Multiple defects in muscle glycogen synthase activity contribute to reduced glycogen synthesis in non insulin dependent diabetes mellitus

Multiple defects in muscle glycogen synthase

activity contribute to reduced glycogen

synthesis in non-insulin dependent diabetes

mellitus.

A W Thorburn, … , G Brechtel, R R Henry

J Clin Invest.

1991;

87(2)

:489-495.

https://doi.org/10.1172/JCI115022

.

To define the mechanisms of impaired muscle glycogen synthase and reduced glycogen

formation in non-insulin dependent diabetes mellitus (NIDDM), glycogen synthase activity

was kinetically analyzed during the basal state and three glucose clamp studies (insulin

approximately equal to 300, 700, and 33,400 pmol/liter) in eight matched nonobese NIDDM

and eight control subjects. Muscle glycogen content was measured in the basal state and

following clamps at insulin levels of 33,400 pmol/liter. NIDDM subjects had glucose uptake

matched to controls in each clamp by raising serum glucose to 15-20 mmol/liter. The insulin

concentration required to half-maximally activate glycogen synthase (ED50) was

approximately fourfold greater for NIDDM than control subjects (1,004 264 vs. 257

+/-110 pmol/liter, P less than 0.02) but the maximal insulin effect was similar. Total glycogen

synthase activity was reduced approximately 38% and glycogen content was approximately

30% lower in NIDDM. A positive correlation was present between glycogen content and

glycogen synthase activity (r = 0.51, P less than 0.01). In summary, defects in muscle

glycogen synthase activity and reduced glycogen content are present in NIDDM. NIDDM

subjects also have less total glycogen synthase activity consistent with reduced functional

mass of the enzyme. These findings and the correlation between glycogen synthase activity

and glycogen content support the theory that multiple defects in glycogen synthase activity

combine to cause […]

Research Article

Find the latest version:

Multiple Defects in Muscle

Glycogen Synthase Activity Contribute

to

Reduced Glycogen Synthesis

in

Non-insulin

Dependent

Diabetes Mellitus

AnneW.Thorbum, Barry Gumbiner,FredBulacan, Ginger Brechtel,and RobertR.Henry Department ofMedicine, University of California, San Diego;and the MetabolismDivision, VeteransAdministration Medical Center, San Diego, California 92161

Abstract

To define the mechanisms ofimpaired muscle glycogen syn-thase andreducedglycogenformationinnon-insulindependent

diabetes mellitus (NIDDM), glycogen synthase activity was kinetically analyzed during the basal state and three glucose

clamp studies (insulin _

300,

700, and33,400 pmol/liter)in

eight matched nonobese NIDDM and eightcontrolsubjects.

Muscle glycogen content was measured in the basal state and

following clamps at insulin levels of 33,400 pmol/liter.

NIDDM subjects had glucose uptake matched to controls in each clamp byraisingserum glucose to15-20 mmol/liter.

The insulin concentrationrequiredtohalf-maximally acti-vate glycogen synthase

_(ED")

wasapproximately fourfold greater for NIDDM than control subjects (1,004±264 vs. 257±110 pmol/liter,P<0.02)butthe maximal insulineffect

was similar. Total glycogen synthase activity was reduced - 38%andglycogencontent was - 30%lower in NIDDM. A positivecorrelationwaspresent betweenglycogencontent and glycogen synthase activity (r=0.51, P <0.01).

In summary, defects in muscle glycogen synthaseactivity

and reduced glycogen content are present in NIDDM. NIDDM

subjectsalsohave less totalglycogensynthaseactivity consis-tentwith reduced functionalmassofthe enzyme.These find-ingsand the correlation between glycogen synthase activity and glycogen content support the theory that multiple defects in glycogen synthase activity combine to cause reduced glycogen

formationin NIDDM.(J. Clin. Invest. 1991. 87:489-495.)

Key words:insulin action-glucosedisposal-nonoxidative

glu-cosemetabolism * enzyme kinetics *

glucose-6-phosphate

Introduction

Glycogen synthesis isamajor pathway ofglucosedisposalin skeletal muscle (1). Changes in the activity of glycogen syn-thase(EC2.4.1.11), either by allosteric regulatorsor covalent

modification,regulate thesynthesis ofglycogen. In muscle, gly-cogensynthase existsmainly in active dephosphorylated forms

orlessactive phosphorylated forms (2).Theseformsare

inter-converted byprotein kinase and phosphatase reactions with glucose-6-phosphate

(G6P)'

stimulatingthe phosphatase

reac-Addressreprintrequests toDr.RobertR.Henry, Veterans

Administra-tion Medical Center

(V-l

IIG), 3350LaJollaVillage Drive,SanDiego, CA 92161.

Receivedforpublication 27 April 1990 and in revisedform 27 Sep-tember 1990.

1.Abbreviationsused in this paper: FFM, fat-free mass;G6P,

glucose-6-phosphate; NIDDM, non-insulin dependent diabetes mellitus; UDPG, uridine5'-diphosphateglucose.

The Journal of Clinical Investigation,Inc.

Volume87, February 1991, 489-495

tion (3-5), while insulin is thought to act by decreasing the kinase(6)aswellasincreasingthephosphatase reactions

(7).

Arecentstudy using

"3C-nuclear

magnetic resonance spec-troscopyhasshown that muscle glycogen formation is reduced by 57% at an insulin concentration of400 pmol/liter in non-in-sulindependent diabetes mellitus (NIDDM) (1). This decrease inglycogen synthesis, however, could be the result ofimpaired glycogensynthase activity,reduced muscleglucoseuptake, ora

combination of these factors. This study was carried out to address whether defective glycogen synthase activity alone plays a major role in impaired glycogen formation in NIDDM. Most current evidenceindicates that diabetes is associated with reduced glycogen synthaseactivity.Inanimals,diabetes induced by streptozotocin, alloxan, or pancreatectomy has been shown to blunt the ability of insulin to promote glucose incorporation into glycogen (8, 9) as well as the ability of insu-lin tolower thesensitivityof the enzyme for G6P(10, 11). A number of recent studies conducted in human subjects have also documented reduced glycogen synthase activity in

insulin-resistant states including NIDDM (7, 12-17). However, the natureand extentofthe glycogensynthasedefectin muscle and

itsrole in reduced glycogenformationin NIDDM has not yet been clearly elucidated. Bogardusand co-workers have

re-ported correlations betweenmuscleglycogensynthaseactivity andinsulin-mediatedglucose uptakein subjectswith varying insulin sensitivity,implyingthat glycogensynthase activityis

impaired in NIDDM, although thedefects were not quanti-tatedat ratesof tissueglucose uptake comparabletonormal subjects(7, 12, 13). In subjects with frank NIDDM, Wright et al.calculatedmuscleglycogensynthaseactivityto be25%less thancontrolsubjects after ingestingameal(14). Althoughthis

studyassessedglycogen synthase activityin response to a physi-ologicstimulus, the fact that it wasconductedunder

non-steady-state conditionswithvarying glucose andinsulin con-centrations did not allow the role of these parameters in the

glycogen synthasedefect tobeascertained. In a recent study thatnormalized glucoseuptake in NIDDMusing hyperglyce-mia during hyperinsulinemic (- 300 pmol/liter) glucose

clamps, reduced muscle glycogen synthase activity (as mea-suredby fractional velocity)wasdocumentedinNIDDM

com-paredwithnondiabetic subjectsundersteady-state conditions

(17).Increasingtheinsulinlevelfourfoldovercame this defect.

Therefore,althoughprevious studies have shown NIDDM to beassociated with impairedmuscle glycogen synthase activity andreduced glycogen formation, little informationis currently

availablethatcharacterizesthe extentofthis enzymaticdefect

over arange of insulinconcentrations ordefines the

mecha-nism for this defectand its potential role in decreased glycogen

formationin NIDDM.

Toaddress thesequestionsinthisstudy,

thekinetics

ofboth

insulin and G6P activation ofglycogen synthase activity, as

diabetes mellitusand nondiabetic (control) subjects over a range of insulin concentrations from basal to 33,400 pmol/

liter. To examineanydefectsateachinsulin level independent

of reduced glucose uptake, studieswere performedat

equiva-lent rates of glucose uptake inthe NIDDM and control sub-jects.

Methods

Subjects. Eight male subjects with NIDDM and eight male nondiabetic (control) subjects with normal glucose tolerance (18) were studied. Ta-bleIlists the clinical characteristics of the subjects. Age, weight, fat-free mass (FFM), and body mass index were not significantly different in the NIDDM and control subjects. The NIDDM subjects did not have significant diabetic complications or hypertension. Five subjects were

onsulfonylurea therapy, one was on insulin, one was on diet therapy only, and one had been recently diagnosed and not yet treated. All treatment for diabetes was withdrawn at least two weeks before the studies.Nosubject was taking any other medication known to affect carbohydrate metabolism. Duration of overt diabetes was 10±3 yr (mean±SEM). Fasting glycosylated hemoglobin in the NIDDM sub-jects was 9.4±12%. The normal reference range for this assay is 4.1-8.1%.The control subjects were screened to ensure they were healthy and hadnofamily history of diabetes. Some of the data derived from these studies has been reported previously (17, 20, 21). Written in-formed consent was obtained from each subject and the experimental protocol approved by the Committee on Human Investigation of the

UniversityofCalifornia, San Diego.

Study protocol. Subjects were admitted to the Special Diagnostic

and TreatmentUnit of the Veterans Administration Medical Center in SanDiego and consumed a weight-maintenance standardized solid diet containing 55% of calories as carbohydrate, 30% as fat, and 15% as protein during hospitalization.

Three studies were carried out in both the NIDDM and control subjects. Each studywasperformed after a12-14-hovernight fast and lasted - 7.5 h: a 2.5-h basal period followed by a 5-h hyperinsulinemic

glucose clamp. Basal serum glucose and insulin concentrations were higher in the NIDDM than in control subjects (P<0.05, Table II). The

serumglucose and insulin levels used during the clamps are also listed in TableII. Clamps A, B, and C were conducted at insulin infusion ratesof 150, 300, and 4,000pmol/m2permin, respectively (equivalent

to - 20,40,and600mU/m2permin). ClampsAandB werecarried

out ateuglycemia (5 mmol/liter) in the control subjects andat hypergly-cemia (15-20 mmol/liter) in theNIDDMsubjects to match glucose uptakeratesin theNIDDMgrouptothoseofthe control group. Clamp C was carriedoutfirst in theNIDDMsubjectsatpharmacologic insulin concentrations andhyperglycemia(15-20 mmol/liter)to maximally stimulate glucose uptake. Inthe control subjects,clamp Cwasthen conductedatthesameinsulin levelastheNIDDMgroup and glucose uptakeratewasmatchedtothe NIDDMsubjects byvaryingthe level of hyperglycemia.

Details of the studyprocedureshavebeen describedpreviously (17,

22). Briefly,

3-(3H]glucose

was infusedasa

45-,gCi

bolus dose followed

TableI. Clinical CharacteristicsoftheSubjects*

NIDDMsubjects Controlsubjects

Age(yr) 59±3 58±5 Weight(kg) 78.8±5.0 74.2±4.0 Fat-free mass(kg)$ 60.5±3.8 56.4±2.7 Bodymassindex

(kg/n2)

25.4±1.1 24.4±1.0 * Data expressedas

mean±SEM.

$ Determined by underwaterweighingwith correlationfor residual lung volume(19).

TableII.Serum Glucose and InsulinConcentrationsandRates

ofGlucose Uptake in the Control andNIDDMSubjects Serum glucose Serum insulin Glucoseuptake

mmol/liter pmol/liter mg/kgFFM permin

Basal

NIDDM 9.9±1.1* 64±+14 3.75±0.23§

Control 4.9±0.1 36±7 2.49±0.10 ClampA

NIDDM 20.7±1.6§ 258±57 8.62±0.49 Control 5.1±0.1 287±43 8.25±0.48 ClampB

NIDDM 17.6±2.2§ 646±65 11.24±0.34 Control 5.1±0.1 740±72 11.76±0.75 Clamp C

NIDDM 18.4±1.2 33136±2369 25.58±1.66 Control 15.3±2.2 34694±5873 27.21±1.28 tP <0.05,* P <0.01, and §P<0.001 for NIDDMvs.control values.

byacontinuous infusion of 0.60MCi/minduring the entire study to isotopically determine rates of hepatic glucose output and glucose ap-pearance (23, 24). Glucose uptake rates (in milligrams per kilogram

FFMper minute) were calculated during the basal state in each subject from the rate of glucose appearance corrected for changes in glucose pool size and urinary glucose loss, if present. Glucose uptakerateswere

calculatedduring each clamp study in each subject from the glucose infusionratecorrected for changes in glucose pool size, urinary glucose loss, and residual HGO. The clampwasstarted withan intravenous infusion of insulin (crystalline human insulin [Humulin R], kindly

suppliedby Eli Lilly & Co., Indianapolis, IN) given asabolus followed byacontinuousrateof insulin infusion. Serum glucose was clamped at thedesired concentration by varying the infusion of 20% glucose. So-matostatin (0.08 pmol/kg per min, cyclic form; Bachem Inc., Torrance, CA) was infused in all studies to suppress endogenous insu-linsecretion. KC1 and K2PO4 were infused at a rate of 0.16 mmol/liter

tomaintainserumpotassiumlevels.Duringthe last30 min ofthe basal andclampperiods,steady-statemeasurements weremadeofratesof glucose uptake,serumglucose, andseruminsulin. Glycogen synthase activity in samples ofvastuslateralis musclewasdetermined four times in eachsubject;atthe endofoneofthe basalperiods andatthe comple-tion of each of the three clamp studies. Glycogencontent wasalso measured in muscle collected frombasaland clampC studiesto deter-mine whether glycogen synthase correlated with the glycogencontent

in muscle from the NIDDM and control subjects in these studies. Gly-cogenwas notmeasured inclampsAandBsince theexpectedincrease in glycogen wouldlikelybewithin thecoefficient of variation of the measurement for muscle glycogen from biopsy samples(25).

Glycogen synthasedetermination. Percutaneous musclebiopsies

wereobtained from thevastuslateralis muscle witha5-mm diameter

side-cuttingneedleusingamodification of theproceduredescribedby

Bergstrom (26). Muscle sampleswereblottedto removeanyblood,

immediately placed intoliquid nitrogen,and storeduntilassayed.

Gly-cogensynthase (EC 2.4.1.11)wasmeasuredusingmodifications of the methodsof Nuttallet al.(27)andThomasetal.(28). This modified method has been described in detailpreviously(14, 17, 29). Glycogen synthaseactivitywasassayed inahomogenate of 50 mg of muscleby

measuring the incorporation of['4C]glucose fromuridine 5'-diphos-phate glucose(UDPG)into glycogen. Proteinwasalsoassayed in the

extract(30) and units of glycogen synthaseactivityexpressedas

nano-molesof

['4C]UDPG

incorporated into glycogen per minute per milli-gram ofprotein.Glycogen synthase activitywasassayedover arange of glucose-6-phosphate concentrations (0, 0.1, 0.3, 0.5,1.0,5.0, and 10.0 mmol/liter G6P) at physiologic levels of substrate (0.3 mmol/liter

UDPG). Resultsare expressed asthe fractional velocity ofglycogen

synthase,i.e., theactivity assayed at 0.1 mmol/liter G6P divided by the activity at 10mmol/liter G6P. Inaddition,glycogensynthaseactivity was measured in eachbiopsyspecimenatsaturating concentrations of UDPG (5mmol/liter)andG6P (10mmol/liter)to assessthe total activ-ityof the enzyme.

Muscleglycogenconcentration. Theglycogencontentofmusclewas

determined usingamodification of the method usedby Kepplerand Decker(31). Before analysis, muscle sampleswerefreeze-dried and powderedto removeconnective tissue and any dilution factor dueto

the presenceof blood.Aportionof dried muscle(8mg)was homoge-nizedwith 80 parts byweight of 0.6 Mperchloricacid,and 40

_,1

of the muscle homogenatewasadded to 20

,d

KHCO3solution(I mol/liter),

385,lacetatebuffer (pH4.8)plus 14.8 Mlglucoamylasesolution(270

U/ml;Sigma ChemicalCo.,St.Louis, MO)andincubatedat40'C for 2hwhile shaking.Afterincubation, another 200 Ml ofperchloricacid was addedbeforecentrifugation and removal of the supernatant. The supernatantwasneutralizedtopH7withsolidKHCO3and recentri-fuged. The concentration of free glucose in the musclehomogenatewas

also measuredafter neutralization. Total and free glucosewere

mea-sured in50Mlofthe samples afteradding ATP/NADP/G6P dehydroge-nasebuffer and 5 Ml ofahexokinasesuspension(433 U/ml,Sigma).

Resultswereexpressed in millimoles glycosyl units perkilogram dry

muscle.

Analyticalmethods.Serum glucosewasdeterminedafter

centrifuga-tion(Eppendorfmicrocentrifuge;Brinkmann Instruments, Inc., West-bury, NY) by theglucose oxidase methodon aModel 23Ainstrument (Yellow Springs Instrument Co., YellowSprings, OH). Blood for

seruminsulinwascollected in untreated tubes and allowedtoclotat

room temperaturebefore the supernatantwasremoved andfrozenat

-20°C. Insulin was measured byaspecificdouble-antibody radioim-munoassay(32). Bloodfor analysis of plasma glucosespecific activity

was collected in tubescontainingpotassium oxalate plus sodium fluo-ride. Total glycosylatedhemoglobinwasdeterminedonfreshlydrawn whole blood using the Glyc-Afin GHb test kit (Isolab, Inc., Akron,OH).

Statisticalanalysis.Kineticanalysisof glycogensynthaseactivity

(expressedasfractionalvelocity)in responsetoinsulinwasdetermined from individual dose-responsecurves.Thesensitivity of glycogen syn-thasetoinsulinwasassessedfrom the insulin concentrationatwhich glycogen synthaseactivitywashalf-maximal

_(EDm).

E..t wasdefined

asthemaximaleffect of insulinonglycogen synthaseactivity. Todetermine the sensitivity of glycogen synthase to G6P, kinetic

analysisof thesigmoidalG6P dose-response curvefor glycogen syn-thasewasalsoperformed. Theconcentration of G6P producing half-maximalactivation of the enzyme(AO.5)wasdetermined byfitting the datato an Eadie-Hofsteeequation: V, = [-Ao.5 X V/[Sj]] + V,&"

where V, is glycogen synthase activity,

_Si

isG6Pconcentration, and

V.

apI is the activity ofthe enzyme at a saturating G6P concentration.

Datacalculations and statistical analysis were performed using the StatViewIIprogram(AbacusConcepts, Inc., Berkeley, CA). All data

areexpressedasmean±SEM. Statistical significance was tested using

analysisofvariancefollowed by the Student's two-tailed paired t test.

Results

Rates

ofglucose uptake

(Table II)

In the basal state, rates of glucose uptake were 51% higher in the NIDDM than controlsubjects(TableII,P <0.001). When insulin levels were raised to - 300 pmol/liter in clamp A per-formedateuglycemia in the control subjects,the glucose

up-takerateincreased threefold frombasal to 8.25 mg/kg FFM per min.Asimilarglucose uptake rate was achieved in clamp A in

theNIDDMsubjectsatthe sameinsulinlevel byincreasing the

serumglucoseconcentration to- _{20 mmol/liter (P < 0.05 vs.}

controls).

In

_clamp

B,

_{glucose uptake}

wasstimulated furtherto

- 11.5mg/kgFFMperminin bothsubjectgroupsby

increas-ing

insulin levels to 700

pmol/liter,

and

increasing

serum

glucose to - 18 mmol/liter in theNIDDM subjectswhile maintainingserumglucoseateuglycemia in the control

sub-jects.Maximalratesofglucose uptakewereachieved inclamp

C in the NIDDMsubjects by raisinginsulin levelsto -

33,000

pmol/literandincreasingserumglucose to - 18mmol/liter. GlucoseuptakeinclampC in the controlgroup wasmatched

tothe NIDDM subjects usingasimilarinsulin level while

in-creasingserumglucoseto - 15mmol/liter.

Insulin

dose-response

curves

for glycogen synthase

activation

(Figs. I and

2)

Effect of insulin onglycogen synthase activityatphysiologic

levels of substrate (UDPG = 0.3 mmol/liter) (Fig. 1). Fig. 1 illustrates the insulin dose-response activation ofglycogen

syn-thaseactivation expressedasfractionalvelocity in the NIDDM andcontrolsubjects. In both the NIDDM and controlgroups,

insulin significantly stimulated glycogen synthase activity.

However,activation ofglycogensynthaseby insulinwas

signifi-cantlyreduced in the NIDDM compared with control subjects in thebasalstateandatmatchedphysiologic stimulatedrates

ofglucose uptake (clamps A and B). Glycogen synthase was

reducedby 52% atbasal (P<0.01),46% during clampA(P

<0.02) and 35% during clamp B (P<0.04). Glycogensynthase

activity was not significantly different, however, during clamp C.

Kinetic analysis of individual dose-response curves

re-vealed asignificantly higher insulin concentration for half-maximal stimulation of glycogen synthase activity

_(EDm)

in NIDDM compared with control subjects (1,004±264 vs.

257±110 pmol/liter, respectively, P <0.02). However, the maximal effect of insulin on glycogen synthase activity

_(Em,,.)

wassimilar in the NIDDM and control subjects (0.230±0.021 vs. 0.206±0.023, respectively, P=NS).

Effect of insulinontotalglycogen synthaseactivityat maxi-mal concentrations of substrate (UDPG = 5 mmol/liter) and

0.31

o > c = a

* 0

O-C

-ID U

0

o0

0 a

o, To LL

21

,Controls

NIDDM

0.2_F

0.1

[

o.cL1

10 100 1000 10000 100000

Insulin (pmol/L)

Figure1.Effect of insulinonmuscleglycogensynthaseactivity

(expressedasfractionalvelocity=theglycogen synthaseactivityat

0.1 mmol/literG6Pdividedbytheactivityat 10mmol/literG6Pat

subsaturatingUDPGof 0.3mmol/liter)inNIDDMand control

subjects during

basalconditions(circles)andthreeglucoseclamps

(clamp

A,squares, 300pmol/liter;clampB,triangles, 700pmol/liter;

clamp C,

diamonds,

33,400 pmol/liter).Significantdifferenceswere

found betweensubjectgroups in the basalstate(P<0.01),andduring

A CONTROLS

*5 100

0

CD 1-1 80 0_{_}

c c

C

-e 0

c _{E 40}

o C

_5

20

0

50_r

I.

Controls

NIDDM

1-F

._

0

0

00

c E

0

c E

0 CD%.

10 100 1000 10000 100000

Insulin (pmol/L)

Figure 2. Effect of insulin on total glycogen synthase activity (in nanomoles per minute per milligram protein) at saturating levels of UDPG (5 mmol/liter) and G6P (10 mmol/liter) in NIDDM and control subjects under basal conditions (circles) and three glucose clamps (clamp A, squares, 300pmol/liter; clampB, triangles, 700 pmol/liter; clamp C, diamonds, 33,400 pmol/liter). Significant differences between subject groups were found in the basal state (P

<0.05) and during clamp C (P<0.03).

activator(G6P = 10 mmol/liter) (Fig. 2). Fig. 2 illustrates the

insulindose-response curvesfortotal glycogen synthase activ-ity (expressed in nanomoles per minute per milligram protein)

inthe NIDDM and control subjects at maximal concentrations

of bothsubstrate(UDPG, 5 mmol/liter) and allosteric

activa-tor(G6P, 10mmol/liter). Within both the NIDDM and con-trol groups, insulin did not significantly increase total glycogen

synthase activity. Mean total glycogen synthase activity was - 38% lowerunder all conditions (basal and stimulated) for the NIDDMcompared with controlsubjects.

Glucose

6-phosphate

dose-response curves for glycogen

synthase

activation at

physiologic levels of substrate

(UDPG

=

0.3

mmol/liter)

(Figs. 3 and 4)

The G6P dose-response curves forglycogen synthaseactivity (in nanomoles per minute per milligram protein) for the NIDDM and control subjects areillustrated inFig. 3. As the

insulin concentration increasedtherewas aleftward shift inthe

dose-responsecurves.Theshiftwaslesspronounced,however,

forNIDDM thancontrol

subjects.

Toanalyzethese

sigmoidal

G6P dose-response curves

(Fig.

3), kinetic

analysis

was

per-formedusingtheEadie-Hofsteeequation (see Methods)to

de-termine the half-maximalactivation of

glycogen

synthase

by

G6P

(AO5).

Fig.4illustratestheinsulin dose-responsecurvesfor the

_A0.5

for G6P in the NIDDM and controlsubjects. Insulin decreased the

AO.5

for G6P in both the NIDDM and control groups. However,

Ao05

values for basal andclampAwere signifi-cantly

higher

for theNIDDMthan control

subjects (4.49±0.97

vs. 2.86±1.09 forbasal, and 1.57±0.23 vs. 0.64±0.08 mmol/

liter forclamp A, both P<0.01). Asinsulin levelsincreased

above 700

pmol/liter

in

clamps

BandC,no

_significant

dif-ferenceswereobservedin

_A0.5

betweenthe NIDDM and

con-trol

subjects.

Muscle

glycogen

content

(Table III)

The muscle glycogencontentin the basalstateand

during

clamp C inNIDDMandcontrol

subjects

isshown inTableIII. In the NIDDM subjects, glycogen was 33% and 29% lower

40

30I

20

101-0 .01 .1 I

G6P (mmol/L)

B NIDDM

I=

0 co

0

0.

0

X

s

C 0

8

0 U

0

0 _Basal

Clamp A -& Clamp B Clamp C

10

Basal

* ClampA

-*-- Clamp B

* Clamp C

.1 I

G6P (mmol/L)

Figure 3. Effect of G6P on glycogen synthase activity (in nanomoles per minute permilligramprotein)atsubsaturating concentrations of UDPG(0.3 mmol/liter) inNIDDMand controlsubjectsunder basalconditions and three glucose clamps (clamp A, 300 pmol/liter; clampB,700pmol/liter;clampC, 33,400 pmol/liter).

6

0

E E

a.

I-0

Cs

0

5

4

3 2

NIDDM Controls

10 100 1000 10000 100000

Insulin

(pmol/L)

Figure4.Effect of insulinonthe half-maximal activation ofglycogen

synthase byG6P

(AO.5)

wasdeterminedusingtheEadie-Hofstee equationand datapointsfromFig. 3. Results from NIDDM and controlsubjectsareshown under basal conditions(circles)and three

glucose clamps (clampA,squares, 300pmol/liter;clampB,triangles,

700pmol/liter; clamp C,diamonds,33,400pmol/liter). Significant

differenceswerefound between NIDDM and controlsubjectsunder basal conditions andduring clampA(bothP<0.01).

TableIII.Muscle GlycogenConcentration (Millimoles Glycosyl UnitsperKilogram Dry Muscle)in the NIDDM and Control Subjects under Basal Conditions andClampC

NIDDM subjects Control subjects

Basal 238.3±20.1 * 357.3±42.6 Clamp C 397.6±42.0$ 561.5±82.6t

* Significantly different from corresponding study in controls.

*_{Significantly different from corresponding basal value.}

under basal and clamp Cconditions, respectively, compared withcontrol subjects, althoughthisreachedsignificance only underbasalconditions (P<0.02). Glycogenincreased signifi-cantlyby67%(P<0.005)and 57%(P<0.05)as insulin levels

increasedfrom basal to clampC forNIDDMandcontrol

sub-jects, respectively.These increments were notsignificantly

dif-ferent(159.3±37.0vs.204.2±84.1mmolglycosylunits/kg dry muscle in the NIDDM and control subjects, respectively, P

= NS). As shown inFig.5, asignificant correlationwasfound

between glycogen content andglycogen synthase activitywhen

individual subjects'dataforthe basal and clamp Cstudieswere plotted (r=0.51, P<0.01).

Discussion

Thisstudy has demonstrated that multipledefects in muscle glycogen synthase activity are present in NIDDM and

asso-ciatedwithreduced muscleglycogenformation.Although

im-pairedmuscle glycogen synthaseactivity inresponse toinsulin

hasbeen welldocumented in a numberof insulin-resistant

states(7, 12-17), little isknownabout the nature or extentof

these defects. In a previous study carried outduring steady-stateglucose clamp conditionsandinsulinlevelsof 300pmol/

liter, we documented that muscle glycogen synthase activity (expressed as fractional velocity) was reduced by 40% in NIDDMbut could be normalized byincreasinginsulin levels

fourfold(17).

1000r

o

E

0

E

E

c .X0)

0

I,

8001

600

400_f

200

[ _Controls

* NIDDM

r=0.51 p<0.01

~~~~~0~~~

0 0

*~~~

.0 ₀ o

0~~~~~~~~~

0 10 20 30 40

Glycogen synthase activity (fractional velocity)

Figure

5.

_Relationship

between muscleglycogen synthase activity

(expressed

asfractional

velocity)

and muscleglycogencontent(in

millimoles

_glycosyl

unitsper

kilogram

dry muscle)in the basalstate

and

_{during clamp}

CforsevenNIDDM andsevencontrolsubjects

(r

=28,_y= 252_{+9.2 x,r}=

0.51,

_P<0.01).

The results of this study led us to examine this defect in greaterdepthandwehavenowshown that the fractional veloc-ity of glycogensynthaseis reducedby 35-52%atinsulin levels ranging from basal to - 700 pmol/liter. As insulinlevels in-creased, the magnitude of the impairment progressively de-creased and at pharmacologic insulin levels of - 33,400pmol/ liter thedefect normalized. Kinetic analysis ofthe insulin dose-response curves showed that the half-maximal activation constant for insulin wasapproximately fourfoldhigherfor NIDDMthan control subjects, but with no difference in the maximaleffectofinsulin on glycogen synthase. One possible explanation for the normalization of maximal glycogen syn-thaseactivityin NIDDM may bestimulation of enzyme activ-ity by G6Psecondary to glucose transport, originally proposed by Lawrence and Larner (33). Therefore, reduced sensitivity of glycogensynthase to physiologic insulin concentrations exists in NIDDM that is overcome when insulin is raised to pharma-cologic levels.

Inagreement with previous reports (25, 34), at each insulin concentration glycogen synthase was activated in the control

subjects withoutchanging the total glycogen synthase activity (i.e., the true maximal activity at saturating concentrations of G6P andUDPG).Similarly,increasing insulin levels from ba-sal to clampC didnotsignificantlychange total glycogen syn-thaseactivityin the NIDDM subjects. Although total activity was less in the NIDDM subjects than in controls over the range

of insulin concentrations studied, itwassignificantlyless only in the basal state andduringclamp C.Alterationsinthe turn-over rateofglycogen synthase,suchas a decrease in the rate of

synthesis,increaseintherateofdegradation,or some combina-tion ofthetwo could beresponsible forreduced total enzyme

activity in NIDDM. Alternatively, NIDDM and control

sub-jectsmaypossess identical amounts of glycogen synthase

pro-teinbut the enzyme simply has less activity in NIDDM. What-everthecause, our data strongly suggest that NIDDM subjects have areduced functional mass of the enzyme, which is not overcome by acutelyincreasingthe serum insulin

concentra-tion.

Insulin acts as a major regulator of glycogen synthase in muscle bydephosphorylatingthe enzyme and reducing the

_AO.

for G6P.G6Pisnot only an allosteric effector but also a messen-ger molecule that canmediatehormone action by stimulating thephosphatase reaction (4, 5). In muscle from diabetic

rab-bits, the half-maximal G6P activation constant for glycogen synthasehasbeen shown to increase by two to threefold com-paredwithnondiabeticanimals (11). Our study has now simi-larlydemonstrated that reduced sensitivity of muscle glycogen synthase toallostericactivation by G6P exists in NIDDM. This is an expected finding since the elevated

Ao

sof glycogen

syn-thasefor G6P islikely the result of the increased

phosphoryla-tion stateofthe enzyme, shown by Roach and Larner to be

associated with reduced fractional glycogen synthase activity (35). Thereduced sensitivity to G6P was overcome at insulin

concentrationsabove 700pmol/liter since insulin presumably

dephosphorylatesglycogen synthase to such an extent that the enzyme nolonger requires G6P for complete activation.

Thisstudy not only examined the extent ofdefects in glyco-gensynthase activationby insulin in NIDDM, but also demon-strated thatdefects inglycogen synthase activity contribute to reduced glycogen content in NIDDM. This study now

con-firms thatglycogen content is reduced in NIDDM, and that

glycogen synthase activity remainsimpaired at physiologic

sulin concentrations even when rates ofglucose uptake are nor-malized. Furthermore, a close correlation between glycogen synthaseactivity and glycogen content has been clearly demon-strated.Thus,these findings taken together areconsistentwith the notion that changes in glycogen synthase activity contrib-utesignificantly to decreased glycogen content in NIDDM.

When adefectin insulin's ability to activateglycogen syn-thase has been reportedpreviously,it has usually been accom-paniedby acorrespondingreduced rateofglucose uptake into

insulin-sensitive tissues (e.g., 7, 12-16). This association has lead to the suggestion that a causal relationship exists between

insulin-stimulation ofglycogen synthase and glucose uptake (7, 13, 14). However, the fact that we found reduced glycogen synthase activity in NIDDM subjects when rates ofglucose uptake werenormalized indicatesthat theeffect of insulinon muscleglycogensynthase is clearly separable fromitseffecton glucose transport (2, 17, 25, 36,37)andconfirmsthatreduced

glycogen synthase activity in NIDDM is not solely a conse-quenceof reduced glucose uptake.Furthermore, thisstudy

providesevidence that activation ofglycogen synthase and glu-cose transportbyinsulin mayoccur through different mecha-nisms.

At present we can only speculate as to the cause of the reducedsensitivity ofglycogen synthase to covalent modifica-tion and allostericactivation by insulinandG6P inNIDDM. The mostlikely explanation isthat glycogen synthase in muscle from NIDDM subjects is morephosphorylated,and hence

in-activated, due to an impairment in the complex pattern of hormonally regulated multisite phosphorylation which con-trolstheactivity of glycogen synthase. Therecent demonstra-tionby Kidaand co-workersthat glycogen synthase

phospha-taseis reducedininsulin-resistant

subjects

supportsthis

con-cept(7).Althoughtheetiology of

impaired

glycogen synthase activity in NIDDMremainstobe

determined,

ithasrecently been demonstrated that this defect can be reversed, atleast

partially, in NIDDM subjects by

eight

weeksof

sulfonylurea

therapy(38).

Insummary,multiple defects in glycogen synthase

activity

and reducedglycogencontentexistin musclefromnonobese NIDDM subjectsover arange of

physiologic

insulin

concen-trations and similarratesof

glucose

uptake.

NIDDMis

asso-ciated withadefectin the

sensitivity

of muscle

glycogen

syn-thase toinsulin thatbecomes

progressively

lessasinsulin

con-centrations increase.

Increasing

the insulinconcentration does not,however, overcome thedefect intotal enzyme

activity

(i.e., functional enzymemass). These defectsin

glycogen

syn-thase

activity

and the correlation between

glycogen synthase

activityandglycogencontentsupport the contention that

mul-tiple defectsin

glycogen

synthase activity

contributetoreduced glycogendepositionin NIDDM.

Acknowledgments

Wegratefully acknowledge the helpful advice received fromDr. Har-veyKaslow,University of SouthernCalifornia,LosAngeles, andDr. TedCiaraldi, Veterans Administration Medical Center, San Diego, and theassistance of PennyWallace, Therese Flynn, Michelle

Gold-berg,CleonTate,theresearchvolunteers,and thenursingstaffof the Division ofEndocrinologyandMetabolism,VAMedicalCenter,San

Diego.

Thisstudywassupported inpartbythe AmericanDiabetes Associa-tion (ADA), the Medical Research Serviceof the Veterans

Administra-tion Medical Center, grant DK-38949 fromtheNationalInstitute of

Arthritis,Diabetes, and Digestive andKidneyDiseases, and grant PHS

RR-00827 from the General Clinical Research Branch, Division of

Research Resources, NationalInstitutes of Health. BarryGumbineris

therecipient ofan ADACaliforniaAffiliate FellowshipGrant. Anne Thorburn is a Neil Hamilton Fairley Fellow funded by the National Health and Medical Research Council of Australia.

References

1.Shulman, G.I., D. L. Rothman, T. Jue, P. Stein, R. A. DeFronzo, and R. G.

Shulman. 1990.Quantitation of muscle glycogen synthesis in normal subjects

and subjectswith non-insulin-dependent diabetesby"3Cnuclearmagnetic reso-nancespectroscopy. N.Engl.J. Med. 322:223-228.

2.Danforth, W. H. 1965.Glycogen synthetase activityinskeletal muscle. Interconversion oftwoformsandcontrolofglycogen synthesis.J. Biol.Chem. 240:588-593.

3.Okubo, M., C. Bogardus, S. Lillioja, and D.M.Mott. 1988. Glucose-6-phosphate stimulationofhuman muscle glycogen synthase phosphatase. Metab. Clin. Exp. 37:1171-1176.

4.Gilboe, D. P., and F. Q. Nuttall. 1972.TheroleofATP and

glucose-6-phos-phatein the regulation of glycogen synthetase D phosphatase. Biochem. Biophys. Res.Commun. 48:898-906.

5.Oron,Y.,andJ.Larner.1980.Insulinaction in intactmousediaphragmI.

Activation ofglycogen synthase through stimulation ofsugar transport and phos-phorylation.Mol. Cell. Biochem. 32:153-160.

6.Okubo, M.,C. Bogardus, S.Lillioja, and D. M. Mott. 1989. Adenosine 3',5'-monophosphate-dependent protein kinase activitydeceasesin human

mus-cleafter insulin infusion. J. Clin. Endocrinol. Metab. 69:798-803.

7. Kida,Y.,A.Esposito-Del Puente, C. Bogardus, andD. M.Mott. 1990. Insulin resistance is associated with reduced fasting and insulin-stimulated

glyco-gensynthase phosphataseactivity inhumanskeletal muscle. J.Clin. Invest. 85:476-481.

8. Kruszynska,Y.T.,and P. D.Home.1988.Liver and muscle insulin sensi-tivity, glycogen concentrationandglycogen synthaseactivity ina ratmodel of non-insulindependent diabetes. Diabetologia. 31:304-309.

9.Rossetti, L., and M.R.Laughlin.1989.Correction of chronic hyperglyce-mia with vanadate, butnotwithphlorizin,normalizesinvivo glycogenrepletion

andinvitro glycogen synthase activity in diabetic skeletal muscle. J.Clin.Invest. 84:892-899.

10.Kaslow,H.R.,andR. D.Eichner. 1984.Fasting and diabetesalteradipose

tissue glycogen synthase.Am.J.Physiol. 247:E581-E584.

1 1.Sheorain,V.S.,B.S.Khatra, andT. R.Soderling. 1982. Hormonal regula-tion of glycogen synthase phosphorylaregula-tion inrabbit skeletal muscle. J. Biol. Chem. 257:3462-3470.

12. Freymond, D., C.Bogardus,M.Okubo,K.Stone, and D.Mott. 1988.

Impairedinsulin-stimulated muscleglycogensynthaseactivation in vivo inman

is relatedtolowfasting glycogensynthasephosphataseactivity.J. Clin. Invest.

82: 1503-1509.

13. Bogardus,C.,S.Lillioja, K.Stone,andD. Mott. 1984. Correlation be-tween muscleglycogensynthaseactivity and in vivo insulin action inman.J. Clin. Invest. 73:1185-1 190.

14.Wright,K.S.,H.Beck-Nielsen,0.G.Kolterman,and L.J.Mandarino.

1988. Decreasedactivation of skeletal muscleglycogensynthase bymixed-meal

ingestion in NIDDM.Diabetes. 37:436-440.

15.Kruszynska, Y. T., G.Petranyi,and K. G. M. M. Alberti. 1988. Decreased

insulinsensitivityand muscleenzymeactivityinelderlysubjects.Eur. J. Clin.

Invest. 18:493-498.

16. Mandarino, L.J., Z. Madar, 0. G. Kolterman,J. M. Bell,and J. M.

Olefsky. 1986.Adipocyte glycogen synthaseand pyruvatedehydrogenase in

obese andtype IIdiabeticsubjects.Am. J. Physiol.251:E489-E496.

17.Thorburn,A.W.,B.Gumbiner,F.Bulacan,P.Wallace,and R. R.Henry.

1990. Intracellularglucose oxidationandglycogen synthase activityarereduced

in non-insulin-dependent (TypeII)diabetesindependent ofimpairedglucose

uptake. J. Clin.Invest.85:522-529.

18.National DiabetesDataGroup. 1979.Classification anddiagnosisof dia-betes mellitusandothercategories ofglucoseintolerance. Diabetes.

28:1039-1057.

19. Goldman, R. F.,and E. R. Buskirk. 1961. A method for underwater

weighing andthedeterminationofbodydensity.InTechniquesforMeasuring

BodyComposition.J.Brozek andA.Dershel,editors. National Research

Coun-cil, NationalAcademyofSciences,Washington,DC. 78-89.

20.Thorburn, A. W.,B. Gumbiner,G. Brechtel,andR. R. Henry. 1990.

Effect ofhyperglycemiaand hyperinsulinemiaonintracellularglucoseand fat

metabolism inhealthysubjects.Diabetes. 39:22-30.

21. Henry,R.R., B.Gumbiner,T.Flynn,and A. W.Thorburn. 1990. The

metabolic fate of intracellularglucosein non-insulindependentdiabetes mellitus: effectsofhyperinsulinemia andhyperglycemia. Diabetes. 39: 149-156.

22.DeFronzo,R.A.,J. D.Tobin,and R. Andres. 1979. Glucoseclamp

tech-nique:amethodforquantifyinginsulin secretion and resistance. Am.J.Physiol.

273:E214-E223.

23. Steele, R. 1959. Influence ofglucose loadingandinjectedinsulin on

he-patic glucose output.Ann. NYAcad. Sci. 82:420-430.

24.Chiasson, J. L., J.E.Liljenquist,W. W.Lacy,A.S.Jennings,and A. D.

Cherrington. 1977.Gluconeogenesis: methodological approachesinvivo. Fed. Proc.36:229-235.

25.Yki-Jarvinen, H.,D.Mott, A. A.Young,K.Stone, and C. Bogardus. 1987. Regulation of glycogensynthase and phosphorylase activities by glucose and insulin in human skeletal muscle. J. Clin. Invest. 80:95-100.

26. Bergstrom, J. 1962. Muscle electrolytes in man. Scand. J. Clin. Lab. In-vest. 14(Suppl.68):7-101.

27.Nuttall,F.Q., J. Barbosa, andM.C. Gannon. 1974.Theglycogen

syn-thasesystemin skeletalmuscleof normalhumans and patientswithmyotonic

dystrophy: effect of glucose and insulin administration. Metab. Clin. Exp. 23:561-568.

28. Thomas, J.A., K. K.Schlender, andJ.Larner.1968. Arapidfilter paper assayforUDPglucose-glycogen glucosyltransferase,including an improved

bio-synthesis of UDP-'4C-glucose. Anal. Biochem. 25:486-499.

29.Mandarino,L.J., K. S.Wright, L. S. Verity,J.Nichols, J.M.Bell, 0. G. Kolterman, andH. Beck-Nielsen. 1987.Effects of insulininfusiononhuman skeletal muscle pyruvate dehydrogenase,phosphofructokinase, and glycogen

syn-thase.J. Clin. Invest. 80:655-663.

30.Lowry,0.H., N. J. Rosebrough, A. L.Farr,and R. J. Randall. 1951. ProteinmeasurementwiththeFolinphenolreagent.J. Biol.Chem.193:265-275. 31. Keppler,D., and K. Decker. 1974. Glycogendetermination with

amylo-glucosidase.In Methods ofEnzymaticAnalysis. H. U.Bergmeyer,editor. Aca-demic Press, New York. 1127-1131.

32.Desbuquois,B., andJ. A. Aurbach. 1971.Useof polyethyleneglycolto separate free andantibodyboundpeptide hormonesinradioimmunoassay.J. Clin. Endocrinol. & Metab. 33:732-738.

33. Lawrence, J. C., andJ. Larner. 1978. Activation ofglycogen synthasein ratadipocytes byinsulin andglucoseinvolvesincreased glucosetransport and

phosphorylation. J.Biol. Chem. 253:2104-2113.

34. Larner,J. 1988. BantingLecture 1987.Insulin signaling mechanisms. Lessons from the oldtestamentofglycogen metabolismand the new testament of

molecular biology. Diabetes. 37:262-275.

35. Roach, P.J., and J. Larner. 1976. Rabbitskeletal muscle glycogen

syn-thase. II. Enzymephosphorylationstate and effector concentrations as interact-ing control parameters. J.Biol. Chem. 251:1920-1925.

36.Chiasson, J-L., M.R.Dietz, H. Shikama, M. Wootten, and J. H.Exton. 1980. Insulinregulation of skeletal muscle glycogen metabolism.Am.J.Physiol.

239:E69-E74.

37.Larner, J., and C.Villar-Palasi. 1971.Glycogen synthaseand its control.

Curr. Top. Cell.Regul.3:195-236.

38.Bak,J.F., 0.Schmitz,N. SchwartzSorensen,and0. Pedersen. 1989.