Relationships between fatty acid synthesis and lipid secretion in the isolated perfused rat liver: Effects of hyperthyroidism, glucose and oleate

Biochimica et Biopto'.~icu Acta, 11186( 1991 ) 197-208 ~, 1991 Elsevier Science Publishers B.V. All rights reserved 11005-2760/91/$(13.511

197

BBALIP 53780

Relationships between fatty acid synthesis and lipid secretion in the isolated perfused rat liver: Effects of hyperthyroidism, glucose and oleate Lawrence W. Castellani, Henry C. Wilcox and Murray Heimberg l~'partment of Pharmacololo; Unirersity of Tennex~ee-MemphL~, The tt~'alth Science Center. Memphis. TN (U.S.A.) (Received 3 April 1991 )

Key words: Hepatic fatty acid synthesis: Hyperlhyroidism: (Pcrfused rat liver)

Various studies on the effects of thyroid status on hepatic fatty acid synthesis have produced conflicting results. Several variables (e.g., plasma free fatty acid and glucose concentrations) are altered simultaneously by thyroid status and can affect fatty acid synthesis. To evaluate the effects of these variables, hepatic fatty acid synthesis (lipogenesis) was studied in isolated perfused livers from normal and triiodothyronine-treated rats. Livers were perfused with media containing either 5.5 or 25 mM glucose without fatty acid, or 5.5 mM glucose and 0.7 mM oleate. Rates of lipogenesis were determined by measurement of incorporation of 3HzO into fatty acids. IApogenesis in livers from hlqperthl~oid animals exceeded that of controls, when perfased with 5.5 mM glucose with or without oleate. Pecfusion with 25 mM glucose increased lip~enesis in both euthyroid and hyperthyroid groups to the same level, abolishing this difference between them. Perfusion with oleate reduced rates of lipogenesis by livers from euthyroid and hyperthyroid rats to a similar extent, but stimulated secretion of radioactive fatty acid in phospholipid and free fatty acid fractious. Oleate increased ketogenesis by livers from normal and triiodothyronine-treated rats, with higher rates of hetogenesis in the triiodothyronine-treated group. When oleate was omitted, ketugenesis in the presence of $.5 mM glucose by the hyperthyroid group was similar to that of euthyroid controls, while ketogenesis was decreased in the hyperthyroid group relative to controls when perfused with 25 mM glucose. About 30% of the radioactivity incorporated into the total fatty acid of both groups was recovered in palmitate, with the remainder in longer chain saturated and unsaturated fatty acids. In both cuthyroid and hyperthyroid groups, the ratio of trincylglycerol:phospholipid fatty acid radioactivity was not only less than predicted (based on synthetic rates of PL and TG) but also was decreased in perfusious with exogenous oleate compared to perfustous without oleate, in perfusions with oleate, both groups incorporated twice as much radioactivity into phospholipid as into triacyiglycerol. The data suggest the following concepts: while hepatic fatty acid synthesis and oxidation are increased simultaneously in the hyperthyroid state, de novo synthesized fatty acids seem to be poorer substrates for oxidation than are exogenous fatty acids, and are preferentially incorporated into phospholipid, while exogenous fatty acids are better substrates for oxidation and esterification to triacylglycerol. The preferential utilization of de novo synthesized fatty acid for phospholipid synthesis may be an important physiologic adaptation insuring a constant source of fatty acid for membrane synthesis.

Introduction Abbreviations: ACAC. aeetoacetate: BHB. ~-hydrox3'l,utyrate: BSA. bovine ~ r u m albumin: C, cholesterol: CE, cholesterol ester: CPT, earnitine palmitoyltransferase: DG, diao'lglycerol; EU, culhyroid: FFA. free fatty acid: MG. n'~moacylgl~.cerol: PI. ph~phalidylinositoL PL. phospholipid: PS. phosphatidylserine: TLC. thin-layer chromatography" TG. Iriao'lglycerol; TOFA, 5-telradec~yloxy-2-furoic acid: T~. trii~thyronine. Corrcsl~mdence: M. Ilcimberg, Dept. of Pharmacology. University of Tennessee-Memphis, The Health Center. Memphis. TN 38163.

U.S.A.

The results of various studies on effects of thyroid status on hepatic synthesis of fatty acids (lipogcncsis) have been conflicting. Hepatic lipogenesis has been reported to be increased [i-13]~ decreased [14-17], or unchanged [18] in the hyperthyroid state, and either decreased [2,3,6,9,12.13,18.19] or unchanged [7,8,19] in hypothyroidism. These investigations had been carried out in liver homogenates [1,2.15,16]. hepatoc~tes

198 [3,4,17], liver slices [5,6,14], the perfused liver [18], or in vivo [7-13], using several radioactive precursors. The only consistent findings have been in the intact animal, in which fatty acid synthesis was observed to be increased in the hyperthyroid state, in vivo, however, several variables (e.g., plasma free fatty acid and glucose concentrations) are altered simultaneously by thyroid status, and can affect fatty acid synthesis. The perfused liver is a useful physiologic model that allows for control of such variables. A previous study with the perfused liver led those authors to conclude that hepatic fatty acid synthesis was unchanged in the hyperthyroid state [18]. Rates of lipogenesis observed in vivo may have been influenced by other variables as mentioned above. Also, increased rates of fatty acid synthesis in the hyperthyroid state in vivo are observed despite the fact that concentrations of plasma free fatty acids are usually elevated in hyperthyroidism [18,2032]. This raises the possibility that exogenous fatty acid does not inhibit lipogenesis in livers from hyperthyroid animals in a manner similar to that demonstrated in euthyroid animals. Since hepatic synthesis of triacyiglycerol and secretion of the very low density lipoprotein are reduced by hyperthyroidism [33], the association of increased fatty acid synthesis with diminished triacylglycerol synthesis would seem to be incongruous, since generally they proceed in the same direction. We therefore initiated this study to evaluate de novo synthesis of fatty acid from 3H O simultaneously with secretion of triacylglycerol by :~olated. peffuscd livers from euthyroid and hyperthyroid rats. and to determine whether inhibition of hepatic lipogenesis by exogenous fatty acid is altered in hyperthyroidism [34]. An abstract of this work was presented at the 74th meeting of the Federation of American Societies for Experimental Biology, Washington. D.C., 1990 [34]. Materials and Methods

Male Sprague-Da~le~., rats (Harlan Industries, Indianapolis, IN), weighing between 2511-275 g, were housed in wire cages at 20°C, on a 12 h ~ight-dark cycle (lights on I16110-1800 h). Animals were allowed access to Purina laboratory chow and water ad libitum, and were kept for i week before initiation of treatment. Animals were made hyperthyroid by administration of T 3 for 7 days via Alzet osmotic minipum0s (model 2001. Alza Corp., Palo Alto. CA) surgically implanted intraperitoneally under anesthesia with diethyl ether [35]. The nominal pumping rate was 1.11 t t i / h and T 3 w a s administered at the rate of 9.6 /ag/rat per day (3.66+0.17 ttg/100 g body wt per day). On the morning of the experiment (0830-1030 h), the liver was removed surgically from the rat and perfused in a recirculating system, using procedures

[36] and apparatus [37] described previously. The perfusate consisted of Krcbs-Henseleit HCO~ buffer (pH 7.4), washed bovine erythrocytcs, 6 g/dl delipidated [27] BSA (Fraction V. Sigma Chemical, St. Louis, MO) and either 5.5 mM or 25 mM glucose. Initial perfusate volume was 72 ml, with a hematocrit of 3{)% (v/v); the perfusion medium ~as gassed continuously with a mixture of 95% 0 2 / 5 % CO z. After a 20 min period of equilibration, 5 mCi of 3H:O 15 mCi/ml 0.9% NaCI) were added at zero time, after which a solution containing Ca 2÷- and Mg2+-free Krebs-Henseleit HCO~ buffer (pH 7.8), 6 g/dl delipidated BSA, and either 1419 ~tmol oleic acid/dl or no fatty acid, was infused at the rate of ii.7 ml/h (166, or 0 /.tmol/h) as described previously [38]. Therefore, lioogenesis was measured under three different sets of perfusion conditions; experiments without the addition of exogenous fatty acid (oleate) in the presence of either 5.5 mM or 25 mM glucose and experiments in the presence of 5.5 mM glucose and 0.7 mM oleate. The perfusion was terminated after 2 h, the liver was flushed with 50 ml ice cold 0.9% NaCI, nonhepatic tissue was removed, and the liver was freeze-clamped in liquid nitrogen. Erythrocytes were removed from the oerfusate by centrifugation. Aliquots of the protein-free oerfusate and liver were analyzed for ketone bodies [39]. Lipids were extracted from the cell-free oerfusate and liver [40] and aliquots were evaporated under vacuum, sealed under N2 and stored at - 7 0 ° C. Lipid classes were separated by TLC on Silica gel G plates (Analtech) using three solvent systems. The plates were developed first in pentane/ethyl ether/methanolglacial acetic acid (110:20: 10: 1, v/v), until the solvent front was 8 em above the origin, to separate MG and PL [41]. The plates were allowed to dry for 10 rain and then developed in the same direction in petroleum ether/ethyl ether/glacial acetic acid (84: 15: 1, v/v), until the solvent front was 15 cm above the origin. These procedures separated all lipid fractions except for C and DG. The C-DG band was eluted into 10 ml chloroform and 8 ml aliquots were dried under vacuum, replated and developed in benzene/ethyl ether/ glacial acetic acid {50:50: I, v/v), until the solvent front was 15 cm above the origin. Lipid bands were visualized under ultraviolet light after spraying with 0.111% rhodamine 6G in methanol, and were scraped from the plates. Radioactivity in each lipid class was determined by liquid scintillation spectrometry using BIOCOUNT (10 ml) (Research Products International, Mount Prospect, IL) added directly to the isolated bands. To determine the radioactivity in the fatty acid moieties of the saponifiable lipids, all classes except PL were eluted from the silica gel with chloroform. Phospholipids were eluted with methanol (3 x 7 ml) followed by chloroform (7 ml). Lipid classes were hydrolyzed [42] and

199 aliquots were taken fi)r counting or chemical analyses, as described previously [39]. Individual PL clas.~s of the liver were resolved by TLC by a modification of the method of Skidmorc and Entenman [43]. Samples of PL were plated on activated Silica gel G plates ( 1 0 0 ° C for 15 rain) and developed in c h l o r o f i ) r m / m e t h a n o l / 7 M ammonium hydroxide (60:35:5, v/v), until the solvent front moved 15 cm. This procedure separated all PL fractions except PS and PI. The PS-PI band was then .scraped from the dry plate (15 rain), elutcd from the gel, dried under vacuum, rcplated and ,~paratcd with chloroform/ m e t h a n o l / 7 M ammonium hydroxide (35 : 60: 5, v/v). After separation, radioactivity in PL fractions was estimated by liquid .scintillation counting. Fatty acid methyl esters were prepared by the method of Morrison and Smith [44]. Fatty acid methyl esters were ,separated by HPLC, using an acctonitrilewater ,solvent system, and a C-18 reverse phase column (Supelcosil strainless steel C-18 cohnmn, 3/z pore size, 4.5 mm i.d. by 15 cm length). The HPLC apparatus consisted of two (Water,~-mode v 6{HX) A) pumps, a liquid sample injector, (Waters- Model U6KL aut,~matic gradient controller (Waters- Model 680L liquid chromatography spectrophotometcr (Watcrs-LambdaMax Model 480), and a recording integrator (LKB model 2220) The eluatc was monitored at 215 nm. Solvent flow was 2 m l / m i n . Solvent A was 7(1~ acetonitrile and .solvent B, I(M)~, acctonitrile. Separation was performed isocratically fi)r 37 rain, with an accl(,nitrile concentration of 73.9c~, that increased linearly to 81.1% between 37-39 rain and increa~d to I(HJ~ between 51-53 rain. The fraction with a retention time of 25.7 rain (2(1:4 and 16: I) was craig)rated under vacuum, resuspendcd in acetonitrile and scparatcd within 7 rain in a second run, using 89.5'~ acctonitrilc. Fatty acid methyl esters wcrc identificd by matching their elution time to that of known standards and quantitated by multiplying peak area by the specific correction factor. The correction factors ( m a s s / a r e a l were established from the peak area of in known mass of each fatty acid methyl ester standard, detcrmincd during a chromatographic run. Since ab~whancc of ultraviolet light varies with degree of unsaturation and chain length, the mass/area was different lot each fatty acid. The column eluates were collected in I rain increments and fractions from duplicate runs wcrc pooled to determine the radioactivity incorporated into each species of fatty acid. PL reference standards were obtained from Sigma Chemical (St. Louis, MO). All fatty acid methyl esters were obtained from Nu-Chck Prep (Elysian, MN) and were more than 99% pure except for Eico~pentacnoic acid (20: 5) which was obtained from Sigma Chemical (St. Louis. MO) and was more than 90~ pure. All data were analyzec using Student's t-test fi~r

unpaired samples. Results arc expressed as means + S.E, Differences between means were considered to be significant with P < 0.05. Results

,4. E]~'cls o f T~ on .~ynthesis and nu'lal~dism o f fully acids Animals treated with "l'~ had increased concentrations of .~rum T~ ct)mpared to EU controls ( I. I I .+_0,117 ~s 3.21 _+ (1.21 ng/ml). In a few animals, plasma insulin and glucagon levels were assayed and found not to be altered by treatment with "I'~(data not shown). Animals treated with T~ were not hyperphagic: weighl gain was less than controls during hormonal treatment, although T~ did not affect the final body weight or the ratio of liver wt/body wt. Final h~)dy weight was not significantly allered, because starting weights varied by up to 25 g. As observed previously, the volume of bile ,~creted was increased by livers from T~-trcated animals per[used with cithcr 5.5 mM [27] or 25 mM glucose, bul was not allcred by gluco.~ concentration or lbe addition of tally acid. Pcrfusate flow rate and liver weight wcrc not affected by either thyroid status or pcrfusion conditions. Treatment with "l'~ increased incorporation of ~H ,O into total (hepatic + .secreted) fatty acids in perfusion with 5.5 mM gluol~c, with or with~mt olcat¢ (Fig. IL reflecting primarily increases in the hepatic lipids, with cs.~ntially no diff~,rence in pcrfusatc lipids. Differcnces in the rates of fatty acid synlhesis were not ob:~ervcd belwccn EU and "l'~ groups in experiments with 25 mM glucose, although, in I~th groups, total synthesis wa~ increased by the higher glucose concentration. In experiments with 5.5 mM glucose, but without fatty acid, treatnlcnt with "I'~ increased incorporation of ~Ft,() into lt)lal (hepatic + secreted) PL larry acids, with incrca,,cs in the hepatic Pl, and no difference in thc secreted PL ('l'ablcs I and ll). Differences in PL mass paralleled 0iffcrcnces in theory)ration o[ ~ l | , O (Tables Ill and IV). Treatment with "l'~ also produced a small decrease in incorporation of ~H 20 into hepatic TG fatty acids, however there was no difference in incorporation into total TG (Tables I and liP. Furthermore, hepatic and total TG were d e c r e a ~ d in per[used livers from hyperthyroid rats (Tables ill and IV). In the hyperthyroid group, a small but statistically significant incrca~ in incorporatkm of radioactivity into total ( ' E falty acids (Tables I and II) was also observed, due to incrca~s in both hepatic and secreted CE. Howc v o . the ma~,~ of total, hepatic and secreted CE was reduced (Tables Ill and IV). In experiments with oleate and 5.5 mM glucose, T~-trealment increased incorporation of ~H:O into hepatic and ~ c r c t e d TG and PL fatty acids ('rabies I

2(~1 TABI.I- 1

hworl~,nttU,,i of ~1t,0 imp latty acM~ oI hcpatic lil,id~ The lipid li-acli~ms in lhc liver were scparaled by TI.(" and h_vdrolyzed. The data represent the radioact vdy in the fatty acid portion of each lipid exprc,~sed as d p m / g li~er/2 h. The panels represent lhe wtrious perfusion condilion~ as indicated. * Indicates significant diflerences between culhymid i l i U t and hyperthyroid (T31 groups within a particular set of pcrfusion c~mdition.,," "' and t, indicate sigpilicant differences due Io exogenous olcale Imlddle ~s top panell and 25 mM glucose (bollom ~s lop panelk respectively. P ~- 11.115. "Fhyloid state

d p m / g liver per 2 h

perl'usi~m conditions

PI.

FFA

T(i

MG

DG

('E

( N o FFA-5.5 mM glu)

I'U (N=

Ihl

1"3

iN:- 151 (11.7 mM FFA-5.5 mM glu) EU

71129 +861

I I 1i8 1 ' + I 157

2315 "

5435 + 331

11005 + 255

326 + 46

184 ± 17

+ I111

45200 ~ + 334

s12 4 122

3007 +56

247* ± 16

234 +47

9s I *' +72

1008" + 121

54" +_ 1'4

'44" + I11

1 482 * " + 134

245 "' + 185

1311" _+411

78" +7

1 46'4 + 1000

1 2119" +- 1711

3113 ±48

I It)l ± 1211

1 4411h ± 133

315 ±31

1600 +43 20000

I , V - 61

+ 141

T3 I N - 61

+ 11,5

171 +411

SS3S +584

144 +2'4

1115~3 * +736

13~ +37

(No |:|:A-25 mM glu) EU IN-7) T3 IN=7)

4112S * "'

I I 235 b + 1263 8472 ..i, %637

observations from this laboratory [31-33], secretion of TG was not decreased by treatment of the rats with T 3 under the current experimental conditions (Table liD. In cxpcriments with 25 mM glucose but without

and !1). The concentration of total and hepatic TG was reduccd by T~, as was both hepatic and secretcd CE (Tables I!1 and IV). As expected, secretion of TG was increased by oleatc: however, in contrast to previous TABI.E II

Incorlu"athm O( ¢11,0 ittto lilttt' achts ot .secreted lipid~ ]'he lipid ffaelions in the per[usate v, erc separated b} TL(" and hydrolyzed. The data represent the radioactivity in the l a i d acid rawlion o f each lipid cxpre,~sed a,, d p m / g liver per 2 h. TL,' panels represent the ~arious perfusion conditions as indicated. * Indicates significant differences belsscen eulh~roid IELII and h.~,perlhymid 113t groups within a parlieular set of perfusion conditions: " and ~' indicate significant differences due Ill CxI)gellOIjs oleatc (middle v., top panel) aad 25 mM gluco,,e lbottom ~.s lop panel), respectively. P (- 11.115. -l'h.~roid slale per fusion conditiom, (No FI:A-5.5 mM gtu) !-I.l I N : Ihl "I'~ ( ,% -= 151

111.7 mM |:FA-5.5 mM glu) ['U

d p m / g li~cr per 2 h PI. FFA

TG

MG

DG

CE

_57 + 47

1`400 +26

_115_ -~ 23'4

184 +_5"

257 +41

24 .+'3

355 -,: 57

24t~ + 55

2 4700 4 5112

2113 + 28

2007 + 311

±9

7411"

2001|8"

445 "

411"

57"

711 *

4"

(N : h)

~ 167

+ 753

+43

± 17

+ I t)

~-2

-1"3 i N = 61

I 174 * "" + 121

3 3002 " +gig

625 * " * 73

50, " _+ 15

88 " +_,'~7

16 " +6

6118tl ~" + 377

5 267 '" +4311

31113 " + 1511

1511 + 211

S4 + II

68 + 14

5S73 ~' +751~

49115 ~' +_727

25001 +512

78 % 1'4

67 ±t~

117 ±8

(No FFA-25 mM glu) EU ( N = 7) T3 I N = 71

201 TABLE Ill

('oncentrations of hepatic lipids Lipid fractions in the liver were separated by T L C aild the mass determined by. chemical analysis. Data ,+~c mcam, + S.I-. ( # m~l,/g li~cr pcl 2 h) The panels represent the various perfusion conditions as indicated. * Indicalcs significant difli:reilcc~ b c t g c c n culhyroid It, l L and hyperthyroid (T3) groups within a particular set of Imrt'usion coilditions. " and b indicate significailt differences duc Its cxogcilous ~*lcalc (middle v,, lop pancD and 25 mM glucose (bad(tom vs top panelL respectively+ P < 11.05. Thyroid slate perfusion conditions (No FFA-5.5 mM glu) EU ( N = 16) T3 ( N = 15) (0.7 raM FFA-5.5 raM glu) EU ( N = 6)

T3

~ m o l / g liver per 2 h PL FFA

TG

18.64 ± 11.31

2.'~ + 0.20

23.71 * +1L21

23.35 " + 2.411

21.14 "'

( N = h)

(No FFA-25 mM glu) EU

+ 2.211

23.21 h

11.60 ~: i1.ll4 11.72 11.1141

ql.t~ + 11.116

11.8~ + 11.111

MG

1.2~, * +11.21

574 " + 11.411

2.41~ .... + 11.411

IXi

(T

1).23 1l.(13

0.113 + 11.01

0.hi + 11.211

11.I ~ + II.l)-I

11.113 -~ li.I)l

11.311 + 0.lhR

(I 29

11.114

0.74

4- li.(14

4- IP.GI

+ 11. IX

11.2q * 1).117

11.1h% + II.lll

1t.32 * II. Iq

11.71

2."4

11.23

( N = 7)

+ 2.711

+ 11.117

+ I1. lil

+ 11.113

T3

24.41 _+ 1.911

(I./49 111.14

I. 711 111.241

11.211

0.114

11.20

111.115

+ (till

+ ILlll

(N

= 7)

11.113 +If.Of

II~7 + l).It+

tnon into hepatic, but not total. PL fatty acids was increased. Under these perfusion conditions, net incorporation of-~lt +O into total fatty acid was not altered by thyroid status, since decreased incorporation into

oleate, T3-treatment did not alter incorporation of 3H.O into total lipid fatty acid (Fig. I )+ although incorporation into total TG fatty acids was decrca~d (Tables I and IlL primarily in the liver, whilc incorporaTABLE IV

Mass of ,~'cretcd lipMs Lipid fractions in the pcrfusatc w'crc separated by TL( + and the mass dclcTmincd b} chcilutal a i l a l ~ s . Dal,i a~c means + S.I.. ( ~ m o t / g h x c r / 2 h). The panels represent the various pcrfu,,itmn conditions as indicated + Indicalc~ ~lgililicailt d i l l c t c n t c s between cuthyroid ( | ' U ) and hyperthyroid (T3) groups within a particular ",el of pcrfu,d,m contlilioh,. ' and +' indlc31c ,,ignilicanl dllfClcilccs duc 1o c,a+gcllOU,, i+'.¢;itc Imiddlc vs top panel) and 25 mM glucose (bottom ~s lop paflclL Icspeclixcl~,. I j • lull+ • l : n d c r (hone coilditi
i x m o l / g li~cr/2 h PL

I+F A

I( +

M( +

D( i

(I

1.32 +11.|6

11.44

It+s_~

( 1.11~

(l+IlX

I~ 14

+ 11.I12

t It I4

+ It.Ill

+ II.(11

. ILl(3

I .¢11 +1LI1)

11.4-1 + 11.111

11.7-I-

(I.Ih'~

II. l l

+ II. Ill

* 11.112

- II.02

11.117 "

+ It.It2

(t).7 mM FFA-5.5 mM g i n )

EU (N

007

II. II

+ 11.115

+ 11.117

1.19 '

+ ILI12

+ILI+Z

+ (1.03

1.44 ._ tl.07

1,115 " + 11JII

If,(IS + 1|.112

11.15 * 11.112

l).l~) + + It.{12

1.34 = 6)

T3 ( N = 6)

(No FFA-25 mM glu) EU ( N = 7) T3 ( N = 7)

(1117

1.27 +l).If~

0.73 + 11+0~

11•6t~ + 111114

11+114 + ll.lll

11.I~ + 11.111

11.14 + 11.114

1.16 ±11.11}

11.511 ÷ 11+111

11159 + It+lib

11.114 + 11.111

IL41~ + 11.III

11.117 q.I)2

2O2

(

NO FFA-5.5 mM GLUCOSE • []

)00(]

fU T3

*

500( 000( $00C 0 L

P

L*P

0 . 7 rnM FFA-5.5 mM GLUCOSE S~2oooo,

iii!

EU I:~ ¥3

a ,

•

L

P

L*P

NO FFA-25 mM G L U C O S E

•

I~

"="

~u

T3

b

b

b

b

IOOG

L

P

L*P

Fig. I. I-ffects of Ireatmcnt v, ilh T~ on incorporation of ~11,O into hepalic and pcrfur, ale fatl~ acids. At Ihe end of the 2 h perfusion, lb.." Iolal lipid exlracl fn~m samples of liver (I.) and cell flee potful, ate (P) wcrc h~dnd~'cd. The fairy acids wcrc i~datcd by TL(" and radioactivity determined in a liquid ~intillation counlcr. The radioactivity incorl~ratcd into the fall} acids of pctfusalc and liver lipids, and their sum. are ~,ho~ n. {A){B) and ((') represent cxpcrimenls carried ouI under Ihc difl~crcnl pcrfusion conditions at, indicaied. * Indicate diffi:rcnces between culhyroid {El.;) and hypcrlhynfid (Ta) gnmps. " and *' indicate dilfcrcnccs duc Io either olcalc {A w B} or 25 m M glucose (A v,, ('). rcspccliv¢ I% P -, 1|.115. In pancb, A and ('. N = S fl~r b , t h c u l h y n , d and h.~perlhynfid gnmps. In panel B. N = 0 for bolh culhynfid and hyperthyroid groups.

total TG fatty acid was offset by the increase into hepatic PL fatty acids. The mass of total TG was reduced in the hyperthyroid group, primarily duc to a decrease in hepatic TG. A slight reduction in the mass of total cholcsteryl ester was also observed. Treatment with T 3 stimulated incorporation of 3H 20 into total PL fatty acids rather than into total TG fatty

acids, under all pcrfusion conditions ('Fables I and il). In pcrfusions with either 5.5 mM or 25 mM glucose in the absence of olcatc, no differences in incorporation between total phospholipid and triacylglycerol fatty acids wcrc observed in the cuthyroid group. In pcrfusions with olcatc however, the cuthyroid group incorporated approximately twice as much radioactivity into the total PL fatty acids as TG fatty acids, analogo,s to the relative distribution in the hyperthyroid groups. The influence of T3-treatment on incorporation of radioactivity into specific classes of hepatic PL was determined in perfusions with oleate. The percentage distribution of radioactivity among the PL subclasses was similar fi)r EU and T3-treated groups (data not shownk although incorporation into total PL was higher in the T3-trcated group (Tables 1 and 11). Thyroid status had a minimal effect on total lipid fatty acid composition in perfusions with oleate (Fig. 2). A reduction in the amount of 211:5 (EPA) was observed in the T3-treated group, due entirely to a reduction in the liver. Administration of T~ increased the incorporation of 3H,O it.to 18:11 and 22:4 of hepatic fatty acids and decreased the incorporation into 211:3 and 22:6 (Fig. 3). Treatment with T~ increased incorporation into 18: 3 and 22: 6 of total fatty acids, while incorporation into 18: I was decreased. As observed previously, oleate increased ketogenesis by livers from either EU or T~ treated animals, and ketogenesis was greater in the T.~ group [32] (Table V). When pcrfusions were performed with 5.5 mM glucose, but without oleate, ketogenesis was similar in EU and T3 groups; in experiments with 25 mM glucose but without oleate, ketogenesis was actually lower in the T 3 group than the EU (Table V). Under all perfusion conditions, the ratio BHB/ACAC was lower in the T~ than in the EU group. Addition of olcate increased this ratio in both groups, while 25 mM glucose increased the ratio only in the EU group (Table V).

I1. Effe,'ts of oh'ate on synthesis and metabolism o f fatty acids The addition of oleate to the medium decreased incorporation of ;H ,O into total fatty acids of both EU and T~ groups (Fig. l) to similar extents (decreased 24 and 28¢;. EU and T~, respectively), although incorporation remained higher in the hyperthyroid group. The amount and relative proportion of secreted radioactivity increased, and that in the liver decreased, with oleate, in both groups, the decrease in hepatic radioactivity was due primarily to TG and PL fatty acids (Table l): oleate also reduced incorporation of 3H,O into CE. MG and DG fatty acids, but these fractions contributed a relatively minor amount to the total fatty acid pool. The increased radioactivity recovered in the perfusate in both groups was in the PL fatty acids and FFA. Oleate increased the concentration of TG in

2113 fatty acid radioactivity to be 1.5. if rates of PL and TG synthesis wcrc similar and the fatty acid supplying each wcrc derived from a common I ~ l . The predicted ratio of T G / P L fatty acid dpm, which reflects differences in rates for TG and PL synthesis, is greater than that actually found, suggesting a preferential utilization of newly synthesized fatty acid for PL+ rathcr than for TG synthesis. Furthermore, cxogcnous olcatc appeared to bc preferentially incorporated into TG, as the proportion of radioactive fatty acid rccovercd in PL incrcascd when pcrfusions used exogenous olcate. The data are, therefore, consistent with t h e existence of at Icast two pools of fatty acid perhaps one derived primarily from de novo synthesis and the other from exogenous

both liver and perfusatc in EU and T s groups and total PL in thc EU group (Tables Ill and IV). The radioactivity recovered in the glycerol moiety can be used to compare the relative rates of de novo TG and PL synthesis, because both pathways share a common precursor (Table VI). We did not determine the radioactivity of the precursor pools and, therefore, can only compare the relative rates of T G and PL synthesis within a given experimental group. A T G / P L glycerol radioactivity ratio of i.0 would indicate identical rates of T G and PL synthesis. These ratios (glycerol moiety) indicate thal, under all cxperimental conditions, rates of T G synthesis wcre similar to or higher than rates of PL synthesis. Considering the molar ratio of FA in T G and PL, we would have expectcd the ratio of T G / P L

~2

=

sources.

•

~

LIVER

2

~

°

.

.

.

2.;,-

l~;u 21.-/10.~, 1 8 . ' / l l S J i t 4 . 4 1 0 . 7 I T3 122.si o.2 2 o . s l t s . 2 1 1 2 . 2 1 o . 6 i

*

. o.2 o.3

o.7 o.s

0.9116+s11.6 o.7 I t , . z l i.i

L2 L4

i.s t.S

3.7 4.3

0.7 10.6 L2 0.~

0,3 o.s

o v TOTAL + ' A r t y ACID

[]

T3

~

•

o

[ E U ]12.5

0.I [10+3 [64.7 ] 7.4

0.2

IT;liLt

0.11..016t.219.6

0.3

M

~

I

I

I

M

NL) ~

0.3 o+s

0.8 o.s

1.6 2.,

0,4 0.3

% o1." TOTAL F'AT'I'Y ACiD

25

Eu

LIVER+PERFUSATE

5

• M

I

Z

M

M

mu l l l t l l i+.z g . +i16.+I*+.911:.+ l 0.+ 0.+ 0.+ 0.+ t+.o l.:~ Lt T.31:9.,Io.:,I,o.+IzTtlIt.:S 0.6 0.', O.S o.t t4.t o.,~ t.4 ~

m

1.3+ +9 1.3+ +.4

% OF TOTAL FATTY ACID

Fig. 2. Tolal lipid fatty acid profile. After pcrfusion fi)r two hours with 5.5 mM glucose and cx~cmm,, olcatc, the total lipid extracted fn~m samples of liver and pcrfusalc were hydrolyzcd and the fall)' acids ~-paratcd hy IIPL('. The lop. middle, and [~lztorn panel,, rcprc,,cm the tally acids in liver, pcrfusalc, and lotal (liver + p~rfusalcl lipid.,,, respectively. The har graph in each panel ,,Imws the ma~s of the individual fatty acids. The table below cach graph sh,nw, what % of the total [any acid comp~e,ilkm each individual fatty acid represents,. * Indicalc valuc~, thai arc significantly' different from control P ~ 0.05: ND. not detected. For both colhyroid (EU) and hypcrtl~/roid T~ group, N - 4.

204

> 1-

[]

BJ

17J T3

20

LIVER

o~1o

0

16:0

16:l

|8:0

tg:t

18:2

18:3 20:1

20:2 20:3 20:4 20:5 22:4 22:5 22;6

FATTY ACIOS

PERFUSATE

~T3

i3°k 2O

¢t

10

.

F16:0

16:1

18:0 18:1

18:2

18:3 20:1

20:2 20:3

20:4 20:5 22:4 22:5 22:6

FATTY ACIDS

Fig. 3. Distribution of radioactivity in liver and perfusate fatty acids. As the fatty acids of the individual lipid classes were separated by HPLC, the column eluate was collected. This eluate was evaporated under vacuum, seintillalkm fluid added, and radioactivity determined by liquid scintillation counting. The radioactivity incorporated into each fatty acid for all lipid classes of perfusate and liver was added, and the % of the total radioactivity of perfusate and liver for each fatty acid was determined. The total FA ladioactivity in euthyroid and hyperthyroid samples was approx, t~ll) and 12IX) DPM, respectively, for liver and 261141and 311111DPM. respectively, for perfusate. The distribution of radioactivity in the individual fatty acids in the total lipid of liver and cell free pcrfusate are shown in the upper and lower panels, respectively. * Indicates values Iha! are significantly different from euthyRfid controls. P ~ 11,115.In both euthyroid and hyperthyroid g n m p s N = 4. TABLE V

I:~/]i'('ts of treatm('nt with T¢ ol~' hepatic k('togc'm'~is Data arc m e a n s ± S . E , fi~r total (hepatic+secreted) acettmcetal¢ ( A C A C I and /3-hydroxybutyrate (BHB) recovered at the end of the 2-h pcrfusion, l','xperiments were carried out under the different perfusion conditions, as indicated. * Indicate values which are significantly different from eulhyroid conlruls vdthin a particular set of perfusion conditions. " and h indicate significant differences due to exogemms oleate {middle vs hip panel) and 25 mM glucose (bollom vs top panel), respectively. P ~<11.t)5. Thyroid state

. u m o l / g liver per 2 h

pcrtusion conditions

A('A("

BI IB

A C A C + BItB

BIIB/ACAC

15.5 mM glu-no FFA) EU ( N = 16)

h.113 +1).28

tL 17 +11.57

15.211 ±11.76

± 11.110

6.72 + 11.411

7.32 * + I).54

14.114 ± 0.75

1.12 * ±11.118

EU (N=¢O

12.67 " + 1.14

41).(~) " +3.76

52.73 ±4.58

T3 i N = h)

23.49 *-" +2.74

49.27 " ±4.05

72.76 +_6.31

2.17 *'" ± 0.211

EU ( N = 7)

5.23 ±11.37

10.71 +_0.92

15.93 +_11.83

2.15 b ± 0.28

T3 ( N = 6l

5.58 + 11.47

h.gl * ± 0,30

12.42 +-_11.37

1.31 * ± 11.211

T3 ( N = 15)

1.53

(5.5 mM glu-11.7 mM F F A ) .I

3.21 " ± IL29

(Z~; mM glu-no F F A )

205 T A B L E VI

Relatit (" rut('.~ o f total PL and 7"(; .~:vntlu'~i.~ The data (dpm) represent the radioactivity fr~)m ~lt zO incorporated into the glycerol or fatty acid moiety of T ( i and PL in liver + perlu~,~te ;dler 2 h of perfusion. ~ T h e predicted ratio ( T G / P L ) is that anticipated lot incorl~)ration of radioactive t:any acid into T G and PI_ if only one p(~d ~1" fatty acid exists for their synthesis, h was obtained by multiplying the T G / P L ratio lor glycerol dpm (a measure of ~ynlhc~i~ rate1 by 1.5 (normalized for three fatty acids in T G and 2 in PL). * Indicate significant difterences between PL and TG, P -: 11.115. Thyroid state (pcrfusion ca)nditions)

D P M (glycerol m o i e t y ) / g liver

D P M (FA m o i e l y ) / g liver

PL

PL

(No FFA-5.5 m M glu) EU ( N = 8)

21152 +_493

6536 * +614

5.117 -+ I.M

7287 + 892

1 741 +549

3985 * +495

4.71 + 1.44

121137 ~ I 191

2973 + 364

111424 * + 982

4.12 -+ 11.gl

2596 -+ 271

61144 * + 784

2.57 -+ 11.54

T3 I N = 8) (0.7 m M FFA-5.5 m M glu) EU ( N = 6) T3

( N = 6)

TG

T(;/PL

C. Effects of high glucose concentration on synthesis and metabolism of fatty acids Glucose (25 mM) stimulated incorporation of ~H,O into total fatty acids (Fig. 1) and PL fatty acids, TG fatty acids, and FFA of both EU and T~ groups (Tables i and 11). The increase in incorporation of 3H20 into total fatty acids was greater for the EU group than the T 3 group (a 118 and 7()% increase in EU and T~, respectively), but both were stimulated to the same level (Fig. 1). 25 mM glucose abolished the higher rates of lipogenesis in the T.~ group observed with 5.5 mM glucose (with and without oleate). The 25 mM glucose increased the amount and the proportion of radioactivity ,secreted by both groups {Fig. !). The increased incorporation into secreted lipids was duc to changes in PL fatty acids and FFA in the T~ group (Table !1), while in the EU group, it was due to increases in TG and PL fatty acids and FFA. The 25 mM glucose increased the total and hepatic concentration of PL in the EU group and, while having no effect on total PL mass, did decrease secretion of PL {Tables !11 and IV} by livers from the T 3 treated rats. Discussion in the presence of a physiologic concentration of glucose (5.5 mM), fatty acid synthesis by livers from hyperthyroid animals exceeded that of euthyroid controls, whether or not oleate was also infused. These data agree with many previous experiments with various tissue preparations [I-13], In our experiments with 25 mM glucose but without fatty acid, differences in rates of hepatic lipogenesis were not observed, in

TG

TG/PL

(predicted) "' TG/PL

74~ + 513

I. I11 +11.111

7.¢~1 + 2.1)1

71HJ7 ' +637

11.61 + 11.1111

7.117 +2.16

31155 + 153

1423 * * I 12

11.4~1 + 11.114

l~.18 + 1.22

5 2112 + 162

2 1117 * + 228

11.411 + 11.114

3.8h + 11.81

agreement with the study of Laker and Maycs [18] in which 15 mM glucose was used. Brunengraber ct al. [45] demonstrated that rates of lipogcnesis increased with glucose concentration in pcrfused livers from norreal rats, with maximal rates of fatty acid synthesis attained at 17 mM gluct~e. Presumably, the higher glucose concentration used by Laker and Mayes [18] was resl~msible for their failure to o~erve increased rates of hepatic lipogcncsis in hyperthyn)idism in the pcrfused liver. Accelerated fatty acid synthesis in the hyperthyroid sta:c has been associated with increased mass of the lipogenic enzymes [I-4,11.13,46]. It was surprising therefore, that fatty acid synthesis by livers from euthyroid and hyperthyroid anim~,[~ became equalized, albeit at a higher rate, in the presence of 25 mM glucose; the mass of the lipogcnie enzymes may not have been tale limiting umlcr these conditions. In agreement with earlier studies 147,48], oleate increased kctogencsis in livers from EU and "I'~ rats. Fatty acid stimulates kctogcnesis directly by increasing the available supply of substrate, and indirectly by inhibiting fatty acid synthesis. Clearly. ketogenesis was stimulated to a greater degree with livers from hyperthyroid animals, as reported in earlier studies [18,27,31,32,49-51], and is a s ~ i a t e d with an increase in activity of CPT (EC 2.3.1.21 ) [52,531. The simultaneous contradictory effects of increa~d hepatic fatty acid synthesis and oxidation in the hyperthyroid state when olcate was present may result from the decreased sensitivity of CPT to the inhibitory effect of mahmylCoA [53]. When oleatc was omitted from the perfu~te. ketogenesis by livers from hyperthyroid rats was either

2(t6 unchanged (with 5.5 mM glucose) or dccrcascd (with 25 mM glucose), compared to cuthyroid controls. The endogenous fatty acids that arc substrates for ketogcnesis unt'er these conditions can be derived by synthcsis or by h~drolysis of lipids. In pcrfusions with physiologic content rations of glucose but without exogenous olcatc, fatty acid synthesis was higher in the hyperthyroid livers than cuthyroid, indicating that a potential endogenous source of fatty acid for oxidation was higher in the hyperthyroid state compared to th cuthyroid. However, since ketogenesis was not increased in the hyperthyroid group under these conditions, it may be that de novo synthesized fatty acids arc not a preferred substrate fiw this pathway. It was expected that the relatively high concentration of exogenous oleatc would compete with endogenous fatty acids fi)r microsomal chain elongation and dcsaturation, spari~ 3 much of the newly synthesized fatty acid from further modification. Palmitate is reported to be the primary product of dc novo fatty acid synthesis. However, in our experiments palmitic acid accounted for less than 31)e~ of the total radioactivity. Instead, most of the radioactivity was recovered in longer chain saturated and unsaturated fatty acids. Wadke et al. [54] u,~d mass spectroscopic analysis to determine the distribution of deuterium into newly synthesized 16:1) and 18:11 in the perfused rat liver. 4(1% as much deuterium was incorporated into stearate as palmitate. Furthermore. the deuterium was distributed throughout the entire chain of the fatty acid and not localized in the terminal portion, in order to explain the distribution of deuterium in stearate, the authors concluded that either stearate was a primary product of de now) fatty acid synthesis or that de novo synthesized palmitate is the preferred substrate for microsomal chain elongation. Because several laboratories have reported that palmitate is the primary product of de novo fatty acid synthesis, it seems reasonable that the de novo syntbesized palmitate was the preferred substratc for microsomal chain elongation. This would also explain the distribution of radioactivity in this study, with most of the radioactive palmitate converted to other fatty acids by chain elongation and desaturation. AI~), Pepin et al. [55] reported that less than 2%, of exogenous fatty acid underwent microsomal modification in cultured hepatocytes, suggesting that exogenous fatty acid was a relatively poor substrate for micro~mal modification. In perfusions with 5.5 mM glucose and exogenous oleate, both euthyroid and hyperthyroid groups incorporated about twice as much radioactivity into fatty acids of total (liver + perfu~te) phospholipid as total triacylglycerol. From the radioactivity incorporated into the glycerol moieties of PL and TG, it appears that the greater proportion of newly synthesized fatty acids incorporated into PL compared to TG can not be

explained by higher rates of PL synthesis. Perhaps de mwo synthesized fatty acids are preferentially incorporated into phospholipid. The addition of exogznous o[eate appeared to stimulate TG synthesis to a greater extent than PL synthesis, however, in spite of this, thc proportion of newly synthesized fatty acid incorporated into PL increase, suggesting that the exogenous fatty acid was preferentially incorporated into TG. Our data, moreover, suggest that exogenous fatty acids are used by the liver preferentially for energy, needs, both immediate through oxidation and future needs by storage of triacylglycen)l. Based on the apparent different utilization of de novo synthesized and exogenous fatty acids for incorporation into PL and TG as well as oxidation and microsomal modification, it is probable that at least two separate pools of free fatty acid exist in the liver. Fukuda and Ontko [56] observed that approximately twice as much newly synthesized fatty acid was incorporated in the liver into phospholipid as into triacyiglycerol, when livers were perfused with 25 mM glucose and without exogenous fatty acid. When TOFA was added to the perfusion to inhibit fatty acid synthesis [56], 8-times as much de novo synthesized fatty acid was incorporated in:o phospholipid as into triacylglycerol. Apparently, when de novo synthesis of fatty acids was reduced, conversion to triacylglycerol was affected much more than to phospholipid. Changes in the ratio of de novo synthesized phospholipid and triacylglycerol fatty acids in perfusate after TOFA treatment were qualitatively the same as those observed in the liver. These observations support the hypothesis that de novo synthesized fatty acids are preferentially used for phospholipid synthesis while exogenous fatty acids are preferentially used for triacylglycerol synthesis. The administration of TOFA almost completely inhibited fatty acid synthesis and increased fatty acid oxidation [56]. Furthermore, TOFA diminished incorporation of radioactive exogenous oleate into triacylglycerol while incorporation into phospholipid was increased. The authors suggested that TOFA altered glycerolipid metabolism at the point of diacylglycerol processing, with more t:xogenous oleate diverted to phospholipid synthesis. An alternative explanation could be that, if newly synthesized fatty acid is the preferred substrate for phospholipid but not triacylglycerol synthesis, inhibition of fatty acid synthesis would require that a gre;,ter proportion of exogenous fatty acid be available /or formation of phospholipid, even though the total supply of exogenous fatty acid might be reduced. None of the findings in the study by Fukuda and Ontko [56] appear to be inconsistent with preferential utilization of de novo synthesized and exogenous fatty acids for phospholipid and triacylglycerol synthesis respectively. However, it is difficult to explain the observation that more of the de

2117 novo synthesized fatty acids are incorporated into phospholipid than into triacylglyceroi, as observed in both studies, as well as the changes in specific activities that occurred after administration of TOFA [56], and changes in the proportion of radioactivity in PL after administration of unlabeled oleate in the present study, if de novo synthesized and exogenous fatty acids form a common pool as proposed by Fukuda and Ontko [56]. While the results of our study and the reevaluation of the report by Fukuda and Ontko [56] do not prove that newly synthesized and exogenous fatty acids form separate precursor pools, we believe that they do indicate that this question is not settled and warrants further investigation. While exogenous fatty acid does diminish hepatic lipogenesis, only a 24-28% decrease was observed despite a relatively high concentration (I}.7 mM) of fatty acid. If de novo synthesized and exogenous fatty acids were used equally by the liver, it might have been expected that a supply of exogenous fatty acid in excess of that needed to meet hepatic esterifieation and oxidation requirements under most physiologic conditions, would have inhibited fatty acid synthesis to a greater extent. It appears that lipogenesis is maintained at a fairly high rate despite the presence of a large supply of exogenous fatty acid, perhaps needed for specific biochemical functions. The relative contributions of newly synthesized and exogenous fatty acids for utilization in the liver appear to be variable and under hormonal and nutritional control. While the data suggest a preferential utilization of de novo synthesized fatty acid for phospholipid synthesis under our experimental conditions, it is clear that de novo synthesized fatty acids do enter alternate pathways of esterification (i.e., triacylglyeerol synthesis) or oxidation. Certainly, during a high carbohydrate diet, much of the carbohydrate in excess of that required to meet immediate energy needs is converted to triacylglycerol fatty acid, although the phospholipid pathway may bc first ~turated. The higher radioactivity in the perfu,~atc free fatty acid fraction in perfusions with oleate compared to perfusions without oleate, cannot be explained by competition between de novo synthesized and exogenously supplied fatty acid for uptake. Uptake of free fatty acid by the liver is linear over the range of concentrations encountered in these perfusions. Other studies from our laboratory did not indicate any differences among uptake of 18:! or several other common fatty acids [57]. The liver may be secreting de novo synthesized fatty acid in response to exogenously supplied oleate, it was suggested by our laboratory a number of years ago that the liver can secrete free fatty acids, albeit the major flow of fatty acids is in the direction of uptake in the presence of exogenous fatty acid [581. We have recently determined that the perfused liver does in-

deed secrete free fatty acid. and thai thc amount and composition of the FFA analyzed by gas-liquid chromatography, is similar whether or not exogenous fatty acid is supplied (Zhang. Z.J. et al. unpublished data). It has been suggested that modifying the fatty acid profile of membrane phospholipids can alter the activities of membrane bound enzymes, which in turn regulates various cellular processes. The hypothesis that de novo synthesized fatty acids may be the preferred substrates for incorporation into phospholipid and microsomal modification through chain elongation and desaturation may be important in maintenance of membrane fluidity. An interesting observation, is the secretion of a pool of phospholipid that was enriched in de novo synthesized fatty acid in response to exogenous oleate. The data suggest that exogenous fatty acid is preferentially incorporated into triacylglycerol. Clearly. the exogenous supply of fatty acid is the primary determinant regulating synthesis and secretion of VLDL in the normal fed animal. However, if de ntwo synthesized fatty acid is preferentially u ~ d for phospholipid synthesis, and phospholipid is necessary for synthesis and secretion of VLDL, de novo synthesized phospholipid fatty acid may a l ~ participate in VLDL secretion. The increased secretion of phospholipid enriched in de novo synthesized fatty acid supports recent reports that newly synthesized phospholipid is required for secretion of the VLDL [59]. Hepatic phospholipid synthesis and turnover is increased in the hyperthyroid state [60-62]. Inc-eased incorrg~ration ~f radioactivity into phospholipids observed in livers from the hyperthyroid animals, probably reflects this increa~d turnover. In addition to the increased demand for fatty acid in ph{rspholipid synthesis, energy demands in the hyperthyroid state are also increased. The simultaneous increase in synthesis and oxidation of fatty acids in the liver in the hyperthyroid slate is generally regarded as a pathophysiologic anomaly representing a futile t3,clc that cont.ributcs to the increased thcrmogenesis a s ~ i a t e d with hyperthyroidism. Nevertheless. rates of lipogenesis in hyperthyroidism appear to respond normally to stimulation by gluco.~ as well as inhibition by exogenous fatty acid. Furthermore, if endogenous fatty acid is used preferentially for incorporation into PL, while exogenous fatty acid is preferentially oxidized, this may be the basis for stimulation of both pathways in the hyperthyroid state. It could a l ~ furnish a teleologic argument for the reduced .sensitivity of CPT to the inhibition by malonyI-CoA observed in hyperthyroidism [53]. The preferential utilization of de nowo synthesized fatty acid for phospholipid synthesis, could be an important adaptation insuring a constant supply of fatty acid for membrane synthesis.

208 Acknowledgements This research was supported by grant HL-27850 from the National Institutes of Health, U.S. Public Health Service. L.W.C. was a postdoctoral trainee supported by training grant HL07641 from the National Institutes of Health. References I Diamant, S., Gorin E. and Shafrir. E. 119721Eur. J. Bi(vchem. 26, 553-559. 2 Landriscina. C.. Gnoni, G.V. and Ouagliariello. E. 119761 Eur. J, Biochem. 71. 135-143. 3 Gnoni. G.V., Landriscina, C. and Ouagliariello, E. 1198111 Biochem. Med. Metab. Biol. 24. 336-347. 4 Gotv,h.~,lge, A.G.. Garay. A. and Silpanata, P. 11074) J. Biol. Chem. 249. 1469-1475. 5 Spirtes. M.A., Grace. M. and Weinhouse. S. (19531J. Biol. Chem. 21M. 705-713. 6 Nciad, N.S.. Chaikoff. I.L and Itill. R. (19h2) Endocrinoh)gy 71. 1117-112. 7 Karp. A. and Sic(ten, D.. Jr. 119491 J. Biol. Chem. 179. 819-830. 8 Lh)bera, M.. Muniesa. A. and Herra, E. 11117('t th)rm. Metab. Res. II. t128-634. 9 Dayton. S., Dayton. J., Drimmer, F. and Forrest. K E . (1960) Am. J. Physiol. 199, 71-76. I11 Marchi. P. and Mayer, J. (1959} Experientia 25. 359-361. 11 Roncari, D. and Murthy. V.K. 119751 J. Biol. Chem. 250: (IlL 4134-4138. 12 Fain, J.N. and Wilhemi, A.E. 11962) End(~,dnology 71,541-548. 13 Das, D.K. and Ganguly. M. (19821 Arch. Biochem. Bit)phys. 218. I, 142-155. 14 Fletcher. K. and Myant, N.B. (19581 J. Physiol. 144. 361-372. 15 Fletcher, K. and Myanl. N.B. 1196111J. Physiol. 154. 145-152. 16 Myant, N.B. and Iliffe, J. 119631 Biochem ~v,:. Syrup. 24. 145. 17 Faas. F.A., Carter. W.J. and Wynn. J. 119721 Endocrim)logy 91: 6. 1481-1492, 18 Laker, M.E. and Mayes, P.A. 119811 Biochem. J. 196. 247-~'~5. 19 Bales. M.W., Zomzeb. C.E. and Mayer. J. (1955) Endocrinology 57. 5115-5118. 20 Rich. C.. Bierman, E.L. and Schwartz. I.L. 1195'41J. Clin. Invest. 38, 275-278. 21 Marks, B.II.. Kiem. I. and Itills, A.G. (191511) Meta~)li~,m 9. 1133-1138. 22 ttarlan. W.R.. Lash). J.. lk)gdor.off. M.D. and Estes E.H. (19f,3) J. Clin. Endoc. Metab. 23. 33-40. 23 Eaton, R.P.. Steinberg. D. and Thompsom. R.tt. 119651 J. Clin. Invest. 44, 247-2611. 24 Tubbling, (i. (19691Clin. Chim. A~.;~ 24. 12~-13',). 25 Nikkila. E.A. and Kekki. M. (19721J. ('lin. Invest. 51. 21113-2114. 26 Jacobsen, B.B. (19731 Acta Endtvzrinol. 72. 443-452. 27 Keyes. W.G. and Heimberg. M. ( 19791J. Clin. Ire'est. 04, 182-1911. 28 Engelken. S.F. and Eaton. R.P. (1980) Endocrinology 1117. 2118214. 29 Engelken. S.F. and Eaton. R.P. (1981} Atherosclerosis 38. 177188. 30 Abrams. J.J.. Grundy, S.M. and Ginsberg. tt. (1981) J. Lip. Res. 22, 3117-322.

31 Olubadewo, J.. Morgan. D.W. and Heimberg, M. 110831 J. Biol. Chem. 258. 938-945. 32 Olubadev, o, J.. Wilcox, H.G. and Heimberg. M. 11984) J. Biol. Chem. 259. 8857-8862. 33 Heimberg, M.. Olubadewo, J. and Wilcox. II.J. 119851 Endocr. Rev. 6, 59t}-6117. 34 Castenani. L.W.. Wilcox, H.G. and tteimberg. M. 1109111 The FASEB Journal. 4 (3). A7811, 35 Frank. R.A.. Wilcox, H.G. and Heimberg, M. (19811 Pharmacologist 23, 187, abs. 394. 36 ileimberg, M.. Weinstein, I., Klausner, It. a , d Watkins, M.L. 119621 Am. J. Physiol. 202. 353-358. 37 tleimberg, M., Fizette. N.B. and Klausner, H. 119641 J. Am. Oil Chem. Soc. 41. 774-779. 38 Hcimberg. M. and Wilcox, H.J. 119721 J. BioL Chem. 247, 8758811. 39 Soler-Argilaga, C. and Heimberg. M. 119761 J. Lip. Res. 17, 605-615. 411 Heimberg. M.. Van Harken, D.R. and Bn)wn, T.O. 119671 Bit~chim. Bioph~,. Acta 137, 435-445. 41 Wets, M.T. and Malik, K.U. (1986) Circ. Res. 59 (6), 694-703. 42 Aveldano, M.I. and Horrocks. L.A. 119831 Lip. Res. 24, I I 0 1 11115. 43 Skidmore, W.D. and Enlenman, C. 119621 Lip. Res. 3, 471-475. 44 Morrison. W.R. and Smith, LM. 119641 J. Lip. Res. 5, 600-608. 45 Brunengraber, H., Boutry, M. and Lowenstein. J.M. 119731 J. Biol. Chem. 248, 2656-2669. 46 Kumar, S.. Das, D.K., Dorfman. A.E. and Asato, N. 119771 Arch. Biochem. Biophys. 178, 507-516. 47 Hagenfeldt. L., Wennlund, A., Felig, P. and Wahren, J. (1981) Life Sci. 28, 1851-1856. 48 Hagenfeldt, L., Wennlund, A., Felig, P. and Wahren, J. 119811J. Clin. Invest. 67, 1672-1677. 49 Barrels. P.D. and Sestoft, L 119801 Biochim. Biophys. Aeta 633, 56-67. 511 Keyes, W.G., Wilcox, t t G . and tteimberg, M. 119811 Metabolism 311. 135-146. 51 Muller, M.J., Koster, H. and Seitz, H.J. (19811 Biochim. Biophys. Aeta 666. 475-481. 52 Stakkestaad, J.A. and Bremer. J. 119821 Biochim. Biophys. Acta 71 I, 911-11111. 53 Stakkestaad, J.A. and Bremer, J. 119831 Biochim. Biophys. Acta 750, 244-252. 54 Wadke. M., Brunengraber, tl., ia)wenstein, J.M., Dolhun, J.J. and Arsenault. G.P. (19731 J. Lip. Res. 14, 403-411. 55 Pepin. D., Bossy, D., Thomas, G., Berziat. G. and Chambaz, J. (19891 Biochim. Biophys. Acta 11106,52-58. 56 Fukuda, N. and Ontko, J.A. (1984) J. Lip. Res. 25, 831-842. 57 Kohout, M.. Kohoutova, B. and Heimberg, M. 119711 J. Biol. Chem. 246, 5067-5974. 58 Van Harken. D.R.. Brown. T.O. and Heimberg, M. (1967) Lipids 2. 231-238. 59 Vance. J.E. (1989) Biochim. Biophys. Acta 111116.59-69. 6d) Flock. E.V.. Bollman, J . L and Berkson. J. 119481 Am. J. Physiol. 155.4tl2. 61 Marino. A.. Fedriani, J.R.. Sancho, M.J. and Macarulla. J.M. (19841 Horm. Metabol. Res. 16. 638-040. 62 Tata. J.R. 119671 Nature II, 51ff)-569.