DECREASED GLUCAGON-STIMULATED CYCLIC AMP

and Ceilufar Endocrinolog);, 32 (1983) 13-26 Elsevier Scientific Publishers Ireland, Ltd.

13

Molecular

MCE 01035

DECREASED GLUCAGON-STIMULATED PRODUCTION BY ISOLATED LIVER TYPE 2 DIABETES Bernard PORTHA a,*, Hilda CHAMRAS LUG PICON a and Gabriel ROSSELIN ’

CYCLIC AMP CELLS OF RATS

b, Yvonne

BROER

WITH

‘,

U Laboratoire de Physiologic du DPveloppement et Laboratoire AssociP au CNRS No. 307, UnioersitP Paris VII, Tour 23, 2 Place Jussieu, 75251 Paris Cedex 0.5, ’ Unite de Recherches de Diabktologie et d’Etudes Radio Immrtnologiques des Hormones ProtPiques (INSERM U.55, CNRS ERA 494), HGpital Saint -Antoine, 184 Rue du Faubourg Saint- Antoine, 75571 Paris Cedex I2 (France) Received

25 March

1983; accepted

3 May 1983

This study was undertaken to investigate the effect of experimental type 2 diabetes in the rat on the insulin and glucagon receptors and on the early steps of glucagon action. The binding of insulin and glucagon and the glucagon-stimulated cyclic AMP accumulation in the presence of a phosphodiesterase inhibitor (IBMX, 0.1 mmoles/l) were studied in liver cells isolated from 7-9-month-old rats with chronic type 2 diabetes and from control rats. No significant change was observed in [ ‘251]insulin binding and [‘251]glucagon binding of diabetic liver cells as compared to controls. Scatchard analysis of the competition experiments indicated that affinity and number of insulin and glucagon receptors were not significantly changed in the liver cells of diabetic rats. The basal cyclic AMP level was significantly lower in the diabetic hepatocytes (2.3kO.9 pmoles/106 cells) than in the controls (4.OkO.6 pmoles/106 cells). Cyclic AMP response to physiological concentrations of glucagon (0.1-l nmoles/l) was about 2 times lower in the diabetic hepatocytes than in the controls. Furthermore, the basal liver membrane adenylate cyclase activity and the fluoride-activatable adenylate cyclase activity were about 2 times lower in the diabetics as compared to control rats, while the liver cyclic AMP and cyclic GMP phosphodiesterase activities were unchanged. The ability of glucagon to stimulate liver membrane adenylate cyclase over a lo- ‘* -lo-$ M concentration range was decreased in diabetic rats. Taken together, these data are consistent with the thesis that the impairment of the liver cyclic AMP response to glucagon in rats with type 2 diabetes is caused by a decrease in the amount of adenylare cyclase in the liver plasma membranes. Keyword.s: experimental type 2 diabetes: insulin receptors; glucagon hepatocytes; cyclic AMP production; liver plasma membranes: cyclase; cyclic nucleotide phosphodiesterases. * To whom all correspondence should be addressed. Present Medical Cell Biology, Box 571, S-75 1 23 Uppsala (Sweden). 0303-7207/83/$03.00

0 1983 Elsevier Scientific

Publishers

Address:

Ireland,

Ltd.

receptors: adenylate

Department

of

14

3. Portha er al.

We have recently developed an experimental model of type 2 diabetes in the adult rat which is obtained as the spontaneous evolution of streptozotocin-induced neonatal diabetes (Portha et al., 1979). This type 2 diabetes in the adult is stable and chronic and is characterized by a slight elevation of basal plasma glucose value and a slightly impaired glucose tolerance (Portha et al., 1979). In vivo and in vitro investigations of the A cell function in this model indicate that (1) basal glucagon levels in the plasma were normal, (2) glucagon release was normally suppressed by increased glucose concentrations but, in response to amino acids or acetylcholine, stimulation of the release was deficient (Giroix et al., 1983). In view of this impairment of A cell function in the rats with type 2 diabetes, the present study was undertaken to determine the characteristics of the glucagon receptors in the liver of rats with type 2 diabetes and the response to glucagon of the hepatic adenylate cyclase-cyclic AMP system. In these experiments we also measured insulin binding to the isolated liver cells.

MATERIALS

AND

METHODS

Albino rats (Wistar strain) were fed ad libitum with pelleted chow (UAR, Villemoisson-sur-Orge, France; carbohydrate 47%, protein 20%, fat 8%). Females were caged with a male for one night (5 p.m. to 9 a.m.) and pregnancy was detected by abdominal palpation 14 days later. Natural birth occurred 22 days after mating. On the day of birth, the newborn rats received streptozotocin (100 p&/g) in 25 ~1 of citrate buffer (0.05 moles/l, pH 4.5) through the saphenous vein directly accessible by transcutaneous puncture. They were left with their own mothers, with the number of animals per litter kept at 8. On day 4 after birth, neonates were tested for glycosuria with Clinistix (Ames Co., Division Miles Labs., Paris, France) and only animals with 3 + values were kept. They were weaned on day 21 after birth. Spontaneous evolution of neontatal diabetes led to a type 2 diabetic state in the adult which was stable and chronic as previously described (Portha et al., 1979). Glucose tolerance was determined in control males and in diabetic males at 7-9 months. Intravenous glucose tolerance tests (0.5 g glucose/ kg body wt) were performed under pentobarbital anaesthesia (4 mg/ 100 g body wt, i.p.). Blood was withdrawn from the tail vein in the fed state. Blood samples (300 ~1) were immediately centrifuged at 4°C; plasma was stored at - 20°C until assayed, The animals were sacrificed 1 week later in the fed state.

Glucagon and CAMP in type 2 diabetic rat liver

15

Highly purified ‘monocomponent’ porcine insulin (MCS 821506) and porcine glucagon (B66) were used for iodination and as standards (supplied by Dr. .I. Schlichtkrull, Novo Research Institute, Copenhagen, Denmark). Carrier-free Na’*‘I, [ 3H]cyclic AMP and f 3H]cyclic GMP were purchased from the Radiochemical Centre (Amersham, U.K.); bovine serum albumin (BSA, fraction V, lot 294) from Miles Laboratories, crude collagenase (type I, 140 IU/mg, lot S&-0075) from Sigma, ~,2-hydroxymethylpiperazide-~~,2-ethanesulphoni~ acid (Hepes, lot 28c5026) from Sigma; streptozotocin from Upjohn Laboratories. The 2’,0succinyl cyclic AMP (cyclic-Sue-AMP), its tyrosine methyl ester (cyclicSue-AMP-Tyr-O-Me), the radioiodinated derivative (‘z I-labelled cyclicSue-AMP-Tyr-O-Me), and antibodies against the albumin-conjugated cyclic-Sue-AMP were prepared in our laboratory (Broer et al., 1972 a,b). Cyclic AMP, cyclic GMP, AMP, ADP, ATP and adenosine were purchased from Calbiochem; bacitracin (lot 124c-01~41), snake venom (Ophiophagus hannah) and soybean trypsin inhibitor from Sigma; Trasy101 8 (6600 kIU/mg) from Bayer AG; and 3-isobutyl-1-methylxanthine (IBMX, lot 12 1247) from Aldrich Chemical. Caffeine and other chemicals were obtained from Merck, and anion exchange resin (AG 1 X 2, 200-400 mesh) from Bio-Rad. Ioslatio~ of Iiver cefk The method of liver cell preparation was the ‘two-step procedure’ described by Seglen (1972, 1973) involving a perfusion of liver first with a calcium-free buffer and second with collagenase buffer. The following modifications of that procedure were made: (1) collagenase buffer (100 ml), containing 50 mg collagenase, was supplemented with 5 mg soybean trypsin inhibitor (Crane and Miller, 1974); (2) after collagenase digestion, the liver was immediately flushed to remove collagenase with lOO- 150 ml calcium-free Krebs-Ringer phosphate (KRP), pH 7.5, containing 0.2 g/ 100 ml bovine serum albumin (BSA) dialysed against bidistilled water; (3) once separated, the cells were centrifuged at 200 rev./min for 1 min. The cell pellet was washed in KRP containing 3 g/100 ml BSA, pH 7.5. Following the third centrifugation the purified parenchymal cell pellet was resuspended in fresh incubation medium (KRP, 3 g/ 100 ml BSA, pH 7.5) at a concentration of 800000 cells/ml, Cells were counted in a Malassez haemocytometer. This procedure yielded 400-500 x lo6 cells/ liver. The viability of the cells was similar in diabetic and control preparations. The proportion of cells excluding trypan blue was about 90-95% before incubation and not lower than 80% after 4 h at 20°C. Extracellular lactic dehydrogenase activity (Neilands, 1955) was not

16

B. Portha et al.

significantly increased after 2 h incubation in either diabetic or control rat liver cells (5.2 i 0.5% and 4.7 k 0.7%, respectively, of the total lactic dehydrogenase activity). The ATP content (Stanley and Williams, 1969) of the diabetic hepatocytes before incubation was not significantly different from that in the controls (11.3 -t 0.8 and 15.3 + 1.4 nmoles/ lo6 cells, respectively). It was 88 f 3% and 86 + 4%, respectively, of initial ATP content after 2 h incubation at 20°C. Hormone-binding assay [‘251]Insulin and [ ‘251]glucagon were monoiodinated, using the chloramine T method under the conditions previously described (Freychet et al., 1971a; Nottey and Rosselin, 1971). The specific activity of labelled insulin and glucagon was 350 and 600 Ci/g, respectively. Studies of insulin and glucagon binding was performed at 20°C in calcium-free KRP buffer as previously described (Freychet et al., 1974) with the following modifications: each incubation tube contained, in a final unvolume of 0.5 ml, ‘251-labelled hormone at about 0.3 nmoles/l, labelled hormone at the concentrations indicated in the figures, BSA at 3 g/100 ml and 0.5 x lo6 cells/ml. The medium of the cells incubated with glucagon also contained 2000 IU/ml of kallikrein inhibitor (Trasylol) and 100 pg/ml of bacitracin (Desbuquois et al., 1974) as inhibitors of glucagon degradation, 0.1 mmoles/l of 3-isobutyl- 1-methylxanthine (IBMX) and 10 mmoles/l of alanine. The cell-bound hormone was separated by centrifugation as described previously (Rodbell et al., 1971; Freychet et al., 1971b). The supernatant was removed and the radioactivity of the washed pellet was determined by gamma spectrometry. Hormone degradation assay Hormone degradation was measured by the method previously dewas scribed (Freychet et al., 1972). ‘251-labelled hormone (0.3 nmoles/l) exposed to liver cells of control and diabetic rats. After 2 h incubation at 20°C the ‘251-labelled hormone remaining intact in the medium was tested for its ability to rebind to specific receptors in liver membranes. Appropriate controls without cells represented 100% of the substrate available for degradation. Measurements of cyclic AMP Endogenous cyclic AMP was measured by radioimmunoassay in the methanol extract of the whole mixture (cells + medium) as previously described (Rosselin et al., 1974). For concentrations lower than 20 nmoles/l, the sensitivity of the radioimmunoassay was increased by the

11

Glucagon and CAMP in type 2 diabetic rat her

succinylation of the cyclic Cailla et al. (1973).

AMP

following

the method

described

by

Adenylate cyclase assay Purified plasma membranes were prepared according to Neville (1968). Adenylate cyclase was measured as previously described (Rosselin and Freychet, 1973). Activity was measured in a final volume of 250 ~1 containing 1 mM ATP, 20 mM creatine phosphate, 1 mg/ml creatine phosphokinase, 0.4 mM IBMX, 4 mg/ml bovine serum albumin, 200 fig/ml bacitracin, and glucagon (from lo-‘* up to lop6 M) or 10 mM sodium fluoride. The assays were initiated by addition of the enzyme preparation (150 pg membrane protein per tube) to reaction tubes and were performed in triplicate at 30°C for 15 min (similar results were obtained at 20°C; data not shown). The cyclic AMP produced was quantitated by radioimmunoassay. Data are expressed as pmoles of cyclic AMP produced per minute and per mg membrane protein. Cyclic nucleotide phosphodiesterase assay The enzyme activity was assayed in crude liver homogenate. Hepatic tissue (about 2 g) was homogenized in 5 ~01s. (w/v) of 40 mM Tris-HCL buffer, pH 8, containing 5 mM MgCl, and 3.75 mM 2-mercaptoethanol. After a 20 min centrifugation, the 20000 x g supernatant was removed and diluted in the same buffer. Phosphodiesterase activities were measured in 250 ~1 assay mixtures containing [ 3H]cyclic AMP (25 000 cpm) or [3H]cyclic GMP (20000 cpm), from 0 up to 64 PM cyclic AMP or cyclic GMP as standards, 50 mM snake venom and 80 pg liver proteins in TrisHCl buffer. The mixture was incubated at 30°C for 15 min for the cyclic AMP phosphodiesterase assay and for 30 min for the cyclic GMP phosphodiesterase assay. The reaction was terminated by the addition of 500 ~1 anion exchange resin (Thompson and Appleman, 1971). Blanks were obtained in the same conditions except that 500 ~1 anion exchange resin was first added and cyclic AMP and cyclic GMP standards were omitted. Data are expressed as nmoles of cyclic AMP or cyclic GMP hydrolysed per minute and per mg protein. Other determinations Plasma immunoreactive glucagon was measured using the Unger glucagon antibody 30 K (Faloona and Unger, 1972). Plasma insulin concentration was estimated by radioimmunoassay (Rosselin et al., 1966), the results being expressed in terms of a rat insulin standard. Plasma glucose was determined using a glucose analyser (Beckman Inc., Palo Alto, CA, U.S.A.). Protein concentration was measured by the method of Lowry et al. (1951) using BSA as standard.

18

B. Portha et al

Expression of results Specific binding was obtained by subtracting from the total binding the amount of labelled hormone which was not displayed by an excess of unlabelled hormone (17 pmoles/l for insulin, 14 pmoles/l for glucagon). The amount of hormone specifically bound per cell was calculated from competition experiments at steady state were ‘251-labelled hormone was incubated in the absence or the presence of increasing concentrations of unlabelled hormone. The number of binding sites per cell and the corresponding apparent dissociation constant were also estimated by Scatchard plots of the same data. Results were expressed as means + SEM. Statistical analysis was performed using Student’s unpaired t-test.

RESULTS Characteristics of experimental animals Body weight and basal plasma glucagon levels in the male diabetic rats were not significantly different from the controls. As indicated in Table 1, the diabetic rats exhibited in the fed basal state a slightly elevated blood glucose level (P < 0.001) and a lower plasma insulin level (P < 0.001) as compared to controls. The diabetic rats had a decreased glucose tolerance and a considerable decline in the glucose-induced insulin release, in accordance with our previous results (Portha et al., 1979). Insulin binding Fig. 1 (left) Table

summarizes

the insulin-binding

studies

performed

on

or control

rats.

1

Plasma glucose, insulin and glucagon in 7-9-month-old Glucose (0.5 g/kg) was administered iv.

Controls

Diabetics

male

diabetic

Insulin 5 min after glucose load

Basal glucose (mg/lOO ml)

Glucose 90 min after glucose load (mg/lOO ml)

127+3

131*4

71+2

241 f 23

(12)

(12)

(12)

(12)

183+9 (9)

*

236+28 (9)

*

Basal insulin ( pU/ml)

Basal glucagon (pg/ml)

( pU/ml)

37+4 (9)

*

110*20 (9)

3575 14 (7) *

348 + 40 (6)

Values are means + SEM. The number of observations is shown in parentheses. * P c 0.001 when compared by Student’s unpaired f-test with controls.

19

Giucagon and CAMP in type 2 diabetic rat her

INSULIN

, nmol/l

i:l’;i__.. l

0

100 BOUND

GLUCAGON

L

200 INSULIN

l

.

, mel f I

1

500 , rmol/

IO* ~111%

BOUND

GLUCAGON

, rmd 1106trh

Fig. 1. Insulin binding to isolated hepatocytes from control (C) and diabetic (D) rats. Incubation conditions were as described in Materials and Methods. The upper panel represents, at each hormone concentration, the amount of insulin bound, calculated from the competition experiments. The lower panel represents the Scatchard analysis of the results. The ratio of bound to free insulin was plotted as a function of bound insulin. Each point is the mean &SEM of 5 separate cell preparations for control and diabetic rats. (Right) Glucagon binding to isolated hepatocytes from control (C) and diabetic (D) rats. Incubation conditions were as described in Materials and Methods. The upper panel represents, at each hormone concentration, the amount of glucagon bound, calculated from the competition experiments. The lower panel represents the Scatchard analysis of the results. The ratio of bound to free glucagon was plotted as a function of bound glucagon. Each point is the mean i SEM of 5 separate cell preparations for control and diabetic rats.

isolated hepatocytes from diabetic and control rats. The binding of [ i2’I]insulin was studied after a 120 min incubation time, since we had previously determined that insulin binding was maximum and stable in both groups between 60 and 180 min of incubation. Binding was expressed as a function of increasing unlabelled insulin concentration. The initial binding of [ ‘251]insulin was not significantly different in cells of diabetic rats (7.4 + 1.2% of total [‘251]insulin, n = 6) from binding in cells of control rats (6.8 f 1.2% of total [lz51Jinsu~n, n = 6). As can be seen when the data are expressed as quantities of insulin bound per lo6 cells

20

B. Porrha

et al.

(Fig. 1, left, top), hepatocytes of diabetic rats bound as much as those of controls at all insulin concentrations. The 50% inhibition of [ ‘*‘I]insulin binding was achieved with about 1.5 nmoles/l insulin in both types of liver cells, suggesting an identical affinity. Scatchard analysis of the data (Fig. 1, left, bottom) revealed similarly shaped curvilinear plots for both groups, thus indicating no change in affinity or number of the receptors for insulin. To rule out the possibility that apparently unchanged insulin binding in diabetic rats could be an artefact due to increased degradation of the hormone in the incubation medium, masking increased insulin binding, the degradation of insulin was examined: there was no significant difference in the amount of insulin degraded after a 120 min incubation at 20°C with hepatocytes of control rats (82 t_ 2%. n = 4) or of diabetic rats (78 f 4%, n = 4). Giucagon binding and glucagon-stimulated cyclic AMP accumulation In both the control and the diabetic groups, maximal level of binding was attained by 45 min and apparent steady state was maintained until 240 min (data not shown). The binding of glucagon was studied as a function of increasing concentrations of unlabelled hormones (l-75 nmoles/l) after a 60 min incubation time (Fig. 1, right, top). The amount of [‘251]glucagon bound after a 60 min incubation period was 5.8 k 0.7% (n = 5) and 6.5 f 0.5% (n = 5) in control and diabetic rat liver cells, respectively. The amount of glucagon bound at equilibrium was calculated from the competition curve and no significant difference in specific glucagon binding was found between diabetic rats and control rats (Fig. 1, right, top). The 50% inhibition of the [‘251]glucagon binding was achieved with about 8 nmoles/l glucagon in both conditions: this suggests that the apparent average affinity is similar in both cases. Scatchard analysis of the data (Fig. 1, right, bottom) showed that diabetic and control curves are curvilinear and superimposed, indicating that the binding affinities and the binding capacities for glucagon are unchanged. The degradation of labelled glucagon that had been exposed to liver cells for 60 min at 20°C was not significantly different in the diabetic group. To determine whether the unchanged binding of glucagon to diabetic rat hepatocytes was accompanied by a corresponding unchanged biological activity, measurements were made of cyclic AMP formation by the hepatocytes in response to glucagon. The glucagon-stimulated cyclic AMP accumulation was studied in the presence of 0.1 mmoles/l 3-isobutyl- 1-methylxanthine. In previous experiments, we found that the addition of IBMX at a concentration higher than 0.1 mmoles/l did not result in any further significant increase of cyclic AMP by 1 nmole/l

21

r_: F=* *

2

> u

0L: oV’o,3

1

10 GLUCAGON, nmol/l

100

Fig. 2. Time-course of glucagon-stimulated cyclic AMP accumulation in isolated liver cells from control (C) and diabetic (D) rats. Cells (0.5 X 106/ml) were incubated at 20°C in the presence of I nmole/l glucagon. The incubation buffer was identical to that used for glucagon binding (see Materials and Methods). Each point is the mean + SEM of 5 separate experiments for control and diabetic rats. Fig. 3. Dose-response of glucagon-stimulated cyclic AMP accumulation in liver cells of control (C) and diabetic (D) rats. Cells (0.5 x 106/ml) were exposed to various concentrations of glucagon and incubated for 30 min at 20°C. Each point is the mean +SEM of 5 separate experiments for control and diabetic rats.

glucagon in both groups of liver cells, suggesting that the phosphodiesterase activity was maximally inhibited under these conditions. In the absence of glucagon the cyclic AMP level was significantly lower (P ( 0.05) in diabetic (2.3 _C0.2 pmoles/ 10h cells, n = 5) than in control (4.0 k 0.6 pmoles/ lo6 cells, n = 5) liver cells. Fig. 2 shows the time-course of the cyclic AMP production in the presence of I nmole/l glucagon. At any time considered during the first hour the cyclic AMP level was lower in the diabetic than in the control cells. In both cases a plateau of cyclic AMP level was reached after 30 min. The results of dose-response studies to glucagon are shown in Fig. 3. Cyclic AMP response to concentrations of glucagon below 15 nmoles/l was significantly lower in diabetics than in normals. At higher concentrations of glucagon very far from the physiological range (28 and 280 nmoles/l), cyclic AMP production in diabetics was not significantly different from that measured in controls. For the concentrations of glucagon (O.l--0.5 nmoles/l) in the physiological range (portal vein), the cyclic AMP level in diabetic hepatocytes was about 2 times lower than in controls. Cyclic nucleotide phosphodiesterases and adenylate cyclase activities The decreased cyclic AMP accumulation in response to glucagon

22

19. Portha et ai.

Table 2 Phosphodiesterase Substrate:

activity

(nmoles cyclic nucleotide

breakdown/min/mg

Cyclic AMP Low K,

Controls

3.5 i: 0.3 (3)

Diabetics

Cyclic GMP High K,

form

K,

Vman

K,

0.21iO.03

24.1k5.5

(3)

3.6 * 0.36 (4)

protein)

0.20 + 0.02 (4)

(3) 27.5 + 5.5 (4)

form V“18.X 1.01 kO.23 (3) 1.511:0.23 (4)

High K, K, 17.8k3.8 (3) 20.3fl.l (6)

form VtnPX 0.22 _t 0.06 (3) 0.29 i 0.02 (6)

The phosphodiesterase activity of control or diabetic liver homogenates was measured using from 0 to 64 PM cyclic AMP or cyclic GMP as substrate. Activity values corresponding to K, and K,,x are given for each form of the enzyme. Values are the mean t_ SEM of 336 separate experiments for control and diabetic rats.

observed in the hepatocytes of diabetic rats may result from a decrease in the rate of cyclic AMP formation and/or from an increase in the rate of cyclic AMP degradation. The possibility of increased phosphodiesterase activity was investigated by assaying for this enzyme in crude extracts prepared from total liver of control or diabetic rat. The results, shown in Table 2, indicate that the low as well as the high Michaelis constant (K,) cyclic AMP phosphodiesterase activities were unaltered in the liver of the diabetics. Moreover, the high K, cyclic GMP phosphodiesterase activities for controls and diabetics were also the same. The hormone-sensitive adenylate cyclase system of the diabetic rat hepatocyte was investigated next in an attempt to identify the basis for \ .O c .6 D

Y

0

1210 8 6

-L~~[GLUCAGON],M Fig. 4. Dose-response of glucagon-stimulated adenylate cyclase activity in purified liver plasma membranes from control (w) and diabetic (A) rats. Adenylate cyclase activity was also measured in the presence of 10 mM NaF (0, A). Incubation conditions were as described in Materials and Methods. Data were expressed as pmoles of cyclic AMP produced per minute and per mg membrane proteins. Each point is the mean of 4 separate experiments for control and diabetic rats.

Glucagon and CAMP in type 2 diabetic rat liver

23

the impaired cyclic AMP response. Adenylate cyclase activity was measured in a purified membrane fraction, both in basal condition and under glucagon stimulation ( 10-‘2- lop6 M) (Fig. 4). Basal adenylate cyclase activity was 2 times lower (5.1 t_ 1 pmoles cyclic AMP/min/mg protein, n = 4) in the liver membranes of the diabetic rats as compared to the value found in the control rats (12.6 + 2 pmoles cyclic AMP/min/mg protein, n = 4). The glucagon-stimulated adenylate cyclase activity in the - 50% in the diabetic rats over a same membranes was diminished 10-‘2-10-6 M range of glucagon concentrations (Fig. 4). These data therefore implicate the adenylate cyclase system as the primary site of the impaired glucagon response of the diabetic rat hepatocyte. The amount of fluoride-stimulatable adenylate cyclase activity was also measured in the same liver membrane preparations to provide an index of the catalytic activity of the cyclase. It was about 50% less in the purified membrane fraction of the diabetics (40 * 5 pmoles cyclic AMP/min/mg protein, n = 4) as compared to controls (87 + 9 pmoles/min/mg protein, n = 4). These data suggest that the reduced glucagon response in the diabetics might reflect a reduced amount of catalytic cyclase in the liver membranes.

DISCUSSION Analysis of the characteristics (affinity and number) of the insulin and glucagon receptors of the liver cells indicated that they were kept unchanged in rats with type 2 diabetes. This is in sharp contrast with numerous reports indicating alterations of insulin binding in diabetes, but all these conclusions were obtained only in acute severe diabetes. Except for a report which concluded that there was no difference in insulin binding to adipocytes from streptozotocin diabetic and control rats (Bennett and Cuatrecasas, 1972), most of the authors found that isolated hepatocytes (Chamras et al., 1980; Samson et al., 1982) or isolated adipocytes (Schoenle et al., 1977; Kasuga et al., 1978; Kobayashi and Olefsky, 1979) from severely diabetic rats bound more insulin than those from controls. The same conclusion was obtained with hepatic plasma membranes from genetically diabetic Chinese hamster (Hepp et al., 1975) or rats with severe streptozotocin diabetes (Davidson and Kaplan, 1977). In most of these studies the basal plasma insulin levels in severely diabetic animals were decreased by 43-80%. So one possibility to explain the increased insulin binding to target tissues from severely diabetic rats involves the decreased plasma insulin concentrations in diabetic rats according to the ‘up- and down-regulation’ theory, since

24

B. Porthu et cd.

several studies have suggested that the insulin concentration can regulate the number of insulin receptors in vivo and in vitro (Gavin et al., 1974; Sol1 et al., 1975). However, our current results regarding the insulin binding do not substantiate the down-regulation concept despite a significantly 50% decreased basal plasma insulin level. Nevertheless it must be noticed that diabetes in our study was a chronic disease (Portha et al., 1979), since the males used had been exposed to 7-9 months of insulinopenia, a condition sharply in contrast with the short duration of the diabetic state (generally 1 or 2 weeks) in most of the studies with acute streptozotocin diabetes. So the absolute level of insulin in the plasma does not seem to be the only determinant of the changes in the insulin receptors. A similar conclusion was obtained with experiments in genetically obese Zucker rats, in obese VMH-lesioned rats and in human diabetic subjects. In obese Zucker rats, binding to isolated hepatocytes had been found to be normal despite chronically elevated (6 times increased as compared to lean controls) plasma insulin levels (Broer et al.: 1977). In obese VMH-lesioned rats, insulin binding to isolated adipocytes was similar to controls despite 3 times increased basal plasma insulin levels (Kasuga et al., 1980). In non-obese adult-onset diabetics, despite low insulin response to glucose, monocytes bound less insulin than cells from normal subjects; however, the authors have correlated this feature with the basal hyperinsulinaemia of their diabetic subjects (Olefsky and Reaven, 1977). Concerning the glucagon, no significant change in binding was observed in liver cells of rats with type 2 diabetes. This has to be considered in the light of the unchanged basal plasma glucagon levels and decreased amino acid stimulated glucagon release reported in these animals (Giroix et al., 1983). In rats with type 2 diabetes, contrasting with unchanged insulin and glucagon receptors in the liver, the glucagon-induced cyclic AMP production was found to be clearly decreased in a physiological range of glucagon concentrations. Furthermore, the basal hepatocyte cyclic AMP level, the basal liver membrane adenylate cyclase activity and the fluoride-activatable adenylate cyclase activity were about 2 times lower in the diabetics as compared to control rats. The cyclic nucleotide phosphodiesterase activities were similar in the two groups. Finally, the glucagon-stimulated adenylate cyclase was decreased in liver membranes of diabetic rats. Taken together, these data are consistent with the thesis that the impairment of the liver cyclic AMP response to glucagon in rats with type 2 diabetes is caused by a decrease in the amount of adenylate cyclase in the liver plasma membranes. The abnormality of glucagon-induced cyclic AMP production in these rats with type 2 diabetes is similar to that found in isolated liver cells or

Ghcagon

and CAMP in type 2 diabetic rut iicer

25

slices of overtly diabetic rats (Bhathena et al., 1978; Chamras et al., 1980; Yamashita et al., 1980). Nevertheless, it may be stressed that in none of the above-mentioned diabetic models at variance with our own has the quantity of adenylate cyclase been reported to be decreased. Our results indicate that this reduction in cyclic AMP response to glucagon is found in the absence of any change in the glucagon binding. Similar observations were also presented in rats with overt diabetes (Chamras et al., 1980; Samson et al., 1982). Moreover this reduced responsiveness to glucagon is found in our model in the absence of hyperglucagonaemia. This observation does not contradict the suggestion that hyperglucagonaemia, always present in acute diabetes (Unger, 1976) induces per se a desensitization to glucagon of the hepatic cyclic AMP production, but it suggests that a different mechanism may also lead to desensitization. We hypothesize that this phenomenon, if it is linked to plasma levels of pancreatic hormones, might be more probably related to the chronic exposure of the liver to low plasma insulin levels. In conclusion, the decreased biological effect of glucagon on the liver of rats with type 2 diabetes might be interpreted from a teleological point of view as a mechanism for reducing to some extent hepatic glucose production and excessive hyperglycaemia. This might partly explain the persistence of a slight glucose intolerance despite a marked reduction in the insulin secretion. liver

ACKNOWLEDGEMENTS We are grateful to Mrs. C. Rouyer-Fessard and N. Vauclin for their excellent technical help and to Mrs. D. Lhenry and Mrs. K. Claesson for their careful preparation of the manuscript. This work was supported by the Institut National de la Sante et de la Recherche Medicale (INSERM (CRL No. 79.1.483.7)).

REFERENCES Bennett, G. and Cuatrecasas, P. (1972) Science 176, 805406. Bhathena, S.J., Voyles, N.R., Smith, S. and Recant, L.. (1978) J. Clin. Invest. 61, 1488I497. Broer, Y., Fouchereau, M. and Rosselin, G. (1972a) CR. Acad. Sci. (Pat-is) 275, 619-622. Broer, Y., Lhiaubet-Grapin, A.M. and Rossehn, G. (1972b) CR. Acad. Sci. (Paris) 275. 883-886. Broer, Y., Freychet, P. and Rosselin, G. (1977) Endocrinology 101, 236249. Cailla, H.L., Racine-Weisbuch, MS. and Delaage, M.A. (1973) Anal. Biochem. 56, 3944407. Chamras, H., Fouchereau-Peron, M. and Rosselin, G. (1980) Diabetologia 19, 74-80.

26

B. Portha et al.

Crane, L.J. and Miller, D.L. (1974)B&hem. Biophys. Res. Commun. 60, 1269% 1277. Davidson, M.B. and Kaplan, S.A. (1977) J. Clin. Invest. 59, 22-30. Desbuquois, B., Krug, F. and Cuatrecasas, P. (1974) Biochim. Biophys. Acta 343. 101-120. Duttagupta, C., Rifas, L. and Makman, M.H. (1978) Biochim. Biophys. Acta 523, 385-394. Faloona, G.R. and Unger, R.H. (1972) In: Methods of Hormone Radioimmunoassay, Eds.: B.M. Jaffe and H.R. Behrman (Academic Press, New York) pp. 317-330. Freychet, P., Roth, J. and Neville Jr., D.M. (1971a) Biochem. Biophys. Res. Commun. 43. 400-408. Freychet, P., Roth, J. and Neville Jr., D.M. (1971b) Proc. Natl. Acad. Sci. (U.S.A.) 68, 1833-1837. Freychet, P., Kahn, CR., Roth, J. and Neville Jr., D.M. (1972) J. Biol. Chem. 247, 3953-3961. Freychet, P., Rosselin, G., Rancon, F., Fouchereau-P&on, M. and Broer, Y. (1974) Hormone Metab. Res., Suppl. 5, 72-78. Gavin, J.R., Roth, J., Deville Jr., D.M., De Meyts, P. and Buell, D.N. (1974) Proc. Natl. Acad. Sci. (U.S.A.) 71, 84-88. Giroix, M.H., Portha, B., Kergoat, M. and Picon, L. (1983) Diabite et Mt-tabolisme (submitted). Hepp, K.D., Lanley, J., Von Funcke, H.J., Renner, R. and Kemmler, W. (1975) Nature (London) 258, 154. Kasuga, M., Akanuma, Y., Iwamoto, Y. and Kosaka, K. (1978) Am. J. Physiol. 235, E175-E182. Kasuga, M., Inoue, S., Akanuma, Y. and Kosaka, K. (1980) Endocrinology 107, 1549-1555. Kobayashi, M. and Olefsky, J.M. (1979) Diabetes 28, 87-95. Lowry, O.H., Rosebrough, N.J.: Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275. Neilands, J.B. (1955) Methods Enzymol. 1, 449-452. Neville, Jr., D.M. (1968) Biochim. Biophys. Acta 154, 540-552. Nottey. J.J. and Rosselin, G. (1971) C.R. Acad. Sci. (Paris) 273, 2118-2121. Olefsky, J.M. and Reaven, G.M. (1977) Diabetes 26, 680-688. Portha, B., Picon, L. and Rosselin, G. (1979) Diabetologia 17, 371-377. Rodbell, M., Krans, M.J.. Pohl, S.L. and Birnbaumer, L. (1971) J. Biol. Chem. 246. 1861-1871. Rosselin, G. and Freychet, P. (1973) Biochim. Biophys. Acta 304, 541-551. Rosselin, G., Assan, R., Yalow, R.S. and Berson, S.A. (1966) Nature (London) 212, 355-357. Rosselin, G., Freychet, P., Fouchereau, M., Rancon, F. and Broer, Y. (1974) Hormone Metab. Res. Suppl. 5, 76-86. Samson, M., Fehlman, M., Morin, O., Dolais-Kitabgi, J. and Freychet, P. (1982) Metabolism 3 1, 766-77 1. Santos, A. and Blazquez, F. (1982) Diabetologia 22, 362-371. Schoenle, E., Zapf, J. and Froesch, E.R. (1977) Diabetologia 13, 243-249. Seglen, P.O. (1972) Exp. Cell. Res. 74, 450-454. Seglen, P.O. (1973) Exp. Cell. Res. 76, 25-30. Soll, A.H., Kahn, CR., Neville Jr., D.M. and Roth, J. (1975) J. Clin. Invest. 56, 769-780. Stanley, P.E. and Williams, S.G. (1969) Anal. Biochem. 29, 381-392. Thompson, W.J. and Appleman, M.M. (1971) Biochemistry 10, 311-316. Unger, R.H. (1976) Diabetes 25, 136-151. Yamashita, K., Yamashita, S., Yasuda, H., Oka, Y. and Ogata, E. (1980) Diabetes 29, 188-192.