMENTAL, PROCEDURE

Isolation and culture ojliver cells

Hepatocytes were isolated from adult male rats (Sprague- Dawley, 250 - 350 g body weight) by the collagenase perfusion method [lo] with modifications as described by Faulstich et al. [16]. Hepatocytes were washed and cultured at a density of 5 x lo6 cells/culture (1 x lo5 cells/cm2) on 100-mm tissue-culture dishes (Falcon) in Hams F12 medium supplemented with 10% fetal calf serum, insulin (1 pg/ml), penicillin (100 Ujml) and streptomycin sulfate (100 pg/ml) for 48 h. Thereafter the cells were incubated in serum-free medium with the inducers as indicated in the figure legends. Dexamethasone and 8-(4-~hlorophenylthio)-cAMP were used at a Concentration of 1 pM and 10 pM respectively.

Isolation of R N A und blot analysis

Total RNA was isolated and analyzed by blot hybridization as described [4, 81. The blots were probed with uniformly 32p-labelled complementary SP6 RNA transcribed from a plasmid containing a TAT cDNA clone [3]. Specific activity of the probe was approximately 5 x lo8 dpm/pg RNA and the probe was used at a concentration of 25 x lo7 dpm/ml. After exposure of the filters, bands were cut out and the radioactivity quantified by liquid scintillation counting.

Analysis o j the relative rate of transcription

Nuclei were prepared from primary hepatocytes by detergent treatment according to Becker et al. [17]. Transcrip- tion reactions and preparation of labelled RNA were done as described by Danesch et al. [18]. For analysis of the transcriptional rate 0.25 pmol cloned single-copy genomic DNA or cloned cDNA were bound to nitrocellulose filters using a Schleicher & Schiill Minifold I. Hybridization was done as described by Danesch et al. [18]. After hybridization the filters were washed four times with 2 x standard saline citrate (NaC1,'Cit) 0.1% SDS at room temperature, then twice for 45 min with 0.2% NaCl/Cit 0.1% SDS at 65"C, followed by RNase digestion (5 pgiml RNase A, 5 U/ml TI RNase) for 30 min at 37 C (in 2 x NaCl/Cit). After a wash in 2 x NaCl/ Cit, 0.1% SDS at 45 C for 30 min, filters were exposed to X-ray film. For quantification the dots were cut out and the bound radioactivity was determined by liquid scintillation counting. A 3H-labelled TAT RNA (SP6-cTAT 3 [3]) was included in the hybridization reactions to correct for hybridization efficiency. The DNA fragments used for hybridization

were subclones of single-copy fragments of the TAT gene: TAT EH 0.95, TAT EH 2.45, TAT EE 1.05 [4]; of the tryptophan oxygenase gene: TO EE 3.6, TO XE 5.4, a subclone of the TO XX 6.0, [18]; of the albumin gene: albumin B, C, D [19] and cDNA clone of the al-antitrypsin mRNA and the cDNA clone, named pliv2 [20].

Synthesis of ( S p ) -cAMP[S] and (Rp) -cAMP[S]

The diastereomeric adenosine cyclic 3',5'-phosphorothioates were synthesized according to published procedures

RESULTS

Expression of TAT mRNA in primary hepatocytes

To establish the usefulness of primary hepatocytes for the intended analysis of the hormonal and cell-specific control of TAT gene expression a comparative analysis of the mRNA content and transcriptional rates of the TAT gene and other genes specifically expressed in liver was performed. Since expression of the albumin gene has previously been intensively studied in hepatocytes [22] we compared expression of the TAT gene with that of the albumin gene. The mRNA levels for TAT and albumin in livers from adrenalectomized rats and in primary hepatocytes were determined in the presence and absence of the inducers dexamethasone, a synthetic glucocorticoid, and 8-(4-~hlorphenylthio)-cAMP, a stable cAMP derivative with high inducing potency. Maximal induction was achieved with 1 pM dexamethasone and 0.05 mM CPT-CAMP (unpublished data). As cell density is an important factor for maintaining various metabolic activities in primary hepatocytes, cells were seeded at high cell density (lo5 cells/cm2) [23]. A representative experiment is shown in Fig. 1. Total RNA from liver and primary hepatocytes was prepared, electrophoresed on a denaturating agarose gel, transferred to nitrocellulose and probed with RNA probes complementary to TAT and albumin mRNA sequences.

As seen in Fig. 1A (lanes 1 and 2) there is no difference in the uninduced levels of TAT mRNA between livers of adrenalectomized rats and primary hepatocytes kept for two days in culture. Application of dexamethasone in vivo leads to a 6-fold (compare lanes 2 and 4) and of cAMP to a 13-fold induction of TAT mRNA (compare lanes 2 and 6); a combination of both inducers is even more effective (lane 8) than dexamethasone alone. Dexamethasone is similarly effective in primary hepatocytes (lanes 1 and 3), whereas the cAMP derivative is an even more potent inducer in vitro than in vivo leading to a 20 - 30-fold increase in TAT mRNA levels. TAT mRNA seems to be maximally induced by cAMP alone, for additional application of dexamethasone is not able to increase the amount of TAT mRNA. In the experiment of Fig. 1B the same filters were probed for albumin mRNA content. It is clearly evident that primary hepatocytes have at least 10-fold less albumin mRNA after two days in culture when compared with intact liver (lanes 1 and 2) confirming observations previously made in mouse hepatocyte cultures [22]. The albumin mRNA level is only marginally increased by dexamethasone in livers of intact animals (lanes 2 and 4) whereas it is raised about 3-fold in primary hepatocytes exposed for 4 h to 1 pM dexamethasone (lanes 1 and 3). The cAMP derivative is without any effect on albumin mRNA concentration in both intact liver (lanes 2 and 6) and primary hepatocytes (lanes 1 and 5) .

501

Dex f

- Dex c A M P c A M P H L H L H L H L

A

A 1 2 3 4 5 7 d a y s + - + - f - + - + - + - O e x

TAT

A L 5

B

A LB

1 2 3 4 5 6 7 8

Fig. 1 . Induction of TAT and albumin mRNA by dexamethasone, CPT- CAMP and a combination of both inducers in primary hepatocytes and in liver. Primary hepatocytes were cultured for two days. Prior to induction the serum-containing medium was replaced by serum-free medium. Dexamethasone (1 pM) and CPT-CAMP (0.05 mM) were given for 4 h and 2 h respectively. Adrenalectomized animals were treated with dexamethasone (10 pg/lOO g body weight) for 2 h. CPT- cAMP (0.2 mg/100 g body weight) was given twice at 30-min intervals and the animals killed 30 min after the last injection. The same pro- tocol was used when both inducers were combined. Total RNA was isolated and 4 pg analyzed as described in Materials and Methods. (A) Analysis of TAT mRNA. (B) Analysis of albumin mRNA. H, hepatocytes; L, liver

Next we determined for how long the inducibility of the TAT gene can be maintained in primary hepatocytes. The effect of time in culture on the albumin mRNA level was also determined. Primary hepatocytes were kept in medium containing 10% fetal calf serum for the times shown in Fig. 2. Before addition of inducers serum-containing medium was replaced by serum-free medium. The results presented in Fig. 2A show that TAT mRNA levels do not change significantly and remain fully inducible by dexamethasone during the first 5 days in culture whereas the albumin mRNA levels progressively decline during 5 days in culture. Surprisingly the albumin mRNA concentration was significantly stimulated when the cultures were incubated for 4 h in dexamethasone- containing medium, suggesting that the transcription of the albumin gene is stimulated by dexamethasone (see below). Inducibility of TAT mRNA by dexamethasone is maintained over the entire period studied, but the maximally induced level declines in parallel with the basal level at later periods. By day seven the morphology of the culture deteriorates. Parenchy- mal cells cluster in islands and fibroblasts begin to overgrow. Thus loss of parenchymal liver cells might explain the observed decrease of basal and induced levels of TAT mRNA at day seven.

In Fig. 2B a similar experiment, in which cAMP was the inducing agent, is depicted. Since insulin has been implicated in control of TAT gene expression [l], the effect of insulin on basal and CAMP-stimulated expression of the TAT gene was also examined. While TAT mRNA levels are fully inducible

1 2 3 4 days f - + - + - + - + - + - + - + - C A M P

T A T

+ In5 - 1 " s T i n 5 - I n 5 +Ins - I n5 +In5 - In5

Fig. 2. Expression of TAT mRNA and albumin mRNA in cultured hepatocytes. Hepatocytes were cultured for 1 ,2 , 3,4, 5 and I days in serum-containing medium. Prior to induction the medium was changed to serum-free conditions. Induction by dexamethasone (I pM) was for 4 h (A), by CPT-CAMP (0.5 mM) for 2 h (B). In the experiment shown in (B) the effect of insulin (1 .O pg/ml) during the inducing period was also tested. 4 pg RNA was analyzed for TAT mRNA (TAT) and albumin mRNA (ALB) content

for four days by CAMP, cAMP has no effect on the amount of albumin RNA. The presence of 1 pM insulin is without effect on either the basal or the cAMP induced levels of both mRNAs.

Time course of induction of TAT mRNA levels by dexamethasone and CPT-CAMP

To establish whether the increase of TAT mRNA levels is a direct effect of the various inducers a time-course experiment was performed, reasoning that a rapid increase in TAT mRNA concentration and changes in TAT gene transcription would argue for a direct effect of the inducers independent of the synthesis of mediating proteins. Primary hepatocytes were kept for 2 days in culture, following which the serum-containing medium was replaced by serum-free medium for induction. A time course of induction of TAT mRNA by dexamethasone and CPT-CAMP is shown in Fig. 3. In the presence of 1 pM dexamethasone the TAT mRNA level increase rapidly, an increase being evident after 1 h and reaching a plateau after 3 h (Fig. 3A). The effect of cAMP on TAT mRNA level is even more rapid (Fig. 3 B). As early as 30 min after addition of the cAMP derivative, the TAT mRNA level is significantly higher than in the uninduced control. After 2 h the TAT mRNA level is increased about 30-fold and remains at this high level for several hours. An equally rapid increase of TAT mRNA levels, compared to what we observe here in primary hepatocytes, has been observed earlier in liver [4] dem- onstrating that hepatocytes are able to react to hormonal signals with comparable kinetics in culture to those in vivo.

502

A

B

Fig. 3. Time coww of TAT rnRNA induction by dexamethasone and CAMP. Hepatocytes were cultured for two days in serum-containing medium. After changing to serum-free conditions, dexamethasone (1 pM) and CPT-CAMP (0.5 mM) was given for the times indicated. 4 pg RNA was anatyxd for TAT mRNAcontent after dexamethasone (A) and CPT-CAMP (B) stimulation

Expression oj’thc TAT gene is regulated at the transcriptional level

The observed changes in the amount of TAT mRNA and albumin mRNA in primary hepatocytes may be the conse- quence of altered stability of the mRNAs or due to a modula- tion of the rate o f transcription. To distinguish between these alternatives, the transcription rate in isolated nuclei prepared from primary hepatocytes was determined. Hepatocytes were induced with dexamethasone or CPT-CAMP for the times indicated and nuclei prepared by detergent treatment were used for in vitro RNA synthesis. To determine relative rates of transcription, RNA labelled by incorporation of [c~-~’P]UTP in vitro was hybridized to cloned single-copy DNA of several genes immobilized on nitrocellulose filters.

A representative dot blot of such a run-on transcription experiment is presented in Fig. 4. Single-copy DNA of the TAT gene, the albumin gene, the al-antitrypsin gene, the tryptophan oxygenase (TO) gene and the liv2 cDNA were utilized as probes. pUC8 DNA served as a background control. As demonstrated in Fig. 4A, dexamethasone leads to a strong increase in transcription of the TAT, TO and albumin genes within 15 min whereas transcription of the liv2 and al- antitrypsin genes is not affected. Transcription of the TAT gene is increased with similar kinetics following cAMP treatment while TO and albumin gene transcription was unaffected as expected (Fig. 4B).

For a quantitative evaluation of transcription rates the dots were cut out and the amount of radiolabelled RNA hybridized was determined by liquid scintillation counting

15 30

45 30

60 60

cAMP 120

DEX

Fig. 4. Efjiects of dexarnethasone and CPT-CAMP on the transcription of liver-speccfic genes. Primary hepatocytes cultured for two days were incubated with dexamethasone (1 FM) or CPT-CAMP (0.05 mM) for various times. After the times indicated nuclei were prepared for in vitro RNA synthesis and the RNA labelled in vitro was used for hybridization as described in Materials and Methods using single copy fragments of various liver-specific genes ;is probes

(Table 1). As the hybridization reaction included a “-labelled TAT probe in addition to the KNA labelled by [32P]UTP incorporation correction could be made for variations in hybridization efficiency. Dexamethasone increases the rate of transcription of the TAT gene 8-fold after 120 min, which very closely fits the increase in mRNA concentration. The cAMP derivative CPT-CAMP enhances TAT gene transcription 13-fold after 60 min. This is lower than the 20 - 30-fold increase of the amount of TAT mRNA, as determined by the RNA blot analysis.

Surprisingly albumin gene transcription is strongly affected by dexamethasone. As seen in Fig. 2 the albumin mRNA concentration increases to a much lower degree following dexamethasone stimulation. Owing to the long half- life and large pool of albumin mRNA, accumulation of the mRNA probably is not yet complete and the steady state not yet reached. The effect on albumin gene transcription is as rapid as that of the TAT gene indicating that the effect of glucocorticoids on albumin gene transcription is primary. CAMP has no influence on albumin gene transcription.

Antagonists of cA MP-dependent protein kinase prevent induction of TAT mRNA by cAMP

The rapidity of TAT induction by cAMP suggests that the inducer has a direct effect on TAT gene transcription. It is, however, not known by which intermediates this effect is exerted. As CAMP acts by activation of CAMP-dependent protein kinases we tested whether these enzymes are involved in TAT mRNA induction by CAMP exploiting the diastereomeric forms of adenosine cyclic 3’3’-phosphorothioates (Sp)-cAMP[S] and (Rp)-cAMP[S] which have agonist and antagonist activities, respectively, on CAMP-dependent protein kinases [ 11 - 151.

To test the suitability of (Rp)-cAMP[S] and (Sp)-cAMP[S] as tools to study the mechanism of induction of specific gene transcription by CAMP, the effect on TAT mRNA level of the cAMP analogues when given separately was tested. Primary hepatocytes were incubated with increasing concentrations of

503

Table 1. Effects of dexamethusone and CPT-CAMP on the transcriptional rate of the TAT and albumin genes 32P-labelled RNA was synthesized in nuclei prepared from primary hepatocytes which were incubated with 1 pM dexamethasone (Dex) and 50 pM CPT-CAMP for various times. The "P-labelled RNA was hybridized together with a uniformly 'H-labelled SP6 TAT cRNA to filter- bound single-copy genomic DNA fragments of the TAT and albumin genes as described in Materials and Methods. After hybridization and washing the dots were cut out and 'H and 32P radioactivity determined by liquid scintillation counting. The hybridization rate of 32P-labelled RNA was corrected according to hybridization efficiency of 3H-labelled TAT cRNA. The transcription rates have not been corrected for the size of the DNA probes

Gene Treatment ~~~~~

lo-' x Input of ["PIRNA Hybridization of Relative rate of TAT [32P]RNA hybridized ['H]SP6 TAT RNA and albumin gene

transcription

TAT control Dex 15 rnin Dex 30 rnin Dex 45 rnin Dex 60 rnin Dex 120 rnin

Dex 15 rnin Dex 30 rnin Dex 45 rnin Dex 60 rnin Dex 3 20 rnin

CTP-CAMP 15 min CTP-CAMP 30 rnin CTP-CAMP 60 rnin

Albumin control

TAT control

Albumin control CTP-CAMP 15 min CTP-CAMP 30 min CTP-CAMP 60 min

5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5

55 212 233 239 41 1 579

8 55

116 167 394 388 67

45 1 668 698

12 4 2

17

28 28 34 32 34 39 28 28 34 32 34 39 48 34 37 31

48 34 37 37

39 151 137 149 242 297

39 68

104 23 1 199 28

265 361 317

5 2.4 1 9.2

5.6

(Sp)-cAMP[S] and (Rp)-cAMP[S] and the TAT mRNA level was followed by RNA blot analysis. As shown in Fig. 5A (Sp)-cAMP[S] strongly induces the TAT mRNA. A concentration of 50 pM leads to maximal induction. In comparison, 5 pM 8-(4-~hlorophenylthio)-cAMP is sufficient to stimulate TAT mRNA production maximally in primary hepatocytes. In contrast, the Rp derivative, (Rp)-cAMP[S], did not influence TAT mRNA levels, even at a concentration of 100 pM, supporting the previous findings that this derivative has no agonist activity [12 - 151. Hybridization signals for each RNA preparation were quantified by cutting out the area of the hybridizing bands for counting. It is clearly seen (Fig. 5B) that (Rp)-cAMP[S] does not alter TAT mRNA levels within the dose range tested, whereas (Sp)-cAMP[S], even at a concentration as low as 1 pM, leads to a significant increase in TAT mRNA concentration. These two cAMP derivatives thus show the same relative behaviour on TAT gene expression as on glucose production [12, 131 making them important tools for dissecting the molecular mechanism of action of cAMP on gene expression.

Even though (Rp)-cAMP[S] binds to the R subunit of protein kinase with an affinity comparable to (Sp)-cAMP[S], it is not capable of increasing CAMP-dependent protein kinase activity above basal levels [ll, 141 because it does not lead to dissociation of the kinase into the regulatory and catalytic subunits. Simultaneous incubation of hepatocytes with (Sp)- cAMP[S] and increasing concentrations of (Rp)-cAMP[S] re- sulted in a concentration-dependent inhibition of glycogen breakdown [12, 131. Analysis of this inhibition suggests that

it is competitive in nature. We determined whether the Rp derivative would also inhbit TAT mRNA production when incubated simultaneously with the agonist derivative, a situa- tion which would strongly implicate the protein kinase as a mediator of the effect of cAMP on gene expression.

Primary hepatocytes were incubated with 10 pM (Sp)- cAMP[S] alone or with 10 pM (Sp)-cAMP[S] and increasing concentrations of (Rp)-cAMP[S]. 10 FM (Sp)-cAMP[S] leads to half-maximal induction of the TAT mRNA (Fig. 5). As seen in the autoradiogram (Fig. 6A) and in the quantitative evaluation of the hybridization signals (Fig. 6B), (Rp)- cAMP[S] inhibits the (Sp)-cAMP[S]-induced TAT mRNA accumulation. It had previously been shown that a 30:l molar ratio of (Rp)-cAMP[S] is necessary for 98% inhibition of glycogenolysis [12,13]. A similar ratio of antagonist to agonist leads to complete inhibition of TAT mRNA induction.

The availability of a cAMP antagonist allowed the exam- ination of whether or not the CAMP-dependent protein kinases are involved in glucocorticoid induction of transcription of the TAT gene. The glucocorticoid receptor can be phosphorylated in vivo and in vitro [24, 251. In one recent report it has been shown that the CAMP-dependent protein kinase can phosphorylate the glucocorticoid receptor in vitro [26]. The physiological significance of the phosphorylation of the receptor is unknown; it can, however, be envisaged that the CAMP-dependent protein kinase carries out this structural alteration of the receptor in vivo. The effect of the protein kinase antagonist on glucocorticoid induction of TAT mRNA levels was, therefore, studied in primary hepatocytes. As can

504

C - R p - C -Sp- t 2 3 4 5 6 7 8 9 1 0 1 1 A

B 2oi

0 10 25

Rpc AMPS

0 10 25 50

Fig. 5. Eflect of the CAMP-agonist. (Sp)-CAMPIS], andcAMPantag- onist, (Rp)-cAMPIS], on TAT mRNA concentration in primary hcpatocytes. Primary rat hepatocytes were incubated in serum-free medium in the absence or presence of increasing concentrations of (5’~)-cAMP[S] and (Rp)-cAMP[S] for 3 h. Total RNA was prepared and analyzed by blot hybridization. (A) The autoradiogram; lanes I and 5, RNA from cells incubated with no CAMP derivative. (C) Lanes 6- 11, RNA from cells incubated with 1, 5, 10, 25, 50 and 100 pM (Sp)-cAMP[S] (Sp); lanes 2-4, RNA from cells incubated with 10, 50 and 100 pM (Rp)-cAMP[S] (Rp). (9) A quantitative evaluation of the data by cutting out the area of the hybridizing bands and sub- sequent counting

be seen (Fig. 7) the synthetic glucocorticoid dexamethasone leads to a 10-fold induction of TAT mRNA. This increase is not prevented, when the cells are incubated with dexamethasone and (Rp)-cAMP[S]. The antiglucocorticoid, RU486 [27], however, prevents the induction of TAT mRNA by dexamethasone, but not that by CAMP. The effect of these two inhibitors thus gives additional evidence that glucocorticoids and CAMP affect transcription of the TAT gene by different mechanisms.

DISCUSSION The results presented here show that expression of the

TAT gene is regulated by dexamethasone and CAMP at the transcrihonal level. The induction of TAT mRNA after dexa-

C Sp -Sp + Rp- c ~ 1 2 3 4 5 6 7 8 9

B

U .- t P 2 1 5 5

p M Rpc AMPS

Fig. 6. Inhibition of’ TAT mRNA induction b.v the CAMP antagonist (Rpj-cAMP[S]. Primary hepatocytes were incubated in serum-free medium in the absence or presence of (Sp)-cAMP[S] alone or (Sp)- cAMP[S] plus increasing concentrations of (Rp)-cAMP[S]. RNA was analyzed as described in Fig. 1 . (A) The autoradiogram; lanes 1 and 9, RNA from cells incubated with no CAMP derivative. (C) Lane 2, RNA from cells incubated with 10 pM (Sp)-cAMP[S] (Sp); lanes 3 - 8, RNA from cells incubated with 10 pM (Sp)-cAMP[S] and 5, 10, 25, 50, I00 and 250 pM (Rp)-cAMP[S] (Sp + Rp). (B) A quantitative evaluation of the data as described in Fig. 1 . Arrow indicates the basal level of TAT mRNA

I 20 c ~ ~ ~ C A : P n

505

The increase in the transcription rate induced by the cAMP derivative is about 2-fold lower than the increase of the mRNA level. Whether this is due to shortcomings of the method or is an indication for a dual mode of action of cAMP remains open. The very low level of transcription in the uninduced state makes an exact determination of the basal level of transcription difficult. The data presented, however, allow the conclusion that the major part of the increased mRNA level is due to an increase in the rate of TAT gene transcription. This increase occurs very rapidly indicating that cAMP induction does not require de navo protein synthesis.

Expression of the albumin gene in contrast to the TAT gene, declines steadily in cultured hepatocytes and is virtually undetectable after 5 days in culture. A similar decay has been described by Darnell and coworkers [20,22], who demonstrated that in mouse primary hepatocytes transcription of liver- specific genes, including the albumin gene, declines very rapidly. In this context it is very interesting that albumin gene transcription can be partially restored by dexamethasone treatment. The induction of albumin gene transcription by dexamethasone is as rapid as the induction of TAT gene transcription, suggesting a primary effect of glucocorticoids on albumin gene expression. Why this strong induction is not detectable in FTO-2B hepatoma cells and in viva is not known but may be related to reduction of factors important for high level transcription of the albumin gene in viva. Similar observations of an influence of glucocorticoids on albumin gene expression have recently been reported [29]. It should be emphasized that the observed effect on albumin gene transcription is specific as other liver-specific genes, like the al-antitrypsin and liv2 genes, are not affected.

The observations made here give strong evidence that the effect of cAMP on gene expression is mediated by the CAMP- dependent protein kinase. The competitive inhibitor of the CAMP-dependent protein kinase, (Rp)-cAMP[S], in appro- priate concentrations completely inhibits the induction of TAT mRNA by the cAMP agonist (Sp)-cAMP[S]. CAMP-dependent protein kinases are the only known physiologically important CAMP-binding proteins in mammalian cells. Therefore it is most likely that they mediate CAMP effects. Results obtained from microinjection of the catalytic subunit of the protein kinase into hepatoma cells with red blood cell ghosts, which generated a 2 - 3-fold increase in tyrosine aminotransferase activity at a surprisingly late time (8 - 16 h) after fusion, are compatible with this interpretation [28]. It can, however, not be excluded that a different CAMP-receptor protein, displaying binding affinities for (Rp)-cAMP[S] and (Sp)-cAMP[S] similar to the protein kinase, is mediating the effect on TAT mRNA accumulation. These results do not address the question of whether the catalytic or the regulatory subunit of the protein kinase transduces the cAMP signal. By analyzing the interaction of proteins from cellular extracts and of the purified kinases with DNA sequences of the TAT gene that are required for cAMP control of expression of TAT-fusion genes after transfer into mammalian cells we hope to be able to identify the component responsible for this transcriptional activation. Preliminary DNA-binding studies (in collaboration with U. Walter and S. Lohmann) using the purified CAMP-dependent kinases or these subunits, respectively, have thus far failed to give evidence for interaction of these proteins with TAT control sequences.

The effect of the glucocorticoid and CAMP antagonists on dexamethasone and cAMP induction, respectively, further demonstrate that these two inducers act by different mechanisms. Induction by cAMP is observed in the presence

of the antiglucocorticoid RU486 excluding the involvement of the glucocorticoid receptor in cAMP induction. Conversely the inhibitor of CAMP-dependent protein kinase, (Rp)- cAMP[S], does not interfere with glucocorticoid induction, which indicates that a phosphorylation of the glucocorticoid receptor by the CAMP-dependent protein kinase is not required for dexamethasone induction of TAT mRNA. Recent experiments in our laboratory, aimed at identification of the DNA sequences required for glucocorticoid and cAMP induction of TAT mRNA, strongly support the results and conclusions obtained with the inhibitors [30].

Our results show that cultures of primary hepatocytes are a valuable tool for exploring the mode of action of factors affecting the expression of liver-specific genes. These cultures will be of great value for future studies on the expression of the TAT gene during development [5] and in mice carrying the albino lethal mutations [7].

We thank Dr K. Krauter for the plasmid pliv-2 and the plasmid containing ccl-antitrypsin sequences. We thank R. Miksicek and F. Steward for critically reading of the manuscript and M. Cole for excellent secretarial work. We thank Dr U. Schubart for helpful suggestions during his visit supported by a NATO travel grant. We also thank Dr Banasch for support and K. Miinter for help in the isolation of hepatocytes. This work was supported by the Deutsche Forschungsgemeinschaft (Schu 51 14-2) and the Fonds der Chemischen Industrie.

REFERENCES 1 . Granner, D. K. & Hargrove, J. L. (1983) Mol. Cell. Biochem.

53/54, 113-128. 2. Hargrove, J. L. & Mackin, R. B. (1984) J . Biol. Chem. 259,386-

393. 3. Scherer, G., Schmid, W., Strange, C. M., Rowekamp, W. &

Schiitz, G. (1982) Proc. Natl Acad. Sci. USA 79,7205 - 7208. 4. Hashimoto, S., Schmid, W. & Schiitz, G. (1984) Proc. Nut1 Acad.

Sci. USA 81,6637-6641. 5. Greengard, 0. (1970) in Biochemical actions of hormones, vol. 1

(Litwack, G., ed.) pp. 53 - 87, Academic Press, New York. 6. Miiller, G., Scherer, G., Zentgraf, H.-W., Ruppert, S., Hermann,

B., Lehrach, H. & Schiitz, G. (1985) J . Mol. Biol. 184,367-373. 7. Gluecksohn-Waelsch, S. (1979) Cell 16, 225-237. 8. Schmid, W., Miiller, G., Schiitz, G. & Gluecksohn-Waelsch, S.

9. Killary, A. M. & Fournier, R. E. K. (1984) Cell 38, 523 - 534. (1985) Proc. Natl Acad. Sci. USA 82, 2866-2869.

10. Seglen, P. 0. (1976) Methods Cell Biol. 13, 29. 11. de Wit, R. J. W., Hoppe, J., Stec, W. J., Baraniak, J. & Jastorff,

B. (1982) Eur. J . Biochem. 122,95-99. 12. Rothermel, J. D., Stec, W. J , Baraniak, J., Jastorff, B. & Bothelo,

L. H. P. (1983) J . Biol. Chem. 258, 12125-12128. 13. Rothermel, J. D., Jastorff, B. & Bothelo, L. H. P. (1984) J. Biol.

Chem. 259, 8151-8155. 14. de Wit, R. J. W., Hekstra, D., Jastorff, B., Stec, W. J., Baraniak,

J., Van Driel, R. &Van Haastert, P. J. M. (1984) Eur. J. Biochem.

15. Van Haastert, P. J. M., Van Driel, P., Jastorff, B., Baraniak, J., Stec, W. J. &de Wit, R. J. W. (1984) J . Biol. Chem. 259,10020- 10024.

16. Faulstich, H., Trischmann, H. & Mayer, D. (1983) Exp. CellRes.

17. Becker, P., Renkawitz, R. & Schiitz, G. (1984) EMBO J. 3,2015-

18. Danesch, U., Hashimoto, S., Renkawit7, R. & Schiitz, G. (1983)

19. Sargent, T. D., Jagodzinski, L. L., Yang, M. & Bonner, J. (1981)

20. Derman, E., Krauter, K., Walling, L., Weinberger, C., Ray, Y. &

21. Baraniak, J., Kinar, R. W., Lesiak, K. & Stec, W. J. (1979)

142, 255-260.

144, 73 - 82.

2020.

J. Biol. Chem. 258,4750-4753.

Mol. Cell. Biol. I, 871 -883.

Darnell, J. E. Jr (1981) Cell 23, 731 -739.

J . Chem. SOC. Chem. Commun., 940-943.

506

22. Clayton, D. F.. 6t Darnell, J. E. (1983) Mol. Cell. Biol. 3, 1552- 1561.

23. Nakamura, T.. Yoshimoto, K., Nakamyama, Y., Tomita, Y. & Ichihara, A. (1983) Proc. Nut1 Acad. Sci. U S A 80, 7229-7233.

24. Hously, J. 1:. & Pratt, W. B. (1983) J. Biol. Chem. 258, 4630- 4635.

25. Miller-Diener, A. M., Schmidt, T. J. & Litwack, G. (1985) Proc. Nut1 Acad. Sci. US.4 82,4003 -4007.

26. Singh, V. N. & Moudgil, V. K. (1985) J . Biol. Chem. 260,3864- 3690.

27. Philipert, D., Deraedt, R., Tournemine, C., Mary, D. & Teutsch, G. (1982) Ahstr. 6th Int. C'ongr. on Hormonal Steroids, Jerusalem, p. 204, J. Steroid Biochem.

28. Boney, C., Fink, D., Schlichter, D., Carr, K. & Wicks, W. D. (1983) J. Biol. Chem. 258,4911 -4918.

29. Moshage, H. J., de Haard, H. J. W., Princen, H. M. G. & Yap, S. H. (1985) Biochim. Biophys. Acta 824,27-33.

30. Jantzen, H. M., Strahle, U., Gloss, B., Stewart, F., Schmid, W., Boshart, M., Miksicek, R. & Schiitz, G. (1987) Cell, in the press.

Transcription activation of the tyrosine aminotransferase gene by glucocorticoids and cAMP in...

Documents

Transcript of Transcription activation of the tyrosine aminotransferase gene by glucocorticoids and cAMP in...