Combined Effects of Insulin and Dexamethasone on Cyclic AMP Phosphodiesterase 3 and Glycogen Metabolism in Cultured Rat Hepatocytes

Cell. Signal. Vol. 10, No. 9, pp. 629–635, 1998 Copyright  1998 Elsevier Science Inc.

ISSN 0898-6568/98 $19.00 PII S0898-6568(98)00003-5

Combined Effects of Insulin and Dexamethasone on Cyclic AMP Phosphodiesterase 3 and Glycogen Metabolism in Cultured Rat Hepatocytes Thomas Hermsdorf and Dietrich Dettmer* Department of Biochemistry, Medical Faculty, University of Leipzig, Liebigstrasse 16, D-04103 Leipzig, Germany

ABSTRACT. Primary cultures of rat hepatocytes were used to study the combined effects of insulin and dexamethasone on cyclic AMP phosphodiesterase 3 (PDE 3) and glycogen metabolism. PDE activity was measured in extracts obtained by hypotonic shock treatment of the particulate fraction from cultured hepatocytes. PDE 3 was identified by inhibition with ICI 118233, Western blotting, immunoprecipitation of the activity with the use of a new PDE 3B-specific anti-peptide antibody and stimulation of the activity after adding insulin, glucagon and okadaic acid to the culture medium. Specific PDE inhibitors were always used to identify the measured PDE activities. Hypotonic extracts contained 30% PDE 3 and 50% PDE 4. Both PDE types show a nearly constant level during cultivation up to 48 h. Long-term exposure of dexamethasone alone has no effect on PDE 3 activity, whereas, in combination with insulin, the insulin stimulation of PDE 3 activity was found to be increased between 48 and 72 h of cultivation. Additionally, db-cAMP was able to stimulate PDE 3. A possible effect of insulin or db-cAMP on PDE 3B expression could not be found. On the other hand, activation of PDE 3B after 48 h of culturing decreased rapidly after removal of insulin or db-cAMP from the culture medium. Insulin-stimulated incorporation of 14C-glucose into glycogen was inhibited by PDE 3- and PDE 4-specific inhibitors as well as by the unspecific PDE inhibitor IMBX. Inhibitions by PDE 3- and PDE 4-specific inhibitors were found to be additive and reached the same extent as with IMBX. Summarising our results, we can conclude that PDE 3 and PDE 4 effectively control the hepatic glycogen metabolism. Insulin effects on PDE activity and glycogen metabolism require the presence of dexamethasone. Insulin-stimulated PDE seems to play an important role in realising insulin effects on hepatic glycogen metabolism. cell signal 10;9:629–635, 1998.  1998 Elsevier Science Inc. KEY WORDS. Cyclic AMP phosphodiesterase type 3, Hepatocytes, Primary culture, Insulin, Dexamethasone, Glycogen metabolism

INTRODUCTION The intracellular second-messenger cyclic adenosine-3⬘,5⬘monophosphate (cAMP) acts as mediator of multiple hormonal signals. cAMP is synthesized by hormone-activated adenylate cyclases and degraded by cyclic nucleotide phosphodiesterases (PDEs). PDEs effectively control the intracellular cAMP level and consequently play an important role in the regulation of cell metabolism. Multiple PDE types can be distinguished, which differ with respect to amino acid sequences, affinity for cyclic nucleotides, sensitivity to * Author to whom all correspondence should be addressed. E-mail: detdi @server3.medizin.uni-leipzig.de Abbreviations: BSA–bovine serum albumin; db-cAMP–dibutyryl-cAMP; PDE–3⬘,5⬘-cyclic nucleotide phosphodiesterase (EC 3.1.4.17); HEPES–N-[2hydroxyethyl]peperazine-N⬘-[2-ethanesulfonic acid]; HPLC–high-performance liquid chromatography; IMBX–3-isobutyl-1-methylxanthine; PBS– phosphate buffered saline; PMSF–phenylmethylsulfonyl-fluoride; SDS–sodium dodecyl sulfate; Tris–2-amino-2-(hydroxymethyl)-1,3-propanediol. Received 15 October 1997; and accepted 15 December 1997.

activators and inhibitors and subcellular location and expression in various cells, tissues and organs [1–3]. At present, PDEs are classified in seven types (1–7), each with different subtypes [3, 4]. Some of these types are stimulated by hormones. Insulin stimulation has been described for the PDE 3 in adipose tissue and liver [1–7]. Furthermore, insulin in rat liver also has been reported to stimulate a PDE 4 [8], a PDE 2 [9] and a PDE 1 [10]. Studies on the PDE 3 in rat adipose tissue and human platelets revealed that phosphorylation of special seryl residues has a role in insulin stimulation [11–15]. On the other hand, effects on PDE have been described that obviously affect the amount of the enzyme. Thus follicle-stimulating hormone (FSH) induced PDE 4 synthesis in Sertoli cells [16], and, in a monocyte cell line, PDE 4 was found to be upregulated after prolonged exposure to agents that increase cAMP content [17]. Moreover, glucocorticoids seem to affect PDE activity. So dexamethasone regulates the hormonal responsiveness of PDE 3 in 3T3-L1

630

adipocytes [18] and, in crude cell lysates of hepatocyte cultures, the addition of dexamethasone lowered low Km PDE activity [19]. The aim of the studies presented here was to introduce primary culture of hepatocytes to study the combined effects of insulin and dexamethasone on PDE 3 and glycogen metabolism at optimised conditions. In comparison with freshly isolated and suspended hepatocytes in primary culture, all experimental conditions can be varied exactly. With the use of hepatocyte cultures, it is possible to study the long-term effects of hormones such as insulin and dexamethasone on PDE 3 activity, as well as the short-term effects of insulin in combination with the long-term effects of dexamethasone. Washing procedures and hormone-free precultivation excluded interference of other influences during the measuring of short-term effects. Our results indicate that PDE 3B—identified by specific inhibitors, Western blotting, immunoprecipitation and stimulation by insulin, glucagon and okadaic acid—plays an important role in the regulation of liver glycogen metabolism. Insulin stimulation of PDE 3B seems to be part of the short-term regulation of glycogen metabolism in rat liver. The effects of insulin on PDE 3B expression could be excluded. It could be demonstrated for the first time that PDE 3- and PDE 4-specific inhibitors show additive effects on liver glycogen synthesis, which is inhibited completely.

MATERIALS AND METHODS Chemicals [3H]cAMP (24 Ci/mmol) and [␥-32P]ATP (3000 Ci/mmol) were from Amersham-Buchler (Braunschweig, Germany). M 199 medium, insulin and cycloheximide were obtained from Serva (Heidelberg, Germany). ICI 118233 and ICI 63197 were generously supplied by D. E. Riley and J. Bebbington (ICI Pharmaceuticals, Macclesfield, Cheshire, UK) and rolipram by Dr. Wachtel (Schering AG, Berlin, Germany). Alkaline phosphatase from calf intestine (3000 U/mg) was supplied by Boehringer (Mannheim, Germany) and Q Sepharose from Pharmacia (Uppsala, Sweden).

Isolation and Primary Culture of Hepatocytes Male Wistar rats, 200–300 g, were kept on a 12-h day/night rhythm and allowed free access to standard diet. Hepatocytes were isolated by recirculating collagenase perfusion (50 mg collagenase H/50 mL Krebs-Ringer bicarbonate buffer containing 15 mM HEPES and 4 mM CaCl2) for 10 min. Primary culture was started by seeding 1.5–2.0 ⫻ 106 viable cells suspended in culture medium M 199 with 25 mM HEPES into 60-mm plastic tissue culture dishes precoated with rat tail collagen. After plating for 4 h at 37⬚C in M 199 containing HEPES and 5% foetal calf serum at an atmosphere of 95% air, 5% CO2, medium was changed and the foetal calf serum was substituted by 0.2% bovine serum albumin. Further medium changes took place after 24 h and 48 h.

T. Hermsdorf and D. Dettmer

Preparation of Hypotonic Extracts from Hepatocytes Cells (two or three dishes) were washed three times with ice-cold homogenisation buffer (10 mM Tris-HCl, 1 mM EDTA, 250 mM sucrose, pH 7.4), scraped off with a rubber policeman and homogenised in about 0.5 mL with 20–30 strokes and 120 r.p.m. by using a conic glass homogeniser (Glas-Col, Terre Haute, IN, USA). The homogenate was transferred to 2-mL Eppendorf cups and adjusted to 2 mL with homogenisation buffer followed by centrifugation at 48,000 ⫻ g for 10 min at 4⬚C. The supernatant was discarded and the pellet was resuspended in 400 ␮L hypotonic extraction buffer (10 mM Tris-HCl, 1 mM EDTA, pH 7.4) by using a 1-mL glass syringe. After extraction for 20 min at 4⬚C, samples were centrifuged at 48,000 ⫻ g for 10 min at 4⬚C. The supernatant, called hypotonic extract, was immediately used for the PDE activity assay. The PDE 3-specific inhibitor ICI 118233 [20] and the PDE 4-specific inhibitors ICI 63197 and rolipram [21, 22] were always used for type identification. Separation of PDE Forms on Q Sepharose PDE isoform separation was carried out on a Q Sepharose column (1 ⫻ 5 cm) equilibrated with 10 mM Tris-HCl (pH 7.4), 5 mM MgCl2, 0.1 mM PMSF, 2 mM benzamidine and 2 mM mercaptoethanol (buffer B). Extracts were applied to the column, which was washed with 20 mL buffer B. Bound PDE activity was eluted with a 100-mL linear sodium chloride gradient (0–600 mM) in buffer B (flow rate of 1 mL/ min). Fractions of 2 mL were collected, and each second fraction was tested for PDE activity. PDE 3 activity was assayed at substrate concentrations of 0.1 ␮M cAMP. The PDE activity of the fractions was further analysed in the presence of 40 ␮M ICI 63197 or 100 ␮M ICI 118233. PDE 3 Antibodies In rabbits, a PDE 3B-specific antiserum was raised to ovalbumin-conjugated peptide NH2-DNLNPKPQRRKGRRRCCOOH by FZB Biotechnik GmbH (Berlin, Germany). The peptide sequence corresponds to amino acid residues 1033 to 1047 at the carboxy terminal of the rat adipocyte PDE 3B [23]. Cysteine was added for conjugation of the peptide to ovalbumin. Sequence analysis using the program blastp of the HUSAR program package (German Cancer Research Center, Heidelberg) revealed that the chosen sequence is highly specific for PDE 3B. The polyclonal antibody was purified by affinity chromatography on a Sepharose 4B column with covalently coupled peptide. Western Blot Analysis Hypotonic extracts from freshly isolated hepatocytes were separated on Q Sepharose as heretofore described. The fractions were precipitated by 5% trichloracetic acid. The protein pellets were resolved in sample buffer (62.5 mM TrisHCl, pH 6.8, 10% glycerol, 5% mercaptoethanol, 2% SDS,

Combined Hormonal Effects on Cyclic AMP Phosphodiesterase 3

0.0025% bromphenol blue). The samples were boiled for 5 min, and proteins (20 ␮g) were separated on a 10% SDSpolyacrylamide gel. The proteins were transferred to an Immobilon nylon membrane by using the trans-blot semi-dry transfer cell (Bio-Rad). Non-specific binding was blocked by incubating the membrane for 30 min in 5% non-fat dry milk and 0.5% BSA dissolved in PBS/0.05% Tween-20. Incubation with the anti-PDE 3B antibody (1:1000 v/v) in PBS/Tween-20 was carried out for 1 h, and, after washing, the membrane was incubated for 1 h with anti-rabbit IgGlinked peroxidase (Boehringer, Mannheim) diluted 1:1000 in PBS/Tween-20. After the washing steps with PBS/Tween, bound antibodies were detected by using 3,3⬘-diaminobenzidine tetrahydrochloride.

631

sample) at 30⬚C for 10 to 20 min. The reaction was terminated by boiling the samples for 3 min. After cooling in an ice bath, samples were incubated for 10 min at 30⬚C with 0.3 U (about 0.1 ␮g) alkaline phosphatase followed by heat inactivation for 3 min in boiling water. To separate [3H]adenosine from [3H]cAMP, 1 mL of a slurry (DOWEX 1 ⫻ 2; 200–400 mesh; Cl⫺ form; resin:water 1:4 v/v) was added to the reaction mixture and mixed for 1 min. After centrifugation at 12,000 r.p.m. for 1 min, 100 ␮L aliquots of the supernatants were transferred to plastic minivials for liquid scintillation counting. If the conversion of cAMP into adenosine did not exceed more than 30%, enzyme activity was proportional to substrate degradation. Recovery of the [3H]-labelled products from DOWEX was estimated by using [3H]adenosine in controls.

Immunoprecipitation Diluted aliquots (180 ␮L) from hypotonic extract and fractions from peak 1 or 2 containing 0.1 mM PMSF, 2 mM benzamidine and 20 ␮M leupeptin were mixed either with 20 ␮L pre-serum or anti-serum. After an overnight incubation at 4⬚C, 6 mg protein A Sepharose was added to the samples. The protein A Sepharose was pelleted after 1 h by centrifugation at 12,000 ⫻ g at 4⬚C for 1 min, and the enzyme activity of the supernatants was determined. The pellets were washed three times for 1 h at 4⬚C in 1 mL buffer B with 0.1 M NaCl, 0.1 mM PMSF and 2 mM benzamidine, and PDE activity was estimated in the pellets. Glycogen Synthesis in Hepatocytes After 48-h culturing in the presence of 10 nM insulin and 0.1 ␮M dexamethasone, dishes were washed two times and then incubated in M 199 containing 0.1 ␮M dexamethasone for 1 h. Incubation was continued with a medium containing dexamethasone, 2 mM lactate and [14C]glucose (0.8 ␮Ci/dish). PDE inhibitors (40 ␮M ICI 63197, 10 ␮M rolipram, 100 ␮M ICI 118233 and 1000 ␮M IBMX) were added after 20 min, and incubation was continued for 10 min. Then zero-time samples were taken, and the experiment was started by the addition of 10 nM insulin. PBS addition served as control. The incubation was terminated after 2 h by washing cultures three times with ice-cold PBS and by the addition of 400 ␮L ice-cold 0.3 M KOH. The dishes were stored on ice, and the cells were harvested with a rubber policeman. The rate of glycogen synthesis was determined by isolation and quantification of the [14C]-labelled glycogen from one dish, as described by Fleig et al. [24], and calculated with respect to the medium glucose content. Cyclic AMP Phosphodiesterase Assay PDE activity was measured by using the two-step assay according to Thompson and Appleman [25]. Incubation was performed in a total volume of 200 ␮L 100 mM Tris-HCl (pH 7.4), 5 mM MgCl2, 3.75 mM mercaptoethanol with 0.1 ␮M or 1 ␮M cAMP and [3H]cAMP (about 500,000 c.p.m./

Protein Determination Protein content was determined by the method of Bradford [26], with the use of bovine serum albumin as standard. RESULTS To study the behaviour of PDE in cultured hepatocytes after short- and long-term application of insulin and dexamethasone alone and in combination, crude membrane preparations were extracted with hypotonic buffer solution. As shown in Figure 1, hypotonic extraction is not specific for PDE 3. After chromatographic separation of hypotonic extracts on Q Sepharose using a linear sodium chloride gradient, PDE 3 inhibited by ICI 118233 in fractions 30 to 40

FIGURE 1. Q Sepharose chromatography of hypotonic extract

from freshly isolated hepatocytes. Hepatocytes (about 4 ⫻ 107 cells) were homogenised in 10 mL homogenisation buffer, adjusted to 30 mL and centrifuged as described in the Materials and Methods section. Particulate fraction was extracted with 10 ml hypotonic buffer after washing two times with 10 mL homogenisation buffer. The extract containing a total activity of 0.3 nmol cAMP/ min was applied to the column. After washing, protein was eluted with a linear sodium chloride gradient. PDE activity was measured at a substrate concentration of 0.1 ␮M cAMP. Open circles, protein; closed circles, non-inhibited PDE activity; closed triangles, ICI 63197-inhibited PDE activity; closed squares, ICI 118233-inhibited PDE activity.

632

T. Hermsdorf and D. Dettmer

FIGURE 2. Western blot analysis of fractions 26–42 of hypo-

tonic extract from freshly isolated hepatocytes after Q Sepharose chromatography.

and PDE 4 inhibited by ICI 63197 in fractions 22 to 32 could be detected. Moreover, small quantities of PDE 2 activated by cGMP could be found in fractions 26 to 28 (data not shown). Other PDE types were not tested. The PDE 3 was further identified by Western blotting (Fig. 2) and immunoprecipitation of the activity with the use of a newly developed PDE 3B-specific anti-peptide antibody (Table 1). A molecular weight of about 57,000 was determined for the specific protein bands in Western blots. The authenticity of the antibody used was further proved by the property of the peptide to successfully compete with the immmunreactive species (data not shown). The levels of PDE 3 (about 30% of total activity) and PDE 4 (about 50% of total activity) specified by the specific inhibitors ICI 63197 and ICI 118233 remained nearly constant during cultivation of hepatocytes up to 48 h (data not shown). A combination of both inhibitors showed additive effects. About 20% of the activity was found to be unaffected and seems to represent other PDE forms. For further characterisation of the PDE 3 in connnection with the PDE 4 short-term effects of insulin, glucagon and okadaic acid on the activity were investigated in hepatocytes cultured for 48 h. As shown in Table 2, both activities were increased, but only the PDE 3 was stimulated significantly. Consequently this PDE could be subtyped as PDE 3B. The PDE 3B activity (calculated as difference in absence and presence of ICI 118233) was found to be stimulated during cultivation up to 72 h in the presence of insulin and in the presence of insulin plus dexamethasone (data not shown). Dexamethasone alone had no effect on PDE 3B activity within the first 48 h, whereas, after 72 h, a small de-

crease could be observed in comparison with untreated cells. But dexamethasone seems to enhance the insulin stimulation at a cultivation time of 48 h. To study this phenomenon, after 24-h precultivation in the presence of dexamethasone, insulin or db-cAMP was added for an additional 24 h, and the PDE 3B activity was measured in the presence and absence of cycloheximide (Table 3). Cycloheximide significantly inhibited insulin- and db-cAMP-stimulated PDE activity after 48 h of incubation. But no effect on PDE expression could be found in Western blot by using our PDE 3B-specific anti-peptide antibody (data not shown). On the other hand, removal of insulin or db-cAMP from the culture medium resulted in a rapid loss of stimulation (Fig. 3). To evaluate the role of PDE 3B and PDE 4 in connection with the regulation of cell metabolism, the influence of the PDE inhibitors ICI 63197, ICI 118233, rolipram and IMBX on insulin-stimulated 14C-glucose incorporation into liver glycogen was studied (Table 4). Stimulating effects of insulin on glycogen synthesis in hepatocytes could be observed only when the cells were precultured in the presence of dexamethasone plus insulin (data not shown). Therefore the hepatocytes were cultured in the presence of 10 nM insulin and 0.1 ␮M dexamethasone. After being washed free of insulin and undergoing a 1.5-h insulin-free incubation period, the cultured hepatocytes were exposed to insulin for 2 h. Insulin increased glycogen synthesis as much as 9-fold. The presence of ICI 63197, ICI 118233 and rolipram led to a partial inhibition of glycogen synthesis. Additive effects were observed after the simultaneous addition of ICI 63197 plus ICI 118233 or of rolipram plus ICI 118233, resulting in a nearly complete inhibition of glycogen synthesis. Additionally, the unspecific PDE inhibitor IMBX totally inhibited 14Cglucose incorporation. Further combination of ICI 63197 plus ICI 118233 or rolipram plus ICI 118233 also showed inhibitory effects on the basal 14C-glucose incorporation. DISCUSSION Primary culture of hepatocytes was used to study combined short-term and long-term effects of dexamethasone and insulin on PDE 3 and glycogen metabolism. PDE activity was measured in hypotonic extracts of membranes. Hypotonic extraction was introduced by Loten et al. [5]. Using this procedure, Pyne et al. [27] purified and characterised the insulin- and glucagon-stimulated “dense vesicle” PDE from rat liver. We also preferred to use this procedure. Chromato-

TABLE 1. Immunoprecipitation of partly purified phosphodiesterase

Hypotonic extract

Total activity ICI 118233-inhibited activity

Peak 1

Peak 2

Pre-serum

Anti-serum

Pre-serum

Anti-serum

Pre-serum

Anti-serum

0.357 0.302

1.156 1.007

0.033 0.012

0.051 0.023

0.020 0.012

0.500 0.430

Hypotonic extracts were prepared from freshly isolated hepatocytes and separated on Q Sepharose by gradient elution. Aliquots from hypotonic extracts and pooled fractions of peak 1 and 2 were used for immunoprecipations with PDE 3B-specific anti-peptide–anti-serum and pre-serum as control (see Materials and Methods). Results are shown as the PDE activity (pmol/min) precipitated by protein A Sepharose.

Combined Hormonal Effects on Cyclic AMP Phosphodiesterase 3

633

TABLE 2. Stimulation of PDE activity by insulin, glucagon and okadaic acid

ICI 63197-inhibited activity ICI 118233-inhibited activity

Control

Insulin

Glucagon

Okadaic acid

4.16 ⫾ 0.52 2.82 ⫾ 0.42

4.99 ⫾ 0.44 5.83 ⫾ 0.34*

4.54 ⫾ 0.19 6.97 ⫾ 0.76*

5.07 ⫾ 0.54 7.59 ⫾ 0.92*

Hepatocytes were cultured for 48 h without hormones. Then 10 nM insulin, 10 nM glucagon and 1 ␮M okadaic acid were added for 10 min. The culture dishes were washed three times with homogenisation buffer, and hypotonic extracts were prepared. PDE activity (pmol/min/mg protein) was assayed at 0.1 ␮M [3H]cAMP in the presence of 40 ␮M ICI 63197 or 100 ␮M 118233. Values represent the mean ⫾ S.D. from four experiments. * P ⬎ 0.05 (Student’s t-test).

graphic separation of the hypotonic extract on Q Sepharose shows that, by using specific inhibitors beside PDE 3, PDE 4 activity also could be found in hypotonic extracts of crude membrane preparations (Fig. 1). Western blot analysis (Fig. 2) and immmunoprecipation with a PDE 3B-specific antipeptide antibody (Table 1) confirm the conclusion that peak 2 contains the PDE 3 activity. The determined molecular weight of 57,000 is in agreement with Pyne et al. [27], who have reported a size of 62,000 Mr, for the intact enzyme and a size of 57,000 Mr for the clipped fragment, which had lost its membrane-anchoring domain. On the other hand, sizes of 73,000 Mr by SDS gel electrophoresis and of 130,000 Mr by HPLC size-exclusion chromatography were determined for the PDE 3 solubilized by mild proteolysis with chymotrypsin [28]. We think that the PDE 3B in hypotonic extracts is an N-terminally truncated form. The C-terminus seems to be complete because the peptide sequence recognised by our anti-peptide antibody is localised closely at the conservative sequence, which most likely contains the active centre of the enzyme. In addition, bands larger than 57,000 Mr could be detected in Western blots by our antibody in earlier preparation steps. Further experiments are neccessary to evaluate this finding. PDE activity was always determined in the absence and presence of the specific inhibitors ICI 118233 and ICI 63197 to precisely differentiate between PDE 3 and PDE 4. A constant relation of 30% PDE 3 and 50% PDE 4 could be found during cultivation of hepatocytes up to 48 h. Kilgour et al. [29] reported that, in hypotonic extracts, about 85% of the PDE activity was

inhibited by the PDE 3-specific inhibitor ICI 118233. Pyne et al. [20] found, in homogenates, a relation between PDE 3 and PDE 4 of 1:1 and 24% of ICI 118233-inhibited PDE activity. These differences may be caused by differences in sample preparation. Important influences seem to have been the homogenisation procedure and time and temperature differences during and after hypotonic treatment. With our standardised experimental conditions, we obtained reproducible results. As reported, PDE 3 could be found to be stimulated by insulin, isoproterenol, ACTH and TSH in adipocytes [18] and by insulin and glucagon in hepatocytes [5, 6, 8]. Okadaic acid, which can mimic the action of the hormones, also increases the PDE 3 activity in adipocytes [30]. Moreover also the plasma membrane cAMP–PDE of type 4 also was found to be stimulated by insulin [7, 8]. We were able to demonstrate a significant PDE 3 stimulation by insulin, glucagon and okadaic acid in agreement with [5], [6], [8] and

TABLE 3. Effects of cycloheximide on ICI 118233-

inhibited PDE activity*

Control Insulin Dibutyryl-cAMP

⫺CHX

⫹CHX

6.6 ⫾ 0.5 14.9 ⫾ 1.9† 13.9 ⫾ 2.6†

6.1 ⫾ 0.4 8.2 ⫾ 0.4† 8.9 ⫾ 0.9†

*During long-term incubation with dexamethasone ⫹ insulin and dexamethasone ⫹ dibutyryl-cAMP after 24-h preincubation of hepatocytes with dexamethasone. Hepatocytes were cultured with 0.1 ␮M dexamethasone for 24 h. Then medium was replaced and 10 nM insulin or 500 ␮M db-cAMP was added. ICI 118233-inhibited PDE activity was determined at 48 h of cultivation in hypotonic extracts after addition of 50 ␮M cycloheximide for 30 min. Data represent the means ⫾ S.D. of PDE activity (pmol/min/mg protein) from four experiments. † Significantly lower in comparison with the control values (P ⬍ 0.05, Student’s t-test).

FIGURE 3. Time-dependent loss of ICI 118233-inhibited PDE

activity after removal of insulin and db-cAMP. Cells were cultured with 0.1 ␮M dexamethasone alone or additionally with 10 nM insuln or 500 ␮M db-cAMP for 24 h. Then culture dishes were washed three times with medium containing only dexamethasone followed by incubation for 30 min. At indicated times, dishes were taken for the preparation of hypotonic extracts for the PDE assay. Results are shown as a percentage of PDE activity after preincubation with insulin (closed triangles) or db-cAMP (open triangles) corresponding to controls (100%). Values are means of two experiments.

634

T. Hermsdorf and D. Dettmer

TABLE 4. Effects of PDE inhibitors on glycogen metabolism

Without No additions ICI 63197 ICI 118233 Rolipram IBMX ICI 63197 and ICI 118233 Rolipram and ICI 118233

6.0 0.7 0.8 5.1 ⫺4.5

⫾ ⫾ ⫾ ⫾ ⫾

2.1 2.8 1.2 2.1 5.0

10 nM Insulin 54.2 32.2 26.8 38.1 0.1

⫾ ⫾ ⫾ ⫾ ⫾

4.9 9.0 11.9 11.2 1.7

⫺3.9 ⫾ 0.7

5.4 ⫾ 1.9

⫺3.3 ⫾ 0.2

4.9 ⫾ 0.5

Hepatocytes were cultured for 48 h with 10 nM insulin and 0.1 ␮M dexamethasone. After being washed free of insulin, the cultures were incubated for 30 min in M 199 medium containing 0.1 ␮M dexamethasone, 2 mM lactate and [14C]-glucose. Ten minutes prior to the addition of PBS as control or 10 nM insulin, PDE inhibitors ICI 63197 (40 ␮M), ICI 118233 (100 ␮M) and IBMX (1 mM) were added. Glycogen turnover was assessed by determination of nmol 14C-glucose incorporated into glycogen after 2 h. Negative values express an attenuation of glycogen labelling compared with time-zero value. Data are shown as means ⫾ S.D. of three experiments.

[30]. This finding indicates that the investigated enzyme is PDE 3B. Regarding the non-significant effects on PDE 4 in connection with the results reported in [7] and [8], we should like to mention that we extracted the so-called dense vesicle PDE with hypotonic buffer, whereas hypertonic extraction was used to purify PDE 4 of plasma membranes [31]. PDE 4 was found in cytosolic and membrane fractions [32, 33]. It is also possible that the insulin effect on PDE 4 may be different, depending on the intracellular localisation of the enzyme, as reported for distinct rolipram inhibition kinetics of cytosolic and particulate PDE 4 expressed in transfected COS7 cells [34]. Moreover, the various cellular fractions may contain different PDE 4 isoforms. Changes in hormone responsiveness and cAMP metabolism in primary cultures of rat hepatocytes were reported by Christoffersen et al. [19]. In 3T3-L1 adipocytes, dexamethasone regulates the responsiveness of a high-affinity hormone-sensitive PDE towards insulin, epinephrine and corticotropine [18]. Permissive as well as “anti-insulin” effects were discussed. Our results show anti-insulin effects at the beginning and a permissive influence of dexamethasone after 48 h of cultivation. Induction of enzyme synthesis by insulin and the cAMP-generating hormone glucagon was described for ornithine decarboxylase [35], and Swinnen et al. [16] were able to detect hormonal induction of PDE 4 by FSH. A possible hormonal induction of PDE 3B could not be detected. On the other hand, a rapid decrease in activity after removal of insulin and db-cAMP (Fig. 3) points to the importance of metabolic short-term regulations, most probably through phosphorylation reactions. Glycogen metabolism was influenced effectively by blocking the PDE 3 activity with specific inhibitors (Table 4). But PDE 4-specific inhibitors also influence glycogen synthesis to the same extent. Moreover, PDE 3- and PDE 4-specific inihibitors show additive effects. The maximal inhibition observed is nearly the same as with the non-spe-

cific PDE inhibitor IMBX alone. This suggests an important role of both PDE types in the regulation of hepatic glycogen metabolism. Although the insulin effect on the PDE 4 in our investigation was not significant (Table 2), the results regarding glycogen synthesis support the idea of a hormoneregulated PDE 4. Summarising our results, we can conclude that dexamethasone and insulin show combined effects on PDE 3 and glycogen synthesis. Insulin stimulation of PDE 3 seems to be a part of metabolic short-term regulation and is most probably realised by phosphorylations. PDE 3 and PDE 4 effectively control hepatic glycogen metabolism. The results suggest that PDEs play an important role in realising insulin signals in rat liver. This study was supported by a grant from Deutsche Forschungsgemeinschaft, Bonn, Germany.

References 1. Houslay M. D. (1991) Eur. J. Biochem. 195, 9–27. 2. Manganiello V. C., Murata T., Taira M., Belfrage B. and Degerman E. (1995) Arch. Biochem. Biophys. 322, 1–13. 3. Beavo J. A. (1995) Physiol. Rev. 75, 725–748. 4. Beavo J. A. and Reifsnyder D. H. (1990) Trends Pharmacol. Sci. 11, 150–155. 5. Loten E. G., Assimapopoulos-Jeanett F. D., Exton J. H. and Park C. R. (1978) J. Biol.Chem. 253, 746–757. 6. Loten E. G., Francis S. H. and Corbin J. D. (1980) J. Biol. Chem. 255, 7838–7844. 7. Marchmont R. J. and Houslay M. D. (1980) Nature 286, 904– 906. 8. Heyworth C. M., Wallace A. V. and Houslay M. D. (1983) Biochem. J. 214, 99–110. 9. Pyne N. J. and Houslay M. D. (1988) Biochem. Biophys. Res. Commun. 156, 290–296. 10. Smoake J. A., Moy G.-M. M., Fang B. and Solomon S. S. (1995) Arch. Biochem. Biophys. 223–232. 11. Degerman E., Smith C. J., Tornqvist H., Vasta V., Belfrage P. and Manganiello V. C (1990) Proc. Natl. Acad. Sci. USA 87, 533–537. 12. Smith C. J., Vasta V., Degerman E., Belfrage P. and Manganiello V. C. (1991) J. Biol. Chem. 266, 13385–13390. 13. Lopez-Aparicio P., Rascon A., Manganiello V. C., Anderson K.-E., Belfrage P. and Degerman E. (1992) Biochem. Biophys. Res. Commun. 186, 517–523. 14. Rascon A., Degerman E., Taira M., Meacci E., Smith C. J., Manganiello V. C., Belfrage P. and Tornqvist H. (1994) J. Biol. Chem. 269, 11962– 11966. 15. Rahn T., Ro¨nnstrand L., Leroy M.-J., Wernstedt C., Tornqvist H., Maganiello V. C., Belfrage P. and Degerman E. (1996) J. Biol. Chem. 271, 11575–11580. 16. Swinnen J. V., Tsikalas K. E. and Conti M. (1991) J. Biol. Chem. 266, 18370–18377. 17. Torphy T. J., Zhou H.-L. and Cieslinski L. B. (1992) J. Pharmacol. Exp. Ther. 263, 1195–1205. 18. Elks M. L., Manganiello V. C. and Vaughan M. (1983) J. Biol. Chem. 258, 8582–8587. 19. Christofferson T., Refsnes M., Bronstad G. O., Huse J., Haffner F., Sand T.-E., Hunt N. H. and Sonne O. (1984) Eur. J. Biochem. 138, 217– 226. 20. Pyne N. J., Anderson N., Lavan B. E., Milligan G., Nimmo H. G. and Houslay M. D. (1987) Biochem. J. 248, 897–901. 21. Nemoz G., Prigent A. F., Moueqqit M., Fougier S., Macovschi O. and Pacheco H. (1985) Biochem. Pharmacol. 34, 2997– 3000.

Combined Hormonal Effects on Cyclic AMP Phosphodiesterase 3 22. Hughes B., Owens R., Perry M., Warrelow G. and Allen R. (1997) DDT 2, 89–101. 23. Taira M., Hochman S. C., Calvo J. C., Taira M., Belfrage P. and Manganiello V. C. (1993) J. Biol. Chem. 268, 18573– 18579. 24. Fleig W.E., No¨ther-Fleig G., Steudter S., Enderle D. and Ditschuneit H. (1985) Biochim. Biophys. Acta 847, 352–361. 25. Thompson W. J. and Appleman M. M. (1971) Biochem. U.S.A. 10, 311–316. 26. Bradford M. M. (1976) Anal. Biochem. 72, 248–254. 27. Pyne N. G., Cooper M. E. and Houslay M. D. (1987) Biochem. J. 242, 33–42. 28. Boyes S. and Loten E. G. (1988) Eur. J. Biochem. 174, 303– 309.

635 29. Kilgour E., Anderson N. G. and Houslay M. D. (1989) Biochem. J. 260, 27–36. 30. Shibata H., Robinson F. W., Soderling T. R. and Kono T. (1991) J. Biol. Chem. 266, 17948–17953. 31. Marchmont R. J., Ayad S. R. and Houslay M. D. (1981) Biochem. J. 195, 645–652. 32. Lavan B. E., Lakey T. and Houslay M. D. (1989) Biochem. Pharmacol. 38, 4123–4136. 33. Flawn P. and Loten E. G. (1990) Int. J. Biochem. 22, 983–988. 34. Huston E., Pooley L. O., Julie P., Scotland G., McPhee I., Sullivan M., Bolger G. and Houslay M. D. (1997) J. Biol. Chem. 271, 31334–31344. 35. Schulz W. A., Gebhardt R. and Mecke D. (1985) Eur. J. Biochem. 146, 549–553.