Effects of triiodothyronine on biosynthesis and secretion of triglyceride by livers perfused in vitro with [3H]oleate and [14C]glycerol.

Livers from euthyroid and hyperthyroid (triiodothyronine-treated) rats were perfused in uitro with tracer amounts of ['4C]glycerol and substrate quantities of r3H]oleate. Hepatic uptake of total free fatty acid, [3H] oleate and ['4C]glycerol were similar regardless of hormonal treatment. Output of triglyceride was 4.46 k 0.48 pmol/g and 2.14 k 0.36 by livers from euthyroid and hyperthyroid animals, respectively. Although total incorporation of ['4C]glycerol into perfusate triglyceride was similar in both groups, the specific radioactivity of triglyceride was approximately 30% higher in the hyperthyroid. The glycero-3-phosphate content of livers from triiodothyronine (T3)-treated rats was 52.3% that of livers from euthyroid rats (0.80 k 0.15 uersus 1.53 f 0.16 pmol/g, respectively). After 4 h of perfusion, the same relationship was observed. The specific radioactivity of glycero-3-phosphate following perfusion with ['4C]glycerol in livers from T3-treated rats was twice that of the euthyroid. Synthesis of triglyceride, calculated from glycero-3-phosphate specific radioactivity, was 4.62 f 0.44 pmol/g/4 h by livers from euthyroid rats and 1.95 f 0.27 after T3 treatment. The incorporation of r3H]oleate into perfusate triglyceride by livers from hyperthyroid rats was approximately 30% of the euthyroid, while the specific radioactivity was approximately equal in both groups. Triglyceride synthesized from r3H]oleate was calculated to be 2.00 k 0.50 pmol/ g of liver/$ h for livers from T3-treated animals and 4.29 f 0.57 for euthyroid controls. The rates of synthesis of perfusate and hepatic triglyceride, and output of perfusate triglyceride, by livers from hyperthyroid rats were lower than in the euthyroid, and similar values were obtained whether calculated by chemical assay, or from radioactive incorporation of [3H]oleate or ['"C] glycerol. Additionally, livers from T3-treated rats had higher rates of ketogenesis, secreted less glucose, and produced more 14C02 from glycerol than did livers from euthyroid rats. Clearly, livers from hyperthyroid rats synthesize and secrete less triglyceride (i.e. very low density lipoprotein) than do livers from euthyroid animals. These conclusions were reached whether glycerol

It was reported previously from our laboratory that livers from rats treated with triiodothyronine secreted less VLDL' triglyceride and had higher rates of ketogenesis than did livers from euthyroid rats, even though the uptake of free fatty acid by the livers was not altered (1,2). Thyroid hormones directed metabolism of fatty acids away from pathways of esterification (triglyceride synthesis) and into oxidative pathways (ketogenesis and COz formation). Observations with the perfused liver in agreement with ours were reported by Laker and Mayes (3); human studies of Abrams et al. (4) were consistent with our findings. However, studies from other laboratories (5)(6)(7) have been at variance with our conclusions. Using cell-free liver preparations (800 X g supernatant), Roncari and Murthy (5) concluded that the synthesis of triglyceride from ["Clsnglycero-3-phosphate, measured as percentage of glycero-3phosphate incorporated into triglyceride, was higher in thyroxine-treated rats than in controls. Total glycerolipid synthesis, however, was not altered. Glenny and Brindley (6) measured the rate of glycerolipid synthesis in vivo 1 min after intraportal administration of ['4C]palmitate and ["H]glycerol to rats; triglyceride synthesis was measured as the percentage of total glycerolipid synthesis. These authors reported that the rate of triglyceride synthesis from [''Clpalmitate was the same in euthyroid and hyperthyroid sthtes, whereas the rate of synthesis from ["H]glycerol was higher in hyperthyroid than in euthyroid states. The absolute rate of triglyceride synthesis by hyperthyroid livers was, however, not different from controls. The rate of triglycerida,: synthesis from ["HI glycerol was previously estimated by Nlkkila and Kekki (7) in man as a function of the plasma triglyceride turnover rate. They concluded that the rate of splanchnic triglyceride production was greater in hyperthyroidism than in the euthyroid state and, moreover, was not reduced on treatment of ithe thyrotoxicosis.
The use of glycerol as a precursor for synthesis of triglyceride may lead to incorrect interpretation of the data, if one neglects to consider the magnitude of the glycero-3-phosphate pool of the liver through which glycerol must pass during biosynthesis of triglyceride. Schimassek et al. (8) and Sestoft et al. (9) had reported earlier that the hepatic content of glycero-3-phosphate in the hyperthyroid state was about half that of the euthyroid. The smaller pool size of G3P in the hyperthyroid liver would dilute a tracer glycerol molecule less, resulting in higher specific activity of the G3P in the hyperthyroid than in the euthyroid state.
The calculated rate of triglyceride synthesis should be identical whether measured with either free fatty acid or glycerol as a radioactive tracer. To avoid erroneous conclusions, it is essential that the actual precursor specific activity be used for calculation of rates of synthesis of triglyceride, and not the specific activity of the added tracer glycerol or fatty acid. To reconcile apparently divergent data, we perfused livers from euthyroid and hyperthyroid rats simultaneously with ["C] glycerol and ['Hloleate, determined the specific activity of hepatic G3P and oleate taken up by the liver, and evaluated the effects of triiodothyronine on hepatic lipid metabolism.

EXPERIMENTAL PROCEDURES
Treatment of Animals a n d Perfusion Conditions-Male Sprague-Dawley rats (Harlan Industries, Indianapolis, IN) weighing 175-200 g were acclimatized for 1 week to 12-h (0600-1800 h) light-dark cycles in our animal facility before the experiment was started. Animals were made hyperthyroid by intraperitoneal administration of TI with Alzet (Model 2001) osmotic minipumps (10). The T.I was administered at the nominal rate of 9.6 pg/rat/day (4.47 & 0.07 pg/100 g, body weight/day) with pumps containing 40 mg T.l/deciliter. The T.1 was dissolved in n-butano1:propylene glycol (0.5:9.5) (v/v). Control EU rats received the pumps containing only the T.1 vehicle. All animals were treated for 7 days and were allowed free access to water and powdered Purina laboratory chow. Food consumption by each rat was measured daily. At the end of the treatment period, the rats were anesthetized with diethyl ether and the livers were removed surgically and perfused in a recirculating system used routinely in this laboratory (11). Immediately before cannulation of the portal vein, 4-5 ml of blood were withdrawn from the abdominal aorta for measurement of serum levels of triglyceride, FFA, and T.I. The composition of the perfusate was described previously (1). The initial volume of the perfusate was 70 ml and the hematocrit was 30%. The liver was equilibrated for 20 min with the initial medium; infusion was then begun of a 6% bovine serum albumin-sodium oleate complex ( 1 ) in Ca2+-and Mg'+-free Krebs-Henseleit buffer, pH 7.8, containing also 50 pCi of ['Hloleic acid and 20 pCi of ["C]glyceroL The infusion rate of 11.7 ml/h supplied the liver with 166 pmol of oleate/h, but only tracer amounts of glycerol. At the start of the infusion, and 2 and 4 h later, aliquots of perfusate were removed for analysis. The experiment was terminated at the end of 4 h. A lobe of liver was tied off with an umbilical suture, excised rapidly, and frozen in liquid nitrogen. The frozen liver was ground into powder, and a weighed quantity was used to extract glycero-3-phosphate. The remaining liver was perfused with about 40 ml of ice-cold 0.9% NaCl to rinse out residual perfusate. Nonhepatic tissues were removed and the liver was blotted dry and weighed. One gram of the liver was homogenized with 4.0 ml of icecold 0.9% NaCI, and 1.0 ml of the homogenate was extracted and analyzed for hepatic content of lipids.
Analytical Procedures-Lipids were extracted from rat serum, perfusate, and liver samples as described previously (11,12). The extracts were fractionated by thin layer chromatography and the free fatty acid and triglyceride were assayed as described previously (13). Aliquots of the triglyceride and free fatty acid fractions were evaporated to dryness in glass scintillation vials and the residues were dissolved in Biocount by shaking overnight. "C and "H radioactivity was measured in a Beckman LS 7500 Liquid Scintillation Spectrometer which had been programmed to calculate dpm for double-labeled samples. To examine the incorporation of ["C]glycerol into triglyceride-fatty acid and that of [,'H]oleate into triglyceride-glycerol, another aliquot of the triglyceride fraction was saponified with 0.5 M alcoholic KOH. After neutralization with 0.4 N HzS04, fatty acid was separated from glycerol by repeated extraction into petroleum ether, by a modification of the methods of Quagliariello et at. (14) and Newman et al. (15). The mass of glycerol (16) and free fatty acid (17) were assayed and the radioactivity in each moiety was measured by liquid scintillation spectrometry. One-ml aliquots of perfusates were deproteinized with Ba(OH),-ZnSO, (13) and after centrifugation the supernatants were analyzed for total ketones (13), glucose (glucose oxidase method, Boehringer Mannheim), and radioactivity of glycerol.
For measurement of [14C]glycerol, samples of deproteinized perfusate were extracted initially with chloroform to remove lipids, and then were treated with Bio-Rad Ag 1-X8 to separate lactate and other anionic compounds. Glucose was separated from glycerol according to the method of Bortz et al. (18) in aliquots from the Bio-Rad treatment in two experiments. Since the radioactivity of ['4C]glucose accounted for only 7-9s of total I4C, glucose was not separated from glycerol in subsequent experiments, and no correction was made for the radioactivity contributed by glucose. Serum T J was assayed by radioimmunoassay (Micromedic, Horsham, PA).
Assay of Glycero-3-phosphate-Samples of liver were excised at times indicated and frozen in liquid nitrogen. The livers were then pulverized on dry ice in a mortar precooled in liquid nitrogen. The powdered liver was weighed into ice-cold 6% perchloric acid (1:4) (w/ v). After centrifugation, the acid soluble extract was decanted into a centrifuge tube; the pellet was resuspended in 6% perchloric acid and centrifuged. The combined supernatant was neutralized with 2.0 M KHCO:I, 0.5 M KC1 (19) using methyl orange as indicator. The perchlorate precipitate was centrifuged and discarded; the supernatant was stored frozen until assayed. The enzymatic method of Hohorst (20) was employed for determination of the G3P content of the extract.
To determine the specific radioactivity of G3P, the G3P extract from livers perfused with radioactive substrates was treated according to the procedure of Conyers et al. (19) to remove cations and perchlorate anions. The G3P was then separated by thin layer chromatography on PEI-cellulose plates. After elution of the G3P band from the plates, the mass (20) and radioactivity of the G3P were determined, and the specific activity (dpm/pmol) was calculated.
To determine the variation of the hepatic content of G3P during the course of a perfusion, additional livers from hyperthyroid and euthyroid rats were perfused as described above. At the same time that aliquots of perfusate were sampled, a lobe of the liver was tied off, excised, and frozen in liquid nitrogen. The frozen tissue was treated as described above and used for extraction and assay for G3P. Additionally, livers from T.3-treated rats, euthyroid rats which received the T:J-vehicle only, and euthyroid rats not given any treatment were assayed for concentrations of G3P. These livers were removed from the animals and analyzed for G3P without being perfused.
Estimation of I4C02-The production of '%On from ['4C]glycerol by the perfused livers was measured by trapping "CO, evolved as described previously (13). The I4CO2 collected was then distilled into hyamine hydroxide, and the radioactivity was determined by liquid scintillation spectrometry (13).

RESULTS
Triiodothyronine-treated and euthyroid rats lost weight the day following intraperitoneal implantation of the osmotic minipumps, after which the body weight gradually returned toward the pretreatment level. Food consumption by the two groups was similar during the treatment period. The liver weights and ratios of liver weight/body weight were identical in both groups ( Table I) In agreement with earlier observations ( 1 , 2 ) , the volume of bile secreted by perfused livers from T.1-treated rats exceeded that of the control group (Table I). Perfusate flow rate was Biosynthesis of Triglyceride identical in both groups. Uptake of infused free fatty acid measured by mass or radioactivity was not affected by treatment with T.7, and was linear with time during the experiment (Fig. 1). The specific radioactivity of the free fatty acid taken up by the liver was similar in both groups (26.4 (x IO4) k 5.2 for the euthyroid and 27.7 (X lo4) -t 5.7 (dpm/pmol) in the hyperthyroid). Furthermore, as shown in Fig. 1, treatment with Ta did not alter the uptake of [14C]glycerol by livers from Ta-treated or euthyroid rats. The concentrations of glycero-3-phosphate was lower in

Functional parameters for perfusion of livers from euthyroid and hyperthyroid rats
Rats were implanted with Alzet pumps containing either T:, (40 mg/deciliter) or vehicle alone, and were fed ad libitum for 7 days.
Livers were perfused in vitro as described in the text. Perfusate flow rate and bile output were measured at Yj-h intervals to monitor the viability of the livers. Values are mean f S.E. There were 10 animals each in EU and T., groups.

T. 4
Body weight (g) 219 f 8.9 213.9 f 5.5 Food consumption (g/day) 13.   ]glycerol and free fatty acid by perfused livers from T3-treated and euthyroid rats. Livers were perfused simultaneously with glycerol and oleic acid as outlined in the text. Free fatty acid was separated from other perfusate lipids, and both mass and radioactivity were measured. The uptake of ["C] glycerol (A), total free fatty acid ( B ) , and ["Hloleic acid (C) were calculated as previously described (11). Per cent uptake was calculated from total amount of substrate infused. Each point represents mean 2 S.E. of 6 perfusion experiments. nonperfused livers from Ta-treated rats than from euthyroid rats (Table 11). Glycero-3-phosphate also was determined in livers from euthyroid and T3-treated rats at the end of perfusion in experiments with radioactive substrates. Another group of livers from euthyroid and Ta-treated rats were perfused without tracer glycerol and with unlabeled oleate, and in these experiments, the variation of glycero-3-phosphate concentration during perfusion was determined after 2 and 4 h of perfusion. For all perfused livers, glycero-3-phosphate levels were lower at the end of the experiment than were the basal levels measured in the nonperfused livers (Table 11). Under the conditions of these experiments, the glycero-3phosphate content of the perfused euthyroid livers declined to approximately 59% ( p < 0.05) of euthyroid basal levels, while it declined to 64% ( p < 0.2 > 0.05) of basal level for the livers from T3-treated animals. Regardless, the concentration of G3P in euthyroid livers at the end of perfusion was higher than in the hyperthyroid livers. In the experiments in which ["C]glycerol and [3H]oleate were infused, the G3P content in livers from Ta-treated rats was 52.9 f 9.4% of that in euthyroid livers. However, the specific activity of the ['4C]glycero-3phosphate purified from liver extracts after 4 h of perfusion was about twice as high in livers from Ta rats compared to the euthyroid (Table 111). Although there was a progressive decrease in the hepatic content of G3P during perfusion, the relative decrease in the hyperthyroid and euthyroid livers calculated from the ratio of G3P concentration at 4 h to that a t 2 was similar in both groups ( Table 11). The specific activity of G3P obtained at 4 h, therefore, was a reasonable estimate of the precursor pool, during the course of the perfusion, and was used for calculation of rates of synthesis of triglyceride.

TABLE I1
Hepatic content of glycero-3-phosphate The G3P content of the nonperfused liver was presumed to represent the basal level. G3P was extracted from the liver at the end of 4 h of perfusion unless indicated otherwise. The specific activity of G3P was calculated from the purified extract in which mass was assayed enzymatically and radioactivity was determined. In Experiment 11, G3P was extracted and assayed from liver samples taken at the indicated times without interruption of the perfusion. Excision of a lobe of the liver for analysis of G3P at 2 h did not adversely affect the subsequent liver perfusion. Values are mean & S.E. Numbers of livers are in parentheses. Significance of differences between basal G3P uersus end of perfusion are: for T:,, p < 0.2 > 0.05, and for EU, p < 0.05. In Experiment 11, significance of differences between G3P concentration at 2 h versus that at 4 h are: for T:j, p < 0.1 > 0.05, and for EU, p < 0.005. EU T.! Nonperfused livers (pmol/g) 1.53 t 0.16 (10) 0.80 f 0.154" (11)   Since 1 mol of G3P is converted to 1 mol of triglyceride, this calculation with ["'C]glycerol is also equal to moles of triglyceride synthesized. Clearly, the rate of synthesis of triglyceride from ['4C]glycerol was higher in livers from ELJ than from Ts-treated animals, even though the specific activity of the triglyceride was higher in the TB group. Although an identical conclusion was reached when the rate of synthesis of triglyceride from ["Hloleate was calculated, certain differences are important.
As shown in Table IV, the total incorporation and fractional conversion of ["Hloleate to triglyceride was higher with livers from euthyroid rats than with those from hyperthyroid animals. The relative specific activity of the per&sate triglycerides were similar in the euthyroid and the T3 group. The specific activity of the ["Hloleate taken up by the liver was identical in both groups.
The incorporation of ["Hloleate into triglyceride (pm01 of oleate/g/4 h) by livers from euthyroid rats was greater than that by T:l livers. Estimation of pmol of triglyceride synthesized from ["Hloleate was calculated from the prnol of oleate incorporated, with the following corrections. Although 3 mol of fatty acid are contained per mol of triglyceride, all the fatty acid moieties are not oleate. It was observed in previous studies from this laboratory (11,21) that the fatty acid composition of triglyceride secreted by livers perfused in the absence of exogenous supply of oleic acid consisted of approximately 30-35% oleic acid. When oleic acid was infused at the rate employed in the present experiments, the fatty acid composition of the secreted triglyceride was approximately 55-65% oleic acid. Clearly, the secreted triglyceride did not derive all its fatty acid from the exogenously supplied oleic acid. At best, the triglycerideoleate concentration increased by a factor of 1.5-2.0 when oleate was infused under these experimental conditions. When the correction factor of I.5 is applied in the calculation of the rate of triglyceride synthesis, the values of 4.29 f 0.57 and 2.00 -+ 0.50 (p < 0.005) are obtained for the euthyroid and hyperthyroid, respectively. These data are similar to those calculated from mass measurement or from incorporation of [ "C]glycerol.
The incorporation of glycerol and oleate into triglyceride was examined further by saponification of the triglyceride isolated from each perfusate and hepatic sample, and fatty acid was separated from glycerol. There was no measurable 'C radioactivity in the isolated triglyceride-fatty acid, and no 'JH radioactivity in the triglyceride-glycerol fraction from either EU or Ts livers. Therefore, under the conditions of these experiments, there was no significant lipogenesis (i.e. fatty acid synthesis) from glycerol by livers from either euthyroid or hyperthyroid rats. At the end of the perfusion experiments, the concentration of triglyceride (pmol/g) in all livers from T:&eated rats was only 46% of that in EU livers (Fig. 2). For those livers perfused with radioactive substrates, the concentration in livers from Ta-treated rats was about 50% of that of the EU group (Table  V). Total incorporation and fractional conversion of [14C]glycerol into hepatic triglyceride were higher in livers from EU rats than in the hyperthyroid (Table V) while specific activity and relative specific activity of the triglyceride tended to be lower in the EU than in TJ group. Incorporation of glycerol (pm01 of glycerol/g/4 h) into triglyceride was higher in the EU than in Ts livers, and the calculated rate of triglyceride synthesis (pmol of triglyceride/g/4 h) was similar to the mass of triglyceride determined by chemical assay. Similarly, as shown in Table VI, the total incorporation and fractional conversion of ["Hloleate into hepatic triglyceride were much higher in the euthyroid than in the hyperthyroid state. However, the relative specific activity of hepatic triglyceride was similar in the two thyroid states. The rate of synthesis of hepatic triglyceride, calculated as for perfusate triglyceride, was higher in the EU than in the T:% livers and similar to that determined by chemical assay. An interesting estimate of the utilization of each radioactive For "CO, production ( C ) , the rate of production (inset) and cumulative output were calculated as previously described (11); values are mean f S.E. for perfusions. For output of triglyceride and ketogenesis, statistical significance of differences between T.j and EU were estimated by Student's t-test: **, p < 0.005; I**, p < 0.001. For "CO, production, statistical significance between T:, and EU was calculated by Student's t-test and rergression analysis: x , p < 0.1 > 0.05; *, p < 0.05.

H O U R S
precursor by the liver for synthesis of perfusate and hepatic triglyceride was obtained by calculating the relative enrichment of each isotope in the triglyceride produced. The ratio "H/I4C in the perfusate and hepatic triglyceride was compared with the ratio of the infused precursors (Table VII). It can be seen that the ratio 'H/14C was greater in the synthesized triglyceride than in the infusate, and that the ratio :3H/'4C in either perfusate or hepatic triglyceride was greater in the EU group than in the hyperthyroid. These data suggest that relatively more fatty acid than glycerol was utilized for triglyceride synthesis by livers from both euthyroid and hyperthyroid rats, that livers from Tj-treated rats converted relatively less free fatty acid into triglyceride than livers from euthyroid rats, and that glycerol enters into many pathways other than lipid metabolism.
Our findings on glucose output confirmed earlier reports from this laboratory. Glucose output, which in these experi-ments, represents primarily glycogenolysis, at t.he end of 4-h perfusion by livers from T:%-treated rats was significantly less than that for EU livers (46.2 + 4.4 pmol/g/ versus 76.3 & 9.4, respectively; p < 0.01).

Triiodothyronine and
Biosynthesis of Triglyceride 943

Use of ['HJoleic acid for measurement of synthesis ofperfusate triglyceride by livers from normal and TI-treated rats
Livers from euthyroid and T:l-treated rats were perfused with a medium containing ["C]glycerol and ['Hloleate. The methods for isolation and assay of mass and radioactivity of triglyceride, and calculations, are described in the text,. Calculation of mol of oleate incorporated into triglyceride is discussed in the text. The estimate of pmol of triglyceride synthesized was obtained by dividing oleate incoroorated bv 1.5. as discussed in the text. Values are mean f S.E. for 6 perfusions in each grot  Student's t test is p < 0.02.

Use of ['HJoleic acid for measurement of synthesis of hepatic triglyceride in livers from normal and T:I-treated rats
At the end of each perfusion with ['4C]glycerol and ["Hjoleate, the concentration and synthesis of triglyceride in livers from normal and T.3-treated rats was determined as described in the text. Calculations of rate of oleate incorporation into triglyceride and amount of triglyceride synthesized from oleate/g of liver/4 h were similar to those described in Table IV

Relative utilization of ('HJoleate and ['4C]glycerol for synthesis of perfusate and hepatic triglyceride
Livers from euthyroid and T.1-treated rats were perfused with ["Cl glycerol and ['Hloleate, and aliquots of perfusat.e were assayed for content of triglyceride. Ratios were calculated from the total incorporation of I4C and "H radioactivity (dpm/g liver/4 h) and the fractional conversion of substrates to triglyceride (dpm/g liver pCi infused). The ratio "H/I4C in the initial infusate was that calculated from radioactivity in the infusate. Values are mean f S.E. for 6 experiments. 3.5 f 0.3" 6.9 f 1. 3 4.3 f 0 . 6
'' Significance of difference from euthyroid is p c 0.05.

DISCUSSION
The influence of thyroid hormones on plasma triglyceride concentration appears to be variable. It was reported that in conditions of elevated thyroid hormones, plasma triglyceride may be increased (7,22,24), decreased, (4,25,26) or normal (4,24,25). The variability may perhaps reflect the severity of the hyperthyroidism. Studies designed to elucidate how thyroid hormones alter degradation and/or biosynthesis of triglyceride have also yielded conflicting conclusions. Sandhofer et al. (25) observed hypotriglyceridemia in their hyperthyroid patients; they also studied splachnic triglyceride synthesis from [l-'4C]palmitate and suggested that the hypotriglyceri-demia may have been due to decreased synthesis of triglyceride, since they did not observe any change in triglyceride turnover or clearance. This suggestion was not supported by Nikkila and Kekki (7) who observed slight hypertriglyceridemia in their hyperthyroid patients; their study led to the conclusion that the rate of triglyceride synthesis from ["HI glycerol was increased. Previous reports from our laboratory (1,2) and the data reported here establish clearly that under controlled conditions, perfused livers from T,l-treated rats directed metabolism of oleic acid away from pathways of esterification (primarily triglyceride synthesis) and secretion of the VLDL and into pathways of oxidation (Con production and ketogenesis). In contrast, livers from hypothyroid rats utilized fatty acids in a direction opposite of those of the hyperthyroid (1,2 (8) and Sestoft et al. (9). The reduced glycero-3-phosphate level could be attributed in part to the enhanced catalytic properties of glycero-3-phosphate dehydrogenase by thyroid hormones (27). Clearly, administration or infusion of a tracer quantity of radioactive glycerol into euthyroid and hyperthyroid livers would mean that the infused ['%]glycerol would be diluted to a lesser extent in the hyperthyroid than in the euthyroid livers. Hence, the specific activity of glycero-3-phosphate would be higher in the hyperthyroid than in the euthyroid liver. In the present experiments, where the specific activity of glycero-3-phosphate in the livers from Trl-treated rats was twice that of the euthyroid, total '% incorporation into triglyceride by the T:% livers would have to be twice that of the euthyroid group to exhibit equal rates of synthesis of triglyceride.
However, as reported above, the total incorporation into perfusate triglyceride was approximately equal in both euthyroid and hyperthyroid livers, while the incorporation into hepatic triglyceride was higher in the euthyroid than in the hyperthyroid animals. Therefore, for both perfusate and hepatic triglyceride, the calculated rate of triglyceride synthesis was lower in the livers from T:{-treated rats than in those from euthyroid controls. The studies reported by Laker and Mayes (3) and Abrams et al. (4) showed, in agreement with our data, that the rate of triglyceride production under hyperthyroid conditions was lower than that in euthyroid. Although the data are not strictly comparable, Ylikahri (28) reached similar conclusions that thyroxine treatment of rats caused inhibition of ethanol-induced elevation of ol-glycerophosphate and the concomitant accumulation of serum and hepatic triglyceride.
Other investigators had calculated, contrary to our data, a higher rate of hepatic triglyceride synthesis in the hyperthyroid state than in the euthyroid (5)(6)(7)22). In the study of Roncari and Murthy (5), triglyceride synthesis was calculated as the percentage of [%]glycero-3-phosphate incorporated into triglyceride.
Glenny and Brindley (6) measured triglyceride synthesis by rats in uiuo in terms of the percentage of total glycerolipid synthesized; in their report, triglyceride synthesis from ['4C]palmitate was not altered by thyroxine treat-ment, whereas synthesis from ["HIglycerol was higher in the hyperthyroid than in the euthyroid condition. These two studies did not consider the endogenous pools of hepatic glycero-3-phosphate.
In comparison with the data reported here, the rates of triglyceride synthesis calculated by Roncari and Mm-thy (5) and Glenny and Brindley (6) would be equivalent to the calculated relative specific activity of triglyceride, shown in Table III. Calculation of triglyceride synthesis based on specific activity of the triglyceride produced from radioactive glycerol in hyperthyroidism is erroneous, because synthesis is an inverse function of the precursor specific activity. The report of Nikkila and Kekki (7) measured the rate of triglyceride synthesis in human patients in terms of turnover rate. It was not clear from their report how the calculated turnover rate translated into rate of triglyceride synthesis. In the three studies referred to above (5-7), radioactive glycerol or glycero-3-phosphate was used as precursor and in none of the studies was the endogenous pool of glycero-3-phosphate measured. Roncari and Murthy (5) incubated cell-free preparation with excess of glycero-3-phosphate, an experimental condition that might also mask the T:s-induced decrease of hepatic glycero-3-phosphate concentration reported here and by others (8,9). In the in uiuo studies, liver biopsies were not done. Therefore, it is appropriate for us to suggest that, although the data reported in these studies are correct and accurate, the conclusions are erroneous.
In addition to hepatic changes induced by thyroid hormone, plasma free fatty acid concentration is elevated by 'I':s-treatment, as shown by others (7,22,25) and in this report; the increased availability of free fatty acid substrate to the liver in uiuo per se may increase triglyceride production (29,30). Even though the hyperthyroid state is also associated with increased oxidation of free fatty acids (l-3, 22, 32), the elevated plasma free fatty acids, nevertheless, may be sufficient to increase plasma triglyceride concentrations above normal values in the hyperthyroid animals. Similar conclusions were obtained in studies of hepatic metabolism of freey fatty acids in experimental insulin deficiency (29). It has also been shown that plasma clearance of triglyceride is enhanced in hyperthyroid patients (4,7,26). This enhanced clearance is another factor which must be considered as contributing to the variability of plasma triglyceride concentrations in the hyperthyroid patient or animal.
The role of hepatic glycero-3-phosphate in substrate-induced stimulation of triglyceride synthesis has been investigated (28,31,32). High concentrations of glycero-3-phosphate may favor esterification of fatty acids to triglyceride, while low concentrations may decrease triglyceride synthesis. The maximal stimulation of triglyceride synthesis seems to be limited by the capacity of diacylglycerol acyltransferase (31). There are indications that the capacity of this enzyme is increased by thyroid hormone treatment (33). Since Ts-treatment reduced hepatic glycero-3-phosphate concentration, it is possible that the in uiuo range of concentration of this metabolite is below that required to saturate the a-glycerophosphate acyltransferase (31) which would lead to decrease of diglyceride concentration (6), which in turn may be lower than that required for saturation of diacylglycerol acyltransferase. This enzyme-substrate relationship is compatible with our observation of a decreased synthesis of triglyceride in the hyperthyroid rats. Moreover, enhanced gluconeogenesis in hyperthyroidism (34) is a frequently reported observation, and this may further decrease the pool of glycero-3-phosphate available for triglyceride synthesis.