Thyroid Hormone Enhances the Ability of Serum to Accept Cellular Cholesterol via the ABCA1 Transporter

Abstract

Objective

The goal of this study was to examine the effects of thyroid hormone status on the ability of serum to accept cellular cholesterol.

Methods and Results

Sera from hypophysectomized rats treated ± T₃ was used to evaluate the role of thyroid hormone on serum efflux capacity. 2D-DIGE analysis of serum proteins showed that T₃ treated rats had increased ApoA-I, ApoA-IV and fetuin A levels with decreased Apo E levels. Microarray and real-time RT-PCR analysis of rat liver revealed large increases in ApoA-I, ApoA-IV, ABCG5, and ABCG8 in response to T₃. J774 macrophages, BHK cells, and Fu5AH rat hepatoma cells were used to measure cholesterol efflux mediated by ABCA1, ABCG1 transporters or SR-BI. Sera from T₃-treated rats stimulated efflux via ABCA1 but not by ABCG1 or SR-BI. Gel filtration chromatography revealed that T₃ treatment caused a decrease in HDL particle size accompanied by higher levels of lipid-poor ApoA-I.

Conclusions

Thyroid hormone enhances the ability of serum to accept cellular cholesterol via the ABCA1 transporter. This effect is most likely attributable to increases in small HDL and lipid poor ApoA-I in response to T₃.

Introduction

It is well established that thyroid hormone acts to lower serum cholesterol levels^{1, 2}. The mechanism by which this occurs appears to involve the modulation of several key processes in cholesterol metabolism. These likely include: increased elimination of neutral sterols and bile acids from the body, decreased rates of intestinal absorption of cholesterol, and increased removal of LDL cholesterol from the circulation by the liver^{3, 4}. These findings have lead to the investigation of thyroid hormone mimetic compounds, thyromimetics, as a treatment for hypercholesterolemia^{5, 6}. Recent clinical trials have found that thyromimetics are capable of lowering plasma LDL cholesterol and increasing bile acid synthesis in the absence of negative cardiac side effects⁷. Furthermore, animal studies utilizing the liver-selective thyromimetic T-0681 have shown that treatment with T-0681 promotes reverse cholesterol transport as measured by fecal sterol excretion⁸. Additional knowledge on the role of thyroid hormone in foam cell formation and cholesterol efflux will be beneficial to the future clinical use of thyromimetics in the treatment of hypercholesterolemia and its associated diseases⁹. However until now, there has been very little information regarding the effects of thyroid hormone on cholesterol efflux capacity - the quintessential anti-atherogenic property of HDL.

Low HDL levels are a strong independent risk factor for death from cardiovascular disease, however HDL cholesterol levels may be an inadequate indicator of HDL function. Strategies that enhance the anti-atherogenic properties of HDL, particularly its ability to promote cholesterol efflux from the macrophage and thereby drive reverse cholesterol transport, are being actively pursued¹⁰. The contributions of the ABCA1 and ABCG1 transporters in macrophage cholesterol transport to HDL and ApoA-I are well-documented^{11, 12}. In this study, we sought to examine the effects of thyroid hormone status on the ability of HDL to promote cholesterol efflux from the macrophage. Using two hypothyroid rat models, we investigated the effect of thyroid hormone on ABCA1, ABCG1, and SR-BI-mediated cholesterol efflux ex vivo. We report that serum from thyroid hormone treated animals promotes cholesterol efflux via the ABCA1 transporter but not ABCG1.

Methods

Experimental Animals

Hypophysectomized, thyroidectomized and normal male Sprague- Dawley rats weighing 125 to 150 g were purchased from Harlan (Indianapolis, IN). Hypophysectomized rats received Tekland Iodine Deficient chow and water ad libitum and were housed in a reverse-cycle light controlled room with a 12-hour light period followed by a 12-hour dark period. The animals were cared for according to the NIH guidelines set forth in the “Guide for the Care and Use of Laboratory Animals” and specifically in accord with protocol 35 71 approved by the University of South Florida Institutional Animal Care and Use Committee. Thyroidectomized rats received 1% calcium gluconate as their drinking water in addition to the Iodine deficient chow. The rats were maintained on the Iodine deficient diet for at least 14 days prior to being used in experiments in order to achieve sufficient turnover of T₄. The normal rats were fed Tekland 22/5 rodent chow ad libitum. Hypophysectomized and thyroidectomized rats were given an initial injection of 1.0 mg/kg T₃ 72 hours prior to harvest of tissue and an additional injection of 0.25 mg/kg T₃ 24 hours prior to harvest of tissue. In both hypothyroid animal models, administration of T₃ alone restored cholesterol efflux via ABCA1 transporters, indicating that other pituitary regulated hormones are not required.

Cholesterol, LDL/VLDL, HDL and T₃ Assays

Blood was collected and centrifuged at 16,000×g for 5 min. The supernatant (serum) was collected and used for these determinations. HDL levels were measured using the HDL and LDL/VLDL Cholesterol Quantification Kit from BioVision (Cat# K613-100; Mountain View, CA). T₃ levels were measured using the Free T₃ ELISA from Calbiotech (Cat# F3106T; Spring Valley, CA).

RNA Isolation

A portion of liver (about 200 mg) was quickly excised and immediately homogenized in 4 ml of Tri-Reagent from Molecular Research Center (Cincinnati, OH) using a Polytron homogenizer at room temperature. The remainder of the isolation steps was carried out using volumes corresponding to 4x the manufacturer’s recommendations.

Microarray Analysis

Isolated RNA was further purified using the RNeasy kit from Qiagen. RNA integrity was confirmed and microarray analysis was performed by the Moffitt Core Facility (Tampa, FL) using the Affymetrix GeneChip instrument system following the protocol established by Affymetrix, Inc. Ten µg each of RNA from the livers of 3 Hx and 3 Hx+T₃ rats was used in the analysis. An Affymetrix GeneChip Rat Genome 230 Plus 2.0 arrays which detects about 28,000 genes was used for the analysis. Multiple oligos are used for each gene with the data averaged. Scanned chip images were analyzed using GeneChip algorithms.

Real Time PCR

To validate the microarray results, we assessed the expression of a subset of genes using real-time PCR. RNA was DNAse treated using the TURBO DNA-Free Kit (Ambion, Austin, TX). cDNA was prepared using the Reverse Transcription System (Promega) per the manufacturer’s protocol. The primer sequences are listed in the Supplementary Materials. Hepatic mRNA was quantified under the following reaction conditions: 95°C for 5 minutes, followed by 40 cycle of 95°C for 15 seconds, 61°C for 1 minute and melt curve 55°C + 0.5°C each 10 seconds, × 80. All samples were run in duplicate on a Bio-Rad Chromo4 DNA Engine thermal cycler using SYBR green chemistry. Relative mRNA was calculated as a function of the internal control 18s using ΔΔ Ct.

Cell Culture and Efflux Studies

Efflux studies were performed as previously described¹³ using either J774 macrophages, Fu5AH rat hepatoma cells, or BHK cells expressing ABCG1 (gift from Dr. Jack Oram, University of Washington School of Medicine). J774 cells were grown in 20%FBS-RPMI, Fu5AH cells in 10%CS-EMEM and BHK-GI cells in 10% FBS-DMEM in the presence of antibiotics. For efflux, cells were plated in 24-well plates (J774 and Fu5AH: 70.000cells/well; BHK:125,000 cells/well) and radiolabeled using [1,2-³H] cholesterol (Perkin Elmer. J774: 2mci/ml in 5%FBS-RPMI; Fu5AH: 2mci/ml in 5%CS-EMEM and BHK:1mci/ml in 2.5% FBS-DMEM) in the presence of an ACAT inhibitor (2 µg/ml, CP 113,818, a gift from Pfizer), also present during efflux, to prevent accumulation of cholesteryl esters. J774 cells were treated with 0.3 mM c-AMP (cpt-AMP, Sigma) in 0.2% BSA for 16–18 hours to upregulate ABCA1. BHK cells were treated with 10 nM mifepristone for 16–18 hours to upregulate ABCG1. Fu5AH were pretreated for 2h with MEM-2% BSA + 1µM BLT to inhibit SR-BI. All efflux medium was prepared using serum from ≥4 Tx or Hx rats treated ±T₃ diluted to 2% in MEM-HEPES and sera were assayed in triplicate. Efflux was measured as the fraction of total radiolabeled cell cholesterol released to the medium in 4h and is corrected for the small amount released to MEM. In all experiments we also measured efflux to a pool of human serum diluted to 2%. ABCA1 efflux is considered to be the difference in efflux between control and c-AMP treated J774 cells and ABCA1 expression was verified by increased efflux to 20 µg/ml ApoA-I from c-AMP treated cells. SR-BI efflux is the BLT-1 sensitive efflux calculated from cells treated ± BLT and the inhibitor effect was monitored as loss of efflux to 25 µg/ml HDL₃. ABCG1 efflux is considered to be the difference in efflux between control and mifepistrone treated BHK cells and ABCG1 expression was verified by increased efflux to 25 µg/ml HDL₃ from mifepristone-treated cells.

2D-DIGE

Two dimensional differential in-gel electrophoresis was performed by Applied Biomics (Hayward, CA). Briefly, equal volumes of serum from 3 Hx rats or 3 Hx + T₃ rats was combined to form a pool for each treatment. A pool of serum from 2 normal rats was included as a control. Each pool was fluorescently labeled with different CyDye (Hx+T₃ - red, Hx-green, normal-blue) for downstream visualization. The samples were run on first dimension isoelectric focusing and second dimension SDS-PAGE. Fluorescent images were captured using a Typhoon image scanner and analyzed using ImageQuant software. Differentially expressed proteins were quantified using DeCyder software analysis, cut out, and subjected to in-gel trypsin digestion followed by protein identification by MALDI-TOF mass spectrometry.

Fast performance liquid chromatography (FPLC) separation of lipoproteins

Pooled rat sera (150 µL) was diluted in running buffer (150 mM NaCl, 1 mM EDTA pH 8.0) and run over a two superpose 6 10/300 GL columns arranged in tandem (Pharmacia Biotech). A total of forty-six fractions (500 µL each) were collected and assayed for total cholesterol using the Cholesterol E kit from WAKO. The following antibodies were used for western blotting of FPLC fractions, serum and livers: Rabbit anti-mouse Apolipoprotein A-I Biodesign K23500R, Rabbit anti-ApoE Biodesign K23100R, a polyclonal antibody to the LDL receptor⁴, and mouse monoclonal anti-ABCA1 (Abcam ab18180).

Statistical Analyses

Data are expressed as means ± standard error for a minimum of four animals per group. Experimental treatments included hypophysectomized or thyroidectomized Sprague–Dawley and hypophysectomized Sprague–Dawley rats treated with T₃. Treated animals were only compared to their non-treated matching surgical counterparts. In comparisons of multiple groups, a one-way between groups analysis of variance (ANOVA) was conducted initially to determine if differences existed. Tukey’s post test was then used to compare between the relevant treatment and control groups.

Results

Thyroid Hormone Treatment Dramatically Changes Levels of Serum Apolipoproteins

We sought to examine the effects of thyroid hormone treatment on the serum proteome in an unbiased manner. To accomplish this we used hypophysectomized (Hx) rats, a surgically modified animal model in which the pituitary gland has been removed to render the animal hypothyroid. To look specifically at the effects of thyroid hormone, these animals were also injected with T₃. Serum proteins from Hx rats and those replenished with thyroid hormone (Hx + T₃) were labeled and subjected to two dimensional differential in-gel electrophoresis (2D-DIGE) (Fig. 1A). Differentially expressed proteins were excised from the gel, digested with trypsin, and identified by MALDI-TOF mass spectrometry. The determined protein sequences were compared to the database and those with high confidence (95–100%) are listed in Table 1. ApoA-I was identified in four separate spots, where its levels were increased 5 to 12 fold by T₃. Serum ApoA-IV was also increased 2.4-fold by T₃. Fetuin-A, serine protease inhibitor alpha 1, fibrinogen alpha subunit, and Clu protein were also increased. Serum Apo E was decreased 3 to 8-fold in response to T₃ treatment, consistent with the known effects of T₃ on LDL receptor expression¹⁴. To confirm the changes in apolipoprotein levels, we examined the serum of these animals by western blot analysis. ApoA-I was substantially reduced in Hx rats relative to normal animals. T₃ treatment fully restored ApoA-I levels to those seen in normal rats. In contrast, ApoE levels were markedly decreased by T₃ treatment (Fig. 1B).

Figure 1.

A. 2D DIGE of Rat Serum. Serum was pooled from 3 hypophysectomized rats treated ± T₃ and fluorescently labeled for analysis by 2D DIGE. This image shows gel analysis of pooled serum from Hx+T₃ rats (red) overlayed with pooled serum from Hx rats (green). Red spots indicate proteins expressed in the Hx+T₃ sera and green spots indicate proteins expressed in the Hx sera. Yellow spots are the result of similar protein expression in both conditions. The fold-change of differentially expressed proteins was quantified using ImageQuant software analysis. Proteins were then cut out, subjected to in-gel trypsin digestion, and identified by MALDI-TOF mass spectrometry. B. ApoA-I western blot analysis. Sera from the individual animals pooled for 2D DIGE analysis in Figure 1A was analyzed by western blot for ApoA-I and ApoE expression. Normal serum is shown as a control.

Table 1.

List of serum proteins altered by thyroid hormone

Spot	Protein	Fold Change
12	ApoA-I	+12.5
5	Fetuin-A	+10.8
44	ApoA-I	+8.54
14	PreproapoA-I	+7.80
4	Fetuin-A	+6.12
13	ApoA-I	+5.02
33	Fetuin-A	+4.57
32	Fetuin-A	+4.12
24	Serine protease inhibitor α1	+3.61
10	Fibrinogen a subunit	+3.48
23	Clu protein	+3.29
22	Clu protein	+2.66
11	ApoA-IV	+2.38
18	ApoE	−3.45
21	ApoE	−8.62

The differentially expressed proteins, by spot number in Fig. 1, were quantified. The fold-increase or decrease in response to T₃ is presented. The spots were cut out and subjected to in-gel trypsin digestion followed by protein identification by MALDI-TOF ass spectrometry.

Hepatic ApoA-I and ApoA-IV are Induced by Thyroid Hormone

We then tested whether thyroid hormone acts to increase serum ApoA-I and ApoA-IV by increasing hepatic transcription of these genes. A microarray analysis of liver RNA isolated from Hx rats treated ± T₃ was performed. Supplementary Table 2 shows a list of selected genes that were found to positively or negatively respond to T₃ treatment. The cholesterol transporter ABCG5 and ApoA-IV showed the greatest activations by T₃. Phospholipase A2 showed the greatest decrease, 142-fold, in response to T₃. This protein plays a crucial role in vascular inflammation and may contribute to atherosclerotic development and progression¹⁵. We did not detect ApoA-I due to limitations of the array. We confirmed the microarray findings using to real-time RT-PCR analysis. As shown in Table 2, ApoA-I and ApoA-IV were both induced about 10-fold in response to T₃ treatment. As controls, we observed 10-fold and 90-fold inductions by T₃ for ABCG5 and ABCG8 respectively. These proteins mediate transport of cholesterol from liver into bile and are known to be induced by T₃³.

Table 2.

RT-PCR analysis of hepatic genes that may contribute to cholesterol efflux

Gene	Hx	Hx + T3
ApoA-I	1.02 ±0.20	12.17 ±6.85*
ApoA-IV	1.19 ±0.80	13.29 ±8.79*
ApoE	2.11 ±0.95	1.04 ±0.16
ABCG5	1.16 ±0.66	13.86 ±1.93*
ABCG8	1.73 ±2.19	166.1 ±44.8*

Relative RNA levels are presented as means ±SD for ≥4 rats per group.

^*p < 0.01 statistical significance relative to Hx.

Thyroid Hormone Promotes Macrophage Cholesterol Efflux via ABCAI

In order to determine if thyroid hormone affects the functionality of HDL in these animals, we performed cholesterol efflux assays. To rule out possible effects by alterations in other hormones in the hypophysectomized rats, we also used the thyroidectomized (Tx) rat for these studies. In our hands, hypophysectomy more reliably ensured a hypothyroid state, which is why most of our studies were performed in this model. Serum was obtained from Hx and Tx hypothyroid animals, as well as Hx and Tx rats replenished with T₃. The sera from these animals was tested for its ability to promote cholesterol efflux from cells. Different cell lines were used to specifically evaluate efflux via ABCA1, ABCG1 and SR-BI as described in the Materials and Methods section. As shown in Fig. 2A, serum from Hx rats treated with T₃ increased ABCA1-mediated efflux over 2-fold as compared to serum from control or Hx rats. Compared to Tx rats, serum from Tx rats treated with T₃ caused a 7-fold increase in efflux mediated by ABCA1 (Fig. 2A). The greater fold increase in Tx rats likely relates to the lower T₃ levels in the Tx rats (Supplementary Table 3). The effect of T₃- treated sera on ABCA1 mediated efflux is almost certainly not due to T₃ itself, or other small molecules, as dialysis had no effect on efflux capacity (Supplementary Figure 1).

Figure 2.

Effects of Thyroid Hormone Deficiency and Administration on Cholesterol Efflux. Serum (2% v/v) from normal, hypophysectomized and thyroidectomized rats treated ± T₃ was used to measure transporter-mediated cholesterol efflux capacity in J774 (ABCA1), BHK (ABCG1), and Fu5AH (SR-BI) cells as described in the Methods (*p<0.05 by ANOVA and Tukey’s Post-test).

We also measured cholesterol efflux via ABCG1 using BHK cells harboring an inducible ABCG1 expression cassette (Fig. 2B). ABCG1 is known to promote cholesterol efflux to a variety of lipoprotein acceptors by a mechanism consistent with driving aqueous diffusion¹⁶. Despite the much lower LDL and VLDL levels in the T₃ treated sera, ABCG1-mediated efflux was unaffected. Recent work has shown that in addition to their known role in promoting ABCA1-mediated efflux, pre-beta HDL can also promote ABCG1-mediated efflux¹⁷. To test whether increased ABCG1-mediated efflux to pre-beta was compensating for the lack of efflux to ApoB lipoproteins (ApoB Lp), we performed additional efflux studies with PEG-precipitated sera. We found no statistically significant difference in the ABCG1-mediated efflux to the PEG supernatants from T₃ treated sera (data not shown), suggesting the specific efflux capacity of the ApoB Lp may be affected. Lastly, we also tested the ability of T₃ treated sera to promote SR-BI mediated efflux using Fu5AH cells treated with and without the SR-BI inhibitor BLT-I. We found that SR-BI mediated efflux to T₃ treated sera was unaffected in this cell system (Fig. 2C).

Thyroid Hormone Replacement Increases Small HDL and Lipid Poor ApoA-I

We then investigated whether changes in HDL levels could explain the large increase in ABCA1-mediated efflux capacity in the serum from T₃ treated animals. Total cholesterol and VLDL/LDL cholesterol levels were substantially lower in animals receiving T₃, consistent with the known actions of this hormone on LDL receptor activity (Supplementary Table 3). Interestingly, despite the large changes in ApoA-I and ApoA-IV protein levels, HDL was not increased with T₃ treatment. We hypothesized that the increase in ABCA1-mediated efflux could be explained by a change in HDL size or lipid poor ApoA-I levels. To address this, we separated serum lipoproteins by gel filtration chromatography to examine the lipoprotein profile (Fig. 3). Interestingly, T₃ treatment reduced LDL cholesterol levels and dramatically shifted the HDL cholesterol peak to the right, indicating an increase in smaller HDL particles at the expense of larger HDL. We also observed considerably higher ApoA-I levels in the “lipid free fractions” (normally fractions 42–46), likely indicating a substantial increase in pre-beta HDL/ lipid poor ApoA-I. Western blotting of livers from these animals revealed increased LDLR and ApoA-I protein levels, strongly supportive of greater LDL clearance and increased ApoA-I production (Supplementary Figure 2).

Figure 3.

A. FPLC Analysis of Rat Serum Serum from normal, hypophysectomized, and hypophysectomized rats treated with T₃ were fractionated and assayed for cholesterol in the upper panel. The lower panel displays Western blots for ApoA-I of selected fractions from 26 to 46. B. Supplementation of Hypothyroid Serum with ApoA-I. Efflux assays using serum from Hyphysectomized rats (Hx) supplemented with varying amounts of ApoA-I.

ApoA-I Promotes ABCA1-mediated Efflux to Serum from Hyphysectomized rats

In order to test whether changes in ApoA-I levels can explain the increased efflux to serum from thyroid hormone treated animals, we performed additional efflux assays using serum from hyphysectomized rats (Hx) supplemented with varying amounts of ApoA-I. Addition of human ApoA-I more than tripled ABCA1-mediated efflux at the highest dose tested (10 ug/mL), indicating that ApoA-I is capable of rescuing the defective efflux seen in hypothyroid sera. However, when added back at lower doses of ApoA-I (2 ug/mL) to the levels seen in T₃ treated rats, efflux was not completely rescued (Figure 3B). This suggests that changes in lipid poor ApoA-I alone may be insufficient to fully explain the effects of T₃ on serum efflux capacity. Alterations in HDL size, lipid and protein composition are likely involved as well.

Discussion

Hypothyroidism has been shown to be associated with hypertriglyceridemia as well as hypercholesterolemia^{18, 19}. Thyroid hormone and thyromimetics are known to exert several actions that lead to lowered serum cholesterol levels, and an improved lipid profile, which is associated with a decreased risk of atherosclerosis^{2, 5, 7, 14}. However, the role of thyroid hormone in promoting cholesterol efflux has not been previously investigated. This study was performed to test the effects of thyroid hormone treatment on the cholesterol efflux capacity of serum.

The contributions of ABCA1 and ABCG1 in macrophage cholesterol transport to HDL and ApoA-I are well-documented^{11, 12}. SR-BI mediates selective uptake of cholesterol and cholesteryl esters from HDL and has been shown to promote efflux of ³H-cholesterol out of the cell^{20, 21}. The present study demonstrates that thyroid hormone (T₃) enhances the ability of serum to accept cellular cholesterol via the ABCA1 transporter, but not the ABCG1 or SR-BI transporters. The enhanced efflux was not solely related to total serum HDL cholesterol levels, supporting the concept that particle size and composition are important factors influencing the cholesterol efflux capacity of HDL.

Analysis of serum lipoproteins by gel filtration chromatography allowed us to examine the effect of thyroid hormone on the lipoprotein profile. We observed a shift to the right in the HDL cholesterol peak in response to T₃, suggesting an increase in small HDL particles as well as lipid poor ApoA-I. Small discoidal pre-beta HDL particles are better cholesterol acceptors than mature lipid-rich HDL particles via the ABCA1-mediated pathway^{22, 23}. In fact, treatment of mice with ApoA-I mimetic peptides was shown to increase the formation of pre-beta high-density lipoprotein and thereby increase cholesterol efflux²⁴. Delipidation of lipid-rich HDL to pre-beta HDL was shown to increase the efficacy of plasma to stimulate ABCA1-mediated cholesterol transfer from monocytic cells to HDL and improved aortic atherosclerosis in African Green Monkeys²⁵. We were surprised to see that despite lower T₃ levels accompanied by lower HDL in the Hx rat sera, ABCA1-mediated efflux was not significantly lower than that for normal sera. While we do not have an obvious explanation for this, it may be due to alterations in the protein or lipid composition of the HDL. Nonetheless, this study presents the first clear evidence that thyroid hormone replacement substantially improves the ability of serum lipoproteins to accept cholesterol via an ABCA1-dependent mechanism.

We also tested whether the efflux we observed was solely attributable to increased ApoA-I. To accomplish this, we performed additional efflux assays in which serum was reconstituted with purified human ApoA-I. Addition of purified human ApoA-I dramatically increased efflux to Hx sera. However, at doses determined to be equivalent to sera from Hx + T₃ animals, ApoA-I could not fully restore the ABCA1-mediated efflux (Figure 3B). Taken together this indicates that while ApoA-I contributes to the T₃-stimulated efflux capacity, the quality and quantity of the HDL are probably also very important. Supporting this idea, it was recently found that HDL efflux capacity is in fact a better predictor of coronary artery disease status than HDL levels alone²⁶. Future studies are needed to test the effects of thyroid hormone on the cholesterol efflux capacity of human serum, as well as the corresponding alterations in the protein and lipid composition of the HDL.

We discovered a number of effects of thyroid hormone that were of significance, but unfortunately out of the scope of this study. Of interest, is the induction of serum fetuin A by thyroid hormone. Fetuin A has been suggested to play a role in the pathophysiology of coronary artery disease²⁷ and is implicated in an increased risk for cardiovascular disease²⁸. On the other hand, fetuin A inhibits pathologic calcification in both the soft tissue and vasculature, even in the setting of atherosclerosis²⁹. The examination of the role of fetuin A in cardiovascular disease has began relatively recently and certainly requires further investigation.

Also of interest is the profound decrease of hepatic phospholipase A2 mRNA in response to T₃ treatment. PLA2 is a small lipolytic enzyme that releases fatty acids from the second carbon of glycerol phospholipids generating lysophosphatidylcholine, a proinflammatory mediator¹⁵. Inflammation has been shown to impair reverse cholesterol transport in vivo³⁰. PLA2 is up-regulated in hepatocytes in response to stress stimuli and is secreted from liver cells. More importantly, overexpression of PLA2 in mice increased the incidence of atherosclerotic lesions, decreased plasma HDL, and increased plasma LDL³¹. PLA2 has been identified in atherosclerotic lesions and has been suggested to aid in the transformation of macrophages into foam-cells by promoting macrophage LDL uptake^{32, 33}. The striking reduction in hepatic expression of this gene in response to thyroid hormone suggests an additional anti-atherogenic property of T₃.

Conclusions

We have provided the first report of the ability of thyroid hormone to enhance the cholesterol-accepting capacity of serum via the ABCAI transporter. Based on an analysis of serum proteins, is seems that increases in the amounts of ApoA-I in serum from the T₃- treated animals may be responsible for the enhanced cholesterol accepting capacity of the associated lipoproteins. These data suggest that T₃ promotes the first step in reverse cholesterol transport in addition to increasing hepatic LDL receptor expression, bile acid synthesis and transport of cholesterol into the bile by ABCG5 and ABCG8^{2, 3}. We also found that thyroid hormone acts to markedly decrease expression of hepatic PLA2 and to increase serum fetuin A levels. These actions suggest protective anti-inflammatory and anti-calcification roles for the thyroid hormone.