Decreased Glucagon Receptors in Diabetic Rat Hepatocytesdm5migu4zj3pb.cloudfront.net/manuscripts/109000/109069/... · 2014-01-30 · 3.75 x 106 hepatocytes in a final volume of 2.0

Decreased Glucagon Receptors inDiabetic Rat Hepatocytes

EVIDENCEFORREGULATIONOF GLUCAGON

RECEPTORSBY HYPERGLUCAGONEMIA

SAMJ. BHATHENA, NANCYR. VOYLES, STEWARTSMITH, and LILLIAN RECANT,Diabetes Research Laboratory, Veterans Administration Hospital andGeorgetown University, Washington, D. C. 20422

A B S T RA C T The effects of endogenous and exoge-nous hyperglucagonemia on the specific binding of glu-cagon to hepatocyte receptors was studied, as was the re-sponse of cAMPto glucagon. In streptozotocin diabeticrats, blood glucose and plasma glucagon increased andplasma insulin decreased as compared with controls.Insulin treatment in diabetic rats restored blood glu-cose and plasma glucagon toward normal and elevatedplasma insulin. Specific binding of 1251-glucagon to iso-lated hepatocytes (106 cells) decreased in diabetic rats(8.17±0.38%) compared to controls (14.05±0.87%) andwas restored by insulin treatment (12.25±0.93%). Spe-cific binding of 1251-insulin in controls was 7.30±10.16%; itincreased in diabetic rats to 12.50±0.86%, and decreasedin diabetic rats after insulin treatment (9.08±0.87%).Scatchard analysis and the competition plots of the dataindicate that decreased glucagon binding and increased'insulin binding in diabetes were due to change in thenumber of receptors rather than a change in their af-finity. Hepatocyte cAMP response to glucagon (0.25-5.0 ng/ml) was almost abolished in diabetic rats and wasrestored with insulin treatment.

Specific glucagon binding by hepatocytes from chron-ically hyperglucagonemic (glucagon injected) rats wasdecreased (P < 0.005) to 8.76±0.61% compared withcontrols (13.20±0.74%) and acutely hyperglucagonemicanimals (13.53±1.33%). The decreased binding was as-sociated with a 70% decrease in hepatocyte cAMP re-sponse to glucagon compared with a normal response inacutely hyperglucagonemic rats.

These data appear to support the concept of receptorregulation by ambient hormone level. In both endog-

This study appeared in abstract form; 1977. Diabetes. 26(Suppl. 1): 387.

Received for publication 28 June 1977 and in revisedform 6 February 1978.

enous and exogenous hyperglucagonemia, however,there was a disproportionately large decrease in cAMPresponse to glucagon compared to the decrease inglucagon binding.

INTRODUCTION

The diabetic state in both animals and man is char-acterized by a relative or absolute deficiency of insulin(1, 2). In the presence of intestinal or pancreatic sourcesof glucagon, diabetes is accompanied by an absolute orrelative increment in plasma immunoreactive glucagon(IRG)' (3). IRG levels progressively rise as the diabeticsyndrome becomes more severe, e.g., in diabetic keto-acidosis (4, 5), the levels of IRG may increase as muchas 100-fold over normal fasting plasma levels of 40-100pglml. The correction of the metabolic abnormalities ofthe diabetic state by insulin treatment is accompaniedby a reduction in plasma IRG toward normal (6, 7).These changes have been documented in both experi-mental animals and human diabetic subjects.

A highly significant inverse correlation has been dem-onstrated between the circulating plasma levels of in-sulin (IRI) and insulin binding to specific receptors inthe experimental diabetic animal (8, 9), as well as in thehyperinsulinemic diabetic patient (10-12). It seemedpertinent, therefore, to determine whether similar rela-tionships existed between glucagon receptors and plasmaIRG levels. The availability of a model for endogenoushyperglucagonemia, namely, isolated hepatocytes ob-tained from the streptozotocin diabetic rat, before andafter treatment with insulin, permitted these studies tobe carried out. To test this hypothesis in a more direct

'Abbreviations used in this paper: B/F, bound glucagonto free glucagon ratio; IRG, immunoreactive glucagon; IRI,immunoreactive insulin.

The Journal of Clinical Investigation Volume 61 June 1978 * 1488-14971488

manner, the effect of' exogenous hyperglucagonemiaupon glucagon binding in normal rats was also studied.

The observations support an inverse correlation be-tween elevated plasma levels of IRG and specific glu-cagon binding under conditions of chronic, but not acute,hyperglucagonemia. Further, the alterations in glu-cagon binding are shown to correlate with thebiological effectiveness of the glucagon molecule asmeasured by the formation of cAMP.

METHODS

Male Sprague-Dawley rats (Flow Laboratories, Dublin,Va.) weighing 120-150 g were used for all studies. Twotypes of studies were conducted. In the first, rats were madediabetic by injecting 85 mg/kg streptozotocin (Calbiochem,San Diego, Calif.) through the femoral vein after an overnightfast. They were allowed free access to f'ood and water but weresupplemented with 5% glucose in water for the first 2 daysafter injection. The animals were sacrificed 5-10 days laterwhen their blood glucose levels rose to >400 mg/100 ml. After5 days, some diabetic animals were treated for an additional 5days with subcutaneously administered isophane insulin, 5 U/day (Eli Lilly and Co., Indianapolis, Ind.). All animals werestudied in the fed state and sacrificed at approximately thesame time of day. The rats were anesthetized with 100 mg/kgi.p. sodium amobarbital (Amytal, Eli Lilly and Co.), and bloodwas collected f'rom the abdominal aorta into tubes containing1,000 U Trasylol (FBA Pharmaceuticals, Inc., New York) and10.5 mg EDTAand hepatocytes were prepared.

In the second study, the effect of acute infusion or chronicinjection of porcine pancreatic glucagon (Eli Lilly and Co.)was assessed on glucagon binding in fed normal rats.

Infusion. The rats were anesthetized by amytal, and glu-cagon (2 mg/2 ml) was infused through the femoral veinover a 2-h period, using a Harvard pump (Harvard ApparatusCo. Inc., Millis, Mass.). Control animals were infused with2 ml saline over a 2-h period.

Injection. The rats were injected intramuscularly with 100,ug glucagon twice daily for 7 days. No injection was given onthe day of sacrifice. The rats were anesthetized and the hepato-cytes were prepared.

Blood glucose was measured by the glucose oxidase methodusing a protein-free filtrate. Plasma immunoreactive insulin(IRI) was measured according to Morgan and Lazarow (13),and plasma IRG was assayed according to Unger et al. (14)using antibody 30K. Hepatocytes were prepared by the col-lagenase (CLS IV, Worthington Biochemical Corp., Freehold,N. J.) digestion of liver according to Zahlten and Stratman(15) with the addition of 5 mg/100 ml soybean trypsin inhibitor(Sigma Chemical Co., St. Louis, Mo.) (16). The viability of thecells was checked by the trypan blue exclusion technique andfound to be >95%.

1251-Glucagon and 1251-insulin were prepared according toHunter and Greenwood (17), as modified by Giorgio et al. (18),sp act = 0.20-0.4 and 0.1-0.2 mCi/nmol, respectively. Carrier-free 1251-iodine was purchased from New England Nuclear,Boston, Mass. Pancreatic porcine insulin and pancreaticporcine glucagon were gifts from Eli Lilly and Co.

The iodinated insulin preparations were demonstrated tohave biological activity equivalent to unlabeled hormone asmeasured by [14C]glucose conversion to "4CO2 in isolatedadipose tissue (19, 20). The activity of 1251-glucagon prepara-tions was demonstrated in isolated hepatocytes by measuringcAMP formation (20, 21).

To measure insulin binding, 106 cells were incubated for 30

min at 20°C in a final volume of 0.2 ml of a 0.3-M potassiumphosphate buffer, pH 7.5, containing 1%bovine serum albumin(Sigma Chemical Co.) and 0.1 ng 1251-insulin in the presence of0-10 gg unlabeled porcine insulin. Glucagon binding wasmeasured by incubating 106 cells for 30 min at 30°C with 0.2 ng1251-glucagon and 0-10 ,ug unlabeled porcine pancreatic glu-cagon in a final volume of 0.2 ml of medium containing 25 mMtris buffer, pH 7.5, with 1% bovine serum albumin, 10 mMglucose, 1 mMEDTA, 1.4 mMsodium acetate, 5.0 mMKCI,120 mMNaCl, and 2.4 mMMgSO4. 100 ,ug/ml Bacitracin (giftfrom The Upjohn Co., Agricultural Prods MKT, Kalamazoo,Mich.) was always used in glucagon-binding studies toprevent the degradation of glucagon by hepatocytes. Sampleswere incubated in a Dubnoff incubator with vigorous shaking.At the end of incubation, the tubes were spun in a microfugeand the supernate was discarded. The pellet was then countedin a gamma scintillation counter. Nonspecific binding, i.e.,binding in the presence of 10 ,tg native hormone (usually15-25% of the total binding) was subtracted from the totalbinding to determine the specific binding. The data weresubjected to Scatchard analysis (22) to determine if there werechanges in number of receptors. Additional plots were alsoconstructed of the percent of maximum specific bindingagainst the total concentration of hormone (tracer plusunlabeled) in the incubation media (23) to determine theapparent affinity.

To measure the degradation of glucagon by hepatocytes, theincubation was carried out as described for binding studies ex-cept that the final volume was 1.0 ml. Cells were spun down inthe cold, and 1.0 ml of mediuni containing 0.5% charcoal,0.25% Dextran T-80 (Pharmacia Fine Chemicals Div. of Phar-macia Inc., Piscataway, N. J.), 0.25% human serum albumin(North American Biologicals, Inc., Miami, Fla.) and 1%sheepserum (North American Biologicals, Inc.) in 0.2 M glycinebuffer (pH 8.8) was added to the supemate. The tubes wereincubated at 4°C for 45 min and spun in the cold. The super-nate and pellet were counted separately. The counts pre-cipitated by charcoal were considered intact glucagon. Theratio of counts in the supernate to the total counts determinedthe percent of added honnone which was degraded. To meas-ure insulin degradation, 10% trichloroacetic acid (TCA) wasused instead of the charcoal-Dextran mixture. The biologicalresponse of cells to glucagon was studied by measuring thecAMPformation in the presence of 0-10 ng porcine glucagon/3.75 x 106 hepatocytes in a final volume of 2.0 ml by themethod described previously (21).

RESULTS

Studies involving endogenous hyperglucagonemia.Blood glucose, plasma IRI, and plasma IRG values innormal controls, streptozotocin-treated diabetic rats,and insulin-treated diabetic rats are shown in Table I.Injecting 85 Ag/kg streptozotocin raised the level ofblood sugar three-fold in 5 days. When these animalswere treated with 5 U isophane insulin/day for 5 days,the blood sugar returned to normal. Plasma IRI sig-nificantly decreased in diabetic rats, and plasma IRGincreased to twice the normal level. On insulin treat-ment, plasma IRI increased significantly and plasmaIRG decreased towards control values.

Hepatocytes obtained from diabetic rats, insulin-treateddiabetics, and normal controls showed linearity forboth glucagon and insulin binding as related to thenumber of cells (25 x 104-2.5 x 106). Insulin binding

Glucagon Receptors in Diabetic Rats 1489

TABLE IEffect of Diabetes on Blood Glucose, Plasma Insulin,

and Glucagon in Fed Rats

Animal No. Bloodpreparation animals glucose Plasma IRI Plasma IRG

mg/lOt) ml jlU/till tiglml

Control (C) 35 128+7* 57.5+4.9 0.13±0.02Diabetic (D) 33 401±+ 13 22.6+7.1 0.34±0.05Insulin RB

diabetic(I-RX) 17 103±3 437±79 0.20±0.06

C vs. D P < 0.001 P < 0.005 P < 0.001D vs. I-RX P < 0.001 P < 0.001 P < 0.05C vs. I-RX NS P < 0.0(1 P < 0.05

* Mean+SEM.

was maximum in all three groups at 20°C and pH 7.5,whereas glucagon binding in all groups was maximum at30°C and pH 7.6. The presence of calcium in the incuba-tion medium decreased the glucagon binding in allthree groups. Insulin and glucagon binding plateauedat 30 min for all groups, hence, binding was measuredfor a 30-min period in all studies. Fig. 1 shows the com-parison between the specific binding of glucagon andinsulin by the hepatocytes obtained from controls, dia-betic, and insulin-treated diabetic rats. There was a40%decrease in glucagon binding by hepatocytes fromdiabetic rats as compared with controls (P < 0.001). Incontrast to glucagon binding, insulin binding was in-creased in diabetic rats (P < 0.05). Insulin treatment in-

18

z 16

0~~~-14-

12-

*s 10

88

6-

4- C 0 I-Rx C 0 I-Rx5 ~~~8) (8) (5) (6) (6) (4)

o 2

GLUCAGONBINDING INSULIN BINDING

FIGURE 1 Specific glucagon and insulin binding tohepatocytes from control (C), streptozotocin diabetic (D), anddiabetic rats treated with insulin (I-RB). The numbers inparentheses indicate number of rats studied. The variationshown is SEM. Binding is calculated as the percent ofadded tracer hormone bound by the hepatocytes. Specificbinding is obtained by subtracting nonspecific binding (i.e.binding in preseinee of 10 ,ug hormone) from total binding.

creased glucagon binding toward normal, resulting inbinding that was not significantly different from that inthe controls. Insulin binding decreased appreciably oninsulin treatment, however, it was still slightly, but notsignificantly, greater than that observed with controls.

Fig. 2A shows the Scatchard plot of glucagon bindingby controls, diabetic, and insulin-treated diabetic rathepatocytes. Both the total glucagon bound, as well asthe initial bound to free (B/F) ratio, were lower in dia-betic rats as compared with controls. On insulin treat-ment, the initial B/F ratio increased toward the valuesobtained with control animals, whereas total glucagonbound increased beyond the normal levels. Since theintercept on the abscissa is taken as a measure of thenumber of receptors, the decreased glucagon bindingin diabetic rats appears to be due to a decrease in thenumber of receptors. The Scatchard plot for insulin-treated diabetic rats is not parallel to that of normal oruntreated diabetics. It is steeper during the initial partof the curve, which suggests that there may be a changein the affinity of receptors. When the same data areplotted as percent of maximum specific binding againsttotal glucagon present in the incubation medium (Fig.2B), there is no significant difference between the threecurves. This indicates that despite the suggestive affinitychange in Fig. 2A, there is no significant difference inthe affinity of receptors among the three groups.

Similar analyses of insulin binding to the hepatocytesfrom the three groups are shown in Fig. 3A and B. Fig.3A shows that the increased insulin binding by hepato-cytes from diabetic rats was due to an increase in thenumber of receptors and that on insulin treatment, insulinbinding decreased due to a decrease in the number ofreceptors. As seen in Fig. 3B, there was no significantdifference in the affinity of hepatocytes for insulin incontrols, diabetic, or insulin-treated diabetic rats.

To rule out the possibility that decreased glucagonbinding in diabetic rats could have been due to factorssuch as increased glucagon degradation, the degrada-tion of both insulin and glucagon by hepatocytes wasexamined. There was no significant difference in theamount of insulin degraded by hepatocytes (Table II)obtained from controls, diabetic, or insulin-treateddiabetic rats. Degradation of glucagon by hepatocyteswas also not altered in diabetes. However, there wassignificantly greater degradation (P < 0.05) of glucagonby hepatocytes from insulin-treated diabetic rats whencompared with hepatocytes from either normal or un-treated diabetic rats. The significance of this observa-tion is not clear.

Since plasma IRG levels were higher in diabetic rats, itseemed pertinent to determine whether binding changescould have been due to differences in the amount ofglucagon endogenously bound to the cells. The livercells of controls and untreated diabetic rats were ex-tracted for glucagon (24) and IRG was measured. There

1490 S. J. Bhathena, N. R. Voylecs, S. Snmith, and L. Recant

B

.nn_Alu

9

0-aC*---x D

S--5I-RX

beODCc._

._

._

00.U,

x

0

0

\ \Uv*

-,. t.

, r S. - i

0 1 2 3Bound Glucagon (ng)

4 5

8

7

6

5

4

3

21

11

)o

to

io

io

0

10

Glucagon (ng/tube)

FIGuRE 2 (A) Scatchard analysis and (B) Competition curves of glucagon binding to hepato-cytes from control (C), diabetic (D), and insulin-treated diabetic rats (I-RX). Each curverepresents the mean of five separate experiments. In competition curves, the abscissarepresents total glucagon (tracer + unlabeled glucagon) per tube in a final volume of 0.2ml. Ordinate indicates specific binding at increasing concentrations of glucagon which are

calculated as the percent of that bound in the presence of tracer only, taken as 100%.

were no significant differences between control anddiabetic hepatocytes.

To determine whether the decreased binding of glu-cagon to diabetic rat hepatocytes was accompanied by a

A

0.20 r

0.18

0.16

0.14

0.12

0.10

0.08

0.06

0.04

0.02

corresponding decrease in biological activity of glu-cagon, measurements were made of cAMPformation bythe hepatocyte in response to glucagon. Since theo-phylline, a known blocker of cAMPphosphodiesterase

10

9

0~ ~ ~

y *~~~~~0---*D

\S \

N~~~~~~~~~~-

xi~~~~~,I "I

______I

0 0.5 1.0 1.5Bound Insulin (ng)

9

.scm10-c

co

0.(0

IU

0

8

7

6

5

4

3

2

11

2.0 2.5

B

MO_0)o

10 _ %

10 \

o 0 1.0 10 10213 14Insulin (ng/tube)

FIGURE 3 (A) Scatchard analysis and (B) Competition curves of insulin binding to hepatocytesfrom control (C), diabetic (D), and insulin-treated diabetic rats (I-R.). The curves for controland diabetic rats are the mean of four separate experiments and the curve for insulin-treated rats is the mean of three separate experiments. For details, refer to Fig. 2.

Glucagon Receptors in Diabetic Rats 1491

0.20r

0.181-0.16k

0.14k

0.12

m 0.10

0.08

0.06

0.04

0.02

A

TABLE IIEffect of Diabetes on Degradation of 125-Insulin

and 125I-Glucagon by Rat Hepatocytes*

No.Animal preparation animals "25I-Insulin 1251-Glucagon

Control (C) 4 1.14±0.731 7.42±0.60Diabetic (D) 4 0.40±0.04 6.06±1.86Insulin R. diabetic

(I-R,) 3 0.82±0.16 13.45±2.25C vs. D NS NSD vs. I-RX NS P < 0.05C vs. I-RX NS P < 0.05

* The percent degradation after a 30-min incubation in themedia described for binding studies of insulin and glucagon,respectively. Bacitracin was omitted.t Mean+SEM.

(25), was present in the incubation media, cAMPforma-tion presumably reflects the changes in the activity ofadenylate cyclase. Fig. 4 shows the cAMP responsesto glucagon of hepatocytes from control, diabetic, andinsulin-treated diabetic rats. Although glucagon bind-ing in hepatocytes from diabetic rats was decreased by40%, there was essentially no increased cAMPforma-tion in these hepatocytes in response to 10 ng of glu-

120

110 _

100 _

x 70_ '/90 - ;7

10 _ D

O~~~~~~~~~~~~~60 1- 678 91

G!ucagon (ng/tubes)

FIGURE 4 Basal- and glucagon-stimulated production ofcAMP by hepatocytes of control (C), diabetic (D), and in-sulin-treated diabetic (I-RX) rats. Each point is the mean offour separate cell preparations for control and diabetic ratsand of two separate preparations for insulin-treated diabeticrats. Each incubation tube had 3.75 x 106 cells in a final vol-ume of 2 ml.

cagon. The response to glucagon returned to normal inhepatocytes from insulin-treated diabetic rats.

Studies involving exogenous hyperglucagonemia.Table III shows blood sugar, plasma IRG, and plasmaIRI levels in control and in glucagon injected and in-fused rats. Since there were no differences in the bloodsugar, plasma IRG, and plasma IRI values between con-trol rats infused with saline for 2 h and control ratsinjected with saline for 7 days, the two sets of con-trols were combined into one control group. There wasno significant difference in blood sugar between thecontrol and glucagon-injected rats, however, in rats in-fused with glucagon, blood sugar was elevated signifi-cantly. Plasma IRG levels were highest in the rats re-ceiving acute infusions of glucagon, although pharma-cologic elevations were also observed in rats injectedwith glucagon. Since a single control and a single ex-perimental animal were studied simultaneously, allstatistical comparisons were done by paired analysis.

There was a significant decrease in glucagon bindingin rats injected with glucagon (Fig. 5) but not in rats in-fused with glucagon, though the latter group had muchhigher values of plasma IRG. Insulin binding was iden-tical in all groups. There was no significant differencein the degradation of either glucagon or insulin be-tween various groups (Table IV).

Scatchard analysis of glucagon binding (Fig. 6A) showsthat decreased binding in glucagon injected rats wasdue to decrease in the number of receptors. The com-petition curves (Fig. 6B) show no change in the ap-parent affinity of glucagon binding in glucagon injectedor infused rats. As seen in Fig. 7A and B, there was nochange in either the receptor number or affinity of in-sulin binding between various groups.

The sensitivity of hepatocytes to glucagon, as meas-

TABLE IIIEffect of Exogenous Glucagon on Blood Glucose, Plasma

Insulin, and Glucagon in Fed Rats

Animal No. Bloodpreparation animals glucose Plasma IRI Plasma IRG

mglOO ml 'UU/ml ng/ml

Control (C) 12 130±7* 65.5±8.2 0.16+0.02G-injected

(G-Inj.) 7 139±11 50.0±7.3 3.28±0.90G-infused

(G-Inf.) 5 164±17 58.2±14.3 > 4.801C vs. G-Inj. NS NS P < 0.005C vs. G-Inf. P < 0.05 NS I

* Mean±SEM.I All values were >4.80 ng/ml. Since samples were not furtherdiluted, the exact values, and therefore, SEMand P valuescannot be calculated. There was no overlap between thevalues in control and glucagon infused animals. Hence Pvalues appear significant.

1492 S. J. Bhathena, N. R. Voyles, S. Smith, and L. Recant

=- 188

a 16

"- 14C= 12

.' 10.5r 8o 6-0tou49 2

Y- Oa

+1

k-I1=

C G4nj G4nf-(11) (6) (5)

C G-lni(11) (6)

61nf.(5)

GLUCAGONBINDING INSUUN BINDINGFIGuRE 5 Specific glucagon and insulin binding of hepato-cytes of control (C), glucagon-injected (G-Inj.), and glucagon-infused (G-Inf.) rats. Saline-injected and saline-infused ratshave been combined into one control group. For details referto Fig. 1.

ured by the production of cAMP (Fig. 8), was sig-nificantly decreased in rats injected with glucagon, butnot in those infused with glucagon.

DISCUSSION

Studies of hormone receptors have been describedutilizing isolated cells as well as cell membranes. Inthis study, isolated hepatocytes were chosen because

A

0.18 F

G-() Control- - G-Inj.-- G-lnf.

mL.m1101

0.08

0.06-

0.041

0.021

TABLE IVEffect of Exogenous Glucagon on Degradation of 1251-Insulin

and 1251-Glucagon by Rat Hepatocytes*

No.Animal preparation animals "25I-Insulin '251-Glucagon

Control (C) 11 1.69+0.44t 7.70±1.86G-injected (G-Inj.) 6 2.06±0.85 12.36±4.35G-infused (G-Inf:) 5 1.87±0.40 6.82±0.68C vs. G-Inj. NS NSC vs. G-Inf: NS NS

* The percent degradation after a 30-min incubation in themedia described for binding studies of glucagon and insulinrespectively. Bacitracin was omitted.t Mean+SEM.

they provided reproducible data permitting correlationof the specific metabolic state of each animal with thecells prepared from that animal. It was also clear in de-veloping this test system, that collagenase treatmenthad no effect upon hormone binding, since cells pre-pared without collagenase provided similar findings interms of glucagon and insulin binding. It should benoted that we explored in great detail the preparationof hepatocyte cell membranes from normal and diabeticanimals. Such preparations were frequently nonstand-ard in the sense that the concentration factor for the5'-nucleotidase, a membrane-marker enzyme, varied

B

100

90

80tOD

X 70

=r 60

e 50E

.E 40x

30

20

10

Bound Glucagon (ng)0 1.0 10 102 103

Glucagon (ng/tube)

FIGURE 6 (A) Scatchard analysis and (B) competition curves of glucagon binding to hepato-cytes from control (C), glucagon-injected (G-Inj.), and glucagon-infused (G-Inf.) rats. Curvesfor control are the mean of 11 separate experiments. The curves for glucagon-injectedrats are the mean of six separate experiments, whereas the curves for glucagon-infusedrats are the mean of five separate experiments. For other details, refer to Fig. 2.

Glucagon Receptors in Diabetic Rats 1493

A

G-o Control*- --0 G-lnj.* G-Inf.

1.0 1.5Bound Insulin (ng)

._

.io

c

*X

x

0

2.0

LUU

90 N80 \\270 8

\ "%60 %

i50 \

40-

30 \

20-

10 \

Oil -Inu In I(tO 1.0 10 102 103 104

Insulin (ng/tube)FIGURE 7 (A) Scatchard analysis and (B) competition curves of insulin binding to hepatocytesfrom control (C), glucagon-injected (G-Inj.), and glucagon-infused (G-Inf.) rats. Rats used are

the same as in Fig. 7. For other details, refer to Fig. 2.

from 12- to 35-fold, and protein yields tended to behigher in diabetic livers. Consultation with Dr. Neville,Laboratory of Neurochemistry, National Institues ofHealth, revealed that the pathologic and metabolicchanges induced in the liver by diabetes, insulin, and(or) glucagon treatment could significantly influence

12

1'8 5c

4

x

in7

0

a.

:E

20

IA10 I30 Sh~~~~~~Gif. ,

D0 - \ -I

70 < ~~~~~~~Control50 -10~~~~~~~~~I

10 --

104

soI10 -*nI IG4n

v0 1 2 3 4 5 6 7 8 9 10Glucagon (ng/tube)

FIGURE 8 Basal- and glucagon-stimulated production ofcAMP by hepatocytes from control (C), glucagon-injected(G-Inj.), and glucagon-infused (G-Inf) rats. Each point is themean of 18 separate cell preparations for control, 4 prepara-tions for glucagon-injected rats, and 5 preparations for gluca-gon-infused rats. Each incubation tube had 3.75 x 106 cellsin a final volume of 2 ml.

the fractionation procedures required to obtain mem-

branes of high quality. One example of this is found inthe observation that in all membranes from diabeticanimals, the 5'-nucleotidase is reduced to 50% of con-

trols. Further, in our hands, insulin treatment restoredthe nucleotidase concentration towards control levels.Details of these studies will be published separately.As a result of these problems in the use of membranes,our studies were conducted with isolated hepatocytes.

In this study evidence is presented indicating thatglucagon binding was significantly decreased in hepato-cytes obtained from streptozotocin diabetic rats andthat with insulin treatment, this binding returned towards normal. Insulin binding, on the other hand, was

increased in hepatocytes from untreated diabetics anddecreased with insulin treatment. This latter observa-tion is wholly consistent with the data of Davidson andKaplan (8) obtained with liver plasma membranes fromstreptozotocin diabetic rats. However, these authorsfailed to observe any change in glucagon binding. Thisfailure may be due to the fact that there is wide varia-tion in hormone binding by membrane preparations bothfrom individual animals and groups (26). Davidson andKaplan (8) examined only one membrane preparationfrom diabetic rats for glucagon binding.

In an effort to understand why glucagon binding was

decreased and insulin binding was increased in experi-mental diabetes, it seemed reasonable to consider thefollowing possibilities: (a) a change in the number ofhormone receptors without any accompanying change in

1494 S. J. Bhathena, N. R. Voyles, S. Smith, and L. Recant

0.10

0.09

0.08

0.07

0.06U-s 0.05

0.04

0.03

0.02

0.01

B

I &rW%I

affinity for the hormone; (b) a change in the affinity ofthe receptors with no corresponding change in theirnumber; (c) a change in the number and affinity of thereceptors; or (d) a change in the amount of degrada-tion of either the hormone or the receptor. These pos-sibilities were approached by subjecting the data to aScatchard analysis and by plotting competition curves(percent maximum specific binding vs. total hormoneconcentration) to obtain an estimate of the apparentaffinity of receptor for hormone. The Scatchard plotsfor glucagon binding in all animal groups were cur-vilinear (Fig. 2A), suggesting the existence of negativecooperativity as described by DeMeyts et al. (27, 28) forinsulin. An additional possible interpretation wouldbe the presence of heterogeneity of receptors or a com-bination of negative cooperativity and heterogeneity.Specific differentiations of these latter possibilities couldnot be easily obtained. Using these analyses, we foundthat the decrease in glucagon binding in untreated strep-tozotocin diabetes was most likely due to a decrease innumber of glucagon receptors. The increase in insulinbinding was also due to a change in the number ofreceptors rather than receptor affinity. This interpreta-tion of' the insulin data is in agreement with that ofDavidson and Kaplan (8). After insulin treatment ofdiabetic animals, glucagon binding was increased to-wards normal and insulin binding was decreased tonormal. Analysis of the glucagon-binding competitioncurves for cells from insulin-treated diabetic rats (Fig.2B) revealed no significant changes in affinity, but onlyan increase in receptor number (Fig. 2A).

The effect of insulin treatment on insulin bindingalso was related to a change in receptor number, namelya decrease, but not a change in affinity (Fig. 3A, B). Sim-ilar findings have been reported by Davidson and Kap-lan (8) in liver plasma membranes of insulin-treateddiabetic rats and by Kobayashi and Olefsky (29) in adi-pocytes of insulin-treated normal rats.

Since the plasma levels of IRG in diabetic rats weremore than twice the levels in control rats, the possibilitywas raised that the hepatocyte receptors in diabetics,relative to receptors from animals with lower plasmaIRG levels, might be partially saturated with glucagon.Our studies of IRG levels extracted from hepatocytesof normal and diabetic animals revealed no differences,making this an unlikely explanation for the reducedglucagon binding in diabetics. With regard to the de-creased number of insulin receptors after insulin treat-ment, the data from several laboratories (30-33) indicatethe ready dissociation of insulin from cells and mem-branes, thus eliminating the possibility of an endogenoussaturation of receptors.

Another factor to be considered in the analysis ofaltered binding relates to degradation of the respectivehormone. Ouir data provided evidence that no alterationin glucagon degradation occurred in untreated diabetic

rat hepatocytes as compared to those of normal rats.Insulin degradation was unaltered in hepatocytes fromeither untreated or insulin-treated diabetic rats. David-son and Kaplan (8) made similar observations on insulindegradation by plasma membranes of control and un-treated diabetic rats. Our data did show increased glu-cagon degradation in cells from insulin-treated diabeticrats. At the present time, we have no explanation forthis phenomenon, though it deserves further explora-tion. The fact that there was no correlation betweenbinding of glucagon and its degradation suggests thatthe two processes are independent of each other.

Having accepted that the number of glucagon re-ceptors was decreased in the diabetic state and normal-ized with insulin treatment, one must raise the questionsof "why" and "how" this occurs. The theory of receptor"down regulation" namely decreased hormone bind-ing by specific receptors in response to elevation ofhormone levels, has been well documented for insulin(8, 20, 34, 35). Our observations with regard to glucagonbinding in diabetics were consistent with a similar reg-ulatory process in that high concentrations of plasmaIRG were associated with decreased numbers of' re-ceptors, and insulin treatment of these diabetic ratswas accompanied by decreased plasma IRG (P < 0.05)and increased numbers of glucagon receptors.

To test this hypothesis of glucagon receptor regulationby hormone levels in a more direct manner, we studiedthe effect of glucagon injection and infusion on re-ceptor binding. It was clear that short-term elevationof plasma glucagon (2 h of hyperglucagonemia as seenin our glucagon-infused rats) did not decrease hepato-cyte glucagon binding. On the other hand. prolongedhyperglucagonemia (normnal rats injected with glucagonfor 1 wk) decreased the glucagon receptor number.A similar time requirement for the effects of hyper-insulinism on specific receptors has been shown byKahn et al. (36) and Gavin et al. (34).

Our studies with exogenous hyperglucagonemiastrongly support, but do not absolutely prove, thehypothesis that the decrease in the receptor numberis due to the regulation of glucagon receptors bythe elevated plasma-glucagon level. Other factors inaddition to hormone concentration and duration ofexposure of the receptor to increased levels of hormonemay influence or regulate binding of hormone. At leastthree possible factors come to mind: (a) The enhancedeffect of endogenous (diabetes) vs. exogenous hyper-glucagonemia (injected) in decreasing receptor num-ber may be due to direct delivery of the forrmerto hepatic cell receptors; (b) Insulin, other than inits effect on decreasing glucagon levels, may influencereceptors in as yet undefined "anabolic" fashion; and(c) Differential effects of various glucagon moieties indecreasing receptor number may depend upon themolecular species which is increased. In diabetic

Glucagon Receptors in Diabetic Rats 1495

plasma and with exogenous glucagon, the increasedmolecular species is the 3,500-mol wt pancreatictype glucagon. This molecule is clearly biologicallyactive. In chronic renal disease, the major componentin plasma is a 9,000-mol wt molecule, probably pro-glucagon (37, 38). This component may not only bebiologically inactive, but also may not be able to bindto the glucagon receptor. Such an effect may explainthe failure to observe a decrease in glucagon receptorsin the hyperglucagonemia of experimental kidneydisease (39). Obviously much work remains to be doneto clarify this area of glucagon receptor regulation.

What are the effects of decreased numbers of gluca-gon receptors in the diabetic and in glucagon injectedrats? Weobserved that cAMPresponse to glucagon wasmarkedly decreased in glucagon-injected and indiabetic rat hepatocytes. Treatment of the diabeticanimals with insulin restored this response to normal.In general, these biological effects correlated with thealterations in glucagon binding. In the diabetic rat,glucagon binding was reduced by 40% and the cAMPresponse to glucagon was essentially abolished. Simi-larly, in the glucagon-injected rat, binding was de-creased by 30% and glucagon-induced cAMP forma-tion was decreased by >70%.

How can one account for the quantitative dis-crepancy in glucagon binding vs. cAMP formation?A number of hypotheses may be suggested: (a)Specific glucagon receptors may be heterogeneous;20%may be true receptors capable of activating cAMPformation, whereas 80% may more properly be calledacceptors in that they bind hormone but cannot activatethe adenylate cyclase, and that the decrease in thenumber of receptors may be more pronounced in thesmall active receptor population rather than in theinactive acceptor population. (b) Problems in the trans-fer of membrane receptor signals to the enzyme mayexist (40). (c) Intracellular alterations such as ATPdeficiency, which may interfere with enzyme action,could occur. Such an alteration might readily be re-stored by insulin.

It is noteworthy that basal hepatic levels of cAMPhave been reported to be elevated in diabetes (41-43)and in starvation (41, 44-46). Fouchereau-Peron et al.(47) reported that glucagon binding and cAMPresponse to glucagon were reduced in starvation, asituation somewhat analagous to diabetes. Thus, themechanisms by which starved and diabetic liversdevelop elevations of cAMP would appear to beglucagon independent.

In conclusion, endogenous hyperglucagonemia inuncontrolled diabetes, as well as prolonged exogenoushyperglucagonemia, are associated with a decrease inthe number of glucagon receptors and impaired cAMPresponse to glucagon. If one were to reason teleologi-cally, the decrease in glucagon receptors and the

marked decrease in the biological effect of glucagon indiabetes could provide a mechanism for reducing theexcessive gluconeogenesis and hyperglycemia as-sociated with the diabetic state.

ACKNOWLEDGMENTS

This research was supported by the Veterans Administrationand National Institute of Health grant RO1 AM 19610 fromthe National Institute of Arthritis, Metabolism and DigestiveDiseases.

REFERENCES

1. Perley, M., and D. M. Kipnis. 1966. Plasma insulinresponses to glucose and tolbutamide of normal weightand obese diabetics and nondiabetic subjects. Diabetes.15: 867-874.

2. Parker, M. L., R. S. Pildes, K. L. Chao, M. Cornblath,and D. M. Kipnis. 1968. Juvenile diabetes mellitus, adeficiency in insulin. Diabetes. 17: 27-32.

3. Unger, R. H., E. Aguilar-Parada, W. A. Muller, and A. M.Eisentraut. 1970. Studies of pancreatic alpha cell func-tion in normal and diabetic subjects. J. Clin. Invest.49: 837-848.

4. Muller, W. A., G. R. Faloona, and R. H. Unger. 1973.Hyperglucagonemia in diabetic ketoacidosis, its preva-lence and significance. Am. J. Med. 54: 52-57.

5. Unger, R. H., and L. Orci. 1976. Physiology and patho-physiology of glucagon. Physiol. Rev. 56: 778-826.

6. Braaten, J. T., G. R. Faloona, and R. H. Unger. 1974.The effect of insulin on the alpha cell response tohyperglycemia in long standing alloxan diabetes.J. Clin.Invest. 53: 1017-1021.

7. Raskin, P., I. Aydin, and R. H. Unger. 1976. The effectof insulin on the exaggerated glucagon response toarginine stimulation in diabetes mellitus. Diabetes. 25:227-229.

8. Davidson, M. B., and S. A. Kaplan. 1977. Increased insu-lin binding by hepatic plasma membranes from diabeticrats. Normalization by insulin therapy. J. Clin. Invest.59: 22-30.

9. Hepp, K. D., J. Langley, M. J. VonFuncke, R. Renner, andW. Kemmler. 1975. Increased insulin binding capacity ofliver membranes from diabetic chinese hamsters. Nature(Lond.). 258: 154.

10. Olefsky, J. M., and G. M. Reaven. 1974. Decreasedinsulin binding to lymphocytes from diabetic patients.

J. Clin. Invest. 54: 1323-1328.11. Goldstein, S., M. Blecher, R. Binder, P. V. Perrino, and

L. Recant. 1975. Hormone Receptors 5. Binding of gluca-gon and insulin to human circulating mononuclear cellsin diabetes mellitus. Endocrinol. Res. Commun. 2: 367-376.

12. Flier, J. S., C. R. Kahn, J. Roth, and R. S. Bar. 1975.Antibodies that impair insulin receptor binding in anunusual diabetic syndrome with severe insulin resistance.Science (Wash. D. C.). 190: 63-65.

13. Morgan, C. R., and A. Lazarow. 1963. Immunoassay ofinsulin: two antibody system. Diabetes. 12: 115-126.

14. Unger, R. H., A. M. Eisentraut, and M. S. McCall. 1959.Glucagon antibodies and immunoassay for glucagon.Proc. Soc. Exp. Biol. Med. 120: 621-623.

15. Zahlten, R. N., and F. W. Stratman. 1943. The isolationof hormone-sensitive rat hepatocytes by a modifiedenzymatic technique. Arch. Biochem. Biophys. 163:600-608.

1496 S. J. Bhathena, N. R. Voyles, S. Smith, and L. Recant

16. Crane, L. J., and D. L. Miller. 1974. Synthesis andsecretion of fibrinogen and albumin by isolated rathepatocytes. Biochem. Biophys. Res. Commun. 60:1269-1277.

17. Hunter, W. M., and F. C. Greenwood. 1962. Preparationof Iodine-131 labelled human growth hormone of highspecific activity. Nature (Lond.). 194: 495-496.

18. Giorgio, N. A., C. B. Johnson, and M. Blecher. 1974. Hor-mone Receptors III. Properties of glucagon-binding pro-teins isolated from liver plasma membranes. J. Biol.Chem. 249: 428-437.

19. Freychet, P., J. Roth, and D. M. Neville, Jr. 1971. Mono-iodo insulin: Demonstration of its biological activity andbinding to fat cells and liver membranes. Biochem.Biophys. Res. Commun. 43: 400-408.

20. Freychet, P. 1976. Interactions of polypeptide hormoneswith cell membrane specific receptors: studies with in-sulin and glucagon. Diabetologia. 12: 83-100.

21. Recant, L., P. V. Perrino, S. J. Bhathena, D. N. Dan-forth, and R. Lavine. 1976. Plasma immunoreactiveglucagon fractions in four cases of glucagonoma: In-creased "Large glucagon-immunoreactivity". Diabeto-logia. 12: 319-326.

22. Scatchard, G. 1949. The attractions of proteins for smallmolecules and ions. Ann. N. Y. Acad. Sci. 51: 660-672.

23. Bhathena, S. J., S. S. Smith, N. R. Voyles, J. C.Penhos, and L. Recant. 1977. Studies on submaxillarygland immunoreactive glucagon. Biochem. Biophys. Res.Commun. 74: 1574-1581.

24. Davoren, P. R. 1962. The isolation of insulin from a singlecat pancreas. Biochim. Biophys. Acta. 63: 150-153.

25. Butcher, R. W., and E. W. Sutherland. 1962. Adenosine3',5'-phosphate in biological material. 1. Purification andproperties of cyclic 3',5'-nucleotide phosphodiesteraseand use of the enzyme to characterize adenosine 3',5'-phosphate in human urine.J. Biol. Chem. 237:1244-1250.

26. Sorrentino, R. N., and J. R. Florini. 1976. Variationsamong individual mice in binding of growth hormone andinsulin to membranes from animals of different ages. Exp.Aging Res. 2: 191-205.

27. Demeyts, P., J. Roth, D. M. Neville, Jr., J. R. Gavin,III, and M. A. Lesniak. 1973. Insulin interactions withits receptors: experimental evidence for negative co-operativity. Biochem. Biophys. Res. Commun. 55: 154-161.

28. DeMeyts, P., A. R. Bianco, and J. Roth. 1976. Site-site interactions among insulin receptors, characterizationof negative cooperativity. J. Biol. Chem. 251: 1877-88.

29. Kobayashi, M., and J. M. Olefsky. 1977. Effect of experi-mental hyperinsulinemia on insulin binding and glucosetransport in isolated adipocytes. Clin. Res. 25: 394A.(Abstr.)

30. Cuatrecasas, P., B. Desbuquois, and F. Krug. 1941. In-sulin-receptor interactions in liver cell membranes. Bio-chem. Biophys. Res. Commun. 44: 333-339.

31. Cuatrecasas, P. 1941. Properties of the insulin receptor ofisolated fat cell membranes. J. Biol. Chem. 246:7265-72 704.

32. Kahn, C. R., P. Freychet, and J. Roth. 1974. Quantitative

aspects of the insulin-receptor interaction in liver plasmamembranes.J. Biol. Chem. 249: 2249-2257.

33. Ginsberg, B. H., C. R. Kahn, J. Roth, and P. DeMeyts.1976. Insulin-induced dissociation of its receptor intosubunits: possible molecular concomitant of negativecooperativity. Biochem. Biophys. Res. Commun. 73:1068-1074.

34. Gavin, J. R., III, J. Roth, D. M. Neville, Jr., P. DeMeyts,and D. N. Buell. 1974. Insulin-dependent regulation ofinsulin receptor concentrations: a direct demonstration incell culture. Proc. Natl. Acad. Sci. U. S. A. 71: 84-88.

35. Olefsky, J. M. 1976. Decreased insulin binding to adipo-cytes and circulating monocytes from obese subjects. J.Clin. Invest. 57: 1165-1172.

36. Kahn, C. R., D. M. Neville, Jr., and J. Roth. 1973.Insulin-receptor interaction in the obese hyperglycemicmouse. A model of insulin resistance. J. Biol. Chem.248: 244-250.

37. Valverde, I., and M. L. Villanueva. 1976. Heterogeneityof plasma immunoreactive glucagon. Metab. Clin. Exlp.25: 1393-1395.

38. Kuku, S. F., A. Zeidler, D. S. Emmanouel, A. I. Katz, A. H.Rubenstein, N. W. Levin, and A. Tello. 1976. Hetero-geneity of plasma glucagon: patterns in patients withchronic renal failure and diabetes. J. Clin. Endocrinol.Metab. 42: 173-176.

39. Soman, P. V., and P. Felig. 1977. Altered glucagon andinsulin binding to hepatic receptors: cellular mechanismof glucose intolerance in uremia. Clin. Res. 25:401A. (Abstr.)

40. Pohl, S. L. 1977. The glucagon receptor and its relation-ship to adenylate cyclase. Fed. Proc. 36: 2115-2118.

41. Jefferson, L. S., J. H. Exton, R. W. Butcher, E. W.Sutherland, and C. R. Park. 1968. Role of adenosine 3',5-monophosphate in the effects of' insulin and anti-insulin serum on liver metabolism. J. Biol. Chem. 243:1031- 1038.

42. Goldberg, N. D., S. B. Dietz, and A. G. O'Toole. 1969.Cyclic guanosine 3',5'-monophosphate in mammaliantissues and urine. J. Biol. Chem. 244: 4458-4466.

43. Pilkis, S. J., J. H. Exton, R. A. Johnson, and C. R. Park.1974. Effects of glucagon on cyclic AMPand carbohydratemetabolism in liver from diabetic rats. Biochimn.Biophys. Acta. 343: 250-267.

44. Exton, J. H., G. A. Robison, E. W. Sutherland, andC. R. Park. 1971. Studies on the role of adenosine 3',5'-monophosphate in the hepatic actions of glucagon andcatechalamines.J. Biol. Chem. 246: 6166-6177.

45. Selawry, H., R. Gutman, G. Fink, and L. Recant. 1973.The effect of starvation on tissue adenosine 3',5'-monophosphate levels. Biochem. Biophys. Res. Commun.51: 198-204.

46. Garrison, J. C., and R. C. Haynes, Jr. 1973. Hormonalcontrol of glucogenolysis and gluconeogenesis in iso-lated rat liver cells. J. Biol. Chem. 248: 5333-5343.

47. Fouchereau-Peron, M., F. Rancon, P. Freychet, and G.Rosselin. 1976. Effect of feeding and fasting on the earlysteps of glucagon action in isolated rat liver cells.Endocrinology. 98: 755-760.

Glucagon Receptors in Diabetic Rats 1497