[email protected]@generic 8158b284c5d3

Hormonal Control of KetogenesisBiochemical Considerations

J. Denis McGarry, PhD, Daniel W. Foster, MD

A two-site, bihormonal concept forthe control of ketone body production isproposed. Thus, ketosis is viewed as theresult of increased mobilization of freefatty acids from adipose tissue (site 1) tothe liver (site 2), coupled with simulta-neous enhancement of the liver's capacityto convert these substrates into aceto-acetic and \g=b\-hydroxybutyricacids. Theformer event is believed to be triggered bya fall in plasma insulin levels while thelatter is considered to be effected primar-ily by the concomitant glucagon excesscharacteristic of the ketotic state. Al-though the precise mechanism wherebyelevation of the circulating [gluca-gon]:[insulin] ratio stimulates hepaticketogenic potential is not known, activa-tion of the carnitine acyltransferase reac-tion, the first step in the oxidation of fattyacids, is an essential feature. Two prereq-uisites for this metabolic adaptation inliver appear to be an elevation in itscarnitine content and depletion of itsglycogen stores. Despite present limita-tions the model (evolved mainly from ratstudies) provides a framework for thedescription of various types of clinicalketosis in biochemical terms and may beuseful for future studies.(Arch Intern Med 137:495-501, 1977)

The fact that in certain instancesthe ketone bodies, acetoaceticand /8-hydroxybutyric acids, subservea critical physiological role in themammalian organism while in othersthey may lead to its death poses animportant problem both from thestandpoint of clinical medicine and

the regulation of intermediary metabolism. While the past two decadeshave brought major advances in ourknowledge of the overall biochemicalphysiology of the ketone bodies (seeWilliamson and Hems' and McGarryand Foster- for detailed reviews) anumber of gaps remain in our understanding of the control of the ketogen-ic process itself. Prominent in thisregard is the question of whichhormones exert primary regulatoryroles and how their effects can bedescribed in biochemical terms. Thematerial presented here is highlyselective and relies heavily on recentinvestigations carried out in our laboratory.

REQUIREMENTS FORDEVELOPMENT OF THE

KETOTIC STATEIt can now be stated with some

certainty that the ketotic staterequires for its development metabolic adaptations in two organ systems. As outlined in Fig 1 there mustbe increased delivery of free fattyacids from their storage site, adiposetissue, to the primary ketogenic organof the body, liver. A substrate transport requirement is readily evidentsince acetoacetic and /Miydroxybu-tyric acids arise almost exclusivelyfrom the hepatic oxidation of fattyacids. Less obvious, however, has beenthe additional requirement for asimultaneous major alteration in livermetabolism which allows the fattyacids taken up to be efficientlyconverted into ketone bodies ratherthan entering the normal pathway oftriglycride synthesis. Evidence forthis change was first presented in thestudies of Mayes and Felts3 who

perfused livers from fed (nonketotic)and fasted (ketotic) rats with radioactive oleic acid and traced the fate ofthe fatty acid among various metabolic products. It was observed thatthe quantity of substrate oxidized toacetoacetic and /3-hydroxybutyricacids was markedly enhanced in thelivers from fasted animals comparedwith the fed group; the converse wastrue as regards the amount of fattyacid incorporated into esterifiedproducts. While this switch in thepattern of hepatic fatty acid metabolism in livers from ketotic animals hassince been extensively confirmed byother groups,4" its biochemical basis isstill unclear and represents the singlemost challenging issue facing investigators in the field of ketogenesis andits regulation. In addition, we mustalso ask how the alteration in "metabolic set" of the liver (control site 2,Fig 1) is coordinated with the activation of lipolysis at the level of adiposetissue (control site 1, Fig 1) in situations of physiological ketosis, and howthe balance between the two eventsmight break down in conditions ofpathological ketosis.

HORMONAL FACTORS IN THECONTROL OF KETOGENESISAll ketotic states, whether physio

logical or pathological, are characterized by a relative or absolute deficiency of insulin. With insulin deficiency fatty acids are mobilized fromfat depots as an alternative fuel formost tissues. The fact that insulinadministration in vivo produces aprompt fall in plasma fatty acidlevels7 coupled with the well documented ability of this hormone topotently inhibit adipose tissue lipo-

From the Departments of Internal Medicineand Biochemistry, University of Texas HealthScience Center at Dallas.Reprint requests to Department of Internal

Medicine, University of Texas Health ScienceCenter at Dallas, Dallas, TX 75235 (Dr McGarry).

Downloaded From: http://archinte.jamanetwork.com/ by a University of California - San Francisco User on 03/08/2015

lysis in vitro8 strongly suggests thatthe elevation of plasma fatty acids inketotic states is the direct consequence of insulin deficiency per se.Can the same be said for the concomitant activation of the liver's fatty acidoxidation machinery? To approachthis question we infused fed, nonke-totic rats either with anti-insulinserum (AIS) or glucagon for periods ofup to three hours during whichsequential measurements of plasmaglucose and ketones were made. Subsequently, the ketogenic capacity ofthe livers was assessed by determining their ability to synthesizeacetoacetate and /3-hydroxybutyratewhen perfused with oleic acid." Thesalient features of these experimentsare summarized in Table 1. Controlserum had little or no effect on thelevels of plasma glucose and ketonebodies or on hepatic ketogenic capacity. The mild elevation in levels ofplasma fatty acids and decrease inlevels of liver glycogen which areevident were undoubtedly caused bythe stress of surgery and physical

restraint. As expected, insulin antibodies resulted in profound andsustained hyperglycemia and a sharpincrease in the concentration of plasma ketone bodies. This ketotic statereflected the fact that both plasmafatty acid levels and hepatic ketogeniccapacity had been markedly increased. In contrast, glucagon treatment produced only a transienthyperglycemia and did not increasethe concentration of plasma free fattyacids or ketone bodies. Despite thedifferent pictures produced in peripheral blood, hepatic effects of glucagonand anti-insulin serum were identical.Both agents resulted in major depletion of glycogen stores and equivalentincreases in ketogenic capacity. Ourinterpretation is that infusion ofglucagon triggered hepatic glycogeno-lysis. The resultant elevation of plasma glucose level stimulated insulinsecretion from the pancreatic /3-cell10which in turn corrected the initialhyperglycemia by enhancing the disposal of glucose in extrahepatictissues. Insulin also blunted lipolysis

in fat depots and thus deprived theliver of ketogenic substrate, preventing the appearance of peripheral ketosis. Nevertheless, despite the elevation in plasma insulin concentrationand the low plasma free fatty acidlevels, the liver had clearly switchedinto a ketogenic mode, suggesting acritical role for glucagon at site 2. Acorollary of this argument would bethat rapid elevation of the plasmafatty acid concentration should resultin a brisk increase in plasma ketonelevels in animals previously treatedwith glucagon as compared to thosereceiving control serum. This wasfound to be the case experimentally."To summarize, the key finding of

the above experiments was that a"ketogenic" liver could be obtainedfrom a nonketotic animal. Stated inanother way, the activation of sites 1and 2 of Fig 1 can be dissociated. Thisoccurs under circumstances wherecirculating insulin levels are eithernormal or increased so that site 1 isblocked, but where an excess ofglucagon is present to activate site 2.For ketosis to develop both sites mustbe "turned on." This requires a deficiency of insulin to activate site 1; theconcomitant hyperglucagonemia isthought to be the signal for activationof site 2. (We do not mean to implythat insulin plays no role in the regulation of site 2. The possibility existsthat the [glucagon]: [insulin] ratiomediates the changes at this site.) Itthus appears that insulin and glucagon constitute a bihormonal systemfor the overall control of the ketogenicprocess, as has been postulated by

Fig 1.Model for the regulation of ketogenesls.ADIPOSE B 00DTISSUE BL00D

F FA

LIVER

r

/TRIGLYCERIDESPHOSPHOLIPIDS

fKETONEBODIES.

Table 1.Effects of Anti-insulin Serum and Glucagon Infusions in Fed Rats*

Hours of Infusion

Plasma Glucose(mg/100 ml)

PlasmaKetones (rail) Plasma Free Hepatic Ketone Liver

Fatty Acids; Production" Glycogent(mM) (mole gm ' 30 min ') (mg gm ')

TreatmentNone 0.25 0.01 24 4 48.3 2.5

Control serum 150 1519

139 13712

0.20:0.01

0.24:0.03

0.25:0.02

0.24 0.54 0.10:0.02

26 3 31.6 6.0

Anti-insulinserum

17710

47845

48332

43838

0.26:0.02

0.92:0.10

1.46:0.17

1.510.16

1.85 0.13 87 9.6 2.0

Glucagon 17013

23625

20921

19221

0.27:0.04

0.26:0.04

0.310.04

0.27t0.02

0.37 0.04 75 3 3.8 0.8

"Animals weighing approximately 100 gm received an intravenous infusion of guinea pig serum containing 1.6-2.1 units of insulin antibody per milliliter,100 fil/min for 5 minutes, followed by 10 /il/min for the remainder of the experiment. Arterial blood samples were analyzed for glucose, ketone bodies, andfree fatty acids. After three hours livers were taken for glycogen determinations or were perfused with 0.7 mM oleic acid for measurement of their ketogeniccapacity. Values are means SEM for 4 to 8 animals in each group. (Data from McGarry et al.s)tDetermlned at the three hour time point.


EXTRA-MITOCHONDRIALSPACE

INNERMITOCHONDRIAL

MEMBRANE

MITOCHONDRIALMATRIX

Fatty acid - Co ASH Co ASH

Fatty acyl-CoA

a-6lycero- >.phosphate >1

T vCarnitine-

Fatty acylcarnitine-

TriglyceridesPhospholipidsetc

-Fatty acyl-CoA

-oxidation

Acetyl-CoA

Ketone bodies

Fig 2Major pathways for fatty acid metabolism in liver.

Unger and colleagues" '- for glucosehomeostasis. (Other hormones mayplay important modulating roles".However, insulin and glucagon are theprincipal mediators of the observedeffects.) Support for this formulationhas come from studies with somato-statin in insulin-dependent diabeticsubjects. Somatostatin is a potentinhibitor of glucagon release from thepancreas and has been shown toretard both the appearance of hyper-glycemia and ketonemia after withdrawal of insulin.14 It has now alsobeen shown that glucagon administration enhances the ketogenic responseto a fatty acid load in diabetic man.'1

INTRAHEPATIC FACTORSIN THE CONTROLOF KETOGENESIS

The next question we shall addressis whether "site 2" of Fig 1 can bedefined in more concrete terms. Forthis purpose it is necessary to examinein greater detail the possible fates of along chain fatty acid once it has beentaken up by the liver cell. Althoughsomewhat oversimplified, the majorfeatures of hepatic fatty acid metabolism are depicted in Fig 2. After activation to its coenzyme A derivative inthe extramitochondrial compartment

Table 2Relationship Between Ketogenic Capacity and Carnitine Content ofRat Liver*

HepaticHepatic Ketone Carnitine Content

State of Production (Free + Esterlfied)Animal (jimole gm ' 30 min ') (nmole gm )

Fed 26+3 103Fed, AlSt 3 hr 87 2 224 9Fed, glucagon 3 hr 87 5 269 21Fasted, 24 hr 118 8 287 27Alloxan diabetic 192 13 487 49

"Animals were treated as indicated. Each liver was then perfused with 0.7 mM olelc acid todetermine its rate of ketogenesis, after which a portion of the tissue was analyzed for its carnitinecontent. Values are means SEM for 6 to 8 animals in each group. (Data from McGarry et al.")tAntl-insulln serum.

the fatty acid may react with a-glyc-erophosphate to form esterifiedproducts (reaction 1) or with carnitinefor transport across the inner mito-chondrial membrane to the site of /3-oxidation in the mitochondrial matrix.The transfer mechanism involves thesequential action of carnitine acyl-transferase I on the outer aspect ofthe inner membrane (reaction 2) andcarnitine acyltransferase II on theinner aspect of the membrane (reaction 3).1(i As indicated earlier, the basicdifference between livers of low andhigh ketogenic potential is that in theformer a fatty acid load is disposed ofprimarily through reaction sequence 1whereas in the latter a much greater

fraction traverses reactions 2 and 3. Itwas early suggested that the dominant factor governing the partitioning of fatty acids between these twopathways was the esterifying capacity of the liver which was considered tobe depressed in ketotic states becauseof inadequate levels of a-glycerophos-phate.'17 However, more recent studies revealed no correlation betweenthe content of a-glycerophosphate inliver and its ketogenic capacity.4'*Moreover, it was shown that in liversfrom ketotic animals blockade of reactions 2 and 3 by ( + )-acylcarnitines(inhibitors of carnitine acyltransfer-ases) resulted in the immediate shunting of fatty acid away from the oxida-


Table 3.Metabolie Characteristics of Neonatal vs Fetal, and Fasted vs FedAdult Rats

Metabolic Measurement

Directional Changeat Birth in

Developing Rat oron Fasting in Adult Rat Reference

Primary fuel for tissues Glucose - Fat 28, 29Circulating [glucagon]:[insulin] ratio t 30Hepatic glycogen stores 30Hepatic fatty acid oxidation capacity 31, 32Blood ketone body concentration 33

Table 4.Relationship Between Ketogenlc Capacity and the Levels of Glycogenand Carnitine in Livers From Female Rats*

Type ofRat

Ketone Body Production(iimole gm Liver ' 30 min ')

LiverCarnitine

(nmole gm ')

LiverGlycogen(mg gm ')

Virgin femalesFedFasted

2.0 :28.8

0.61.5

160361

16 55.0 1.634 0.9 0.4

Nursing mothersFedFasted

6.026.8

2.61.9

417 60487 26

52.60.8

4.10.3

"Livers from virgin female and 1 to 2 day nursing mother rats were perfused with 0.7 mM oleic acidto determine their ketogenic capacity and were subsequently analyzed for their content of glycogenand total carnitine. Values are means SEM for four livers in each group. (Data from Robles-Valdeset al.3')

tive sequence and into the esterifica-tion pathway, indicating no fundamental defect in the latter system.1"-'21This led us to suggest that primarycontrol on the flow of fatty acidsthrough this metabolic branchpointwas exerted at some locus in theoxidation sequence.-'4''' The findingthat octanoic acid, which enters themitochondrion by a carnitine-inde-pendent mechanism, was oxidized toacetyl-CoA at equal rates in liversfrom fed, fasted, and diabetic rats,--'in contrast to long chain fatty acids,41"indicated that the carnitine acyltrans-ferase step constitutes a key controlsite in the hepatic metabolism of longchain fatty acids.1" Consistent withthis thesis was a series of studies onthe metabolism of (-)-octanoylcarni-tine. It was observed that variousmanipulations used to enhance thecapacity of rat liver to oxidize oleicacid (eg, starvation, short-term treatment with insulin antibodies or glucagon) resulted in proportional increasesin the oxidation of (-)-octanoylcarni-tine."'-'3 The fact that octanoic acidcould be made to act like a long chainfatty acid by esterifying it with carnitine and thereby imposing a requirement for the action of carnitine acyl-

transferase II prior to its oxidationstrongly implicated this enzyme system as a potential regulatory site inthe ketogenic process.

THE ROLE OF CARNITINEIf the carnitine acyltransferase step

plays a central role in the control ofhepatic fatty acid oxidation, how is itsactivity regulated? Two possibilitiesexist. The first is that the quantity ofthe enzyme is increased in the ketoticliver. While reports to this effect haveappeared-1 -> the observed incrementsin the amount of enzyme in liversfrom fasted and diabetic animals weresmall and, furthermore, could not bereproduced by others'11 (our unpublished data). The second possibility,which we favor, is that the flow offatty acids through this step in theintact liver is governed by factorsother than alterations in the amountof enzyme. One such factor nowappears to be carnitine itself, asubstrate in the first of the two transferase steps. This conclusion followsfrom the observation that all treatments used to stimulate the ketogeniccapacity of rat liver were accompaniedby increases in hepatic carnitineconcentration (Table 2).J7 Moreover,

addition of carnitine to the mediumperfusing livers from fed animalsmarkedly stimulated ketogenesisfrom oleic acid (Fig 3). It should benoted, however, that even in thepresence of added carnitine the rateof ketone production in the fed liverswas still significantly lower than thatin the fasted group, indicating thatadditional factors are involved in thecontrol of fatty acid oxidation in theintact organ.To gain further insight into this

problem we recently turned to anothermodel of physiological ketosis, theneonatal rat, which appears to requirethe ketone bodies for survival andnormal development. It was reasonedthat at birth many of the hormonaland metabolic adaptations that areseen during the development of ketosis in adult animals occur abruptly inthe transition from intrauterine toextrauterine life (Table 3) making themodel attractive for further investigation of the relationship betweencarnitine and fatty acid oxidation inthe liver. The characteristic surge ofketogenesis in the newborn rat (asreflected by the plasma ketone level)and its gradual decline during thesuckling period is illustrated in Fig 4.Particularly striking is the parallelprofile of liver carnitine content in theneonate, pointing again to the closelinkage between ketosis and livercarnitine concentration. A curiousfinding, however, was the fact thatdespite the absence of significantketonemia in the nursing mother ratsthey too showed a marked increase inliver carnitine levels around the timeof parturition. One explanation mightbe that the elevation in maternal livercarnitine concentration functionednot to enhance ketogenesis in themother but as a source of carnitine forexport to the suckling pups for stimulation of their capacity to synthesizeketone bodies. Consistent with thispossibility was the finding that thecarnitine content of maternal milkwas high at the beginning of thenursing period and fell thereafter inparallel with that of maternal liver.The fate of the maternal carnitinewas then studied directly by injectingnursing mother rats with butyrobe-taine ]1C, the penultimate interme-


c

Em

r- 5

O J3rra oLu o

30

20

O inUJ o*

L***

o>

10

"Sfr**"*-^Fasted (4)/

Fed + Carnitine (7)

_L J_0 20 40 60 80

MINUTES OF PERFUSION

Fig 3.Effect of carnitine on ketogenesisfrom oleic acid in perfused livers from fedrats. Livers were perfused with nonrecir-culating medium containing 0.7 mM oleicacid. Where indicated L-carnitine wasinfused at a concentration of 0.5 mM fromthe 15 minute time point. Values aremeans SEM for the number of liversshown. See McGarry et al27 for fulldetails.

diate in carnitine biosynthesis. (Conversion to carnitine occurs only in theliver and the process is extremelyrapid.) After various intervals maternal and neonatal tissues were examined for their content of carnitine14C.'4 The results of these studies,which will not be described in detail,indicated a movement of carnitinefrom maternal liver > maternalplasma - milk > neonatal liver (andother tissues).While the above findings in neona

tal rats provided strong support forthe central role of carnitine and carnitine acyltransferase in the intrahe-patic control of ketogenesis, the studies with nursing mother rats presented an intriguing paradox: whywas there no ketosis in the face ofsuch an elevated liver carnitine content? A simple explanation mighthave been the absence of substratesince measured free fatty acid concentrations were low, presumably because the animals were consuming ahigh-carbohydrate diet and had significant circulating insulin levels withresultant inhibition of site 1. Theproblem was more complicated, however, since measurement of theketogenic capacity of these livers byperfusion with oleic acid in vitroshowed inappropriately low values forthe measured carnitine content. As

100h

J_L

MOTHERS'/^LIVER

_I_L

X2_L _L

250

200

150

100

50

LU

H en

CC Q.

-3 -1 10 15 20 25 30t

BIRTH DAYS WEANINGFig 4.Plasma ketone and tissue carnitine concentrations in maternal and developingrats. See Robles-Valdes et al34 for full details.

seen in Table 4, ketone production bylivers from fed, nursing mother ratswas only modestly elevated over thatof fed virgin rats. Rates for both thefed virgin and fed mothers were lowwhen compared with those seen ineither fasted virgin or maternalanimals. Despite the fact that fastingproduced no significant further elevation in liver carnitine content in thenursing rats, ketogenic capacity increased fourfold. Conversely, the expected increase in carnitine did occurin the virgin group, concomitantwithin a 14-fold increase in ketogenicrate. Taken together, these studiessuggest that an increase in liver carnitine concentration is a necessary butnot sufficient requirement for theswitch of liver metabolism from a lowto a high ketogenic profile. Obviously,some other factor(s) is involved, thenature of which is not yet clear.However, we have been struck by thefact that a liver with high ketogeniccapacity is invariably characterized

both by an elevated carnitine contentand a depletion of its glycogen stores(see tables). A unique feature of theliver from fed nursing mother ratswas that the carnitine content washigh but that glycogen was notdepleted (Table 4). It thus seems likelythat both carnitine enrichment andglycogen depletion must occur for fullactivation of hepatic fatty acid oxidation. In this regard it is interesting tonote that compared with normal individuals patients with type 1 glycogenstorage disease have been found toexhibit relatively low levels of bloodketone bodies in the fasting state,despite normal elevation in levels ofplasma free fatty acids. '"' Conversely,the degree of lipemia was muchgreater in the type 1 patients, indicating that esterification rather thanoxidation was the major disposal routefor fatty acids in the liver. Preciseunderstanding of the interrelationships between glycogen, carnitine,and fatty acid oxidation is not yet


possible and several major issuesremain to be resolved: first, what isthe source of the increased liver carnitine in the ketotic (and nonketotic)animals? Second, does glycogen play adirect role or does its concentrationmerely reflect the level of anotherfactor that acts as a repressor for theflow of fatty acids through the trans-ferase step? Studies to answer thesequestions are in progress.

CLINICAL IMPLICATIONSIf the above model for the regula

tion of ketogenesis, which has evolvedlargely from studies with the rat, is tobe applicable to humans then it mustprovide reasonable explanations forthe etiology of the several distinctvarieties of ketosis commonly encountered in clinical medicine. An attemptto rationalize some of the clinicalobservations with the predictions ofthe model follows.

STARVATION KETOSISDuring the first few days of starva

tion in man the blood concentration ofacetoacetate plus /3-hydroxybutyrateclimbs to a value of about 2 mM andseldom exceeds 5 to 6 mM even withprolonged fasting."11? This physiological steady state is likely due to thefact that circulating insulin levels,though diminished, are sufficient toprevent the plasma free fatty acidsfrom exceeding a concentration ofabout 1.0 mM."1 The self-limiting characteristic of the physiological ketosisof fasting probably is due to theability of the ketone bodies to stimulate insulin release from the pancreatic /J-cell,ls although a direct antilipo-lytic effect of ketones on adiposetissue may play a role.1" In terms ofthe model, site 2 is fully operativewhile site 1 is only partially activated.The limiting factor in the rate ofketogenesis is substrate availability.

DIABETIC KETOACIDOSISIn severe diabetic ketoacidosis the

negative feedback loop of ketonebodies on insulin secretion cannotoperate, with the result that plasmafree fatty acid levels rise to extremelyhigh values (2.5 to 3.5 mM). Underthese circumstances the liver is drivento produce ketone bodies at a rate that

exceeds the capacity of extrahepatictissues to utilize them. Consequently,their concentration in blood can soarto the region of 20 mM producing alife-threatening metabolic acidosis.-In this case hepatic fatty acid oxidation capacity (site 2) is fully activatedas a result of the increased [gluca-gon]:[insulin] ratio and ketogenesis ismaximal because of full (nondamped)activation of site 1.

ALCOHOLIC KETOACIDOSISThis syndrome occurs in nondia-

betic alcoholic subjects in whom bloodketone concentrations equivalent tothose seen in diabetic ketoacidosis arefrequently encountered.4" The hallmark of the syndrome is free fattyacid levels much higher than thoseseen in ordinary starvation. In descriptive terms both sites 1 and 2 arefully active. Why site 1 should function in nondamped fashion with apresumably intact insulin loop remains to be explained. Two possibilities seem reasonable. The first is thatthe hormone-sensitive lipase of adipose tissue is abnormal in a selectedgenetic subset of alcoholics (diminished sensitivity to insulin suppression or accentuated sensitivity to lipo-lytic hormones or ethyl alcohol).Alternatively, insulin release from the-cell could be abnormally sensitive tocatecholamine inhibition in these patients. Since the syndrome is rapidlyreversed by giving glucose, presumably because of insulin release, anabnormal responsiveness of the adipo-cyte would seem most likely.It is of interest that the several

types of ketosis differ markedly intheir ease of reversibility in both manand experimental animals. Thus, starvation ketosis and alcoholic ketoacidosis are rapidly corrected by administration of small quantities of insulinand/or glucose. The acute response isundoubtedly due to the antilipolyticaction of insulin on the adipocyte (site1) which deprives the liver of ketogenic substrate. In the case of carbohydrate feeding this is followed by aresetting of the liver into a nonke-togenic mode as it becomes repletedwith glycogen (or a glycogen-relatedantiketogenic factor) and its carnitinecontent returns to normal levels. The

latter events are considered to befacilitated by the concomitant fall inplasma glucagon levels although, asindicated earlier, a direct effect ofinsulin on the liver is not precluded. Incontrast, severe diabetic ketosis isextremely resistant to insulin, even invery large doses.- Studies with alloxandiabetic rats indicate that the problemhere resides within the liver itselfsince free fatty acid levels fall rapidlywith insulin treatment.1" -'" The persistent ketone production stems from thefact that livers from these animals areengorged with fat such that they cansustain high rates of ketogenesis forseveral hours even in the absence ofan external supply of fatty acids (onlyby blockade of hepatic carnitine acyl-transferase with ( + )-acylcarnitinescan experimental diabetic ketoaci-dosis be rapidly reversed-11-1). Furthermore, after three hours of constant insulin infusion, during whichblood glucose levels were in excess of400 mg/100 ml, the livers showedessentially no glycogen deposition andstill possessed a high ketogenic capacity (J. D. McGarry and D. W. Foster,unpublished data).

To what extent this phenomenonresults from a prolonged period oftotal insulin deficiency per se or to thegreatly elevated glucagon levels seenboth before and for several hoursafter insulin therapy"-1- remains to beestablished. Also unknown at presentis the time-course over which hepaticcarnitine content falls to normalvalues during insulin administration.In summary, the two site model for

control of ketogenesis derived fromanimal studies appears to be fullyapplicable in principle to ketotic statesin man. Those factors that can betested in humans (plasma hormonelevels, and ketone and fatty acidconcentrations) change in patternspredicted by the theoretical system inphysiological and pathological ketosis.For the present, therefore, it wouldseem to be an adequate foundationupon which to build future biochemical and clinical investigations.

This investigation was supported in part byPublic Health Service grant AM-18573 and by agrant from the American Diabetes Association.Dr. McGarry is the recipient of research careerdevelopment award 1-K04-AM0763.


References1. Williamson DH, Hems R: Metabolism and

function of ketone bodies, in Bartley W, Korn-berg HL, Quayle JR (eds): Essays in Cell Metab-olism. John Wiley and Sons Ltd, 1970, pp 257\x=req-\281.

2. McGarry JD, Foster DW: Regulation ofketogenesis and clinical aspects of the ketoticstate. Metabolism 21:471-489, 1972.

3. Mayes PA, Felts JM: Regulation of fatmetabolism in the liver. Nature 215:716-718,1967.

4. McGarry JD, Foster DW: The regulation ofketogenesis from oleic acid and the influence ofantiketogenic agents. J Biol Chem 246:6247-6253,1971.

5. Ontko JA: Metabolism of free fatty acids inisolated liver cells. J Biol Chem 247:1788-1800,1972.

6. Woodside WF, Heimberg M: Hepatic me-tabolism of free fatty acids in experimentaldiabetes. Isr J Med Sci 8:309-316, 1972.

7. Foster DW: Studies in the ketosis of fasting.J Clin Invest 46:1283-1296, 1967.

8. Fain JN, Kovacev VP, Scow RO: Effect ofgrowth hormone and dexamethasone on lipolysisand metabolism in isolated fat cells of the rat.J Biol Chem 240:3522-3529, 1965.

9. McGarry JD, Wright PH, Foster DW:Hormonal control of ketogenesis: Rapid activa-tion of hepatic ketogenic capacity in fed rats byanti-insulin serum and glucagon. J Clin Invest55:1202-1209, 1975.

10. Liljenquist JE, Bomboy JD, Lewis SB, etal: Effects of glucagon on lipolysis and ketogen-esis in normal and diabetic man. J Clin Invest53:190-197, 1974.

11. Unger RH: Alpha- and beta- cell interrela-tionships in health and disease. Metabolism23:581-593, 1974.

12. Unger RH, Orci L: The essential role ofglucagon in the pathogenesis of diabetes melli-tus. Lancet 1:14-16, 1975.

13. Engel FL, Amatruda TT Jr: Hormonalaspects of ketosis. Ann NY Acad Sci 104:753-771,1963.

14. Gerich JE, Lorenzi M, Bier DM, et al:Prevention of human diabetic ketoacidosis bysomatostatin. Evidence for an essential role ofglucagon. N Engl J Med 292:985-989, 1975.

15. Schade DS, Eaton RP: Glucagon regulationof plasma ketone body concentration in humandiabetes. J Clin Invest 56:1340-1344, 1975.

16. Kopec B, Fritz IB: Comparison of proper-

ties of carnitine palmitoyl-transferase I withthose of carnitine palmitoyltransferase II, andpreparation of antibodies to carnitine palmitoyl\x=req-\transferases. J Biol Chem 248:4069-4074, 1973.

17. Tzur R, Tal E, Shapiro B: \g=a\-Glycerophos-phate as regulatory factor in fatty acid esterifi-cation. Biochim Biophys Acta 84:18-23, 1964.

18. Williamson DH, Veloso D, Ellington EV, etal: Changes in the concentration of hepatic me-tabolites on administration of dihydroxyacetoneor glycerol to starved rats and their relationshipto the control of ketogenesis. Biochem J 114:575\x=req-\584, 1969.

19. McGarry JD, Meier JM, Foster DW:Effects of starvation and refeeding on carbohy-drate and lipid metabolism in vivo and in theperfused rat liver. J Biol Chem 248:270-278,1973.

20. McGarry JD, Foster DW: Acute reversal ofexperimental diabetic ketoacidosis in the ratwith ( + )-decanoylcarnitine. J Clin Invest 52:877\x=req-\884, 1973.

21. McGarry JD, Foster DW: Studies with ( + )\x=req-\octanoylcarnitine in experimental diabetic keto-acidosis. Diabetes 23:485-493, 1974.

22. McGarry JD, Foster DW: The regulation ofketogenesis from octanoic acid: The role of thetricarboxylic acid cycle and fatty acid synthesis.J Biol Chem 246:1149-1159, 1971.

23. McGarry JD, Foster DW: The metabolismof (\p=m-\)-octanoylcarnitinein perfused livers fromfed and fasted rats: Evidence for a possibleregulatory role of carnitine acyltransferase inthe control of ketogenesis. J Biol Chem 249:7984\x=req-\7990, 1974.

24. Norum KR: Activation of palmityl-CoAcarnitine palmityltransferase in livers fromfasted, fat-fed or diabetic rats. Biochim BiophysActa 98:652-654, 1965.

25. Aas M, Daae LNW: Fatty acid activationand acyl transfer in organs from rats in differentnutritional states. Biochim Biophys Acta 239:208\x=req-\216, 1971.26. DiMarco JP, Hoppel C: Hepatic mitochon-

drial function in ketogenic states: Diabetes, star-vation and after growth hormone administration.J Clin Invest 55:1237-1244, 1975.

27. McGarry JD, Robles-Valdes C, Foster DW:The role of carnitine in hepatic ketogenesis. ProcNatl Acad Sci USA 72:4385-4388, 1975.

28. Dymsza HA, Czajka DM, Miller SA:Influence of artificial diet on weight gain andbody composition of the neonatal rat. J Nutr

84:100-106, 1964.29. Bailey E, Lockwood EA: Some aspects of

fatty acid oxidation and ketone body formationand utilization during development of the rat.Enzyme 15:239-253, 1973.

30. Girard JR, Cuendet GS, Marliss EB, et al:Fuels, hormones, and liver metabolism at termand during the early postnatal period in the rat.J Clin Invest 52:3190-3198, 1973.

31. Augenfeld J, Fritz IB: Carnitine palmitoyl\x=req-\transferase activity and fatty acid oxidation bylivers from fetal and neonatal rats. Can JBiochem 48:288-294, 1970.

32. Lockwood EA, Bailey E: Fatty acid utiliza-tion during development of the rat. Biochem J120:49-54, 1970.

33. Page MA, Krebs HA, Williamson DH:Activities of enzymes of ketone-body utilizationin brain and other tissues of suckling rats.Biochem J 121:49-53, 1971.

34. Robles-Valdes C, McGarry JD, Foster DW:Maternal-fetal carnitine relationships and neo-natal ketosis in the rat. J Biol Chem 251:6007\x=req-\6012, 1976.

35. Havel RJ, Balasse EO, Williams EE, et al:Splanchnic metabolism in Von Gierke's disease(glycogenosis type I). Trans Assoc Am Physi-cians 82:305-323, 1969.

36. Owen OE, Felig P, Morgan AP, et al: Liverand kidney metabolism during prolonged starva-tion. J Clin Invest 48:574-583, 1969.

37. Owen OE, Reichard GA Jr, Markus H, et al:Rapid intravenous sodium acetoacetate infusionin man: Metabolic and kinetic responses. J ClinInvest 52:2606-2616, 1973.

38. Madison LL, Mebane D, Unger RH, et al:The hypoglycemic action of ketones: II. Evidencefor a stimulatory feedback of ketones on thepancreatic beta cells. J Clin Invest 43:408-415,1964.

39. Bj\l=o"\rntorpP, Scherst\l=e'\nT: Effect of \g=b\\x=req-\hydroxybutyrate on lipid mobilization. Amer JPhysiol 212:683-687, 1967.

40. Cooperman MT, Davidoff F, Spark RS, etal: Clinical studies of alcoholic ketoacidosis.Diabetes 23:433-439, 1974.

41. Meier JM, McGarry JD, Faloona GR, et al:Studies of the development of diabetic ketosis inthe rat. J Lipid Res 13:228-233, 1972.

42. M\l=u"\llerWA, Faloona GR, Unger RH:Hyperglucagonemia in diabetic ketoacidosis: Itsprevalence and significance. Am J Med 54:52-57,1972.


[email protected]@generic 8158b284c5d3

Documents