Mechanism by which hyperglycemia inhibits
hepatic glucose production in conscious rats.
Implications for the pathophysiology of fasting
hyperglycemia in diabetes.
L Rossetti, … , G Sebel, M Hu
J Clin Invest. 1993;92(3):1126-1134. https://doi.org/10.1172/JCI116681.
To examine the relationship between the plasma glucose concentration (PG) and the pathways of hepatic glucose production (HGP), five groups of conscious rats were studied after a 6-h fast: (a) control rats (PG = 8.0 +/- 0.2 mM); (b) control rats (PG = 7.9 +/- 0.2 mM) with somatostatin and insulin replaced at the basal level; (c) control rats (PG = 18.1 +/- 0.2 mM) with somatostatin, insulin replaced at the basal level, and glucose infused to acutely raise plasma glucose by 10 mM; (d) control rats (PG = 18.0 +/- 0.2 mM) with somatostatin and glucose infusions to acutely reproduce the metabolic conditions of diabetic rats, i.e., hyperglycemia and moderate hypoinsulinemia; (e) diabetic rats (PG = 18.4 +/- 2.3 mM). All rats received an infusion of [3-3H]glucose and [U-14C]lactate. The ratio between hepatic [14C]UDP-glucose sp act (SA) and 2X [14C]-phosphoenolpyruvate (PEP) SA (the former reflecting glucose-6-phosphate SA) measured the portion of total glucose output derived from PEP-gluconeogenesis. In control rats, HGP was decreased by 58% in hyperglycemic compared to euglycemic conditions (4.5 +/- 0.3 vs. 10.6 +/- 0.2 mg/kg.min; P < 0.01). When evaluated under identical glycemic conditions, HGP was significantly increased in diabetic rats (18.9 +/- 1.4 vs. 6.2 +/- 0.4 mg/kg.min; P < 0.01). In control rats, hyperglycemia
increased glucose cycling (by 2.5-fold) and the contribution of […]
Research Article
Find the latest version:
Mechanism
by
Which Hyperglycemia Inhibits
Hepatic
Glucose
Production
in
Conscious Rats
Implications for the Pathophysiology of Fasting Hyperglycemia in Diabetes
LucianoRossetti, Andrea Giaccari, Nir Barzilai, Kathleen Howard, Gary Sebel, and Meizhu Hu
Division ofEndocrinology, DepartmentofMedicine,Albert Einstein College ofMedicine, Bronx, New York 10461; and Diabetes Division, DepartmentofMedicine, UniversityofTexasHealth Science Center, San Antonio, Texas 78284
Abstract Introduction
Toexaminetherelationship between the plasma glucose con-centration(PG) and the pathways of hepatic glucose produc-tion(HGP), fivegroupsof consciousrats werestudied aftera 6-h fast:(a) controlrats(PG=8.0±0.2mM); (b)controlrats (PG=7.9±0.2 mM)with somatostatin and insulinreplaced at the basal level; (c) control rats (PG = 18.1±0.2 mM) with somatostatin, insulin replacedatthe basal level,and glucose infusedtoacutely raise plasmaglucose by 10 mM;(d)control rats(PG= 18.0±0.2mM)with somatostatin and glucose infu-sionstoacutelyreproduce the metabolic conditions of diabetic rats,i.e.,hyperglycemiaand moderatehypoinsulinemia; (e) dia-beticrats(PG= 18.4±2.3mM).All ratsreceivedaninfusion
of
13-3HIglucose
and [U-14C
Ilactate.
The ratio betweenhepatic
I14CJUDP-glucose
sp act (SA) and 2X114CJ-phosphoenolpyruvate (PEP) SA (the former reflecting
glu-cose-6-phosphate SA) measured the portionof total glucose output derived from PEP-gluconeogenesis. In control rats, HGPwasdecreased by 58% inhyperglycemic comparedto eug-lycemic conditions (4.5±0.3 vs. 10.6±0.2 mg/kg min; P
<0.01).When evaluated underidenticalglycemicconditions,
HGPwassignificantlyincreased in diabeticrats(18.9±1.4vs. 6.2±0.4
mg/kg.
min; P<0.01).Incontrolrats,hyperglyce-mia increased glucosecycling(by2.5-fold)and the contribu-tion ofgluconeogenesistoHGP
(91%
vs.45%),
whiledecreas-ing that of glycogenolysis (9% vs.
55%).
Under identicalplasmaglucose and insulin concentrations, glucosecyclingin diabeticratswasdecreased(by
21%)
and thepercent contribu-tionofgluconeogenesistoHGP(73%)
was similartothat of controls(84%). These data indicate that:(a) hyperglycemia
causes a markedinhibition of HGP
mainly
through
the sup-pression ofglycogenolysisandtheincrease inglucokinaseflux,
withnoapparentchangesin thefluxesthrough
gluconeogene-sis and
glucose-6-phosphatase;
under similarhyperglycemic
hypoinsulinemicconditions:
(b)
HGP ismarkedly
increased in diabetic rats; however,(c)
the contribution ofglycogenolysis
andgluconeogenesistoHGP is similartocontrolanimals.
(J.
Clin.Invest. 1993,92:1126-1134.) Keywords:diabetes melli-tus *glycogenolysis*glucokinase *
gluconeogenesis
*glucose
cycling .
glucose-6-phosphatase
AddressreprintrequeststoDr.LucianoRossetti,Divisionof Endocri-nology, Department ofMedicine,Albert EinsteinCollegeofMedicine,
1300MorrisPark Avenue,Bronx,NY 10461.
Receivedfor publication 12January1993 andin revisedform 9
April1993.
Overproduction of glucose is the majorcauseoffasting hyper-glycemia in both insulin-dependent (1) and non-insulin-de-pendent diabetes mellitus (NIDDM)
(2).'
Gluconeogenesis composes anhigherpercentageofhepatic glucose production (HGP) indiabeticascomparedwith nondiabetic individuals(3-6),andit has been postulatedtobe the "primary" determi-nantofincreased HGP in diabetes(5). However, recent experi-mentalevidencesuggeststhatchangesinthe rates offormation ofhepaticglucose-6-phosphate (G6P) through gluconeogene-sis donotalter HGP inconsciousdogs (7), healthy volunteers (8-10), and NIDDMsubjects(11).The"finalcommon path-way"forthe netreleaseofglucoseintothecirculationis regu-lated bythe balanceoffluxesthroughglucokinaseand glucose-6-phosphatase(G6Pase). Theassessmentof in vivo metabolic fluxescontributingtoG6Pformation andtohepaticglucose output, in combinationwith the steady-state changes in the
hepaticG6Pconcentration,may allow one todiscern ifthe key regulatory siteforHGP is theformation of G6Poritsnet
de-phosphorylationtoglucose.SeeFig.1foraschematic
represen-tation.
Soskin and Levine ( 12) first proposed that mammalian liver can rapidly change its glucose output in response to changes in thecirculating glucoseconcentration independently from hormonalsignals. It isnow recognizedthatthe plasma glucose concentration per se regulates HGP ( 13-18)and as much as50% ofthedecline in plasma glucose concentration after glucose administrationmaybe duetothecombined effect
ofhyperglycemiaper se on glucosedisposalandHGP(18).
However, relatively little information is available about the
mechanism(s)bywhichhyperglycemia,independent of insu-lin,inhibits HGP. Basal HGP is markedlyincreasedin some
diabetic states despite the presence of hyperglycemia and normo- orhyperinsulinemia,bothof whichareknownto sup-pressHGP(2, 13-18).Thismaysuggest arolefor defective
glucose-induced suppression ofHGP in thepathophysiologyof
fastinghyperglycemiaindiabetes.
HGP iscomposed ofglycogenmobilizationand
gluconeo-genesis. However, underpostabsorptiveconditions,i.e.,basal insulin, a sizeable portion of the flux through G6Pase also comesfromglucosecycling ( 19-23). Althoughhyperglycemia
per seinhibitsHGP, suchaneffecthasnotbeendemonstrated for totalglucoseoutput(TGO) (21),i.e.,fluxthroughG6Pase. Thus, it ispossiblethatacomponentoftheeffect
ofhyperglyce-miaper se insuppressingHGPis mediated by the enhanced
1.Abbreviations used in this paper: G6P, glucose-6-phosphate; G6Pase,
glucose-6-phosphatase;GC, glucose cycling; HGP, hepatic glucose
pro-duction; NIDDM, non-insulin-dependent diabetes mellitus; PEP, phosphoenolpyruvate; SRIF, somatostatin; TGO, total glucose output. J.Clin. Invest.
© The AmericanSocietyfor ClinicalInvestigation,Inc.
0021-9738/93/09/1126/09 $2.00
Figure1.Schematicrepresentation ofthemajorpatways and enzy-matic stepswhichregulatehepatic glucose production.Ina netsense, thehepatic glucose-6-phosphate (glucose-6-P)poolreceives three
major inputs: (a)plasma-derivedglucose,(b) gluconeogenesis,and
(c)glycogenolysis. Thefinalcommonpathwayforhepatic glucose
output is thenetdephosphorylationofG6P,whichisregulated by
thebalanceofglucokinase(GK)andG6Pase activities.Similarly,the
netcontribution ofhepatic glycogentothe G6Ppoolrepresents the balance of thefluxes throughglycogen synthaseandglycogen phos-phorylase.Therelative contribution ofplasma glucoseand gluconeo-genesistothehepaticG6Ppoolcanbedirectlymeasuredbytracer
methodology. Infact,after[3-3H]glucoseinfusion,theratio of spe-cificactivities oftritiatedhepatic UDP-glucose ( UDPG)andplasma glucoserepresents the percentage of thehepaticG6Ppoolwhich is derived fromplasmaglucose.Similarly,after[U-'4C]lactateinfusion,
theproportionoftheG6Ppoolwhich isformedthrough PEP-gluco-neogenesiscanbe calculatedastheratio of '4C-labeledUDP-glucose
and PEP.
phosphorylationofplasma glucose,i.e., increased fluxthrough
glucokinase. Inthatamarkeddecreaseinhepaticglucokinase
activityhas been shown in several animal models of diabetes mellitus(24),an impairmentin thehepatic capacityto
phos-phorylate glucosemaycause abluntedglucose-inducedrise in
glucose cyclingandadiminished inhibition ofHGP.
The short-term in vivo regulation of glucokinase and G6Pasefluxes andactivityremains controversial.Inparticular,
themechanism(s) bywhichacutechangesin theplasma
glu-cose concentration cause alterations in HGP have not been delineated. Because several hepaticenzymesare defectivein thediabeticliver,thecombinedassessmentof invivoglucose
fluxes,
substrateconcentrations,and invitroenzymeactivitiesisrequiredtodistinguishtherelative role of these biochemical alterationsoninvivohepatic glucosemetabolism. Our results suggestthat(a) acutehyperglycemia (in thepresenceof low
insulin)
causes a marked inhibition of HGP in normal ratsmainly throughthesuppressionofglycogenolysisandthrough
a markedincrease inglucokinase flux(increased glucose
cy-cling),withno apparentchangesin the fluxesthrough
gluco-neogenesis
andG6Pase and(b)
undersimilar conditions of hyperglycemiaandhypoinsulinemia, increased glycogenolysis and gluconeogenesis and decreased glucosecycling(GC) all contributetotheenhanced HGP indiabeticrats.Methods
Animals. Twogroups of male Sprague-Dawley rats (Charles River
Breeding Laboratories,Inc.,Wilmington,MA)werestudied: groupI,
controls(n=44);groupII,partiallypancreatectomizedrats(n= 10).
At3-4wkof age,allrats(80-100 g)wereanesthetized with
pentobar-bital(50 mg/kg bodywti.p.) and in group II 90% oftheir pancreaswas
removed accordingtothetechniqueofFoglia (25), asmodifiedby
Bonner-Weir et al. (26). Immediately after surgery rats were housed in individual cages and subjectedto astandard light (6a.m. to6p.m.)/
dark(6 p.m.to6a.m.) cycle. After surgeryrats wereweighed twice weekly and tail vein blood was collected for the determination of
non-fasting plasma glucose and insulin concentrations at the same time (8 a.m.). The fasting plasma glucose and insulin concentrations alsowere
determined weekly on tail vein blood. 5 wk after pancreatectomyrats
were anesthetized with intraperitoneal injection of pentobarbital(50 mg/kg bodyweight) andindwellingcatheterswereinsertedinto the
rightinternal jugular vein and in the left carotid artery,aspreviously
described(23, 27, 28).
Insulin/somatostatin/infusions. Studies were performed inawake,
unstressed, chronically catheterized rats using the euglycemic or hy-perglycemic clamp technique in combination with [2-3H]glucoseor
[3-3H]glucose infusions as previously described (23, 27, 28). Rats were fasted for 6 h before thein vivo studies. Briefly, a prime-con-tinuousinfusion of somatostatin (0.8jig/kg*min)andregular insulin (-0.4mU/kg. min) or saline (protocols 1 and 4) were administered, and avariable infusion of a 25% glucose solution was started at time
zeroandperiodically adjustedtoclamp the plasmaglucose
concentra-tion at - 8 mM(euglycemic studies; protocol 2) or - 18mM
(hy-perglycemic studies; protocol 3A and 3B).
During the euglycemic study the insulin infusion was adjustedto
maintainnormoglycemia during the somatostatin infusion. The aver-ageinsulin infusion required to maintain normoglycemia in control ratswas0.4±0.1 mU/kg.min,without need for glucose infusion. 80 minbefore starting the insulin / somatostatin infusion, a prime-continu-ousinfusion of[3H-3 ]glucose (New England Nuclear, Boston, MA; 40
jCibolus,0.4,Ci/min)wasinitiated and maintainedthroughoutthe basalperiodandstopped beforeinitiatingthesomatostatin / insulin in-fusions. At t = 30 min into the infusion study (a time when the tritiated glucose infusedduring the basal period was not detectable in plasma andanew"metabolic"steady-statewasachieved)aprime-continuous
infusion of[3-3H]glucose(15-40,uCi bolus, 0.4jCi/min)was initi-ated andmaintained throughout the remainder of the study. [U-
14C]-lactate(20MCibolus/1.0,Ci/min)wasinfusedduring the last 10 min of thestudy. A subgroup of nine 6-h fastedratsreceivedan hyperglyce-mic clamp study in combination with somatostatin infusion (asin protocol3B), but[2-3H]glucoseratherthan
[3-II]glucose
wasusedto measuretherateofglucoseturnover.Plasma samples for
determina-tion of[3H]glucose specific activitywereobtainedat I0-min intervals throughout the basal period and the insulin/somatostatin infusions. Plasma samples fordetermination of plasma insulin and glucagon con-centrationswereobtained at time -30, 0, 30, 60, 90, 120 min during thestudy. The total volume of blood withdrawnwas - 3.0 ml per
study;topreventvolumedepletionandanemia, asolution ( 1:1 vol / vol)of - 4.0 mlof fresh blood (obtained by heart puncture from a
littermate of the testanimal)andheparinized saline (10 U/ml) was infused. All determinations were also performed on portal vein blood obtained at the end of the experiment. At the end of the insulin
infu-sion,rats wereanesthetized (pentobarbital60 mg/ kg body wt, i.v.), the abdomenwasquicklyopened, portal vein bloodobtained and liver was
freeze-clampedinsitu with aluminum tongs precooled in liquid nitro-gen. Thetime from the injection ofthe anesthetic until freeze-clamping of the liverwas <45 s.Alltissue samples were stored at -80°Cfor
subsequentanalysis.
The study protocol was reviewedandapproved by theInstitutional Animal Care and Use Committees of the University of Texas Health Science CenteratSan Antonio andof the Albert Einstein College of Medicine.
Glycogen synthase. Hepaticglycogen synthase activitywas
mea-suredbyamodification (29, 30) of the method of Thomas et al. (31 ) andis basedonthe measurementof the incorporation of radioactivity
intoglycogenfromUDP-[U- '4C]glucose. Tissue samples (20-30 mg)
werehomogenizedin 2.0mlofTris/HCl buffer,pH7.8,containing10
liver glycogen typeIII. The homogenate was centrifuged at 2,000 g for 15 min (at40C) and the supernatant used for glycogen synthase assay bymeasuringtheincorporationof UDP-[U-
'4C]glucose
into glycogen at30'C. Synthase Iactivity is definedas thatactivityassayable in the presenceof0. I mMG6P.Inorder toapproximatethe invivoglyco-genicratesincubationsarealso carried out in presenceofthe UDP-glu-coseconcentrations (125-500MM)withinthephysiologicalrangefor
ratliver. Total glycogen synthase D activity is measured in the presence
of7.2 mMG6P.The percentageofsynthaseintheactive(FVO.
I)
form is definedas theratioof activities(I/D)X 100.Glycogenphosphorylase. Liver glycogen phosphorylase activity was
measured aspreviously described(32). This assayisbased on the mea-surementof the incorporation of 14C intoglycogenfromlabeled glu-cose-I-phosphate.Glycogen phosphorlasea,theactivephosphorylated enzyme, wasassayedin the absenceofAMP.Tissuehomogenates (20-30mg)werepreparedasdescribedabove. The supernatant was used
forglycogenphosphorylaseassaybymeasuringtheincorporation of ['4Cjglucose-l-phosphateinto glycogenat300Cinamixture contain-ing33mMMes, 200mMKF, 0.45% mercaptoethanol, 15mM
glu-cose-l-phosphate (50
ACi/mmol),
and 3.4mg/ml glycogen. Glucokinase.Hepatic glucokinase activitywas measuredbythecon-tinuousassayspectrophotometric method asdescribed by Davidson
andArion(33), withsomemodifications. Liver homogenates (- 200
mg) were prepared in50mMHepes, 100mMKC1, 1 mMEDTA, 5
mMMgCI2,and 2.5 mMdithioerythreitol. Homogenateswere centri-fuged at 100,000gfor45 min to sedimentthemicrosomal fraction (which willbe usedfortheG6Paseassay). Thepostmicrosomal frac-tion is assayedin amedium, pH7.4at37°C,containing50 mMHepes,
100mMKCI,7.5mMMgCI2,5mMATP, 2.5 mMdithioerythreitol,
10mg/mlalbumin, and 0.5 mM(exokinasesactivity)or7, 18, and
100 mMglucose (totalphosphorylating activity),0.5 mMNAD+,4U
of G6P dehydrogenase (L. mesenteroides)andtheequivalentof - I
mgofwetliver.Thereaction is initiated bytheadditionof ATP,and the rateof NAD+ reduction willbe recordedat340nm.Glucose phos-phorylation is determinedastheabsorbancechangein thecomplete medium minustheabsorbancechange in theabsenceofATPunder
conditionsin which the absorbance is increasing linearlywith time (generally from 15 to40min). Thisimprovedassayprocedure
pro-videsan accurate measureofthehepatic glucosephosphorylating ca-pacity and, in the presenceofphysiological glucose concentrations (7-18 mM), allow one to approximate the in vivo glucokinase activity (33).
Glucose-6-phosphatase. Theassaywasperformedasdescribedby
Burchell et al.(34)andis basedonthedissociation ofPifrom G6Pby
thetissue microsomal fractioncontainingthe G6Pase. Themicrosomal fractionwasincubated withI mMG6P.Thereactionwasstoppedafter
20 minwith solutioncontainingacidmolybdate,with2/9volof10%
SDSandI/9volof10%ascorbic acid.Itwasthenincubated for20 min at450C, and theabsorbance readat 820nM. Standardcurveis
ob-tained from different concentrations of
Pi.
Analytical procedures.Plasmaglucosewasmeasuredbytheglucose oxidase method (Glucose Analyzer II, Beckman Instruments, Inc.,
PaloAlto,
CA.)
andplasma insulin byradioimmunoassayusingratandporcine insulin standards.Plasma[3H]glucose radioactivitywas
mea-sured induplicateon thesupernatants ofBa(OH)2andZnSO4
precipi-tates(Somogyi procedure) of plasma samples afterevaporationto
dry-ness toeliminate tritiated water. Liver G6P concentrationswere
mea-sured spectrophotometrically as described by Michal (35). Liver
glycogen concentrationwasdeterminedaspreviously described(23, 27).Uridinediphosphoglucose(UDP-glucose)and
phosphoenolpyru-vate (PEP)concentrations andspecific activitiesin theliverwere ob-tainedthroughtwosequentialchromatographic separations,as previ-ously reported (23, 36).Differences between groupsweredetermined
byANOVAanalysis ofvariance.
Terminology.Inthe present manuscripttheterm "totalglucose output" (TGO)is intendedastotal in vivofluxthrough G6Paseas
measuredby[2-3H]glucoseturnover, and theterm"hepatic glucose production" (HGP)is intendedasthenet ratesofG6P
dephosphoryla-tion to glucose (balance of the in vivofluxes through glucokinaseand G6Pase) as measured by[3-3H]glucose turnover. Finally, "glucose cycling(GC) is definedasimput ofextracellular glucoseintothe G6P poolfollowed by exit ofplasma-derivedG6P backintothe extracellular pool.
Calculations. The HGP was calculated as thedifferencebetween thetracer-derivedrateofappearance and theinfusionrateofglucose.
Gluconeogenesis (G)wasestimated fromthespecific activities (SA) of '4C-labeledhepaticUDP-glucose(this isassumed toreflectthespecific activityofhepatic G6P),andhepaticPEPaftertheinfusion of[ U-
'4CJ-lactate and[3-3H]glucose:G=TGOX
[`4C]UDP-glucose
SA/[4C]-PEP SA x 2.
Glycogenolysiswas calculated as the difference between the HGP and thegluconeogenesis.Anadditional estimate ofhepatic
glycogenoly-siswasobtained fromthe decreasein liver glycogen concentration (be-low the basal valuesobtainedduring the saline infusions) during the in vivo study.
The percentage ofthe hepatic G6P pool directly derived from plasma glucose can becalculated as the ratio of[3H]UDP-glucose and plasma [3_3H ]glucosespecific activities.Thus,this ratio also measures the percentcontribution ofplasma glucosetothe G6Pase flux(i.e.,
GC). Since TGOisequal to the sum ofthe HGP + GC (and GC
= [3H]UDP-glucose SA/plasma [3-3H]glucose SA X TGO), the
equation can be resolved to calculate both GC and TGO: TGO
=HGP/( 1 -
[3H]UDP-glucose
SA/plasma[3-3H]glucoseSA) and GC=[3H]UDP-glucoseSA/plasma[3-3H]glucose
SAxTGO (23). The abovecalculationsarebased on theassumptionsthat (a) thespecific activity of hepatic UDP-glucose reflectsthatof G6P and(b)
that complete detritiation of the
[3-3H]glucose
iscompleted at the level ofthe triose-phosphate pool. In separatestudiesin 6-h fastedconsciousrats wehaveverifiedthecorrespondenceof4C specific activi-tiesofthehepatic UDP-glucoseandG6Paftertheinfusion of [U- 4C
]-lactate. The absence oftritiatedlabel at the levelofthehepaticPEP was
confirmedin all thestudies, indicatingthat allofthe3Hlabelislost in themetabolism ofthe[3-3H]glucosethroughglycolysis. However,a smallpercentage of
[3-_3H]
glucosemay alsoenterthe pentose phos-phatepathway, undergo detritiationattheribulose-5-phosphate level,and reenter in the G6P pool. In fact,for each molecule of
[3-3H]-glucose that enters the pentosephosphate pathway (generating3CO2
and I glyceraldehyde-3-phosphate),about two otherdetritiated mole-culesof[3-3H]G6Pmay reenter the G6Ppoolandtherefore dilute its specificactivity. Thus, thistracerwilldetermineanunderestimationof
the glucosecycling which is in direct proportion totherate ofthe pentosephosphate cycle(23). Similarly,if there wassignificant fruc-tose-6-phosphatecycling duringthe presentstudies,theuseof [
3-3H]-glucose (rather than[6-3H]glucose)would cause anunderestimation oftherateofGC andanoverestimationofHGP.Althoughwehave
previouslyshownsimilarrelative UDP-glucosespecific activitiesafter
[3-3H]glucose
or [6-3H]glucose infusions under hyperinsulinemicandhyperglycemic conditions (23), thiswas not verifiedunder the presentexperimentalconditions. It should also bepointedout that this tracermethodologywill measurePEP-gluconeogenesis,which repre-sents thegreatmajorityofthegluconeogenicfluxundermost
experi-TableI. General CharacteristicsofControl and DiabeticRats
Group Control Diabetic
n 35 10
Body wt(g) 322±7 305±12
Fasting plasmaglucose(mM) 5.6±0.2 7.2±0.4*
Fastingplasma insulin
(gU/ml)
26±3 20±6Nonfasting plasma glucose(mM) 7.7±0.4 18.5±1.3*
Nonfastingplasma insulin
(MU/m)
51±3 26±5*TableII. Steady-StatePlasma Glucose, Insulin, and Glucagon ConcentrationsDuring the In Vivo Studies
Group Glucose Insulin Glucagon
mM sU/ml pg/mi
1. Control(C) 8.0±0.2 46±5 118±11
2. C/euglycemic 7.9±0.2 43±3 115±14
3A. C/hyperglycemic 18.1±0.2* 45±3 103±10
3B. C/hyperglycemic 18.0±0.2* 24±4* 106±19
4. 90%
pancreatecto-mized rats 18.4±2.3* 27±6* 119±16
* P<
0.01
vs.2-C/euglycemic.mental conditions. However, itmaysignificantly underestimatethe overall gluconeogenic rate underexperimental conditions in which non-PEP-gluconeogenesisissignificantlyincreased.
Results
Basalmetabolicparameters.Therewere nodifferences inthe meanbodyweights between control and diabeticrats(Table I). Both the fasting (P < 0.05) and nonfasting (P < 0.01) plasma glucoseconcentrations duringthe 2-wkperiod before thein vivo studiesweresignificantly elevated in the diabetic comparedtothe controlgroup(TableI).The fastingplasma
insulin concentrations were similar, while the nonfasting plasmainsulin concentrations weresignificantlydecreasedin
diabetic(P<0.01 ).
Effect
of hyperglycemia onhepatic glucosefluxes and he-patic substrate levels and enzyme activities in control rats (study protocols 1,2, and3A; Tables II-V; Figs. 2 and 3). The steady-state plasma glucoseconcentrationwaskeptatthe basal level(- 8 mM) throughout the study protocols1 and 2,whileitwasraised by -10mMduringthe study protocols 3A and B. In studies 2 and 3A the plasma insulin concentration was maintained atthe levels measured during study protocol 1,
usingcombined somatostatin and insulin infusions. Instudy 3B the infusionof somatostatin ledto adecrease inthe plasma insulin concentrationto alevelsimilarto thatof diabeticrats (Table II). Plasma glucagonconcentration wassimilar in the
experimentalgroups(Table II).
TableIIIdisplays the[3H]-and
['4CIUDP-glucose
specificactivities,the
[14C]
PEPand the[3H]glucose
specific activities whichare used to calculate thecontribution of plasma-derivedglucose ("Direct" in Table III) and thecontributionof PEP-derived G6P("Indirect"inTable III) to the hepatic G6P pool. TheUDP-galactose specific activities confirmedthe values ob-tainedwithUDP-glucose, suggesting rapidandcomplete isoto-picequilibrationbetween the twointracellularpools.
HGPwassimilarinthetwoeuglycemicstudies(10.6±0.6 and 10.6±0.2 mg/kg- mininprotocols 1 and 2, respectively), but it was significantly suppressed by hyperglycemia, in the presenceof basal insulin (4.5±0.3vs. 10.6±0.2 mg/kg min,P <0.01; protocol 3Avs.protocol2; Fig.2A) and with hypoin-sulinemia (6.2±0.4 mg/kg min, P < 0.01; protocol 3B vs. protocol 2). TGO (G6Pase flux; Fig. 2 B)wassimilar in the presence of normoglycemia (protocol 1 = 13.4±0.6 mg/ kg.min and 2 = 14.2±0.7 mg/kg.min) or hyperglycemia (protocol 3A= 13.5±0.5mg/kg min;P=NS). Under basal insulin conditions, glucose cycling (Fig. 2; C) was markedly enhancedbyhyperglycemia (9.0±0.6vs.3.6±0.6mg/kg.min,
P <0.01 protocol 3Avs. 2). Under basal insulin conditions, hyperglycemia ledto a decreasein thehepatic G6P and UDP-glucoseconcentrations (TablesIVand V). The reductionin thehepaticG6P concentrationwassignificant between proto-cols 3A and 1, but didnotreachstatisticalsignificancebetween groups 3A and 2. Fig. 3 displays the contribution of hepatic gluconeogenesis(A)and glycogenolysis (B)to HGP in the pres-enceofnormo-orhyperglycemiaand basalinsulinemia. Hyper-glycemia ledto amarkedreduction in hepatic glycogenolysis
(0.4±0.2vs. 5.8±0.5 mg/kg- min, P< 0.01 study 3Avs. 2), whereasgluconeogenesiswasonlyslightly decreased (4.0±0.5 vs. 4.8±0.5 mg/kg-min,P=0.07 study 3A vs. 2).Consistent with this finding,theresidual hepaticglycogen at the endofthe infusion studieswasgreaterin thehyperglycemic than inthe normoglycemic studies(Table IV; P <0.01 study3A vs. 2). Theeffect of the plasma glucose concentrationonhepatic gly-cogensynthase and phosphorylase activities is displayed in
Ta-bleIV.
FVo.I
wasmeasuredasthemeanof the valuesobtainedatthreeUDP-glucoseconcentrations withinthephysiological range for rat liver (36), i.e., 0.125, 0.250, and 0.500 mM. Under basal insulinconditions, hyperglycemia activated the hepatic glycogen synthaseanddecreased the activity of the gly-cogen phosphorylase a (P< 0.01 study 3A vs. 2). Hepatic glucokinase and G6Pase activities wereunchanged by short-termhyperglycemia (TableV).
Comparisonofcontrol and diabeticratsundersimilar hy-poinsulinemic and hyperglycemic conditions(study protocols 3B and 4; Figs. 4 and 5). HGP (18.9±1.4 vs. 6.2±0.4 mg/ TableIII. Substrate
Specifc
Activities atthe Endofthe[3-3H]Glucose-[U-'4C]Lactate
InfusionIndirect Direct
Group ['4C]PEP ['4C]UDPGlu ['4CJUDPGal UDPGlu UDPGal [3H]Glu [3H]UDPGlu [3H]UDPGal UDPGlu UDPGal
dpm/nmol % dpm/nmol %
1. Control(C) 9.9±1.9 7.9±1.8 7.7±2.1 38.7±3.1 37.9±3.4 41.5±5.8 9.0±1.3 9.5±1.4 20.9±1.2 22.7±1.2
2. C/euglycemic 11.3±1.4 7.6±1.0 7.6±1.2 34.0±2.2 34.3±2.7 39.5±3.9 9.9±0.9 10.2±1.3 25.4±1.6 25.7±1.8
3. C/hyperglycemic 9.7±1.3 6.5±1.1 6.4±0.8 33.0±3.3 33.4±3.4 14.3±1.3 9.0±1.1 9.3±1.0 60.7±2.7 62.0±3.0
4. PANX 10.1±1.9 10.8±2.5 10.6±3.0 52.5±4.8 51.4±5.0 15.7±1.4 4.5±0.3 4.8±0.5 28.9±2.0 29.4±1.5
Abbreviations: UDPGlu, uridinediphosphoglucose; UDPGal=uridinediphosphogalactose; Glu, plasma glucose; Direct=percentage of the
he-paticG6Ppool derived from plasma glucose, calculatedastheratio of thespecificactivities of
[3H]UDPGlu
(Glu)or[3H]UDPGal(Gal)and[3H]Glu;
Indirect=percentage of thehepaticG6P pool derived fromPEP-gluconeogenesis,calculatedastheratioof thespecificactivities ofkg min; P < 0.01 protocol 4 vs. 3B; Fig. 4 A) and TGO (26.5±1.5 vs. 15.8±0.9 mg/kg- min; P < 0.01, group 4 vs. 3B; Fig. 4 B) were markedly increased in diabetic compared to control rats. In thepresence of similarplasmainsulin and glu-cose concentrations, GC (Fig. 4 C) was significantly less in diabetic animals (7.6±0.7 vs. 9.6±0.7 mg/kg.min,P <0.01 group 4 vs. 3B). ThehepaticUDP-glucoseconcentrationwas similar indiabetic and control rats(Table IV), but the G6P concentration(Table V) was significantly lower in diabetic rats (P < 0.01, group 4 vs. 3B). Fig. 5 displays thecontribution of
Az 15
0
0
0C -10
en --m
0 Ng
wE
CI)b
E20
0.
B0
E 15
o 10
en Cl
E 5
0
-J
I-E-I
cnl
0)
C81
z
0-C)
-J
C.'_
j
zControl(C)
C+SRIF+Euglycemia C4SRIF+Hyperglycemia/A
Control(C)
C+SRIF+Euglycemia C+SRIF+Hyperglycemia/A
Control(C)
C+SRIF+Euglycemia C+SRIF+Hyperglycemia/A
*
l1-Figure 2. Effect ofhyperglycemia,underbasalinsulinconditions,on
(A)HGP, (B) TGO, and (C) GC in control rats. *P< 0.01 vs.C
+somatostatin(SRIF)+euglycemia.
TableIV. Fractional Velocity (GSFVO.) of Hepatic Glycogen
Synthase, Activity ofAMP-independent (Pho a) Form ofHepatic Glycogen Phosphorylase, Hepatic Uridinediphosphoglucose (UDPGlu), and Glycogen Concentrations (Glycogen) at the End
oftheIn Vivo Studies
Group GSFVo, Phoa UDPGlu Glycogen
% mol/g*min nmol/g JAmotlg
1. Control(C) 29.4±1.9 14.1±1.4 358±11 88±13
2. C/euglycemic 28.5±1.4 17.6±2.2 390±14 84±5
3A. C/hyperglycemic 40.2±1.8* 9.7±0.9* 259±10* 139±11* 3B. C/hyperglycemic 38.6±2.9* 11.3±0.7* 276±12* 130±9*
4. PANX 17.9±1.1t 9.1±1.0* 279±24* 116±12*
TheFVOIfor thehepatic glycogen synthase represent themeanof the
activity ratios in the presence of threephysiologic UDPGlu
concen-trations(0.125, 0.25, and 0.5 mM)with 0.11 or7.2mMG6P.The
Phoais thehepatic glycogenphosphorylaseactivityin the absence of AMP. * P < 0.01 vs.2-C/euglycemic; tP<0.01 vs. 3B-C/hy-perglycemic.
hepaticgluconeogenesis (A) and glycogenolysis (B) toHGP. Both enhanced hepatic glycogenolysis (4.9±1.0 vs. 0.9±0.3 mg/kg min, P< 0.01 group 4 vs. 3B)and gluconeogenesis (13.9±1.8 vs. 5.2±0.3 mg/kg min,P<0.01 study4vs. 3B) contributedtothe marked increase in HGP in the diabetic rats. The TGO (16.0±1.1 mg/kg.min), the gluconeogenesis (5.6±0.4), and the GC (9.8 mg/kg min)weresimilar in the subgroupof controlrats,receiving protocol3B in combination with[2-3H
]glucose
ratherthan[3-3HIglucose.
Despite lower plasma insulin concentrations, hyperglyce-mia activatedthe hepatic glycogen synthase and decreased the activity ofthe glycogen phosphorylase a (Table IV; P < 0.01 study 3Bvs.2) incontrol rats. The activity ofthe hepatic glyco-gen synthase was severely impaired in diabetic compared to controlrats(Table IV). Hepatic glucokinase activity was mark-edlydecreased (abouttwofold) and G6Pase activity increased (about twofold) in diabetic compared to control rats (Table V).
Correlations. Inthe control groups (I,II, IIIAand B), the HGP was inversely correlated to the rate of glucose cycling (Fig. 6, left), whilethere was no significant correlation with the totalglucose output. In the control groups, the HGP was highly correlated totherateof hepaticglycogenolysis, while a much weaker correlation was present between the HGP and the rate of gluconeogenesis (Fig. 6,right). In diabetic rats (group IV), the HGP was highly correlated to the TGO (r2 = 0.904; P
< 0.01) and to the rates ofgluconeogenesis (r2 = 0.624; P
< 0.05), while there was no significant correlation between HGPand glucosecycling( r2 = 0.204; P=NS) and glycogenoly-sis(r2=0.014;P=NS).
Discussion
PostabsorptiveHGP and GC are markedly elevated in hyper-glycemic NIDDM individuals despite normal or elevated plasma insulin(2-5, 20).Becausetheenhanced contribution ofgluconeogenesishas beenshowntocompletely accountfor the increased HGP(3,5),severalinvestigatorshavesuggested
that theoveractivityof the
gluconeogenic pathway
is thepri-marydefectresponsibleforfastinghyperglycemiain NIDDM
TableV. Activity ofHepatic Glucokinase (GK) and G6Pase andHepatic G6P and Glucose(Glu)Concentration attheEnd ofthe In Vivo Studies
Group GK (7 mM) GK(18mM) G6Pase G6P Glu
Asmol/g/min nmol/g Amol/g
1. Control (C) 3.8±0.5 6.3±0.7 4.2±0.7 564±62 7.8±0.2
2. C/euglycemic 4.0±0.4 6.4±0.8 4.7±0.5 522±28 7.5±0.3
3A. C/hyperglycemic 3.9±0.3 6.3±0.9 4.1±0.7 494±18 16.1±0.3*
3B. C/hyperglycemic 3.6±0.4 5.9±0.5 5.2±0.4 465±24* 15.9±0.2*
4. PANX 1.4±0.1*8 2.3±0.1*8 10.4±1.1* 342±23*t 16.6±2.1*
Theactivity of the hepatic glucokinase is calculated in the presence of 7 or 18 mM glucose to approximate the in vivo plasma glucose concen-trationin the fourexperimentalgroups.The G6Pase activity is measured in the presence of a physiologic(1 mM)G6P concentration.
*P<
0.01
vs.2-C/euglycemic; tP<0.01
vs.3B-C/hyperglycemic.(2, 3, 5). However, the comparison of diabeticand control
subjectsunderdifferentglycemic conditionsappears to under-estimate thepotentialroleofthe plasma glucoseconcentration
per sein theregulationof the pathways of HGP.
Although a major role for the circulating plasma glucose concentration in theregulation of HGP has long been recog-nized ( 12-18), the mechanism(s)bywhichchangesinglucose level per se inhibit HGPand the impactoftheconcomitant hyperglycemia intheinterpretation of hepaticglucosefluxesin diabetes havenotbeendelineated. Thus, thepresentstudy ex-amined the mechanisms by which hyperglycemia regulates HGP in consciousrats.
Themajor metabolic steps which regulate HGP may be
schematicallydivided intotwocategories (see Fig. 1): (a)
for-A
.99r
E 6
cn Cl)
z
wo.
z
0
-j
0 0
IControl(C)
C+SRIF+Euglycemia
C+SRIF+HyperglycemiaA
B
Control
(C) *C|SRIF+Euglycemia.X' IC+SRIF+Hyperglycemia/A
U)
0
z3
0
0*
Figure 3. Effect of hyperglycemia, under basal insulin conditions, on theratesof (A) gluconeogenesis and (B) hepatic glycogenolysis. *P
<0.01 vs.C+SRIF+euglycemia.
mation ofhepatic G6P; and (b) phosphorylation/de-phos-phorylation ofG6P. The hepatic G6P pool can derive form three major sources: gluconeogenesis, glycogenolysis and plasma glucose,i.e.,GC. Thetwokeyenzymes whichregulate
thenetdephosphorylation of the hepaticG6Pto glucoseare glucokinase and G6Pase. The relative contribution ofplasma
glucose andgluconeogenesisto thehepatic G6Ppool canbe directlymeasured bytracermethodology. Infact,the propor-tion of the G6Ppoolwhich isderived fromplasma glucosecan
be calculatedasthe ratio ofspecific activities of tritiated he-patic UDP-glucose andplasma glucoseafter [ 3-3H] glucose in-fusion and the percentage of the G6P pool which is formed through PEP-gluconeogenesis can be calculated as the ratio
of '4C-labeled UDP-glucose and PEP after [U- '4C]lactate infusion (23).
To examine the effect ofhyperglycemia, independent of
hormonalsignals,onhepatic glucose fluxes,wecombined the infusion of somatostatin and insulin, with either euglycemia
(- 8 mM)orhyperglycemia(- 18 mM).Inthepresenceof
euglycemia, the infusion of somatostatin and insulin did not causeanysignificantalteration inHGP,TGO, GC, gluconeo-genesis, or glycogenolysis (Figs. 2 and 3). Hyperglycemia
caused a 58% decrease in HGP comparedtothe euglycemic condition, while total G6Pase flux was not significantly de-creased. Thisdiscrepancybetweentotal andnetglucoseoutput was due to a 2.5-fold elevation in glucose cycling. Such an increase inthecontribution ofplasma-derivedglucose carbons totheG6Pasefluxwasproportionaltothe - 2.3-fold increase
in the circulatingglucose concentration, suggesting that it is largely mediated through the increase in substrate for hepatic glucokinase. This interpretation of the in vivo fluxes is sup-ported by the enzyme activities and substrate concentrations
displayedin TableV. Infact,a2.2-fold increase inthehepatic glucose concentration didnotchangetheglucokinase andthe G6Paseactivities. However, the glucokinase activity measured in thepresenceof 18mMglucosewasincreasedby about two-fold comparedtotheenzyme'sactivityin the presenceof7mM
glucose, suggestingthat theincreased availability ofsubstrate ratherthan anincrease in enzymeactivityaccounted forthe enhancedcontribution ofplasma-derivedglucose to the G6P pool. Similarly, the lack of effect ofacute hyperglycemiaon G6Pase activity andhepatic G6P concentration is consistent withthelackofglucose-inducedinhibition of in vivo total
glu-cose output. The ability of hyperglycemiaper se to suppress HGPwasconfirmedin a subgroupofrats in whichinsulinwas notreplaced duringthehyperglycemicstudy.Despitethe
C+SRIF+Hyperglycemia/B
PANX
C+SRIF+Hyperglycemia/B
PANX
..
.:.. ....:...
:.:
*.:
....
89%ofthedecline inHGP. Although gluconeogenesiswasalso slightlydiminished( 16%) comparedtothe euglycemic control, thisdeclinewasinconsistently observedanddidnot reach sta-tisticalsignificance.Similarly, in thepresenceof hypoinsuline-mia, the hyperglycemia-induced suppressionof hepatic
glyco-genolysisaccountedforallofthe decrease in HGP, while gluco-neogenesiswasmarginally increased(by8%).Consistent with the in vivo metabolic fluxes, hyperglycemia was associated with enhanced activation of hepatic glycogen synthase, de-creased glycogen phosphorylaseaactivity, and enhanced
he-paticglycogen content compared to the euglycemiccontrols (Table IV).
Thus, anacuteelevation in theplasma glucose concentra-tioninhibitsHGP mainly through the stimulation ofhepatic glycogen synthase and the inhibition of hepaticglycogen phos-phorylase, while the activities ofbothglucokinaseand G6Pase are not affected. However, a marked substrate-mediated in-creaseinglucokinase flux(asreflectedby the enhanced
contri-bution of plasma-derivedglucosetototal G6Pase flux)allows theinhibitionof HGP in the absenceofanychange inG6Pase flux andmaycontributetotheinhibition of HGPby maintain-ingtheG6Pconcentrationnearthe basal leveldespitethe com-pleteinhibition ofglycogen breakdown (Fig.1). Thismay facil-itate the effectson the activityofhepatic glycogen enzymes, and perhaps more importantly, may contributeto maintain
carbon flowthroughhepatic glycolysisandprevent anincrease
ingluconeogenic flux.Although suchascenariois speculative, it is inkeepingwithrecentexperimentalevidence supporting
A
Z16CD CM12
E xcn
LU8
0
0
4. 0
C)
CD G.1
*C@SRIF+Hyperglycemia/B T
[3PANX.-.---
-.---|---....
Figure4.(A) HGP, (B) TGO,and(C)GCincontrol and 90%
pan-createctomizeddiabeticrats,inthepresenceof similarconditions of hyperglycemiaand moderatehypoinsulinemia (controlrats received
SRIFwithout insulin replacement; protocol 3B). *P < 0.01 vs. C
+SRIF+hyperglycemia/B.
ence ofhypoinsulinemia, HGP was inhibited by 42%
com-paredtoeuglycemiccontrol.Hyperglycemiaincreasedglucose cycling by 2.7-fold, and slightly (by 12%) TGO, similarlyto
what observedinthepresenceof basalinsulin.
The decrease in HGP induced byacutehyperglycemiacan betheconsequence ofdiminishedflux fromglycogen,
gluco-neogenesis,orboth.Inthepresenceofbasalinsulin
concentra-tions, theincrease incirculating glucosecausedanalmost
com-plete inhibition of hepatic glycogenolysis,whichaccountedfor
B b C+SRIF+Hyperglycemia/B
* .p4 PANX
E4
U)
C/i
-j
0
Z 2
-0 C)
-j
00 _
Figure5. Rates of(A) gluconeogenesisand(B) hepatic glycogenolysis
incontrol and90%pancreatectomizeddiabeticrats,inthepresence
of similarconditions ofhyperglycemiaand moderate
hypoinsuline-mia(controlratsreceivedSRIF without insulinreplacement; protocol 3B).*P<0.01 vs.C+SRIF+hyperglycemia/B.
A
21z
0
F
0
0
0
a. .5
wE
OY
C,76
a. U'a
DE
z
0
7
-B
.'
-3
30
E
I-:3 20
C.
I-0
CL
U)
0
8.
10
-J -J
I-0
C
E12E 0
-.- 8
-z
-0
4-0 C)
-j
thepresence ofanhomeostaticregulation of hepatic glucose fluxesatthe level of the G6Ppool (7-11).
Incontrol animals, the changes in HGP induced by changes
inplasmaglucoseconcentrationcorrelatedpositivelywith the rate ofglycogenolysis and negatively with the GC (Fig. 6). Thesecorrelationsappear tofurthersupportthe notion that in normalratstheeffect of hyperglycemiaonHGP ismainly me-diatedthrough its effectsonhepatic glycogen breakdown and onglucosephosphorylation.
Since HGP is the maincauseofpostabsorptive hyperglyce-mia indiabetes, it is difficulttodissociate the enhanced HGP andhyperglycemiainthe diabeticstate.Thus,in theattemptto
delineate the mechanism(s)responsible for enhanced HGP in diabetes, comparisonshave been made between normoglyce-mic controls andhyperglycemicdiabeticsubjects (3-6). Due tothespecific effects ofhyperglycemiaper se onhepatic
glu-cosefluxes,the relative contribution of various metabolic
path-ways totheenhanced HGP in diabetesmaybeinfluencedby either chronic alterations in hepatic glucose metabolism or
acute metabolic conditions at the time of the experimental
study.
Similarto what has been previously reported in human NIDDM(3, 5),diabeticrats,in thepresentstudy,were
charac-terizedbyamarked increase in HGP(- 2-fold), glucose
cy-cling(2.7-fold),andgluconeogenesis(2.7-fold),while therate
of hepatic glycogenolysiswassimilartothat in control animals (group 4 vs. 1). Furthermore, theincreament in HGP, above control levels, could be accounted for entirely by the marked increase in gluconeogenicflux.
However, whendiabetic and control rats were compared under similar conditions ofhyperglycemia and moderate hy-poinsulinemia, the relative contribution of the pathways of he-patic glucose metabolismtothe marked increase in HGPwere dramatically different. Although, in the diabeticgroup, HGP (3-fold) and TGO(1.7-fold)weremarkedlyincreased above control levels, therateof GCwasdecreasedby 21% (Fig. 4). Furthermore, both increased gluconeogenesis (+8.7 mg/ kg. min) and hepatic glycogenolysis (+4.0 mg/kg- min) con-tributedtotheincreased HGP. Thesevereimpairment in
he-patic glucokinaseandglycogensynthaseactivityand the about
twofold increaseinG6Paseactivityseem tosupportthe
multi-factorial origin and complexity of the alterations in hepatic glucosehomeostasiswhichsustain thestrikingincrease in
post-absorptiveHGP andfastinghyperglycemia.
Finally,thehepaticconcentration ofG6Pmayhelpto fur-therspeculateontherate-determining step(s) responsiblefor theenhancedhepatic glucoseoutputinthis diabeticratmodel. Infact,if therateofformationof G6P exceeded therateof its
dephosphorylation oneshould be abletodemonstratean in-crease inthe-steady-state level of G6P in the liver of diabetic
B
0 0 0 0
0 0
00
0 0
'09
0
0o
oe9
(b 0
3 6 9
Hepatic
Glucose
Un
Z
0
c
cm 0
>u
- 1
12 15
Prodctlon
R-2 = 0.645 0
0
0
8
0 0
0
0
0
3 6 9 12 15
Hepatic Glucose
Productlon
10
R-2 - -0.714 0
0
0
(9000
._ 8
U, 0) c
0 6
C 0
_ 4
00
R-2 = 0.11 8
0 0
0 00
0
3 6 9 12
Hepatic Glucose Production
Figure6. (A) Correlation between the rates of HGP and TGO (upper panel) andglucose/glycogen
cy-cling (lowerpanel) in control rats. (B) Correla-tion between the rates of HGP and glycogenolysis
(upperpanel)and gluco-- neogenesis (lower panel) 15 in control rats.
A
271 R-2 = 0.010
06b
0.
0
In
o 18.
u
0
c U
u
In
0 u
LO
3 6 9 12 1
rats. On the contrary, if therate of net dephosphorylation of
hepatic
G6Ptoglucose
exceeded therateofitsformation (fromgluconeogenesis
and glycogenolysis), the steady-state G6Pconcentrationshould be lower indiabeticthan in controlliver. Sincethe
hepatic
G6Pconcentrationwasconsistentlylower indiabeticvs.control
animals,
thismay suggest thatthe enhancedrate of
dephosphorylation
of G6P, rather than the increasedratesofits
formation,
representedtherate-determining step for the increased HGP in the diabetic group.Furthermore,
the markeddifferencesobserved in the correlates of HGP in thediabetic
compared
with the control groups suggest that in-creasedfluxthrough
G6Pase is themajor
determinantoftheincreasedHGP inthisgroup and that theacute
regulation
ofHGP
by glucose
isimpaired
after chronichypoinsulinemia
andhyperglycemia.
Theinvestigation
of thelatterissuewillrequire
furtherstudiesinwhich
hepatic
glucose
fluxesindiabeticani-malsareexaminedin the presenceof either
normoglycemia
orhyperglycemia.
In
conclusion,
ourresults indicatethat(a)
hyperglycemia,
independent
ofchanges
in insulinconcentration,
causes amarked inhibition ofHGP
through
thesuppression
ofnetgly-cogenolysis
andthrough
theincreaseinglucokinase flux,
whilethefluxes
through
gluconeogenesis
andG6Pasearenotsignifi-cantly
affected;
and(b)
undersimilar conditions ofhyperglyce-miaand
hypoinsulinemia,
increasedglycogenolysis
andgluco-neogenesis
and decreased GC all contribute tothe enhanced HGP indiabetic animals.Acknowledaments
Theauthors thank Dr. Harry Shamoon (Albert Einstein College of
Medicine) for critical readingof themanuscriptand Drs. S. Pilkis and A.Lang(StonyBrookMedicalSchool)fortheir assistance in setting up the glucokinase and glucose-6-phosphatase assays. We also thank DonnaBanduchforherexcellenttechnical assistance.
This work was supported by grantsfromtheNational Institutes of Health (R029-DK 42177), the Juvenile Diabetes Foundation (No. 1911127), and theAmericanDiabetes Association and by the Albert
Einstein DiabetesResearchandTrainingCenter (DK20541).
References
1.Bell,P.M.,R.G.Firth,and R.A.Rizza. 1986.Assessmentof insulin action ininsulin-dependentdiabetes mellitususing[6'4C]glucose,[33HIglucose,and
[23H~glucose.
J.Clin.Invest.78:1479-86.2.DeFronzo,R. A.Thetriumvirate. 1988.B-cell, muscle,liver:acollusion
responsibleforNIDDM.Diabetes. 37:667-687.
3. Magnusson, I.,D. L. Rothman,L. D.Katz, R. G.Shulman,and G. I.
Shulman. 1992. IncreasedrateofgluconeogenesisintypeIIdiabetes mellitus:a
'3Cnuclearmagneticresonancestudy.J.Clin.Invest.90:1323-1327.
4.Consoli,A.,N.Nurjhan,J. J.Reilley,D. M.Bier,and J. E.Gerich. 1990.
Mechanismof increasedgluconeogenesisinnon-insulin-dependentdiabetes
mel-litus. J.Clin.Invest.86:2038-2045.
5.Consoli,A., N.Nurjhan,F.Capani,and J. E. Gerich. 1989.Predominant
roleofgluconeogenesisin increasedhepatic glucoseproductionin NIDDM.
Dia-betes.38:550-557.
6.Zawadzki,J.K.,R. R.Wolfe,D. M.Mott,S.Lillioja,B. V.Howard,andC.
Bogardus. 1988. IncreasedrateofCoricyclein obesesubjectswith NIDDM and
effect ofweightreduction.Diabetes.37:154-159.
7.Diamond,M.P.,R.C.Rollins,K. E.Steiner,P. E.Williams,W.W.Lacy,
andA. D.Cherrington. 1988.Effect of alanine concentrationindependentof
changes ininsulin andglucagonon alanine andglucose homeostasis in the
consciousdog.Metab. Clin. Exp. 37:28-33.
8.Jennsen,T.,N.Nurjhan,A.Consoli,and J. E.Gerich. 1990.Failure of substrate-inducedgluconeogenesistoincrease overallglucoseappearance in
nor-mal humans. J.Clin.Invest.86:489-497.
9.Jahoor, F.,E. J.Peters,and R.R.Wolfe. 1990. Therelationshipbetween
gluconeogenicsubstratesupplyandglucose productionin humans.Am.J.
Phys-iol.258:E288-E296.
10.Wolfe,R.B.,F.Jahoor,and J. H. F. Shaw.1986. Effect ofalanine infusion
onglucoseandureaproductioninman.J. Parenter.EnteralNutr. I11:109-11 1. 11.Puhakainen, I.,V.A.Koivisto,and H.Yki-Jarvinen. 1991.Noreduction in totalhepaticglucoseoutputbyinhibition ofgluconeogenesiswithethanol in
NIDDMpatients.Diabetes. 40:1319-27.
12.Soskin, S.,and R.Levine. 1946.CarbohydrateMetabolism.Universityof ChicagoPress,Chicago.247-263.
13.Glinsman,W.H.,E. P.Hem,and A.Lynch. 1969.Intrinsicregulationof
glucoseoutputbyratliver.Am.J.Physiol.216:698-703.
14.Sacca,L.,P. E.Hendler,and R.S.Sherwin.1979.Hyperglycemiainhibits
glucoseproductioninmanindependentofchangesinglucoregulatoryhormones.
J.Clin. Endocrinol. Metabol. 47:1160-1163.
15.Shulman,G.I.,J. E.Lijenquist,P. E.Williams,W. W.Lacy,andA.D.
Cherrington. 1978. Glucosedisposalduring insulinopeniasomatostatin-treated dogs:therolesofglucoseandglucagon.J. Clin.Invest.62:487-491.
16.Shulman,G.I.,W. W.Lacy,J. E.Lijenquist,U.Keller,P.E.Williams,
andA. D.Cherrington.1980. Effectofglucose, independentofchangesininsulin
andglucagon secretion,onalanine metabolism intheconsciousdog.J. Clin.
Invest.65:496-505.
17.Bell,P.M.,R.G.Firth,and R. A.Rizza. 1986.Effects ofHyperglycemia
onglucoseproductionandutilizationin humans:measurementswith[23H]-, [33H]-,and[6 '4C]-glucose.Diabetes. 35:642-648.
18.Ader,M.,G.Pacini,Y. J.Ysug,and R. N.Bergman.1985.Importanceof
glucose perse tointravenousglucosetolerance: comparisonoftheminimal
modelpredictionwithdirectmeasurements.Diabetes. 34:1092-1103.
19.Karlander,S.,A.Roovete,M.Vranic,andS.Efendic.1986.Glucose and
fructose-6-phosphate cyclein humans.Am.J.Physiol.25l:E530-E536.
20. Efendic, S.,S. Karlander,and M.Vranic. 1988.MildtypeIIdiabetes markedlyincreasesglucosecyclingin thepost-absorptivestateandduringglucose infusionirrespectiveofobesity.J.Clin.Invest.81:1953-61.
21.Miyoshi,H.,G. 1.Shulman,E. J.Peters,M. H.Wolfe,D.Elahi,and R. R.
Wolfe.1988. Hormonal controlofsubstratecyclingin humans. J.Clin.Invest. 81:1545-55.
22.Wajngot,A.,A.Kahn,A.Giacca,M.Vranic,andS.Efendic. 1990.
Dexa-methasoneincreasesglucosecycling,butnotglucose production,inhealthy
sub-jects.Am.J.Physiol.259:E626-E632.
23.Giaccari,A.,and L.Rossetti.1992.Predominant roleofgluconeogenesis
in thehepatic glycogen repletionof diabeticrats. J.Clin.Invest.89:36-45.
24.Granner,D.K.,and T. L. Andreone. 1985.Insulin modulation ofgene
expression.Diabetes Metab.Rev. 1:139-170.
25.Foglia,V.G. 1944.Caracterhsticasde ladiabetesenlarata.Rev.Soc.
Argent.Bio.20:21-37.
26.Bonner-Weir, S.,D. F.Trent,and G.C.Weir. 1983. Partial pancreatec-tomy intheratandsubsequentdefectinglucose-inducedinsulinrelease.J.Clin.
Invest.71:1544-1553.
27. Rossetti,L.,andA.Giaccari. 1990.Relative contributionofglycogen synthesisandglycolysistoinsulin-mediatedglucose uptake:adose-response
eu-glycemic clamp studyin normal anddiabeticrats.J. Clin.Invest.85:1785-1792. 28.Rossetti,L.,D.Smith,G. I.Shulman,D.Papachristou,andR.A.
De-Fronzo. 1987.Correction ofhyperglycemiawithphlorizinnormalizestissue
sensi-tivitytoinsulinindiabeticrats.J.Clin.Invest.79:1510-1515.
29.Rossetti,L.,and M. R.Laughlin.1989.Correction of chronic
hyperglyce-mia withvanadate,butnotphlorizin,normalizes in vivoglycogen repletionand
invitroglycogensynthaseactivityindiabeticskeletalmuscle. J. Clin.Invest.
84:892-899.
30.Farrace, S.,and L. Rossetti. 1992. Hyperglycemiamarkedlyenhances
skeletal muscle glycogen synthase activity in diabetic, but not in normal
consciousrats.Diabetes. 41:1453-1463.
31.Thomas,J.A.,K. K.Schlender,and J.Larner.1968.Arapidfilterpaper
assayforUDPG-glycogen glucosyltransferase,includinganimproved
biosynthe-sisofUDP["'4CIglucose.Anal. Biochem. 25:486-499.
32.Rossetti, L.,S.Farrace,S. B.Choi,A.Giaccari,L.Sloan,S.Frontoni,and
M.S. Katz. 1993.Multiplemetabolic effects of calcitoningene-relatedpeptide
(CGRP)inconsciousrats:relationshiptoitsregulationofglycogen synthaseand
adenylate cyclase.Am.J.Physiol.
264:El-E1O.
33.Davidson,A.L., andW.J.Arion. 1987. Factorsunderlying significant underestimations ofglucokinaseactivityin crude liverextracts: physiological implicationsofhighercellularactivity.Arch. Biochem.Biophys.253:156-167.
34.Burchell, A.,R.Hume,and B.Burchell. 1988.Anewmicrotechniquefor
theanalysisof the humanhepaticmicrosomalglucose-6-phosphatasesystem. Clin. Chim.Acta. 173:183-192.
35.Michal,G. 1985. Methods ofEnzymatic Analysis.VolumeVI.N.U.
Berg-meyer, editor. VCHVerlagsgesellschaftmbH, Weinheim,FRG.191-198.
36. Giaccari, A.,and L. Rossetti. 1989. Isocratichigh-performanceliquid chromatographicdeterminationoftheconcentrationandspecific radioactivityof