Increased plasma non-esterified fatty acids and ... - Clinical Science

Clinical Science (2004) 106, 421–432 (Printed in Great Britain)

Increased plasma non-esterified fatty acids and platelet-activating factor acetylhydrolase are associated with susceptibility to atherosclerosis in mice Uma SINGH∗ , Shumei ZHONG∗ , Momiao XIONG†‡, Tong-bin LI∗ , Allan SNIDERMAN§ and Ba-Bie TENG∗ ‡ ∗

Research Center for Human Genetics, Institute of Molecular Medicine, The University of Texas Health Science Center at Houston, 2121 W. Holcombe Blvd, Houston, TX 77030, U.S.A., †Human Genetics Center, School of Public Health, University of Texas Health Science Center at Houston, 2121 W. Holcombe Blvd, Houston, TX 77030, U.S.A., ‡University of Texas Graduate School of Biomedical Sciences at Houston, Houston, TX 77030, U.S.A., and §Mike Rosenbloom Laboratory for Cardiovascular Research, McGill University, Montreal, Quebec, Canada H3A 1A1

A

B

S

T

R

A

C

T

Animal models provide vital tools to explicate the pathogenesis of atherosclerosis. Accordingly, we established two atherosclerosis-prone mice models: (i) mice lacking the LDL (low-density lipoprotein) receptor (LDLR) and the ability to edit apo (apolipoprotein) B mRNA (Apobec1; designated LDb: LDLR−/− Apobec1−/− ), and (ii) mice with the LDb background, who also overexpressed human apoB100 (designated LTp: LDLR−/− Apobec1−/− ERhB+/+ ). Both LDb and LTp mice had markedly elevated levels of LDL and increased levels of NEFAs (non-esterified fatty acids) compared with C57BL/6 wild-type mice. However, fasting glucose and insulin levels in both animals were not different than those in C57BL/6 wild-type mice. It has been suggested that PAF-AH (platelet-activating factor acetylhydrolase) increases susceptibility to vascular disease. Both LDb and LTp mice had significantly higher PAF-AH mRNA levels compared with C57BL/6 wild-type mice. PAF-AH gene expression was also significantly influenced by age and sex. Interestingly, PAF-AH mRNA levels were significantly higher in both LTp male and female mice than in the LDb mice. This increased PAF-AH gene expression was associated with elevated plasma PAF-AH enzyme activities (LTp > LDb > C57BL/6). Moreover, a greater proportion of PAF-AH activity was associated with the apoB-containing lipoproteins: 29 % in LTp and 13 % in LDb mice compared with C57BL/6 wild-type animals (6.7 %). This may explain why LTp mice developed more atherosclerotic lesions than LDb mice by 8 months of age. In summary, increased plasma NEFAs, PAF-AH mRNA and enzyme activities are associated with accelerated atherogenesis in these animal models.

Key words: atherosclerosis, atherosclerotic-prone mice, apolipoprotein B mRNA editing enzyme (Apobec1), low-density lipoprotein (LDL) receptor, non-esterified fatty acid (NEFA), platelet-activating factor acetylhydrolase (PAF-AH). Abbreviations: apo, apolipoprotein; Apobec1, apoB mRNA editing enzyme; ECL, enhanced chemiluminescence; ERhB, transgenic mice expressing human apoB; HDL, high-density lipoprotein; IDL, intermediate-density lipoprotein; LDL, low-density lipoprotein; LDLR, LDL receptor; LDb, LDLR−/− Apobec1−/− mice; LPC, lysophosphatidylcholine; LTp, LDLR−/− Apobec1−/− ERhB+/+ mice; NEFA, non-esterified fatty acid; NIH, National Institutes of Health; PAF-AH, platelet-activating factor acetylhydrolase; RT, reverse transcriptase; VLDL, very-LDL. Correspondence: Dr Ba-Bie Teng, Research Center for Human Genetics, Institute of Molecular Medicine, The University of Texas Health Science Center at Houston, 2121 W. Holcombe Blvd, Houston, TX 77030, U.S.A. (e-mail [email protected]).

C

2004 The Biochemical Society

421

422

U. Singh and others

INTRODUCTION

MATERIALS AND METHODS

It is well recognized that the plasma apo (apolipoprotein) B level is a strong predictor of atherosclerosis and coronary artery disease [1,2], and the commonest cause of an increased plasma apoB is over-secretion of apoBcontaining lipoprotein particles by the liver [3]. Moreover, over-secretion of apoB in humans has also been associated with insulin resistance [4,5]. Therefore extensive clinical and molecular investigations have been carried out to understand the factors that regulate hepatic apoB secretion. Genetically engineered mice have become important tools to dissect the relationship of apoB with atherosclerosis. Sanan et al. [6] generated mice expressing human apoB100 on the background of deficiency in LDL (low-density lipoprotein) receptor (LDLR). These mice have markedly elevated levels of LDL and reduced levels of HDL (high-density lipoprotein) and develop complex atherosclerotic lesions on a chow diet. The mouse liver, unlike the human liver, expresses Apobec1 (apoB mRNA editing enzyme) [7] and produces apoB48-containing lipoproteins. This difference is recognized as one of the main reasons that wild-type mice are resistant to atherosclerosis. Recently, Powell-Braxton et al. [8] crossed mice lacking Apobec1 (Apobec1−/− ) with mice deficient in LDLR (LDLR−/− ). This mouse model (LDLR−/− Apobec1−/− ) secreted mouse apoB100containing lipoproteins only, had markedly elevated LDL-cholesterol levels and developed extensive atherosclerosis on regular chow diet by 8 months of age. Both animal models develop atherosclerosis on a chow diet; however, mice expressing human apoB100 had much higher levels of plasma triacylglycerol (triglyceride) and lower levels of HDL compared with the mice expressing mouse apoB100. In humans, over-secretion of apoB is associated with hypertriglyceridaemia and insulin resistance. We sought to generate a mouse model that overexpresses human and mouse apoB100 to compare with mice which express mouse apoB100 only (LDLR−/− Apobec1−/− ; designated LDb) to determine whether these animals were insulin-resistant. To do so, it was necessary to engineer a rare triple genetically modified mouse line that overexpresses human apoB on the background of deficiency in both LDLR and Apobec1 genes (LDLR−/− Apobec1−/− ERhB+/+ ; designated LTp). In the present study, we show that these animals are not insulin-resistant on a chow diet. However, they have markedly elevated levels of plasma NEFAs (nonesterified fatty acids) and increased gene expression of PAF-AH (platelet-activating factor acetylhydrolase) and plasma PAF-AH activity. Of importance, both NEFAs and PAF-AH were positively associated with the development of atherosclerosis.

Generation of LDb and LTp mice

C


To generate the LDb mice (LDLR−/− Apobec1−/− ), Apobec1−/− mice (kindly provided by Dr Lawrence Chan, Department of Medicine, Baylor College of Medicine, Houston, TX, U.S.A.) [9] were backcrossed with C57BL/6 to five generations of backcross mating (96.88 % of the C57BL/6 genome). The progeny were then cross-bred with LDLR−/− mice (The Jackson Laboratory, Bar Harbor, ME, U.S.A.), yielding a double-mutant mouse line. To generate the LTp mice (LDLR−/− Apobec1−/− ERhB+/+ ), Apobec1−/− ERhB+/+ mice [10] were backcrossed with C57BL/6 to ten generations of backcross mating and cross-bred with LDLR−/− mice, yielding a triple genetically modified mouse line. The genotypes of the mouse lines were confirmed by PCR and Southern-blot analysis using mouse tail genomic DNA as the template [9,11].

Animal experiments All animal experiments were conducted in accordance with the Guidelines of the Animal Protocol Review Committee of the University of Texas Health Science Center at Houston, Houston, TX, U.S.A. Male and female LDb and LTp mice were kept on a regular chow diet (4 % mouse/rat diet 7001; Harlan Teklad, Indianapolis, IN, U.S.A.) for the duration of the study. At the end of the indicated time, the mice were fasted for approx. 10 h, weighed, anaesthetized using 0.02 ml of 2.5 % (v/v) Avertin/g of body weight, via intraperitoneal injection, exsanguinated and the blood samples were collected in EDTA tubes. The aorta was excised. The liver and other tissues were collected and snap-frozen in liquid nitrogen and stored at − 80 ◦ C until use.

Analysis of plasma lipid and lipoprotein levels Pooled plasma (200 µl) was separated by FPLC on two Superose 6 columns (Amersham Biosciences, Piscataway, NJ, U.S.A.). Fractions (0.5 ml) were collected using an elution buffer [154 mM NaCl, 1 mM EDTA and 0.02 % NaN3 (pH 8.2)] as described previously [10,12]. Very-LDL, LDL and HDL lipoproteins were separated by this technique. The cholesterol and triacylglycerol concentrations in each fraction and in plasma were determined using InfinityTM Cholesterol and Triglyceride reagents respectively (Sigma–Aldrich, St. Louis, MO, U.S.A.). Plasma glucose and NEFAs were determined using glucose and NEFA commercial kits obtained from Wako Chemicals (Richmond, VA, U.S.A.). Plasma insulin concentrations were measured by a commercial ELISA kit obtained from Crystal Chem Inc. (Downers Grove, IL, U.S.A.). The concentration of human apoB100 in

Atherosclerosis-susceptible mouse models

mouse plasma was determined by ELISA, as described previously [10,13], using monoclonal antibody 4G3 (University of Ottawa Heart Institute, Ottawa, Ontario, Canada), which recognizes the C-terminal residues 2980– 3084, i.e. full-length human apoB100 [14].

Western-blot analysis of mouse apoB, apoE and apoAI The mouse apoB, apoE and apoAI in mouse plasma and FPLC fractions were determined by Western-blot analysis. Briefly, mouse plasma samples (1 µl of a 1:10 dilution) or FPLC fractions (5 µl) were electrophoresed on a 6 % ProSieve 50 gel for apoB (FMC BioProducts, Rockland, ME, U.S.A.) and on a SDS/ 12.5 % (w/v) polyacrylamide gel for apoE and apoAI, and transferred on to an Immobilon P membrane (Millipore, Danvers, MA, U.S.A.), as described previously [10]. The membrane was then incubated with anti-(mouse apoB) polyclonal antibodies (generously given by Dr Thomas Innerarity, Gladeston Institute for Cardiovascular Disease, San Francisco, CA, U.S.A. [15]), or with anti-(mouse apoE) or -(mouse apoAI) polyclonal antibodies (generously given by Dr Brian Ishida, Cardiovascular Research Institute, University of California, San Francisco, CA, U.S.A.), followed by HRP (horseradish peroxidase)-conjugated rabbit anti-(mouse IgG) polyclonal antibodies (Amersham Biosciences). Bound secondary antibody was detected using the ECL (enhanced chemiluminescence) detection system (Amersham Biosciences). The specificity of the anti(mouse apoB) antibodies was confirmed by Western-blot analysis, as described previously [10], where no crossreactivity with human apoB was observed. The relative intensities of the apoB100, apoE and apoAI bands on the autoradiogram were determined by densitometry and calculated using the NIH (National Institutes of Health) Image program 1.58 f.

PAF-AH enzyme activity Blood samples collected as described above were diluted 1:150 (v/v) and FPLC fractions were diluted 1:4 (v/v) before carrying out the assay. PAF-AH activity was determined by the release of [3 H]acetate from the substrate 1-O-hexadecyl-2-[acetyl-3 H]-sn-glycerol-3phosphocholine ([3 H]PAF; 22 Ci/mmol; PerkinElmer– NewEngland Nuclear, Boston, MA, U.S.A.) at a final concentration of 100 µM, as described by McCall et al. [16]. Results are expressed as nmol of acetate released · h−1 · ml−1 of plasma or nmol of acetate released · h−1 · ml−1 of FPLC fraction.

Real-time quantitative RT (reverse transcriptase)-PCR for PAF-AH mRNA and 18 S RNA Total RNA from the mouse liver was extracted using TRIzol® reagent (Invitrogen, Carlsbad, CA, U.S.A.).

The sequence-specific primers and probes used for PAFAH and the endogenous control 18 S ribosomal RNA were designed using Primer Express Software (Applied Biosystems, Foster City, CA, U.S.A.). The nucleotide sequences used were as follows: mouse PAF-AH mRNA, 5 -ATGAGAGCGTCTTCGTGCGT (forward primer), 5 -TTTGGGATCCAAACAGTGTCG (reverse primer) and 5 -FAM-ACTACCCAGCTCAAGATCAAGGTCGCC-TAMRA (probe; where FAM is fluorescein aminohexylamidite and TAMRA is tetramethylrhodamine); mouse 18 S RNA, 5 -TAACGAACGAGACTCTGGCAT (forward primer), 5 -CGGACATCTAAGGGCATCACAG (reverse primer) and, 5 -FAM-TGGCTGAACGCCACTTGTCCCTCTAATAMRA (probe). The nucleotide sequences of each primer and probe were Blast searched against the GenBank® database to confirm the uniqueness of each primer. Real-time quantitative RT-PCR was performed on ABI Prism 7700 Sequence Detection System (Applied Biosystems). We used TaqMan One-Step RT-PCR Master Mix reagent (Applied Biosystem) to quantify RNA. The RNA standard curve for the specific gene was generated from T7-cDNA fragment using Lig’nScribe kit (Ambion, Austin, TX, U.S.A.), and serial dilutions of 103 –109 molecules were employed in duplicate for the assay. Total RNA was treated with DNase (DNA-free kit; Ambion) to remove DNA contamination. The optimum RNA concentration for each gene was determined initially using real-time quantitative RT-PCR. Each RNA sample was then diluted accordingly. We used 100 ng and 10 pg of total RNA to quantify PAF-AH mRNA and 18 S RNA respectively. Briefly, each 10 µl reaction contained 1 µl of diluted total RNA, 5 µl of 2× mastermix, 0.25 µl of 40× multiscribe and RNase inhibitor mix, 900 nM each of forward and reverse primers and 250 nM probe. A serial dilution of synthetic standard RNA was included with each experiment. The reaction was employed at 48 ◦ C for 30 min and 95 ◦ C for 10 min, followed by 40 cycles at 95 ◦ C for 15 s and 60 ◦ C for 1 min. At the end of the reaction, the result was analysed using the Sequence Detection System Software version 1.7 (Applied Biosystems). Each of the RNA samples was normalized with the endogenous control of 18 S ribosomal RNA. The copy numbers were calculated from the standard curve. The results are expressed as the ratio of specific RNA/18 S RNA.

En-face analysis of atherosclerotic lesions The mice were anaesthetized, as described above, exsanguinated and the aorta was carefully excised with part of the heart still attached. Under the stereomicroscope (Leica MZ60), all of the fat and adventitious tissues were removed. With the major branching vessels still attached, the aorta was opened longitudinally from the iliac bifurcation to the aortic arch and all the branching C


423

424

U. Singh and others

vessels and the heart were then removed. The aorta was pinned flat on a white wax surface, fixed overnight in 10 % (v/v) formalin and stained with freshly prepared filtered Oil Red O solution [17]. The aorta was scanned using the Polaroid Sprint Scan 35 Plus with Geoscan Enabler, the image was captured using Adobe Photoshop 5.0, and the background was removed with the guidance of the stereomicroscope. The total areas of the aorta and the atherosclerotic plaques were quantified using SigmaScan Pro 4.0 imaging software (SPSS Science, Chicago, IL, U.S.A.). The results were presented as the percentage of the aortic surface covered by lesions (mm2 ) divided by the total surface area of the aorta (mm2 ).

Statistical analysis

Results are presented as the means + − S.D. It is known that factors, such as the genotype of the animal, sex and age, all influence the levels of plasma lipids and apos, the levels of mRNAs and the severity of atheorosclerotic lesion development. The interactions between each factor often mask the thresholds for significance of the main factor. In the present study, we used ANOVA to analyse the effects of genotypes, sex and age on parameters including body weight, plasma cholesterol, triacylglycerol and human apoB (see, Table 1), glucose, insulin and NEFAs (see, Table 2), PAF-AH mRNA and PAF-AH activity (see Figure 3, and Table 3) and atherosclerotic lesions (see, Table 4). This method would reflect the true strength of the main factor and the interaction effects of several factors on each parameter, unlike the t statistic, which is only suitable for analysis of the effect of the main factor in the absence of interactions [18]. The corresponding P values are tabulated in Tables 1(B), 2(B), 3 and 4(B). A P value of < 0.05 is considered significantly different. The program for implementing ANOVA was written in MATLAB. First, we used the three main factors, genotype, sex and age, to analyse simultaneously their effects on each parameter. The main effect of each factor was to measure the changes of the mean value of each parameter at each level, e.g. three genotypes (C57BL/6, LDb and LTp), sex (male and female) and age (2, 5 and 8 months). The effect of the interaction was to measure the dependence of the main effect of each factor on the levels of the other factors.

RESULTS Plasma lipids, lipoproteins, glucose, insulin and NEFA levels in mice In Table 1(A), the plasma parameters of male and female mice of wild-type C57BL/6, LDb (mice expressing mouse apoB100) and LTp (mice expressing human and mouse apoB100) on a regular chow diet are shown. The results of the ANOVA are shown in Table 1(B). The body weight of each genotype increased steadily from C


2 months to 8 months of age in both males and females (Table 1; P < 10−12 ), and the males were heavier than the females (P = 7.7−8 ). Moreover, the body weights of LDb and LTp were significantly different from C57BL/6 wild-type mice (Table 1; P = 0.0003), whereas the differences between LDb and LTp were not significant (P = 0.1664). The levels of plasma cholesterol and triacylglycerols in LDb and LTp were at least 5 times higher than that of C57BL/6 mice. When the parameters of LTp mice to LDb mice were compared, the LTp animals had significantly higher levels of total cholesterol (P < 10−16 ) and triacylglycerols (P < 10−12 ) than LDb animals (Table 1) as a result of overexpressing human apoB100. The concentrations of human apoB levels in LTp mice were not influenced by age and gender (Table 1). The markedly elevated levels of plasma lipids were the result of increased concentrations of VLDL and IDL (intermediate density lipoprotein)/LDL with decreased levels of HDL, as determined by FPLC analysis (Figure 1). Compared with LDb mice, the LTp mice had more cholesterol and triacylglycerol in the VLDL and IDL/LDL fractions. The distribution of cholesterol, especially triacylglycerol, in the IDL/LDL fraction in LTp mice showed a more heterogeneous pattern; the peak of the curve shifted towards the right, which indicated the presence of smaller LDL. Interestingly, LTp mice also had lower HDL-cholesterol but increased triacylglycerol compared with LDb mice. These findings suggest that overexpression of human apoB100 produced a more heterogeneous pool of lipoproteins in plasma. In comparison with LDb animals, Western-blot analysis of FPLC fractions (Figure 2A) revealed LTp animals had less mouse apoE (total densitometric units: LTp = 656 compared with LDb = 1016) in VLDL and IDL/LDL fractions, as well as less mouse apoAI in HDL fractions (total densitometric units: LTp = 165 compared with LDb = 432). In contrast, mouse apoB was increased in the fractions of VLDL and IDL/LDL in LTp mice compared with LDb mice (total densitometric units: LTp = 345 compared with LDb = 153). These observations were consistent with the results shown in the mouse plasma of apoAI and apoE (Figure 2B). The LTp animals had 3.0- and 1.9-fold lower levels of apoAI and apoE respectively, than that of LDb animals (average densitometric units, apoAI: LDb = 78.0 + − 18.2 compared with LTp = 26.1 + − 2.2; apoE: LDb = 34.4 + − 5.6 compared with LTp = 18.3 + − 1.24). However, the mouse plasma apoB levels between these two genotypes remained relatively the same (average densitometric units, apoB: LDb = 26.0 + − 1.1 compared with LTp = 21.3 + − 2.5). Since the elevated levels of plasma triacylglycerols and the reduced concentrations of HDL are very common in patients with diabetes, we measured the fasting plasma glucose, insulin and NEFA levels (Table 2A) in these mice. Plasma glucose levels were not consistently different


Table 1 Levels of body weight, plasma cholesterol, triacylglycerol and human apoB in wild-type C57BL/6, LDb and LTp mice (A), and the strength of the individual factor (genotype, age and sex) and the interaction between each factor on body weight, plasma cholesterol, triacylglycerol and human apoB (B)

In (A), each value represents the average fasting plasma concentrations of 7–17 mice at the indicated age. The results are expressed as means + − S.D. In (B), results were analysed using ANOVA as described in the Materials and methods section. P values are shown, with P < 0.05 in bold. Three-genotype, C57BL/6, LDb and LTp; Two-genotype, LDb and LTp. (A)

C57BL/6 Age (months). . . 2 Body weight (g) Males Females Cholesterol (mg/dl) Males Females Triaclyglycerols (mg/dl) Males Females Human apoB (mg/dl) Males Females

LDb 2

LTp

5

8

5

8

2

5

8

21 + − 1.2 18 + − 0.4

31 + − 2.6 21 + − 0.4

36 + − 2.8 27 + − 2.5

26 + − 4.8 21 + − 2.9

29 + − 6.5 28 + − 6.1

32 + − 5.1 29 + − 2.8

23 + − 4.2 21 + − 5.4

33 + − 5.0 26 + − 2.9

40 + − 4.7 38 + − 5.1

73 + − 3.5 63 + − 11

70 + − 13 55 + − 7.7

59 + − 15 36 + − 12

375 + − 66 326 + − 61

468 + − 95 421 + − 82

371 + − 96 391 + − 23

503 + − 136 418 + − 56

811 + − 141 729 + − 208

821 + − 149 770 + − 235

74 + − 14 56 + − 7.8

77 + − 18 51 + − 3.1

56 + − 12 42 + − 19

208 + − 71 140 + − 27

284 + − 73 302 + − 118

324 + − 61 281 + − 35

455 + − 219 421 + − 88

560 + − 110 470 + − 95

525 + − 138 615 + − 145

– –

– –

– –

– –

– –

– –

214 + − 52 218 + − 45

249 + − 29 292 + − 55

329 + − 62 361 + − 103

(B) Main factor

Body weight Three-genotype Two-genotype C57BL/6 LDb LTp Cholesterol Three-genotype Two-genotype C57BL/6 LDb LTp Triacylglycerol Three-genotype Two-genotype C57BL/6 LDb LTp Human ApoB LTp

Interaction Sex

Genotype × age

Genotype × sex

Age × sex

Genotype × age × sex

7.7−8

0.0009

0.0005

0.0003

0.2660 0.7504

< 10−11

< 10−10

0.1023 0.3643 – – –

0.0008

0.5768 0.1366

0.0899 0.1148 – – –

0.4558 0.4894 – – –

0.7936 0.4344 0.9750 0.0870 0.4678

0.9564 0.7281 – – –

0.9591 0.9778 – – –

0.6699 0.4901 0.6434 0.1616 0.2696

0.1688 0.1227 – – –

–

0.5070

–

Genotype

Age

0.0003

< 10−12 −15

0.1664 – – –

0.0016

0.063

< 10−13

0.0028

– – –

< 10−16

< 10−13

1.9−6

9.4−9

−16

−12

0.0006

5.8−6

< 10

– – –

< 10

< 10

0.001

0.0005

0.003

0.0210

7.7−9

0.080

< 10−12

4.2−5

−12

0.2694

0.1115 0.3780

< 10

– – – –

– – – 0.0066

0.0258

0.0007

4.9−6 0.001

0.2973 0.3074

0.1942 – – –

0.2340

0.1953

–

among the wild-type C57BL/6, LDb and LTp mice (P = 0.1693). Similarly, plasma insulin levels were also not consistently different among the three genotypes (P = 0.6393). In contrast, when compared with C57BL/6 wild-type mice, the plasma NEFA levels in both

LDb and LTp mice increased significantly (P < 10−10 ; Table 2B) and the difference was influenced by sex (P = 0.0007). Moreover, NEFA levels were significantly different between LDb and LTp (P = 2.1−5 ), where LTp females had a significantly higher level than LDb females C


425

426

U. Singh and others

Figure 1 Distribution of cholesterol and triacylglycerols (triglyceride) in plasma separated by FPLC from male LDb (䊉) and male LTp (䊊) mice

Pooled plasma samples (200 µl) from 4–6 animals in each age group (2, 5 and 8 months) were fractionated by FPLC. Total cholesterol and triacylglycerols from each fraction were measured and expressed as mg/dl. The elution fraction for each lipoprotein class is shown. (P = 0.0010). Therefore the results suggest that, on a regular chow diet, despite marked hyperlipidaemia, these two genotypes (LDb and LTp) were not insulin-resistant, although they did have markedly elevated levels of plasma NEFA compared with C57BL/6 wild-type mice.

Altered activities of PAF-AH in LDb and LTp mice PAF-AH is a serine lipase, which can hydrolyse PAFlike molecules e.g. phospholipids, to produce LPC (lysophosphatidylcholine) and NEFAs. In mice, PAFAH is associated primarily with HDL, whereas, in humans, PAF-AH is associated preferentially with LDL, especially small-dense LDL [16,19–21]. Since LTp mice had a heterogeneous pool of LDL with the indication of the presence of small LDL, we measured the PAFAH mRNA levels, plasma enzyme activities and the distribution in lipoprotein fractions in these animals. When compared with C57BL/6 wild-type mice, both LDb and LTp mice (males and females at 5 and 8 months of age) had significantly elevated PAF-AH mRNA levels (P < 10−16 , P < 0.0124, and P < 0.0002 for genotype, age C


and sex respectively; Figure 3, and Table 3). Furthermore, PAF-AH mRNA levels were significantly higher in LTp mice than LDb mice (Figure 3, and Table 3). Next, we determined PAF-AH enzyme activity in the plasma of 5- and 8-month-old mice (the activity of 8-month-old C57BL/6 mice was not determined). The results reveal that both LDb and LTp mice had significantly higher (P = 0.0001) plasma PAF-AH enzyme activity than C57BL/6 mice (Figure 3, and Table 3). Similar to the results obtained for PAF-AH mRNA, LTp mice also had significantly higher (P = 0.0097) enzyme activity compared with LDb (Figure 3, and Table 3). Therefore the two atherosclerosis-prone animals had higher levels of PAF-AH mRNA and plasma PAF-AH activities compared with C57BL/6 wild-type mice. The distribution of PAF-AH activities among the lipoprotein fractions separated by FPLC was then determined in C57BL/6 wild-type, LDb and LTp mice. The cholesterol concentration and PAF-AH enzymic activity from each fraction were measured and the results were expressed as the ratio of FPLC fractions 1– 20 (VLDL + IDL/LDL fractions) divided by the total FPLC fractions (1–34). As expected, approx. 7 % of the total PAF-AH activity and approx. 3 % of the total cholesterol was associated with the VLDL/LDL fraction in the C57BL/6 wild-type mice. In contrast, most of the cholesterol in LDb and LTp mice (82 % and 90 % respectively) was in the VLDL/LDL fraction, and there was an increased association of PAF-AH activities in these fractions in LDb and LTp mice (13 % and 29 % respectively). The increased PAF-AH activity in LTp mice could be the result of overexpression of human apoB100 in these mice.

Atherogenesis Next, we used the en-face technique to quantify the atherosclerotic lesions in the whole aorta in order to determine the acceleration of atherosclerotic lesions in LTp and LDb mice. As shown in Table 4(A), atherosclerotic lesions developed more quickly in LTp mice than in LDb mice (P = 0.0066; Table 4B); it increased 4.6- and 4.3-fold from 5 months to 8 months of age in LTp males and females respectively (P < 10−6 ; Table 4B). On the other hand, in LDb mice, lesions extended 2.8- and 2-fold from 5 months to 8 months of age in males and females respectively (P < 10−6 ; Table 4B). At 8 months of age, LDb males had significantly more lesions (P = 0.006) than females, whereas, in LTp mice, there was no significant difference in lesion development (P = 0.1202) between males and females (Table 4B). LTp female mice had more lesions than the age-matched female LDb mice (1.9-fold more lesions). Therefore the gender difference in atherosclerosis susceptibility shown in LDb mice was lost in LTp mice. Taken together, overexpression of human apoB in the LTp mice accelerated lesion development from 5 months


Figure 2 Western-blot analyses of FPLC-fractionated plasma lipoproteins (A) and mouse plasma (B)

(A) Pooled plasma samples from 5-month-old male LDb and LTp mice were fractionated by FPLC. Fractions (5 µl) of VLDL (fractions 7–11), IDL/LDL (fractions 12–20) and HDL (fractions 25–34) were analysed by SDS/PAGE, followed by Western blotting with anti-(mouse apoE) (mApoE), anti-(mouse apoB) (mApoB) and anti-(mouse apoAI) (mApoAI) polyclonal antibodies. Immunoreactive bands were detected using the ECL detection system, as described in the Materials and methods section. The intensity of each band was determined by densitometry and analysed using the NIH Image program 1.58 f. (B) Pooled plasma samples (1 µl) from LDb and LTp male mice 2, 5 and 8 months of age were separated by SDS/PAGE, followed by Western blotting as described in (A). This experiment was performed three times. Similar results were obtained from the female mice. The intensity of each immunoreactive band was determined by densitometry and analysed using the NIH Image program 1.58 f. to 8 months of age, and the relative protection of gender from atherosclerosis was lost in the LTp female mice.

DISCUSSION The principal findings of the present study were that combined hyperlipidaemia was produced in mice lacking

both Apobec1 and LDLR, which was not associated with significant abnormalities in glucose or insulin metabolism, but was associated with elevated levels of the plasma levels of NEFA and PAF-AH and increased levels of PAF-AH mRNA. To the best of our knowledge, this is the first report of these findings in a mouse atherosclerosis model. C


427

428

U. Singh and others

Table 2 Plasma glucose, insulin and NEFA levels in wild-type C57BL/6, LDb and LTp mice (A), and the strength of the individual factor (genotype, age and sex) and the interaction between each factor on glucose, insulin and NEFAs (B)

In (A), each value represents the average fasting plasma concentrations of 7–15 mice at the indicated age. Results are expressed as means + − S.D. In (B), results were analysed using ANOVA as described in the Materials and methods section. P values are shown. Three-genotype, C57BL/6, LDb and LTp; Two-genotype, LDb and LTp. P < 0.05 is in bold. (A)

C57BL/6 Age (months). . . 2 Glucose (mg/dl) Males Females Insulin (ng/ml) Males Females NEFA (mmol/l) Males Females

187 + − 27 201 + − 28

LDb

LTp

5

8

2

5

8

2

5

8

195 + − 23 203 + − 32

214 + − 30 202 + − 36

198 + − 31 244 + − 39

182 + − 76 145 + − 37

184 + − 50 147 + − 26

180 + − 38 193 + − 18

200 + − 44 214 + − 46

173 + − 36 200 + − 62

0.89 + − 0.74 1.20 + − 0.68 0.68 + − 0.32 0.80 + − 0.18 0.42 + − 0.16 0.80 + − 0.26 0.93 + − 0.49 0.99 + − 0.41 1.01 + − 0.49 + + + + + + + + 0.66 + 0.43 0.61 0.34 0.25 0.05 1.08 0.63 0.54 0.49 0.64 0.26 0.58 0.40 0.59 0.29 0.57 − − − − − − − − − 0.29 0.57 + − 0.10 0.64 + − 0.1 + 0.38 + 0.09 0.38 − − 0.1

0.39 + − 0.1 0.32 + − 0.1

1.03 + − 0.14 1.70 + − 0.54 1.44 + − 0.96 1.94 + − 0.63 1.40 + − 0.41 1.20 + − 0.42 + + + + + 1.09 + 0.26 1.03 0.06 1.15 0.43 1.82 0.86 2.04 0.59 2.06 − − − − − − 0.96

(B) Main factor Genotype Glucose Three-genotype Two-genotype C57BL/6 LDb LTp Insulin Three-genotype Two-genotype C57BL/6 LDb LTp NEFA Three-genotype Two-genotype C57BL/6 LDb LTp

0.1693 0.2169 – – – 0.6393 0.3911 – – – < 10−10 2.1−5

– – –

Interaction Age

Sex

Genotype × age

Genotype × sex

Age × sex


0.2739

0.5660 0.5869 0.7519 0.6169 0.0945

0.0009

0.2134 0.0979 – – –

0.1344 0.0921 0.6058 0.8910

0.1299 0.0566 – – –

0.3626 0.3795 0.7390 0.1813 0.9580

0.8461 0.6229 – – –

0.4281 0.4474

0.0326

0.0434

– – –

0.0272

0.5707 0.0004

0.1968

0.0013

– – –

0.1773 0.0584 0.1552

0.0012

0.0135

0.0063

0.0394

0.0259

0.0024

0.0307

0.0009

0.1276

0.9496

0.0009

– – –

– – –

0.6561 0.8145 0.0013 0.1756 0.2948

0.9704 0.4786

0.2720 0.1387 – – –

0.0007

1.1−5 0.0293 0.0123

The abnormal fatty acid metabolism observed in both genetically altered mice (LDb and LTp) would not be predicted based on their predecessors. Mice deficient in Apobec1 have normal chylomicron transport [22] and normal levels of plasma triacylglycerols [8,9,22] and NEFA (males, 0.49 + − 0.23 mmol/l; females, 0.42 + − 0.19 mmol/l). Similarly, mice deficient in LDLR have normal ranges of plasma NEFAs (males, 0.68 + − 0.30 mmol/l; females, 0.53 + − 0.11 mmol/l) and normal chylomicron transport and normal levels of plasma triacylglycerols [8]. However, when crossbred, these C


0.0010

– – –

0.0134

0.1596 0.0672

0.0153

two strains, as well as overexpressing human apoB, had elevated plasma NEFAs, elevated plasma cholesterol and triacylglycerol, increased VLDL and LDL levels and accelerated atherosclerotic lesion development. This combination of findings points to the fact that combined hyperlipidaemia is due to increased hepatic VLDL secretion, which is due, in turn, to increased fatty acid flux to the liver. Of interest, NEFA levels were even higher in the LTp mice, and this was associated with even higher levels of VLDL and LDL. These observations support the hypothesis that fatty acid flux is a critical determinant


Figure 3 Levels of PAF-AH mRNA and plasma PAF-AH enzyme activity in atherosclerosis-prone mice (LDb and LTp )

Upper panels, PAF-AH mRNA was determined by real-time quantitative RT-PCR. The results are normalized with 18 S RNA and are presented as means + − S.D. PAF-AH mRNA levels of C57BL/6 wild-type (c57), LDb and LTp of males and females at 5- and 8-months-old are shown. Lower panels, plasma PAF-AH activities of 5- and 8-month-old animals are shown: C57BL/6 is depicted in dotted bar, LDb mice in shaded bar, and LTp in dark-shaded bar. The results are presented as means + − S.D. in nmol of acetate · h−1 · ml−1 of plasma. Statistical analysis using ANOVA of these results is shown in Table 3. Table 3 Strength of the individual factor (genotype, age and sex) and the interaction between each factor on PAF-AH mRNA and plasma PAF-AH activity

Results were analysed using ANOVA as described in the Materials and methods section. P values are shown, with P < 0.05 in bold. Three-genotype, C57BL/6, LDb and LTp; Two-genotype, LDb and LTp. Main factor

PAF-AH mRNA Three-genotype Two-genotype C57BL/6 LDb LTp PAF-AH activity Three-genotype Two-genotype C57BL/6 LDb LTp

Interaction

Genotype

Age

Sex

< 10−16

0.0124

0.0002

4.7

−9

– – – 9.7−7 0.0001

– – –

Genotype × age

Genotype × sex

Age × sex


0.0043

6.3−5

0.5742 – – –

0.1626

0.0586 0.1662 – – –

0.9305 0.7685 – – –

0.6363 0.6426 – 0.5864 0.8013

0.9791 0.9810 – – –

0.3696

0.7117

0.0036

4.0−5

0.0785

0.0859

0.1897 0.8956 – – –

0.0532 0.8505 – 0.8895 0.8250

0.6744 0.8791 0.6505 0.9688 0.7164

0.7950 0.8134 – – –

0.0119

7.7

−5

of hepatic triacylglycerol and cholesterol synthesis and, therefore, of the rate of secretion of the apoB-containing lipoproteins [23,24]. In these animals, increased secretion of apoB100containing lipoproteins resulted in elevated levels of

0.0003

0.1180 4.3−5

plasma NEFAs because of either impaired fatty acid uptake or increased fatty acid release from peripheral tissues [24]. The result is a vicious cycle, the consequence of which is the accumulation in plasma of large amounts of VLDL and IDL/LDL particles. However, it is not clear C


429

430

U. Singh and others

Table 4 Atherosclerotic lesions in mice on a regular chow diet (A) and strength of main factors (genotype, sex and age) and their interactions between each factor on lesion development (B)

In (A), results are presented as means + − S.D. for the indicated numbers of each genotype. n values are shown in parentheses. nd, not determined. Percentage lesions, total atherosclerotic lesion area in mm2 /total surface area of the aorta in mm2 . In (B), results were analysed using ANOVA as described in the Materials and methods section. P values are shown, with P < 0.05 in bold. Two-genotype, LDb and LTp. (A)

C57BL/6 Age (months). . . Percentage lesion Males Females

LDb

LTp

2

5

8

2

5

8

2

5

8

nd nd

0 (5) 0 (5)

0 (5) 0 (5)

nd nd

6.5 + − 2.0 (7) 5.4 + − 1.8 (7)

18.5 + − 3.3 (7) 12.1 + − 4.1 (7)

nd nd

5.3 + − 2.8 (7) 5.0 + − 3.2 (7)

23.5 + − 5.8 (7) 22.3 + − 6.1 (7)

(B) Main factor Lesion (%) Two-genotype LDb LTp

Genotype 0.0066

– –

Interaction Age

Sex

Genotype × age

Genotype × sex

Age × sex


0.5270 – –

0.1165 – –

0.7597 – –

< 10−14

0.0416

0.0011

< 10−6

0.006

< 10−6

0.1202

– –

why the combination of genetic deletions should interfere with the normal peripheral clearance of triacylglycerols and uptake of fatty acids by adipose tissue or why they should interfere with the normal clearance of chylomicron remnants by the liver. Nevertheless, one, or other, or both, must have occurred to account for the concurrent elevation in NEFAs, cholesterol and triacylglycerols in both the LDb and LTp mice. The second major finding was that, on a chow diet, notwithstanding marked combined hyperlipidaemia and elevated NEFA levels, there was no marked disturbance of plasma glucose or insulin in either LDb or LTp mice. Thus there was no evidence of insulin resistance in mice with obviously abnormal fatty acid flux and metabolism. This is in contrast with the association between dyslipidaemia and dysglycaemia, which has been observed so frequently in humans. However, since the measurements of glucose and insulin levels were done in mouse plasma after 10 h fasting and since we did not directly test the glucose tolerance [25], we cannot rule out that these animals may still have insulin resistance and/or they might develop insulin resistance on a high-fat diet. The third major finding was the relationship of the increased levels of NEFA and PAF-AH in the development of atherosclerotic lesions in these animals. As shown in Figure 4, the atherosclerosis-susceptible animals clearly had elevated levels of NEFA and PAF-AH activity. These three genotypes had distinctly different levels of PAF-AH activities, with LTp mice (mice overexpressing human apoB) having the highest levels of PAF-AH activity. Elevated NEFA levels in humans have been shown to impair nitric oxide release from the endothelium [26]. C


Furthermore, increased NEFAs might also contribute to the induction of reactive oxygen species generation in the endothelial and vascular smooth muscle cells [27]. Thus the elevated levels of NEFA in both LDb and LTp mice may directly contribute to the development of atherosclerosis. A current perspective has listed the levels of circulating NEFAs as a factor to predict the patient’s risk of acute cardiovascular complications [28]. In contrast, the association of PAF-AH with disease has been controversial. Increased PAF-AH activity has been demonstrated in human and rabbit atherosclerotic lesions [29], but its relationship to risk in human clinical studies has been variable. The strongest results, to date, have come from the study from the West of Scotland Coronary Prevention Study [30]. In a nested casecontrolled study, Packard et al. [30] have shown that the level of PAF-AH in plasma is positively associated with coronary artery disease. Two other studies have supported this observation. In univariate analyses of the Women’s Heart Study [31], PAF-AH was a significant predictor of cardiovascular risk, but the strength of the relationship diminished considerably on multivariate analyses. Preliminary results from the ARIC (Atherosclerosis Risk in Communities Study) also point to a positive relationship with coronary disease, at least at low levels of LDL-cholesterol [32,33]. In contrast with these results in Caucasians, deficiency of PAF-AH in plasma in Japanese patients has been associated with an increased risk of stroke and coronary artery disease [34]. These contradictory results may be explained by the fact that PAF-AH is carried in both LDL and HDL. The relationship to risk is positive in LDL, whereas the


Figure 4 Relationships between NEFAs, PAF-AH mRNA and PAF-AH activity with atherosclerotic lesion development in C57BL/6 wild-type, LDb and LTp mice

Values are means + − S.D. relationship to risk is negative for HDL. Our present study has documented that both LDb and LTp mice had an increased association of PAF-AH with apoBcontaining lipoproteins and, therefore, the risk would have been predicted to be increased. There are at least two mechanisms by which LDL-associated PAF-AH could increase the risk of vascular disease. First, PAFAH appears to promote the formation of smaller dense LDL particles [32,35], which enter the arterial wall more easily and bind to the proteoglycans within the arterial wall more avidly than the larger, more buoyant, LDL particles. The results from our FPLC analyses of the PAF-AH activities on lipoproteins in both LDb and LTp mice show the peak shift towards the small LDL regions. Secondly, LDL particles to which PAF-AH is attached will generate more LPC and oxidatively modified NEFAs [36]. This will accelerate LDL oxidation and provoke an inflammatory response within the arterial wall, both of which will accelerate atherogenesis. In summary, the present study indicates for the first time in a murine model that elevation of plasma NEFAs is associated with an increased risk of atherosclersosis and that increased NEFAs are not associated with insulin resistance. The mechanisms by which NEFAs contribute to atherosclerosis are currently being actively studied in our laboratory.

ACKNOWLEDGMENTS This work was supported by a grant from the NIH (HL 53441), and an Established Investigator Grant from the American Heart Association (9740192N). We are grateful to Dr Edward Rubin (University of

California, Berkeley) for providing the transgenic human apoB (ERhB+/+ ) mice and Dr Lawrence Chan (Baylor College of Medicine, Houston) for generously giving the Apobec1−/− mice. We are also grateful to Dr Trudy Forte (University of California, Berkeley) for providing the established assay of the determination of PAF-AH activities in plasma and FPLC fractions.

REFERENCES 1 Sniderman, A. D., Furberg, C. D., Keech, A. et al. (2003) Apolipoproteins versus lipids as indices of coronary risk and as targets for statin treatment. Lancet 361, 777–780 2 Williams, K., Sniderman, A. D., Sattar, N., D’Agostino, Jr, R., Wagenknecht, L. E. and Haffner, S. M. (2003) Comparison of the associations of apolipoprotein b and low-density lipoprotein cholesterol with other cardiovascular risk factors in the Insulin Resistance Atherosclerosis Study (IRAS). Circulation 108, 2312–2316 3 Teng, B.-B., Sniderman, A. D., Soutar, A. K. and Thompson, G. R. (1986) Metabolic basis of hyperapobetalipoproteinemia: turnover of apoB in LDL and its precursors and subfractions compared with normal and familial hypercholesterolemia. J. Clin. Invest. 77, 663–672 4 Kissebah, A. H. (1987) Low density lipoprotein metabolism in non-insulin-dependent diabetes mellitus. Diabetes Metab. Rev. 3, 619–651 5 Ginsberg, H. N. (1996) Diabetic dyslipidemia: basic mechanisms underlying the common hypertriglyceridemia and low HDL cholesterol levels. Diabetes 45 (Suppl 3), S27–S30 6 Sanan, D. A., Newland, D. L., Tao, R. et al. (1998) Low density lipoprotein receptor-negative mice expressing human apolipoprotein B-100 develop complex atherosclerotic lesions on a chow diet: no accentuation by apolipoprotein(a). Proc. Natl. Acad. Sci. U.S.A. 95, 4544–4549 7 Teng, B., Burant, C. F. and Davidson, N. O. (1993) Molecular cloning of an apolipoprotein B messenger RNA editing protein. Science (Washington, D.C.) 260, 1816–1819

C


431

432

U. Singh and others

8 Powell-Braxton, L., Veniant, M., Latvala, R. D. et al. (1998) A mouse model of human familial hypercholesterolemia: markedly elevated low density lipoprotein cholesterol levels and severe atherosclerosis on a low-fat chow diet. Nat. Med. 4, 934–938 9 Nakamuta, M., Chang, B. H.-J., Zsigmond, E. et al. (1996) Complete phenotypic characterization of apobec-1 knockout mice with a wild-type genetic background and a human apolipoprotein B mRNA editing by somatic gene transfer of apobec1. J. Biol. Chem. 271, 25981–25988 10 Enjoji, M., Wang, F., Nakamuta, M., Chan, L. and Teng, B. B. (2000) Hammerhead ribozyme as a therapeutic agent for hyperlipidemia: production of truncated apolipoprotein B and hypolipidemic effects in a dyslipidemia murine model. Hum. Gene Ther. 11, 2415–2430 11 Callow, M. J., Stoltzfus, L. J., Lawn, R. M. and Rubin, E. M. (1994) Expression of human apolipoprotein B and assembly of lipoprotein(a) in transgenic mice. Proc. Natl. Acad. Sci. U.S.A. 91, 2130–2134 12 Cole, T., Kitchens, R. T., Daugherty, A. and Schonfeld, G. (1988) FPLC. BioCommunique 4, 4–6 13 Pullinger, C. R., North, J. D., Teng, B. B., Rifici, V. A., Ronhild de Brito, A. E. and Scott, J. (1989) The apolipoprotein B gene is constitutively expressed in HepG2 cells: regulation of secretion by oleic acid, albumin, and insulin, and measurement of the mRNA half-life. J. Lipid Res. 30, 1065–1077 14 Pease, R. J., Milne, R. W., Jessup, W. K. et al. (1990) Use of bacterial expression cloning to localize the epitopes for a series of monoclonal antibodies against apolipoprotein B100. J. Biol. Chem. 265, 553–568 15 Boren, J., Lee, I., Zhu, W., Arnold, K., Taylor, S. and Innerarity, T. L. (1998) Identification of the low density lipoprotein receptor-binding site in apolipoprotein B100 and the modulation of its binding activity by the carboxyl terminus in familial defective apoB-100. J. Clin. Invest. 101, 1084–1093 16 McCall, M. R., La Belle, M., Forte, T. M., Krauss, R. M., Takanami, Y. and Tribble, D. L. (1999) Dissociable and nondissociable forms of platelet-activating factor acetylhydrolase in human plasma LDL: implications for LDL oxidative susceptibility. Biochim. Biophys. Acta 1437, 23–36 17 Tangirala, R. K., Rubin, E. M. and Palinski, W. (1995) Quantitation of atherosclerosis in murine models: correlation between lesions in the aortic origin and in the entire aorta, and differences in the extent of lesions between sexes in LDL receptor-deficient and apolipoprotein E-deficient mice. J. Lipid Res. 36, 2320–2328 18 Montgomery, D. C. (1991) Design and Analysis of Experiments, John Wiley & Sons, New York 19 Karabina, S. A., Liapikos, T. A., Grekas, G., Goudevenos, J. and Tselepis, A. D. (1994) Distribution of PAFacetylhydrolase activity in human plasma low-density lipoprotein subfractions. Biochim. Biophys. Acta 1213, 34–38 20 Tselepis, A. D., Dentan, C., Karabina, S. A., Chapman, M. J. and Ninio, E. (1995) PAF-degrading acetylhydrolase is preferentially associated with dense LDL and VHDL-1 in human plasma. Catalytic characteristics and relation to the monocyte-derived enzyme. Arterioscler., Thromb., Vasc. Biol. 15, 1764–1773 21 Tsimihodimos, V., Karabina, S. A., Tambaki, A. P. et al. (2002) Altered distribution of platelet-activating factoracetylhydrolase activity between LDL and HDL as a function of the severity of hypercholesterolemia. J. Lipid Res. 43, 256–263

22 Morrison, J. R., Paszty, C., Stevens, M. E. et al. (1996) Apolipoprotein B RNA editing enzyme-deficient mice are viable despite alterations in lipoprotein metabolism. Proc. Natl. Acad. Sci. U.S.A. 93, 7154–7159 23 Adeli, K., Taghibiglou, C., Van Iderstine, S. C. and Lewis, G. F. (2001) Mechanisms of hepatic very low-density lipoprotein overproduction in insulin resistance. Trends Cardiovasc. Med. 11, 170–176 24 Sniderman, A. D., Cianflone, K., Arner, P., Summers, L. K. and Frayn, K. N. (1998) The adipocyte, fatty acid trapping, and atherogenesis. Arterioscler., Thromb., Vasc. Biol. 18, 147–151 25 Ahren, B., Oosterwijk, C., Lips, C. J. and Hoppener, J. W. (1998) Transgenic overexpression of human islet amyloid polypeptide inhibits insulin secretion and glucose elimination after gastric glucose gavage in mice. Diabetologia 41, 1374–1380 26 Steinberg, H. O. and Baron, A. D. (2002) Vascular function, insulin resistance and fatty acids. Diabetologia 45, 623–634 27 Inoguchi, T., Li, P., Umeda, F. et al. (2000) High glucose level and free fatty acid stimulate reactive oxygen species production through protein kinase C–dependent activation of NAD(P)H oxidase in cultured vascular cells. Diabetes 49, 1939–1945 28 Naghavi, M., Libby, P., Falk, E. et al. (2003) From vulnerable plaque to vulnerable patient: a call for new definitions and risk assessment strategies: Part II. Circulation 108, 1772–1778 29 Hakkinen, T., Luoma, J. S., Hiltunen, M. O. et al. (1999) Lipoprotein-associated phospholipase A2 , plateletactivating factor acetylhydrolase, is expressed by macrophages in human and rabbit atherosclerotic lesions. Arterioscler., Thromb., Vasc. Biol. 19, 2909–2917 30 Packard, C. J., O’Reilly, D. S., Caslake, M. J. et al. (2000) Lipoprotein-associated phospholipase A2 as an independent predictor of coronary heart disease. West of Scotland Coronary Prevention Study Group. N. Engl. J. Med. 343, 1148–1155 31 Blake, G. J., Dada, N., Fox, J. C., Manson, J. E. and Ridker, P. M. (2001) A prospective evaluation of lipoproteinassociated phospholipase A2 levels and the risk of future cardiovascular events in women. J. Am. Coll. Cardiol. 38, 1302–1306 32 Caslake, M. J. and Packard, C. J. (2003) Lipoproteinassociated phospholipase A2 (platelet-activating factor acetylhydrolase) and cardiovascular disease. Curr. Opin. Lipidol. 14, 347–352 33 Ballantyne, C., Hoogeveen, R., Bank, H. et al. (2004) Lipoprotein-associated phospholipase A2 , high sensitivity C-reactive protein and risk for incident coronary heart disease in middle-aged men and women in the Atherosclerosis Risk in Communities (ARIC) Study. Circulation, in the press 34 Yamada, Y., Ichihara, S., Fujimura, T. and Yokota, M. (1998) Identification of the G994 → T missense in exon 9 of the plasma platelet-activating factor acetylhydrolase gene as an independent risk factor for coronary artery disease in Japanese men. Metabolism 47, 177–181 35 Hurt-Camejo, E., Camejo, G. and Sartipy, P. (2000) Phospholipase A2 and small, dense low-density lipoprotein. Curr. Opin. Lipidol. 11, 465–471 36 MacPhee, C. H., Moores, K. E., Boyd, H. F. et al. (1999) Lipoprotein-associated phospholipase A2 , plateletactivating factor acetylhydrolase, generates two bioactive products during the oxidation of low-density lipoprotein: use of a novel inhibitor. Biochem. J. 338, 479–487

Received 14 November 2003/5 January 2004; accepted 13 January 2004 Published as Immediate Publication 13 January 2004, DOI 10.1042/CS20030375

C
