HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 5, 3511-3519, February 1, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/5/3511    most recent M109883200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Furbee, J. W. Articles by Parks, J. S. Search for Related Content PubMed PubMed Citation Articles by Furbee, J. W., Jr. Articles by Parks, J. S. Social Bookmarking                      What's this?

Lecithin:Cholesterol Acyltransferase Deficiency Increases Atherosclerosis in the Low Density Lipoprotein Receptor and Apolipoprotein E Knockout Mice^*

James W. Furbee Jr., Janet K. Sawyer, and John S. Parks

From the Department of Pathology, Section on Comparative Medicine, Wake Forest University School of Medicine, Medical Center Blvd., Winston-Salem, North Carolina 27157-1040

Received for publication, October 12, 2001, and in revised form, November 16, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The purpose of the present study was to test the hypothesis that lecithin:cholesterol acyltransferase (LCAT) deficiency would accelerate atherosclerosis development in low density lipoprotein (LDL) receptor (LDLr/) and apoE (apoE/) knockout mice. After 16 weeks of atherogenic diet (0.1% cholesterol, 10% calories from palm oil) consumption, LDLr/ LCAT/ double knockout mice, compared with LDLr/ mice, had similar plasma concentrations of free (FC), esterified (EC), and apoB lipoprotein cholesterol, increased plasma concentrations of phospholipid and triglyceride, decreased HDL cholesterol, and 2-fold more aortic FC (142 ± 28 versus 61 ± 20 mg/g protein) and EC (102 ± 27 versus 61 ± 27 mg/g). ApoE/ LCAT/ mice fed the atherogenic diet, compared with apoE/ mice, had higher concentrations of plasma FC, EC, apoB lipoprotein cholesterol, and phospholipid, and significantly more aortic FC (149 ± 62 versus 109 ± 33 mg/g) and EC (101 ± 23 versus 69 ± 20 mg/g) than did the apoE/ mice. LCAT deficiency resulted in a 12-fold increase in the ratio of saturated + monounsaturated to polyunsaturated cholesteryl esters in apoB lipoproteins in LDLr/ mice and a 3-fold increase in the apoE/ mice compared with their counterparts with active LCAT. We conclude that LCAT deficiency in LDLr/ and apoE/ mice fed an atherogenic diet resulted in increased aortic cholesterol deposition, likely due to a reduction in plasma HDL, an increased saturation of cholesteryl esters in apoB lipoproteins and, in the apoE/ background, an increased plasma concentration of apoB lipoproteins.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Lecithin:cholesterol acyltransferase (LCAT,¹ EC 2.3.1.43) is a 65-kilodalton glycoprotein that is responsible for the esterification of cholesterol in plasma lipoproteins (1, 2). LCAT catalyzes a two-step reaction that hydrolyzes the sn-2 fatty acid of phospholipid, followed by a transacylation step that transfers the fatty acid to the 3-hydroxyl group of cholesterol, generating cholesteryl ester and lysolecithin. LCAT is activated by apolipoprotein A-I, which is the major apolipoprotein on HDL particles (3). The LCAT enzyme is important in the process of reverse cholesterol transport, which involves the movement of cholesterol from peripheral tissues back to the liver for excretion and is critical for the maintenance of normal concentration, size, and shape of plasma lipoproteins (1, 4).

There are two types of LCAT deficiency in the human population, familial LCAT deficiency (FLD) and fish eye disease (FED). FLD patients are characterized by a complete lack of plasma LCAT activity, corneal opacification, anemia, proteinurea, and kidney dysfunction that can lead to renal failure (5, 6). These patients have decreased plasma total and esterified cholesterol, HDL cholesterol, apoA-I, and apoA-II concentrations and increased plasma concentrations of triglyceride, phospholipid, free cholesterol, and VLDL cholesterol. FED is characterized by a partial LCAT deficiency in which patients have corneal opacification, but no anemia or renal disease. Lipid and lipoprotein abnormalities are milder than those of FLD and include decreased concentrations of total and esterified cholesterol, and HDL cholesterol and increased plasma triglyceride and VLDL cholesterol. In both FLD and FED, plasma HDL particles are abnormal in composition and shape. HDL in FLD are discs or small spherical particles that never mature into the larger HDL₃ and HDL₂ particles because LCAT activity is required for normal HDL maturation. LCAT deficiency may also cause a breakdown in reverse cholesterol transport resulting in the increased cholesterol concentrations observed in peripheral tissues. Most patients with FLD do not show signs of premature coronary artery disease, although the number of patients studied is low. However, premature coronary heart disease has been documented in several male family members with FED (6).

Two LCAT deficient mouse lines have been generated by gene targeting and both show a similar phenotype. LCAT/ mice have no detectable cholesterol esterification in plasma and very low concentrations of total and esterified cholesterol, HDL cholesterol, apoA-I and apoA-II. The small amount of apoA-I in plasma is associated with small HDL particles with pre- mobility on agarose gels (7) and appear discoidal in shape by electron microscopy (8). There is also a decrease in plasma phospholipid, free cholesterol, and non-HDL cholesterol and an increase in triglyceride concentrations. Lipoproteins of abnormal size and shape, including discoidal HDL and LpX-like particles have been observed in the plasma of LCAT/ mice. These mice also develop many of the clinical characteristics of FLD patients, including anemia, proteinuria, and glomerulosclerosis, when fed an atherogenic diet (9). Depletion of cholesterol stores in the adrenal gland has also been observed in LCAT/ mice (8).

The role of LCAT in atherosclerosis development has been controversial. The transgenic overexpression of human LCAT in mice results in no significant difference in atherosclerosis development compared with non-transgenic controls (10) or greater atherosclerosis (11). Transgenic overexpression of human LCAT in rabbits results in decrease atherosclerosis (12). The conflicting results in mice may be due to the impact of LCAT overexpression on HDL concentration and subfraction size. Very high levels of LCAT overexpression (50-100-fold) result in a 2-fold increase in plasma HDL concentrations and the appearance of large HDL₁ particles and ultimately in greater atherosclerotic lesion development (11). However, more modest levels of LCAT overexpression (10-20-fold) result in a smaller increase in plasma HDL cholesterol, no change in HDL particle size, and no effect on atherosclerotic lesion size compared with non-transgenic mice (10). An apparent explanation for the increased atherosclerosis with high levels of LCAT overexpression is that the HDL₁ particles that result are unable to support reverse cholesterol transport.

To further address the question of whether LCAT is pro- or anti-atherogenic in mice, several studies, including the present one, have been designed to determine whether LCAT deficiency will result in more or less atherosclerosis. A recently published study investigated the influence of LCAT deficiency on atherosclerosis development in four different genetic backgrounds of mice (C57Bl/6, LDLr/, apoE/, and cholesteryl ester transfer protein transgenic) (9). The C57Bl/6, LDLr/, and cholesteryl ester transfer protein transgenic mice were fed a cholic acid-containing atherogenic diet, whereas the apoE/ mice consumed a chow diet. LCAT deficiency resulted in a decreased concentration of plasma cholesteryl esters and apoB lipoprotein cholesterol, which should be anti-atherogenic, as well as the expected decrease of plasma HDL, which should be a pro-atherogenic. Despite the lack of plasma HDL, all four genotypes of mice developed less atherosclerosis compared with their respective controls with functioning LCAT.

Our lab has recently generated LCAT/ mice in the LDLr/ and apoE/ backgrounds. However, our results showed an increase in apoB lipoprotein cholesterol in chow-fed LCAT/ mice in both the LDLr/ and apoE/ backgrounds.² We also documented a significant enrichment of plasma apoB lipoproteins with saturated and monounsaturated cholesteryl esters, at the expense of polyunsaturated species, in mice lacking functional LCAT in both genetic backgrounds. We hypothesized that these lipoprotein changes, along with the loss of HDL, would result in more atherosclerosis in the LCAT/ mice. The purpose of the present study was to test this hypothesis using LCAT/ mice in the LDLr/ or apoE/ backgrounds consuming an atherogenic diet without cholic acid. We also measured atherosclerosis as the aortic accumulation of free and esterified cholesterol. Our results support the hypothesis that loss of functioning plasma LCAT leads to greater aortic cholesterol deposition.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Generation of LDLr/ and ApoE/ Mice Lacking Endogenous Mouse LCAT-- The LCAT/ (75% C57Bl/6, 25% 129/Ola) mice were kindly provided by Dr. Omar Francone (Pfizer, Inc.) and apoE/ (100% C57Bl/6) mice were provided by Dr. Nobuyo Maeda (UNC-Chapel Hill). The LDLr/ (100% C57Bl/6) and C57Bl/6J mice were obtained from Jackson Labs (Bar Harbor, ME). The LDLr/ LCAT/ or apoE/ LCAT/ mice were created in a two-step breeding process. In the first step, LDL receptor knockout (i.e. LDLr/ LCAT+/+) mice were crossed with LCAT knockout (LDLr+/+, LCAT/) mice to generate compound heterozygotes (LDLr+/ LCAT+/). In the second step, the F1 generation was then intercrossed to generate LDLr/ LCAT/ mice that were ~90% in the C57Bl/6 background. The apoE/ LCAT/ mice were generated in a similar fashion.

At each step of the breeding protocol, pups were screened for total plasma cholesterol (TPC) and exogenous LCAT activity. PCR of genomic DNA was also used to confirm LDLr/, LCAT/, and apoE/ genotypes. Primers and conditions for PCR have been described previously (14). The PCR conditions for the screening of apoE/ mice were similar and used the following primers: EKO-F, 5'-GTCTCGGCTCTGAACTACATAG-3' and EKO-R, 5'-GCAAGAGGTGATGGTACTCG-3'. The primers generated a 600-bp band for the wild type allele and a 1600-bp band for the targeted allele. PCR products were analyzed on 0.8% TAE-agarose gels.

Experimental Design and Plasma Lipoprotein and Lipid Measurements-- The mice in this study were housed at the Wake Forest University School of Medicine. All protocols and procedures were approved by the Animal Care and Use Committee of the Wake Forest University School of Medicine. At 3 weeks of age the mice were weaned onto a chow diet. The animals were tail bled at 5 weeks of age and the following measurements were made on the isolated plasma: TPC, free cholesterol (FC), phospholipid (PL), triglyceride (TG), and exogenous LCAT activity (14). At 6 weeks of age the animals were switched to an atherogenic diet (0.1% cholesterol, 10% of calories from palm oil) for an additional 16 weeks. At 22 weeks of age, mice were anesthetized with ketamine/HCl/xylazine mixture (1:1 ratio; 120 mg/kg ketamine/HCl, 16 mg/kg xylazine) prior to euthanasia. After anesthesia was assured, a terminal blood sample was taken via cardiac puncture. The animal was then opened via midline laparotomy to expose the thoracic and abdominal cavities. The liver was removed, quick frozen in liquid N₂, and stored at 80 °C. The aorta and heart were removed en bloc up to the aortic iliac bifurcation. The cardiovascular system was immediately placed in 10% neutral buffered formalin for fixation.

All plasma assays performed at 5 weeks of age were repeated at the end of the 16-week atherosclerosis induction phase. In addition, apoB lipoproteins and HDL were isolated from 100 to 150 µl of plasma using Superose 6 (1 × 30 cm) and Superose 12 (1 × 30 cm) FPLC columns in series as described previously (14). Lipoprotein cholesterol distribution of FPLC column fractions was determined using an enzymatic cholesterol assay and aliquots of the apoB lipoproteins and HDL were used to quantify cholesteryl ester and phospholipid fatty acid composition as described previously (14). Plasma apolipoproteins were analyzed by SDS-polyacrylamide gradient gel electrophoresis. Lipoproteins from plasma (100 µl) were isolated by ultracentrifugation at d = 1.21 g/ml using a Beckman TLA 100.2 rotor operated at 1 × 10⁵ rpm for 18 h (15 °C). The top fraction (lipoproteins) was isolated by tube slicing and the samples were extensively dialyzed against 0.05% EDTA and 0.05% sodium azide. Fifteen µg of protein, determined by the Lowry assay (15), were lyophilized and loaded on a 4-16% SDS-polyacrylamide gradient gel in SDS sample buffer (16). The gel was run at 75 V for 30 min, then 150 V for 2 h (10 °C), stained with 0.2% Coomassie Blue in 50% methanol, 10% acetic acid and destained with 50% methanol, 10% acetic acid.

Aorta and Liver Analysis-- Liver samples were thawed, blotted dry, weighed, minced into small pieces, and 20-50-mg aliquots were incubated in 5.0 ml of chloroform:methanol (1:1 v/v) for a minimum of 72 h. 5-Cholestane was added as an internal standard and an aliquot of the lipid extract was dried down, brought up in chloroform, and used for phospholipid and cholesteryl ester fatty acid composition as described previously (14). Another aliquot was used to determine free and total cholesterol concentration in the liver. For free cholesterol analysis, the lipid extract was dried down, brought up in hexane and analyzed using a glc column (J. & W. Scientific DB-17 column) as described previously (17). For total cholesterol analysis, the lipid extract was dried down, brought-up in 1 ml of 100% ethanol and 200 µl of 50% KOH, incubated at 65 °C for 30 min, and then allowed to cool to room temperature. To extract the cholesterol from the sample, 1.0 ml of hexane and 1.0 ml of water were added and the sample was vortexed vigorously and centrifuged at 1500 rpm for 5 min. The upper phase (hexane) was removed, placed in a clean tube, dried down, brought up in a small volume (100-200 µl) of hexane and analyzed as described above for free cholesterol. Esterified cholesterol was calculated as total minus free cholesterol. The lipid-extracted liver was then completely digested with 1.0 M NaOH and protein concentration was assayed using the Lowry method (15).

After at least 72 h fixation in 10% neutral buffered formalin, the aorta was removed from the heart proximal to the aortic arch. The aorta was cleaned of all adventitia and arteries coming off the aorta were removed at their origin. For quantification of aortic FC and EC, the cleaned aorta was incubated in 3.0 ml of chloroform:methanol (1:1 v/v) for a minimum of 72 h. 5-Cholestane was added as an internal standard. Total and free cholesterol were measured by glc as described above for the liver samples and esterified cholesterol was calculated as the difference between total and free cholesterol content. Protein in the lipid-extracted tissue was determined using the Lowry assay.

In a subset of animals, a 3-mm section was taken from the aorta just distal to the origin of the left subclavian artery for histological evaluation of lesion characteristics. Histological sections were stained with Verhoeff-van Gieson's stain.

Data Analysis-- Data are presented as mean ± S.D. The Statview program was used to analyze data using unpaired t tests or ANOVA. When statistical differences were observed by ANOVA, a Fisher's least significant difference test was used for post-hoc analysis to identify the source of the difference.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The current study was undertaken to determine the effect of LCAT deficiency on atherosclerosis development in LDLr/ and apoE/ knockout mice. Table I summarizes the plasma lipid concentrations for all the study animals after consumption of the atherogenic diet (0.1% cholesterol and 10% of calories from palm oil) for 16 weeks. When LDLr/ LCAT/ mice were compared with LDLr/ mice, there was no difference in plasma total cholesterol, free cholesterol, esterified cholesterol, but there was a significant increase in phospholipid (45%) and triglyceride (2.7-fold) concentrations. When apoE/ LCAT/ mice were compared with apoE/ mice, there were significant increases in the plasma concentrations of total cholesterol (46%), free cholesterol (74%), and phospholipid (53%), with no significant changes in esterified cholesterol or triglyceride concentrations. Compared with the apoE/ mice, the LDLr/ mice had significantly higher plasma concentrations of total cholesterol (2.6-fold), free cholesterol (2.1-fold), esterified cholesterol (2.8-fold), phospholipid (2.5-fold), and triglyceride (4.7-fold). When LDLr/ LCAT/ mice were compared with apoE/ LCAT/ mice, there was a significant increase in the plasma concentrations of total cholesterol (66%), free cholesterol (27%), esterified cholesterol (87%), phospholipid (2.4-fold), and triglyceride (4.1-fold). The EC/TC ratio was lower for both LDLr/ and apoE/ mice without LCAT compared with their counterparts with active LCAT.

Plasma exogenous LCAT cholesterol esterification activity was measured using a 1-16:0, 2-18:1 PL rHDL substrate. The esterification values (nanomole of CE formed per h/ml of plasma; n = 5 per group) were similar for LDLr/ (58 ± 9), apoE/ (58 ± 3) and C57Bl/6 controls (63 ± 7), whereas background activity was observed for the LDLr/ LCAT/ (1 ± 0) and apoE/ LCAT/ (2 ± 1) mice.

Fig. 1 shows an FPLC cholesterol elution profile for plasma of each genotype of mouse in the study. LDLr/ LCAT/ mice had less cholesterol in the LDL size range (fractions 35-43) and very little HDL cholesterol (fractions 49-58) compared with the LDLr/ mice. The apoE/ LCAT/ mice had more cholesterol in VLDL (fractions 30-34), a similar amount in LDL, and less in the HDL region compared with the apoE/ mice. The LDLr/ and LDLr/ LCAT/ mice had considerably higher amounts of VLDL and LDL than did the apoE/ or apoE/ LCAT/ mice.

View larger version (24K): [in this window] [in a new window]   Fig. 1.   FPLC cholesterol elution profiles of whole plasma from mice of the indicated genotypes. One-hundred µl of plasma was injected onto a Superose 6 (1 × 30 cm) and a Superose 12 (1 × 30 cm) column connected in series with a flow rate of 0.5 ml/min. One-min fractions were collected and total cholesterol was measured using an enzymatic assay. The inset shows a vertical expansion of the HDL elution region.

The lipoproteins from FPLC separations were pooled into apoB lipoproteins and HDL fractions for further analysis. The distribution of total cholesterol in the lipoprotein fractions and the EC/TC ratio are shown in Table II. Compared with the LDLr/ mice, LDLr/ LCAT/ mice had a significant reduction in plasma HDL cholesterol (28 ± 10 versus 99 ± 7 mg/dl) and in the fraction of cholesterol that was esterified (0.51 ± 0.05 versus 0.86 ± 0.02), whereas the concentration and EC/TC ratio of apoB lipoproteins was similar between the two genotypes of mice. In contrast, the apoE/ LCAT/ mice had significantly higher concentrations of apoB lipoproteins and no difference in HDL concentrations compared with apoE/ mice. However, the ratio of HDL EC/TC was reduced in the apoE/ LCAT/ compared with apoE/ mice. The reason there was measurable HDL EC in animals without functional LCAT likely resulted from the co-elution of some LDL in the HDL region of the FPLC column because of the large disparity in plasma concentration between apoB lipoproteins and HDL in these animals (Fig. 1, inset). Unexpectedly, the concentration of apoB lipoproteins in the LDLr/ mice consuming the atherogenic diet was considerably higher than that of the apoE/ mice.

Table III contains the percentage lipid composition of the plasma apoB lipoproteins isolated by FPLC. The absence of plasma LCAT resulted in an increase in plasma apoB lipoprotein PL and TG for both LDLr/ and apoE/ groups. The concentration of apoB lipoprotein FC and CE was similar in LDLr/ and LDLr/ LCAT/ mice, whereas the concentration of these lipid constituents was 2-fold greater in apoE/ LCAT/ mice compared with apoE/ animals.

Fig. 2 shows the SDS-PAGE separation of d < 1.21 g/ml lipoproteins for the four genotypes of mice consuming the atherogenic diet. Lipoproteins from the LDLr/ mice contained predominantly apoB100, apoE, and apoA-I, with a small amount of apoB48 (lane 1). The apolipoprotein profile was similar in LDLr/ LCAT/ mice, except for the near absence of apoA-I (lane 2). Lipoproteins from the apoE/ mice had predominantly apoA-I and apoB48_, with smaller amounts of apoB100 and an unidentified apolipoprotein, presumably apoA-IV, which appeared in the 43-67-kDa range (lane 3). The apolipoprotein pattern was similar for the apoE/ LCAT/ mice, except for a relative reduction in the proportion of apoA-I in plasma (lane 4).

View larger version (32K): [in this window] [in a new window]   Fig. 2.   4-16% SDS-polyacrylamide gel separation of plasma lipoproteins floated at a density = 1.21 g/ml in the ultracentrifuge. Plasma was isolated from mice of the indicated genotypes (top of gel) after 16 weeks of atherogenic diet consumption and plasma lipoproteins were isolated as indicated under "Experimental Procedures." Fifteen µg of protein were applied per lane. Molecular weights for a set of low molecular weight standards (Pharmacia Corp.) is shown on the left side of the gel. Estimated migration position of apolipoproteins B100, B48, E, and A-I is indicated on the right side of the gel.

Previous studies have shown a strong correlation between LDL CE fatty acid composition and the extent of atherosclerosis (18-20). Thus, the plasma apoB lipoproteins were isolated by FPLC, the lipids were extracted, and the CE fatty acid distribution was quantified. Fig. 3 shows the relative amounts of saturated, monounsaturated, and polyunsaturated CE species in the plasma apoB lipoproteins of the study mice. Compared with the LDLr/ mice, the LDLr/ LCAT/ mice had an increase in the proportion of monounsaturated CE (81.7 ± 3.9 versus 56.3 ± 1.7%), a decrease in polyunsaturated CE (3.2 ± 0.4 versus 28.8 ± 4.3%), and a similar proportion of saturated CE species (15.1 ± 4.0 versus 13.7 ± 2.6%). A similar, but less striking trend was observed in the apoE knockout background. Compared with the apoE/ mice, the apoE/ LCAT/ mice had increased monounsaturated CE (64.0 ± 3.9 versus 55.9 ± 5.5%), a decrease in polyunsaturated CE species (5.6 ± 1.5 versus 17.5 ± 10.8%), and similar saturated CE distribution (30.4 ± 3.4 versus 26.6 ± 6.6%). The differences in CE fatty acid distribution can be simplified by expressing the data as a ratio of saturated + monounsaturated to polyunsaturated CE species (CEFA ratio) (21). Loss of functional LCAT in these mice resulted in a 12-fold increase in the CEFA ratio of the LDL/ LCAT/ mice compared with LDLr/ mice and a 3-fold increase in the apoE/ LCAT/ mice compared with their apoE/ counterparts. Thus, LCAT deficiency in these two mouse models of atherosclerosis resulted in a significant increase in the saturation of CEs in plasma apoB lipoproteins.

View larger version (22K): [in this window] [in a new window]   Fig. 3.   Percentage distribution of saturated (Sat'd), monounsaturated (Mono), and polyunsaturated (Poly) cholesteryl ester species in plasma apoB lipoproteins. Plasma was obtained from mice of the indicated genotype after 16 weeks of atherogenic diet consumption and apoB lipoproteins were isolated by FPLC and CE fatty acids were quantified as detailed previously (14). Bars represent the mean ± S.D. (n = 6-8 mice per genotype).

To determine the role of LCAT on atherosclerosis development, the aortic cholesterol content was quantified for each group of animals. LDLr/ LCAT/ mice, compared with LDLr/ mice, had significantly more TC (244 ± 52 versus 122 ± 46 mg of cholesterol/g protein, p < 0.0001, Fig. 4A), EC (102 ± 27 versus 61 ± 27 mg of cholesterol/g protein, p = 0.0003, Fig. 4B), and FC (142 ± 28 versus 61 ± 20 mg of cholesterol/g protein, p < 0.0001, Fig. 4C). Similar results were observed when apoE/ LCAT/ mice were compared with apoE/ mice, with significantly higher concentrations of aortic TC (249 ± 79 versus 178 ± 52 mg of cholesterol/g wet weight, p = 0.0076), EC (101 ± 23 versus 69 ± 20 mg of cholesterol/g protein, p < 0.0047), and FC (149 ± 62 versus 109 ± 33 mg of cholesterol/g wet weight, p = 0.025) in the double knockout mice. For comparative purposes, C57Bl/6 mice (n = 21) consuming a 15% palm oil, 1.0% cholesterol, and 0.5% cholic acid diet for 24 weeks accumulated 1.2 ± 1.0 mg of EC/g of aortic protein and 14.4 ± 6.8 mg of FC/g of aortic protein. Loss of functional LCAT resulted in a similar amount of aortic lipid accumulation in both atherosclerosis susceptible backgrounds, as there was no statistically significant difference for aortic TC, EC, or FC between LDLr/ LCAT/ and apoE/ LCAT/ mice. However, the LDLr/ mice did have significantly less aortic TC (p = 0.032) and FC (p = 0.0076) compared with the apoE/ mice, despite the fact that the LDLr/ mice had higher plasma concentrations of apoB lipoproteins. The proportion of EC and FC that accumulated in the aortas was nearly equivalent among all groups, ranging from 0.39 ± 0.03 (apoE/) to 0.49 ± 0.06 (LDLr/).

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Aortic total (top panel), esterified (middle panel), and free (bottom panel) cholesterol content in mice of the indicated genotypes after 16 weeks of atherogenic diet consumption. Aortic total and free cholesterol content was determined by gas-liquid chromatography as described under "Experimental Procedures"; esterified cholesterol content was calculated as total minus free. Symbols represent data from individual animals and the horizontal line for each genotype represents the group mean. Data are expressed as milligrams of cholesterol/g of aortic protein.

Verhoeff-van Gieson-stained cross-sections of aortas from the four genotypes of mice are shown in Fig. 5. The sections are representative of the extent and severity of atherosclerosis among the four experimental groups and were chosen to illustrate the characteristics of the atherosclerotic lesions. Atherosclerotic lesions were found to be non-circumferential in all groups, with mostly small foam cells. No fibrous plaques or areas of necrosis were observed in the aortas but extensive areas of free cholesterol were observed in the aortas from LCAT-deficient mice, especially in the apoE/ LCAT/ aorta.

View larger version (67K): [in this window] [in a new window]   Fig. 5.   Verhoeff-van Gieson-stained cross-sections of aortas from mice of the indicated genotype after 16 weeks of atherogenic diet consumption. A 3-mm section was taken from the aorta just distal to the origin of the left subclavian artery for histological evaluation. Sections are representative of the extent and severity of atherosclerosis among the four experimental groups and were chosen to illustrate the character of the atherosclerotic lesions. Lesions were non-circumferential for all groups and show extensive areas of free cholesterol deposition in the LCAT deficient mice, particularly the apoE/ LCAT/ mouse.

Previous studies in non-human primates have documented a strong correlation between hepatic CE content and hepatic CE secretion rate (22) and LDL particle size (23), both of which are highly correlated with coronary artery atherosclerosis (22, 24, 25). To determine whether LCAT deficiency, which resulted in increased aortic cholesterol deposition, also resulted in increased hepatic cholesterol deposition, we quantified EC and FC in the livers of the study animals (Fig. 6). The LDLr/ LCAT/ mice had reduced levels of hepatic EC (48.5 ± 14.2 versus 78.0 ± 32.2 mg/g protein, p = 0.036), but similar concentrations of TC (75.9 ± 17.3 versus 99.4 ± 38.2, p = 0.065) and FC (27.4 ± 6.5 versus 21.4 ± 6.3) compared with the LDLr/ mice. The apoE/ LCAT/ mice and apoE/ mice had similar concentrations (mg/g protein) of hepatic TC (52.7 ± 15.3 versus 44.2 ± 15.1, respectively), EC (39.8 ± 13.0 versus 31.8 ± 16.0), and FC (12.9 ± 2.4 versus 12.5 ± 1.9). The LDLr/ mice, compared with the apoE/ mice, had significantly greater concentrations of hepatic TC (99 versus 44 mg/g protein), EC (78 versus 32 mg/g protein), and FC (21 versus 13 mg/g protein).

View larger version (20K): [in this window] [in a new window]   Fig. 6.   Liver esterified and free cholesterol content in mice of the indicated genotypes after 16 weeks of atherogenic diet consumption. Total and free cholesterol was measured by gas-liquid chromatography of aortic lipid extracts as described under "Experimental Procedures." Cholesterol was calculated as total-free. Values are the mean ± S.D. (n = 5 mice per group) and are expressed as milligrams of cholesterol/g of liver protein.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The purpose of this study was to determine the effect of LCAT deficiency on atherosclerosis development. Previous studies using LCAT-deficient mice and LCAT transgenic mice that overexpress human LCAT have yielded mixed results as regards the pro- or anti-atherogenic potential of LCAT. Studies performed by Lambert et al. (9) showed that LCAT/ mice in the C57Bl/6 or LDLr/ background fed a cholic acid-containing atherogenic diet had smaller atherosclerotic lesions compared with their counterparts with active LCAT. Similar results were observed for apoE/ LCAT/ compared with apoE/ mice fed a chow diet. Studies by Mehlum et al. (10, 26) have shown that moderate levels of LCAT overexpression (10-40-fold) in transgenic mice show no difference in atherosclerosis development compared with non-transgenic controls. Other studies by Berard et al. (11) have shown that high levels of LCAT overexpression (100-250-fold) results in the generation of very large HDL₁ particles and an increase in atherosclerosis compared with C57Bl/6J controls. In our study, we examined the effect of LCAT deficiency on atherosclerosis development in LDLr/ and apoE/ mice using an atherogenic diet that did not contain cholic acid and by quantifying aortic free and esterified cholesterol. We found that LCAT-deficient mice had increased aortic total, free and esterified cholesterol accumulation in both the LDLr/ and apoE/ atherosclerosis susceptible backgrounds, compared with LDLr/ and apoE/ mice with functioning LCAT. Although the outcome of LCAT deficiency on atherosclerosis development was similar in both LDLr/ and apoE/ mice, the mechanisms of the increased aortic cholesterol accumulation likely were different for each background. The LDLr/ LCAT/ mice had increased aortic lipid deposition despite having no difference in the concentration of apoB lipoproteins in plasma. However, LDLr/ LCAT/ mice did have reduced HDL cholesterol concentrations and an increase in the saturation of plasma apoB lipoprotein CEs that likely resulted in increased atherosclerosis compared with LDLr/ mice. The apoE/ LCAT/ mice also had decreased HDL concentrations and an increase in the saturation of apoB lipoprotein CEs compared with apoE/ mice, but in addition, had an increased concentration of apoB lipoproteins in plasma. Thus, we conclude that LCAT deficiency in LDLr/ and apoE/ mice fed an atherogenic diet resulted in increased aortic cholesterol deposition, likely due to a severe reduction in plasma HDL, an increased saturation of cholesteryl esters in apoB lipoproteins and, in the apoE/ background, an increased plasma concentration of apoB lipoproteins.

Our main finding that LCAT deficiency in mice resulted in increased aortic atherosclerosis is opposite to that observed in the study by Lambert et al. (9). There are several likely explanations for disparity between the two studies. First, different diets were fed in the two studies and this led to different plasma lipid responses. Lambert et al. (9) fed a 0.5% cholic acid, 1.25% cholesterol, 10% fat-containing diet to the C57Bl/6 and LDLr/ mice or a chow diet to the apoE/ mice. The LCAT/ mice responded with decreased concentrations of TC, CE, and non-HDL cholesterol, compared with control mice with functioning LCAT, in all three backgrounds. The decrease in lipid concentrations, particularly apoB lipoprotein cholesterol, was hypothesized to be the cause of decreased lesion area in the LCAT-deficient mice (9). The animals in our study consumed a diet with no cholic acid and a moderate amount of cholesterol (0.1%) and dietary fat (10% wt). Using this dietary regimen, LDLr/ and LDLr/ LCAT/ mice had similar TPC, EC, and apoB lipoprotein cholesterol concentrations and there was actually an increase in total plasma and apoB lipoprotein cholesterol in apoE/ LCAT/ mice compared with apoE/ mice (Tables I and II). Thus, the most likely explanation for the disparate results is the difference in plasma lipid response to the atherogenic diet between the two studies, with decreased apoB lipoproteins in LCAT-deficient mice in the Lambert study and increased or unchanged apoB lipoprotein concentrations in our study. The method for atherosclerosis quantification was also different between the two studies. Lambert et al. (9) used a histological technique of measuring lesion area in the aortic root that examines a well defined, but small area of the aorta very near the heart, whereas we used a direct measurement of free and esterified cholesterol accumulation in the whole aorta. However, it is unlikely that the differences in methodology could explain the divergent experimental outcomes, since the correlation between aortic root intimal area and extent of aortic atherosclerosis, measured as percentage surface stained with Sudan IV, is high (r = 0.77) (27) and the association of percentage of surface with lesion involvement and aortic cholesterol content is also high (r = 0.85) (28). Thus, whether LCAT is pro-atherogenic or anti-atherogenic may well depend on other environmental and genetic variables that affect plasma lipoproteins.

A trivial explanation for the discrepancy in atherosclerosis results between our study and those of Lambert et al. (9) pertains to genetic background of the mouse strains. All mice in the study by Lambert et al. (9) had been back-crossed at least eight generations with C57Bl/6 mice, a strain that is susceptible to atherosclerosis when fed a cholic acid-containing diet. In our study the LDLr/ and apoE/ were in a pure C57Bl/6 background, whereas the LDLr/ LCAT/ and apoE/ LCAT/ mice were ~90% in the C57Bl/6 background. There are several reasons why we believe this does not explain the differences in atherosclerosis development. First, the 10% contaminating background of our double knockout mice came from the 129 background, a strain that is resistant to atherosclerosis when challenged with a cholic acid-containing diet (29). Thus, the contaminating 129 background should decrease, not increase atherosclerosis. Furthermore, the atherosclerosis response in both the LDLr/ and apoE/ mice was similar in direction (i.e. increased with LCAT deficiency), magnitude, and variability (Fig. 4). In fact, the coefficients of variation of measurements of aortic TC, FC, and EC were similar among all four groups of mice, ranging from 20 to 44%, with no consistent pattern for animals with or without LCAT. Random genomic fragments introduced from another strain would be expected to increase the variability in atherosclerosis compared with a pure genetic strain (30), but this was not the case in our study. This would argue against a random chromosomal fragment from the 129 background that caused the observed increase in atherosclerosis, since the LDLr/ LCAT/ and apoE/ LCAT/ mice would have different random genomic fragments from the two 129 embryonic stem cell lines (E14TG2a and AB-1) used to develop the LDLr/ (31) and apoE/ (32) mice, respectively. Finally, the genetic susceptibility of the C57Bl/6 mice is defined by the amount of lesion area in the aortic root of animals fed a cholic acid-containing diet and cannot be extrapolated to non-cholic acid-containing diets, which do not result in atherosclerosis in this strain of mice (33).

One potential mechanism for increased atherosclerosis in LCAT-deficient mice is decreased plasma HDL concentrations, which could result in defective or retarded reverse cholesterol transport. While there are consistent results in the literature demonstrating that transgenic overexpression of apoA-I results in less atherosclerosis (34-36), the atherosclerosis results in apoA-I/ mice with decreased plasma apoA-I concentrations are variable. Li et al. (37) first documented that apoA-I/ mice were not at increased risk of developing atherosclerosis compare with wild type controls. Atherosclerosis extent was also similar between apoE/ and apoE/ apoA-I/ mice, although the latter had significantly reduced concentrations of TPC and HDL cholesterol (38). However, Boisvert et al. (39) demonstrated greater aortic FC accumulation in apoA-I/ mice compared with wild type mice, both of which were in an apoE/ background with apoE-expressing macrophages derived from bone marrow transplantation. Other studies using human apoB100 transgenic mice without apoA-I have shown significantly more atherosclerosis development compared with their counterparts expressing apoA-I (40, 41); however, atherosclerosis development was minimal and similar between control and apoA-I/ not expressing the human apoB100 transgene (41). Other studies that use genetic manipulation to lower HDL concentrations, such as overexpression of scavenger receptor BI, have also yielded inconsistent results. For instance, transient overexpression of scavenger receptor BI with adenoviral gene transfer resulted in reduced HDL concentrations and atherosclerosis in LDLr/ mice (42), whereas transgenic overexpression of scavenger receptor BI at low or high levels resulted in striking decreases in plasma HDL concentrations, but only in the former case was atherosclerosis extent reduced (43). Arai et al. (44) also demonstrated that while transgenic overexpression of scavenger receptor BI uniformly resulted in low HDL concentrations, atherosclerosis outcome depended on the composition of the experimental diet and gene dosage of LDLr. Thus, genetic manipulations that lower HDL concentrations do not necessarily result in increased atherosclerosis, but rather depend on their overall impact on plasma lipoprotein metabolism. Given the conflicting nature of the published results on genetic models that result in decreased HDL concentrations and the relatively low proportion of plasma cholesterol distributed in the HDL fraction in our study animals, it seems unlikely that the decreased atherosclerosis observed with LCAT deficiency can be explained completely by the decrease in plasma HDL concentrations.

Another possible explanation for increased atherosclerosis in LCAT-deficient mice could be the increased fatty acyl saturation of the apoB lipoprotein CEs, which has been indirectly linked to atherosclerosis susceptibility in humans (45, 46) and animal models (17, 20, 47-49). Consumption of diets enriched in polyunsaturated fatty acids results in increased polyunsaturated fatty acids in LDL CEs, in LDL particles with lower CE melting temperatures, and in decreased atherosclerosis (18-20). In the current study LCAT deficiency resulted in a striking reduction of polyunsaturated fatty acids in the CE fraction of apoB lipoproteins, from 29 to 3% in the LDLr/ background and from 18 to 6% in the apoE/ background (Fig. 3). The mechanism by which an increased polyunsaturated fatty acyl CE content of apoB lipoproteins would result in decreased atherosclerosis is poorly understood. Polyunsaturated fat-enriched apoB lipoproteins may bind less avidly to extracellular matrix components of the artery wall resulting in less cholesterol ester accumulation (50). Polyunsaturated CE within cells of the artery wall may efflux more rapidly due to a lower melting temperature compared with more saturated CE species (51, 52). LDL particles isolated from animals fed polyunsaturated fat are smaller on average, contain fewer CE molecules per particle, contain less apoE, and bind with lower affinity to cells in culture (24, 53). One or more on these changes may decrease the accumulation of LDL particles in the artery wall, resulting in less cholesterol accumulation for animals fed polyunsaturated compared with saturated fat. LCAT may function to buffer the intracellular acyl-CoA:cholesterol acyltransferase generated saturated and monounsaturated CE in plasma LDL through the generation of polyunsaturated CE species. Thus, factors that alter the polyunsaturated CE content of LDL, whether induced by diet or genetic manipulation, may affect the structure and metabolism of LDL and affect atherosclerosis outcome.

In our study, apoE/ mice had more aortic FC than LDLr/ mice (Fig. 4), despite the latter having plasma apoB lipoprotein cholesterol concentrations that were 3-fold higher (Table II). One possible explanation for this observation could be the 3-fold higher concentration of plasma HDL in the LDLr/ mice compared with apoE/ mice (Table II), which may have stimulated reverse cholesterol transport in the LDLr/ mice and prevented further exacerbation of atherosclerosis. Another possibility is the higher saturated and lower polyunsaturated CE content of apoB lipoproteins in the apoE/ mice (Fig. 3) resulted in more aortic FC accumulation compared with LDLr/ mice, despite the lower concentration of apoB lipoprotein cholesterol concentrations in the apoE/ mice. Finally, apoE may have an atheroprotective effect that is independent of its role in the clearance of plasma lipoproteins. Several studies have shown that replacement of macrophage apoE via bone marrow transplantation in apoE/ mice retarded atherosclerosis development (54-56). Conversely, C57Bl/6 mice transplanted with bone marrow from apoE/ mice developed more atherosclerosis compared with control mice (57). More recently, it has been shown that transgenic expression of small amounts of apoE in apoE/ mice does not correct the defective clearance of plasma lipoproteins, but is sufficient to reduce atherosclerosis (58, 59). It is likely that all of these differences in the apoE/ mice result in aortic FC accumulation that is greater than that for LDLr/ mice despite differences in apoB lipoprotein concentration that would predict less FC accumulation.

The apoE/ LCAT/ had significantly higher plasma concentrations of apoB lipoproteins compared with apoE/ mice (Table I). The increase was due entirely to lipoproteins eluting in the void volume region of the FPLC column (Fig. 1). It is reasonable to conclude that this was the result of accumulation of LpX-like particles in plasma that have been described in LCAT-deficient mice fed a chow diet (9) and in familial LCAT-deficient patients (5). However, the compositional data in Table III argue against this interpretation because there is no observable increase in the proportion of FC and PL in the plasma apoB lipoproteins. Our previous studies in non-human primates have shown a strong correlation between hepatic CE content and the concentration of VLDL CE in recirculating liver perfusate, suggesting that hepatic CE storage pools are coupled to VLDL CE secretion (22, 23). However, the amount of hepatic CE was similar for both apoE/ and apoE/ LCAT/ mice (Fig. 6), suggesting that VLDL CE production was not likely different between the two stains of mice and that the accumulation of VLDL size lipoproteins in the plasma of apoE/ LCAT/ mice might result from an impaired clearance of these particles. LCAT has been implicated in the clearance of apoB lipoprotein particles via the LDLr pathway; rabbits with transgenic overexpression of human LCAT demonstrated a more rapid decay of plasma LDL compared with non-transgenic controls, but only in animals expressing active LDLr (13). Further studies are necessary to define the role of LCAT in the clearance of the apoB particles in the apoE/ mice.

In conclusion, we have described an increase in atherosclerosis in LCAT-deficient mice in two different atherosclerosis susceptible backgrounds. To our knowledge, this is the first report that directly supports an anti-atherogenic role for LCAT in mice, although the concept that LCAT plays a key role in reverse cholesterol transport and may be anti-atherogenic was proposed many years ago (1) and transgenic overexpression of human LCAT in rabbits is anti-atherogenic (12). The mechanism by which LCAT may perform its function in preventing arterial cholesterol accumulation appears multifactorial and likely involves its role in maintaining plasma HDL concentrations, its contribution to the polyunsaturated CE pool of apoB lipoproteins, and, in the case of apoE/ mice, its role in the metabolism of plasma apoB lipoproteins.

	ACKNOWLEDGEMENTS

We gratefully acknowledge the technical assistance of Abraham Gebre and Ellen Burleson.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grants HL-54176 (to J. S. P.) and HL-49373 (to J. S. P.) and a National Research Service Award Institutional Grant HL-07115 (to J. W. F.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: Dept. of Pathology, Section on Comparative Medicine, Wake Forest University School of Medicine, Medical Center Blvd., Winston-Salem, NC 27157-1040. Tel.: 336-716-2145; Fax: 336-716-6279; E-mail: jparks@wfubmc.edu.

Published, JBC Papers in Press, November 21, 2001, DOI 10.1074/jbc.M109883200

² Furbee, J. W., Jr., Francone, O., and Parks, J. S. (2002) J. Lipid Res. 43, in press.

	ABBREVIATIONS

The abbreviations used are: LCAT, lecithin:cholesterol acyltransferase; HDL, high density lipoprotein; LDL, low density lipoprotein; apoA-I, apolipoprotein A-I; apoB, apolipoprotein B; apoE, apolipoprotein E; LDLr, low density lipoprotein receptor; TPC, total plasma cholesterol; FC, free cholesterol; EC, esterified cholesterol; CE, cholesteryl ester; PL, phospholipid; TG, triglyceride; FPLC, fast protein liquid chromatography; FLD, familial LCAT deficiency; FED, fish eye disease.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				D. J. Rader, E. T. Alexander, G. L. Weibel, J. Billheimer, and G. H. Rothblat The role of reverse cholesterol transport in animals and humans and relationship to atherosclerosis J. Lipid Res., April 1, 2009; 50(Supplement): S189 - S194. [Abstract] [Full Text] [PDF]

				F. Domoki, B. Kis, T. Gaspar, J. A. Snipes, J. S. Parks, F. Bari, and D. W. Busija Rosuvastatin induces delayed preconditioning against oxygen-glucose deprivation in cultured cortical neurons Am J Physiol Cell Physiol, January 1, 2009; 296(1): C97 - C105. [Abstract] [Full Text] [PDF]

				X. Zhu, J.-Y. Lee, J. M. Timmins, J. M. Brown, E. Boudyguina, A. Mulya, A. K. Gebre, M. C. Willingham, E. M. Hiltbold, N. Mishra, et al. Increased Cellular Free Cholesterol in Macrophage-specific Abca1 Knock-out Mice Enhances Pro-inflammatory Response of Macrophages J. Biol. Chem., August 22, 2008; 283(34): 22930 - 22941. [Abstract] [Full Text] [PDF]

				R. E. Temel, L. Hou, L. L. Rudel, and G. S. Shelness ACAT2 stimulates cholesteryl ester secretion in apoB-containing lipoproteins J. Lipid Res., July 1, 2007; 48(7): 1618 - 1627. [Abstract] [Full Text] [PDF]

				M. Zabalawi, M. Bharadwaj, H. Horton, M. Cline, M. Willingham, M. J. Thomas, and M. G. Sorci-Thomas Inflammation and skin cholesterol in LDLr-/-, apoA-I-/- mice: link between cholesterol homeostasis and self-tolerance? J. Lipid Res., January 1, 2007; 48(1): 52 - 65. [Abstract] [Full Text] [PDF]

				M. Cuchel and D. J. Rader Macrophage Reverse Cholesterol Transport: Key to the Regression of Atherosclerosis? Circulation, May 30, 2006; 113(21): 2548 - 2555. [Full Text] [PDF]

				H. Mezdour, G. Larigauderie, G. Castro, G. Torpier, J. Fruchart, M. Nowak, J.-C. Fruchart, M. Rouis, and N. Maeda Characterization of a new mouse model for human apolipoprotein A-I/C-III/A-IV deficiency J. Lipid Res., May 1, 2006; 47(5): 912 - 920. [Abstract] [Full Text] [PDF]

				R. G. Lee, R. Shah, J. K. Sawyer, R. L. Hamilton, J. S. Parks, and L. L. Rudel ACAT2 contributes cholesteryl esters to newly secreted VLDL, whereas LCAT adds cholesteryl ester to LDL in mice J. Lipid Res., June 1, 2005; 46(6): 1205 - 1212. [Abstract] [Full Text] [PDF]

				R. G. Lee, K. L. Kelley, J. K. Sawyer, R. V. Farese Jr, J. S. Parks, and L. L. Rudel Plasma Cholesteryl Esters Provided by Lecithin:Cholesterol Acyltransferase and Acyl-Coenzyme A:Cholesterol Acyltransferase 2 Have Opposite Atherosclerotic Potential Circ. Res., November 12, 2004; 95(10): 998 - 1004. [Abstract] [Full Text] [PDF]

				M. Zabalawi, S. Bhat, T. Loughlin, M. J. Thomas, E. Alexander, M. Cline, B. Bullock, M. Willingham, and M. G. Sorci-Thomas Induction of Fatal Inflammation in LDL Receptor and ApoA-I Double-Knockout Mice Fed Dietary Fat and Cholesterol Am. J. Pathol., September 1, 2003; 163(3): 1201 - 1213. [Abstract] [Full Text] [PDF]

				L. L. Rudel, M. Davis, J. Sawyer, R. Shah, and J. Wallace Primates Highly Responsive to Dietary Cholesterol Up-regulate Hepatic ACAT2, and Less Responsive Primates Do Not J. Biol. Chem., August 23, 2002; 277(35): 31401 - 31406. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 277/5/3511    most recent M109883200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Furbee, J. W. Articles by Parks, J. S. Search for Related Content PubMed PubMed Citation Articles by Furbee, J. W., Jr. Articles by Parks, J. S. Social Bookmarking                      What's this?