Hepatic Lipogenesis and Cholesterol Synthesis in Hyperthyroid Patients

To determine the effect of hyperthyroidism on hepatic lipogenesis and cholesterol synthesis we measured these metabolic pathways (deuterated water method) in euthyroid and hyperthyroid subjects investigated in the postabsorptive state. Hyperthyroid patients had increased concentrations of glucose (P < 0.05), insulin (P < 0.05), nonesterified fatty acids (P < 0.01), and triglycerides (P < 0.05) and decreased levels of plasma cholesterol (P < 0.01). The contribution of hepatic lipogenesis to plasma triglycerides was largely increased in hyperthyroid subjects (23.0 ± 1.8% vs. 7.5± 0.2%; P < 0.001), whereas the fractional synthetic rate of cholesterol was moderately higher (5.0 ± 0.8% vs. 3.3 ± 0.2%; P < 0.05). mRNA levels of β-hydroxy-β-methyl glutaryl-coenzyme A reductase, measured in circulating mononuclear cells, were increased (P < 0.05), whereas those of low density lipoprotein (LDL) receptor and LDL receptor-related protein were unchanged. Sterol responsive element binding protein-1c mRNAs were undetectable in mononuclear cells from both groups of subjects. The large stimulation of hepatic lipogenesis in hyperthyroid patients is probably explained by both a direct action of thyroid hormones and the increase in insulin. It could contribute to their moderate rise in triglycerides levels. The decreased plasma cholesterol level is observed despite an enhanced synthetic rate and is thus related to an increased clearance rate. The lack of increased expression of LDL receptor and LDL receptor-related protein suggests that other receptors are implicated.

HUMAN HYPERTHYROIDISM is accompanied by multiple metabolic abnormalities, with increased energy expenditure and excessive mobilization and utilization of metabolic substrates (1–3). With respect to lipid metabolism it is clear that the breakdown of triglycerides (TG) stored in adipose tissue is enhanced by thyroid hormones excess, resulting in increased concentration and turnover of nonesterified fatty acids (NEFA) (3, 4). This increased availability of fatty acids is associated with a rise in the lipid oxidation rate (1). At the hepatic level this contributes to an elevated production rate of ketone bodies (3, 5). Modifications of fatty acids and TG synthesis are less well characterized. In vitro experiments clearly showed that the synthesis and secretion of TG by the liver is decreased by thyroid hormone excess (6, 7). In vivo, however, TG levels are often found to be elevated in hyperthyroid patients, and increased secretion rates have been reported (8). This could be explained by the in vivo enhanced turnover rate (3) and delivery of NEFA to the liver, resulting in an increased availability of fatty acyl-coenzyme A (acyl-CoA) for hepatic reesterification, as plasma NEFA delivery to liver is a major contributor to hepatic TG synthesis and secretion (9, 10). There is also evidence for a stimulation of hepatic lipogenesis by thyroid hormones. This is supported by in vitro and in vivo experiments (11), although some discordant results have been reported (12). Moreover, thyroid hormones stimulate the gene expression (13) and increase the amount and activities (14) of enzymes involved in the lipogenic pathway. However, all of these results were obtained in vitro or in experimental models of hyperthyroidism, and to our knowledge hepatic lipogenesis has never been measured in human hyperthyroidism. Therefore, our first aim was, taking advantage of stable isotope methodology, to measure this metabolic pathway in normal subjects and hyperthyroid patients to determine whether enhanced hepatic lipogenesis could contribute to the increased TG concentrations often observed in hyperthyroidism.

Hyperthyroid patients have also low cholesterol levels. This has been ascribed in part to an increased biliary excretion of cholesterol (15). In addition, the clearance of plasma cholesterol could be enhanced through thyroid hormone stimulation of the expression of the low density lipoprotein (LDL) receptor gene (16). There is no evidence for decreased cholesterol synthesis. On the contrary, there is some evidence for an increase in the activity of HMG-CoA reductase, the key regulatory enzyme for cholesterol synthesis (17, 18). No kinetic data of cholesterol metabolism during hyperthyroidism are available. Therefore, our second aim was to measure the fractional synthetic rate (FSR) of plasma free cholesterol in euthyroid and hyperthyroid subjects. In addition, we measured the mRNA concentrations of HMG-CoA reductase and LDL receptor. We also measured the mRNA level of LDL receptor-related protein (LRP), as it is also involved in the clearance of circulating lipoproteins (19). These measurements were performed in blood mononuclear cells, because there is evidence that basic regulatory mechanisms are similar in these cells and in hepatocytes, at least for HMG-CoA reductase and the LDL receptor (20, 21).

Subjects and Methods

Materials and methods

Deuterated water (99% mole percent excess) was obtained from Cambridge Isotope Laboratory (Andover, MA). Chemical and reactifs were purchased from Sigma (St. Louis, MO), Roche (Mannheim, Germany), or Pierce Chemical Co. (Rockford, IL).

Subjects

Informed written consent was obtained from 10 healthy volunteers and 5 hyperthyroid patients with Graves’ disease after full explanation of the nature, purpose, and possible risks of the study. The control group consisted of 6 women and 4 men (aged 20–42 yr; body mass index, 18–25). No control subject had a personal or familial history of diabetes, dyslipidemia, or obesity or was taking any medication; all had normal physical examinations and normal plasma glucose and lipid concentrations (Table 1). Subjects with unusual dietary habits were excluded. The hyperthyroid group consisted of 3 women and 2 men (aged 27–50 yr; body mass index, 19–25), free of other disease and taking no medication. Thyroid status was assessed clinically and by determination of thyroid hormone levels. All subjects consumed their usual diet on the days before the study.

Table 1.

Hormonal and metabolic parameters in the postabsorptive state

	Control subjects	Hyperthyroid patients
Glucose (mmol/liter)	4.46 ± 0.09	5.31 ± 0.59a
NEFA (μmol/liter)	367 ± 50	909 ± 92b
TG (mmol/liter)	0.81 ± 0.08	1.17 ± 0.12a
Total cholesterol (mmol/liter)	4.98 ± 0.28	3.77 ± 0.07b
HDL cholesterol (mmol/liter)	1.55 ± 0.11	1.33 ± 0.07a
Free cholesterol (mmol/liter)	1.10 ± 0.04	0.85 ± 0.02a
Insulin (pmol/liter)	41 ± 8	78 ± 14a
Glucagon (ng/liter)	167 ± 29	112 ± 23

^a

P < 0.05 vs. control subjects.

^b

P < 0.01 vs. control subjects.

Protocols

The protocol of the study was approved by the ethical committee of Lyon and by INSERM, and the study was conducted according to the Hurriet law. All tests were performed in the Centre de Recherche en Nutrition Humaine of Lyon.

For women the tests were performed during the first 10 d of the menstrual cycle to take into account the known variations in lipogenesis during the menstrual cycle (there are no menstrual variations in cholesterol synthesis) (22). On the evening before the test subjects drank a loading dose of deuterated water (3 g/kg body water; half after the evening meal, ingested between 1900–1930 h, and half at 2200 h). From then until the end of the study, they drank only water enriched with ²H₂O (4.5 g ²H₂O/liter drinking water). The following morning at 0730 h, in the postabsorptive state (i.e. after an overnight fast), an indwelling catheter was placed in a forearm vein, and blood samples were drawn for measurements of various concentrations and enrichments and for separation of circulating monocytes.

Analytical procedures

Metabolites were assayed with enzymatic methods on neutralized perchloric extracts of plasma (glucose) or on plasma (FFA and TG) (10). Plasma insulin and glucagon concentrations were determined by RIA (23, 24). Total cholesterol was measured by enzymatic assay, and the high density lipoprotein fraction was measured as previously described (25). Plasma lipids were extracted by the method described by Folch et al. (26). Free and esterified cholesterol, FFA, and TG were separated by TLC and scraped off the silica plates. Cholesterol was eluted from silica with ether before preparing its trimethylsilyl derivative (27). The methylated derivative of the palmitate of TG and FFA was prepared according to the method described by Morrison and Smith (28). Deuterium enrichment determinations were performed as previously described (29) on a gas chromatograph (HP5890, Hewlett-Packard Co., Palo Alto, CA) equipped with a 25-m fused silica capillary column (OV1701, Chrompack, Bridgewater, NJ) and interfaced with a mass spectrometer (HP5971A, Hewlett-Packard Co.) operating in the electronic impact ionization mode (70 eV). The carrier gas was helium. Ions 368–370 were selectively monitored for cholesterol, and ions 270–272 were monitored for palmitate. Deuterium enrichment in plasma water was measured by the method of Yang et al. (30). Special care was taken to obtain comparable ion peak areas between standard and biological samples, adjusting the volume injected or diluting the sample when necessary.

Measurement of LDL receptor and HMG-CoA reductase mRNA concentrations

Mononuclear cells were immediately isolated by centrifugation of whole venous blood on a Ficoll gradient at 4 C as previously described (31) and were stored at −80 C. Total RNA was prepared from frozen samples as described previously (25), then quantified by electrophoresis on 1% agarose gel of serial dilutions compared with a known amount of standard RNA (Roche). LDL receptor and HMG-CoA reductase mRNA copy numbers were determined by competitive RT-PCR. Total cellular RNA was reverse transcribed into cDNA in the presence of dilution series of PAW 109 RNA (Perkin-Elmer Corp.) an internal standard, which contains primer sites for the LDL receptor and HMG-CoA reductase. Reactions were performed in a final volume of 100 μl containing 10 mm Tris-HCl (pH 8.3), 50 mm KCl, 2.5 mm MgCl_2, 150 μm of each of the four deoxyribonucleoside triphosphates (Pharmacia LKB Biotechnology Inc., Piscataway, NJ), and 0.65 μm LDL receptor or HMG-CoA reductase downstream primers. Primer sequences were identical to those described by Powell et al. (21). The RT-PCR protocol consisted of 1 cycle at 42 C for 40 min, followed by 33 cycles of 94 C for 2 min, 58 C for 1 min, and 72 C for 3 min. All PCR included a negative control. The absence of genomic DNA contamination in RNA samples was confirmed by using RT-negative RNA samples. PCR products were analyzed on 2% agarose gel stained with ethidium bromide. For quantitation, densitometric scanning of photographic images was performed, and the relative amounts of target and competitor product in each sample were compared. The initial amounts of target and competitor were assumed to be equal when the molar ratio of target and competitor was equal to 1. The results were expressed as copy number per μg total cellular RNA.

Measurement of LRP mRNA concentrations

The LRP competitor transcript is a 410-bp long synthetic gene, constructed by PCR (Boucher, P., et al., submitted). Specific LRP mRNA was quantified by RT, followed by competitive PCR, which consists of the coamplification of known amounts of competitor DNA with the target cDNA in the same tube. Briefly, the reaction consists of a specific first strand cDNA synthesis, performed from 0.5–1 μg total RNA with 2.5 U thermostable reverse transcriptase (Tth DNA polymerase, Promega Corp., Charbonniere, France). The reaction was carried out for 3 min at 60 C, followed by 15 min at 70 C. The whole RT reaction was then added to 80 μl of a PCR mix and subjected to 120 sec at 94 C, followed by 40 cycles of 40 sec at 95 C, 50 sec at 55 C, and 50 sec at 72 C. Fifteen microliters of each PCR were separated on a 2% agarose gel stained with ethidium bromide and analyzed as described above for LDL receptor and HMG-CoA reductase genes.

Measurement of sterol responsive element binding protein (SREBP)-1c mRNA levels

As SREBP-1c is the main transcription factor controlling the expression of lipogenic genes (32), we also measured its mRNA levels in circulating mononuclear cells. A 311-nucleotide-long cDNA fragment was synthesized by RT-PCR from human adipose tissue total RNA using 5′-₈GCGGAGCCATGGATTGCAC₁₁-3′ as sense primer (specific for exon 1c of the SREBP-1 gene) and 5′-₃₁₁CTCTTCCTTGATACCAGGCCC₂₉₁-3′ as antisense primer. The competitor was obtained by adding 20 bp in the SREBP-1c cDNA fragment. The RT-competitive PCR assay was validated using synthetic mRNA. For assays of mRNAs in cell samples, serial dilutions of the competitor plasmids were performed. A RT reaction was performed from 0.5 μg total RNA with 2.5 U thermostable reverse transcriptase (Tth DNA polymerase, Promega Corp.) in 10 mm Tris-HCl (pH 8.3), 90 mm KCl, 1 mm MnCl₂, 0.2 mm deoxynucleoside triphosphates, and 15 pmol of the specific antisense primer in a final volume of 20 μl. The reaction consisted of 3 min at 60 C, followed by 15 min at 72 C and then 5 min at 99 C. Sense primers labeled in the 5′ position with Cy-5 fluorescent dye (Eurogentec, Seraing, Belgium) were used during the PCR. For this reaction, 20 μl RT medium were added to 80 μl PCR mix [10 μl 10 mm Tris-HCl (pH 8.3), 100 mm KCl, 0.75 mm EGTA, and 5% glycerol] containing 0.2 mm deoxynucleoside triphosphates, 5 U Taq polymerase (Life Technologies, Inc., Cergy Pontoise, France), 45 pmol of the corresponding sense primer, and 30 pmol of the antisense primer. Four 20-μl aliquots were then transferred in PCR tubes containing 5 μl of defined working solutions of the competitor cDNA. The PCR conditions were 2 min at 94 C, followed by 40 cycles (40 sec at 94 C, 60 sec at 55 C, and 40 sec at 72 C), and finally 10 min at 72 C. The PCR products were analyzed with an automated laser fluorescence DNA sequencer (ALFexpress, Pharmacia Biotech, Uppsala, Sweden) in 4% denaturing polyacrylamide gels. The amounts of PCR products (competitor and target) were calculated by integrating peak areas using the Fragment Manager software from Pharmacia Biotech. The initial concentration of target mRNA was determined at the competition equivalence point.

Calculations

The fractional contribution of cholesterol synthesis to the plasma free cholesterol pool was calculated from the deuterium enrichments in free cholesterol and plasma water, as previously described (27, 29). In short, the deuterium enrichments that would have been obtained if endogenous synthesis were the only source of plasma cholesterol were calculated from plasma water enrichment. Comparison of the actual enrichments observed with these theoretical values gives the contribution, expressed as FSR, of endogenous synthesis to the pool of rapidly exchangeable free cholesterol during the time between the ingestion of the loading dose of deuterated water and blood sampling (12 h). This value was then transformed as the absolute synthetic rate (ASR), expressed as miligrams per 12 h. For this calculation we calculated first the total pool of rapidly exchangeable cholesterol (M₁, which comprises cholesterol in plasma, liver, intestine, and blood cells) according to the equation developed by Goodman et al. (33). ASR was then calculated first as ASR = FSR × M_1. As M₁ comprises both esterified and free cholesterol, and we found deuterium enrichment only in free cholesterol, we calculated ASR in the rapidly exchangeable free cholesterol pool, estimating that the ratio in plasma of free over total cholesterol concentrations is representative of this ratio in the whole pool. Results are shown as the mean and sem. Comparisons between values of the euthyroid and hyperthyroid groups were performed with Student’s two-tailed t test for nonpaired values.

Results

Hormonal and metabolic parameters (Table 1)

Hyperthyroid patients in the postabsorptive state had slightly increased levels of glucose and insulin (P < 0.05), whereas glucagon concentrations were comparable. TG levels were also slightly higher (P < 0.05), and NEFA were greatly increased (P < 0.01). Total and free cholesterol concentrations as well as high density lipoprotein cholesterol were decreased in hyperthyroid patients (P < 0.01).

Hepatic lipogenesis and cholesterol synthesis

Deuterium enrichments in plasma water were 0.34 ± 0.01% and 0.27 ± 0.01% in hyperthyroid and control subjects, respectively. The corresponding enrichments were 0.47 ± 0.05% and 0.26 ± 0.01% in plasma free cholesterol and 1.75 ± 0.18% and 0.44± 0.01% in the palmitate of plasma TG (these enrichments can be higher than in plasma water, because there are multiple possible incorporation sites of deuterium in the molecules of cholesterol and palmitate during their synthesis). No enrichment was found in the palmitate of FFA, ruling out any possible contribution of the uptake and reesterification of plasma FFA to the deuterium enrichment found in the palmitate of plasma TG. Figure 1 shows the individual values of the fractional contribution of hepatic lipogenesis and cholesterol synthesis to the plasma pools of TG and free cholesterol in the two groups of subjects. In hyperthyroid patients there was a moderate increase in the FSR for cholesterol (5.02 ± 0.80% vs. 3.30 ± 0.20%; P < 0.05) and a large increase in hepatic lipogenesis (23.02 ± 1.83% vs. 7.47 ± 0.19%; P < 0.001). The ASR of cholesterol calculated during the period of the study (12 h) with the total pool of rapidly exchangeable cholesterol were 1116 ± 240 and 746 ± 105 mg in hyperthyroid and control subjects, respectively (P< 0.05). When the ASR were calculated for the fraction of free cholesterol in this pool the values were 253 ± 56 and 165 ± 19, respectively (P < 0.05).

Figure 1.

Individual values for hepatic lipogenesis and cholesterol synthesis in control subjects (•) and hyperthyroid patients (▵).

mRNA values

Table 2 shows the concentrations of HMG-CoA reductase, LDL receptors, and LRP mRNAs measured in circulating mononuclear cells, shown as absolute values. Compared with control subjects, hyperthyroid patients had increased mRNA levels for HMG-CoA reductase (P < 0.05). The decreased values for LDL receptor mRNA in hyperthyroid patients were borderline (P = 0.08). The values for LRP were comparable. Despite the use of a large amount of total RNA during the RT-competitive PCR assay, SREBP-1c mRNA was undetectable in mononuclear cells of control and hyperthyroid subjects, even when using the lowest concentration of competitor.

Table 2.

mRNA concentrations in circulating mononuclear cells

	HMG-CoA reductase	LDL receptor	LRP
Control subjects	972 ± 174	38 ± 7	114 ± 18
Hyperthyroid patients	2843 ± 557a	22 ± 7	108 ± 36

Results are shown as number of copy × 10^d perμ g total RNA.

^a

P < 0.05 vs. control subjects.

Discussion

The present results show that hepatic lipogenesis is largely increased in human hyperthyroidism. This is in agreement with previous data obtained in vitro or in experimental animal models of thyroid hormone excess (11, 34). This stimulation of fatty acid synthesis occurs simultaneously with enhanced lipolysis and lipid oxidation rate (1, 3, 4). The parallel stimulation of synthesis and degradation represents another enhanced metabolic cycle that could contribute to the increased energy expenditure of hyperthyroid patients. Hepatic lipogenesis is usually only a minor contributor to liver TG synthesis and secretion (9, 10). However, the 3-fold rise in lipogenesis we observed in hyperthyroid patients suggests that this metabolic pathway contributes, in addition to the enhanced NEFA delivery to the liver, to the increased TG levels found in our patients and to the enhanced TG secretion rate previously reported during hyperthyroidism (8).

Several mechanisms can intervene in this stimulation of lipogenesis. First, thyroid hormones could act directly; they have been shown to stimulate the activity of the promoter I of acetyl-CoA carboxylase I (the isoform of acetyl-CoA carboxylase involved in fatty acid synthesis) (35) and that of the human fatty acid synthase promoter (36). In addition, thyroid hormone increases the transcription rate of malic enzyme, ATP citrate lyase, and l-pyruvate kinase, genes coding for enzymes involved in the control of lipogenesis through the production of acetyl-CoA and NADPH (37). Second, thyroid hormone could act through modifications of the metabolic and hormonal environment of the liver. Both insulin and glucose are implicated in regulation of the expression of key regulatory genes of hepatic glycolytic and lipogenic pathways (38, 39). Although moderate, the increase in glucose and insulin concentrations of hyperthyroid patients could thus have played a role. This is all the more possible, because glucose has been shown to act in synergy with thyroid hormone to stimulate the expression of lipogenic genes (37, 40). This action of insulin/glucose is mediated in part through the transcription factor SREBP-1c (32, 39). We tried therefore to use mononuclear cells as a substitute for hepatocytes for the determination of SREBP-1c mRNA levels. However, in both groups of subjects SREBP-1c mRNA in these circulating cells was undetectable even when using the most sensitive conditions for the assay. Therefore, with respect to SREBP-1c expression, mononuclear cells cannot be used as a substitute for hepatocytes.

Cholesterol synthesis, as appreciated by the deuterated water method, was also stimulated, as both the fractional and absolute synthetic rates were moderately increased in hyperthyroid patients. Our finding of increased mRNA levels for HMG-CoA reductase in circulating mononuclear cells suggests that this rise in cholesterol synthesis is linked not only to increased substrate availability through stimulation of the glycolytic pathway, but also to an increase in the expression of regulatory genes. Moreover, in vitro studies showed that thyroid hormone increases the mRNA, protein, and activity of HMG-CoA reductase (17, 18). As cholesterol synthesis is stimulated, the lower cholesterol level of hyperthyroid patients is explained only by an increased removal. Although we cannot exclude in the absence of direct measurement an increase in their activity, our finding that both LDL receptor and LRP mRNA levels were not increased does not support a role for these lipoprotein receptors in this increased clearance. Other proteins involved in the reverse transport of cholesterol, such as scavenger receptor B-I (41) and ATP binding cassette 1 (42), could be responsible for this increased removal of plasma cholesterol. It is also possible that mononuclear cells are not representative of all aspects of liver cholesterol metabolism during hyperthyroidism. Actually thyroid hormones stimulate the expression of cholesterol 7α-hydroxylase, a key regulatory gene of the synthesis of bile acids (43), and the biliary secretion of cholesterol and bile acids (44). In addition to contributing to the decrease in plasma cholesterol, these effects result in a secondary stimulation of the expression of HMG-CoA reductase and LDL receptor in liver. Such a mechanism is not operative in mononuclear cells. Therefore, these cells could be more representative of the effects of thyroid hormones on the expression of key regulatory genes of cholesterol metabolism in extrahepatic tissues than in liver.

In conclusion, we found that both hepatic lipogenesis and, to a lesser extent, cholesterol synthesis are increased in hyperthyroid patients. These stimulations are probably explained by an increased expression of the genes involved in lipid synthesis. The low plasma cholesterol level of hyperthyroid patients is due to an increased clearance rate.

Abbreviations:

 
• ASR,

  Absolute synthetic rate;

 
• CoA,

  coenzyme A;

 
• FSR,

  fractional synthetic rate;

 
• HMG,

  β-hydroxy-β-methyl glutaryl;

 
• LDL,

  low density lipoprotein;

 
• LRP,

  low density lipoprotein receptor-related protein;

 
• NEFA,

  nonesterified fatty acids;

 
• SREBP,

  sterol response element binding protein;

 
• TG,

  triglycerides.