Zei-Shung Huang and Bor-Luen Chiang 1

Transcription

1 CORRELATION BETWEEN SERUM LIPID PROFILES AND THE RATIO AND COUNT OF THE CD16+ MONOCYTE SUBSET IN PERIPHERAL BLOOD OF APPARENTLY HEALTHY ADULTS Zei-Shung Huang and Bor-Luen Chiang 1 Background and Purpose: Deposited vascular oxidized low-density lipoproteins (LDLs) are important triggers of the transformation of circulatory monocytes into macrophages. CD16+ monocytes have been reported to be the precursors of tissue macrophages. In this study, we sought to determine the relationship between serum LDL-cholesterol and the percentage and count of the CD16+ monocyte subset in the peripheral blood of healthy adults. Methods: We studied the correlations between serum lipid profiles and both peripheral CD16+ and CD36+ monocyte subset ratios and counts in apparently healthy adults (50 men and 50 women). Monocyte surface antigens CD16 and CD36 on CD14+ monocytes were detected using fluorescent triple staining and flow-cytometry. Surface staining was performed by incubating 1 x 10 6 blood mononuclear cells with phycocrythrin-conjugated anti-cd14, fluorescein isothiocyanate-conjugated anti- CD36, and respective control isotopes (mouse IgGs). A total of 5,000 cells were counted and the frequency of surface antigens was determined by FACscan. Results: A significant positive link between LDL-cholesterol and the CD16+ subset ratio was found by linear correlation analysis (p < 0.05) but not by multivariate regression analysis. Both linear correlation analysis and ANOVA revealed a significant inverse link between high-density lipoprotein (HDL)-cholesterol and the CD16+ subset ratio (both p < 0.01). By multivariate regression analysis, gender was the main significant determinant for the CD16+ subset ratio. When serum total cholesterol (TC) was excluded from the analysis to avoid the interference from collinearity between serum TC and LDL (r = 0.84), HDL-cholesterol became independently and inversely linked to the CD16+ subset ratio. There were independent inverse links between HDL-cholesterol and the counts of all monocytes, CD16+ monocytes, and CD36+ monocytes. Conclusions: Our results suggest that circulating HDL-cholesterol may be much more important than LDL-cholesterol in affecting the transformation of circulatory monocytes into macrophages. The inverse link between HDL-cholesterol and the number of macrophage precursors in peripheral blood might contribute partly to the well-known antiatherogenic effect of HDL-cholesterol. (J Formos Med Assoc 2002;101:11 7) Key words: CD16 HDL-cholesterol LDL-cholesterol monocytes atherosclerosis Monocytes and monocyte-derived macrophages play important roles in hypercholesterolemia-induced atherosclerosis [1 3]. Accumulated evidence has suggested that deposited vascular oxidized low-density lipoproteins (LDLs) are important triggers of the transforma- tion of circulatory monocytes into macrophages that adhere to the vascular wall and migrate into the intimal space [2 6]. Since a high serum LDL-cholesterol is associated with a high atherosclerotic risk, and circulating LDLs are generally thought to be the main source Departments of Internal Medicine and 1 Pediatrics, College of Medicine, National Taiwan University, Taipei. Received: 13 July Revised: 7 September Accepted: 6 November Reprint requests and correspondence to: Dr. Bor-Luen Chiang, Department of Pediatrics, National Taiwan University Hospital, 7 Chung-Shan South Road, Taipei, Taiwan. J Formos Med Assoc 2002 Vol 101 No 1 11

2 Z.S. Huang and B.L. Chiang of vascular oxidized LDLs [4 6], the percentage of macrophage precursors in peripheral blood may have a positive link with serum LDL-cholesterol. Macrophage precursors are monocytes owning certain characteristics of macrophages including expression of surface antigen CD16, the Fc-gamma receptor III, which has been reported to be a marker of the tissue macrophages [7, 8]. This study investigated the relationship between various serum lipid levels and the percentage of CD16+ monocytes (CD14+CD16+/all-CD14+, CD16+ subset ratio) in apparently healthy adults to determine if there is a significant link between the CD16+ subset ratio and serum LDL-cholesterol or other lipids. In addition, because CD36 has been reported to be a lipidrelated scavenger receptor and may play a role in foam cell formation [9, 10], our study also included the measurement of the percentage of CD36+ monocytes (CD14+CD36+/all-CD14+, CD36+ subset ratio). 12 Materials and Methods Study subjects A total of 100 apparently healthy adults, 50 men and 50 women, were selected from participants in a regular health check program at our hospital. Peripheral blood cell analysis, including leukocyte classification, and lipid profile study were regular items in the health check program. The CD16+ and CD36+ monocyte subset ratios were measured using the remaining blood sample after leukocyte classification, and thus no extra blood sampling was needed. The age group distributions of the men and women were the same: 35 to 49 years, 15 subjects each; 50 to 59 years, 29 subjects each; and at least 60 years, six subjects each. The mean age of the men was ± 1.07 years and of the women was ± 0.95 years (mean ± standard error of the mean). For each subject, an overnight (> 8 hr) fasting blood sample was collected from a peripheral vein, and the laboratory measurements listed below were performed within 2 hours. Lipid measurement Serum levels of total cholesterol (TC, mg/dl), LDLcholesterol, high-density lipoprotein-cholesterol (HDL-cholesterol, mg/dl), and triglycerides (TGs, mg/dl) were determined in all specimens using an Automatic Multichannel Biochemical Analyzer (Hitachi-7450, Hitachi, Tokyo, Japan) following routine laboratory procedures in our hospital. Subjects with a serum TG of more than 450 mg/dl were excluded from the study because of its profound influence on the measurement of serum LDL-cholesterol. Monocyte measurement Monocyte measurements included monocyte count (x 10 9 /L), CD16+ subset ratio (%), and CD36+ subset ratio (%). Leukocyte composition was analyzed using a Sysmex Cell Counter (NE-8000, TOA Medical Electronics, Kobe, Japan), and the monocyte count was calculated as the product of the total leukocyte count and the monocyte differential percentage. Surface antigens CD16 and CD36 on CD14+ monocytes were detected using fluorescent triple staining and flowcytometry regularly calibrated with CaliBRITE beads (Becton Dickinson, Mountain View, CA, USA). The procedures were briefly as follows. Surface staining was performed by incubating 1 x 10 6 peripheral blood mononuclear cells with phycocrythrin (PE)-conjugated anti-cd14 (mouse IgG2a) and fluorescein isothiocyanate (FITC)-conjugated anti-cd36 (mouse IgG1), cychrome 5 (Cy5)-conjugated anti-cd16 (mouse IgG1) and mouse IgG2a isotope control, PE-conjugated; mouse IgG1 isotope control, Cy5-conjugated; mouse IgG1 isotope control, FITC-conjugated (Immunotech, Marseille Cedex, France). All antibody incubations were performed at 4 C for 30 minutes. The cells were washed and resuspended in 0.5 ml phosphate buffered saline (PBS) with 0.1% sodium azide and subjected to FACscan analysis. A total of 5,000 cells were counted and the frequency of each cell surface marker was determined using appropriate software (FACscan, Becton Dickinson). Only cells suspended in the medium served as controls. Statistical methods Relationships between monocyte measurements and lipid measurements were analyzed using an SAS statistical program. Comparisons of means of age, lipid measurements, and monocyte measurements by gender were performed using Student s t-test. Comparisons of means of age and lipid measurements by three levels of monocyte measurement were analyzed by ANOVA. Linear correlations were analyzed by Pearson s method. Multivariate regression analyses were calculated using the ordinary least squares method. Results All study subjects had a CD36+ subset ratio larger than 85.0% (range, %) except for one woman who had a very low CD36+ subset ratio of 0.90%. This subject was thought to have CD36 deficiency and was excluded from the statistical analyses of the CD36+ subset ratio. The monocyte count and CD16+ subset ratio of this CD36-deficient subject were x 10 9 /L J Formos Med Assoc 2002 Vol 101 No 1

3 and 9.31%, respectively. Among the 100 subjects, the monocyte counts ranged from to x 10 9 /L, the CD16+ subset ratios ranged from 5.74 to 45.77%, the serum TC ranged from 121 to 275 mg/dl, and the TG ranged from 29 to 362 mg/dl. In six subjects, the serum TC was more than 240 mg/dl; in 35, it was 200 to 239 mg/dl; and in 59, it was less than 200 mg/dl. In 23 subjects, the serum TG was more than 170 mg/dl; in 26, it was 130 to 169 mg/dl; and in 51, it was less than 130 mg/dl. Comparisons of the means of age, monocyte measurements, and lipid measurements by gender are shown in Table 1. The means of the monocyte count (p < 0.01), CD16+ subset ratio (p < 0.01), CD36+ subset ratio (p < 0.01), and LDL-cholesterol (p < 0.05) in men were markedly higher than those in women. In contrast, the mean HDL-cholesterol in men was markedly lower than that in women (p < 0.01). Comparisons of the means of age and serum TC, LDL-cholesterol, HDL-cholesterol, and TGs for three levels of monocyte measurements, analyzed by ANOVA, are shown in Table 2. A consistent inverse-trend relationship between HDL-cholesterol and the three monocyte measurements was found: monocyte count (p < 0.01), CD16+ subset ratio (p < 0.01), and CD36+ subset ratio (p < 0.05). Although a significant association was found between LDL-cholesterol and three levels of CD16+ subset ratio (p < 0.01), this relationship was neither a positive nor an inverse trend but, instead, was concave-shaped, i.e., the group with the middle level of CD16+ subset ratio had the lowest LDL-cholesterol. Table 1. Difference between male and female subjects in means of age, monocyte measurements, serum total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), and triglycerides (TGs) (n = 100) Males Females p * (n = 50) (n = 50) Age (yr) ± ± Monocyte count ± ± < 0.01 (x 10 9 /L) CD16+ subset ratio ± ± 0.96 < 0.01 (%) CD36+ subset ratio ± ± 0.51 < 0.01 (%) TC ± ± LDL-C ± ± 3.81 < 0.05 HDL-C ± ± 1.52 < 0.01 TGs ± ± * Estimated using Student s t-test with mean ± standard error of the mean. A female subject with CD36 deficiency was not included in the analyses concerning CD36+ subset ratio. Table 3 shows the results of linear correlation analyses between monocyte measurements, age and various serum lipid levels. The following results were significant: The monocyte count was inversely correlated with HDL-cholesterol (p < 0.01, r = 0.48); and the CD16+ subset ratio was positively correlated with LDL-cholesterol and TG (both p < 0.05, r = 0.20), and inversely correlated with HDL-cholesterol (p < 0.01, r = 0.45). The CD36+ subset ratio correlated inversely with HDLcholesterol (p < 0.05, r = 0.24). The correlation coefficient between the CD16+ subset ratio and HDL-cholesterol (r = 0.45) was larger than that between the CD16+ subset ratio and LDL-cholesterol (r = 0.20), indicating that the strength of the correlation between the CD16+ subset ratio and HDL-cholesterol was greater than that between the CD16+ subset ratio and LDL-cholesterol. The results of multivariate regression analysis of monocyte measurements on age, gender, and serum TC, LDL-cholesterol, HDL-cholesterol, and TGs are shown in Table 4. Only HDL-cholesterol was significantly correlated with monocyte count (p < 0.01). Gender was the most significant determinant of both the CD16+ subset ratio and the CD36+ subset ratio. When serum TC was excluded from the analysis to avoid the interference of collinearity from a strong linear correlation between serum TC and serum LDLcholesterol (r = 0.84, p < 0.01 by Pearson s method), HDL-cholesterol became independently and inversely linked to the CD16+ subset ratio (p < 0.05), but not to the CD36+ subset ratio (p = 0.56). LDL-cholesterol and TG had no significant association with either the CD16+ subset ratio or the CD36+ subset ratio. Table 5 shows the results of additional multivariate regression analysis of the CD16+ and CD36+ monocyte counts on age, gender, and serum lipid levels. This analysis revealed independent inverse links between HDL-cholesterol and both the CD16+ monocyte count and the CD36+ monocyte count (both p < 0.01 when serum TC was excluded from the analysis), while no significant link was found between these monocyte subset counts and LDL-cholesterol. Discussion Our study revealed that the CD16+ monocyte subset ratio in peripheral blood correlated much better with serum HDL-cholesterol than with serum LDLcholesterol (Tables 2 4). Since circulating CD16+ monocytes are likely the precursors of macrophages [7, 8] and circulating LDLs are generally thought to be the main source of vascular oxidized LDLs, our results seem incompatible with previous findings that vascular J Formos Med Assoc 2002 Vol 101 No 1 13

4 Z.S. Huang and B.L. Chiang Table 2. Comparisons of the means of age, serum total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), and triglycerides (TGs) by monocyte count, CD16+ subset ratio, and CD36+ subset ratio Monocyte count (x 10 9 /L) CD16+ subset ratio (%) CD36+ subset ratio (%) (n = 30) (n = 34) (n = 36) p * (n = 29) (n = 40) (n = 31) p * (n = 33) (n = 32) (n = 34) p * Age ± ± ± ± ± ± ± ± ± (yr) TC ± ± ± ± ± ± ± ± ± LDL-C ± ± ± ± ± ± 4.76 < ± ± ± HDL-C ± ± ± 1.74 < ± ± ± 1.26 < ± ± ± 1.63 < 0.05 TGs ± ± ± ± ± ± ± ± ± Values are mean ± standard error of the mean. * Calculated by ANOVA. A female subject with CD36 deficiency was not included in the analyses concerning CD36+ subset ratio. 14 J Formos Med Assoc 2002 Vol 101 No 1

5 oxidized LDLs are important triggers of the transformation of circulatory monocytes into macrophages [2 6]. There are two possible explanations for such an incompatibility. First, more circulating HDL-cholesterol might reflect a more effective transfer of deposited vascular cholesterols to the liver for excretion by HDLs [11]. The resultant decrease in retained vascular cholesterols would stimulate less macrophage colonystimulating factor and, thus, lessen the transformation of circulating monocytes into macrophages [12 14]. The second possible explanation is that circulating HDLs may have a prominent inhibitory effect on the oxidation of circulating LDLs, as has been recently reported [15 17], and, thus, the circulating HDL concentration and not the circulating LDL concentration might determine the amount of circulating LDLs to be oxidized and deposited in the vascular wall. The finding of a positive link between the CD16+ subset ratio and male gender was unexpected (Tables 1 and 4). This link was stronger than that between the CD16+ subset ratio and HDL-cholesterol (Table 4). Several studies have noted that men are at increased risk for atherosclerosis, especially before the age of 50 years [18, 19], and there is an obvious sex difference in the mechanism and pathophysiology of atherosclerosis [18 21]. A higher CD16+ subset ratio in men, as found in the study, may in part contribute to the reported male predominance in atherosclerosis. Further in-depth studies are needed to test this hypothesis. An interesting finding in this study was a persistent significant inverse link between HDL-cholesterol and the monocyte count by various statistical methods (Tables 2 4). This inverse link was not affected by gender (Table 4) and was probably related to the antiinflammatory effect of HDLs. Several other studies have shown the abilities of HDLs to bind inflammatory mediators, neutralize their leukocyte chemotactic activities, and, thus, attenuate the inflammation- or Table 3. Linear correlation analyses between monocyte measurements and age, serum total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), and triglycerides (TGs) (n = 100) Monocyte count (x 10 9 /L) CD16+ subset ratio (%) CD36+ subset ratio (%) r value p * r value p * r value p * Age (yr) TC LDL-C < HDL-C 0.48 < < < 0.05 TGs < 0.05 < * Correlation coefficients (r values) were estimated by Pearson s method, and probability values (p) were then calculated by t-test. A female subject with CD36 deficiency was not included in the analyses concerning CD36+ subset ratio. Table 4. Multivariate regression analyses of monocyte measurements on age, gender, serum total cholesterol (TC), lowdensity lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), and triglycerides (TGs) (n = 100) Monocyte count (x 10 9 /L) CD16+ subset ratio (%) CD36+ subset ratio (%) Regression coefficient p * Regression coefficient p * Regression coefficient p * (SE) (SE) (SE) Age (yr) (0.0016) (0.1401) (0.0491) 0.54 Gender (0.0255) (2.1917) < (0.7652) < 0.01 TC (0.0011) (0.0959) (0.0335) 0.92 LDL-C (0.0010) (0.0871) (0.0303) 0.73 HDL-C (0.0017) < (0.1414) (0.0493) 0.72 TGs (0.0003) (0.0256) (0.0089) 0.64 Excluding TC Age (yr) (0.0016) (0.1355) (0.0471) 0.55 Gender (0.0251) (2.1668) < (0.7517) < 0.01 LDL-C (0.0004) (0.0340) (0.0118) 0.26 HDL-C (0.0012) < (0.1027) (0.0357) 0.56 TGs (0.0002) (0.0152) (0.0053) 0.36 * Regression coefficients and P values were estimated by ordinary least squares method and t-test. A female subject with CD36 deficiency was not included in the analyses concerning CD36 subset ratio. Because there was a highly significant linear correlation between serum TC and serum LDL-C (r = , p < by Pearson s method), the analysis was performed a second time after excluding serum TC. SE = standard error of mean. J Formos Med Assoc 2002 Vol 101 No 1 15

6 Z.S. Huang and B.L. Chiang Table 5. Multivariate regression analysis of peripheral CD16+ and CD36+ monocyte counts on age, gender, serum total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), and triglycerides (TGs) (n = 100) CD16+ monocyte count (x 10 9 /L) CD36+ monocyte count (x 10 9 /L) Regression coefficient (SE) p * Regression coefficient (SE) p * Age (yr) (0.0007) (0.1574) 0.13 Gender (0.0108) (2.4556) 0.28 TC (0.0005) (0.1073) 0.86 LDL-C (0.0004) (0.0974) 0.71 HDL-C (0.0007) (0.1581) < 0.01 TGs (0.0001) (0.0286) 0.63 Excluding TC Age (yr) (0.0666) (0.1512) 0.12 Gender (1.0648) (2.4124) 0.26 LDL-C (0.0167) (0.0379) 0.58 HDL-C (0.0505) < (0.1144) < 0.01 TGs (0.0075) (0.0169) 0.56 CD16+ monocyte count = (peripheral monocyte count, x 10 9 /L) x (CD16+ subset ratio, %); CD36+ monocyte count = (peripheral monocyte count, x 10 9 /L) x (CD36+ subset ratio, %). * Regression coefficients and P values were estimated by ordinary least squares method and t-test. A female subject with CD36 deficiency was not included in the analyses concerning CD36 subset ratio. Because there was a highly significant linear correlation between serum TC and serum LDL-C (r = , p < by Pearson s method), the analysis was performed a second time after excluding serum TC. SE = standard error of mean. infection-induced increase in the count or recruitment of leukocytes, especially monocytes and neutrophils [22, 23]. In this way, HDLs may ultimately suppress the peripheral monocyte count. This may be one reason why more circulating HDL-cholesterol is antiatherogenic [24, 25]. The coexistence of inverse links between serum HDL-cholesterol and both the peripheral monocyte count and its CD16+ subset ratio (Tables 2 4) led us to perform another multivariate regression analysis to evaluate the relationship between HDL-cholesterol and the CD16+ monocyte count, calculated as the product of the monocyte count and the CD16+ subset ratio, as shown in Table 5. This age- and genderadjusted analysis revealed that HDL-cholesterol, but not LDL-cholesterol, had an independent inverse link with the CD16+ monocyte count (p < 0.05). These results further support the hypothesis that the circulating HDL concentration might have an important influence on the transformation of circulating monocytes into macrophages. Since monocytes are closely related to lymphocytes in immunoinflammatory function, and hypercholesterolemia has been associated with elevated CD4 and CD8 lymphocyte subset counts [26, 27], significant interrelationships among different lymphocyte subset counts, different monocyte subset counts, and serum HDL-cholesterol may exist. Some of the numerous cytokines produced by monocytes and lymphocytes are also likely to have significant relationships with serum HDL-cholesterol [28, 29]. 16 The generally high CD36+ monocyte subset ratio in peripheral blood (85.94% 100.0% except for one subject with CD36 deficiency), together with the findings of other studies that CD36 is an important receptor in the regulation of HDL, LDL, and very low-density lipoprotein (VLDL) [9, 10, 30], suggests that circulating monocytes may be active in lipid metabolism. The significant inverse links between HDL-cholesterol and both the CD36+ subset ratio (Tables 2 4) and CD36+ monocyte count (Table 5) suggest that the participation of monocytes in lipid metabolism may be downregulated in the presence of more circulating HDLs. Although the CD36+ subset ratio was significantly higher in men (Tables 1 and 4), the difference in the absolute values between men and women was actually very small (98.13 ± 0.31% and ± 0.51%, respectively; Table 1). The clinical and biological significance of such a small difference may be limited. In summary, our study found that serum HDLcholesterol had an independent inverse link with peripheral monocyte count (Table 4), probably related to the antiinflammatory property of HDLs that might suppress production and mobilization of monocytes [22, 23]. Serum HDL-cholesterol, but not LDLcholesterol, had independent inverse links with both the ratio and count of the CD16+ monocyte subset in peripheral blood (Tables 4 and 5). These findings suggest that the circulating HDL concentration may be much more important than the circulating LDL concentration in affecting the transformation of J Formos Med Assoc 2002 Vol 101 No 1

7 circulating monocytes into macrophages, probably either because HDLs can effectively remove vascular and tissue cholesterols to the liver for excretion [11, 24, 25], or because HDLs can effectively inhibit oxidation of circulating LDLs [14 17]. The resultant lower deposition of vascular cholesterols or oxidized LDLs would subsequently lessen the need for macrophages. In this study, we also found that men had a significantly higher ratio and count of CD16+ monocyte subset in peripheral blood (Tables 1, 4, and 5). This situation may contribute partly to the reported male predominance of atherosclerosis [18, 19] and may be valuable in studying the mechanism of sex difference in atherosclerosis [20, 21]. References 1. Gerrity RG: The role of the monocyte in atherosclerosis. I. Transition of blood-borne monocytes into foam cells in fatty lesions. Am J Pathol 1981;103: Joris I, Zand T, Nunnari JJ, et al: Studies on the pathogenesis of atherosclerosis. I. Adhesion and emigration of mononuclear cells in the aorta of hypercholesterolemic rats. Am J Pathol 1983;113: Faggiotto A, Ross R, Harker L: Studies of hypercholesterolemia in the nonhuman primate. I. Changes that lead to fatty streak formation. Arteriosclerosis 1984;4: Aviram M: Interaction of oxidized low density lipoprotein with macrophages in atherosclerosis, and the antiatherogenicity of antioxidants. Eur J Clin Chem Clin Biochem 1996;34: Westhuyzen J: The oxidation hypothesis of atherosclerosis: an update. Ann Clin Lab Sci 1997;27: Kruth HS: The fate of lipoprotein cholesterol entering the arterial wall. Curr Opin Lipidol 1997;8: Ziegler-Heitbrock HWL, Fingerle G, Ströbel M, et al: The novel subset of CD14+/CD16+ blood monocytes exhibits features of tissue macrophages. Eur J Immunol 1993;23: Frankenberger M, Sternsdorf T, Pechumer H, et al: Differential cytokine expression in human blood monocyte subpopulation: a polymerase chain reaction analysis. Blood 1996;87: Huh HY, Pearce SF, Yesner LM, et al: Regulated expression of CD36 during monocyte-to-macrophage differentiation: potential role of CD36 in foam cell formation. Blood 1996;87: Pietsch A, Erl W, Lorenz RL: Lovastatin reduces expression of the combined adhesion and scavenger receptor CD36 in human monocytic cells. Biochem Pharmacol 1996;52: Oram JF, Yokoyama S: Apolipoprotein-mediated removal of cellular cholesterol and phospholipids. J Lipid Res 1996;37: Gerrity RG: The role of the monocyte in atherosclerosis. II. Migration of foam cells from atherosclerotic lesions. Am J Pathol 1981;103: Koren E, Koscec M, McConathy WJ, et al: Possible role of macrophages in regression of atherosclerosis. Prog Lipid Res 1991;30: Watanabe Y, Inaba T, Gotoda T, et al: Role of macrophage colony-stimulating factor in the initial process of atherosclerosis. Ann NY Acad Sci 1995;748: Hasselwander O, McEneny J, McMaster D, et al: HDL composition and HDL antioxidant capacity in patients on regular haemodialysis. Atherosclerosis 1999;143: Holvoet P, Collen D: Oxidation of low density lipoproteins in the pathogenesis of atherosclerosis. Atherosclerosis 1998;137(Suppl):S Yoshikawa M, Sakuma N, Hibino T, et al: HDL3 exerts more powerful anti-oxidative, protective effects against copper-catalyzed LDL oxidation than HDL2. Clin Biochem 1997;30: McCrohon JA, Death AK, Nakhla S, et al: Androgen receptor expression is greater in macrophages from male than from female donors. A sex difference with implications for atherogenesis. Circulation 2000;101: Joakimsen O, Bonaa KH, Stensland-Bugge E, et al: Age and sex differences in the distribution and ultrasound morphology of carotid atherosclerosis: the Tromso Study. Arterioscler Thromb Vasc Biol 1999;19: Reilly SL, Ferrell RE, Kottke BA, et al: The gender-specific apolipoprotein E genotype influence on the distribution of plasma lipids and apolipoproteins in the population of Rochester, Minnesota. II. Regression relationships with concomitants. Am J Hum Genet 1992;51: Barrett-Connor E: Sex differences in coronary heart disease. Why are women so superior? The 1995 Ancel Keys Lecture. Circulation 1997;95: Pajkrt D, Doran JE, Koster F, et al: Antiinflammatory effects of reconstituted high-density lipoprotein during human endotoxemia. J Exp Med 1996;184: Badolato R, Wang JM, Murphy WJ, et al: Serum amyloid A is a chemoattractant: induction of migration, adhesion, and tissue infiltration of monocytes and polymorphonuclear leukocytes. J Exp Med 1994;180: Eisenberg S: High density lipoprotein metabolism. J Lipid Res 1984;25: Stein O, Stein Y: Atheroprotective mechanisms of HDL. Atherosclerosis 1999;144: Muldoon MF, Marsland A, Flory JD, et al: Immune system differences in men with hypo- or hypercholesterolemia. Clin Immunol Immunopathol 1997;84: Moreno LA, Sarria A, Lazaro A, et al: Lymphocyte T subset counts in children with hypercholesterolemia receiving dietary therapy. Ann Nutr Metab 1998;42: Eggesbo JB, Hjermann I, Joo GB, et al: LPS-induced release of EGF, GM-CSF, GRO alpha, LIF, MIP-1 alpha and PDGF-AB in PBMC from persons with high or low levels of HDL lipoprotein. Cytokine 1995;7: Cockerill GW, Saklatvala J, Ridley SH, et al: High-density lipoproteins differentially modulate cytokine-induced expression of E-selectin and cyclooxygenase-2. Arterioscler Thromb Vasc Biol 1999;19: Calvo D, Gomez-Coronado D, Suarez Y, et al: Human CD36 is a high affinity receptor for the native lipoproteins HDL, LDL, and VLDL. J Lipid Res 1998;39: J Formos Med Assoc 2002 Vol 101 No 1 17